The State and Life Chances in Urban China: Redistribution and Stratification, 1949-1994

This page intentionally left blank The State and Life Chances in Urban China The State and Life Chances in Urban Ch...

15 downloads 406 Views 2MB Size Report

This content was uploaded by our users and we assume good faith they have the permission to share this book. If you own the copyright to this book and it is wrongfully on our website, we offer a simple DMCA procedure to remove your content from our site. Start by pressing the button below!

Report copyright / DMCA form

DOWNLOAD PDF

This page intentionally left blank

The State and Life Chances in Urban China

The State and Life Chances in Urban China Redistribution and Stratiﬁcation, 1949–1994

XUEGUANG ZHOU Duke University

   Cambridge, New York, Melbourne, Madrid, Cape Town, Singapore, São Paulo Cambridge University Press The Edinburgh Building, Cambridge  , UK Published in the United States of America by Cambridge University Press, New York www.cambridge.org Information on this title: www.cambridge.org/9780521835077 © Xueguang Zhou 2004 This book is in copyright. Subject to statutory exception and to the provision of relevant collective licensing agreements, no reproduction of any part may take place without the written permission of Cambridge University Press. First published in print format 2005 - -

---- eBook (EBL) --- eBook (EBL)

- -

---- hardback --- hardback

Cambridge University Press has no responsibility for the persistence or accuracy of s for external or third-party internet websites referred to in this book, and does not guarantee that any content on such websites is, or will remain, accurate or appropriate.

To Zhaohui and Katie

Contents

List of Figures List of Tables

page xiii xv

Preface Credit Line 1

2

xix xxiii

Redistribution and Stratiﬁcation Dynamics Under State Socialism Introduction Social Stratiﬁcation in a Comparative Perspective Redistribution and Stratiﬁcation Dynamics Under State Socialism The Political Logic of Redistribution The Bureaucratic Class Thesis Redistributive Institutions Thesis Stratiﬁcation Dynamics Under State Socialism Institutional Transformation and Stratiﬁcation Processes in the Reform Era The Theoretical Debate Coevolution Between Politics and Markets Looking Ahead Overview: Historical Context and Research Design Changing Historical Contexts in Urban China The 1949–1965 Period The 1966–1979 Period The 1980–1994 Period

vii

1 1 5 7 7 9 13 16 26 27 30 33 35 35 36 38 39

viii

Contents The Evolution of Redistributive Institutions and Life Chances Labor Force Composition Composition of Industrial Output Across Sectors Educational Opportunities The Life History Data From “The State and Life Chances” Project Sampling Research Design Considerations Focus of Analysis: Different Aspects of the Stratiﬁcation System Main Covariates of Theoretical Interest Statistical Analysis: Models and Methods The Concept of the “Risk Set” Periodization Models Missing Values Chapter Summary

3

4

part one. redistribution and stratiﬁcation dynamics: empirical evidence Educational Stratiﬁcation Education and Social Stratiﬁcation: The Issues Historical Evolution of Educational Opportunities State Policies, Social Origins, and Educational Attainment: Further Analyses Overview: Do Social Origins Matter? Effects of Gender and Social Origins on Transition Rates Across Historical Periods Effects of Family Class Background Chapter Summary Entry Into the Labor Force: Patterns of First-Job Attainment The Concept of Job in a Comparative Perspective Occupational Status The Hierarchy of Work Organizations Descriptive Patterns of Entry Into the Labor Force Entry Into First Occupation Entry Into First Organization Explaining Entry Into First Jobs: The Overall Pattern Overall Patterns of First-Job Attainment Period-Speciﬁc Job-Attainment Patterns Entry Into First Occupation

40 40 46 47 50 51 52 53 54 60 60 61 61 64 64

69 71 75 80 83 85 93 95 98 99 100 102 103 104 105 107 108 111 112

Contents

5

6

7

Entry Into First Organization An Alternative Speciﬁcation of Social Origins Chapter Summary Occupation Versus Work Organization Hierarchies Stratiﬁcation Dynamics Children of the Cultural Revolution: The Send-Down Episode The Issues The Send-Down Experience in Historical Context Historical Patterns of the Send-Down Episode Different Impacts of State Policies on Social Groups Who Was More Likely to Be Sent Down? Return to Urban Areas Consequences of the Sent-Down Experience Experience of Subsequent Life-Course Events Income Determinants Chapter Summary Climbing the Political Ladder: Bureaucratic Career Patterns Introduction: The Personnel System in the Chinese Bureaucracy The Bureaucratic System Policies and Practice of Recruitment and Promotion Explaining Bureaucratic Career Patterns in Urban China: Two Models Walder’s Dual-Path Model A Model of Stratiﬁcation Dynamics Implications for Empirical Examination Descriptive Patterns Across Historical Periods Entry Into Cadre and Professional Careers Recruitment Into the Communist Party Promotion Patterns in the Chinese Bureaucracy Promotion Patterns in Formal Bureaucratic Ranks Promotion Patterns in Work Organizations Chapter Summary part two. assessing institutional changes in the post-mao era Economic Transformation and Changes in Income Inequality Redistribution and Marketization: Implications for Income Inequality

ix

116 119 122 122 122 124 126 128 131 133 133 138 142 142 146 151 154 156 157 159 162 162 164 167 171 173 179 182 183 188 192

197 199

x

Contents Cross-Sectional Analyses of Income Determinants Comparing Changes Across Two Eras: A Mixed Model for Panel Data Exploration of Regional Variations Comparison With the OLS Results and Other Studies Chapter Summary

8

9

Redistribution and Latent Economic Beneﬁts Redistribution and Latent Economic Beneﬁts Policy on Maternity Leave: An Illustration The Issues Distribution of Latent Economic Beneﬁts: Historical Overview A Case Study: Determinants of Housing Distribution Determinants of Housing Space Determinants of Housing Quality Chapter Summary Implications for Understanding Redistribution and Stratiﬁcation Dynamics Implications for Understanding Recent Economic Transformation Institutional Changes and Patterns of Job Shifts Job-Shift Patterns as a Focus of Study Why Study Job Shifts? Explaining Job Shifts: Some Considerations Analytical Strategies in Modeling Job Shifts Historical Trends of Job Shift Patterns Explaining Job Shift Patterns Across Historical Periods External Job Shifts: Moves Across Types of Organizations Internal Job Shifts: Moves Within Types of Organizations External Job Shifts: Moves Across Economic Sectors Chapter Summary

10 Economic Transformation and Life Chances: A Life-Course Perspective Institutional Changes and Life-Course Implications Empirical Implications Job-Shift Patterns: Cohort Difference in Response to Opportunities Determinants of Personal Income: Cohort Variations in Economic Resources Historical Context, Cohort Attributes, and Evolving Opportunity Structures

202 209 216 221 221 225 226 226 227 229 233 234 239 240 240 243 246 247 247 249 251 253 255 255 261 265 269 272 274 277 277 279 280

Contents Response to Opportunities and Risks: Cohort Variations in Job-Shift Patterns Cohort Variations in Economic Rewards Chapter Summary part three. summary and conclusion 11 Social Stratiﬁcation Under State Socialism: Summary and Assessment The Bureaucratic Class Thesis Revisited Socioeconomic Status and Bureaucratic Privileges Authority Relationships in Work Organizations Intergenerational Status Inheritance A Summing Up Human Capital Versus Political Capital Redistributive Institutions in Social Stratiﬁcation Stratiﬁcation Dynamics Under State Socialism: Theoretical Implications Further Issues

xi

284 290 293

299 300 300 303 306 308 309 311 312 315

12 Stratiﬁcation Dynamics and Institutional Transformation Rethinking State–Society Relationship Under State Socialism Imageries of Society under State Socialism An Institutional Model of Unorganized Interests and Collective Action Stratiﬁcation Dynamics and State–Society Relationship: Further Considerations The Social Basis of Institutional Transformation in China Erosion of the Organizational Basis of Redistribution Life Experiences and Ideological Crises From Social Stratiﬁcation to Institutional Transformation Beyond China and Beyond State Socialism

317

References Index

337 353

317 318 320 322 324 325 327 330 333

Figures

2.1a. Composition of urban labor force across sectors in ofﬁcial statistics, 1952–94 page 41 2.1b. Composition of urban labor force across sectors in the sample, 1951–91 (5-year average) 42 2.2a. Composition of type of organizations in the state sector (5-year average) 43 2.2b. Composition of type of organizations in the nonstate sector (5-year average) 44 2.3a. Composition of cadres, professionals and ofﬁce workers, 1951–91 (5-year average) 45 2.3b. Composition of workers and self-employed, 1951–91 (5-year average) 45 2.4. Composition of industrial output across sectors, 1952–94 47 2.5. Enrollment rates at three educational levels, 1949–94 (ofﬁcial statistics) 48 2.6. Distribution of educational levels in the sample, 1951–91 (5-year average) 49 3.1. Rates of entry into four educational levels in the sample, 1953–88 (5-year average) 76 3.2. Number of higher education institutions, 1949–94 78 3.3a. Rates of entry into high school, by the father’s occupation (5-year average) 79 3.3b. Rates of entry into college, by the father’s occupation (5-year average) 80 3.4. Transition rates by period and gender 89 xiii

xiv

Figures

3.5a. Transition rates to high school by period and father education 3.5b. Transition rates to college by period and father education 3.6a. Transition rates to high school by period and class label 3.6b. Transition rates to college by period and class label 4.1. Hazard rates of entry into ﬁrst occupations, 1949–92 4.2a. Hazard rates of entry into ﬁrst organizations, the state and collective sectors, 1949–92 4.2b. Hazard rates of entry into ﬁrst organizations, hybrid ﬁrm and farm, 1949–92 4.3. Rates of entering types of occupations for college graduates, by period 4.4. Rates of entering types of organizations for college graduates, by period 5.1. Hazard rates of entry into ﬁrst job, 1967–78 5.2. Hazard rates of return to urban areas, 1967–87 5.3. Distribution of age at marriage among three groups 5.4. Distribution of age at birth of ﬁrst child among three groups 6.1. The administrative rank system 6.2. The professional rank system 6.3. Hazard rates of entry into CCP, administrative, and professional occupations (1950–93) 6.4. Hazard rates of promotion in national bureaucratic system (1950–93) 6.5. Age effects on entry into CCP, administrative, and professional occupations (1980–94) 6.6. Age effects on promotion rates, 1980–94 7.1. Comparison of income in national and sample statistics 7.2. Log-rates of income for female and college education (selected years) 7.3. Log-rates of income for cadres and professionals (selected years) 7.4. Log-rates of income for type of organizations (selected years) 9.1 Hazard rates of job shifts to the state sector 9.2 Hazard rates of job shifts to collective and nonstate ﬁrms 10.1 Distribution of ﬁrst-job destinations, by cohort 10.2 Distribution of job destinations in 1993, by cohort

91 92 94 95 104 106 106 116 119 132 132 144 144 158 160 171 173 177 186 203 206 207 208 254 254 281 282

Tables

1.1. A partial list of political campaigns in China, 1949–1994 2.1. Distribution of the sample size across provinces and cities 3.1. Likelihood ratio tests of improvement in model ﬁt 3.2. Parameter estimates of logistic regression models of probability of entering three educational levels, by historical period 3.3. Parameter estimates of class background effects on probability of entering three educational levels, by historical period 4.1. Parameter estimates of logistic model for entry into ﬁrst occupation (reference category: worker) 4.2. Parameter estimates of logistic model for entry into ﬁrst organization (reference category: collective ﬁrm) 4.3. Parameter estimates of logistic model for entry into ﬁrst occupation, by period (reference category: worker) 4.4. Parameter estimates of logistic model for entry into ﬁrst organization, by period (reference category: collective ﬁrm) 4.5. Parameter estimates of logistic model for entry into ﬁrst work organization, using “family class background” by period 5.1. Parameter estimates of logistic model for Probability of Being Sent Down: Sent-down = 1, 1967–1978 xv

page 17 53 84

86

94 109 111 113

117

121 135

xvi

Tables

5.2. Parameter estimates of logistic model for Probability of Returning to Urban Areas, return to urban areas = 1, 1967–1977 5.3. Descriptive and test statistics of group differences in experiencing life-course events 5.4. OLS estimates of the determinants of income, selected years 6.1. Parameter estimates of logistic model for entry into the cadre occupation, by period 6.2. Parameter estimates of logistic model for entry into the Communist Party, by period 6.3. GEE estimates of the determinants of promotion patterns in the Chinese bureaucracy 6.4. GEE estimates of the determinants of promotion patterns in work organizations 7.1. OLS estimates of the determinants of log(income), selected years 7.2. Parameter estimates of the determinants of log(income) in the full mixed model 7.3. Parameter estimates of the determinants of log(income) in the mixed model, by region 8.1. Distribution of welfare expenditure, by type of organization 8.2. Distribution of expenditure on housing construction, by type of organization 8.3. OLS estimates of the determinants of living space (log[square meter]), selected years 8.4. Parameter estimates of the Poisson model for number of facilities, selected years 9.1. Parameter estimates of logistic model for job shifts outside type of work organization, by period 9.2. Parameter estimates of logistic model for job shifts within type of organization, by period 9.3. Parameter estimates of logistic model for job shifts outside economic sectors, by period 10.1. Descriptive statistics of three cohorts of workers in urban China, 1993 10.2. Parameter estimates of logistic model for job shifts across type of organizations in the reform era (1980–94), by cohort

139 143 147 175 180 184 190 205 211 217 230 231 236 241 256 263 266 284

286

Tables 10.3. OLS Parameter estimates of determinants of 1993 income for three cohorts of workers in urban China 11.1. Summary of returns to bureaucratic positions 11.2. Gini Coefﬁcients of income inequality across countries 11.3. Summary of ﬁndings on intergenerational status inheritance 11.4. Summary of ﬁndings on returns to education 11.5. Summary of ﬁndings on returns to organizations (the state sector)

xvii

291 301 304 307 310 312

Preface

This book is about the impact of the state and state policies on ordinary citizens, and how individual lives have been shaped and reshaped by the political twists and turns in the People’s Republic of China, from 1949 to 1994. Behind the lifeless numbers, statistics, and academic jargon in this book, I want to tell stories of real people – there were millions of them – who lived through a political environment that witnessed major and often volatile changes in the forty-ﬁve-year history of the People’s Republic of China. My academic training has taught me to be abstract in theoretical arguments, rigorous in statistical analysis, and systematic in presenting empirical evidence, often, unfortunately, at the expense of real people and real lives. While I was working on research for this book, different images often came to my mind from my personal experience when I grew up in China and from the recollections of many others with whom I am acquainted. It is these real life stories that made up the statistical patterns reported in this book. I want to share some of these stories with the reader as an introduction to the context of this study. Let me begin with a Chinese scholar I met in the 1980s. He was a talented student in a university in Beijing in the 1950s. During the Hundred-Flower period, he responded to the government’s advocacy to criticize the misconducts of the ofﬁcials in his university. One thing led to another, and for reasons no one could comprehend, he was put in prison – for more than twenty years. Even today, his wife told me, because of the imprinting from his long prison experience, he only sleeps on one side of the bed and turns over in bed on the same spot, no matter how big the bed is. xix

xx

Preface

A friend of mine went to a labor farm in a remote region with his family when he was only six, because his father was purged in the political campaign and sent there. The economic disaster in the early 1960s hit that farm especially hard; many died of starvation. He vividly recalled, witnessed from a child’s eyes, that even those who carried the dead bodies for burial were trembling, because they were so hungry that they could barely support their legs. When he left the farm a few years later, of several hundred laborers working there, only a few dozen remained. Zhaohui, my wife, told me her own experience. When she was a little girl during the Cultural Revolution, her mother was routinely called to the meetings by the “red guards” to be “criticized” simply because of her “exploiting class” family background. Often Zhaohui would walk alone to the corner of the street late at night, waiting for her mother to return. “Many times,” she said in a low voice, “I had a feeling that my mother would never come back.” I still have a vivid memory of the humiliation and despair I felt when I saw my father’s name on a “Big Character Poster” in a public place in the Cultural Revolution, revealing his “exploiting class” background. As I walked home that day, I felt as if the sky were falling on me, and I tried to avoid anyone on the street. I was only seven or eight years old at the time. When I ﬁrst began working on this book, my daughter was about that age. Often times when I saw her innocent eyes, I could not help but realize how fast the political turbulence has pushed a young generation to mature! These stories are by no means extraordinary ones that I deliberately sought for sensational reasons. Any individual who lived through the political turbulence in China can narrate many similar and even more dramatic experiences. They were part of everyday life experience in those years! Indeed, through the lens of social stratiﬁcation patterns in this book, we can observe many different, often contradicting, images of state socialist China in the second half of the twentieth century. On the one hand, it is undeniable that, in the forty-ﬁve-year history of the People’s Republic of China, there were impressive improvements in different aspects of living standards for the Chinese people and in equalizing opportunities among social groups. At the same time, we also witnessed the ups and downs of individual career trajectories, the losses and regains of opportunities, the break-ups and reunions of families, and the political purges and rehabilitations of cadres and intellectuals. This book reports my efforts to describe, analyze, and explain these ironies and multifaceted patterns of political and social changes and their impacts on individuals’ lives.

Preface

xxi

In the course of working on the research reported in this book, I have accumulated much debt to colleagues and friends. I want to especially thank Phyllis Moen and Nancy Brandon Tuma, who collaborated with me on this project from the very beginning and who co-authored several publications that are adapted into this book. I learned statistical modeling methods from Nancy and life-course research from Phyllis, and they left their marks on many aspects of this book. But their inﬂuence on me far exceeds this research project. Through our collaborations over the years, they continued to educate me, nurture my professional development, and push me to think clearly about the theoretical and analytical issues pursued in this book. To them I am forever in debt. I thank Liren Hou, Lin Lu, Olga Suhomlinova, and Wei Zhao – my research assistants, friends, collaborators, and colleagues – who have provided not only valuable research assistance but also invaluable friendship throughout this process. I thank Yanjie Bian, Ronald Breiger, Claudia Buchmann, Debbie Davis, Tom DiPrete, Ken Land, Tom Gold, Nan Lin, Victor Nee, Angie O’Rand, Bill Parish, Ken Spenner, David Stark, Andy Walder, Feng Wang, Marty Whyte, John Wilson, and Yu Xie who have commented and helped me improve on my work over the years. A special thanks goes to David Grusky, whose initial encouragement jump started this book project. Although this book is on China’ my scholarship has been built on many years of learning from my teachers Jim Macch, John Meyer, Dick Scott, and Buzz Zelditch. I am grateful to Ms. Alia Winters, my editor at Cambridge University Press, whose guidance and help has made this publishing process an enjoyable experience! This project ﬁrst took shape when I taught in the Department of Sociology, Cornell University. My current institution, Duke University and the Department of Sociology there, has provided generous support for my research over the years. This project began with a pilot study in which we collected data in Beijing and Shanghai in the summer of 1993. The pilot study was funded by small grants from ASA/NSF Fund for the Advancement of the Discipline, the Lam Fund, and Halpern Fund from the East Asian Program at Cornell University. The main data collection for this project was carried out from June to December 1994. It was funded by a grant from NSF (SBR-9413540) and carried out with the collaboration of the Institute of Sociology at Tianjin Academy of Social Sciences, and the Departments of Sociology at Fudan University and at the People’s University. I am grateful to Fan Weida, Li Qiang, Pan Yunkang, Peng Xizhe, and Wang Hui, as well as to many students and researchers in these institutions, for their help in the data collection. A Spencer Fellowship

xxii

Preface

from the National Academy of Education allowed me to take a leave from Duke University to organize the survey in the ﬁeld. I thank the funding institutions for their support of this project. As usual, they are not responsible for the analyses and opinions expressed in this book. My greatest gratitude goes to my parents. In those difﬁcult days when most books were forbidden in my native country, they carefully protected and nurtured a young, curious mind. Without their inspiration, my intellectual career would have ended after high school. Finally, I want to thank Zhaohui Xue, my wife, and Katie, my daughter. They have quietly but unfailingly supported me and my work, and tolerated my late hours, disrupted schedules, and absences from birthdays and holidays when I was on research trips. To them I dedicate this book.

Credit Line

I thank several publishers for their copyright permission to use the following materials from their publications: Part of Table 11.2 in Chapter 11 is reproduced from Economic Transformation in Eastern Europe and the Distribution of Income, edited by C 1992, with permisAnthony Barnes Atkinson and John Micklewright. sion from Cambridge University Press. Table 4.1 on page 81. Part of Table 11.2 in Chapter 11 is reproduced from Changing Patterns in the Distribution of Economic Welfare, edited by Peter Gottschalk, C 1997, with permission Bjorn A. Gustafsson, and Edward E. Palmer. from Cambridge University Press. Table 3.1 on page 39. Part of Table 11.2 in Chapter 11 is reproduced from China’s Retreat C 2001, from Equality, edited by Carl Riskin, Zhao Renwei, and Li Shi. with permission from M. E. Sharpe, Inc. Tables 2.1 and 2.2 on p. 28. Quotation from Franz Schurmann at the beginning of Chapter 1 is reproduced from Ideology and Organization in Communist China, by C 1968, with permission from University of California Franz Schurmann. Press. P. xxxvi. Quotation from Ivan Szel´enyi at the beginning of Chapter 1 is reproduced from International Journal of Comparative Sociology, 16, Ivan C 1978, with perSzel´enyi, “Social Inequalities under State Socialism” mission from Brill Academic Publishers. P. 1. Quotation from Max Weber at the beginning of Chapter 5 is reproduced from Max Weber, Economy and Society. 2 vols. Edited by Guenther C 1968, with permission from University of CalRoth and Claus Wittich. ifornia Press. P. 975. xxiii

xxiv

Credit Line

I also thank the following publishers and academic associations for their copyright permissions to reprint or adapt portions of the following pieces in this book: American Sociological Association, 1999. Xueguang Zhou and Liren Hou. “Children of the Cultural Revolution: The State and the Life Course in the People’s Republic of China.” American Sociological Review. 64: 12–36. University of North Carolina Press, 1996. Xueguang Zhou, Nancy B. Tuma, and Phyllis Moen. “Stratiﬁcation Dynamics Under State Socialism: The Case of Urban China, 1949–1993.” Social Forces 74: 759–96. Sage Publications, Inc., 2001. Xueguang Zhou. “Political Dynamics and Bureaucratic Career Patterns in the People’s Republic of China.” Comparative Political Studies 34 (9): 1036–62. Academic Press, 2001. Xueguang Zhou and Phyllis Moen. “Explaining Life Chances in China’s Economic Transformation: A Life Course Approach.” Social Science Research 30 (4): 552–77. University of Chicago Press, 2000. Xueguang Zhou. “Economic Transformation and Income Inequality in Urban China: Evidence from a Panel Data.” American Journal of Sociology 105: 1135–74. Finally, I note that almost all statistical information reported was reanalyzed using updated data and alternative statistical models when I adapted these previous publications into this book. In addition Chapters 3 and 9 were also based on previous publications but were thoroughly revised for this book.

1 Redistribution and Stratiﬁcation Dynamics Under State Socialism

The Chinese Revolution is for the latter half of the twentieth century what the Russian Revolution was for the ﬁrst half. By transforming Chinese society, it has brought a great power into being which proclaims itself the revolutionary and developmental model for the poor countries of the world. Franz Schurmann (1968, p. xxxvi) In state socialist societies social inequalities are basically created and structured by redistributive mechanisms. Ivan Szel´enyi (1978, p. 1)

introduction One evening in 1985, I found myself at a dinner table in Palo Alto, California, with Professor Arthur Wolf, a distinguished anthropologist of China studies, and several students of his. During that conversation, Professor Wolf asked this question: “How can we explain the phenomenon that, ever since population data have been recorded in China’s history, the Chinese population continued to rise, but there was a sharp drop in the late 1950s and early 1960s?” My heart sank as I followed Professor Wolf’s waving arm and visualized the long and upward trajectory and then a sudden, deep slump. Many images and stories rushed into my mind – the recollections of the so-called “Great-Leap-Forward” episode and the subsequent famine period that I heard about over and over as I grew up, from my parents, grandparents, my friends’ parents, and from the peasants in the village where I once worked. Indeed, over so many dynasties, emperors, wars, famines, and other disasters in China’s long history, why did the 1959–1961 famine strike the 1

2

The State and Life Chances in Urban China

Chinese people so mercilessly, when an estimated thirty million perished in a few years (Banister 1987; Yang 1996)? As we know by now, this was not at all a “natural” disaster, but a “man-made” one, caused by policy mistakes by the top leaders! And this was by no means an accidental or isolated event. In the past ﬁve decades in the People’s Republic of China, this was but one of many episodes in which state policies affected individual life chances in dramatic and violent ways. The larger issue behind Professor Wolf’s question is this: Why did “man-made mistakes” have such far-reaching consequences? To address this and other related issues, we must examine social stratiﬁcation processes under state socialism. Social stratiﬁcation – the structure of opportunities and social positions, and the processes of allocating individuals to these opportunities and positions – provides an important lens through which we understand the link between the state and individual life chances under state socialism. To understand social stratiﬁcation patterns, we are confronted with these basic questions: How are the hierarchies of social positions and, ultimately, the social stratiﬁcation system, constructed and maintained? What are the allocative mechanisms through which individuals are channeled to these positions and resources? Why are there noticeable variations in social stratiﬁcation patterns across societies? By seeking answers to these questions, we come to recognize the importance of institutional arrangements that deﬁne and construct the structures and processes through which resources are allocated and transferred among social groups and across generations. In the twentieth century, especially after World War II, industrialized market societies and state socialist societies have presented two distinct and prominent models of social stratiﬁcation. This book is a study of how structures of opportunities and individual life chances evolved over time, especially the role of the state in this process, in the People’s Republic of China (PRC). We trace the historical evolution of the social stratiﬁcation system in urban China over a fortyﬁve-year span under state socialist governance, from 1949 to 1994. We accomplish this goal by examining patterns of state socialist redistribution and their impacts on individual life chances in such areas as educational attainment, job mobility, bureaucratic promotions, and the distribution of economic beneﬁts, among others. There are two main themes that organize this book. The ﬁrst theme addresses the interplay between redistribution and stratiﬁcation dynamics under state socialism. At the core of social stratiﬁcation

Redistribution and Stratiﬁcation Dynamics Under State Socialism 3 are redistributive institutions that provide stable organizational links between the state and its citizens. On the one hand, we observe stable structures of positions and patterns of resource allocation. On the other hand, as I elaborate in the following, the political logic of redistribution also leads to organizational failures that cause drastic changes in state policies and in macro-political processes, thereby introducing stratiﬁcation dynamics that are often noncumulative, disruptive, and generate twists and turns in individuals’ life chances across historical periods and over their life course. The interplay between the stable organizational basis of redistribution and stratiﬁcation dynamics provides a key to understanding the relationship between the state and life chances. A main contribution of this study is to develop theoretical arguments about, and empirically study, the link between the state and life chances under state socialism. The second theme addresses institutional changes in the transformation of state socialism. Since the 1980s, China and other former state socialist societies have embarked on the great transformation of state socialism. The extent of institutional changes and the mechanisms underlying these changes have generated enormous interest and debates among social scientists. Major theoretical work on the transformation of state socialist economies has been developed in the Chinese context. These theoretical arguments point to different sources and directions of institutional changes. A useful angle from which to address these theoretical issues is to examine how stratiﬁcation patterns have evolved over time. Social stratiﬁcation patterns reﬂect fundamental institutional arrangements of a society. If such arrangements have undergone signiﬁcant changes, they should be ﬁrst and foremost captured in changes in patterns of allocative mechanisms, hence patterns of social stratiﬁcation. The study of social stratiﬁcation patterns reported in this book addresses this set of issues by examining historical patterns of resource allocation over time, especially contrasting those before and after the economic reform, and by assessing changes in the mechanisms of social stratiﬁcation in areas of job shift patterns and the distribution of economic resources. The exploration of these two themes, we hope, can shed light on the evolution and decline of state socialism as a political institution. The rise and decline of state socialism was one of the major political events of the twentieth century. The Bolshevik Revolution of 1917 created the ﬁrst state socialist country in Russia. After World War II, state socialist governments spread across continents, providing a seemingly strong, competitive societal model to challenge capitalist market societies. But by the end of

4

The State and Life Chances in Urban China

the twentieth century, most societies have abandoned the state socialist model; others are in deep crises or undergoing fundamental changes. To explain the evolution and crisis of state socialism, one needs to seek answers not only in the political structure of the authoritarian states or in the economic inefﬁciency of the command economy but also in the ways in which the state and society interact with each other. It is, after all, the popular uprisings of the late 1980s involving millions of citizens and a wide spectrum of social groups that pronounced the demise of state socialism as a worldwide political institution. This recognition calls for a political sociology of state socialism that centers its explanations in the institutional structures of state–society relationships, which are sustained by and reﬂected in social stratiﬁcation processes. Moreover, although state socialism as a worldwide political system is withering away, its legacy has not lost contemporary relevance. Since World War II and with the expansion of the modern states, redistribution through the welfare state has greatly expanded in industrialized nations as well as in newly developed or developing societies in Asia, Africa, and Latin America. The institutional arrangements based on redistribution persist to various degrees in the former state socialist countries as well, shaping their unique paths of social transformation. Therefore, an understanding of the evolution of state socialism, its redistributive institutions, and their impact on social stratiﬁcation is not a mere historical curiosity. This study is in the sociological tradition of comparative social stratiﬁcation. Throughout this book, we make explicit or implicit comparisons between market and redistributive institutions and between China and other state socialist societies in their patterns of resource transfer across generations and among social groups. It is in this comparative perspective, we believe, that the Chinese experience can best contribute to our understanding of social stratiﬁcation processes in contemporary societies. There are two main tasks in the rest of this chapter. First, we compare and contrast the main characteristics of social stratiﬁcation patterns between industrialized market societies and those in state socialist societies. This discussion develops a comparative framework and highlights a set of distinctive issues in understanding social stratiﬁcation in state socialist societies. Second, we present theoretical ideas and research issues on the two themes of this study – redistribution and stratiﬁcation dynamics under state socialism and the institutional transformation in China – that will guide the empirical studies reported in this book. We outline the main structure of this book toward the end of this chapter.

Redistribution and Stratiﬁcation Dynamics Under State Socialism 5

social stratiﬁcation in a comparative perspective Our knowledge of social stratiﬁcation processes is largely informed by studies of market societies. In a market economy, political and economic transactions are based on the principle of exchange. In this institutional structure, initial endowments of resources have a lasting effect on social positions because private property rights stabilize and reinforce the relative opportunities of various groups based on their preexisting social and economic resources. It is in this sense that Weber (1978, p. 928) viewed the stratiﬁcation structure as ultimately resting on market position: “[T]he kind of chance in the market is the decisive movement which presents a common condition for the individual’s fate. Class situation is, in this sense, ultimately market situation.” Not surprisingly, then, social mobility and stratiﬁcation in industrialized market societies have been characterized by persistent advantages of initial resource endowments. Studies invariably ﬁnd that social mobility in these societies mainly operates through individual and family-based social inheritance and achievement. Improved socioeconomic status has resulted largely from the emergence of new opportunities due to industrialization and concomitant structural changes (Featherman, Jones, and Hauser 1975; Treiman 1970). These ﬁndings are broadly consistent with the historical pattern of the gradual extension of citizenship and social rights to lower classes in Western Europe and North America (Bendix 1964; Marshall 1950). The importance of an individual’s market position leads to a theoretical focus on inter- and intragenerational social mobility and status attainment in industrialized market societies (e.g., Blau and Duncan 1967; Goldthorpe 1987; Hauser and Featherman 1977; Sørensen 1977). There have been extensive studies of channels of social mobility – the relative contribution to upward mobility of social status inherited from parents versus acquired human and social capital (e.g., Coleman 1988; Hout 1988; Yamaguchi 1983). The focus on structural conditions such as social origins and occupational position implicitly assumes that the stratiﬁcation mechanisms are relatively stable across generations and over an individual’s life course. Indeed, in industrialized market societies, patterns of social mobility are relatively consistent over time and place (see, e.g., Erikson and Goldthorpe 1992; Grusky and Hauser 1984; Hauser and Grusky 1988; Sørensen 1992), even though some cross-national variations in mobility regimes clearly exist. Father’s social status and son’s education, for instance, have remarkably consistent positive effects on intergenerational mobility over time (see, e.g., DiPrete and Grusky

6

The State and Life Chances in Urban China

1990; Featherman and Hauser 1978; Hout 1988). For a theoretical reassessment, see Ganzeboom, Treiman, and Ultee (1991), and Sørensen (1986). Even when social groups in industrialized market societies advance their interests through organized interests, they often engage in marketlike competition. In the United States, for instance, Larson (1977) showed how powerful professional groups have advanced their common interests through “collective projects” to establish protective boundaries and social closures. These interest organizations play an important role in “carving out a labor-market shelter, a social closure, or a sinecure for its members in the labor market” (Freidson 1986, p. 59) and they provide the basis for collective action in interest articulation (Grusky and Sørensen 1998). They may use their resources to gain competitive advantages and adopt political means to acquire legislation in their favor (Zhou 1993a). But, in contrast to state socialist societies, the relative positions of social groups and classes can be appropriately seen as the outcome of private ordering through market-like lateral competition among groups endowed with varying economic and political resources.1 Tilly’s (1998) theorizing on durable inequality based on social relations and social closures highlights an image of structural stability in social stratiﬁcation systems in market societies. Models of stratiﬁcation in industrialized market societies provide a useful starting point for a comparative framework, but they are inadequate for understanding stratiﬁcation in state socialist societies. In these societies, stratiﬁcation is organized around the state-socialist redistributive economy rather than around the market economy, by political authorities rather than by market mechanisms. As we shift our attention from social stratiﬁcation patterns in industrialized market societies to those in state socialist societies, the issues that demand explanations change accordingly. First, we must shift the analytical focus from individuals’ “market situations” to the broader redistributive institutions. Second, in addition to structural locations, we must pay attention to the stratiﬁcation dynamics in order to explain individual life chances. 1

The image portrayed here is heavily inﬂuenced by social stratiﬁcation studies in North America. In a variety of governance structures in Western Europe, the states play a much more active role in constructing relationships among groups and organizations, thereby limiting market-like mechanisms in social stratiﬁcation processes (see Esping-Andersen 1990). However, even the corporatist governance in European societies differs substantively from the state socialist societies in the role of state dominance and control over other organizations and groups.

Redistribution and Stratiﬁcation Dynamics Under State Socialism 7

redistribution and stratiﬁcation dynamics under state socialism Redistribution, or the allocation of resources through a centralized authority, has been a distinctive mode of economic institutions in history (Polanyi 1965). It evolved into its full-ﬂedged form in the Soviet-type state socialist societies, where almost all resources were subject to allocation by the state. To understand social stratiﬁcation under state socialism, one must place squarely the role of the state and its redistributive institutions at the center of theoretical explanations. We begin by considering the main characteristics of the stratiﬁcation system governed by state socialist redistributive institutions.

The Political Logic of Redistribution A deﬁning characteristic of state socialist political structure is the encompassing role of the state. As Lindblom (1977, pp. 238–39) explained: [T]he scope of government is near all-encompassing – wider than in any other politico-economic system. Government owns most productive assets of the society – private property in the means of production is not the general rule – and government immediately and directly organizes the economy. But it reaches as well into the control of religion, all education, family, labor unions, all organizations, and details of personal behavior usually outside the scope of government in other systems.

If the stratiﬁcation processes in capitalist societies reﬂect a market logic, redistribution under state socialism is governed by a political logic in the sense that “the political rather than ‘economic’ deﬁnition of the surplus means that wage levels and the extent of inequality among different categories of the labour force express ﬁrst of all political considerations” (Szel´enyi 1978, p. 78). In this institutional context, the central authority, the Communist Party in power, plays an ultimate and decisive role in resource allocation. Allocative as well as redistributive priorities are decided through political processes structured by the monopolistic partystate. These observations provided the starting point for most studies of state socialist stratiﬁcation in the literature. The political logic of redistribution sheds light on distinctive institutional arrangements and authority relationships in these societies. In the economic arena, industries are prioritized based on the political goals of the state – heavy industry is favored and industries for consumer goods

8

The State and Life Chances in Urban China

are neglected, leading to unique patterns of resource transfers across economic sectors (Kornai 1959, 1972, 1992; Szel´enyi, Beckett, and King 1994; Walder 1992). Authority relationships also reﬂect such a logic. For instance, at each level of the government, the Communist Party organization (e.g., the party headquarters in a city) exercises authority over administrative ofﬁces (e.g., the municipal government) and other organizations. Similarly, in the workplaces, managerial positions associated with the political career line (e.g., the head of the party branch) are granted higher authority over administrators or technocrats. Not surprisingly, the political logic of redistribution leads to distinctive stratiﬁcation patterns in state socialist societies. Consider the socioeconomic status of occupations in Szel´enyi’s (1978) study of Hungary. As Szel´enyi (1978:75) observed: “Under State Socialism the State redistributes surplus, surplus which was never accumulated in personal income, but was directly centralized in the State budget and reallocated according to centrally deﬁned goals.” Although the occupational categories in Hungary are similar to those in industrialized market societies (i.e., professionals, managers, clerks, and skilled and unskilled workers), the stratiﬁcation patterns differ substantially. The privileges and beneﬁts of various social groups depend systematically on their relationships ´ and Szel´enyi 1979). As a result, to the state (Bauman 1974; Konrad labor market structures in these societies differ markedly from those in market societies (Burawoy and Lukacs 1985; Connor 1979; Stark 1986). The political logic of redistribution points to a distinctive set of mechanisms in resource allocation under state socialism. Consider the role of political capital versus human capital in social stratiﬁcation. In a market economy where factors of production are allocated through competitive market transactions, it is argued that human capital plays an important role in determining one’s economic rewards (Becker 1964; Mincer 1974). In contrast, the political logic of redistribution establishes a qualitatively different reward system. The central authority places a premium on political loyalty and the command economy demands the effective implementation of the administrative directives from the above. Accordingly, the reward system favors those with political status and loyalty, such as Communist Party membership or those who are closer to the redistributive power. In contrast, those with human capital (e.g., educational qualiﬁcations) are subordinate to the political authority. This line of argument has been especially advanced in the Chinese context. Many scholars observed the common practice of rewarding political loyalty rather than

Redistribution and Stratiﬁcation Dynamics Under State Socialism 9 competence in the Chinese industries and bureaucracies (Harding 1981; Lee 1991; Nee 1989; Walder 1986; Zhou 1995). The Bureaucratic Class Thesis In line with the political logic of inequality under socialism, the most commonly noted social boundaries have been those between the bureaucrats (cadres) on the one hand and all other social groups on the other. Following Djilas’s (1957) bureaucratic class thesis, students of socialism have emphasized the fundamental divide between “redistributors” and “immediate producers” (Szel´enyi 1978) and the importance of bureaucratic positions in acquiring economic beneﬁts. The main argument is that, because of their privileged positions in the command economy, bureaucrats possess enormous authority in the redistribution of economic resources and enjoy better access to economic beneﬁts. Similar arguments have been advanced in the Chinese context. Nee (1989, 1991), in particular, emphasized the central role of cadres in the stratiﬁcation order in China. The redistribution of resources through central planning led to the rise of bureaucratic organizations and hierarchical structures from industries and economic sectors down to work organizations. Cadres – those administrators, managers, or personnel belonging to the bureaucratic apparatus – are advantageous in their political positions, promotion ladders, and economic beneﬁts. But who belongs to the bureaucratic class, if such a class indeed exists? In what ways can we identify the bureaucratic class? One insight, often labeled as the Weberian approach, highlights the proximity in socioeconomic situations among members of a social class. The notion of common “market situation” occupies a central place in the contemporary discussions of the Weberian approach to social classes. More generally, these market situations are reﬂected in individual life chances. In this light, we expect to observe signiﬁcant differences between the bureaucratic class on the one hand and other social groups on the other with respect to important aspects of life chances. The Marxian approach to social class provides the second insight: social classes are rooted in the property right relations in the production processes. The ownership of money, property, and other means of production is the major divide between the exploiting class and the exploited class. However, there is a fundamental difﬁculty to apply this approach in state socialist societies, where all means of production belong to the state. This recognition led to major revisions in the neo-Marxist approach: a

10

The State and Life Chances in Urban China

shift of focus from property ownership to “authority relationships” in the work environments. Wright (1989:16–17) argued: In state bureaucratic socialism, organization assets assume a much greater importance. Controlling the technical division of labor – the coordination of productive activities within and across labor processes – becomes a societal task organized at the center. The control over organization assets is no longer simply the task of ﬁrm-level managers but extends into the central organs of planning within the state. Exploitation in such societies is thus based on bureaucratic power: the control over organization assets deﬁnes the material basis for class relations and exploitation.

In this light, authority relationships within work organizations are another key analytical focus for understanding the role of the bureaucratic class. We expect to uncover systematic evidence of the advantages conferred in positional power over workers on the work ﬂoor. Finally, the third aspect is related to “class reproduction” – intergenerational resource transfer, especially in terms of status inheritance – the central issue in class analysis and studies of social stratiﬁcation. Insofar as classes are stable, recurring social groups, the issue of class reproduction is central to any explanation of class formation. In the transition from the aristocratic society to the capitalist society, principles of class reproduction changed from property-based inheritance to credentialist strategies and legal monopoly through state-sponsored occupational licensure (Parkin 1979). As cumulative evidence has shown, as long as opportunities are limited and resources matter in social mobility, intergenerational resource transfer has been and still is the most direct way of generating status inheritance and the reproduction of social classes. Using these criteria, we can develop a working deﬁnition of the bureaucratic class in state socialist societies as a group whose members (1) enjoy signiﬁcantly better socioeconomic beneﬁts relative to other social groups, (2) exercise power and control over organizational assets and in “authority relationships” in the work organizations, and (3) can reproduce their class through status inheritance across generations. These three dimensions provide the basic criteria for us to evaluate the bureaucratic class thesis. As Parkin (1979, p. 53) pointed out succinctly: “The relevant question is not whether surplus extraction occurs, but whether the state confers rights upon a limited circle of eligibles to deny access to the ‘means of life and labour’ to the rest of the community.” At the empirical level, the notion of a bureaucratic class becomes muddy. In the early critique of the bureaucratic class in the Soviet Union and Eastern Europe (e.g., Djilas 1957), the top echelon of party and state leaders were conceptualized as the bureaucratic class proper. Although

Redistribution and Stratiﬁcation Dynamics Under State Socialism 11 we are not aware of systematic, empirical studies in this regard, anecdotal evidence seems to be consistent with the three dimensions discussed previously. That is, the top leaders in these societies had unlimited power in imposing their wills on the society; they enjoyed economic beneﬁts and privileges (privileged access to villas, special goods, and vehicles); and their children advanced through sponsored mobility. But to conﬁne the bureaucratic class only to the top echelon of the state apparatus is problematic in several ways. Theoretically, the usefulness of a theory of the bureaucratic class is limited when it excludes the majority of the bureaucrats who exercise the bureaucratic control in these societies. Such a narrow deﬁnition would greatly weaken the signiﬁcance of the bureaucratic class thesis, because we are still left with the daunting task of explaining how the top echelon of the leadership can successfully ´ and monopolize resources and exercise their power in the society. Konrad Szel´enyi (1979) and Walder (1995a) addressed this issue by arguing the incorporation of intellectuals/professionals into the political elites in these societies. Their arguments logically direct our attention to those mid-level bureaucrats who are in charge of routine work functions, manage production processes, and run the workplaces. Empirically, most studies in the literature focused on mid- and lowlevel bureaucrats (heads of work organizations and “cadres” within work organizations) in developing their theoretical arguments about, and empirically studying, the bureaucratic class thesis. For example, Szel´enyi’s major theoretical arguments on the role of redistributive power in state socialist stratiﬁcation was empirically based on social surveys of housing conditions and income differentials among broadly deﬁned occupational groups – high-rank bureaucrats and leading professionals, technicians, skilled workers, unskilled workers, and agricultural laborers, to name a few (Szel´enyi 1978). In the Chinese context, much of the empirical evidence on “positional power” and “redistributive beneﬁts” was gathered by analyzing the socioeconomic conditions of cadres (political leaders, administrators, and managers) in the workplaces in the urban areas (Bian and Logan 1996; Peng 1992; Walder, Li, and Treiman 2001; Zhou 2000a) or rural cadres in villages (Nee 1989, 1992, 1996; Parish and Michelson 1996; Walder 2002). In all these studies, the empirical evidence on the bureaucratic class was based on cadres and professionals at the mid-level or low-level of the bureaucratic system. In the recent literature, many researchers no longer use the notion of “the bureaucratic class” explicitly in their writings. However, the imagery is still prevalent: There exists a privileged social group based on positional power and with stable boundaries.

12

The State and Life Chances in Urban China

In contrast to the consensus on the bureaucratic class thesis, the class position of professionals has been a controversial issue in the literature. Some scholars treat professionals as a group markedly different from redistributors, whom they identify as high party and governmental ofﬁcials and top managers of the state ﬁrms (Djilas 1957; Katz 1972; Moore 1987). Others contend that the rationalization of socialist bureaucracies has increased the role of professionals in the economic and political processes and argue that the intelligentsia has been incorporated into the bureau´ and Szel´enyi 1979; Zukin 1978). Yet cratic class (Chirot 1978; Konrad Wright (1984) suggests the position of professionals as indeterminate, treating the intelligentsia as occupying a “contradictory location” in the class structure of state socialist societies. There are also scholars who argue that high-rank professionals, as a group closely associated with the bureaucrats, should be distinguished from rank-and-ﬁle professionals, who do not share the same privileged position (Matthews 1978; Voslensky 1984). In recent years, Walder and his colleagues (Li and Walder 2001; Walder 1995a; Walder, Li, and Treiman 2000) developed a dual-path model to explain cadre and professional trajectories in state socialist societies. Walder’s model aims to address an apparent paradox in the political life chances in state socialist societies. On the one hand, the state has to rely on the bureaucratic apparatus at various levels of the administrative system to exercise its political control and implement its policies. Thus, political loyalty and adherence to the party line were the most important criteria in recruitment and promotion in the Chinese bureaucracy, often at the expense of competence and efﬁciency (Harding 1981; Lee 1991; Walder 1986). On the other hand, the “new class” theory originated in the East ´ and Szel´enyi 1979) argued that state socialist European context (Konrad governments actively recruited intellectuals into their governance structures; as a result, educational credentials play a prominent role in one’s upward mobility. The paradox is that political loyalty and competence are often at odds with each other in the allocation of life chances. If educational credential is associated with competence and independent thinking, it poses a direct threat to the centralized authority in state socialist societies. How could such a political system reward both loyalty and competence at the same time? Walder’s (1995a) dual path model is an attempt to reconcile these paradoxical aspects by calling attention to the institutional channels of bureaucratic recruitment and promotion. Walder argues that political

Redistribution and Stratiﬁcation Dynamics Under State Socialism 13 selection processes under state socialism involve both political loyalties and educational qualiﬁcations. He proposes a dual-path model in which one path emphasizes “expertise and education (but not politics) and led to professional occupations (with little decision-making authority),” and the other emphasizes “political loyalty (but not higher education) and led to inﬂuential leadership positions” (Walder, Li, and Treiman 2001, p. 194). Thus, the two segmented bureaucratic career lines were developed to recruit and promote bureaucrats using distinctive criteria and institutional processes. In this book we have followed the empirical tradition in the literature to deﬁne the bureaucratic class as consisting of those who work as heads of the work organizations or branch ofﬁces, or who hold administrative ranks in the bureaucratic apparatus. For a long time, the Chinese “cadre” system did not make a sharp distinction between administrative cadres and professionals – both belong to the same “cadre” category. Because of this fact and because of the theoretical debates on the role of professionals in the communist political order, we have also examined the role of professionals in our study. The broadly deﬁned “bureaucrats” in our study are mainly conﬁned to mid-level or lower level cadres/professionals in the Chinese bureaucracy. The majority of those labeled “high-rank cadres” in our sample are at “chu” (or associate “chu”) level, which belongs to the mid-level cadre in China’s ofﬁcial system, such as managers in charge of a workplace, or administrators in government agencies. As we noted earlier, the mid- and lower level bureaucrats are comprised of the majority of the bureaucrats in China and other state socialist societies. It is difﬁcult to imagine a theory of the bureaucratic class that excludes them. It is interesting to note that, beyond the heated discussion on the distinctions or connections between bureaucrats and professionals, there has been little attention in the literature to the relative positions of other social groups in the stratiﬁcation order. Scholars have noted the vast gap between rural and urban residents (McAuley 1979; Parish and Whyte 1978; Whyte 1995), and between manual and nonmanual workers in the Soviet context (e.g., Asselain 1987; Connor 1979), but these differences have not been conceptualized as a major demarcation of group boundaries. Redistributive Institutions Thesis Studies of state socialism have drawn particular attention to the importance of such institutions as industries/economic sectors and work

14

The State and Life Chances in Urban China

organizations in the redistribution of resources. It has long been recognized that, under state socialism, the state allocates a large proportion of resources not on the basis of occupational groups, but on the basis of industries or work organizations, largely in the form of latent beneﬁts such as housing and welfare programs. The state growth strategies traditionally favored heavy industry and sectors of “material production” (manufacturing, construction, transport) because they were seen as more crucial for rapid industrialization and the creation of national wealth than light industry (e.g., food, apparel) or the “nonmaterial sector” (services and public sector). These considerations imply that those who work in manufacturing tend to receive larger economic rewards than do those in other economic sectors. A major theoretical contribution in the studies of the Chinese social stratiﬁcation system is the recognition of another structural feature of stratiﬁcation under state socialism – the hierarchy of work organizations based on state socialist property rights. In China, the relations between work organizations and the state are hierarchically ordered in terms of property rights. Firms closer to the state in their property rights relationships enjoy preferential political treatment, better economic beneﬁts, and higher prestige than other ﬁrms. Lin and Bian (1991) distinguished three sectors of work organizations: (1) governmental agencies, public organizations, and state-owned ﬁrms, which constitute the state sector and enjoy a larger proportion of redistribution beneﬁts; (2) collective ﬁrms, often afﬁliated or sponsored by a local government, which are less privileged in the redistribution of economic resources; and (3) private ﬁrms, which are excluded from beneﬁting from state redistribution. Employees enjoy better welfare programs in the state sector than in the collective and private sectors. The organizational hierarchy is a key feature that helps legitimate the state socialist system. As Walder (1986, p. 248) observed: “The extraordinary job security and beneﬁts, the goods and services distributed directly by the state enterprise in a situation of scarcity that affects other sectors of the workforce more severely, is an important source of acceptance of the system.” These ideas have stimulated a large number of empirical studies. The study by Whyte and Parish (1984) was the earliest to recognize these distinctive types of work organizations. Bian (1994) provided the ﬁrst comprehensive discussion of the importance of work organizations and redistributive systems in urban China. Bian and Logan (1996) and Zhou (2000a) offered systematic examinations of the role of work organizations on income distribution in the reform era. Several studies examined

Redistribution and Stratiﬁcation Dynamics Under State Socialism 15 patterns of job mobility across work organizations along the hierarchical order (Blau and Ruan 1990; Davis 1990, 1992b; Lin and Bian 1991; Zhou, Tuma, and Moen 1996, 1997). Walder’s work (1992) was one of the few studies that focused on the importance of economic sectors in China. He found signiﬁcant variations in economic beneﬁts across sectors – work organizations in government, manufacturing, and educational sectors offered better beneﬁts to their employees than did the commerce sector (in the USSR context, see Titma and Tuma 1993).2 It is worth noting that these studies of the Chinese society and its stratiﬁcation processes were partly informed by and partly parallel with recent analyses of labor markets and their roles in social stratiﬁcation processes in market societies. Segmented labor market arguments pointed to discrimination in the marketplace that allocates men and women into different career lines and industries with differential economic beneﬁts (Bielby and Baron 1984, 1986; Jacobs 1989; Reskin and Roos 1990). These lines of research highlighted the importance of institutional arrangements in structuring authority relationships and patterns of resource allocation in work organizations, industries, and nation states. In the last two decades, a growing literature has also emphasized the role of organizational context and the state in stratiﬁcation in industrialized market societies. See, for example, Althauser and Kalleberg (1981), Baron and Bielby (1980) and O’Rand (1986) on the impact of organizational context; and DiPrete and Grusky (1990) and Mayer and Schoepﬂin (1989) on the role of the state. In retrospect, we can say with some conﬁdence that, since World War II, theoretical arguments about social stratiﬁcation in state socialist societies have made important contributions to comparative stratiﬁcation research by identifying distinctive structures and processes of social stratiﬁcation under state socialism (see Bian 2002 for a recent review of research on China). However, our knowledge of social stratiﬁcation processes under state socialism is far from satisfactory. After an extensive review of the literature, Zhou and Suhomlinova (2000) concluded that there were few systematic studies of these multiple channels of redistribution or studies that scrutinized the empirical implications of these theoretical arguments. This was partly due to the fact that there had been a scarcity of individuallevel information on social positions and beneﬁts in these societies before the collapse of the state socialist regimes in the late 1980s. Another contributing factor was that there was a lack of intellectual dialog among 2

There is also a large literature in Chinese. See Li (1993), Li and Li (2000), Li, Zhou, and Li (1996), Li (1992), Li and Zhang (2000), and Lu (1989).

16

The State and Life Chances in Urban China

scholars conducting research in different societal contexts. As a result, we have not seen the kind of systematic and cumulative research or theoretical and empirical progress in this area that are typical in other areas of social stratiﬁcation research. Stratiﬁcation Dynamics Under State Socialism Although the institutional structure of state socialism is important, an emphasis on these structural conditions alone is inadequate for understanding stratiﬁcation processes in these societies. In contrast to industrialized market societies, in China structural conditions – occupational and organizational locations – are subject to change, especially to political processes directed by the state and state policies. A central theme of this study is the importance of stratiﬁcation dynamics in understanding the consequences of state socialist redistribution on individual life chances. We use the concept of stratiﬁcation dynamics to refer to a distinctive pattern of changes in stratiﬁcation structures and processes in China that are often imposed from the top down and that often have noncumulative and disruptive impacts on individuals’ life chances. Our central argument can be summarized as follows: Life chances under state socialism, in contrast to those in industrialized market societies, are affected not only by the distinctive structural conditions of the state redistributive system (e.g., locations in the organizational hierarchy), but also, and to a much greater extent, by political dynamics and concomitant shifts in state policies that directly and immediately alter the opportunity structures and re-value various social resources for individuals’ life chances. To develop this line of argument, we need to explicate the mechanisms through which such dynamics affect individual life chances and patterns of social stratiﬁcation outcomes, and to draw empirically testable implications. Let us begin with an empirical observation: The political twists and turns induced by shifting state policies have been a salient social fact throughout the political history of the People’s Republic of China. A useful angle from which to appreciate this phenomenon is to consider the political/economic campaigns in the history of the PRC (see Table 1.1). Political and economic campaigns here refer to those state-initiated mobilizations to carry out new policy initiatives. Such mobilizations were often implemented in the form of large-scale mass participation in political and economic processes, and beyond routine-based activities organized by state bureaucracies. Several features of these campaigns are relevant to our discussion here.

Redistribution and Stratiﬁcation Dynamics Under State Socialism 17 table 1.1. A Partial List of Political Campaigns in China, 1949–1994 Time May 1950–Dec. 1950 Oct. 1950–Dec. 1952 May 1951 Oct. 1951 Nov. 1951–Fall 1952 Sept. 1951 Dec. 1951–Oct. 1952 Jan. 1952–Oct. 1952 Feb. 1952–Spring 1954 Jan. 1953 Sept. 1953 Oct. 1954 Jan. 1955 July 1955–Dec. 1957 Apr. 1957 June 1957 Aug. 1957 May 1958 Aug. 1958 July 1959–Aug. 1959 Dec. 1960 March 1963 Sept. 1963 Dec. 1963 Sept. 1964–Spring 1966 1966–1976

Campaigns/Events CCP rectiﬁcation campaign The movement to suppress counterrevolutionaries Critique of movie “Wuxunzhuan” Campaign for production enhancement and thrift The thought reform movement Mutual assistance and cooperation movement The “san fan” movement The “wu fan” movement CCP rectiﬁcation The new “san fan” movement The Anti-Gao-Rao clique campaign Critique of Yu Ping Bo’s Hong Lou Meng Study The Anti-Hu Feng clique Campaign The campaign to eliminate counterrevolutionaries CCP rectiﬁcation campaign The anti-rightist’s campaign The socialist education movement The Great-Leap-Forward movement Commune and collectivization movement The Anti-Peng clique campaign Rectiﬁcation movement in rural areas The “5-Anti” movement The socialist education movement Rectiﬁcation movement in literary areas The “4-Clean” campaign The Cultural Revolution

Affected Area Political Political Cultural Economic Education and culture Rural Political, economic Political, economic Political Political Political Culture Cultural Political Political Political, cultural Political, economic Economic Political, economic Political Rural Political, economic (urban) Rural Cultural Rural All areas (continued)

18

The State and Life Chances in Urban China table 1.1 (continued)

Time May 1978 Dec. 1982 Aug. 1983–Nov. 1983 Oct. 1983 1986 1989 1992

Campaigns/Events The “practice and truth” debate The “cleaning 3-type” movement The anti-criminal activities campaign The rectiﬁcation movement Anti-spiritual pollution campaign Anti-westernization campaign Economic liberalization

Affected Area Political Political Economic, social Political Political, cultural Political, cultural Economic

Sources: Information is drawn from Research Ofﬁce (1991) and from Feng and Liu (1992).

First, political and economic campaigns were not ad hoc solutions but an integral part of political and economic organization in China. As Table 1.1 shows, these campaigns took place throughout the entire history of the People’s Republic of China, serving as a critical and institutionalized means for political mobilization and resource allocation. They were often initiated by state policy shifts, such as changes in development priorities, new initiatives, or organizational responses to crises. There have been noticeable changes in the use of political campaigns in the post-Mao era, indicating signiﬁcant changes in state-society relationships. We will discuss this issue in Chapter 12. Second, because of the nature of such political mobilizations, they often drew a large number of individuals into the nonroutine processes and affected their lives directly and often disruptively. Each campaign generated different political opportunities, together with its victors and victims. Activists adhering to the political line of the campaign were rewarded with promotions and gains in political status and economic beneﬁts. Victims of the political campaigns, on the other hand, were often purged from their positions and lost redistributive beneﬁts associated with these positions (Rosen 1982; Shirk 1982; Walder 1986). For instance, the “Anti-Rightist Campaign” purged over 530,000 “rightists,” over 10 percent of the intellectuals at the time, who were rehabilitated only twenty years later (Shao 1998). The Cultural Revolution and the “Great-Leap-Forward” episodes had even wider impacts. Third, these campaigns were not merely conﬁned to the political arena but affected all walks of life. For example, the nationalization and collectivization campaigns of the 1950s and the “Great-Leap-Forward” of 1958 took place in the economic arena. But, like political campaigns, they affected individual life chances in a dramatic manner and in a broad scope. The targets of these campaigns also varied over time. At different

Redistribution and Stratiﬁcation Dynamics Under State Socialism 19 points in time, different social groups – capitalist property owners, intellectuals, and even bureaucrats – became the targets of political campaigns. Consequently, a focus on stable group boundaries fails to capture these stratiﬁcation dynamics. That state policies have a direct and immediate impact on individual life chances has been well recognized in studies of politics and society in China. Parish (1984) called attention to the radical destratiﬁcation state policies since the 1950s that drastically changed the stratiﬁcation patterns and the life chances of individuals. Whyte (1985) discussed how state policy changes drastically altered opportunity structures, generating distinctive cohort patterns. Davis (1990; 1992b) conducted systematic studies of the impacts of radical state policies on the lives of urban residents and found that these policies severely disrupted the trajectories of the urban youth. Gold (1991) also developed this theme in his study of the impact of the state on youth’s life course. China is not alone in this regard: Similar policy cycles and political dynamics were also documented in the former Soviet Union and Eastern Europe, in the form of political successions (Bunce 1980), cycles of social protests (Hankiss 1989), cycles of reforms (Jermakowicz and Follis 1988), or downward mobility due to political purges (Viola 1987). Szel´enyi and Manchin (1987) showed that the state’s repressive policy of collectivization interrupted embourgeoisement in rural Hungary in the earlier period. These studies notwithstanding, main theoretical models and arguments in the literature have focused primarily on the structural, but not the dynamic, aspects of state socialist redistribution. Considering the profound impacts of these macro-political events on individuals’ lives, it is surprising that there were no systematic theoretical and empirical efforts to develop explanations of the implications of these macro-political processes for social stratiﬁcation. Are these political dynamics so idiosyncratic, ad hoc, context-speciﬁc, and their impacts on the society so transient that they defy sociological theorizing; and should the best be left to a historical account? To the contrary, we argue that the observed political turbulence is rooted in the institutional arrangements of state socialism. The inherent tensions in state socialism generate several types of organizational failures in the formation and implementation of state policies. These organizational failures and consequent policy disasters, when put together, generate large-scale, often-unexpected, and bottom-up political pressures on top leaders. And they trigger fractional conﬂicts and force sudden, dramatic shifts in state policies and induce political twists and turns that

20

The State and Life Chances in Urban China

drive the observed stratiﬁcation dynamics. Such political dynamics often take the form of policy ﬂuctuations or oscillations between centralization and decentralization (Lardy 1975), between political control and economic development, between “Red” and “Expert,” and among different policy alternatives (Nathan 1976; Zhou 1992a). The discussion thus far has motivated two main observations: First, based on earlier studies of state socialist institutions, we recognize the enormous capacities of the state and state policies to affect individual life chances as the result of the monopoly of resources by the centralized authority, the close ties between the state and individual citizens, and the disruptive nature of the political instruments used to implement these policies. Second, by considering the inherent tensions within state socialist institutions we can also identify political processes that induce unexpected shifts in state policies and generate the observed stratiﬁcation dynamics. These considerations call attention to the consequences of unexpected, dramatic, and disruptive shifts in state policies for individuals living in these political environments. The stratiﬁcation dynamics are induced by frequent shifts in state policies and dramatic changes in macro-political processes, which are closely related to the following characteristics of state socialism as we discussed earlier: The centralized state possesses the organizational capacities for resource allocation and transfer across economic sectors and among social groups at a scale much larger and at a pace much faster than market-based economic institutions. At the same time, mass mobilization, often in the form of political campaigns, brings individual citizens in direct contact with the state and the political processes through the implementation of state policies. We now turn to discuss the processes and empirical regularities through which the state and state policies affect stratiﬁcation dynamics and, in turn, individual life chances. These implications will guide the empirical studies reported in this book. Our ﬁrst proposition is that state policies can drastically alter opportunity structures and individual life chances regardless of the socioeconomic status of various groups. In all societies, external shocks (e.g., economic crises, wars, and natural disasters) adversely affect all social groups, but the extent of the effects varies with social groups (Elder 1974). To appreciate the implications of this proposition in the Chinese context, it is important to recognize that stratiﬁcation dynamics are often secondary consequences of state policy shifts in other domains. That is, the dynamics of state policy shifts are often organizational responses to crises or

Redistribution and Stratiﬁcation Dynamics Under State Socialism 21 changes in political agenda and hence they often operate independently of the stratiﬁcation processes in place. For instance, drastic transfer of resources across economic areas is often triggered by changes in developmental priorities, or as response to failures of previous policies. Such policy shifts have immediate and direct impacts on individual life chances because these changes simultaneously alter opportunity structures as well as preexisting access to resources. Consequently, we should observe that the distribution of opportunities varies systematically with shifts in state policies. That is, state policy shifts often expand or reduce opportunities in ways that affect all social groups and individuals in a similar manner. Earlier studies often treat these policy changes as contexts and focus on variations in life chances based on group-speciﬁc factors (e.g., differences in positional power and access to resources). Given the dominance of the state and political dynamics under state socialism, however, it is important for us to treat state policies and macro-political processes as crucial causal mechanisms in explaining individual life chances. Because these policy changes are often independent of social stratiﬁcation processes, the capacities of those in beneﬁcial social positions to buffer these policy shocks and to protect their self-interests are limited; as a result, policy impacts are much greater, less predictable, and more uniform in the centralized state socialist societies than in other institutional contexts. In discussing the “politics of life chances” in China, Whyte (1985) traced four historical periods in state socialist China and observed that state policies affected all social groups in comparable ways. Sometimes they raised the general level of economic development and opportunities, as occurred during the 1950s when expansion of the state and economy created more opportunities and widespread upward mobility. Contrarily, policy disasters (e.g., the “Great-Leap-Forward” and its aftermath in the 1958–1963 period) reduced living standards and economic development and adversely affected the life chances of individuals across organizational and occupational boundaries. For most citizens, variations in social positions in the social structure and the economic beneﬁts associated with returns to these positions were minor relative to dramatic changes in the overall environment and in opportunities induced by state policy shifts. Therefore, to explain patterns of social stratiﬁcation under state socialism, we need to go beyond a narrow focus on the conventional stratiﬁcation arena such as intergenerational mobility or status-attainment processes and to pay special attention to macro-political processes and larger historical contexts.

22

The State and Life Chances in Urban China

Our second proposition is that the state can change the relative socioeconomic status of social groups by drastically reevaluating their resources and by shifting the criteria of political selection. In state socialist societies, resource transfers are typically based on ideological and political considerations rather than the demand and supply interactions in the marketplace. In the past ﬁve decades of state socialist China, state policies changed frequently and substantially, for example, between political control and economic development in governance priorities, between competence and political loyalty in promotion criteria, and between centralization and decentralization in administrative structure. Drastic shifts in the relative status of social groups resulted. For example, state policies that redirected resources to children of the working class (e.g., in educational opportunities, entering into higher status organizations and jobs, and recruitment into the Communist Party) substantially changed the patterns of social mobility (Parish 1984). This line of argument is based on the observation that stratiﬁcation under state socialism operates as a top-down process. In this context, boundaries of social groups – occupational categories – are constructed not by market-like competition, but through deliberate state policies. In the Mao era, the state systematically constructed classes in terms of occupations, in particular, workers, peasants, and intellectuals (White 1989). For example, a long-standing policy in the Mao era was to favor those in workingclass occupations in educational attainment, promotion, and placement in bureaucratic positions (Lee 1991). By eliminating autonomous intermediate layers of social groups between the state and the society, the state socialist institutional structure directly linked each citizen with the state, thus reducing all social groups to a similar structural position in subordination to the state and its bureaucratic organizations. In stateowned ﬁrms, workers’ wage grades, promotion opportunities, ﬁnancing for housing construction, and health insurance were decided by ministries in Beijing. Although local collective enterprises are not under the direct administrative control of the state, they nonetheless are subject to state intervention. As a result, boundaries and the relative positions of social groups (classes), organizations, and localities are extremely sensitive to state policy considerations rather than to market-like transactions (Zhou 1993b). Mass mobilization and resource transfers across sectors subject different social groups to the same state policy vibration, and social boundaries are ﬂuid and vulnerable to state policy shifts. In tandem with state policy shifts, resources are redistributed across sectors, and the life chances of

Redistribution and Stratiﬁcation Dynamics Under State Socialism 23 social groups and individuals in various social sectors change accordingly. Social positions are important, but their capacities are more limited and less effective than they are in other societal contexts, where a more direct link between social positions and resources is established. In summary, macro-political dynamics directly and decisively affect the life chances of all social groups and individuals in China, regardless of their structural locations. In a politicized society, symbolic boundaries between “us” and “them” are critical, and shifts in state policies frequently redraw such boundaries. Indeed, state policies may change the very meaning of “resources,” thus altering the nature of social inheritance. For example, educational qualiﬁcations were treated as political liabilities rather than as personal resources during the Cultural Revolution. This is evident in the fact that white-collar occupations had not consistently delivered the kinds of economic beneﬁts and prestige common in other societies (Davis 1992b). Therefore, to understand the meaning and effect of resources associated with social positions requires substantive interpretations of the historical context. One important empirical implication of this process is that the effects of individual or family-based attributes and capital on life chances vary over time. This is in sharp contrast with the literature on social stratiﬁcation in industrialized market societies that rests on the core idea that human, social, cultural, and political capitals (e.g., parental socioeconomic status and education) invariably have positive returns and enhance individuals’ upward mobility. Chinese political dynamics imply that conventional capitals do not have consistent impacts on life chances over time: In the Chinese experience, today’s capital may become tomorrow’s liability, and vice versa. Take education as an example. In industrialized market societies, education contributes signiﬁcantly to individuals’ upward mobility (Blau and Duncan 1967; Hout 1988). In China, however, the effect of education on individuals’ life chances has varied dramatically depending on contemporaneous state policies. In the early years of communist China, the Communist Party stressed candidates’ “revolutionary experience” and downplayed education (Lee 1991). During the Cultural Revolution, Maoist leaders treated intellectuals with suspicion and hostility and made serious attempts to recruit those with peasant or worker backgrounds into management positions. Indeed, those with higher education were often the target of political purges, as occurred in the Anti-rightist Campaign (MacFarquhar 1960) and the Cultural Revolution (Tsou 1986). In the

24

The State and Life Chances in Urban China

post-Mao era, education has again become an important factor for upward mobility. The new cadre policies adopted in 1980 emphasize educational qualiﬁcations and competence in recruitment and promotion. These observations suggest that, in a ﬂuctuating environment, human, social, and political capital does not necessarily accumulate and give consistently positive returns. As another example, consider the common observation that the Chinese political system rewards those who are politically loyal and clients of powerful persons (Lee 1991; Walder 1986). Frequent shifts in state policies mean that the political activists committed to the political line of the top leaders in one historical context may become the targets of purges after a policy reversal. Ties to the ousted elite become a liability rather than an asset. For instance, those who actively followed Mao’s radical political course have been systematically discredited in the post-Mao era. And political activists rewarded with admission to colleges during the Cultural Revolution (gongnongbing xueyuan) have been discriminated against in the post-Mao era. The second empirical implication is that the effects of structural positions in occupational and organizational hierarchies on life chances vary over time. In contrast to the gradual expansion of opportunities with industrialization in market societies, occupational and organizational hierarchies in China are constructed by political processes and shifts in state policies dramatically alter life chances of social groups from time to time. For instance, the social status of occupational groups is politically constructed in accordance with ofﬁcial class identiﬁcation in Communist ideology. The high political status of industrial workers (chanye gongren) was not based on their prestige or income, but on the ideological considerations of the socialist state. But as a political construction, the status of occupational groups is sensitive to state policy shifts. Workers were favored in Mao’s era but their political status has declined substantially, and even reversed, in the reform era. In contrast, intellectuals had low status in the Mao era but enjoy a higher status in the reform era. These dramatic changes across historical periods suggest a pattern of inconsistent effects of structural conditions over time on individual life chances. Attention to stratiﬁcation dynamics raises important theoretical issues in the comparative study of social stratiﬁcation. First, to develop satisfactory explanations of stratiﬁcation patterns, we must take seriously evolving political processes in a societal context. In fact, these considerations raise a series of questions about the exclusive focus of the earlier

Redistribution and Stratiﬁcation Dynamics Under State Socialism 25 theoretical models on stable structural positions and associated advantages or disadvantages in social stratiﬁcation processes under state socialism. For instance, if stratiﬁcation dynamics are introduced by macropolitical processes that are independent of the conventional stratiﬁcation mechanisms, the advantages of social positions to resist these forces are likely to be limited. If the making of social classes is heavily inﬂuenced by a top-down process initiated by the state and the boundary and status of the bureaucratic class are to a large extent dependent on the state and state policies, then an important question is whether and to what extent the bureaucratic class can maintain and protect its interests, and reproduce its class position through intergenerational mobility. In other words, the proposed arguments on stratiﬁcation dynamics point to the limitation of the bureaucratic class to accomplish these goals and predict a less effective, less cumulative role of social positions. Second, stratiﬁcation dynamics have important implications for understanding institutional changes in transitions from state socialism. As is well known, political loyalty has been an important mechanism for upward mobility in state socialist societies. However, given shifting state policies, political loyalty in one policy “regime” may become suspect when policies change. Over time, different cohorts of political activists may be recruited and promoted into the bureaucratic system with divided loyalties and internal conﬂicts. This may lead to political instability and provide impetus for political change. More important, recent studies of the transformation of state socialism have observed the path-dependent characteristics of institutional changes (Rona-Tas 1994; Stark 1996; Stark and Bruszt 1998). Individual biography may be an important mechanism to trigger a recombination of existing routines and generate emergent institutions. We will return to these issues in the concluding chapter of this book. The recognition of stratiﬁcation dynamics also raises important methodological issues in studying social stratiﬁcation processes under state socialism. Clearly, snapshots of a particular period or sector are not adequate and may be misleading to capture the processes of social stratiﬁcation in these societies. We need a historical approach to understand the patterns of social stratiﬁcation and changes in individuals’ life chances; that is, we need to bring in historical contexts and changes over time. Moreover, the impact of state policy shifts on life chances may vary for individuals at different stages of the life course. Thus, we need to consider both historical periods as well as life course stages to examine the consequences of stratiﬁcation dynamics. In this study, we address these

26

The State and Life Chances in Urban China

issues by analyzing individual life history data and by adopting analytical strategies that examine changes over time. Finally, we point out that the main arguments on stratiﬁcation dynamics developed here are theoretical hypotheses, which need to be tested, revised, or even refuted in empirical studies. And there are considerable indeterminacies in the proposed arguments. For example, to what extent do the stable redistributive institutions countervail the stratiﬁcation dynamics when the two are at odds to each other? To what extent can the bureaucratic class succeed in resisting state policies that adversely affect their life chances? These are empirical questions that have important implications for the proposed theoretical arguments here, and they are to be examined in empirical studies in the subsequent chapters.

institutional transformation and stratiﬁcation processes in the reform era Since the 1980s, China has been at the center of social science research on the transformation of state socialism. There are several reasons for the enormous scholarly interest in China’s economic transformation: The most important one is, of course, China itself. The unique Chinese experience – its impressive and sustained economic growth for over two decades, and the continuing, central role of the socialist state – provides an important case for the study of transitional economies. In sociology, important theoretical ideas about transitional economies – Nee’s market transition theory and Walder’s arguments on the role of local governments are developed in the Chinese context and have stimulated active research. An ongoing debate in recent studies of the transformation of state socialist societies centers on how to explain (and predict) the processes and mechanisms of large-scale social changes and their implications for the social stratiﬁcation order in these societies. The main issues under debate are closely related to long-standing sociological interest in explaining social changes and processes of social stratiﬁcation. They also reﬂect sociologists’ renewed interest in the role of social and political institutions in economic activities and how social institutions persist and change. The study of social stratiﬁcation processes provides an important lens from which we can address important issues about sources of institutional changes and continuity. That is, fundamental changes in institutions in the transformation of state socialism should be reﬂected in changes in social stratiﬁcation processes. Not surprisingly, comparative sociologists have embraced the

Redistribution and Stratiﬁcation Dynamics Under State Socialism 27 opportunity, and a large number of studies emerged to address these issues (Bian and Logan 1996; Cao 2001; Gerber and Hout 1998; Lee 1995; Nee 1989, 1991, 1992, 1996; Peng 1992; Rona-Tas 1994; Stark 1996; Tang and Parish 2000; Walder 1995a, 1996b, 2002; Xie and Hannum 1996; Zhou 2000a, 2000b; Zhou and Moen 2001a; Zhou, Tuma, and Moen 1997). Situated in this context, the second theme of this book is to understand sources and extent of institutional transformation in post-Mao China. To accomplish this task, we analyze and contrast changes and continuity in social stratiﬁcation processes over time. In the second part of the book, we report a series of empirical studies that look into the evolution of allocative processes and individual life chances before and in the reform era. Our study focus on urban China, where China’s redistributive economy has reached its most elaborate form, provides a more appropriate examination of the changes in state socialist redistribution. In this section, we provide an overview of the general issues in this ﬁeld that guided our empirical studies in the second part of this book. The Theoretical Debate To risk oversimpliﬁcation, one may characterize two general lines of arguments in this debate. On one hand, Nee’s (1989) market transition theory emphasizes the role of emerging markets that compete with and undermine state socialist redistributive economy and alter the stratiﬁcation order. On the other hand, other scholars emphasize various aspects of the political economy of transformation in which existing political and economic institutions mold emerging economic institutions. Based on a series of studies of income inequality in rural China, Nee’s market transition theory (1989, 1991, 1992, 1996) was one of the earliest attempts to theorize about sources of social changes in transitions from state socialism and their implications for changes in the social stratiﬁcation order. Nee’s bold theoretical statements and insistent defense of his positions have stimulated active research and debates on both alternative theoretical explanations and empirical assessments of the competing theoretical arguments. Central to market transition theory is its emphasis on the importance of emerging market economies. In Nee’s view, markets and state socialist redistribution represent two fundamentally different logics of resource allocation. The emergence of a market economy introduces new mechanisms of resource allocation that challenge and undermine state socialist

28

The State and Life Chances in Urban China

redistributive institutions. Nee’s earlier arguments (1989, 1991) predicted two fundamental changes: First, the emergence of a market economy alters the opportunity structure shaped by a redistributive economy; and second, changes in mechanisms of resource allocation lead to changes in social stratiﬁcation order. On the other hand, other scholars have developed several alternative, competing explanations. These explanations consist of several streams of arguments emphasizing the increasing role of local governments (Walder 1995b), the persistence of political power (Bian and Logan 1996), the conversion of political power to economic resources (Rona-Tas 1994), changes in “political markets” (Parish and Michelson 1996), institutional reconﬁguration (Stark 1996), and the institutional arrangements of work organizations (Zhou et al. 1997). Although they differ in emphases, these arguments share the main disposition that the ongoing transformation processes and emerging economic institutions are circumscribed by and intertwined with existing political, economic, and social institutions. Thus, an explanation of the sources and directions of the transformation must take seriously existing institutional arrangements and explain how they interact with and shape emerging new institutions. A central issue in the debate is the role of institutions in the processes of social changes. Nee’s focus is on new market institutions whose advance forges new interests and relentlessly pushes aside, if not dismantle, “old” economic and political institutions. On this account, he criticized his critics as to “emphasize structural continuity with unreformed state socialism” and to focus “only on changing incentives for political actors in the state organizational hierarchy.” Yet a similar critique can be made of market transition theory. Despite the claim to “incorporate the state as a potent causal force,” Nee’s work has made little effort to consider how the interactions between existing political and economic institutions and emerging institutions affect patterns of social stratiﬁcation. For instance, Nee (1996) advocated an institutional analysis of market transition in China, but the path of change he emphasized is a process in which “power – control over resources – shifts progressively from political disposition to market institutions” (p. 910). Nee’s arguments leave the impression that the entire outcome of institutional changes can be explained by the advance of market economies, despite the resistance of “old” institutions. The second major difference is related to the role of agency and interests. Nee emphasized the importance of new interests associated with emerging markets. As Nee (1996, p. 910) argued: “Whereas opportunities

Redistribution and Stratiﬁcation Dynamics Under State Socialism 29 for advancement were previously centered solely on decisions made by the redistributive bureaucracy and within the economy controlled by it, markets open up alternative avenues for mobility through emergent entrepreneurship and labor markets.” Along with these changes, “economic actors strive to institute new rules of competition and cooperation that serve their interests, both through informal arrangements and through formal institutional channels.” In contrast, the emphases on existing social and political institutions led other scholars to recognize the importance of incumbents and vested interests in the transformation processes. The prevalence of political authority, bureaucratic bargaining, governmental intervention in risk and beneﬁt sharing have been noted in several studies and commentaries (Bian 1997; Bian and Logan 1996; Naughton 1996; Oberschall 1996; Oi 1992; Parish and Michelson 1996; Shirk 1993; Walder 1995b, 1996b). The incumbents may use their positional power and privileged access to decision-making processes to inﬂuence state policies and governmental regulations in favor of their interests. Or they may be in an advantageous position to capitalize on opportunities and economic beneﬁts, relative to other social groups. Therefore, their positional power and advantages in stratiﬁcation processes may persist even when the role of market institutions increases (in the East European context, see Gerber 2000; Rona-Tas and Guseva 2001). We should point out that, over time, Nee has modiﬁed his arguments and incorporated in his arguments the role of the state, local corporatism, networks, and the persistence of political power (Nee and Cao 1999, 2002; Nee and Lian 1994). In our view, however, the deﬁning characteristic of market transition theory is its primary emphasis on the institutional logic of markets. As Nee and Mathews (1996, p. 422) put it: “According to market transition theory, the causes giving rise to a change in the mechanisms of stratiﬁcation in reforming and postcommunist societies are linked inextricably to the expansion of market institutions.” Over time, scholars came to appreciate the increasingly important role of emerging markets in China’s economic transformation as emphasized in Nee’s market transition theory, but they still differ on the speciﬁc mechanisms through which the emerging markets affect institutional changes in China. Substantively, the debate in part stems from different emphases given to different processes of social changes. Like any large-scale social changes, the transformation of state socialism involves multifaceted processes, in which the expansion of markets is but one, albeit a very important one,

30

The State and Life Chances in Urban China

of many. Such transformations inevitably produce a variety of patterns of changes that defy a single theoretical logic. In this light, there is room for a theoretical synthesis (Szel´enyi and Kostello 1996). In the following, we propose a conceptual model that emphasizes the interplay between politics and markets as a coevolutionary process and advocates a focus on the processes of institutional change that provides a basis for a theoretical synthesis among competing arguments. Coevolution Between Politics and Markets As a basis for a theoretical synthesis, we propose a conceptual model that treats institutional changes as processes of coevolution between politics and markets. By “politics” we refer to patterns of interest articulation in the political arena associated with both current and emerging political and economic institutions. By “markets” we refer to the modes of resource allocation and economic transactions that take place through price systems and involve “autonomous” economic agents. The interplay between politics and markets in transitional economies has been emphasized in the recent literature (e.g., Bian and Logan 1996; Guthrie 1997; Parish and Michelson 1996; Rona-Tas 1994; Stark 1996; Walder 1996). We highlight some theoretical considerations along this line of reasoning. Using the proposed model, we intend to portray an image of intrinsic interconnectedness between politics and emerging markets in the course of institutional changes in China, where state-initiated reforms and emerging markets have been two major sources of economic transformation. Both interest politics and markets co-evolve in interactions with each other: In some areas, they compete with and constrain each other; in other areas, they mutually reinforce each other; in still others, they adapt to and transform each other in the process. The central idea in the imagery is that neither one can be understood alone without a careful and substantive understanding of the other. The premise of the proposed model is based on an insight from institutional theories in sociology and economics that politics and markets are not antithetical and that the state plays a critical role in setting up institutional rules within which markets operate (Campbell, Hollingsworth, and Lindberg 1991; Esping-Andersen 1990; Evans, Rueschemeyer, and Skocpol 1985; Wade 1990). Two ﬁndings in this literature are especially relevant for our discussion here. First, the expansion of markets is not a self-evolving process. Rather, economic activities are embedded in and

Redistribution and Stratiﬁcation Dynamics Under State Socialism 31 constrained by the social context and historical path of change (Granovetter 1985; Hamilton and Biggart 1988; Stark and Bruszt 1998). Second, in the coevolution of politics and markets, the state is an active actor, with its own interests and preferences. As a result, changes in economic institutions inevitably reﬂect a compromise between competing interests and institutions. We are in complete agreement with the view of new institutional economics that “the conception of economy is of an evolving, open system in historical time, subject to processes of cumulative causation – instead of approaches to theorizing that focus exclusively on mechanical equilibria” (Hodgson 1994, p. 69). What are the mechanisms that govern this coevolutionary process? Nee’s market transition theory best elaborated one such mechanism – the competition between existing and new economic institutions. The main mechanism here is that of efﬁciency in market competition. The expansion of markets introduces new mechanisms of resource allocation and new ways of organizing economic activities. Redistributive economic institutions, in contrast, rely on the political authorities and favorable regulatory policies to hold on to their advantages in economic production and transactions. So long as market-induced economic activities outperform those of the traditional (redistributive) institutions, they tend to grow and increase their share in the national economy. In this sense, the emerging market economies compete with and undermine the state socialist redistributive economy. An equally important mechanism is interest politics in the political arena. Social and economic institutions are associated with organized interests. The course of any institutional change ultimately depends on how various interests play out in the political arena. In comparison with the new interests associated with emerging markets, the vested interests in current political and economic institutions have considerable advantage (1) in their access to political authorities and to the processes of policy making and implementation, (2) in their organizing capacity (bureaucratic politics) in protecting and advancing their interests (Lieberthal and Lampton 1992; Shirk 1993). As Parish and Michelson (1996, p. 1043) pointed out: “[T]he type of market compromises that will be arrived at depends as much on bargaining in more participatory political systems as it does on some underlying economic dynamic.” This leads us to consider the third mechanism – the central role of the state – in this coevolutionary process. Economic activities, including markets, operate within the institutional rules set by political authorities. From the logic of markets, it is conceivable that the competitive

32

The State and Life Chances in Urban China

advantages of market-based institutions would force the state and state policies to evolve in favor of market institutions and the new interest groups associated with these institutions. So long as the state has its own interests in economic growth and national wealth, the more competitive market institutions, according to this logic, would eventually give the “new interests” an upper hand in “capturing” the state. However, the state may have its own interests that are not necessarily in congruence with the economic agents in the market place. As North and Thomas (1973) noted, “the ﬁscal needs of government may induce the protection of certain property rights which hinder rather than promote growth; therefore we have no guarantee that productive institutional arrangements will emerge” (p. 8). Indeed, concerns about political stability, legitimation, and historical traditions have led to a variety of institutional rules that facilitate as well as constrain markets even in industrialized market economies. Given the historical role of the state in China and the prevalence of vested interests associated with existing institutions, there is no reason to doubt that the remaking of institutional rules in China’s economic transformations will be heavily inﬂuenced by the vested interests and the state’s own interests. Even the dismantling of the redistributive state would not lead to a break with the past, and vested interests and “old” institutions are likely to continue to exert their inﬂuence, as is evident in the experience of other former state socialist societies (Gerber 2002; Gerber and Hout 1998; Rona-Tas 1994; Stark 1996). Consequently, both markets and politics coevolve in response to each other. Given the constraints of institutional rules and the prevalence of political authorities, market activities take distinctive forms in China’s transitional economy: Many private enterprises were disguised as “collective” ﬁrms in the reform era; all kinds of economic agents cultivated relations with political authorities and engaged in rent-seeking behavior (Parish and Michelson 1996); economic transactions, even when operated through the price system, were often brokered by political authorities (Boisot and Child 1996). These demands for and beneﬁts from political authorities reinforce the latter’s role in economic life and foster emerging interests that promote the roles of both politics and markets. In this sense, economic activities in the marketplace are molded by politics. On the other hand, the nonstate sector’s increasing contribution to state revenue provides positive feedback for the state to adopt policies to encourage market expansion; governmental agencies are gradually changed from redistributors to regulators; state-owned ﬁrms, though constrained by political authorities, are being pushed into market competition. In this process,

Redistribution and Stratiﬁcation Dynamics Under State Socialism 33 both politics and markets have been transformed by each other. The proposed conceptual model, we hope, serves the purpose to highlight the interconnectedness among these processes and to insist on understanding them in relation to one another. How do we assess the outcomes of this coevolutionary process? An important implication of the proposed model (as well as other models that emphasize one or another particular process) is that there is a large area of model indeterminacy with respect to its power of empirical predictions. For instance, in the processes of institutional change, the recognition of the presence of a market institution (e.g., contract, labor market) does not necessarily lead to any predictive power without a substantive understanding of how it interacts with and is constrained by existing institutions. In this light, we should expect both continuity and changes in the transformation of state socialism in China, governed by multiple processes. Our empirical studies focusing on this theme aim to examine and assess the extent and direction of such institutional changes and continuity and to understand the underlying mechanisms that govern the transformation processes.

looking ahead In summary, we pursue two themes in this book: (1) a study of the interplay between redistribution institutions and stratiﬁcation dynamics under state socialism and (2) an exploration of institutional changes in the recent economic transformation in China. In what follows, we will report a series of studies based on the life histories of a national sample of urban residents, whose information will be provided in detail in the next chapter. We will follow this sample of urban residents over different stages of their life course and over the forty-ﬁve-year history of the People’s Republic of China, examining their life experiences in educational attainment, their entries into the labor force, their workforce experiences such as patterns of promotions and job shifts, and variations in their economic beneﬁts such as personal income and housing space. Corresponding to the two themes mentioned previously, the main body of this book is divided into three parts. After an overview of the historical context and description of the data and analytical strategies for this book in Chapter 2, Part One provides systematic evidence on social stratiﬁcation patterns. We examine the impact of multiple stratiﬁcation processes – occupational groups, hierarchies of organizations, human and political capital, and, above all, the impacts of shifts in state policies. The chapters

34

The State and Life Chances in Urban China

are organized roughly along the sequence of stages in one’s life course: educational stratiﬁcation (Chapter 3), patterns of entry into the labor force (Chapter 4), a close look at the life experiences of one cohort – the children of the Cultural Revolution (Chapter 5), and bureaucratic career patterns (Chapter 6). These empirical studies attend to key research issues related to the interplay between redistribution and stratiﬁcation dynamics. In Part Two, we focus on the second theme of this study – assessing institutional changes in the era of economic transformation since the 1980s. The chapters in this part are organized by selected topics of theoretical interest. We begin with an examination of how patterns of redistribution have changed in the reform era in those areas of manifest economic beneﬁts (personal income, Chapter 7) and latent beneﬁts (housing allocation, Chapter 8). We then examine changes in patterns of job shifts across types of work organizations and economic sectors to assess the extent of institutional changes (Chapter 9) and how the institutional transformation affects individuals at different stages of the life course (Chapter 10). In Part Three, we summarize the main ﬁndings of this study and draw implications for social stratiﬁcation research and studies of state socialism as well as other redistributive systems. Based on the preceding chapters and corroborative evidence, we revisit and assess main theoretical issues in the literature of state socialist redistribution and stratiﬁcation (Chapter 11). We conclude this book by drawing its implications for understanding institutional changes in the transformation of state socialism (Chapter 12). Finally, a note on the style of presentation in this book. The studies reported in this book are conducted in the tradition of sociological research. Consistent with the research style in this discipline, oftentimes we develop theoretical arguments and hypotheses to guide our research design and statistical analyses on a particular research topic. The statistical ﬁndings and results are then interpreted in light of these theoretical ideas and hypotheses. We tried to minimize academic jargon whenever possible. Nevertheless, the presentation style will unavoidably reveal the traits of academic training and practice in which these studies were situated. We hope that the reader is patient and forgiving as we take on this journey to make sense of the historical changes in social stratiﬁcation processes in urban China.

2 Overview: Historical Context and Research Design

This chapter provides an overview of the evolution of the state socialist redistributive institutions in the People’s Republic of China, from 1949 to the 1990s. We accomplish this goal in two ways: First, we use ofﬁcial statistics and documentary evidence to introduce the historical context in which our study is situated and to familiarize the reader with major historical events and episodes in China. Second, we use more detailed life history information of a national sample of urban residents – the main data for the studies reported in this book – to describe historical changes in individual life chances in such areas as educational attainment, labor force composition in types of organizations and occupations, as well as across economic sectors. For anyone familiar with the Chinese context, it would be difﬁcult to discuss the People’s Republic of China without paying attention to, and indeed giving prominence to, the distinctive, evolving historical contexts in its brief but turbulent history. Therefore, we begin with a discussion of the historical context by ﬁrst describing – and emphasizing – the distinctive historical periods of the People’s Republic of China in the second half of the twentieth century. We conclude this chapter with discussions about general issues regarding data sources and research designs in our study, paving the way for the presentation of empirical evidence in the subsequent chapters.

changing historical contexts in urban china When the People’s Republic of China was founded in 1949, the ruling Chinese Communist Party and the new government were characteristic 35

36

The State and Life Chances in Urban China

of semi-military organizations evolved out of several decades of armed struggles with the KMT government it had just toppled, with no experience of national, political, and economic governance. By the time China embarked on its course of economic transformation in the early 1980s, the socialist state had established an elaborate and centralized system of political, economic, and social institutions, unprecedented in China’s history, that entrenched into every aspect of urban life. It is these pervasive redistributive institutions that have provided stable organizational links between the state and society and that carried out state policies in different arenas of political, economic, and social life. Since 1949, the Chinese political and economic structures have been largely based on the Soviet-style party-state. China witnessed an increasing expansion of the central authority and the growth of the redistributive institutions (Whyte and Parish 1984). This path, however, has by no means been a linear one. In the ﬁve decades of history, China also experienced frequent and large-scale political and economic changes, as evidenced by political campaigns and drastic shifts of state policies. For the purpose of understanding important variations in the political and economic environments, we distinguish three broadly deﬁned historical periods that characterize distinct changes in state policies and political processes: 1949–1965, 1965–1979, and 1980–1994. The 1949–1965 Period The ﬁrst period is between the inception of the People’s Republic of China in 1949 and the eve of the Cultural Revolution. Overall, this was a period of nation building, centralization, and economic development. State policies were largely geared toward industrialization and economic development, often characteristic of “rushed growth,” leading to an overall expansion of opportunities. But there were notable political ﬂuctuations and campaigns during this period (Whyte 1985). Institutionally, the central government underwent a state building process, in which it centralized the control of resources and its power through collectivization and nationalization, especially in urban China . In 1950, the government nationalized foreign-owned companies in China. Beginning in 1954, private ﬁrms were transformed to being ﬁrms jointly operated by the state and private owners. In the same period, the government also launched the collectivization of the craft industry. As a result, the sales by private retailers declined from 57.2 percent in 1952 to 25.6 percent in

Overview: Historical Context and Research Design

37

1954. By 1956, ninety-nine percent of the private industrial enterprises and 85 percent of the private commerce shops were jointly managed by the government and private owners (Research Ofﬁce 1991). Over time, these ﬁrms or shops evolved into collective ﬁrms under the administration of local governments. In the rural areas, starting in 1953, the government adopted the national planning system for the procurement and sale of agricultural produce, thus beginning to exert control over the allocation of agricultural production (Feng and Liu 1992). Similar to other state socialist societies (Kornai 1992), the Chinese state adopted “rushed growth” strategies in industrialization, culminating in the Great-Leap-Forward episode (1958–1960), when the government mobilized enormous resources for rapid economic growth. The urban labor force expanded greatly in response to demand from industrial growth; so did educational institutions and production facilities under the “rushed growth” strategy. The “rushed growth” policies resulted in the economic disaster of the early 1960s. Between 1961 and 1965, China experienced a period of economic contraction characterized by policies of “economic adjustment.” Between 1961 and 1963, the urban labor force was reduced by 18 million; the urban population was reduced by 26 million (Research Ofﬁce 1991, p. 261). The economic disaster also had a signiﬁcant impact on the centralization process in China. In response to earlier policy mistakes, the central government adopted new policies that encouraged local initiatives and allowed some extent of decentralization. For example, unlike in the USSR where state ﬁrms were directly controlled by ministries in the central government, local governments in China have retained certain authorities over the state ﬁrms within their jurisdiction. Moreover, there has been a signiﬁcant proportion of the urban labor force in the collective sector outside the direct control of the central planning apparatus. In the political arena, this period witnessed frequent political campaigns that altered the life chances of a large number of people. The most dramatic one was the “anti-rightist” campaign of 1957–1958, which purged over 550,000 intellectuals – about 10 percent of the total intellectuals at the time – who were rehabilitated only twenty years later. In 1960, the Communist Party “sent down” over one million cadres to the lower levels – about 15 percent of the total number of cadres at that time (Zhao 1987). Obviously, for ordinary workers, intellectuals as well as cadres, life chances were sensitive to shifting state policies and political campaigns.

38

The State and Life Chances in Urban China The 1966–1979 Period

This was the period of the Cultural Revolution and its immediate aftermath. During the Cultural Revolution, Mao Zedong and Maoist leaders adopted radical policies and used political campaigns to implement Mao’s vision of socialism and to overcome resistance from the bureaucracies (Tsou 1986; Whyte 1973). Large-scale political purges ensued. According to one estimate, 17 percent of the seventeen million cadres and 75 percent of high-level ofﬁcials were interrogated and purged. Including their family members who also bore the consequences of downward mobility, it is estimated that a hundred million people suffered from political persecution in this period (Chen and Liu 1991). Given the relatively small proportion of urban population in China during this period, the extent of political purges is striking! Political mobilizations and campaigns severely disrupted economic activities and labor force participation. Food, cooking oil, meat, cloth, and even soap were rationed, and there was little job mobility across immediate administrative jurisdictions (e.g., county). Although the whole nation suffered adverse economic consequences, children of the Cultural Revolution – those who entered the labor force during this period – experienced particular adverse effects of state policies. Because of the disruption of the urban economy, a large number of middle school and high school graduates could not be assigned jobs by the government agencies. As a result, between 1968 and 1979, the government adopted the “senddown” policies that forced over seventeen million middle and high school graduates to work in the rural areas. One immediate consequence of the chaos during this period was that labor force participation and other opportunities were highly politicized and severely affected by sharp shifts in state policies. In particular, the class label of family class background (jiating chushen), became a main political criterion for school enrollment, job allocation, promotion, and the distribution of economic resources among social groups. For example, between 1966 and 1970, admission into colleges was closed. When some colleges were reopened in 1970, students were recruited from those already in the labor force, and “political recommendation” replaced entrance examination as the criterion of admission. The Cultural Revolution period, then, was an extraordinary period of time in which the impacts of the state and state policies on life chances reached their height. This episode provides the opportunity for us to

Overview: Historical Context and Research Design

39

examine how social stratiﬁcation processes changed in response to state policies and how state policies were moderated by the structures and processes of social stratiﬁcation. The Cultural Revolution ofﬁcially ended in 1976, shortly after Mao Zedong died. However, most radical policies as well as the institutional arrangements remained in effect until major changes began to take place in the early 1980s. Therefore, for our research purpose we include the years between 1977 and 1979 in the second period. The 1980–1994 Period This was the post-Mao era or the reform era when radical policies adopted during the Cultural Revolution were abandoned and economic reforms were introduced, initially in the rural areas in the late 1970s and then in the urban areas starting in the early 1980s. In many ways, this period represents, by now we can conﬁdently say, a fundamental break from the state socialist redistributive system in the Mao era. These changes took place only gradually and are still ongoing. Although the directions of these changes are by no means unambiguous, the overall trends are clear: There has been a gradual decline of the centralization of authority, especially with regard to the state control over economic sectors, labor markets, and work organizations. This is evidenced by the gradual shrinking of the state sector, in both the proportion of industrial output and that of employees, compared with the nonstate sector. By 1994, the state sector employed 67 percent of the urban labor force, and the collective sector and “other sector” (including the private sector, foreign ﬁrms, and ﬁrms with hybrid property rights) employed 20 percent and 13 percent, respectively. In terms of industrial output, however, the proportions are 34, 41, and 25 percent for the state sector, the collective sector, and the nonstate sector, respectively (SSB 1995, pp. 84, 377). Multiple processes are underlying China’s economic transformation and institutional changes. One important process is the emergence of market economies and market mechanisms that led to the erosion of the redistributive state and changes in both mechanisms of resource allocation and distribution of opportunities (Nee 1989). But, unlike the former Soviet Union and the East European societies, the Chinese state has been actively involved in the transformation process (Naughton 1996; Shirk 1993). Local governments also play an important role in forging diverse paths of regional development (Oi 1989, 1999; Walder 1995b).

40

The State and Life Chances in Urban China

the evolution of redistributive institutions and life chances State socialism is sustained by a distinctive system of political, economic, and social institutions. Despite considerable variations across societal contexts, these institutional arrangements can be summarized in the “Soviet Model,” in which the party-state as the central political authority monopolizes and redistributes resources (Kornai 1992). All walks of life are organized into the web of state organizations: peasants into state farms or communes, workers into state enterprises, teachers and students into educational organizations, all under the control of the state apparatus. To sustain this elaborate control over society and implement state policies, there grew an enormous bureaucratic apparatus. To illustrate the evolution of redistributive institutions in China’s urban economy, we now turn to a series of empirical patterns in major areas of stratiﬁcation. The purpose is to impress the reader with the striking changes in life chances over historical periods and to invite the reader to take on the journey through this book to appreciate the impacts of state and state policies on individual life chances. To do so, we use both national statistics and the sample statistics from the life histories of a national sample of urban residents drawn from twenty cities to demonstrate varying opportunities across social groups and over time. The technical details of the sample are described in the latter part of this chapter.1 Labor Force Composition In urban China, there are distinctive types of work organizations with respect to their property right relationships to the state: (1) the state sector, which comprises government agencies, public organizations, and state ﬁrms; (2) the collective sector, which comprises those ﬁrms that do not belong to state central planning but often sponsored by a local (e.g., county or town) government; (3) the private (self-employed) sector; and (4) the “other” sector, including foreign ﬁrms, joint venture ﬁrms, and hybrid ﬁrms with mixed property rights, say, between state and collective, or between collective and private. 1

Because we retrospectively gathered the life histories of a sample of urban residents, the respondents enter our sample when they become available to experience a speciﬁc event (e.g., transition into different educational levels, transition to the labor force, etc.). As a result, it is useful to treat our data as a sample that evolves over time, reﬂecting changes in historical context and internal demographics (e.g., aging).

Overview: Historical Context and Research Design

41

100%

percent

80%

60%

40%

20%

0% 50

55

60

65

70

75

80

85

90

95

year state sector

collective sector

other sector

ﬁgure 2.1a. Composition of urban labor force across sectors in ofﬁcial statistics, 1952–1994.

This dimension of the labor force composition is of particular interest because it provides information on the types of institution-based links between the state and individual life chances in the work force. There are distinctive differences in terms of economic resources, welfare programs, as well as social status associated with types of work organizations (Bian 1994; Walder 1986, 1992). Changes in the composition along types of organization provide clues to changes in the state responsibilities for the employment and economic well-being of the labor force over time. Figures 2.1a and 2.1b display the distribution of employees across types of organizations in urban areas, based on national statistics and our sample, respectively.2 The proportions employed in state organizations (government agencies, public organizations, and state-owned ﬁrms), collective ﬁrms, and nonstate ﬁrms (private or hybrid ﬁrms) are useful measures of the extent 2

The national statistics are based on SSB (1990a, p. 3), and SSB (1995, pp. 84–85) information before 1952 is unavailable. State organizations include governmental agencies, public organizations, and state-owned ﬁrms. “Other sector” includes both hybrid ﬁrms and private ﬁrms.

The State and Life Chances in Urban China

42 100%

percent

80%

60%

40%

20%

0% 50

55

60

65

70

75

80

85

90

95

year state sector

collective sector

other sector

ﬁgure 2.1b. Composition of urban labor force across sectors in the sample, 1951–1991 (ﬁve-year average).

of the redistributive economy. According to Figure 2.1a, the proportion of the labor force working in state organizations rose dramatically in the period of nationalization and collectivization in the late 1950s, and there was an associated sharp decline in the proportion working in nonstate ﬁrms. After the peak of the late 1950s, employment in the state sector was steady with a slight decline in the 1980s. Since the mid-1950s, collective ﬁrms have had an important presence in urban China, employing between 20 and 25 percent of the urban labor force. Finally, since the early 1980s, the proportion working in nonstate organizations has increased noticeably. However, based on changes in the proportions in state and nonstate organizations, a market economy had emerged in urban China to only a limited extent by the mid-1990s. These patterns show the dominance of the state sector in China’s urban labor force since 1949. The historical trends also show signiﬁcant changes over time, especially the ﬂuctuations in the late 1950s and early 1960s during the period of collectivization and nationalization, and the noticeable rise of nonstate sectors since the 1980s. Because we have in our sample more detailed information on respondents’ location in types of work organizations, we now examine in some detail the trend of changes in the composition of the labor force as

Overview: Historical Context and Research Design

43

30

percent

20

10

0 50

55

60

65

70

75

80

85

90

95

year government

public

central gov't firm

local gov't firm

ﬁgure 2.2a. Composition of type of organizations in the state sector, 1951–1991 (ﬁve-year average).

reﬂected in our sample, using more reﬁned categories (see Figures 2.2a and 2.2b). The state sector comprises three types of organizations: government agencies, public organizations (e.g., nonproﬁt organizations such as entertainment, media, research institutions, hospitals, etc.), and state ﬁrms. Among state ﬁrms, we further distinguish those owned by the central government (or provincial government) and those owned by local (county) governments. Although both types are state ﬁrms, they have different welfare programs and latent beneﬁts. As Figure 2.2a shows, in the state sector, there has been a gradual decrease in the proportion of those working in government agencies and an increase in public organizations and state ﬁrms. However, the impacts of historical periods were evident, as indicated by the noticeable higher proportion of state ﬁrms in the late 1950s during the period of rushed growth, the relative low or steady state of the whole state sector during the Cultural Revolution, and the noticeable increase in state ﬁrms in the early 1980s as urban China was absorbing the return of millions of the sent-down youth. Figure 2.2b shows changes in the composition of labor forces in the nonstate sector, which clearly employed less labor force than the state sector did. The impacts of shifting state policy are most evident in changes in the collective and nonstate sectors. First, there was a sharp drop in

The State and Life Chances in Urban China

44 30

percent

20

10

0

50

55

60

65

70

75

80

85

90

95

year collective firm

private firm

join venture

hybrid firm

farm

ﬁgure 2.2b. Composition of type of organizations in the nonstate sector, 1951– 1991 (ﬁve-year average).

self-employed or employment in the private arena in the Mao era and it increased since the late-1970s. Second, the collective ﬁrm sector expanded since the Cultural Revolution to create employment opportunities for the urban population. The proportion of those on the farm partly reﬂect those in our sample who grew up and worked in the rural areas before they moved into the urban areas and fall within our sampling scheme. However, the noticeable, high proportion of those on the farm during the Cultural Revolution resulted from state sent-down policy between 1968 and 1979. Since the mid-1980s, a new type of organization – “joint venture” ﬁrms between foreign capital and Chinese ﬁrms – emerged in the reform era. Overall, during the period of our study, the redistributive economy is the main employer of the urban labor force. By the early 1990s, the end of our observation span, there were noticeable increases in employment in the nonstate sector, but proportion-wise, these changes were still small. Another dimension of the labor force is the composition of occupations. Figures 2.3a and 2.3b, respectively, show the changing compositions among occupations based on our sample. Among white-collar occupations, we have distinguished high-rank cadres and professionals, low-rank cadres and professionals, and ofﬁce workers (clerks). Among blue-collar workers, we distinguish service workers, skilled manual workers and unskilled manual workers, and self-employed individuals (private entrepreneurs).

Overview: Historical Context and Research Design

45

20

percent

15

10

5

0 50

55

60

65

70

75

80

85

90

95

year high-rank cadre low-rank prof

low-rank cadre clerk

high-rank prof

ﬁgure 2.3a. Composition of cadres, professionals, and ofﬁce workers, 1951– 1991 (ﬁve-year average).

Figure 2.3a shows that there was an overall increase in the proportion of white-collar jobs over time. The meanings of high rank and low rank will be deﬁned precisely in the next section. These occupations experienced little growth and even decreased during the Cultural Revolution. 30

percent

20

10

0 50

55

60

65

70

75

80

85

90

95

year service worker

skilled worker

unskilled worker

self-employeed

ﬁgure 2.3b. Composition of workers and self-employed, 1951–1991 (ﬁve-year average).

46

The State and Life Chances in Urban China

The ﬂuctuations among skilled and unskilled workers, as shown in Figure 2.3b, also captured nonlinear changes in the labor force. These patterns were most noticeable during the Cultural Revolution, as millions of the urban youth were sent down to the rural areas. In the reform era, the proportion for both unskilled and skilled workers declined, whereas service workers and self-employed had a higher presence in labor force composition. The trend shows noticeable structural changes over time, but the ﬂuctuating patterns cannot be satisfactorily explained by the process of industrialization and the gradual expansion of opportunities along with economic growth; rather, they reﬂect signiﬁcant impacts of political dynamics and shifting state policies. Composition of Industrial Output Across Sectors Another dimension to measure the impacts of the state and redistributive economy in China is to examine their roles in economic activities. Figure 2.4 shows the proportion of industrial outputs in China across types of sectors.3 These patterns show, in a stark manner, drastic changes in economic activities across these sectors over time. Similar to the patterns in labor force composition, the prominent role of the state sector reached its peak in the late 1950s, when the Chinese state attempted to push for rushed growth and failed miserably. Since then, the proportion of industrial outputs by the state sector declined over time, and the declining trend accelerated in the 1990s. In contrast, industrial output by the collective sector increased steadily over time.4 Contribution by private and hybrid ﬁrms has increased noticeably since the mid-1980s. By the early 1990s, the state sector contributed less than half of the total industrial output. These variations show that there have been considerable changes in the development of state redistributive institutions over time. On one hand, we observe the prominent role of the state in urban life, with the state sector employing the majority of the urban labor force in the past ﬁve decades. On the other hand, there have been considerable variations across historical periods. It seems that the centralization in the economic 3 4

The information is based on SSB (1990a), which was corrected and updated using more recent sources in ofﬁcial statistics. The information used here is drawn from national statistics that also include rural China. Collective ﬁrms here also include rural township enterprises.

Overview: Historical Context and Research Design

47

100%

percent

80%

60%

40%

20%

0%

50

55

60

65

70

75

80

85

90

95

year state sector

collective sector

other

ﬁgure 2.4. Composition of industrial output across sectors, 1952–1994.

arena reached its peak in the late 1950s and was followed by a steady state (in employment) or a gradual decline (in proportion of industrial output) thereafter.

Educational Opportunities We have learned from research on other parts of the world that the general trend of educational attainment across societies follows a pattern of gradual expansion of opportunities along with industrialization and economic development. That is, although education opportunities are under the control of the privileged classes, signiﬁcant upward mobility takes place over time because increasing educational opportunities are gradually extended to lower classes. The Chinese experience tells a different story – educational opportunities varied greatly and nonlinearly over time. Let us illustrate this point by displaying the enrollment rates at three educational levels over historical periods based on ofﬁcial statistics (SSB 1999; see Figure 2.5). First look at the enrollment rate into college. The extremely high rate in the early 1950s (over 100 percent) was, we suspect, due to the fact that the government sent a large number of on-the-job cadres to colleges for

The State and Life Chances in Urban China

48 250%

percent

200%

150%

100%

50%

0%

49

54

59

64

69

74

79

84

89

94

year entry to college

entry to high school

entry to middle school

ﬁgure 2.5. Enrollment rates at three educational levels, 1949–1994 (ofﬁcial statistics).

training. As a result, the number of those entering college during this period greatly exceeded the number who graduated from high school. The rate witnessed a second peak in the late 1950s amid the GreatLeap-Forward period of the rushed growth. However, during the entire Cultural Revolution (1966–1979) the enrollment rate to college was extremely low when radical state policies severely disrupted the educational sector. Enrollment rates to middle school and high school also followed highly ﬂuctuating, nonlinear trajectories. For example, there were noticeably lower enrollment rates at these two levels in the early 1960s when the economic disasters severely interrupted many individuals’ education. These changes show distinctive historical contexts over time with varying opportunities. The composition of educational levels evolved over time. Figure 2.6 displays the changing composition of education in our sample. As one can see, the urban population became more educated, as indicated by the decreasing percentage of respondents with elementary or no education and the increasing percentage of respondents with advanced educational levels (at or above middle school). Since the mid-1970s, technical schools

Overview: Historical Context and Research Design

49

60

50

percent

40

30

20

10

0 50

55

60

65

70

year

75

80

85

90

95

elementary

junior high

technical school

senior high

community college

college

ﬁgure 2.6. Distribution of educational levels in the sample, 1951–1991 (ﬁveyear average).

have become an important component of the educational system. “Informal college” education (dazhuan), similar to community college in the United States is a category for those who received some specialty training at college level through a less than four-year college education. For example, some cadres were sent back to special programs for retraining, or those already in the labor force attended night-school programs offered by colleges. However, it is noticeable that these increases do not follow a linear trend but ﬂuctuated over time. Behind these statistics and descriptions are the real lives of the Chinese citizens who lived through these events of political turbulence, and many of them saw their lives changed overnight. Consider a college student in the 1950s. The rare high-education opportunities of that time meant that he or she would be assigned to a desirable job in a governmental agency or a public organization, with job security and social status. However, the mere venture of criticizing the party leaders in the college may have led him to be labeled a rightist in the anti-rightist campaign. His life course and the life experiences of his family members would be changed forever from then on!

50

The State and Life Chances in Urban China

the life history data from “the state and life chances” project As we stated in Chapter 1, the main purpose of this book is to examine patterns of social stratiﬁcation processes and their consequences in state socialist China. To accomplish this goal, we adopt the analytical strategy of examining how individual life chances are realized over time and over individuals’ life courses in major areas of social stratiﬁcation. We ask several interrelated questions: How do state policies and historical context affect life chances through changes in opportunities and in allocative mechanisms over time? How do the early life events affect an individual’s subsequent status and economic well-being over his or her life course? In what ways does the link between the state and life chances vary across social groups and over time? These issues call for analyzing the evolution of the stratiﬁcation processes over time. Ideally, we would like to adopt a prospective research design that follows a representative sample of individuals and record their life experiences over time, for the entire observation period, from 1949 to the present. Alternatively, one can use repeated samples over time to gauge information about changes and continuity in individual life chances over time, and to explain these patterns by identifying the underlying social stratiﬁcation processes. Unfortunately, systematic, historical, individual-level data in the area of social stratiﬁcation and mobility in former state socialist societies were extremely scarce because of tight control of information by the government and because of the stalled social science research in these countries. Data collection for social science research became possible only after fundamental changes in these societies in the late 1980s. As a result, the only feasible strategy available to researchers is to adopt the method of retrospective data collection – by interviewing individuals and record retrospectively their life events over their life course so as to examine these causal processes using such retrospective data. This is the approach adopted in this study. The main data used for this study include the life history of 5,512 respondents from twenty cities selected from six provinces and two municipalities (Beijing and Shanghai) in China. The data were collected through a large-scale survey (4,071 respondents) conducted in 1994 and a pilot study (1,041 respondents) conducted in Beijing and Shanghai in the summer of 1993. These data were collected using a pretested questionnaire by trained interviewers through face-to-face interviews. The data include

Overview: Historical Context and Research Design

51

retrospective information on changes in respondents’ organizations, occupations, and geographic locations, as well as their education, family structure, political status over time, and other related information. In addition, if the respondent is married, we also collected the same information from the spouse of the respondent using a similar questionnaire. Because our respondents were between twenty-ﬁve and sixty-ﬁve years old at the time of interview, most of them were married. Therefore, a total of 4,562 spouses are in our data. Although respondents are drawn from a multistage sampling scheme, spouses are included based on their ties to the respondents. Therefore, it is inappropriate to treat the spouse sample as a random sample. In most analyses, we only focused on the sample of respondents. On some occasions where we needed to use the spouse data to increase statistical power and where we ascertained that assortive marriage is not an important statistical concern, we also included spouses in some of the analyses. Sampling Six provinces were selected for data collection, each representing a conventional geographic region in China. The selection of these provinces was made using purposive sampling based on our knowledge of the political, economic, and social environments across regions in China, and in consultation with our Chinese collaborators with regard to the research facilities and manpower in these provinces to carry out the interviews. In each province, we chose the capital city to represent large cities (population over one million). We randomly selected a medium-sized city (population between two hundred thousand and one million) and a small city (population under two hundred thousand) based on the 1990 Yearbook of Chinese Cities (SSB 1990b). The sample size in each city was proportional to the population in that size of city in that province.5 We also included Beijing, China’s capital, and Shanghai, the largest industrial city in China, in our sample. These twenty cities cover a variety of geographic locations and types of urban economies. We selected a stratiﬁed random sample of each city’s residents. In China, a metropolitan city is composed of residential districts. Each district is composed of residential “streets,” which in turn is organized into residential blocks (juweihui). We selected residential blocks in each city 5

There were minor adjustments in the sample quota for each province in the data collection process.

52

The State and Life Chances in Urban China

using a systematic sampling scheme. That is, we selected every nth residential block based on ofﬁcial residential statistics. We used an analogous sampling scheme to select households in each residential block. Finally, we randomly selected a member of the household between the ages of twenty-ﬁve and sixty-ﬁve to be interviewed. Respondents included both current labor force participants and retirees. Table 2.1 reports the distribution of sample sizes in the selected provinces and cities. A noticeable feature of this data set is that large cities are overrepresented in the sample: over 60% of our sample is drawn from cities that have a population of more than one million. This is partly because the pilot sample in 1993 collected data only in Beijing and Shanghai, both large cities. Therefore, the empirical ﬁndings in this book are more likely to reﬂect patterns in large, metropolitan areas rather than in smaller cities. For our research purposes, the disproportional sample of large cities also has advantages: Large cities are more representative of the state redistributive system and they are more sensitive to state policy shifts, facilitating our exploration of the related research issues. Moreover, large cities are more prone to changes in the reform era, allowing us to detect changes in the economic transformation in urban China. It is also worth noting that, despite noticeable disparities across urban areas, urban China was tightly controlled by the state bureaucratic organizations. In this sense, rhythms of changes and patterns of stratiﬁcation are likely to be similar across different regions of the urban areas. The data yield a sample of urban residents whose composition varies over time. An individual enters our data when she or he begins formal education (or labor force if illiterate) and leaves when she or he leaves the labor force (e.g., retires). A potential problem with our sampling scheme is that residents of the twenty cities who died or migrated before the interviews in 1993–1994 were not in our sampling frame, but we have no reason to think that this fact would cause systematic biases in our ﬁndings.

research design considerations Subsequent chapters in this book report empirical patterns of life chances in major areas of social stratiﬁcation in urban China, from 1949 to 1994. The speciﬁc issues being examined – educational attainment, job mobility, promotion patterns, or income distribution – dictate distinctive research designs and analytical strategies, which will be attended to when

Overview: Historical Context and Research Design

53

table 2.1. Distribution of the Sample Size Across Provinces and Cities Respondent

Spouse

Province/City

N

%

N

%

Beijing Shanghai Hebei Shijiazhuang Congzhou Hengshui Heilongjiang Haerbin Mudanjiang Anda Jiangsu Nanjing Suzhou Huaian Guangdong Guangzhou Chaozhou Huizhou Sichuan Chengdu Neijiang Guanghan Gansu Lanzhou Tianshui Jiayuguan total

1,000 975 467 103 264 100 516 217 190 109 529 195 188 146 642 277 256 109 510 150 185 175 473 205 105 163 5,112

19.6 19.1 9.1

930 899 455 100 256 99 465 193 171 101 453 169 157 127 498 188 209 101 451 136 158 157 411 172 90 149 4,562

20.4 19.5 10.0

10.1

10.3

12.6

10.0

9.3

100.0

10.2

10.0

11.0

9.9

9.0

100.0

Note: The samples include both the 1994 survey data and the 1993 data collected in a pilot study.

we consider these speciﬁc investigations. In this section, we outline general considerations in research design for the studies reported in this book. Focus of Analysis: Different Aspects of the Stratiﬁcation System In the subsequent chapters, we will examine several aspects of the stratiﬁcation structure: (1) educational stratiﬁcation, (2) patterns of job mobility with respect to both entry into ﬁrst job locations and subsequent job changes, (3) promotion patterns in the workplaces, (4) the distribution of

54

The State and Life Chances in Urban China

economic beneﬁts (e.g., personal income and housing), and (5) the interactions between stratiﬁcation dynamics and life-course factors. There are several advantages in examining these areas in relation to one another. First, these areas are connected sequentially over one’s life course. For instance, paths of educational attainment channel individuals into different patterns of job attainment, which lead to different promotion opportunities and economic beneﬁts. These aspects are likely to be interrelated to a great extent. More important, the processes that generate these patterns are interconnected, due to macro-political environments and redistributive processes governed by a centralized state. Therefore, the subsequent chapters that examine speciﬁc areas are an integral part of a larger, concerted effort to assess social stratiﬁcation patterns in urban China. Another important aspect of the research design is to take into consideration multiple stratiﬁcation processes. In market societies, social origins – parental socioeconomic status and resources – are often the most important channels of social stratiﬁcation. One’s own achieved status, education, or previous work experience, may also contribute to social mobility in later stages of the life course. These micro-level relationships may be structured and reshaped by macro-political processes in the Chinese context. The relationships between social origins and status attainment, for instance, may depend on state policies that deﬁne the nature and the strength of such links. The role of education in one’s later achievement may also depend on how it was valued by state policies. This recognition calls attention to multiple stratiﬁcation processes underlying the observed patterns. Main Covariates of Theoretical Interest We now describe the main covariates of theoretical interest used in the analyses. Additional covariates are devised for speciﬁc purposes in some analyses, and will be introduced in the context of these speciﬁc analyses. Human Capital and Political Capital Following the convention, we use two measures of human capital: formal education and work experience (on-the-job training). In most cases, the former is measured by a set dummy variables indicating the respondent’s level of education: elementary school or below (including no education, i.e., illiterate), middle school, high school, and college (or above). Because of their nature and the smaller number of cases, we combine “technical

Overview: Historical Context and Research Design

55

school” (zhongzhuan) into “high school” and “college granting associate degree” (dazhuan) into “college.” Work experience is measured by the respondent’s age. In most cases, we consider both the ﬁrst- and second-order effects of age. On a few occasions, when the data warrant, we calculate an approximate measure of labor force duration by using the formula as follows: Labor force duration = age − years of formal education −6. To measure political capital, we use Communist Party membership as an indicator of one’s political status in some analyses.6 Occupational Status Occupational status has been the conventional measure of one’s political, economic, and social status in studies of social stratiﬁcation. To examine the effects of one’s occupational status on stratiﬁcation outcomes, we will use the following broad occupational categories: Cadres include managers in enterprises as well as bureaucrats in party and government agencies. In contrast to the standard international occupational classiﬁcation system, in state socialist China, it is not always theoretically meaningful and empirically feasible to distinguish the category of managers (in ﬁrms) and that of bureaucrats (in government ofﬁces), because both belong to the same bureaucratic rank system (see Chapter 6). In many analyses, we distinguish high-rank cadres and low-rank cadres. The Chinese bureaucratic hierarchy has four main levels: bu (ministry), ju (bureau), chu (division), and ke (section). These ranking positions are ofﬁcially standardized across organizations to a large extent, especially in the state sector. We classify those holding ranks at or above the chu level as high-rank cadres and those at or below the ke level as low-rank cadres. According to ofﬁcial bureaucratic ranks, a chu-rank is equivalent to the head of a county, although substantive authorities vary greatly depending on the speciﬁc situations. In large government agencies, for example, a chu-rank employee is likely to be only a mid-level ofﬁcial; but in a small-size enterprise, the head of the enterprise may be a ke-rank cadre. 6

Because party membership information is missing for two provinces in our data (about 20 percent of our sample), using the “party membership” variable in the statistical analysis leads to the omission of this subset of the data. Because “political capital” is more appropriately measured by one’s cadre position that is associated with positional power, and because of the importance of ensuring national representativeness of the sample, we used the variable “party membership” only in some selected analyses when theoretical issues warrant special attention to this variable.

56

The State and Life Chances in Urban China

The low-rank cadres include mostly those who do not have formal administrative positions, but belong to the administrative system and at the bottom of the bureaucratic hierarchy. For example, most ofﬁce workers in government agencies are unranked cadres, so are mid-level managers in a small state ﬁrm. Professionals are occupations that involve specialized training and skills, such as physicians, engineers, university professors, teachers, and researchers. In some analyses, we distinguish high-rank professional and low-rank professionals. In the Chinese professional system there are senior engineer, engineer, assistant engineer, and technician levels (or equivalent levels in other professional occupations). We classify those at or above engineer level as high-rank professionals and those at or below assistant engineer level as low-rank professionals. The high-rank professionals are typically specialists in speciﬁc ﬁelds, such as doctors, journalists, actors, editors, and university professors. The category of low-rank professional is equivalent to technician or semiprofessional groups in industrialized societies, including nurses, technicians, teaching assistants in universities, and so forth. (See Chapter 6 for detailed information on the Chinese cadre and professional hierarchies.) Ofﬁce workers (clerks), in some sense, is a group similar to “clerks” in the industrialized societies in that they hold “white-collar” jobs, such as typists, communication personnel, and bank clerks. In China, individuals in this group do not formally belong to the cadre administrative system, but they often have an important status because of their closeness to the ofﬁces of redistributive authorities. Service workers include those in wholesale or retail trade, and other service-related jobs. Production workers include all types of workers, regardless of their location in the organizational hierarchy. In some analyses, we combine “service workers” and “production workers” into a single category; in some more reﬁned analyses, we distinguish skilled and unskilled manual workers among “production workers.” Military personnel, in some analyses, are included because, in the Mao era, military service was an important path for urban youth to acquire political capital and a good civilian job later when they left the military service. Farmers. Although our sample was drawn from urban areas, throughout a respondent’s life course, he or she may have been sent to work in the rural areas and become a farmer. Or a respondent of rural origin may

Overview: Historical Context and Research Design

57

have migrated into the urban areas later in his or her life course. Thus, we include the farmers category in our sample frame. Types of Work Organizations We now describe types of work organizations in China’s urban economy and of economic sectors. The organizational types are those ofﬁcially recognized in governmental classiﬁcations of work places and are frequently mentioned in scholarly studies. organizations in the state sector. Government agencies (zhengfu bumen) include ministries, commissions, bureaus, and various ofﬁces in the Communist Party and state bureaucracies. Both the Communist Party and the administrative apparatus (which are interwoven at every level) fall in this category. Public organizations (shiye danwei) are nonproﬁt organizations in the public domain. They include educational and research institutions, and organizations in the medical, publishing, broadcasting, and entertainment sectors. Although not administrative organs of the state, most of these organizations are closely linked with the central government or with local bureaucracies.7 State-owned ﬁrms (guoyou qiye) are enterprises in production areas (e.g., manufacturing, processing, etc.) or services that are owned by central or municipal governments. Like public organizations, they are governed by state administrators and state policies, but they also enjoy extensive beneﬁts in housing, health care, and welfare subsidies. In some analyses, we distinguish the ﬁrms owned by central government (including provincial government) and those owned by local government (city or districts within cities). organizations in semistate and nonstate sectors. Collective ﬁrms (jiti qiye) are ﬁrms not directly administered or ﬁnanced by central planners, but they are often sponsored by a local government (e.g., a district or county government). They receive fewer redistributive beneﬁts than state-owned ﬁrms. Private ﬁrms include ﬁrms owned solely by individuals or partners in private organizations. Mostly comprised of family-based or individual businesses, they do not receive redistributive beneﬁts. They are the least 7

A 1993 publication of the Ofﬁce of the Central Government Structure Committee estimated that about 63 percent of employees in public organization are totally ﬁnanced by the government budget, 21 percent are partially ﬁnanced by the government budget, and only 11 percent are self-ﬁnanced (Ofﬁce of Organization Committee 1993, p. 500).

58

The State and Life Chances in Urban China

protected and also the least constrained by state institutions and state policies with regard to production, resource allocation, and job security. Hybrid ﬁrms are ﬁrms with mixed property rights, such as those jointly owned by the central state and private owners, or by collective and private owners. Such hybrid forms existed as a legacy of the collectivization processes since the 1950s, and they gained momentum in China’s economic transformation since the 1980s. Joint venture ﬁrms (sanzi qiye) belong to a new type of ﬁrm that emerged in the reform era. This category includes ﬁrms consisting of partnerships between Chinese ﬁrms and foreign ﬁrms, ﬁrms with both Chinese and foreign investments, and wholly foreign-owned ﬁrms. Because the proportion of employees in this category is small, we combine this category with hybrid ﬁrms in our analyses on the basis that both represent emerging organizational forms and are gaining importance in the reform era. Farm (village) refers to rural areas in China. Before the late 1970s, rural China was organized into Communes and production brigade (villages). Although this study focuses on urban China and selected respondents from urban residents, some went to the rural area as their job destinations before and later returned to the urban areas; others with rural origins entered the urban labor force in their job attainment. Both types are represented in the sample. In some analyses, we distinguish the following economic sectors on the basis of the service or product provided and the relation to market activities and the redistributive economy.8 Among the economic sectors, we distinguish heavy industry, which includes the mining, steel, machinery, and power industries; light industry, consisting of industries producing textiles and other consumer goods, such as food products; and service industry, which includes wholesale and retail trade, food service, and the transportation and communication industries. Parental Status/Social Origins refer to family-based social status. In some analyses, we examine intergenerational status inheritance by including measures of social origins. We use the father’s education and occupational status as indicators of social origins based on the fact that family status in China has historically been associated with the father’s

8

Organization types and economic sectors are determined on the basis of two different questions in the questionnaire. Economic sector is based on the product of, or service provided by, the workplace. Type of organization is based on the workplace’s property rights ownership.

Overview: Historical Context and Research Design

59

status. We use a set of dummy variables to indicate the father’s educational level: elementary school or no education, middle school, high school, and college. In some analyses, we may combine some of these categories for speciﬁc research purpose.9 We use two measures of the father’s sociopolitical status. Because these two measures overlap to some extent, in our analyses we use only one of these two alternative measures at a time. One measure is based on the father’s occupation. We collected information on the father’s occupation in selected years (year of entry into ﬁrst job, 1965, 1975, 1978, 1986, and 1993). We updated the father’s speciﬁc occupation over time as a time-varying covariate. Although previous studies of socialist societies have measured parental occupation using categories similar to those used in advanced Western societies (Gerber and Hout 1995; S. Szel´enyi and Aschaffenburg 1993), we differentiate between cadres and professionals on the basis of the political logic of stratiﬁcation under state socialism. Because of the tension between the state and the intellectuals in China, fathers who were cadres (managers, bureaucrats) were in a privileged position to facilitate their children’s life chances, compared with the fathers who were professionals. As we have elaborated the status of cadres and professionals earlier, we classiﬁed those at (or above) the chu level as high-rank cadres and those at (or below) the ke level as low-rank cadres. We classify those at (or above) the engineer level as high-rank professionals and those at (or below) the assistant engineer level as low-rank professionals. In addition, we also code the father’s occupation in the following categories: workers and farmers. We may combine some of these categories for the purposes of speciﬁc analyses. To further examine the political selection mechanisms, in some analyses, we also use an alternative measure of social origins based on the ofﬁcial label of family class background (jiating chushen). This class label refers to a person’s class location based on his or her parent’s or grandparent’s class before 1949. This ofﬁcially labeled family class background may not coincide with father’s current occupation and was directly deﬁned by state policies as a critical basis of political discrimination in the Mao era. Before its abolition in 1979, individuals were routinely required to report this label on applications for educational enrollment, jobs, and promotions. We use the following dummy indicators of family 9

For the father’s education, it is desirable to distinguish high school and college credentials. However, the number of cases with fathers having a college degree in the earlier periods is too small to allow meaningful statistical analyses.

60

The State and Life Chances in Urban China

class background: (1) workers/poor peasants, (2) cadre background (include military), (3) middle class (small business owners, middle-rich peasants), (4) intellectuals, and (5) the “exploiting class” (large business owners, landlords, “rightists,” “bad elements,” etc.). In our sample, the respective distributions for these categories are: 69.5, 9, 14.4, 2, and 2.6 percent. Because of the small numbers, we combine intellectual background with middle class in the analyses.

statistical analysis: models and methods Given our theoretical interest in historical patterns of redistribution and stratiﬁcation dynamics over time, most of our analyses involve modeling dynamic processes based on longitudinal data. In the following, we outline the main issues related to statistical modeling and analyses in this research context. The Concept of the “Risk Set” To examine the effects of state policies on individual life chances, an important analytical issue is to establish a common ground on which the life chances of individuals and social groups can be evaluated and statistical relationships estimated. We adopt the concept of the “risk set” – a statistical concept originated in event history analysis (Tuma and Hannan 1984) to address this issue. A risk set refers to those individuals who are available for experiencing certain events. For our analyses, the main consideration is whether individuals are exposed to the same state policies or life events in a similar way. For instance, the close-down of colleges during the Cultural Revolution had severe impacts on those who graduated from high school in those years, but this event had no effects on those who were already in the labor force, as far as educational achievement is concerned. As another example, the event of recruitment into the Communist Party is meaningful only for those who are eighteen years old or above because of the age restrictions in the Communist Party Charter. Therefore, for studying the event of joining the communist party, the risk set should include only those who are eighteen years old or above. Careful construction of the risk set for speciﬁc analysis is an important design consideration for conducting meaningful statistical analysis. For this reason, different groups of respondents are included in different analyses, depending on research purposes.

Overview: Historical Context and Research Design

61

Periodization To examine the evolution of the state socialist redistributive system in urban China and to assess changes in state policies and social institutions over time, we rely on examining changes across historical periods associated with distinctive state policies. Because state policies are often interrelated and there are considerable delays in implementations across regions and over time, in most cases it is impractical to directly measure the impacts of speciﬁc policies. In this study, we adopt a strategy of using historical periods to gauge the impacts of state policies on individual life chances. That is, we ﬁrst identify those historical periods that are associated with distinctive state policies and then examine how associations among observable relationships, such as those between social origins and educational attainment or those between type of work organizations and economic beneﬁts, vary across these periods in accordance with changes in state policies. In most analyses, we use a three-period scheme – 1949–1965, 1966– 1979, and 1980–1994 – for reasons discussed in the preceding sections. Needless to say, any attempt at periodization runs the risk of oversimpliﬁcation. More reﬁned periodization would allow better measurements of more speciﬁc changes in state policies or other contextual dimensions. However, given the sample size, partitioning of samples into more reﬁned categories is likely to cause the loss of statistical power and make the statistical analyses unfeasible. Thus, the three-period scheme is a decision that captures the main shifts in state policies and at the same time allows reasonable sample size for meaningful statistical analyses. In a few cases, we employ different periodization schemes for particular analytical purposes. For instance, in our analysis of educational attainment (Chapter 3), 1977 is a landmark for the beginning of a new period because this was the year when the college entrance examination was resumed after the Cultural Revolution. These decisions are described in the chapters when these speciﬁc studies are introduced.

Models Most of the analytical issues concern changes in one’s status such as movement across educational levels, across work organizations, or across hierarchical levels. These changes are characteristically qualitative outcomes. Moreover, to model changes in status, one needs to take into consideration

62

The State and Life Chances in Urban China

several features of the data and the underlying processes. First, the composition of the sample evolves over time. For instance, because of stages in the life course, only those within a certain age range are likely to enroll into certain educational levels and hence should be included in the analysis of the entry into these educational levels. Second, the values of the covariates may change over time. For instance, one’s occupational status in the labor force may change over time, thus affecting the probability of experiencing certain events (e.g., becoming a cadre). These considerations call for a dynamic modeling of a continuous-time, discrete-event process (Tuma and Hannan 1984). An important advantage of event history analysis is that information on previous life history and on time-varying covariates can be incorporated in analyses of the time-dependent hazard rate of the occurrence of various life events. There is a variety of event history models designated for different patterns of hazard rates, or different parametric or semiparametric assumptions. In this study, we primarily adopt a logistic model for discrete event history (Allison 1982, 1995). Our decision is based on the following considerations: First, our experience working with a variety of event history models is that, in most cases, these models yield similar results. Second, for those parametric models, the parametric assumptions made in these models are difﬁcult to assess. Third, although the logistic model does not specify any form of the hazard rate, variations in hazard rates are captured by time-varying covariates and time trend can be explicitly modeled through the incorporation of trend covariates. Moreover, the discreteevent history model allows great ﬂexibility in incorporating time-varying covariates and comparison across historical periods. Finally, most covariates in our data are measured on an annual basis, so a discrete-time event history model is appropriate. The logistic regression model for event history analysis can be expressed as follows:

Pit log 1 − Pit

= αt + β1 χit1 + · · · + βkχitk,

where Pit is the conditional probability that individual i has an event (e.g., promotion or job change) at time t. This probability can be called a discrete-time hazard. In most cases, historical context is taken into consideration by analyzing these transitions in separate periods. In logistic regression models, the estimated coefﬁcient is the log-odds, or “logit,” of experiencing an event to not experiencing such an event. The log-odds of

Overview: Historical Context and Research Design

63

an outcome or event are assumed to be a linear function of the covariates, with the regression coefﬁcients giving the effects of covariates on the log odds of the outcome. An advantage of the logistic model is that the estimated effects of covariates are not affected by changes in the marginal distribution of the variable. In other words, an estimated coefﬁcient in the model captures the association between the outcome and the corresponding covariate, net of marginal changes that reﬂect evolving opportunities. The parameter estimate for log-odds can be transformed into “odds,” or “probability” of events. In the context of the dynamic-event history model, the “logodds” and “odds” are sometimes interpreted as “log-rate” and “rate” of an event occurring, respectively. The relative magnitudes and direction of changes of the parameter estimates remain the same under these transformations (but not the magnitudes). Therefore, we may choose to interpret the coefﬁcients along one of these dimensions depending on their intuitive appeal. The logarithm transformation has some unique properties: The logarithm of the rate (logit) has the same trend as the rate. That is, an increasing log-rate implies an increasing rate, and vice versa. When logit is zero, it means that there is no difference between the group (e.g., women) in question and the reference group (e.g., men); when logit is greater than zero, the log-rate (as well as the rate) is increasing; when logit is less than zero, the log-rate is decreasing. For those events of status changes that may be repeated, one needs to take into consideration the complication of repeated events. For example, in our study of promotion patterns (Chapter 6), a cadre (manager) may experience more than one promotion in his or her career, and these repeated events are likely to be correlated over time. For such analyses, we adopt the generalized estimation equation (GEE) approach to model repeated events. The GEE approach has the advantage of not requiring parametric assumptions about the form of the covariance structures among multiple promotion events (Diggle, Liang, and Zeger 1994). The β is estimated by solving the estimation equation: U(β) =

∂µi ∂β

[Vt (α)]−1 (yi − µi ) = 0 p ,

where µi is the expectation of yi (e.g., the promotion event), which is linked to a linear combination of the covariates and the corresponding estimates through the logit function. The estimated variance is robust for repeated measures.

64

The State and Life Chances in Urban China

We also adopted other models to address speciﬁc analytical issues. For example, the examination of income distribution requires panel data analysis of continuous outcomes. We will discuss these models in those chapters where they are used for analysis. Missing Values Because our survey collected information over one’s life course, it is inevitable that some information is missing, especially with regard to information on one’s parents over time. The decision rule for treating missing values in the statistical analyses is based on “pairwise” deletion. That is, cases with missing values in any of the covariates in a speciﬁc statistical analysis are deleted from that analysis. In some instances when the missing values in particular variables are of large proportion and these variables are of theoretical importance, it is important to keep these variables but control for the effects of missing values. In such cases, we adopted a strategy to include these variables in the model while using indicator variables to control for the effects of those cases with missing information.

chapter summary Like in the former USSR, the redistributive institutions in China controlled the main channels of resource allocation and redistribution in the urban areas. The central government put urban residents into the web of state-owned or semistate (collective) work organizations. From wage grades and educational opportunities to welfare, health care, and family planning, the state exerted a strong grip on the society (Whyte and Parish 1984). In the rural areas, production and sales were also largely controlled by central planning. Along with the centralization processes, there grew a bureaucratic apparatus, which allocated resources across regions, economic sectors, and social groups in accordance with state policies. Redistributive institutions include ministries in the central government that governed speciﬁc industries across regions, local governments at different levels of the geographically based administrative system (provinces, counties, towns), as well as work organizations that were either owned by governmental agencies (state ﬁrms) or sponsored by local governments (collective ﬁrms). The dependency relationship of workers on the bureaucratic apparatus has become a salient feature of the redistributive system in urban China (Bian 1994; Walder 1986).

Overview: Historical Context and Research Design

65

On the other hand, there are several noticeable historical contingencies in China. First, unlike in the USSR, there exist signiﬁcant nonstate and collective sectors in China. The state control in the Chinese agricultural sector has never reached the same extent as in its Soviet counterpart. In the urban areas, there has been a signiﬁcant proportion (about 25 percent) of the labor force working in collective ﬁrms. Although collective ﬁrms are subject to state regulations and administrative intervention, they have never been an integral part of the state budgetary or administrative system, and thus have been less controlled by the central government in terms of their production and employment decisions. Second, the Chinese state was more thorough than most state socialist societies in its control of individual mobility, which greatly strengthened political control and the role of work organizations in China. Third, unlike the USSR, the Chinese state had pursued an egalitarian policy of “destratiﬁcation” before and during the Cultural Revolution (Parish 1984). As a result, there were noticeable differences in the distribution of economic resources between China and the USSR or Eastern European countries (see Zhou and Suhomlinova 2000). From a semimilitary Party organization to a full-ﬂedged bureaucratic governance structure, from land reform experience to an elaborate system of redistributive institutions, the Chinese state and its political and economic institutions have evolved tremendously over the last ﬁve decades. This chapter traced the evolution of the state socialist redistributive institutions, in particular political twists and turns accompanying shifts in state policies in the history of the People’s Republic of China. In this process, we encounter a series of questions: How were the life chances of the Chinese citizens affected by the political turbulence? What were the social stratiﬁcation processes that allocate and relocate individuals into different social positions over their life course in this turbulent political environment? How were the political dynamics mediated by one’s location in the social stratiﬁcation structure? The subsequent chapters are our efforts to address these issues.

part one REDISTRIBUTION AND STRATIFICATION DYNAMICS: EMPIRICAL EVIDENCE

I

n this part, we focus on the ﬁrst theme – redistribution and stratiﬁcation dynamics in the 45-year history of the People’s Republic of China. We present a series of empirical studies to document, illustrate, and assess the impacts of redistributive institutions of state socialism and shifting state policies on individual life chances over time. The order of these chapters follows more or less the sequential stages of one’s life course. We begin with patterns of educational attainment, especially focusing on variations in the association between social origins and educational attainment across historical periods (Chapter 3). We then examine the next major life event in one’s life course – patterns of entry into the labor force (Chapter 4). Again, our focus is on how the association between social origins and ﬁrst-job attainment varied over time, reﬂecting the interplay between redistribution and shifting state policies. In Chapter 5, we take a closer look at a special cohort – the children of the Cultural Revolution – and examine the dramatic impacts of state policies on their entry into the labor force and how this event affected their subsequent life experiences. Finally, we present a systematic examination of bureaucratic career patterns in work organizations, including patterns of recruitment into the Communist Party, entry into and promotion within the Chinese bureaucracy (Chapter 6). These studies shed light on the stratiﬁcation dynamics, and the critical and time-varying impacts of state policies on all social groups, through changes in both opportunity structures and allocative mechanisms.

67

3 Educational Stratiﬁcation

Educational achievement leads one into ofﬁcialdom. Confucius The more educated, the more “antirevolutionary” they became. Slogan in the Cultural Revolution

To investigate the link between the state and life chances, we begin our journey in the area of educational attainment – patterns of progression along formal education levels over time and among social groups. Educational attainment is an appropriate starting point in many respects. As the Confucius teaching above indicates, throughout Chinese history, education and knowledge earned respect, conferred status, and led to desirable careers in ofﬁcialdom. As many vivid folklores and solemn classics literature recorded, since the Sui dynasty (587 a.d.), the educational examination system in China has provided the most effective and direct channel of upward mobility, especially for individuals of lower social strata. This sentiment is still echoed in today’s China, as the children of peasants in remote areas see entrance into college through the national college entrance examination as the only feasible way to permanently break away from their rural roots. China was not alone in this regard. In contemporary societies, education has become the most important channel for upward mobility and economic well-being. Across societies all over the world, education helps individuals open doors to prestigious jobs, gain higher salaries, and have

69

70

The State and Life Chances in Urban China

better promotion chances in organizations. Because of the critical role of education in social mobility, patterns of educational attainment shed light on the operation of the social stratiﬁcation system – how educational opportunities are distributed and what the stratiﬁcation mechanisms are that generate the observed patterns of educational attainment. In particular, because formal education typically starts in the early stage of one’s life course and is heavily inﬂuenced by family-based social, economic, and cultural resources, patterns of educational stratiﬁcation make transparent the association between social origins and educational achievement in intergenerational status inheritance. Educational attainment across generations is also indicative of the relative openness of both societies and social stratiﬁcation systems. But, in the political history of the People’s Republic of China, education was at times a political liability. On the one hand, the socialist state recognized the importance of education and knowledge in economic development and in rational economic planning. Like other state socialist societies, the Chinese state has consciously used educational institutions in “social engineering” to allocate educational opportunities among social groups. On the other hand, the top leaders, especially Mao Zedong, had a deep suspicion and political mistrust toward intellectuals, with good reason: most political “disturbances,” from the “Hundred Flower” episode of 1957 to the student protests in 1989, were initiated by intellectuals or on college campuses. It is no small wonder that major political campaigns in China especially targeted intellectuals, and took place in those areas related to education. In the Mao era, educational credentials were often synonymous with being politically suspect, which may have had a negative effect on one’s career development. In the post-Mao era, education has once again become the most important institutionalized channel for upward mobility, especially for the children of peasants and workers. Fluctuations in educational attainment and the role of education in social stratiﬁcation in China mirrored its turbulent political history and captured variations in life chances for individuals who entered the educational system. This chapter examines patterns of educational stratiﬁcation in this context. We begin with a set of issues that this chapter intends to address. In the subsequent empirical analyses, we ﬁrst examine the distribution of opportunities over time and then focus on the analytical issues of assessing the association between social origins and educational attainment, and especially how this association varied across historical periods in response to shifts in state policies.

Educational Stratiﬁcation

71

education and social stratiﬁcation: the issues That there is a strong association between social origins and educational attainment in contemporary societies is a well-established ﬁnding in the social science literature (Psachargopoulos 1994). Here, social origins refer to family backgrounds and family resources as indicated by parents’ education, occupational status, or income. In the sociological research in the United States, it is Blau and Duncan’s (1967) status attainment model that ﬁrmly established the analytical focus of education as a critical mechanism in intergenerational status transfer. Since then, a central issue in stratiﬁcation research has been the assessment of the strength of association between social origins and educational attainment, that is, the role of parents’ economic, social, and cultural capital in advancing children’s educational attainment and upward mobility. Why do we observe marked stability in educational stratiﬁcation over time and across societies? A major explanation developed by sociologists is that patterns of educational stratiﬁcation reﬂect key processes of class reproduction. Historically, in the transformation from the feudal society to the capitalist market society, mechanisms of intergenerational status inheritance were fundamentally altered. In the former, the propertied class could transfer their family status from one generation to the next through direct inheritance of family titles and properties. In the capitalist market society, as Max Weber (1946, p. 182) pointed out: “the kind of chance in the market is the decisive moment which presents a common condition for the individual’s fate. ‘Class situation’ is, in this sense, ultimately ‘market situation.’ ” The rise of the middle class – managers, professionals, and other white-collar workers – posed a serious challenge: Without the tangible estates or recognizable status titles, how can the middle class reproduce itself through the transmission of their status across generations? According to Weber, it is through the acquisition of “marketable skills” that the middle class can reproduce itself and education plays a vital role in this process. Parkin (1979, pp. 63, 55) was most explicit on this point: In systems based on aristocratic, caste, or racial exclusion, families of the dominant group can expect to pass on their privileged status to their own descendants as a direct result of the closure rules in operation, however socially lethargic those families might be. The bourgeois family, by contrast, cannot rest comfortably on the assumption of automatic class succession; its must make deﬁnite social exertions of its own or face the very real prospect of generational decline . . . Formal qualiﬁcations and certiﬁcates would appear to be a handy device for ensuring that those who possess “cultural capital” are given the best opportunity to transmit the beneﬁts of professional status to their own children.

72

The State and Life Chances in Urban China

Scholars differ in their identiﬁcation of and emphasis on mechanisms of intergenerational status inheritance. One tradition, mostly represented by the status attainment model, emphasizes the role of individual-based family resources – a conception of atomized customers optimizing their utility (reproduction) through investment in their children’s education in the marketplace. In this light, class reproduction is an aggregate outcome of “spontaneous” individual or family decision making in the marketplace. Family social origins (in particular, parental socioeconomic status) are a good indicator of individuals’ social and cultural capital in market societies: Parents who can afford to send their children to expensive private schools can thereby foster their children’s accumulation of social and cultural capital, which in turn promotes their educational and career advancement. At the macro level, two models of educational stratiﬁcation are developed in sociological research. One image is based on the “industrialization” thesis, which posits a general trend of gradual increasing of educational opportunities along with industrialization (Treiman 1970). That is, along with economic development, the expansion of educational opportunities gradually loosens the link between social origins and educational attainment, as more and more children of the lower classes enter the educational system at all levels (Hout 1988). Using this argument, we would expect a gradual increase in the rates of school enrollment and at progressive levels over time. In contrast, the second image is the persistence of inequality in educational stratiﬁcation. In a major comparison of educational processes in thirteen countries, researchers documented “persistent inequality” in the processes of educational attainment in a variety of social contexts (Shavit and Blossfeld 1993). Raftery and Hout (1993) proposed the “maximally maintained inequality” (MMI) hypothesis to characterize this pattern in which “transition rates and inequality remain constant unless forced to change by increasing enrollments” (p. 42). Maximally maintained inequality may stem from several sources. Raftery and Hout (1993) developed a rational choice explanation, arguing that such a pattern can be accounted for by individual choices in response to costs and beneﬁts associated with educational opportunities. Other scholars highlight the role of social classes in promoting their “collective projects” of class reproduction. This view is exempliﬁed in Bourdieu and Passeron (1977), who argued that the educational system is institutionalized into a stratifying machinery that controls access to social class destinations through the distribution of cultural capital – the ability to encode and decode

Educational Stratiﬁcation

73

information, hence the competitive edge, in the marketplace. In this view, education is no longer a neutral enterprise of objective knowledge and marketable skills. Rather, it serves as a sorting mechanism designed for class reproduction. This view leads to an emphasis on patterns of collective action, educational institutions, and macro-political processes through which interest groups promote their own agenda. These two imageries point to two different mechanisms: One is based on changes in opportunity structures and the other on allocative mechanisms, and they are not necessarily incompatible: The mechanisms of educational stratiﬁcation can remain intact, but we can observe gains in educational attainment across social groups resulting from the expansion of opportunities. Scholars studying stratiﬁcation under state socialism have developed a similar line of arguments on the role of the bureaucratic class. They pointed out that a new class of bureaucrats emerged in state socialist societies that acted in ways similar to the ruling classes in a capitalist economy (Djilas 1957). This elite class used state policies to secure privileges for themselves and their families, including access to better education for their children. In this light, patterns of educational stratiﬁcation under state socialism should resemble that in a capitalist economy. For example, research shows that although educational opportunities for lower classes initially expanded following the establishment of the socialist states, the effects of social origins reemerged in a later stage of the socialist development, especially for those progressing to advanced educational levels. But there is some uneasiness in developing such parallel arguments between capitalist and state socialist societies. In the former, the marketplace provides the arena for families to transmit their status to their children through “marketable skills.” But in state socialist societies, both educational opportunities and labor “markets” are organized by the state. Macro-political processes and state policies play a decisive role in creating and allocating educational opportunities in socialist societies. For the Soviet Union, see Gerber and Hout (1995); for Eastern European societies, see Hanley and McKeever (1997), Heyns and Bialecki (1993), Mateju (1993), Simkus and Andorka (1982), and Szel´enyi et al. (1998); in the Chinese context, see Deng and Treiman (1997), Hannum and Xie (1994), Kraus (1981), Rosen (1982), Unger (1982), and Whyte and Parish (1984). Moreover, the stratiﬁcation dynamics discussed in Chapter 1 point to the impact of the socialist state on both the opportunity structure and the mechanisms used to allocate these opportunities to individuals (see also Zhou, Moen, and Tuma 1998). We have reason to suspect that, because of the state’s strong grip on entrance into and departure from the educational

74

The State and Life Chances in Urban China

system in urban China, parental economic resources are far less important than parental political status for educational attainment. Moreover, as we noted earlier the state routinely and systematically redistributes resources and opportunities. As a result, family social origins are inﬂuential only to the extent that they are supported by state policies. The suspicion and mistrust of intellectuals in the Mao era raise questions about the role of parental education on children’s educational attainment (Yao 1984). The political value of parents’ occupational or educational status may vary with state policies over time, since macro-political processes affect the very notion of “capital” as well as how it is transferred across generations. These processes in turn affect the relationship between parental social origins and educational attainment. As we shift our attention from market societies to state socialist societies, we move from a focus on the association between family resources and educational attainment in the marketplace to the recognition of the central role of the state and macro political processes that exert strong impacts on this association. In this light, we argue that neither the industrialization thesis nor the MMI thesis adequately captures the impacts of stratiﬁcation dynamics; instead, we propose a third view that emphasizes the direct role of the state and state policies on patterns of educational attainment. A main hypothesis in this regard – the hypothesis of stratiﬁcation dynamics – is that, under state socialism, the effects of individual-based family resources and interest groups are contingent on shifting state policies. Speciﬁcally, we expect that the relationship between social origins and children’s educational attainment in China varies across historical periods, and that these variations are best explained by macro-political processes directed by the state. Take the role of local bureaucrats as an example. When the state depended on the bureaucratic apparatus to implement its policies, the bureaucrats played a key role in resource allocation and were in a strategic position to enhance their children’s educational attainment. In these circumstances, we expect to ﬁnd a strong association between social origins and educational attainment. But in times of political mobilizations against bureaucratic apparatus, as in the Cultural Revolution period, the bureaucrats’ status and privileges were greatly weakened by the prevailing state policies of the time. In these circumstances, we expect the association between social origins and educational attainment to become weakened, and possibly even reversed. The rest of this chapter focuses on the empirical evidence in light of the issues discussed here. We draw on detailed life history information in our data on parental socioeconomic status, historical context, and the timing

Educational Stratiﬁcation

75

of individuals’ educational attainment to examine education stratiﬁcation in urban China and how changing state policies and redistributive processes affected the relationship between social origins and educational attainment. Our analytical focus is to address the following questions: (1) How is the relationship between social origins and educational attainment shaped by the institutional arrangements and political processes in China? (2) How do varying state policies and political dynamics affect educational opportunities over time?

historical evolution of educational opportunities [In 1963, Yaoyao, the daughter of a famous actress, graduated from the special music school in Shanghai.] She did not pass the entrance examination to the piano specialty in Shanghai Music Conservatory. Her family background was not that of the working class, so she could not get the quota set up for political recommendation, either. Her school asked her to go to state farms in the Xinjiang Province and to spend her youth in the remote areas of the Motherland. But Yaoyao did not go . . . A few months later, her mother helped her enroll in the vocal music department in Shanghai Music Conservatory. . . . (D. Chen 2000, p. 82). On the eve of the Cultural Revolution, the political atmosphere was tense, with slogans about class struggles and family backgrounds all over the place. Because I had bad family class background, I could not enter college after the college entrance examination. Nor was I allocated a job by the government. I was idle at home at the age of 17, having to be supported by my mother . . . Because of this situation, I volunteered to enter a special training program for rural work and eventually went to the rural areas [for the next 9 years]. – Recollection by an ordinary citizen (Yang 1992, p. 66).

The thesis of industrialization portrays a process of gradually increasing opportunities for lower classes in tandem with the expansion of the educational system. But, as the previous two examples show, a striking characteristic of China’s post-1949 history is that educational opportunities have ﬂuctuated dramatically over time, especially in response to political shifts such as the Cultural Revolution. To demonstrate this point, let us ﬁrst take a look at the historical trend of educational opportunities in urban China. Figure 3.1 shows the proportions of those in the risk set who entered four educational levels – elementary, middle school, high school, and college – over time, based on the data from our sample.1 Notice that 1

To avoid exaggerating ﬂuctuations in a speciﬁc year due to sampling variability, Figure 3.1 plots the ﬁve-year moving average of the proportion that entered the next educational level.

The State and Life Chances in Urban China

76 1

proportion

0.8

0.6

0.4

0.2

0

53

58

63

68

73

78

83

88

year elementary

junior high

senior high

college

ﬁgure 3.1. Rates of entry into four educational levels in the sample, 1953–1988 (ﬁve-year average).

this ﬁgure is different from Figure 2.5 in Chapter 2 – Figure 2.5 reports national statistics (including rural areas), whereas Figure 3.1 displays the patterns for only urban residents in our sample. As Figure 3.1 shows, the trends of entry into elementary and middle schools revealed a path of steady growth: There was a steady increase in the rates of entry to both lower levels of education. By contrast, the proportions of those entering high school and college reveal marked ﬂuctuation over time. Entry into high school declined steadily during the economic contractions associated with the aftermath of the “Great-LeapForward” episode and the early years of the Cultural Revolution (1962– 1968). The pattern was similar at the college level. Moreover, due to disruptions of the educational system (e.g., colleges being closed down for several years) caused by radical state policies, the proportion entering college during the entire period of the Cultural Revolution (1966–1976) was much lower than that in earlier and later periods.2 2

The apparently high proportion of entry into college in the late 1950s occurred partly because the government selected some college entrants from outside the pool of high school graduates, that is, cadres and workers already in the labor force.

Educational Stratiﬁcation

77

We use the term transition rate to describe the rate of entering an educational level among those who are eligible for such a transition (i.e., the risk set of those who graduated from the immediate lower level of education). Obviously, transitions into different educational levels differed signiﬁcantly in their historical trends. There are two distinctive patterns. First, transitions into the two lower educational levels (elementary and middle school) appeared to follow a path of gradual expansion of opportunities for children of all urban families and to be not vulnerable to changes in state policies over time. Second, the sharp ﬂuctuations associated with transition rates into high school and college levels were far from the conventional image of gradual expansion of educational opportunities along with industrialization as seen in other parts of the world; rather, the evidence showed drastic and nonlinear changes in opportunities, suggesting that individuals moving into adolescence and early adulthood experienced very different educational opportunities across historical periods. The ﬂuctuations in transition rates also cast doubts on the image of persistent inequality as suggested in the maximally maintained inequality hypothesis. State policies played a decisive role in shaping educational stratiﬁcation, which directly generate or indirectly induce dramatic changes in educational opportunities. As an illustration, let us take a look at the number of higher education institutions over time (see Figure 3.2). The great expansion of educational opportunities in the early period of the state socialist China (from 1949 to 1959) was the result of state policies in nation building and economic development. During this period, the new state socialist government’s primary goal was to consolidate its control through nationalization and collectivization and to promote economic development, leading to the “Great-Leap-Forward” episode (1958–1959). As part of the “rushed growth” strategies, the government greatly expanded educational institutions at all levels to meet the increasing demand for an educated work force, hence induced an exponential expansion of educational opportunities. For example, China had only 229 formal colleges in 1957, but grew to 1,289 by 1960, a ﬁve-time increase in three years (SSB 1999, p. 75). Contrast the preceding pattern with that in the Cultural Revolution period, from 1966 to 1977. In this period, many high schools and colleges closed their doors. A large proportion of graduates from middle and high schools were forced to work in rural areas (see Chapter 5). As a consequence, enrollments at the high school or higher level sharply declined

The State and Life Chances in Urban China

78 1500

number of colleges

1250 1000 750 500 250 0 49

54

59

64

69

74

79

84

89

94

year

ﬁgure 3.2. Number of higher education institutions, 1949–1994.

between 1966 and 1968. Although the number of colleges appeared to be steady in this period (Figure 3.2), in fact all colleges were closed at the beginning of the Cultural Revolution period. When a limited number of colleges reopened in 1972, only a small number of students were allowed to enter college based on political selection criteria. Moreover, recruitment into college was selected from those already in the labor force and the normal progression from high school to college was discontinued in this period (1968–1977). Not only have the opportunities of education changed greatly over time, so have the selection criteria that were used for recruitment. In the early years, merit-based selection criteria were used in advancement through formal educational levels, favoring those bourgeois families and other middle-class families that could provide better economic or cultural resources for school preparation. In contrast, during the Cultural Revolution political selection dominated educational progression, which favored the children of workers and peasants. However, the severe limitation of educational opportunities in this period adversely affected all social groups. These striking patterns raise important questions: In what ways have state policies affected educational attainments for the children of different social groups and the association between social origins and educational attainment over time? In particular, have the high-status groups been able to shelter their children from adverse state policies? If the high-status

Educational Stratiﬁcation

79

0.5

proportion

0.4

0.3

0.2

0.1

0 53

58

63

68

year

father worker

73

78

83

88

father cadre/prof.

ﬁgure 3.3a. Rates of entry into high school, by the father’s occupation (ﬁve-year average).

groups could signiﬁcantly protect their children from the adverse inﬂuence of state policies, we should expect to ﬁnd much less ﬂuctuations of life chances for the children of the bureaucratic class. We plotted the rate of entry into high school and college for two groups: (1) those whose fathers had a cadre or professional occupation (highstatus groups), and (2) those whose fathers were workers or peasants; see Figures 3.3a and 3.3b. We observed several characteristics: First, educational inequality based on social origins was present throughout the entire history of the People’s Republic of China. At both the high school and the college levels (except for college enrollment brieﬂy in the mid-1970s), the proportion of those advancing in education was higher for those whose fathers were cadres or professionals than for those whose fathers were workers or peasants. Second, the extent of inequality varied over time. At both high school and college levels, inequality was high in the earlier periods (before the mid-1960s), much lower during the Cultural Revolution, and again substantial in the post-Mao era (after 1977). It is interesting to note that, although all social groups were affected by radical social policies, ﬂuctuations were greatest for those having high-status social origin. For example, of those with fathers in higher status occupations, more

The State and Life Chances in Urban China

80 0.2

proportion

0.15

0.1

0.05

0 53

58

63

68

73

78

83

88

year father worker

father cadre/prof.

ﬁgure 3.3b. Rates of entry into college, by the father’s occupation (ﬁve-year average).

than 30 percent entered the high school level in the early 1960s, whereas only about 10 percent did so in the mid-1960s; we found similar patterns at the college level. Over time, state policies both altered the overall level of opportunities and re-allocated opportunities among social groups. The sharp declines in the transition rates for children of cadres and professionals to the next educational level during certain periods suggest that state policies might have redeﬁned what constituted privileged groups (e.g., favoring such family class background as worker or peasant), generating drastic changes in life chances among children of different social groups.

state policies, social origins, and educational attainment: further analyses In the last section, we argued and showed descriptive evidence that the state and state policies played a major role in allocating educational opportunities. We now take a closer look at the processes of educational attainment over time. The main task in this section is to examine the role

Educational Stratiﬁcation

81

of state policies in mediating the association between social origins and educational attainment. Following the strategies outlined in Chapter 2, we incorporate distinctive historical contexts into the analysis to capture the effects of varying state policies over time. Here, we examine how the association between social origins and educational outcomes varied across historical periods and with distinctive state policies. To accomplish this goal, we want to enhance our analytical power by adopting formal, statistical models to scrutinize the data in a more rigorous manner. There are several reasons for doing so. First, as is well recognized in studies of educational stratiﬁcation (Mare 1981), students’ attainment of various educational levels reﬂects two processes: One is marginal changes due to the expansion of opportunities; the other is changes in allocative mechanisms, such as the association between social origins and educational attainment. Our substantive interest is on the issue of allocative mechanisms. The adoption of appropriate statistical models allows us to disentangle these two distinct processes. Second, although our emphasis is on the association between social origins and educational attainment, we need to consider other confounding processes. For example, the number of siblings in a family may affect the distribution of family-based resources for children’s educational achievement. So we need to take into consideration these additional factors through statistical modeling. To set up the research and analyses reported in the following sections, we brieﬂy outline some design considerations.3 To study educational advancement empirically, we focus on the transition rate of entry into different educational levels (e.g., entrance into middle school, high school, or college) rather than on departure or graduation from each level. The focus on entry is consistent with the fact that state policies of exclusion and discrimination in China occurred mainly at the point of admission into educational institutions. Moreover, state policies affecting the content and length of educational programs frequently shifted across historical periods. For many students, graduation was often delayed or rushed in response to policy changes. For this reason, the timing of graduation is less useful than the timing of entry in assessing the effects of historical context and social origins on educational attainment.

3

In the analyses of educational transitions, we included the data for both respondents and their spouses. This decision is partly based on the observation that most respondents completed their education prior to marriage and partly based on the practical consideration that the larger sample allows enough subsample to achieve reasonable statistical power for model estimation at different levels of educational transition and across historical periods.

82

The State and Life Chances in Urban China

Following conventions, we use the father’s educational level and occupational status as measures of social origin. In the Mao era, family class background (juating chushen) was used as the main indicator of one’s class location in the political selection process, so we also use an alternative set of class labels in some analyses. To capture varying state policies and ﬂuctuations in macropolitical processes over time, in the analyses of educational stratiﬁcation, we speciﬁcally distinguish four historical periods based on the educational experiences of urban residents in China. Both educational opportunities as well as selection criteria varied across these historical periods. The ﬁrst period is between 1949 and 1959. This was a period of nation building and economic development, with an expansion of educational opportunities, and rushed growth. As Figure 3.1 shows, during this period, the trends of educational progression were increasing at various educational levels. The second period (1960–1965) was a time of economic contraction following the aftermath of the economic disaster in the early 1960s. During this period, economic development was seriously disrupted and educational opportunities at the high school and college levels shrank. Political intervention increased at this time with the initiation of state destratiﬁcation policies. That is, policies explicitly aimed at reducing or eliminating social inequalities among social groups. The third period (1966–1977) was the period of the Cultural Revolution, which paralyzed educational institutions and severely disrupted the educational progression of millions of school-age youth. Political selection processes reached its peak during this period. In the fourth period (1978–1994), the post-Mao era, the resumption of college entrance examination in 1977 signaled a new beginning in China’s educational system: Merit-based educational qualiﬁcations became the primary factor in college admission and upward mobility. In this period, technical schools, vocational colleges, and other types of college programs mushroomed, widening the doors to higher education in an unprecedented pace in Chinese history. Before we leave the discussion of research design, we want to note that our analytical focus on historical periods differs from the birth cohort-based analyses in the previous research on changes in educational opportunities. In a ﬂuctuating social environment like China’s, a cohort approach sometimes leads to misspeciﬁcation of the timing of educational opportunities. As we observed before, many educational institutions (especially colleges) were closed during the Cultural Revolution. Students

Educational Stratiﬁcation

83

who had graduated from middle and high schools during this period were forced to postpone further education until after the Cultural Revolution. A design based on the historical period in which an individual is at risk of an educational transition directly links patterns of educational advancement to period-speciﬁc state policies affecting educational opportunities. Accordingly, we constructed “risk sets” (i.e., those available/eligible for entry into the next higher level of education) that incorporate both interruptions in educational progression and ﬂuctuations in the number of school years needed to complete a given educational level (e.g., middle school, high school). Speciﬁcally, we included all individuals who graduated from the previous educational level in the risk set for advancing to the next level in the current and subsequent periods and examined the probability of entering different educational levels in speciﬁc historical periods. For example, those graduating from high school during the Cultural Revolution were in the risk set for attending college during and after the Cultural Revolution. One exception to the general rule in constructing risk sets is the one for college entry. We included respondents who graduated from middle school between 1966 and 1968 in the risk set for college entry after the Cultural Revolution. This decision was based on our knowledge of the historical context. Although this group was deprived of the opportunity to attend high school during the Cultural Revolution, a large number of individuals in this group went on to take college entrance examinations after the Cultural Revolution (see Zhou, Moen, and Tuma 1998, for additional information on research design considerations). Overview: Do Social Origins Matter? Let us begin with this basic question: Do social origins, such as father’s education and father’s occupational status (or family class background) play a signiﬁcant role in children’s entries into different educational levels in urban China? One way to address this question is to ask: How likely is it to observe the kinds of patterns in our sample, if there is no association between social origins and educational outcomes? If we begin with this working “hypothesis,” we can use a formal statistical test (technically, the likelihood ratio test based on estimated statistical models) to explicitly test this hypothesis against the observed data. If there is no association between the two, then variables measuring social origins should have no statistically signiﬁcant

The State and Life Chances in Urban China

84

table 3.1. Likelihood Ratio Tests of Improvement in Model Fit Entry Into Middle School

(1): Intercept, duration, number of siblings (2): (1) + Female (3): (2) + Father’s education (4a): (3) + Father’s occupation (4b): (3) + Family class background (5): (4a) + Residential location ∗∗ ∗

Entry Into High School

Entry Into College

χ2 Statistic

d.f.

χ2 Statistic d.f. χ2 Statistic d.f.

—

—

—

—

—

—

124.1∗∗ 353.3∗∗ 160.3∗∗ 87.3∗∗

4 16 26 16

31.6∗∗ 289.2∗∗ 71.0∗∗ 120.0∗∗

4 16 26 16

51.7∗∗ 193.5∗∗ 72.5∗∗ 101.3∗∗

4 16 26 16

255.0∗∗

16

113.7∗∗

16

135.8∗∗

16

p < 0.01 p < 0.05

effects on the transition rates to different educational levels. We conducted such formal tests for transition rates to middle school, high school, and college levels. To summarize the main results: For all three transitions, social origins, as measured by father’s education, father’s occupation, or family class background, respectively, contribute signiﬁcantly to the transition rates to these levels. Gender and residential location also play a signiﬁcant role in one’s access to educational opportunities at all three levels. Technically, as Table 3.1 shows, we begin with a baseline model that includes the grand mean, one’s duration since the graduation from the lower level and number of siblings. Model 2 adds gender. Model 3 includes father’s education. Models 4a and 4b use two alternative measures of father’s status – occupational status or family class background, respectively. Because father’s occupation and family class background overlap considerably, we do not include both sets of variables in the same model. Finally, Model 5 adds residential location. All these variables contribute to a signiﬁcant improvement of model ﬁt. It is noteworthy, however, that the improvement in model ﬁt for entry into high school and college is appreciably greater for family class background variables than for the father’s occupation. The χ 2 values for adding family class background variables are 120.0 and 101.3 with 16 degrees of freedom (d.f.) for high school and college, respectively; whereas the χ 2 values for adding the father’s occupational status variables are 71.0 and 72.5 with 26 d.f. for high school and college, respectively. This pattern suggests that access to

Educational Stratiﬁcation

85

higher education in China depended more on family class background than on the father’s occupational status. The general conclusion is that there is indeed a signiﬁcant association between social origins (as measured by father’s educational levels and the father’s occupational status, or, alternatively, family class labels) and educational attainment for transitions into the three educational levels. Although this ﬁrst step ascertains the signiﬁcant role of social origins in educational attainment, it does not give us information on the speciﬁc impacts of social origins, or whether such impacts vary across historical periods. To address these key issues, we need to take a closer look at the patterns through more detailed analyses. Effects of Gender and Social Origins on Transition Rates Across Historical Periods To examine the speciﬁc roles of social origins, among other factors, on the transition rates to educational levels, we used logistic regression model for discrete event history data to estimate the effects of these variables. In these analyses, the outcome or event is the entry into a speciﬁc educational level, given that the person has completed the preceding level. The statistical estimates are reported in Table 3.2.4 To put these results in a historical perspective, note that before the 1980s, high school was practically the highest educational level possible for the majority of Chinese urban youth, and college was a rare opportunity. We demonstrate the main ﬁndings using more intuitive graphic displays. To better understand the statistical patterns in the tables and ﬁgures presented here, let us begin with some preliminaries. To measure the transition rate, we use the “logit” – the log-odds of experiencing a transition for one group compared with the reference group. For example, the logit associated with the variable “female” in the transition rate to college gives the log-odds (the logarithm of odds) of experiencing the event of entering college for women compared with men (the reference group). In the context of changes over time (as in our case here), we often interpret the log-odds as the log-rate of transition. Therefore, the magnitude of logit gives an intuitive interpretation as a transformed (in logarithm) transition rate into an educational level for a speciﬁc group. 4

Because the percentages of fathers who were cadres or professionals were fairly small in the earlier years, we combined low- and high-rank cadres into a single category of cadre, and low- and high-rank professionals into a single category of professional in the ﬁrst period (1949–1959).

86

The State and Life Chances in Urban China

table 3.2. Parameter Estimates of Logistic Regression Models of Probability of Entering Three Educational Levels, by Historical Period Historical Period Covariates entry into middle school Female Years since prev. graduation father’s education Elementary Middle school High school or above Missing father’s occupation High-rank cadre Low-rank cadre High-rank professional Low-rank professional Farmer Others Missing residential location Medium-sized city Small-sized city Rural area Missing Number siblings Intercept entry into high school Female Years since prev. graduation father’s education Elementary Middle school High school or above Missing father’s occupation High-rank cadre Low-rank cadre High-rank professional Low-rank professional Farmer Other Missing

1949–1959

1960–1965

1966–1977

1978–1994

−0.873∗∗ −1.387∗∗

−0.628∗∗ −2.519∗∗

−0.376∗∗ −2.379∗∗

−0.069 −0.997∗∗

0.390∗∗ 0.511∗∗ 1.319∗∗ −0.027

0.768∗∗ 1.179∗∗ 2.161∗∗ 0.491†

0.266∗ 0.722∗∗ 0.445∗ 0.550∗

0.287 0.635∗∗ 1.561∗∗ 1.267∗∗

0.639∗,a — 1.392∗∗,a — −0.092 0.494∗ −0.051

−0.014 0.024 −0.575 −0.125 −0.627∗∗ −0.239 −0.297

0.854† 0.395∗ 0.317 0.125 −0.326† 0.034 −0.404∗

0.176 −0.071 2.699∗∗ −0.591∗ −0.668∗∗ −0.269 −0.626∗

−0.794∗∗ −0.510∗∗ −0.682∗∗ 0.516∗∗ 0.006 0.749∗∗

−0.883∗∗ −0.839∗∗ −0.733∗∗ 0.956∗∗ −0.044 1.678∗∗

−1.132∗∗ −1.382∗∗ −1.453∗∗ −0.833∗∗ −0.082∗∗ 2.975

−0.427∗ −0.626∗∗ −0.399† −0.038 −0.084∗ 2.107∗∗

−0.446∗∗ −1.234∗∗

−0.359∗∗ −1.444∗∗

−0.065 −1.509∗∗

−0.266∗ −0.518∗∗

0.389∗ 0.596∗ 0.716∗∗ −0.181

0.529∗∗ 0.686∗∗ 1.182∗∗ 0.613∗

0.382∗∗ 0.373∗∗ 0.432∗∗ 0.371∗

0.308∗∗ 0.751∗∗ 1.053∗∗ 0.530∗∗

0.728†,a — −0.417a — 0.073 0.025 0.341

0.350 0.177 0.850 0.400 0.180 0.305 0.048

0.571∗∗ 0.198 0.115 0.064 0.172 −0.078 0.156

0.876∗∗ 0.402∗∗ 0.520∗∗ 0.404∗∗ 0.180 0.225∗ 0.395∗∗

Educational Stratiﬁcation

87

Historical Period Covariates residential location Medium-sized city Small-sized city Rural area Missing Number siblings Intercept entry into college Female Years since prev. graduation father’s education Elementary Middle school High school or above Missing father’s occupation High-rank cadre Low-rank cadre High-rank professional Low-rank professional Farmer Other Missing residential location Medium-sized city Small-sized city Rural area Missing Middle school graduate Number siblings Intercept

1949–1959

1960–1965

1966–1977

1978–1994

−0.329 −0.078 −0.974∗∗ 0.200 −0.059† −0.040

−0.220 −0.207 −1.087∗∗ 0.074 −0.029 −0.517∗

0.051 0.138 0.209 −0.296∗ −0.105∗ −0.648∗∗

−0.452∗∗ −0.493∗∗ −0.437∗∗ 0.016 −0.085∗∗ −0.642∗∗

−0.994∗∗ −0.858∗∗

−0.389† −0.769∗∗

−0.709∗∗ −0.114∗∗

−0.408∗∗ −0.109∗∗

0.508 0.170 0.728 0.382

0.605† 0.844∗ 1.354∗∗ 0.620

−0.120 0.154 −0.226 0.171

0.226 0.665∗∗ 1.010∗∗ 0.566∗∗ 0.790∗∗ 0.555∗∗ 0.928∗∗ 0.617∗∗ 0.973∗∗ 0.240 0.539∗∗

0.615a — 0.509a — 0.120 −0.218 −0.154

0.658 −0.332 −0.041 0.298 0.634† 0.339 0.364

1.012∗ 0.271 1.052† 0.767† 0.588 −0.048 0.455

0.012 −0.721∗ −1.970∗∗ 1.099∗∗ −2.327∗∗ 0.097† −1.455∗∗

−0.603 −0.069 −0.546 0.731∗ −3.018∗∗ −0.010 −2.482∗∗

−0.570 −0.836∗ −1.105∗ 0.003 −0.754∗∗ 0.071 −4.686∗∗

−0.677∗∗ −0.022 −1.173∗∗ 0.140 −1.224∗∗ −0.053∗ −3.784∗∗

∗∗

p. < 0.01 p. < 0.05 † p. < 0.10 a High- and low-rank cadres and professionals are combined, respectively, in the earliest period due to small number of cases in the cadre and professional categories. Note: Illiteracy is the reference category for father’s education; worker, for father’s occupation; and large city, for residential location. ∗

Now we are in a position to take a closer look at the ﬁndings revealed in Table 3.2. Recall that our primary concern is about the association between social origins and educational outcome, and how this association changes over

88

The State and Life Chances in Urban China

time with varying state policies. So our analytical focus is on changes in the association between social origins and transition rates across historical periods. Before we consider social origins, let us ﬁrst consider a more familiar case – gender inequality in educational attainment over time. Our purpose here is to introduce basic statistical patterns through this example. But evolving gender roles are of substantive interest as well. Elimination of gender inequality was one of the proclaimed social policies of the socialist state. Previous studies have found marked improvements in women’s educational attainment in the Soviet Union and Eastern Europe (Gerber and Hout 1995; S. Szel´enyi 1998), as well as noticeable variations over time in access to elementary and middle school in China (Hannum and Xie 1994; Lavely, Xiao, Li, and Freedman 1990). But, there were also contradictory ﬁndings with regard to gender inequality (Whyte 1984; Wolf 1985). If our emphasis on the role of the state and state policies is sensible, we expect that the gender role will vary with changing social policies over time. Let us begin with an imagined “ideal” scenario consistent with the industrialization thesis. Suppose we observe an initial state of gender inequality. Over time, the advancement of industrialization would gradually expand educational opportunities, resulting in the narrowing of the gender gap. If this is indeed the case, we should observe the transition rates (in logit) to different educational levels for women to increase and approach zero (e.g., no gender difference) over time. Moreover, we may also expect variations in these changes across the education levels, with faster paces in the narrowing of gender gap at lower levels of education. The patterns in gender inequality over time show a much more complicated picture as seen in Figure 3.4, which is based on the estimated coefﬁcients for females in Table 3.2. The four periods, again, are as follows: 1949–1959, 1960–1965, 1966–1977, and 1978–1994. We observe several interesting patterns. First, females’ transition rate for middle school followed a path that closely resembles the ideal scenario: The rate of transition for females increased steadily over the four periods and in the last observation period approaches the “zero” level, a state of no gender difference. This pattern indicates that, at the lower level, educational attainment did follow the path of the industrialization hypothesis with respect to gender. That is, increases in the transition rate for women were due to the expansion of opportunities and were not affected by changes in mechanisms of educational stratiﬁcation. In contrast, gender differences in the transition rates to high school and college followed strikingly different patterns. Starting with the initial

Educational Stratiﬁcation

89

transition rate (logit)

0.5

0

-0.5

-1 1949-1959

1960-1965

1966-1977

1978-1994

period middle school

high school

college

ﬁgure 3.4. Transition rates by period and gender.

gender gap in the ﬁrst period, there was a narrowing of gender gap at both high school and college levels in the second period; given the initial state, it appeared that the narrowing of gender gap was faster at the college level than at the high school level. However, during the third period of the Cultural Revolution, the gender gap was the narrowest at the high school level but largest at the college level. In the reform era (the fourth period), gender gap in transition to high school reappeared, whereas gender gap in transition to college narrowed. To explain these zigzags of transition rates over time, we need knowledge about the historical context and varying state policies across these periods. Recall that the ﬁrst period witnessed the great expansion of educational opportunities in the process of nation building. The second one was a period of economic contraction; as a result, educational opportunities shrank greatly during this period (see Figure 3.1 for the historical trend). So it is surprising to see that the gender gap was still narrowing during this period, especially at the college level. This is inconsistent with the MMI thesis, which would predict a widening gender gap in favor of men in the time of shrinking opportunities. It is likely that those state policies that promoted gender equality played a signiﬁcant role in leveling the

90

The State and Life Chances in Urban China

ground for women. During this period, there were conﬂicting “red versus expert” educational policies (Yao 1984). On the one hand, the failure of the rushed growth policies led to more emphasis on merit-based admission and formal educational curricula; on the other hand, Mao Zedong repeatedly called for the reform of the educational system. We suspect that increased gender equality reﬂected strong state intervention into the educational processes. The Cultural Revolution era (the third period) shows striking ironies. On the one hand, because of the destratiﬁcation policies, gender inequality at the high school level was virtually eliminated; on the other hand, such equality was achieved at the expense of the deprivation of educational opportunities for all men and women (see Figure 3.1). In contrast, college opportunities were extremely limited during this period, and the selection mechanism was stringently in favor of men. If gender parity was a proclaimed state policy in the Cultural Revolution, it was implemented inconsistently. In the fourth period, gender inequality reemerged at the high school level and appeared to return to the patterns in the pre-Cultural Revolution period. It seems that both the expansion of education opportunities and the state policies affect gender inequality in education stratiﬁcation over time. The monotonic trend in increasing gender equality for education at middle school level was mainly due to the former process, whereas the zigzag patterns for education at high school and college level were largely attributable to the latter one. We now turn to the primary focus of our analyses: How does the association between social origins and transition rates vary over time? We begin with the father’s education level and its impacts on children’s educational attainment over time. As we noted before, in the Mao-era the top leaders had a deep distrust of those with an intellectual background, so educational credential has been an ambiguous asset in China. We expect that respondents’ educational attainment is not always positively associated with fathers’ education. Figure 3.5a shows the transition rates (in logit form) to high school at different levels of the father’s education (the reference group being those fathers who were illiterate) across periods. Because education for the parent generation was limited, we combined fathers who had high school education and fathers with a college degree. With the exception of the third period, transition rates varied systematically with father’s educational levels, with children of better-educated father having a higher transition rate. However, there were noticeable variations across these

Educational Stratiﬁcation

91

1.4

transition rate (logit)

1.2 1 0.8 0.6 0.4 0.2 0 1949-1959

1960-1965

1966-1977

1978-1994

period elementary

middle school

high school or above

ﬁgure 3.5a. Transition rates to high school by period and father’s education.

periods that cannot be explained by either the industrialization thesis or the MMI thesis. For example, in the Mao era (the ﬁrst three periods), educational inequality on the basis of father’s education was the largest in the second period, when there was a drastic decline in educational opportunities and when the government adopted policies that “emphasize students’ academic records and merits and no longer adopt the criteria of students’ political qualiﬁcation and class background” (Yao 1984, p. 59). In the second period, compared with those whose fathers are illiterate (the reference category), the odds of entering high school were about 1.7 times higher for those whose fathers had an elementary education, about twice as high for those whose fathers had a middle school education, and about 3.2 times higher for those whose fathers had a high school or college education. In contrast, during the Cultural Revolution, when destratiﬁcation policies were most effective and political criteria prevailed, children of all social groups had similar transition rates to high school and the positive effect of father’s education was drastically reduced or even reversed. The father’s education regained its importance in facilitating educational advancement in the post-Mao era.

The State and Life Chances in Urban China

92 1.5

transition rate (logit)

1

0.5

0

-0.5 1949-1959

1966-1977

1960-1965

1978-1994

period elementary

middle school

high school or above

ﬁgure 3.5b. Transition rates to college by period and father’s education.

The sharpest shifts in educational selection occurred at the college level (see Figure 3.5b). In the early period, entry into college was heavily inﬂuenced by the policies of rushed growth and not affected signiﬁcantly by father’s education (as evidenced by statistical insigniﬁcance among educational levels). In the 1960–1965 period, however, the odds of entering college rose systematically with the father’s education. Having an educated father (one with a high school or college education) rather than an illiterate father increased the odds of entering college by more than three times. In sharp contrast, the father’s education had no signiﬁcant effect on the odds of college entry during the political turmoil of the Cultural Revolution. Those with educated fathers actually had lower (though not statistically signiﬁcant) odds of entering college. In the post-Mao era, the father’s higher education again signiﬁcantly increased the odds of entering college. These results clearly demonstrate changing selection criteria in educational stratiﬁcation across historical periods. These time-varying effects suggest that the role of social and cultural capital (as indicated by father’s education) in facilitating access to educational opportunities has been highly sensitive to state policies.

Educational Stratiﬁcation

93

Effects of Family Class Background In our statistical analysis, the father’s occupational status as a measure of social origin does not appear to have played a signiﬁcant role over time, so we will not discuss its effect here. In contrast, family class background, an alternative measure of parental political status, has played a salient role. In the Mao era, the Communist Party used family class background as an explicit criterion for political selection. Family class background was based on the ofﬁcial classiﬁcation of the social class of one’s father or grandfather before the establishment of the People’s Republic of China in 1949. This dimension of class background is especially interesting because, unlike the father’s occupational status, which may change over time (e.g., from professional to cadre), family class background became ﬁxed once it was ofﬁcially certiﬁed. Was there any association between the political label of class background and educational attainment? If there was, did the association change over time? We estimated the effects of family class background on the rates of entry into the three educational levels (Table 3.3). Because the class background labels overlap with the father’s occupational status and education, we included only the “class background” variables and controlled provinces and types of cities in the model. Figures 3.6a and 3.6b demonstrate this relationship over the four periods. Figure 3.6a shows transition rates to high school by family class background, with “worker/peasant” as the reference category. Here political selection mechanisms varied noticeably across historical periods. In the ﬁrst two periods, children with a “cadre” background and those of “exploiting class” background had similar advantages; in contrast, children with a “worker/peasant” background (the reference group) were disadvantaged. This trend was exacerbated in the second period, as all three groups witnessed higher rates relative to the children of “worker/peasant”; children of cadre background gained most in this period. But during the Cultural Revolution, political selection became stringent, leading to a sharp drop in the transition rate for those with “middle-class” and “exploiting class” backgrounds; children of cadres still managed to hold on to their advantages during this period of shrinking opportunities. In other words, the children of workers/peasants (the reference category) had the largest gain in this period. In the reform era, those from the cadre background regained their lost ground, but there were no statistically signiﬁcant differences among children of other groups.

table 3.3. Parameter Estimates of Class Background Effects on Probability of Entering Three Educational Levels, by Historical Period Historical Period Covariates

1949–1959 1960–1965 1966–1977 1978–1994

entry into middle school family class background Cadre Middle class Exploiting class entry into high school family class background Cadre Middle class Exploiting class entry into college family class background Cadre Middle class Exploiting class

1.895∗∗ 0.710∗∗ 0.748∗∗

2.490∗∗ 0.320∗∗ −0.394†

1.920∗∗ −0.244∗∗ −0.888∗∗

2.394∗∗ −0.282∗∗ −0.926∗∗

0.597∗ 0.513∗∗ 0.550∗

1.425∗∗ 0.758∗∗ 0.704∗∗

0.932∗∗ −0.314∗∗ −0.986∗∗

1.322∗∗ 0.026 −0.073

0.941∗ 0.618∗∗ 0.820∗

0.627 0.577∗∗ 0.119

0.934∗∗ 0.008 −1.075

1.598∗∗ 0.662∗∗ 0.376

∗∗

p < .01 p < .05 Note: Worker/peasant is the omitted category for family class background; middle class includes small business owners, middle-rich peasants, and intellectual family background. Exploiting class includes large business owners, landlords, rightists, and bad elements. Estimates are obtained while controlling for provinces and types of city.

∗

2

transition rate (logit)

1.5 1 0.5 0 -0.5 -1 -1.5 1949-1959

1966-1977

1960-1965

1978-1994

period cadre

middle class

exploiting class

ﬁgure 3.6a. Transition rates to high school by period and class label.

Educational Stratiﬁcation

95

2

transition rate (logit)

1.5 1 0.5 0 -0.5 -1 -1.5 1949-1959

1960-1965

1966-1977

1978-1994

period cadre

middle class

exploiting class

ﬁgure 3.6b. Transition rates to college by period and class label.

Transition rates to college show more dramatic patterns (see Figure 3.6b). Here, we ﬁnd that children with exploiting class backgrounds experienced a declining rate from the very beginning. In contrast, those with middle-class backgrounds were advantageous before the Cultural Revolution, probably due to the cultural capital in those families. Children of the cadre families had advantages in all those periods, except for the second one, when severe shrinking of opportunities eliminated their privileges (as indicated by no statistically discernable differences between this group and the reference category). In the reform era, all three groups appeared to have gained advantages over the children of worker/peasant (the reference category). These results again show strong evidence of the impacts of varying state policies on the links between family social origins and educational attainment. But the conspicuous, relative advantages for the children with “cadre” background through the whole history in China also show some evidence of persistent inequality.

chapter summary Numerous studies of educational attainment in industrialized societies have documented persistent inequality over time and across societies (see Hallinan 1988). By distinguishing the effect of marginal changes, Mare

96

The State and Life Chances in Urban China

(1981) found a trend of increasing association between social origin and schooling progression across the cohorts who were born in the ﬁrst half of the twentieth century in the United States. Shavit and Blossfeld (1993) revealed persistent educational inequality in a variety of institutional contexts (see also Treiman and Yip 1989). Against this background, this chapter revealed the distinctive Chinese experience and shed new light on the mechanisms that govern the association between social origins and educational attainment. We have considered three lines of arguments to guide our examination of educational stratiﬁcation in urban China. The ﬁrst is the “industrialization” theme, which hypothesizes a gradual opening up of educational opportunities for all social groups. The second is the “persistence of inequality” theme, which predicts a state of maximally maintained inequality based on class boundaries. The third approach emphasizes the role of the state and state policies that transcend group boundaries, leading to a nonmonotonic trend of changes over time. The evidence presented in the preceding discussions shows two distinctive patterns of educational attainment. On the one hand, at the lower levels of elementary and middle school enrollment, we observed a gradual expansion of educational opportunities that renders social origins irrelevant. On the other hand, transition rates into high school and college, where opportunities were scarce, show a strong, nonlinear trend of selection processes. It is clear that both educational opportunities and the association between social origins and educational attainment varied signiﬁcantly over time in the history of the People’s Republic of China, indicating the active role of the state and state policies in altering opportunities and stratifying processes. A further issue arising from our discussion of the stratiﬁcation dynamics is the role of the bureaucratic class in securing privileged access to educational opportunities for their children. Comparative studies of Eastern Europe and the Soviet Union have documented how political elites adopted social policies to shift access to education to different social groups and especially how they provided better access for their own children. Our ﬁndings for urban China showed that the relationship between social origins as measured by family class background and educational advancement was noticeable, but it also varied in response to changes in state policies. On the one hand, relative to other social groups, the bureaucratic class in China had advantages in providing access to education for their children; individuals with “middle class” or “exploiting class” family backgrounds were vulnerable to discriminatory state policies. On the other hand, bureaucratic privileges have at times been limited by state

Educational Stratiﬁcation

97

policies. Although having a father with high-rank cadre status had signiﬁcant, positive effects on one’s odds of entering college during the Cultural Revolution, college-level education was so severely limited then that the overall chances for young people from all social classes were tiny. Less than 3 percent of individuals in the risk set entered college during the Cultural Revolution era. During this period of shrinking opportunities, all social groups were adversely affected by state policies and the advantaged group (the high-rank cadre group) experienced a larger proportion of losses. Over time, we observed considerable variations in the relationship between social origins and educational advancement in urban China. Even though industrialization generated demand for the expansion of education, in urban China it has clearly been disrupted and shaped by political processes not always congruent with industrialization. The linkages between social origins and educational advancement also varied across historical periods because of the frequent shifting of selection policies between merit-based and political-based criteria. Finally, the ﬁndings provide strong evidence of increasing educational inequality in urban China in the most recent period (1978–1994), a time of rapid economic growth and momentous economic reform. For entry into high school and college, the effects of social origin, gender, and residential location are highly signiﬁcant and their magnitudes sizable, along with the remarkable expansion of educational opportunities. This era of burgeoning economic growth and transformation appears to favor the most advantaged groups in the population: the children of high-rank cadres and professionals, residents of large cities, and men relative to women. Because educational attainment is the earliest major event and major driving force in one’s status attainment, ﬂuctuations in educational opportunities and selection mechanisms are likely to have a lasting impact on one’s subsequent life course. And this is indeed the case, as will be shown in the areas of job attainment, promotions, and economic wellbeing in the following chapters. According to a report from Forbes Global magazine on October 26, 2001, among the top one hundred richest entrepreneurs in 2001 in China (the one hundreth one owns ﬁve hundred million RMB [Chinese yuan]), seventy-three of them experienced the disruption of education during the Cultural Revolution, and except for fourteen of them, all others received their college degrees after the Cultural Revolution (http://ﬁnance.sina.com.cn, October 26, 2001). The resilience of the Chinese youth in their life course is remarkable and the role of education in their careers will be examined in the subsequent chapters.

4 Entry Into the Labor Force: Patterns of First-Job Attainment

In Chapter 3, we examined the impacts of state policies on patterns of educational experience in urban China. As urban youth came to the end of their educational journey, delayed, rushed, interrupted, or otherwise completed, they faced the next critical stage in their life course: entry into the labor force. This is a critical stage because, until the 1990s, job mobility across work organizations in urban China was very low (see Chapter 9). For many employees, location in the ﬁrst workplace often meant their only, life-long work destination. Moreover, job mobility, if there was any, is likely to be path dependent on the previous job location. For example, it is easier to transfer from one state ﬁrm to another than from a nonstate ﬁrm to a state ﬁrm. As we discussed in Chapter 2, types of work organizations are associated with signiﬁcantly different economic beneﬁts, welfare programs, and career prospects. Therefore, entry into the labor force often had lasting impacts on subsequent events in one’s life course, and patterns of ﬁrst-job attainment shed light on stratiﬁcation processes in the distribution of life chances. This chapter centers on patterns of ﬁrst job attainment – one’s access to types of occupations and work organizations – and the allocative mechanisms underlying these processes. Following the general analytical strategy discussed earlier, we examine how opportunities and the association between social origins and ﬁrst-job attainment varied across historical periods as a function of shifting state policies. Before we look into the empirical patterns, it is useful to ﬁrst consider the concept of “job” in the Chinese context and its comparative signiﬁcance.

98

Entry Into the Labor Force: Patterns of First-Job Attainment

99

the concept of job in a comparative perspective In How Institutions Think, Mary Douglas (1986, p. 69) observed: “Institutions create shadowed places in which nothing can be seen and no questions asked. They make other areas show ﬁnely discriminated detail, which is closely scrutinized and ordered.” The distribution of attention in an institution, then, informs us of the ways in which that institution operates. The deliberation and struggles over the classiﬁcation of people and information are well documented in the social science research, from the American Census (Alonso and Starr 1987; Anderson 1988), to industries (Stone 1973), to occupations and organizations (Baron and Bielby 1986; Strang and Baron 1990). A telling observation can be made about the labor statistics in the China Statistical Yearbook, an annual ofﬁcial publication on major statistics compiled by the State Statistics Bureau (SSB). If one looks into the ofﬁcial statistics of the labor force and economic activities, one is surprised to ﬁnd that there is little information on the distribution of employees across occupational categories. For instance, in the China Statistical Yearbook, 1995 (SSB 1995), there is no information on the labor force classiﬁed by occupational groups, such as manager, professional, machine operator, and so forth. In contrast, there is detailed information on the distribution of the labor force across economic arenas/industries (e.g., heavy industry, transportation, communication, etc.), types of organizations in terms of their property rights relationships to the state (e.g., belonging to the state sector, the collective sector, and to “other” nonstate sectors) and administrative regions (e.g., by province and city, etc.). Even China’s Census only uses broadly deﬁned occupational categories. This is in sharp contrast with the detailed classiﬁcation of occupations in industrialized market societies. This pattern of statistics compilation tells us the ways the Chinese state “sees” and “thinks” about the labor force. These ofﬁcial categories highlight the ofﬁcial views about how economic activities are organized and carried out and the ways in which the central authority makes social and economic policies regarding the labor force. In brief, economic sectors and types of work organizations have occupied a prominent place in China’s planned economy, whereas occupations played a far less important role. The rationale is not difﬁcult to fathom: In the planned economy, the central authority organizes the production, sets economic priorities, and directs the ﬂow of resources mainly on the basis of economic sectors

100

The State and Life Chances in Urban China

and work organizations rather than through occupational categories. From the point of view of the central planner, information on the former is much more important and relevant than the latter. It is only logical, then, that such ofﬁcial constructions affected individuals’ job-seeking behaviors and, ultimately, the structure of the social stratiﬁcation system. Because the ofﬁcial construction of organizations and occupations was closely related to patterns of resource ﬂows across economic sectors and work organizations, such constructions also created a hierarchical order of jobs and workplaces, in terms of economic beneﬁts or political status and accordingly induced social behaviors in the pursuit of these opportunities and status. To analyze and understand patterns of job attainment in China, one needs to begin with an understanding of how the meaning of job is shared in this particular context, and how the ofﬁcial construction affects the processes that match individuals and positions in the social stratiﬁcation structure organized around work organizations. But let us begin on familiar ground – occupational status. Occupational Status In industrialized market societies, the concept of “job” usually refers to one’s occupational status. Occupation is associated with social status, prestige, economic beneﬁts, and different mobility patterns. The kind of job one does is the single, most important indicator of one’s socioeconomic status in the social stratiﬁcation structure. Theoretical models of life chances are developed largely in this light. For example, one interpretation of the role of job focuses on the supply side of the labor market. Human capital theory (Becker 1964) sees skills, like any other form of capital, as allocated through market processes to realize their highest value possible. Individual-based skills acquired through formal education and on-the-job training are the most important human assets that channel an individual into certain occupational careers, which in turn determine one’s economic rewards and social status. In this view, the importance of job is driven by market transactions that reward the most valuable human capital. What one does (type of job) is seen as the ultimately decisive factor in one’s market position, hence his or her location in the social stratiﬁcation structure. For a long time, this perspective also inﬂuenced the study of social stratiﬁcation processes in market societies. Consider the inﬂuential statusattainment model in the stratiﬁcation literature (Blau and Duncan 1967). The basic structure of the model – the effects of parental educational and

Entry Into the Labor Force: Patterns of First-Job Attainment 101 occupational status on children’s education and subsequent occupational status – focuses on family-based resources and individual attributes associated with “marketable skills.” In this model, socioeconomic status is typically measured as a function of income and education, both of which are closely related to one’s occupational status in conceptualization as well as in practice. Moreover, the entire literature on intergenerational mobility is based on occupational categories, conceptualized one way or another, as an approximation of status groups or social classes (Erikson and Goldthorpe 1992; Featherman and Hauser 1978). Overall, jobs are deﬁned largely by occupation, and labor forces are segmented along this dimension in industrial market societies (e.g., Althauser 1989). Until recently, occupational categories in China were of limited importance in one’s entry into the labor force for two reasons: First, state policies based on occupational groups were deﬁned very broadly (e.g., workers, cadres, and professionals); as a result, status differentiation was less important among speciﬁc occupations than in market societies. Second, within a workplace, boundaries across occupations (jobs) were relatively low, making shifts across occupations (e.g., from a cadre to a professional, from a clerk to a cadre) easy relative to market societies where occupational barriers cultivated by organized interests were much higher. For these reasons, the occupational categories used in this study are broadly deﬁned rather than ﬁne-grained, compared with studies in other societal contexts. But this does not mean that occupational groups are meaningless in the Chinese context. The broadly deﬁned occupations (e.g., workers, cadres, and professionals) were also related to the redistribution and state policies in important ways. Bureaucrats promoted from the broad basis of cadres were the major players in the redistributive economy who controlled the resource allocation, and professionals were the crucial force in production. State wage policies, although often varied with economic sectors and industries, were partly established on these broad occupational categories. For instance, cadre and professional salary grades promulgated by the central government took into consideration both type of job (occupation) and economic sector (Chen and Zhang 1992), with distinctive career lines for cadres and professionals (Walder 1992). More important, state policies and political processes often make use of these occupational categories to discriminate among social groups. For example, in the Mao era, the working class, especially those in manufacturing industries (chanye gongren), was proclaimed as the “leading class”; as a result, state policies favored children of this group in educational attainment and job

102

The State and Life Chances in Urban China

opportunities. In contrast, those in “white-collar” occupations – those in education and literary sectors – were frequent targets of political purge and discrimination. The Hierarchy of Work Organizations In industrialized societies, the importance of work organizations in social stratiﬁcation processes has gained recognition in sociological and economic studies of labor markets since the 1980s. These studies have shown that the allocation of job opportunities and economic beneﬁts varies systematically with those ﬁrms in the core industries versus periphery industries, and with those in the primary labor markets versus those in the secondary labor markets (Althauser and Kalleberg 1981; Doeringer and Piore 1971; Hodson and Kaufman 1982). Formal organizations also play a critical role in social stratiﬁcation processes (Baron 1984; Rosenbaum 1984; Sørensen and Kalleberg 1981). In this light, a model of life chances that ignores the role of work organizations would have grossly misspeciﬁed or overlooked critical mechanisms of social stratiﬁcation. In the history of the People’s Republic of China, the role of work organizations has attained a prominent role that is unprecedented historically and unrivaled in other societies, partly due to the importance of such organizations in resource allocation and partly due to restrictions of job mobility across organizational boundaries in the Chinese economy. We have summarized the theoretical discussion of the organizational hierarchy in China in Chapter 1 and outlined the types of organizations in urban China in Chapter 2. In brief, work organizations in China have been organized into a hierarchical structure with respect to their property-rights relationships to the state (Bian 1994; Lin and Bian 1991; Walder 1992; Zhou, Tuma, and Moen 1997). A central characteristic of the redistributive state socialism is that a large proportion of economic beneﬁts are redistributed through work organizations rather than in marketplaces. Those organizations that are closer to the redistributive authority (e.g., government agencies and state-owned ﬁrms) have better access to resources, enjoy higher status, and offer better economic beneﬁts. In contrast, collective ﬁrms and especially private ﬁrms are excluded from the central planning system and hence have no redistributive beneﬁts. Therefore, locations in types of organizations are important indicators of one’s socioeconomic status and patterns of access to organizational hierarchies reﬂect critical stratiﬁcation mechanisms.

Entry Into the Labor Force: Patterns of First-Job Attainment 103 The dual structure of workplace and occupation and their relative importance is also consistent with the everyday experience of urban youth. In the Mao era, graduates from middle school or high school were allocated into jobs by the labor bureau of the local government (usually a county or municipal government) that governed their residential areas. Most prospective job candidates were assigned to different types of work organizations under the broad umbrella of “workers” with little speciﬁcation of particular jobs (or occupations). Authorities in the work organization then decided on the particular jobs and career tracks of the new employees, giving rise to the importance of local authorities on individual life chances. The main exception is for the group of college graduates. According to the governmental personnel policy, college graduates entered the ofﬁcial administrative system as cadres or professionals (broadly called “state cadres” [guojia ganbu] for both types of jobs), regardless of which work organizations they were allocated into. Finally, our emphasis on the political dynamics of social stratiﬁcation under state socialism also points to the systematic variations in state policies that allocate individuals into different types of jobs, in both occupations and types of work organizations. Types of work organizations and occupations are hierarchies constructed by the state; as a result, the meanings and status of jobs are sensitive to changes in state policies that direct the allocation of resources and of life chances across these jobs. State policies may directly affect the allocation of individuals into different types of work organizations and/or occupations; or, state policy shifts may indirectly lead to the transfer of resources across economic sectors and generate varying opportunities across types of organizations and occupations. For instance, the emphasis on “heavy industry” as the priority of economic development in the Mao era channeled disproportional resources into the manufacturing sectors, where large state-owned ﬁrms concentrated. Employees in these types of work organizations enjoyed better welfare programs and other fringe beneﬁts. In contrast, the neglect of consumer goods in the Mao era left those ﬁrms in the service sector (a disproportionally large number of collective ﬁrms) with meager resources for their employees.

descriptive patterns of entry into the labor force We ﬁrst report the descriptive patterns of ﬁrst-job attainment over time, as reﬂected in the hazard rate of job entries into different types of occupations and work organizations. The hazard rate measures the annual

The State and Life Chances in Urban China

104

instantaneous rate of entering a speciﬁc type of job for those in the risk set (i.e., individuals who graduated from their highest educational levels). These patterns provide an overview of the evolution of the labor force in urban China, as well as changes in social/economic institutions over time. As we noted earlier, until recently, most ﬁrst jobs in urban China were assigned by governmental agencies. Therefore, patterns of job entries were also closely associated with shifts in state policies over time.

Entry into First Occupation To describe the overall pattern of ﬁrst-job attainment over time, we ﬁrst report the estimated hazard rates of entry into different types of work organizations for the entire risk set (i.e., everyone who is available for entry into the labor force in a given year). Figure 4.1 reports the hazard rates of entry into three major occupations between 1949 and 1992. Some occupations are omitted due to their small number of cases. The ﬂuctuations in Figure 4.1 are partly characteristic of hazard rates in modeling event history processes, resulting from sampling variability and evolving composition of the risk set over time; but they also

0.5

hazard rate

0.4

0.3

0.2

0.1

0

49

55

61

Cadre

67

73 year Professional

79

85

Worker

ﬁgure 4.1. Hazard rates of entry into ﬁrst occupations, 1949–1992.

91

Entry Into the Labor Force: Patterns of First-Job Attainment 105 reﬂect the substantive effects of shifting state policies on individual life chances. Consider the rate of becoming a worker over time. The sharp increase in the rate in the mid- to late-1950s reﬂected the period of nationalization and collectivization when a large number of self-employed were transferred from the private sector to become “workers” in state enterprises; at the same time, the government also recruited millions of rural laborers into the urban labor force to pursue rushed growth. The subsequent decline in the early 1960s marked a period of the economic contraction in which the government reduced the urban labor force by returning twentysix million migrant urban residents to rural areas. The low rate and ﬂuctuations between the mid-1960s and 1979 captured the impacts of the Cultural Revolution, when a large number of graduates from secondary schools were sent to the rural areas (see Figure 4.2b). The rise in the rate in the late 1970s mainly resulted from the return of the sent-down youth to the urban areas as well as from the absorption of unemployed youth cumulated in urban areas. Since the late 1980s, it appears that the rate of becoming workers was declining, indicating a more diverse labor force. Trends of cadres and professionals appear to be similar, reﬂecting the fact that both occupations were under similar regulations by the Chinese bureaucracy. There were relatively low and stable rates of entering these two occupations in the Mao era, but the trends increased in the reform era. The rate of becoming a professional increased substantially since the mid-1980s, reﬂecting the expansion of higher education and the evolving composition of the labor force in urban China. Entry into First Organization Fluctuations are more salient and systematic in entry into types of work organizations, because, as we noted earlier, entries into organizations are more sensitive to state policies and political processes. Figure 4.2a displays the hazard rates of ﬁrst entry into different types of work organizations in the state and collective sectors. It appears that a large proportion of labor force entrants were employed in state ﬁrms. As one can tell, the hazard rate for the state ﬁrms resembles that for “workers” closely (cf. Table 4.1), because most workers were employed in state ﬁrms. Moreover, it is clear from Figure 4.2a that the ﬂuctuations for all four types of work organizations tract one another over time, suggesting that they all respond to similar macro-environmental changes – shifts in state policies and political processes. Overall, there were two waves of the expansion of

The State and Life Chances in Urban China

106

hazard rate

0.3

0.2

0.1

0 49

55

61

67

73

79

85

91

year Government

Public

State firm

Collective firm

ﬁgure 4.2a. Hazard rates of entry into ﬁrst organizations, the state, and collective sectors (1949–1992).

hazard rate

0.3

0.2

0.1

0 49

55

61

67

73 year

Hybrid firm

79

85

91

Farm

ﬁgure 4.2b. Hazard rates of entry into ﬁrst organizations, hybrid ﬁrm, and farm 1949–1992.

Entry Into the Labor Force: Patterns of First-Job Attainment 107 the urban labor force. The ﬁrst was in the second half of the 1950s, when nationalization, collectivization, and rushed-growth-oriented industrialization took place. The second wave was in the 1980s, when the state sector was expanded to absorb urban unemployed and the return of the sent-down youth in the aftermath of the Cultural Revolution. The rates of entry into government and public organizations appear to be similar and increased noticeably since the 1980s. The rate of entry into the collective ﬁrms, especially since the 1980s, indicates the signiﬁcant presence of the collective sector in the urban labor force. Figure 4.2b shows the hazard rates of entry into hybrid ﬁrms and farms. The rate of entry into the rural labor force was low before the 1960s. In most cases, it reﬂected a small portion of our sample that grew up and began their work in rural areas but later joined the urban labor force (through college education, job hiring, or as the result of urbanization). But the hazard rate also reﬂected the impacts of state policies: It rose abruptly in the early 1960s, capturing the state efforts to reduce the urban population by returning some migrants back to rural areas. The rate of becoming a farmer reached its peak in 1968, when the send-down policy was ﬁrst implemented, which forced seventeen million urban youth to rural areas (see Chapter 5). The rate dropped sharply in late 1970s, when the send-down policy was ofﬁcially abandoned. In contrast, the rate of entry into hybrid ﬁrms was low in the Mao era, reﬂecting the small size of the nonstate sector. Since the 1980s, the rate of entering hybrid ﬁrms increased noticeably and coincided with the expansion of the private sector in the reform era. To summarize the patterns of ﬁrst-job attainment, variations in the hazard rates have captured marked ﬂuctuations in the urban labor force that did not reﬂect a trend of gradual expansion of the labor force along with industrialization; rather, they were consistent with the documented shifts in state policies and political interventions across these periods.

explaining entry into ﬁrst jobs: the overall pattern Given the distinctive types of occupations and work organizations, and the considerable variations of opportunities over time, a natural question is: What determines the destination of one’s entry into the labor force? Again, the central issue is the link between state policies and individual life chances. To explore this link, we adopt a strategy similar to that in the last chapter. That is, we examine the association between social origins and job destination and how this association varied across historical periods

108

The State and Life Chances in Urban China

as a function of shifts in state policies. We include both spouses in the analyses to increase the statistical power in the analyses of speciﬁc destinations. As a ﬁrst step, we begin with an analysis of the overall pattern, regardless of historical periods, to establish the baseline for comparison with further analyses that focus on period-speciﬁc patterns. Overall Patterns of First-Job Attainment Do social origins affect patterns of entry into the labor force? To address this question, we conduct statistical analyses of the association between social origins and ﬁrst-job locations. In this set of analyses, we do not take into consideration the effects of historical periods in the entire forty-ﬁveyear history, from 1949 to 1994. Technically, one can interpret the statistical models adopted here as assuming (or constraining) the covariates of theoretical interest (e.g., social origins) to be the same across historical periods. We include a separate intercept for each of the three historical periods (see below), thus allowing the overall level to vary across the periods to capture possible structural changes over time. This set of analyses provides the baseline for us to compare and assess the impacts of varying state policies in period-speciﬁc models reported later. To set up the statistical analyses, we adopt the same logistic model of discrete-time event history used in the analyses of educational attainment in Chapter 3, which allows us to take into consideration changing compositions of the risk set and time-varying covariates. Similar to our analyses of educational attainment, we use the father’s education and the father’s occupational status to measure social origins. In addition, we also include other variables such as gender, age, and residential location for the purpose of statistical control. Table 4.1 reports the parameter estimates of the logistic model for the overall pattern of entry into the ﬁrst occupation. Here the reference category for this set of analyses is “worker” – the main job destination for urban youth. The coefﬁcients in the table refer to the effects of the corresponding variables on the “log-rate” (or log-odds) of entering the occupation in that column, relative to the probability of entering the reference category of worker. To facilitate interpretation, we may treat these occupational categories as reﬂecting some hierarchical order descending from cadre, professional, worker, to farmer. Thus, a move to cadre or professional occupations can be seen as upward mobility relative to the majority of those entering the worker occupation, whereas a move to the “farmer” occupation is downward mobility.

Entry Into the Labor Force: Patterns of First-Job Attainment 109 table 4.1. Parameter Estimates of Logistic Model for Entry Into First Occupation (Reference Category: Worker) Cadre intercept 1949–1965 1966–1979 1980–1994 Age Female education Middle school High school College father’s education Middle school High school College father’s occupation High-rank cadre Low-rank cadre High-rank prof. Low-rank prof. Farmer χ2 d.f. N of events

Professional

Farmer

−4.158∗∗ −5.069∗∗ −5.153∗∗ 0.045∗ −0.236∗

−4.490∗∗ −5.468∗∗ −5.842∗∗ 0.053∗∗ 0.310∗∗

1.748∗∗ 2.881∗∗ 0.085 −0.242∗∗ 0.372

0.761∗∗ 1.592∗∗ 3.924∗∗

1.182∗∗ 2.848∗∗ 5.514∗∗

0.249∗∗ 0.425∗∗ 1.205∗∗

0.455∗∗ 0.657∗∗ 0.471†

0.153 0.262 0.549∗∗

0.557† 0.431∗ −0.189 0.129 0.612∗∗ 3338.2 27 533

0.391 0.401∗∗ 0.657∗ 1.143∗∗ 0.899∗∗ 3274.6 27 1234

−0.165 0.009 0.352 0.222 −0.191 0.371 0.351† 1.267∗∗ 2657.5 27 995

∗∗

p < 0.01 p < 0.05 † p < 0.10 Note: Elementary or no education is the reference category for the respondent’s and father’s education; worker, for the father’s occupation. All models also include indicator variables for cases where the father’s occupation and education information is missing. ∗

The intercepts in Table 4.1 capture the rate of the reference group of males with low education in large cities whose fathers are workers with low education. For this group, the intercepts show that, not surprisingly, there have been low rates of becoming a cadre or a professional, compared with becoming a worker. In the ﬁrst two periods, there were higher rates of becoming a farmer, mainly due to state policies that sent millions of urban youth to the rural areas in the Mao era (see Chapter 5). These variations notwithstanding, if we focus on the columns for the high-status cadre and professional occupations, Table 4.1 portrays a picture that is not much different from the conventional model of social

110

The State and Life Chances in Urban China

mobility in other societies. First, education played a major role in entering the two high-status occupations. Second, parental status, as indicated by father’s education and occupational status, also contributed to children’s initial location in the labor force. College-educated fathers increased the probability of their children becoming a professional; this was not as important as becoming a cadre. There was also evidence of status inheritance, in that children of cadres were more likely to become cadres and those of professionals were more likely to become professionals. But an anomaly appeared when we considered the “farmer” column. Here, higher educational credentials increased the rate of downward mobility and advantageous family backgrounds did not help prevent such a slide at all.1 We turn now to the overall patterns of entry into types of work organizations (see Table 4.2). Here the reference category is the events of “entry into collective ﬁrms.” In general, government, public organizations, and state ﬁrms belong to the state sector and are high-status organizations, more or less in that descending order. Collective ﬁrms and hybrid ﬁrms were low-status ﬁrms in the nonstate sector, especially in the Mao era. We observed patterns similar to those reported in the analyses of entry into occupations. First, education increased the rate of entering highstatus organizations in the state sector. Second, in terms of status inheritance, it seems that the father’s education facilitated the respondents’ entry into public organizations. Children from educated families were more likely to enter hybrid ﬁrms, compared with those entering collective ﬁrms (the reference category). But, as we will see later, this was mainly due to political discrimination in the Cultural Revolution period. Highrank cadres were successful in placing their children into the state sector, compared with those fathers who were workers. Low-rank professionals appeared to have some advantages as well. That there were parallel patterns between attainment in types of occupations and of work organizations is not surprising, because occupation and organization are strongly, though imperfectly, associated. Cadres were especially concentrated in government agencies; professionals in public and organizations; workers in state or collective ﬁrms. These 1

As we noted earlier, the indicator variable for the “father’s farmer origin” captures sampling design effect and is used here for statistical control. Because farmer occupation and farm as type of work organizations are identical, we only report one set of analyses. Note that, because of some ambiguity in job boundaries (for instance, some respondents regarded working on state farms as a “worker” occupation), there is some discrepancy between the coding of farmer as an occupation and farm as a type of work organization.

Entry Into the Labor Force: Patterns of First-Job Attainment 111 table 4.2. Parameter Estimates of Logistic Model for Entry Into First Organization (Reference Category: Collective Firm)

Government intercept 1949–1965 1966–1979 1980–1994 Age Female education Middle school High school College father’s education Middle school High school College father’s occupation High-rank cadre Low-rank cadre High-rank prof. Low-rank prof. Farmer χ2 d.f. N of events

Public Organization

State Firm

1.082∗ 0.643 0.617 0.072∗∗ −1.513∗∗

−0.988† −2.019∗∗ −1.819∗∗ 0.029 0.261∗

2.713∗∗ 2.367∗∗ 2.420∗∗ −0.063∗∗ −0.505∗∗

0.330∗ 0.879∗∗ 2.278∗∗

0.961∗∗ 2.220∗∗ 4.031∗∗

0.164 0.480∗∗ 1.193∗∗

0.227 0.273 0.266

0.161 0.559∗ 0.634∗

0.007 0.279† 0.011

1.239∗∗ 0.293 0.981∗ 0.886∗∗ 1.260∗∗ 826.0 27 736

0.718† 0.266 0.513 0.912∗∗ 0.522∗∗ 1,118.4 27 654

0.754∗∗ −0.136 0.058 0.528∗∗ −0.058 1,163.3 27 2,901

Hybrid Firm −0.381 −0.381 0.124 −0.021 −0.452 −0.257 −0.412∗ −0.030 0.415∗ 0.518∗ 0.736∗ 0.327 −0.244 −0.269 0.141 0.155 740.4 27 463

∗∗

p < 0.01 p < 0.05 † p < 0.10 Note: Elementary or no education is the reference category for the respondent’s and father’s education, worker for the father’s occupation. All models also include indicator variables for cases where the father’s occupation and education information is missing. ∗

ﬁndings revealed some interesting and familiar patterns of job attainment and provided a baseline model for us to assess the importance of attention to stratiﬁcation dynamics – changes over time in allocation criteria that result from shifting state policies, the focus of our next set of analyses.

period-speciﬁc job-attainment patterns To consider changes in allocation mechanisms over time, we now examine the patterns of entry into ﬁrst occupation and ﬁrst organization across speciﬁc historical periods (1949–1965, 1966–1979, and 1980–1994). If

112

The State and Life Chances in Urban China

our argument about shifting state policies over time is valid, we should observe that the effects of education and social origins on one’s labor force locations vary across these periods that were associated with distinct state policies and political processes. Entry Into First Occupation Table 4.3 gives the estimates of the effects of covariates on entry into ﬁrst occupation. It is useful to ﬁrst assess model ﬁt in this set of analyses. Although the likelihood ratio statistics across these models are not directly comparable, the large reduction in the period-speciﬁc models, compared with those in Table 4.1, suggest that the period-speciﬁc models greatly improve model ﬁt. In substantive terms, this means that the criteria used to allocate labor market entrants to their ﬁrst occupation do vary across the three historical periods. Because the patterns reported in Table 4.3 were generated by and reﬂect major stratiﬁcation processes in China’s state socialist practice, we inspect these results in some detail. The reference category is entry into the worker occupation. Entry Into the Cadre Occupation There was no signiﬁcant gender difference in becoming a cadre or a worker (the reference category) across these three periods. We suspect that this pattern resulted partly from the egalitarian policies and partly from gendered job lines in the Chinese economy where women tended to be concentrated in nonmanual (e.g., clerical or technician) jobs. The impacts of political dynamics were reﬂected in the educational effects. In the second period, education played a noticeably smaller role in entering the cadre occupation. The effects of a college or high school education were large and signiﬁcant in the 1980–1994 period, indicating the increasing importance of educational credentials in the post-Mao era. The ﬁnding that education still played an important role in entering the cadre profession during the Cultural Revolution should be interpreted with caution. In China’s bureaucratic system, especially in the Mao era, all college graduates were by default allocated to cadre jobs, but these jobs were not necessarily associated with redistributive power. There were no systematic effects of father’s educational or occupational status in the Mao era (the ﬁrst two periods), indicating that the link between social origins and job attainments, controlling for education, was weak. In the post-Mao era, children of those with secondary education and of the low-rank cadres had a signiﬁcant and higher rate of becoming

113

0.498† 0.896∗∗ 0.648

0.498† 0.473 −0.127

0.624 0.079 −0.383 0.094 0.540† 89.2 24 139

0.376 0.140 0.314

0.471a

0.709∗∗ 198.5 22 197

0.154a

0.377 1.923† 4.355∗∗

0.687† 1.180∗∗ 2.973∗∗

0.808∗∗ 1.767∗∗ 4.028∗∗

0.533 0.637∗ −0.248 0.231 0.800† 334.2 24 197

−5.370∗∗ 0.054 −0.187

−5.156∗∗ 0.067∗ −0.292

−4.005∗∗ 0.018 −0.240

0.755∗∗ 886.6 22 486

0.774∗,a

0.593a

0.088 0.493 −0.141

1.199∗∗ 3.644∗∗ 6.338∗∗

−3.756∗∗ 0.008 0.413∗∗

1949–1965

0.717† 0.458† 0.764 1.227∗∗ 1.146∗∗ 330.9 24 307

0.069 −0.506 0.563†

0.447 1.687∗∗ 3.808∗∗

−4.637∗∗ 0.063 0.337∗

1966–1979

Professional

0.024 0.400† 0.563 1.382∗∗ 1.141∗∗ 731.9 24 441

0.410† 0.685∗∗ 0.916∗∗

−0.249 1.533∗ 4.376∗∗

−5.210∗∗ 0.074∗ 0.184

1980–1994

1.725∗∗ 301.41 22 236

0.348a

0.903†,a

−0.025 0.222 −1.473

0.015 0.187 2.178∗∗

2.121∗∗ −0.303∗∗ 0.827∗∗

1949–1965

0.128 −0.231 0.386 0.318 0.783∗∗ 213.4 24 728

−0.150 0.049 0.491∗

0.626∗∗ 0.683∗∗ 1.223∗∗

2.209∗∗ −0.209∗∗ 0.191∗

1966–1979

Farmer

∗

p < 0.01 p < 0.05 † p < 0.10 a Because of the fewer number of cases in the ﬁrst period, we combined high-rank cadre and low-rank cadre into one category of cadre, and high-rank professional and low-rank professional into one category of professional. Note: Elementary school or below is the reference category for father’s education. Worker is the reference category for the father’s occupation. All model estimations also include a set of dummy variables indicating respondents’ province and city size, and for missing values for the father’s education and occupation.

∗∗

Intercept Age Female education Middle school High school College father’s education Middle school High School College father’s occupation High-rank cadre1 Low-rank cadre High-rank prof. Low-rank prof. Farmer χ2 d.f. N of events

1980–1994

1966–1979

1949–1965

Cadre

table 4.3. Parameter Estimates of Logistic Model for Entry Into First Occupation, by Period (Reference Category: Worker)

114

The State and Life Chances in Urban China

a cadre. But the overall evidence of intergenerational status inheritance with respect to this important job position is weak at best. The general impression of patterns of entry into the cadre occupation is that, during the 1980–1994 period, the most important criterion for selection into this occupation was having a higher education. This pattern is consistent with the new cadre policies in the post-Mao era that emphasize educational qualiﬁcations. Entry Into the Professional Occupation Females had a higher rate of entering professional jobs in the ﬁrst two periods, compared with becoming a worker. As we noted previously, the “professional” occupations included semiprofessional jobs (e.g., nurse, school teacher, technician), and female employees tend to concentrate in these jobs. It is interesting to note that, as the labor force became more professionalized in the reform era, females lost their advantages. Education was very important for entry into a professional occupation in China, as in industrialized market societies. Table 4.3 shows that having a higher educational level (especially with a college and to a less extent a high school education) signiﬁcantly increased the rate of becoming a professional in all three periods. The effects of high school and college education increased after the Cultural Revolution period, but did not have the level of effects as they had before the Cultural Revolution. This result probably reﬂected the changing educational composition: Over time, the pool of highly educated youths became much larger, thus dampening the effect of educational levels. The smaller effect of education in the second period captured the disruption of the Cultural Revolution, when educational credentials were signiﬁcantly downplayed. Children of cadres and of low-rank professionals had a signiﬁcant and higher rate of becoming professionals in the Cultural Revolution period, when such opportunities were especially scarce – compared with millions of urban youth who had to work in rural areas. In the reform era, children of educated families had a signiﬁcant, higher rate of entering professional occupations. This was the strongest evidence of status inheritance. Entry Into the Farmer Occupation Note that, because there are few entries into “farmer” jobs in the reform era, we only report the ﬁndings for the ﬁrst two periods. In the ﬁrst period, there were some movements from urban areas to rural areas, especially in the early 1960s. Table 4.3 shows that college graduates were more likely

Entry Into the Labor Force: Patterns of First-Job Attainment 115 to experience downward mobility to rural areas, which was related to the political purge of “old” intellectuals and the sent-down policy during the Cultural Revolution. Children of high-rank cadres also had a marginally signiﬁcant and higher rate of becoming a farmer in the ﬁrst period. One speculation is that, during the earlier period, many cadres had families in the rural areas. Their children often started working in the rural areas and moved to urban areas later. During the Cultural Revolution, becoming a farmer resulted mainly from the send-down policies. Here we ﬁnd the most striking evidence of the inconsistent effects of education on individual life chances. Every educational level has a signiﬁcant and high rate of becoming a farmer than the reference category – an elementary or no education. The send-down policy especially affected the life chances of youth with secondary and college education. Under radical state policies, some college graduates have been sent to rural areas upon their graduation, and this may have led to a signiﬁcant, higher rate for this group in both periods. In the Cultural Revolution, children of educated families – those with a college educated father – had a signiﬁcant, higher rate of being sent to rural areas compared with the children from low-educational families. Other than this, there were no statistically discernible differences among different social origins. Children of cadres, professionals as well as workers, all had similar fates. These results showed the strong political pressures of state policies in this period, which signiﬁcantly suppressed the effects of the intergenerational status inheritance. We will examine this episode in greater detail in the next chapter. Education played an important role in allocating individuals into different types of jobs. The effects of education also varied over time in a nonlinear way. Figure 4.3 displays the log-odds for college graduates, as compared with those with elementary or lower education. As one can see, the effects of education on entering cadre and professional occupations were noticeably lower in the Cultural Revolution period. In the reform era, college education had similar effects on becoming either a cadre or a professional. Regarding the signiﬁcant positive effects of the father’s farmer occupation on becoming cadre and professional across periods, we should keep in mind that the sample only included urban residents, and no farmer at the time of interview was included. A large proportion of those who moved into cities with a rural farmer origin in the early period were revolutionary cadres and military ofﬁcers. Others came to cities through their

116

The State and Life Chances in Urban China

8

log-odds

6

4

2

0 1949-1965

1966-1979 period cadre

professional

1980-1994

farmer

ﬁgure 4.3. Rates of entering types of occupations for college graduates, by period.

college education. We suspect this was the main reason to explain their advantages in becoming cadre and professional, compared with those in the worker family (the reference category). Entry Into First Organization We now examine patterns of entry into the ﬁrst organization, considering especially how the criteria used to allocate urban youth to their ﬁrst work organization varied across the three historical periods. Again, we use “entry into collective ﬁrm” as the reference category in this set of analyses. The ﬁndings are reported in Table 4.4. As in the case of entry into ﬁrst occupation, the period-speciﬁc models in Table 4.4 greatly improve the model ﬁt, as indicated by log-likelihood statistic. Hence, there is strong evidence that the criteria used to allocate youths to their ﬁrst organizations varied across these three historical periods. We summarize the main ﬁndings as follows: First, education played an important role in entering high-status organizations – government agencies, public organizations, and state ﬁrms. Figure 4.4 shows that college education and to a lesser extent middle school and high school had signiﬁcant, positive effects on entering high-status organizations, compared with those with only elementary

117

Intercept Age Female education Middle school High school College father’s education Middle school High school College father’s occupation High-rank cadre Low-rank cadre High-rank prof. Low-rank prof. Farmer χ2 d.f. N of events

0.992∗∗,† 242.0 22 309

0.245a

0.371a

0.492† 0.480 0.228 0.574 0.788∗ 1.523∗ 1.526∗∗ 1.582∗∗ 220.4 24 197

2.086∗∗ −0.155 −0.035 0.822† 1.269∗∗ 198.7 24 230

−0.375 0.614 2.175∗∗

0.253 −0.065 −1.222∗∗

1980–1994

0.315 0.043 −0.098

0.623∗∗ 0.907∗∗ 2.181∗∗

−0.225 0.034 0.853

0.317 0.868∗∗ 1.839∗∗

−1.755∗∗

1966–1979 0.915 −0.087∗ −1.720∗∗

−0.061∗

1.439∗

1949–1965

Government

0.208 371.4 22 259

0.203a

−0.080a

0.115 0.182 0.642

0.849∗ 2.726∗∗ 4.076∗∗

0.765 −0.075† −0.348

1949–1965

2.087∗ −0.295 0.617 1.085∗ 0.255 165.7 24 151

0.304 0.215 0.105

0.936† 2.213∗∗ 3.158∗∗

0.008 −0.274

−2.534∗∗

1966–1979

Public Organization

(continued)

0.155 0.772∗ 0.789 1.100∗ 1.354∗∗ 343.6 24 244

0.146 1.064∗∗ 0.893†

0.984 1.979† 4.403∗∗

−2.665† −0.005 −0.428†

1980–1994

table 4.4. Parameter Estimates of Logistic Model for Entry Into First Organization, by Period (Reference Category: Collective Firm)

118 0.119 −0.055 0.438 1.049∗∗ −0.294 256.7 24 872

1.434∗ −0.237 −0.012 0.632 −0.014 246.5 24 1100

−0.155a

−0.022 163.9 22 929

0.089a

0.121 0.595∗ 0.027

0.495 1.089∗ 1.895∗∗

−0.050† −0.598∗∗

−0.084 −0.138 −0.080

0.407∗ 0.642∗∗ 1.072∗∗

−0.094∗∗ −0.468∗∗

1.569∗

1980–1994

−0.060 0.357 0.148

0.237∗∗ 0.190∗∗ 0.975∗∗

−0.039† −0.509∗∗

2.767∗∗

2.344∗∗

0.092 106.1 22 128

−0.295a

0.565a

−0.486 0.096 0.313

−0.259 −0.244 0.531

0.922 −0.085∗ −0.777∗∗

1949–1965

1.526 −0.654† −1.311 −0.087 −0.126 147.4 24 149

0.862∗∗ 0.572 1.625∗∗

0.384 −0.205 −0.865

0.012 −0.547∗∗

−1.492†

1966–1979

Hybrid Firm

−0.158 −0.075 0.143 0.446 0.555 103.5 24 186

0.506† 0.741∗ 0.408

−1.322∗∗ −1.316∗∗ −1.144∗

0.474 0.009 −0.287

1980–1994

∗

p < 0.01 p < 0.05 † p < 0.10 a Because of the fewer number of cases in the ﬁrst period, we combined high-rank cadre and low-rank cadre into one category of cadre, and high-rank professional and low-rank professional into one category of professional. Note: Elementary school education or below is the reference category for respondents’ and father’s education. Worker is the reference category for the father’s occupation. All model estimations also include a set of dummy variables indicating respondents’ province and city size.

∗∗

Intercept Age Female education Middle school High school College father’s education Middle school High school College father’s occupation High-rank cadre Low-rank cadre High-rank prof. Low-rank prof. Farmer χ2 d.f. N of events

1966–1979

1949–1965

State Firm

table 4.4 (continued)

Entry Into the Labor Force: Patterns of First-Job Attainment 119 5 4.5 4

log-odds

3.5 3 2.5 2 1.5 1 0.5 0 1949-1965

1966-1979 period government

public org'n

1980-1994

state firm

ﬁgure 4.4. Rates of entering types of organizations for college graduates, by period.

education. But the role of education was noticeably lower in the second period, reﬂecting the downplay of education by the radical state policies at the time. The negative effect of education on entering hybrid ﬁrms in the reform era suggests that this type of ﬁrm, on average, had a lower status than the reference category of collective ﬁrms. Second, the father’s education had no systematic effects on entering the types of organization throughout these historical periods. The strongest evidence of social origins is that, compared with children of the working class, children of high-rank cadres still had a higher rate of entering government, public organizations, or state ﬁrms in the Cultural Revolution period, suggesting that some bureaucrats did have advantages in buffering the negative effects of political processes. An Alternative Speciﬁcation of Social Origins The previous analyses show somewhat surprisingly that political selection processes have not shown the kind of strong, nonlinear patterns we had anticipated. We suspect that part of the reason is that there are many “noises” in the model speciﬁcation – there are many processes involved

120

The State and Life Chances in Urban China

in the previous models, some overlapping each other. For example, the cadre’s positional advantages may be overshadowed by adverse effects of antibureaucratic campaigns. To take another look at this issue, we now focus on one dimension – the family class background labels used in political selection in the Mao era. As we noted in the analyses of educational attainment, class labels played a critical role in the allocation of opportunities in the Mao era. This was also the case for allocating job opportunities. Among the four categories – worker/peasant (the reference category), cadre, middle class, and exploiting class background – state policies favored cadre class background and discriminated against the last two groups, and these policies were most effective during the Cultural Revolution. Table 4.5 reports the determinants of ﬁrst-job attainment, using family class background as indicators. Now the effects of political selection become salient. For the two occupational categories, the ﬁrst period witnessed no discernible political selection: There was no signiﬁcant difference among these class backgrounds in entering the cadre occupation, and all had a higher rate in entering professional occupations than the reference category of worker/peasant background in the Mao era. This suggests that professional occupations were less subject to the political screening. But political selection emerged in the Cultural Revolution when cadre background had advantages in both occupations and children of the exploiting class lost their advantages. Patterns of entering work organizations showed discriminatory practice in government agencies in the ﬁrst period, when children from the cadre background had a higher rate of entering government and children of the exploiting class had a lower rate. In this period, children of the middle class and exploiting class had a higher rate of entering public organizations, mainly due to their educational qualiﬁcations. But in the Cultural Revolution, their advantages disappeared and the children of the cadre background gained advantages. Children of the exploiting class also had a higher rate of going to rural areas in this period. In sum, there were noticeable variations with class backgrounds in entries into ﬁrst jobs. Children of the exploiting class were systematically discriminated against – they were denied some high-status jobs and organizations, whereas children of the “cadre” background were favored. These discriminatory policies were most effective during the Cultural Revolution. But life chances of the children of the middle class were not statistically different from those of the worker/peasant group.

121

0.072 0.377† 0.476

1949–1965

−0.589 0.066 0.031

1949–1965

1.182∗ 0.149 −1.314∗

0.608† 0.328 —

1966–1979

Cadre

0.292 0.050 −0.203

1966–1979

State Firm

−0.155 0.244 0.289

1966–1979

0.874∗∗ 0.324 1.578∗

1980–1994

0.615∗∗ 0.009 −0.738

1980–1994

1.054∗∗ 0.201 0.329

1980–1994

1.091∗∗ 0.877∗∗ 1.396∗∗

1949–1965

−0.394 −0.262 −0.208

1949–1965

0.718 0.897∗∗ 1.501∗∗

1949–1965

0.652∗∗ 0.757∗∗ −0.015

1966–1979

Professional

−0.109 0.143 0.682∗

1966–1979

Farmer

0.791∗ 0.461 0.819

1966–1979

Public Organization

1.176∗∗ 0.852∗∗ 1.721∗∗

1980–1994

1980–1994

1.337∗∗ 0.589† −0.207

1980–1994

∗

p < 0.01 p < 0.05 † p < 0.10 Note: The reference category for family class background is worker/peasant. Middle class includes small business owners, middle-rich peasants, and intellectual family background. Exploiting class includes large business owners, landlords, rightists, bad elements. All model estimations also include a set of dummy variables indicating respondents’ province and city size.

∗∗

family class background Cadre Middle class Exploiting class

family class background Cadre Middle class Exploiting class

family class background Cadre Middle class Exploiting class

1949–1965

Government

table 4.5. Parameter Estimates of Logistic Model for Entry Into First Work Organization, Using “Family Class Background” by Period

122

The State and Life Chances in Urban China

chapter summary In this chapter, we examined patterns of entry into ﬁrst jobs – types of occupations and of work organizations – across historical periods in urban China. As we noted before, because of low job mobility, ﬁrst-job attainment had a lasting impact on one’s life chances in state socialist China; this is especially true for those in the Mao era. We now summarize the main ﬁndings in our analyses. Occupation Versus Work Organization Hierarchies We began this chapter with the arguments that there are two distinctive hierarchies related to the social construction of jobs in urban China: (1) the occupational hierarchy and (2) the organizational hierarchy based on state redistributive institutions. The ﬁndings show that, overall, the hierarchy of occupational status and that of type of work organizations showed two parallel patterns. The patterns of entry into cadre and professional occupations were similar to those for entry into government and public organizations. This is because there were considerable overlaps between the two: A large proportion of cadres were concentrated in governmental agencies, and a large proportion of professionals were in public organizations. In most cases, however, state policies in resource allocation pertained more to economic sectors than to occupational categories. As we observed in this set of analyses, education played a major role in ﬁrst-job destinations. This was even so under the radical policies in the Cultural Revolution period. This analysis gives us an opportunity to pause and reﬂect on the implications of our ﬁndings on educational attainment revealed in the last chapter. Recall that, as shown in Chapter 3, social origins, especially political selections based on family class background, played an important role in facilitating children’s educational attainment. This was consistent with the educational policy under state socialism epitomized in the slogan “both red and expert.” We can see the consequences of such educational attainment in the early stages of one’s life course: It channeled individuals into different types of jobs and work organizations, and exerted a lasting impact on their life chances in the subsequent stages of their life course. Stratiﬁcation Dynamics There is evidence that the processes of social stratiﬁcation have changed substantively across historical periods. Shifting state policies affect

Entry Into the Labor Force: Patterns of First-Job Attainment 123 individual life chances in two ways: First, state policies may drastically expand or reduce opportunities and affect those in a particular stage of the life course. For instance, the Cultural Revolution severely disrupted urban economies, forcing a large number of secondary school graduates to leave their home cities and go to the rural areas. On the other hand, the recent economic reform has greatly expanded avenues of employment, leading to divergent job trajectories. Second, shifting state policies also led to changes in the mechanisms of social stratiﬁcation, as evident in changes in the effects of education and social origins across periods. For instance, children of highly educated parents had a lower rate of entering the urban labor force during the Cultural Revolution. Children of cadre origin, on the other hand, had some advantages. It appears that the most consistent predictor of one’s job attainment is one’s own educational level. But recall that educational attainment was also sensitive to state policy shifts and the resulting stratiﬁcation dynamics. Therefore, educational attainment plays an important role in mediating the effects of stratiﬁcation dynamics on individual life chances. After arriving at their initial location in work organizations, the urban youth have moved into a new phase of their life experiences in the context of work organizations. Events such as promotions, job changes, and distribution in income and housing associated with workplace will be our focus of inquiry in the following chapters. Before we turn attention to these aspects, in the next chapter we want to conduct a microscopic analysis of one particular generation – the children of the Cultural Revolution – in their entry into the labor force. The radical state policies during that period have had lasting impacts on families in urban China and especially on this generation.

5 Children of the Cultural Revolution: The Send-Down Episode

It is necessary for the educated youth to go to the countryside, and be reeducated by the poor peasants. We need to persuade cadres and others in urban areas to send their children who graduated from middle school, high school and college to rural areas. Let us have a mobilization. Comrades in rural areas should welcome them. Mao Zedong, 1968

In retrospect, 1968 can be labeled as “the year of student movements.” In that year, college students in Paris took to the streets to protest the deterioration of the educational system in France; student protests broke out at UC Berkeley and other campuses in the United States against the U.S. government’s involvement in the Vietnam War. In that same year, however, high school and college students in China were in turmoil. Only two years before, the “Red Guard Rebellion” on college campuses and in secondary schools swept the nation, drew worldwide attention, and initiated the momentum of the Cultural Revolution. Under attacks from the Red Guards, government administrations and educational institutions were paralyzed: Government ofﬁces were taken over by the Red Guards, and no classes were offered in schools (Pepper 1996; Unger 1982). By 1968, the cold reality set in: Students could not graduate from school or advance to the next educational level; nor could they be assigned to a job in urban areas. More than six million secondary school students accumulated in schools, waiting to graduate and to be assigned a job by the government (Shi 1996). Unemployment loomed large in a stagnant urban economy. 124

Children of the Cultural Revolution

125

In this context, Mao Zedong’s instruction, quoted earlier, often labeled as the send-down policy in China, ofﬁcially began one of the most intensive political and social mobilizations during the Cultural Revolution (Bernstein 1977). In a twelve-year period, over seventeen million of the so-called urban “educated youth” – mostly graduates from middle school and high school – were forced to live and work in rural areas. The send-down policy has had profound impacts on the life course of a generation of urban youth in China. About one-third of the children of the Cultural Revolution – those urban youth who entered the labor force between 1967 and 1978 – were sent to rural areas; many stayed there for more than ten years. In the analyses of educational and ﬁrst-job attainment patterns in the last two chapters, we observed the salient impacts of the Cultural Revolution period that shaped distinctive patterns of social stratiﬁcation processes. In this chapter, we focus on the dramatic send-down episode in this era and address two major issues. First, we consider how large-scale sociopolitical events are mediated by the social stratiﬁcation structure and processes. Studies of social stratiﬁcation have focused on stable institutional structures and processes. But a highly volatile political environment and shifting state policies have been main characteristics of social stratiﬁcation processes in China and other state socialist societies. The send-down episode demonstrates, in a dramatic and transparent way, the interaction between state policies and the social stratiﬁcation structure in this unique historical context. Second, we examine how state policies have shaped and altered individual life courses. We hope to gain insight into how policy-induced life events like the sent-down experience affect individuals’ subsequent life course and economic well-being. Our emphasis on the impacts of historical events on the life course calls attention to the analytical importance of cohorts. As Ryder (1965, p. 843) pointed out: “Each birth cohort acquires coherence and continuity from the distinctive development of its constituents and from its own persistent macroanalytic features. Successive cohorts are differentiated by the changing content of formal education, by peer-group socialization, and by idiosyncratic historical experience.” The send-down policy in China targeted those urban youth that completed their secondary education and began to enter the labor force. It especially affected the cohort of the young people who were at the stage of transition into adulthood. Therefore, the sent-down experience is cohort-speciﬁc and associated with a particular historical context.

126

The State and Life Chances in Urban China

the issues In his seminal work, Children of the Great Depression, Glen Elder (1974) conducted a detailed examination of how the Great Depression affected children of different social groups and their subsequent life course, and how locations in the social stratiﬁcation structure mediated this dramatic economic event. Elder’s study pioneered a sociological approach to link large-scale social changes to the life course and set the research agenda for understanding social changes through their impacts on individuals’ life experience. Since then, the reciprocal relationship between social change and the life course has gained great attention in the sociological literature (Elder 1985b; George 1993; Hogan 1981; Moen 1985). In recent years, research has increasingly paid attention to the role of the state in shaping the life course. In industrialized market societies, as Mayer and Muller (1986) pointed out, “the state has assumed an increasing amount of responsibility for the life of individuals in society” (p. 225). They discussed three modes of state actions that affect individual life course: regulations and laws (e.g., the compulsory schooling law), various types of ﬁscal policies (e.g., social security), and the provision of services (e.g., social programs related to the aged). Through these social policies, the welfare state tends to produce an institutionalized life course in industrialized market societies. In this light, Mayer and Schoepﬂin (1989, p. 203) observed: [T]he welfare state provides continuity over the life course by preventing sudden and steep income losses through health and unemployment insurance, and by redistributing income over the life time through old age insurance. Also, the welfare state provides for people working the public sector speciﬁc education-occupation linkages, stable employment, orderly career lines with secure and progressing income.

Scholars studying China and other socialist societies have long recognized that life chances of social groups are decisively affected by state policies that shape both opportunity structures and the status of social groups. However, in contrast to the role of the welfare state in industrialized market societies, shifting state policies in China often dramatically interrupt and alter individuals’ life courses (Davis 1992b; Gold 1991; Whyte 1985; see Szel´enyi and Manchin 1987 in the Hungarian context). Following Elder’s lead, our ﬁrst goal is to examine how impacts of state policies on the life course are mediated by the existing social stratiﬁcation processes. Dramatic social and political events (economic recessions, wars, and social movements) often have differential impacts on

Children of the Cultural Revolution

127

different social groups. Elder (1974), for instance, found noticeable differences in the effects of the Great Depression experience on the children of the working class and those of the middle class. Social groups with advantageous socioeconomic status or with more resources are able to protect themselves from disastrous events or minimize the negative impact of such events. These differential effects shed light on the structure of social stratiﬁcation and the mediating processes. Indeed, the send-down policy had different impacts on children of the Cultural Revolution. On the one hand, the send-down episode was part of massive state efforts in “destratiﬁcation” that affected all social groups (Parish 1984). On the other hand, the send-down policy was by no means uniformly implemented. Not all urban youth in this cohort went to rural areas. Except in the ﬁrst few years of policy implementation, a majority of the urban youth stayed in urban areas and was employed in the urban labor force. Some sent-down youth were able to return to urban areas sooner than others. Thus, the send-down episode raises a set of interesting issues: How did state policies change individuals’ life courses? How were groups of different social origins affected by state policies? Which group, if any, was able to protect their children from adverse state policies? One central issue that we have pursued in the previous chapters is the role of social origins for individual life chances in the episode of dramatic political turmoil. In the state socialist societies, the bureaucratic class was at the center of the political authority and resource redistribution. We would expect that the bureaucrats were in an advantageous position to protect their children from the negative impact of state policies. However, as we know, during the Cultural Revolution the radical top leaders made serious efforts to dismantle the existing bureaucratic apparatus and the bureaucratic class. A large proportion of ofﬁcials and managers (administrators) in various organizations were purged and replaced during this period. In this ﬂuctuating political environment, the extent to which the bureaucrats were able to protect their children from adverse state policies is questionable. These considerations led us to expect that the role of social origins in mediating the impacts of state policies on the life course was limited. Our second goal is to assess the consequences of the sent-down experience on the children of the Cultural Revolution. Large-scale social changes inevitably shape and alter individuals’ life course, producing lasting effects on their perceptions and behaviors (Alwin, Cohen, and Newcomb 1991; Elder and Clipp 1988; McAdam 1989). Life-course events at the stage of

128

The State and Life Chances in Urban China

early adulthood seem especially critical, as Elder (1985a, p. 35) observed: “variations in the transition to adulthood leave a durable imprint on the course that follows.” For the sent-down youth, their duration in rural areas interrupted their educational attainment and delayed their accumulation of skills and work experience in the urban labor force. Therefore, their experience of life events in the subsequent life course, such as the timing of marriage and ﬁrst birth, was likely to be interrupted or delayed. Moreover, duration in rural areas may have deprived the sent-down youth of the opportunities that were available to urban residents, such as job assignments in favorable occupations and organizations. We expect that the sent-down experience had signiﬁcant and negative effects on the individuals’ life course and that the negative consequences were exacerbated with longer rural duration. Accordingly, our analytical focus is on the life experience of one cohort group – those who entered the labor force between 1967 and 1978. Speciﬁcally, we compare and contrast the group who had the sent-down experience with the group who stayed in urban areas during this period. We ﬁrst consider how state policies differentially affected different social groups by examining those factors that affected the probability of being sent down to rural areas and, among those who were sent to rural areas, the duration before returning to urban areas. We then examine how the sent-down experience in their early adulthood affected the sent-down youth’s subsequent life course and economic well-being. For this purpose, we compare the group who was sent down and the group who stayed in urban areas during major life events and the determinants of income (in selected years) in the subsequent life course.

the send-down experience in historical context The send-down policy was rooted in the unique historical context in the People’s Republic China. The Cultural Revolution severely disrupted all walks of life, especially the educational institutions. Since 1966, most governmental agencies were paralyzed and educational institutions closed, with Red Guards accumulating in schools and colleges and nowhere to go. Although Mao Zedong’s 1968 instruction marked the ofﬁcial beginning of the send-down policy and initiated the large-scale mobilization across urban areas in China, this episode had begun a year earlier. In 1967, some “Red Guard” students volunteered to go to rural areas to participate in peasants’ ﬁeldwork. The Central Committee of the Communist

Children of the Cultural Revolution

129

Party seized the opportunity and quickly issued a document to ofﬁcially endorse their endeavor (Deng 1993, p. 55). By the end of 1967, some partially functioning governmental agencies were actively involved in organizing the send-down arrangements across regions. The implementation of the state’s send-down policy varied over the course of the Cultural Revolution. After intensive mobilizations in the ﬁrst few years, the government allowed a larger proportion of secondary school graduates to enter the urban labor force. In later years, the senddown policy was also adjusted to allow one child per family to stay in urban areas. From time to time, the government selectively recruited the sent-down youth into college or the military, or employed some back into the urban labor force. There were also policies that allowed the sentdown youth to return to urban areas on an individual basis if one had the medical proof that he or she had lost the ability to endure manual labor in rural areas. The reopening of college based on the national entrance examination by the end of 1977 greatly widened opportunities for many sent-down youth to return to urban areas. Since then, the government no longer strictly enforced its send-down policies. For this reason, we conﬁne our attention to the period of 1967–1978 for the send-down events, and the period of 1967–1977 for the return events. Policy-directed large-scale migration from urban to rural areas had occurred in China before the Cultural Revolution. For instance, during the economic disaster in the early 1960s, the government reduced the urban population by twenty-six million between 1961 and 1963. But those policies mainly targeted those families who had recently moved from rural to urban areas. Hence, the proportion of the urban population affected was relatively small. In the send-down mobilization, however, all urban families with children who graduated from secondary schools were expected to follow Mao’s instruction and send their children to the designated rural areas. Some “volunteered,” but most were forced to do so under political pressures. The send-down policy especially targeted those social groups who were discriminated against during the Cultural Revolution. As two sent-down youths recounted their experience more than ten years later (Deng 1993, p. 60): My family class background was bad and my father was labeled a ‘rightist.’ So when Chairman Mao’s instruction was announced, my father ‘voluntarily’ applied for going to rural areas on behalf of his three children . . . When he came home that day, he held the three of us in his arms and cried: ‘It is not that I don’t want to keep you in the city. But I dare not do so . . .’ I can never forget my father’s eyes that day, ﬁlled with torture, fear, sadness, and guilt . . . Sending his three children to rural

130

The State and Life Chances in Urban China

areas did not save my father’s soul. Instead, he put upon himself an enormous psychological burden. He died of lung infection in the winter of 1971. I was only 15 when I was sent down. No one wanted to go, but no one could resist. When I refused to go, those in charge of the residential committee came to our home everyday, asking us to study Chairman Mao’s instructions. A member of the worker’s propaganda team came to live in our home and organized a study team for my family. My father was a cadre. He was locked up in a study team in his workplace and was not allowed to return home until his children agreed to go to the rural area. In the end, my mother begged me to go to the rural area.

The send-down experience was a dramatic, and for many a traumatic event in one’s life course. There were vast disparities in living environments between the urban and the rural areas in China. Most urban youth had never lived in rural areas before their sent-down experience. They were sent to designated rural areas far away from their home cities. In many cases, they were only allowed to visit their families for a few weeks every three years. They had to endure harsh physical labor in the ﬁeld, often more than twelve hours a day, seven days a week (Jiang, Li, and Shi 1996; Yang 1992; Yu and Wang 1993). Most sent-down youth lived and worked in rural areas for an extended period of time. In our sample, the average duration in rural areas was six years. Nineteen percent of them stayed more than ten years; 39 percent, between ﬁve and ten years, and 41 percent had ﬁve or fewer years of experience. In this context, opportunities of staying in urban areas and of returning to urban areas after one was sent down were scarce and valuable. Variations in access to these opportunities shed light on mechanisms of social stratiﬁcation, especially the distribution of political resources among different social groups. For instance, there is evidence that high-rank cadres succeeded in getting their children back to urban areas through various “back-door” channels (Shi 1996). According to one estimate, between 1972 and 1976, seventy percent of the college students admitted based on the system of political recommendation were the children of cadres and those with political connections (Deng 1993, p. 163). In 1979, three years after Mao Zedong died, and under enormous pressures of continuous riots and protests by the sent-down youth, the reform-minded leadership denounced the send-down policy and allowed all the sent-down youth to return to their home cities (except for those who were married to local residents or employed in nonrural jobs in the local areas). This signaled the ofﬁcial end of the send-down episode. The end of the send-down era also marked the beginning of the great transformation of China’s state socialist economy. Major socioeconomic

Children of the Cultural Revolution

131

changes were shaping the landscape of the urban economy in which the sent-down youth sought to relocate their social positions. The economic reform signiﬁcantly altered opportunity structures and generated a variety of job trajectories. In particular, since the late 1970s, the pressure of unemployment in urban areas has forced the central government to legitimize self-employment that gave rise to a striving private sector (Bureau of Policies and Regulations, 1990, Vol. 1, p. 159). Moreover, patterns of economic beneﬁts (e.g., personal income) associated with different occupations and work organizations began to change in emerging market economies. Private entrepreneurs and those working in the nonstate ﬁrms have signiﬁcantly higher income than do employees in the state sector. It is in this historical context that we interpret the consequences of the sent-down experience later.

historical patterns of the send-down episode We ﬁrst describe the historical patterns of the send-down episode during the Cultural Revolution period. Figure 5.1 reports the estimates of the hazard rate of entry into ﬁrst jobs, 1966–1978. To recall, the hazard rates measure the annual instantaneous rate of entering a speciﬁc type of job for those in the risk set (i.e., individuals who completed their education). As we described before, although the ofﬁcial policy was not in effect before the end of 1968, send-down movement was initiated and ofﬁcially sanctioned in 1967. The large-scale send-down mobilization began in 1968, as reﬂected in the peak of the hazard rate (the solid, bold line) between 1968 and 1969. The hazard rates for other job destinations indicate different job opportunities in urban areas. Figure 5.2 shows the hazard rate of returning to urban areas among the sent-down youth. Before the ofﬁcial termination of the “send-down” policy in 1979, some sent-down youth returned to urban areas through various channels. The peak in 1979 reﬂected the end of the send-down policy and most sent-down youth returned to urban areas by the early 1980s. The increase in the hazard rate in the mid-1980s coincided with the beginning of the urban reform in China, when the state control over the urban labor force was weakened. Thus, the urban reform may have created new opportunities for those still left behind to return. Notice that, although the second peak in the mid-1980s is high, this is mainly due to the relatively small number of individuals still in the risk set (i.e., still in rural areas) at the time.

The State and Life Chances in Urban China

132 0.16

hazard rate

0.12

0.08

0.04

0

66

68

government

70

72 year

public org'n

state firm

74

76

collective firm

78 farm

ﬁgure 5.1. Hazard rates of entry into ﬁrst job, 1967–1978.

0.6

0.5

hazard rate

0.4

0.3

0.2

0.1

0

66

68

70

72

74

76 78 year

80

82

84

ﬁgure 5.2. Hazard rates of return to urban areas, 1967–1987.

86

Children of the Cultural Revolution

133

Variations in the hazard rates of entering different work destinations and of returning to urban areas indicate that the implementation of the send-down policy was not uniform over time or across social groups. We now turn to examine and explain these variations.

different impacts of state policies on social groups Who Was More Likely to Be Sent Down? During the Cultural Revolution episode, the major divide for the urban youth was two distinctive ﬁrst-job destinations: a job in the urban area and being sent to rural areas. We ﬁrst examine how the children of the Cultural Revolution were allocated into these two types of jobs. We included in the model the covariates that measure social origins and related individual attributes such as gender and educational credentials. As before, we consider the father’s occupational status and family class background as two alternative measures of social origins. In addition, we also included a measure of urban job opportunities1 and the number of siblings in the respondent’s family as control variables, as the probability of being sent down was related to both factors. We analyzed the probability of the send-down events for the entire 1967–1978 episode as well as for two reﬁned periods, 1967–1971 and 1972–1978, so that we could examine period-speciﬁc variations. The urban youth that entered the labor force during the ﬁrst period were often labeled as the old sent-down youth in Chinese literature, whereas those who entered the labor force in the second period were labeled as the new sentdown youth. There were important contextual changes across the two 1

The send-down policy was in part the government’s response to unemployment. Therefore, we expect that the probability of the send-down events varied with urban economies and employment conditions across regions. For the analyses of the sent-down and return events, the most important contextual information is about changes in urban employment opportunities. We devised a measure of percentage change in urban employment opportunities in a province in two adjacent years. It would be ideal to use city-level information on employment opportunities. But, such data are not available for all the cities across the years. We used the provincial-level information as a proxy. This measure is deﬁned as follows:

N i,t = (N i,t − N i,t−1 )/N i,t−1 . where N i,t refers to the number of urban employees in province i, in year t. N i,t−1 refers to the same statistic but in the previous year (t − 1). N i,t measures the percentage change in the urban labor force in the particular province between the adjacent years, t − 1 and t. We used a three-year average measure to smooth random ﬂuctuations in any particular year.

134

The State and Life Chances in Urban China

periods: Colleges were reopened in 1972 on a limited basis; changes in the send-down policies also allowed greater local variations in the second period. These changes provided opportunities for some selected send-down youth to enter college, or to be recruited into the urban labor force. Analyzing these two periods separately allows us to detect potential changes in stratiﬁcation processes that mediate state policies and individual life chances. Table 5.1 reports the estimates of the probability (in log-odds form) of being sent down to rural areas. Model A used the father’s occupation as a measure of social origins, whereas Model B used family class background as the alternative measure of social origins. The Overall Pattern The analysis in Panel 1 examines the determinants of being sent to rural areas in the entire Cultural Revolution period (1967–1978). The ﬁndings show that the probability of being sent down decreased with age. There was no signiﬁcant gender difference. That is, the adverse send-down policy affected both men and women in a similar manner. As for the effect of educational credentials, the odds of being sent down for high school graduates was 80 percent higher (exp[.586] = 1.80) than that for those with middle school or lower education (the reference group). There was no statistically discernible difference between the college graduates and the reference group. It appears that the senddown policy especially affected the urban youth that graduated from high school. An important issue for our attention is the role of parental political status in mediating the effect of adverse state policies. In Model A, we included the father’s education and occupational status as indicators of the father’s political status. The most salient evidence of political selection is that children of those who had a college education had a signiﬁcant, higher probability of being sent to rural areas: The odds for this group was 40 percent higher than that for the reference group whose fathers had elementary or no education. This ﬁnding is consistent with the state policies in this period that discriminated against those with an intellectual family background. In the state socialist political order, the father’s occupation (e.g., cadre status) was closely associated with his political status. In this period, however, we ﬁnd no statistically discernible differences in the effects of father’s occupational status on the probability of experiencing the senddown event. Thus, there was no evidence that high-rank cadres, a large

135

Intercept Age Female education High school College father’s education Middle school High school College father’s occupation Cadre, high-rank Cadre, low-rank Professional, high-rank Professional, low-rank family class background Cadre Middle class Exploiting class Number of brothers Number of sisters Percent change in urban labor force −0.011 −1.093∗ −0.028 0.355† 0.342† 0.115 −0.302† −0.043 −0.167 — — — 0.015 0.035 −0.120∗∗

0.587∗∗ 0.294 −0.226∗ 0.037 0.334∗∗ — — — — −0.024 0.110 0.636∗ 0.043 0.048† −0.117∗∗

−0.211† 0.061 0.333∗

−0.051 −0.079 0.073 0.071

— — — 0.045 0.048† −0.117∗∗

−2.322 −0.009 0.043

∗∗

0.586∗∗ 0.304

0.790 −0.221∗∗ 0.049

0.775 −0.222∗∗ 0.050

A ∗∗

B

B ∗∗

0.197 −0.124 0.372 0.013 0.036 −0.121∗∗

— — — —

−0.038 0.223 0.245

−0.019 −1.128∗

−2.353 −0.012 0.033

1967–1971

∗∗

Model A

1967–1978

†

— — — 0.076† 0.050 −0.056∗

−0.117 0.183 0.242 0.315

−0.579∗∗ −0.231 0.181

−0.913∗∗ 1.075∗

0.637 −0.238∗∗ 0.049

A

B

(continued)

−0.277 0.071 0.516 0.070 0.045 −0.052

— — — —

−0.492∗∗ −0.069 0.385∗

−0.926∗∗ 1.126∗∗

0.713† −0.241∗∗ 0.060

1972–1978

table 5.1. Parameter Estimates of Logistic Model for Probability of Being Sent-Down: Sent-Down = 1, 1967–1978

136

16 765

15 765

−0.161 — −0.196 1004.9∗∗

−0.042 −0.213† — 1001.9∗∗ 16 425

−0.215 −0.288 — 89.1∗∗

A

B

15 425

−0.238 — −0.075 85.4∗∗.

1967–1971

16 340

−0.203 −0.347 — 320.0∗∗

A

B

15 340

−0.362 — −0.542 317.8∗∗

1972–1978

∗

p < 0.01 p < 0.05 † p < 0.10 Note: Middle school education or below is the reference category for respondent’s education; Elementary education or below, for father’s education; worker, for father’s occupation; and middle class, for family class background.

∗∗

Father education missing Father occupation missing Family background missing χ 2 (compared with the baseline model with no covariates) d.f. N events

B

Model A

1967–1978

table 5.1 (continued)

Children of the Cultural Revolution

137

proportion of whom were the target of purge during the Cultural Revolution, were able to help their children evade adverse state policies. Not surprisingly, the probability of experiencing the send-down event was closely associated with urban job opportunities. An increase in employment in urban areas signiﬁcantly reduced the probability of being sent to rural areas. The number of siblings did not have a signiﬁcant effect. In Model B, we estimate the effect of family class background on the send-down event. The ﬁndings show that children of the exploiting class had a signiﬁcant, higher probability of being sent to the rural area. The odds for this group was 89 percent higher than that for the children of the working class (the reference group). This is again consistent with the proclaimed state policies during this period that used family class background labels to discriminate among social groups. Other covariates show similar effects as in Model A. Variations Over the Two Periods We now examine variations, if any, across the two periods. Panels 2 and 3 in Table 5.1 report the estimated effects of the covariates for two reﬁned periods: 1967–1971 and 1972–1978. Comparing across these two periods, there are a few noticeable differences with regard to the effects of social origins. For instance, the father’s occupational status or family class background had no signiﬁcant effect on the sent-down rate in both periods. In the earlier period (1967–1971), when the send-down policy implementation was stringent, children of those with high school or college education were more likely to be sent to rural areas (Model A). In the second period, there was no difference with regard to the father’s education. However, once family class background is controlled for (Model B), children of those with a college degree had a higher probability of being sent to rural areas. College graduates had a higher rate of being sent to rural areas in the second period. We suspect that this was mainly because of the radical policies in the 1970s that advocated that college graduates under the “new educational system” participate in production processes in workshops and ﬁelds. Another interesting ﬁnding is that the effect of changes in the urban labor force was smaller and statistically insigniﬁcant in the second period. It appears that, as the send-down policy was institutionalized over time, it became a political ritual and was implemented regardless of urban job opportunities. Overall, although there is some evidence of political selection in the sent-down events, the general picture is that all social groups have had

138

The State and Life Chances in Urban China

similar experiences, and children of the bureaucratic class were unable to be sheltered from the adverse effects of state policies. Return to Urban Areas Before 1979, most sent-down youth were forced to stay in the countryside. But, as we noted before, there were channels that permitted some individuals to return to urban areas. The events of acquiring permission to return provide information about the political or social resources possessed by different social groups to minimize the adverse impacts of state policies. We analyzed the rate of return events between 1967 and 1977, as well as the two reﬁned periods within this episode. Table 5.2 reports the logit model estimates for this set of analyses.2 The Overall Pattern Our ﬁndings for the entire Cultural Revolution period (1967–1977) show that age and gender had no signiﬁcant effects on the return rate. But duration in rural areas signiﬁcantly increased the probability of experiencing the returning event. Those with higher educational levels tended to return earlier than those with low educational levels. Compared with the sent-down youth who had middle school or less education, graduates from high school had 90 percent higher odds of returning to urban areas; those with college education had even higher odds. We suspect that this pattern resulted from those policies of college admission, urban employment, or military recruitment that favored candidates with higher educational levels. For our purpose, the effects of social origins are especially illuminating. The father’s education had no signiﬁcant effect on the return rate. On the other hand, the effects of the father’s occupation clearly indicate the importance of political status. Children of high-rank ofﬁcials had a signiﬁcant, higher probability of returning to urban areas: The odds of returning to urban areas for this group was twice as high as those whose fathers were workers (the reference group). Children of low-rank cadres had a 95 percent higher rate than children of the working class, with children of professionals having no advantage. Again, an increase in urban

2

Because our sample frame is in the urban areas, our analyses of the return events focused on the sent-down youth who eventually returned to urban areas. To the best of our knowledge, most sent-down youth returned to their home cities. The proportion that remained in rural areas is very small and should not affect our analyses appreciably.

139

Intercept Age Duration in rural areas Female education High school College father’s education Middle school High school College father’s occupation Cadre, high-rank Cadre, low-rank Professional, high-rank Professional, low-rank

−4.563 0.029 0.288∗∗ −0.072 0.675∗∗ 1.733∗∗

0.660∗∗ 1.757∗∗ −0.117 −0.327 0.008 — — — —

−0.234 −0.523 0.005

0.711∗ 0.680∗∗ −0.533 −0.228

−4.707 0.028 0.307∗∗ −0.055

∗∗

B

∗∗

Model A

1967–1977 ∗∗

0.327 1.347∗∗ −1.517 —a

−0.701 0.178 −0.129

0.696 3.888∗∗

−7.146 0.145 0.567∗∗ −0.894∗

A

B ∗∗

— — — —

−0.329 0.686 −0.174

0.728† 3.173∗∗

−7.045 0.145 0.522∗∗ −0.892∗

1967–1971 ∗∗

0.767∗ 0.515∗ −0.418 −0.106

−0.156 −0.697∗ −0.006

0.584∗∗ 1.520∗∗

−4.234 0.017 0.258∗∗ 0.058

A

B

(continued)

— — — —

−0.077 −0.502 0.071

0.602∗∗ 1.535∗∗

−4.067∗∗ 0.015 0.243∗∗ 0.038

1972–1977

table 5.2. Parameter Estimates of Logistic Model for Probability of Returning to Urban Areas, Return to Urban Areas = 1, 1967–1977

140

— — — −0.068 0.025 0.085∗∗ −0.156 −0.361 — 220.7∗∗ 17 273

−0.098 −0.151 −0.381 −0.079 −0.013 0.029∗∗ −0.203 — −0.268 197.4∗∗ 16 273

B — — — −0.270† 0.041 0.068 —a −1.337 — 68.7∗∗ 15 41

A

B 0.228 −1.477† −1.009 −0.266† 0.075 0.076† —a — —a 61.1∗∗ 14 41

1967–71

— — — −0.037 0.025 0.083∗ −0.039 −0.323 — 114.6∗∗ 17 232

A

B −0.189 0.001 −0.282 −0.045 0.012 0.088∗∗ −0.069 — −0.043 100.0∗∗ 16 232

1972–77

∗

p < 0.01 p < 0.05 † p < 0.10 a This category is not estimated because of no variation (no return event) for this period. Note: Middle school education or below is the reference category for respondent’s education; elementary education or below, for father’s education; worker, for father’s occupation; and middle class, for family class background.

∗∗

family class background Cadre Middle class Exploiting class Number of brothers Number of sisters Percent change in urban labor force Father education missing Father occupation missing Family background missing χ 2 (compared with the baseline model with no covariates) d.f. N events

Model A

1967–77

table 5.2 (continued)

Children of the Cultural Revolution

141

employment opportunities had a signiﬁcant, positive effect, increasing the probability of returning to urban areas. Family class backgrounds (Model B) did not show any signiﬁcant effect for the entire period. Obviously, access to the return opportunities was more associated with more immediate positional power (father’s occupational status) than with symbolic labels (the label of family class background). Variations Over the Two Periods Our additional analyses for the two reﬁned periods show that there were considerable variations over time. In the 1967–1971 period (Model A in Panel 2), only children of low-rank cadres had a signiﬁcant and higher rate of returning to urban areas. We suspect that this is because, in the earlier years of the Cultural Revolution, high-rank cadres were especially targeted in the political purge. On the other hand, family class background showed some signiﬁcant effects consistent with the proclaimed political selection criteria (Model B in Panel 2): Compared with a worker/poor peasant family class background (the reference category), the middle class background had a lower return rate – only 24 percent of those with a worker/poor peasant family class background. As we have expected, the effect of the exploiting class background was negative but not statistically signiﬁcant, probably due to the fewer number of cases in this category. Female youth had a signiﬁcant and lower probability of return in the early period. In the 1972–1977 period, the effects of family class background disappeared. Instead, the father’s occupational status became important. The odds of return for the children of high-rank ofﬁcials was twice as high as that for those whose fathers were workers. Children of low-rank cadres also had a signiﬁcant and higher odds (67 percent higher). It is obvious that, in the second period, broadly deﬁned class labels for family background were no longer effective in discriminating among social groups. Rather, parent’s political status (father’s cadre status) had a more direct effect on children’s return to urban areas. Urban employment opportunities became a signiﬁcant factor in increasing the return rate in the second period. To sum up, patterns of returning to urban areas show distinctive effects of political selections and varying effects of political criteria. In the early period, when political pressures were at their height, symbolic class labels of worker/poor peasant family background contributed signiﬁcantly to the return probability. In the second period, positional power reemerged

142

The State and Life Chances in Urban China

and played a signiﬁcant role, as evidenced by the effects of cadres in Model A in the second period.

consequences of the sent-down experience How did the sent-down experience affect the sent-down youth’s subsequent life course? Given their long duration in rural areas (an average of six years in our sample), we expect that this episode had a lasting effect on this group. We now turn to the second theme of the analysis – assessing the consequences of the sent-down experience. Experience of Subsequent Life-Course Events To examine the consequence of the sent-down experience on the subsequent life course, we ﬁrst compare those who stayed in urban areas and those who were sent down in their experience of following major life course events: (1) age at marriage, (2) age at ﬁrst birth, (3) educational (college) attainment, and (4) location in type of workplaces and occupations in one’s ﬁrst urban job. Given the considerable variations in the length of duration in rural areas, it is likely that the effects of the sent-down experience varied with one’s duration in rural areas. For those who were able to return earlier, the negative effects of the sent-down experience may be minimized. For those who stayed for a longer duration, the negative consequences may be exacerbated. Based on this consideration, we further divide the sent-down youth into two groups: (1) those who stayed in rural areas for less than six years and (2) those who stayed in rural areas for six or more years.3 Age at Marriage and First Birth To explore the effects of the sent-down experience, we divide the respondents in each category into three age groups (eighteen to twenty-ﬁve, twenty-six to thirty, thirty-one or above). Figures 5.3 and 5.4 display the distribution of age at marriage and at the birth of ﬁrst child among the three groups (based on Table 5.3). It is interesting to note that there were only minor differences between those who stayed in urban areas and those who had less than six years of the sent-down experience. But, for those who worked for more than six years in rural areas, a much larger 3

The choice of six years is because it coincides with both the mean and the median of the sent-down duration for the sent-down group in our sample.

Children of the Cultural Revolution

143

table 5.3. Descriptive and Test Statistics of Group Differences in Experiencing Life-Course Events

Covariate Age at marriage (mean age) age group (%) 18–25 26–30 31+ Sample N Pearson’s χ 2 test of independence Age at ﬁrst birth (mean age) age group (%) 18–25 26–30 31+ Sample N Pearson’s χ 2 test of independence Percentage obtaining college degree after 1977 Sample N Pearson’s χ 2 test of independence type of workplace at ﬁrst urban job (%) Government Public organization State ﬁrm Collective ﬁrm Hybrid ﬁrm Private entrepreneur Sample N Pearson’s χ 2 test of independence occupation at ﬁrst urban job (%) Cadre Professional Clerk Service worker Production worker Military, police Sample N Pearson’s χ 2 test of independence

No Sent-Down Experience

Sent Down for Less Than 6 Years

Sent Down for More Than 6 Years

26.5

26.6

28.1

38.6 36.1 52.9 56.5 8.4 7.4 1623 407 χ 2 = 88.1, d.f. = 4, p < 0.01 27.4 27.4

16.5 65.6 18.0 401

27.1 25.0 57.7 63.3 15.2 11.7 1403 376 χ 2 = 56.4, d.f. = 4, p < 0.01 8.6 14.1

13.2 59.5 27.3 363

1612 403 χ 2 = 11.6, d.f. = 2, p < 0.01

404

10.4 6.7 50.3 28.1 1.0 3.8 1967 χ 2 = 41.8, d.f. = 10,

11.6 8.1 52.8 25.5 0.5 1.5 396 p < 0.01

4.9 11.7 45.3 32.4 2.6 3.1 386

3.1 5.9 13.8 11.3 3.6 5.7 11.1 13.1 62.8 58.1 5.7 5.9 1805 389 χ 2 = 59.7, d.f. = 10, p < 0.01

9.2 10.3 6.2 12.7 61.0 0.5 369

28.9

9.4

The State and Life Chances in Urban China

144

70 60

percent

50 40 30 20 10 0 18-25

not sent down

26-30 age sent down < 6 yrs

31+

sent down > 6 yrs

ﬁgure 5.3. Distribution of age at marriage among three groups.

70 60

percent

50 40 30 20 10 0 18-25

not sent down

26-30 age sent down < 6 yrs

31+

sent down > 6 yrs

ﬁgure 5.4. Distribution of age at birth of ﬁrst child among three groups.

Children of the Cultural Revolution

145

proportion of them delayed their marriage and child rearing after thirty years old. It is worth pointing out that, since the 1970s, all urban youth delayed their marriage and child rearing as a result of the government’s family planning policies. We suspect that the observed differences among these groups would be more salient were these policies absent. Educational Attainment Because most urban youth graduated from secondary schools during the Cultural Revolution, we focus on their attainment of a college education. We compared college enrollment among different groups since 1977 when national examination was resumed. Since 1977, a large number of the children of the Cultural Revolution (including graduates from both middle school and high school) took college entrance examinations in an effort to resume their disrupted education. There were signiﬁcant differences among the three groups (see Table 5.3): The sent-down youth with a shorter duration had the highest proportion of college education (14.1 percent). But the resumption of the college entrance examination in 1977 may have come too late for some of those who stayed in rural areas for more than six years and whose enrollment rate was noticeably lower than those with shorter rural duration (9.4 percent). Interestingly, only 8.6 percent of those who stayed in urban areas went to college after 1977. It is likely that the negative life experience had fostered the sent-down youth’s determination to use educational achievement to relocate their social positions. Types of Occupations and Work Organizations in the First Urban Job Location in the labor force plays an important role in obtaining economic beneﬁts. Because an individual’s job location may vary over time, we compared types of ﬁrst urban jobs among the three groups. For the sentdown youth this job refers to the ﬁrst job after they returned to urban areas. The descriptive statistics in Table 5.3 indicate, surprisingly, that those with the sent-down experience tended to have similar proportions in government and/or public organizations and in occupations such as cadres and professionals, compared with those who stayed in urban areas. This may be due to the fact that a larger proportion of the sent-down youth received a college education, which channeled them into favorable organizations and occupations. It might also have resulted from favorable state policies in the early 1980s. Overall, the empirical evidence suggests that the sent-down experience had a signiﬁcant effect on various life events in one’s subsequent life

146

The State and Life Chances in Urban China

course. It appears that the longer one stayed in the rural area, the more severe its impact on the subsequent life course. However, it is also noticeable that the rural experience may have fostered the sent-down youth’s determination to change their social dislocation. This is in part reﬂected in the higher proportion of college education for this group, which led them into high-status work organizations and occupations. Income Determinants To assess the impacts of dramatic life events on individuals’ economic well-being, we now examine the effects of the sent-down experience on personal income in the post-Mao era. More systematic examination of personal income over time and across social groups is reported in Chapter 7; here we take a glimpse into this speciﬁc cohort through the lens of a life-course perspective. We use the logarithm of a respondent’s total income – the sum of basic income, bonuses, and income from second jobs – as the dependent variable. We include in the model the covariates that measure gender, labor force experience, education, occupation, work organizations, as well as residential locations. We estimate the income models in three years, 1978, 1987, and 1993, to examine variations of these effects over time. To contrast the two groups with and without the sent-down experience, we create a set of interaction variables to indicate the additional effect of the sent-down experience. The use of the interaction terms allows us to detect the speciﬁc channels through which the sent-down experience affects personal income. Determinants of Income in 1978 Column 1 of Table 5.4 reports the estimated income determinants in 1978, the year before the send-down policy was ofﬁcially abolished. To set up the historical context, this was the year at the beginning of the postMao era, but was still inﬂuenced by the aftermath of the radical policies in the Cultural Revolution. There were 304 respondents in this analysis that returned to urban areas by 1978. Before we examine the speciﬁc effects of the sent-down experience as indicated by the interaction effects, we ﬁrst describe the general patterns of income determinants for the whole cohort at the beginning of the post-Mao era, which are estimated by the “main effects” in the ﬁrst block of the variables in the model. A female employee earned 12 percent less than a male employee. Labor force duration, as measured by the ﬁrst- and second-order effects of years

Children of the Cultural Revolution

147

table 5.4. OLS Estimates of the Determinants of Income, Selected Years Covariates Intercept Female Labor force duration Labor force duration2 /100 education High school College work organization Government Public organization Central govt ﬁrm Local govt ﬁrm Hybrid ﬁrm Private occupation Cadre Professional Clerk Service worker sd (dummy: sent-down = 1) sd duration (years) interaction with sent-down experience Female∗ sd Labor force duration∗ sd Labor force duration2∗ sd/100 education Senior High∗ sd College∗ sd work organization Government∗ sd Public org’n∗ sd Central gov’t ﬁrm∗ sd Local gov’t ﬁrm∗ sd Collective ﬁrm∗ sd hybrid ﬁrm∗ sd Private∗ sd occupation Cadre∗ sd Professional∗ sd Clerk∗ sd

1978

1987

1993

3.416∗∗ −0.123∗∗ −0.033 −0.059∗

4.661∗∗ −0.186∗∗ 0.013 0.012

4.646∗∗ −0.208∗∗ 0.014 0.009

0.003 0.048

0.036 0.070†

0.106∗ 0.109∗ 0.179∗∗ 0.149∗∗ 0.270∗∗ —

0.052 0.071 0.136∗∗ 0.130∗∗ 0.152∗∗ 0.593∗∗

0.218∗∗ 0.260∗∗ 0.255∗∗ 0.127∗∗ 0.184∗∗ 0.647∗∗

0.037 0.072† 0.012 0.109∗∗ −1.995 −0.007

0.141∗∗ 0.094∗ 0.084 0.082∗ −0.762 −0.004

0.204∗∗ 0.111∗∗ 0.050 0.046 −0.098 −0.007

0.103† −0.074 −0.059

0.073 −0.034 −0.042

0.076† −0.016 −0.019

0.007 0.086

−0.063 0.025

0.022 0.033

−0.289 −0.013 −0.156 −0.226 −0.140 −0.235 —

0.231† 0.087 0.093 0.080 0.070 0.162 −0.002

−0.288∗∗ −0.246∗ −0.172† −0.139 −0.207∗ 0.020 −0.301∗

0.151 −0.285 −0.182

−0.034 0.032 −0.046

−0.059 0.024 0.116

−0.026 0.003

(continued)

148

The State and Life Chances in Urban China table 5.4 (continued)

Covariates Service worker∗ sd Production worker∗ sd Adjusted R2 N (total/with sd experience)

1978

1987

1993

−0.019 0.003 0.24 2220/304

−0.020 −0.065 0.26 2407/683

−0.004 0.003 0.43 2738/773

∗∗

p < 0.01 p < 0.05 † p < 0.10 Note: Middle school or below is the reference category for education; employee in collective ﬁrm with no sent-down experience, for work organization; and production worker with no sent-down experience, for occupation. The models also include a set of dummy variables to indicate respondents’ provincial and city-size locations. ∗

in the urban labor force, has a downward slope. That is, for this group, the earlier one entered the labor force, the lower the wage rate, other things being equal. There were no signiﬁcant differences in returns to educational credential for this group in 1978. Because our sample includes only one cohort, differentiation in returns to labor force experience and to education may be relatively limited by 1978. With respect to the effect of one’s location in type of work organizations and occupations, those working in government, state ﬁrms, and hybrid ﬁrms all had signiﬁcantly higher income than those in collective ﬁrms and with no sent-down experience (the reference category).4 There was less income difference among occupational groups. Only service workers had signiﬁcantly higher income than production workers with no sent-down experience (the reference category), perhaps reﬂecting the commercial activities in this sector in the early phase of the reform. Due to the small number of private entrepreneurs, we omitted this group in this analysis. Our research interest centers on the effects of the sent-down experience on income. The ﬁndings show that, on average, the sent-down event has no signiﬁcant effect, as indicated by the effect of the “sd” dummy variable; nor does the duration in rural areas. With regard to the interaction effects, most covariates (types of work organizations and occupations) show negative effects on income. But they are not statistically signiﬁcant. Before 4

Because we include the interaction term between collective ﬁrm and the sent-down dummy variable in the model, only those employees in collective ﬁrms who had no sent-down experience were in the reference group for organizational categories. Similarly, only those production workers who did not have the sent-down experience were included in the reference group for occupational categories.

Children of the Cultural Revolution

149

the send-down policy was abolished, a larger proportion of those who returned earlier had better educational credentials, and many were children of high-rank cadres (cf. Table 5.2). Their human capital and parental political resources may have facilitated them to relocate in those advantageous workplaces or occupations and thus reduced the negative consequences of the sent-down experience. Determinants of Income in 1987 By 1987, most sent-down youth had returned to urban areas for an extended period of time and urban China was undergoing dramatic changes in the process of urban reform that had begun in 1985. Thus, patterns of income determinants in the analysis reﬂected both the impact of the sent-down experience and that of the urban reform under way. In contrast to the patterns in 1978, gender difference increased, but labor force duration and education had no signiﬁcant effects in 1987 (Column 2 of Table 5.4). With regard to work organizations, those in state ﬁrms had higher income; so did those in hybrid ﬁrms, compared with employees in collective ﬁrms and with no sent-down experience (the reference category). Private entrepreneurs had the highest earning among all occupational groups. These results reﬂect the impacts of the initial urban reform, where the decentralization of authority in industrial and service organizations led to higher bonuses for their employees, relative to those in government and public organizations (Walder 1989). Among the occupational groups, cadres, professionals, and service workers earned more than production workers did. The effects of the interaction between the main covariates and the sentdown experience were similar to those in 1978: There were no statistically discernible effects of the sent-down experience. It appears that the sentdown experience had no signiﬁcant impacts on the returned youth’s economic well-being, at least with respect to personal income. These results are somewhat surprising, given the social dislocation that these sent-down youth experienced. A plausible explanation is that, in the 1980s, the government had adopted a series of policies to facilitate the sent-down youth’s transition into the urban labor force (Bureau of Policies and Regulations 1990b, pp. 371–374). For example, the Ministry of Labor and Personnel issued a directive on June 28, 1985, which stated that: For those who were sent down to rural areas according to governmental planning and administration, their work duration in the rural area can be counted as part

150

The State and Life Chances in Urban China

of their work experience in their urban job after they began working in the urban labor force . . . From now on, their wages and welfare beneﬁts, if related to length of work experience, should be based on this new calculation, and be treated in the same way as those workers with the same length of work experience. (Bureau of Policies and Regulations, 1990, Vol. 2, p. 374).

The most signiﬁcant point in this directive required that the sent-down youth’s duration in rural areas be counted as work experience in the workplace for wage grade and promotions. These policies may have helped to reduce income disparity between the two groups. Determinants of Income, 1993 By 1993, there were signiﬁcant changes in the effects of income determinants (Column 3 of Table 5.4). Overall, gender-based income differences increased. But it is noticeable that a female employee with the sent-down experience had a signiﬁcant and positive effect on her income. This may indicate that female workers with the sent-down experience were more active in changing their social positions, compared with their counterparts who remained in the urban area. College education signiﬁcantly increased one’s income in 1993. Location in type of work organizations became increasingly important. For individuals who stayed in urban areas, those working in government, public organizations, state ﬁrms, and hybrid ﬁrms, all had higher income, compared with those in collective ﬁrms with no sent-down experience (the reference category). But for the sent-down youth, working in government, public organizations, central-government-owned ﬁrms, and collective ﬁrms all had signiﬁcant, lower income than their counterparts in the same type of workplaces. Even private entrepreneurs who had the sent-down experience had a lower income, relative to their counterparts with no sent-down experience. There were no signiﬁcant occupational differences based on the sent-down experience. Variations in income determinants over the three years show signiﬁcant changes in urban China in the post-Mao era. Among the children of the Cultural Revolution, college-level education has become an important factor in recent years. Cadres and professionals earned considerably more now than before. More important, types of work organizations became increasingly important in determining one’s economic well-being. The rise of the private sector has had a profound impact on the distribution of income, as private entrepreneurs had the highest income among urban residents. These patterns reﬂect the multifaceted reform processes in which both redistributive institutions and market mechanisms coexist.

Children of the Cultural Revolution

151

These fundamental changes have affected the economic well-being of the sent-down youth in several ways. Until recently, it appears that the sent-down experience did not cause signiﬁcant income disparity for the sent-down youth. We think that this may largely result from China’s economic reform. The emergence of new economic sectors and the expansion of a market economy may have alleviated the impacts of the send-down events by creating diverse job opportunities and avenues of obtaining economic resources. In the 1980s, the state also adopted a series of social policies that aimed at smoothing the sent-down youth’s relocation in the urban labor force and equalizing opportunities for the sent-down youth in the workplace. Ironically, the emergence of marked work organization-based disparity for the sent-down youth in 1993 also reﬂected the impacts of the economic reform in recent years. Workplace-based disparity coincided with the trend of work organizations moving away from the command economy. The weakening of the state may have undermined the effectiveness of those social policies that helped the sent-down youth’s transition into the urban labor force before. As a result, the lack of skills and of work experience in the urban labor force now became a salient factor affecting the sent-down youth’s economic well-being. The increasing disparity in personal income in recent years indicates the lasting effects of the negative life-course experience, which appear to have accelerated along with the ongoing social changes in China.

chapter summary We now return to the central issues we intended to address in this chapter: How do state policies affect the life course? How do social stratiﬁcation processes mediate these impacts? What are the consequences of large-scale social changes for one’s subsequent life course and economic well-being? The ﬁndings in this chapter and preceding ones point to the critical impacts of the state and state policy shifts that disrupt and restructure individuals’ life course in China. The twelve-year send-down mobilization in China has fundamentally changed the life course of a generation of urban youth. Evidence on patterns of labor force entry into and of return from rural areas show that, when an authoritarian state was highly mobilized to implement its policy, no social groups could resist its intrusion into their lives. This is indicated by the fact that, among children of different social origins, there were no signiﬁcant differences in the probability of being sent to rural areas. In this sense, the send-down policies

152

The State and Life Chances in Urban China

had indiscriminate and adverse effects on all social groups. The extent to which the bureaucratic class could resist state policies was indeed limited. Among the social groups, children of those with higher education or with an exploiting class background were especially targeted, consistent with the state policies of the time. However, there were also signiﬁcant effects of social origins on the probability of returning to urban areas. As the political tides subsided in later years, high-rank cadres did have an advantage in reducing the adverse impact of state policies by channeling their children back earlier from rural areas. Such early-return opportunities were extremely valuable, as these ﬁndings show that those who had a short sent-down duration were signiﬁcantly better off in terms of experiencing later life-course events, compared with those with a longer duration. These ﬁndings remind us that, when state’s coercion was loosened, the existing social stratiﬁcation structures and processes continued to play an important mediating role between state policies and individual life chances, even in a highly volatile political environment. State policy-induced social changes have produced lasting effects on individuals’ subsequent life course and their behavioral patterns. Comparing the group that had no sent-down experience to those that had different rural durations, we ﬁnd signiﬁcant variations in their experience of major life-course events. It appears that the send-down episode had an especially large and negative effect on those who had an extended rural duration: Their marriage and child rearing were delayed signiﬁcantly. They also had fewer advantageous locations in the urban labor force (with respect to certain occupations and work organizations), compared with those who had a shorter rural duration. With regard to the determinants of personal income, on average, the sent-down experience did not appear to have signiﬁcant effect in the three years we selected for analysis. However, the signiﬁcant effects of the interactions between types of workplaces and the sent-down experience in 1993 indicate that the negative consequences of the sent-down experience emerged and may have become more salient in recent years. Ironically, the weakening of the central government is partly responsible for the negative effects of the sent-down experience in recent years. In the aftermath of the sent-down policy, the post-Mao government did adopt a series of policies to stabilize and smooth the subsequent life course of sent-down youth. The evidence of the absence of income differences is consistent with these policies. However, the subsequent advance of economic reforms, especially the increasing autonomy of work organizations

Children of the Cultural Revolution

153

in the determination of economic beneﬁts, may have rendered these governmental policies more or less irrelevant. The ﬁndings also show some positive consequences of the sent-down experience. For instance, a noticeably higher proportion of the sent-down youth had higher educational attainment, as measured by college attainment after 1977. Partly as a result of their educational attainment, we ﬁnd that the sent-down youth, especially those with a shorter rural duration, appeared to have favorable positions (type of occupations and work organizations) in the urban labor force, despite their relative short urban labor force experience. Women with the sent-down experience also had higher income than those women who stayed in urban areas. Two main factors may have accounted for these positive consequences. First, major social changes in the economic reform era have had a great impact on the reshaping of the sent-down youth’s life course on their return to urban areas. The resumption of college entrance examination and the expansion of the urban economy may have smoothed the sent-down youth’s relocation in urban areas. The emerging market economies opened new opportunities in jobs and new economic sectors (the growth of the service sector), which may have alleviated the negative impacts of the sent-down experience. The sent-down experience played a pivotal role in China’s economic transformation in the early years of the post-Mao era. The waves of sentdown youth returning to wracked urban economies demanded, protested, or even used violence in their search for jobs and put enormous pressures on the government. The initial state policy of legitimizing and encouraging the expansion of a private sector in the early 1980s was largely in response to employment pressures generated by the return of the sent-down youth. We will return to these issues when we consider the reciprocal role of ﬂuctuating life chances in the engineering of social changes in the last chapter of this book.

6 Climbing the Political Ladder: Bureaucratic Career Patterns

Only bureaucracy has established the foundation for the administration of the modern law. Max Weber (1978, p. 975) Once the political course is set, cadre is the decisive factor. Mao Zedong (1938, p. 514)

No concept has occupied a more central place than that of the bureaucratic class in studies of social stratiﬁcation under state socialism. Since Djilas’s (1957) early critique of the emerging “new class” in state socialist societies in Eastern Europe, the notion of a powerful and privileged bureaucratic class has been a widely accepted imagery among students of state socialism and it became the starting point for most research on social stratiﬁcation in these societies. This intellectual consensus is not without justiﬁcation, for bureaucrats and bureaucratic organizations have been the cornerstone of the political order in state socialist societies. They serve as the “organizational weapons” of the communist state (Bauman 1974; Schurmann 1968; Selznick 1952) and as the ruling class with the monopoly of power and privileges (Szel´enyi 1978). They provide the institutional basis for accomplishing economic goals in the command economy and are major channels through which the state redistributes manifest and latent economic beneﬁts. Ultimately, it is through bureaucrats and bureaucratic organizations that the socialist state exercises its dominance over the society. As a result, many studies of social stratiﬁcation under state socialism started from and were centered on a critique of the bureaucratic class. 154

Climbing the Political Ladder: Bureaucratic Career Patterns

155

But there is a problem here. We know little about the making of the bureaucratic class with regard to how individuals are selected into, and promoted within, the organizational hierarchy. Let us consider this simple question: Does the bureaucratic class reproduce and perpetuate itself through intergenerational or sponsored mobility, or is it formed through some other basis (e.g., seniority, merit, etc.)? Obviously, implications for class formation and social stratiﬁcation processes are qualitatively different depending on answers to this and other related questions. Although students of Chinese politics have long been interested in bureaucratic career patterns, most studies have focused on national or provincial-level political elites (Barnett 1969; Barnett and Vogel 1967; Lee 1991; Li and Bachman 1989; Oksenberg 1968; Scalapino and Bennett 1972; Vogel 1967). Until recently, there were few comprehensive studies of the career patterns of those lower-level bureaucrats who consist of the majority in the Chinese bureaucracy. In recent years, signiﬁcant progress has been made in theoretical explanations of bureaucratic career patterns, which shed light on the making and transformation of the bureaucratic class in China. In the sociological literature, Walder’s dual-path model (Li and Walder 2001; Walder 1995a; Walder, Li, and Treiman 2000) emphasizes the institutional basis of segmented political and professional career lines. My colleagues and I have also developed arguments that highlight the impacts of macropolitical processes and shifting state policies on bureaucratic career patterns (Zhou 1995, 2001; Zhou, Tuma, and Moen 1996). This chapter reports a systematic examination of the historical evolution of bureaucratic career patterns in China. We focus on a series of issues: (1) the acquisition of political status – Communist Party membership, (2) patterns of entry into the Chinese bureaucracy, and (3) promotion patterns within the Chinese bureaucracy. Until the major changes starting in the early 1990s, most work organizations in China were under the administrative apparatus of the state. Access to and promotion within these organizations as well as redistributive beneﬁts have been regulated by government agencies or appointed authorities. This political control was reﬂected not only in the intervention of supervising agencies but also in the direct presence and dominance of political cadres (e.g., the heads of the local Communist Party branches) in the workplace. As a result, the state and state policies affect bureaucratic career patterns not only in those agencies within the central or local governments but also in those public organizations and business ﬁrms in the state and collective

156

The State and Life Chances in Urban China

sectors. Accordingly, we deﬁne the concept of the Chinese bureaucracy broadly to include government agencies and work organizations such as nonproﬁt organizations and ﬁrms in the state and collective sectors. Similarly, until recent years, administrators, managers, and professionals were all labeled “cadres” in China, and the distinction between the two was not strictly maintained. We use the notion of “bureaucrats” to refer to both administrators and political cadres in the public domain (e.g., heads of Communist Party branches in work organizations) and professionals, because professionals are also an integral part of the Chinese “cadre” system.

introduction: the personnel system in the chinese bureaucracy As one goes through the numerous regulations on the multiple and parallel hierarchical structures of the Chinese bureaucratic system and those policies on recruitment, promotions, and salaries, one cannot help but be amazed at the elaborate, all-encompassing nature of the bureaucracy as if it left no corner untouched, no employee unaccounted for, and no matter unregulated.1 From the head of the state down to the typist in a town government ofﬁce, as long as one is a formal employee subject to the administration of the government system (both central and local governments), one is covered and regulated in this system. And this is only the beginning. As long as one is a cadre or a professional in a work organization owned by the government – no matter what one does or what occupation one belongs to (e.g. a university professor, a journalist, a writer, an engineer, or a librarian), or which sector one works in (e.g., manufacturing, service, education, or residential administration) – one is covered under this system. The cadre system in China is a distinctive bureaucratic system that consists of an administrative hierarchy and a professional hierarchy.

1

The best historical sources on labor policies and regulations in the People’s Republic of China are the three volumes of the Encyclopedia of Labor Policies and Regulations in the PRC, compiled by the Bureau of Policy and Regulation in the Ministry of Labor (Bureau of Policies and Regulations 1990). Chen and Zhang (1992) provide an excellent description of the evolution of the salary system in government agencies and public organizations. Xu and Zhang (1992) is a good source of the personnel policies and practice in the reform era. The descriptions in this section draw on these sources.

Climbing the Political Ladder: Bureaucratic Career Patterns

157

According to the Dictionary of the Modern Chinese (Xiandai Hanyu Cidian), the concept of cadre broadly covers “formal employees in government agencies, military, and public organizations (except for soldiers or those who provide subsidiary assistance).” Until the late 1980s, all administrators and professionals in these two hierarchies were labeled cadres in the ofﬁcial bureaucratic system. Accordingly, there are two distinctive career tracks in the Chinese bureaucracy: One is the ladder of administrative ranks and the other is the ladder of professional ranks, each having its own ascending positions and salary grades for promotion. At the lowest level of the bureaucratic system, all employees are routinely labeled as cadres; the distinction between a cadre or a professional is not carefully maintained. For example, according to the established policies and standard practice, a college graduate who is allocated into a research institution is automatically assigned to a cadre grade in the cadre system, often without any administrative rank. Whether he or she would be promoted along the administrative or professional track depends on further designation of his or her job in that workplace. Many managers hold dual track titles such that the head of a company may have both an administrative rank and a professional title (e.g., senior engineer). The Bureaucratic System Although the bureaucratic system evolved considerably in the forty-ﬁveyear span of our study, its basic structure took shape in the early years of the People’s Republic of China. In 1955, the State Council formally established the salary system intended to uniformly apply to all government employees. The basis of this salary system was a hierarchical system with thirty administrative (salary) grades that covered all administrative positions in the Chinese bureaucracy. Figure 6.1 shows the basic structure of this hierarchical system. First, the grade system is structured along a ladder of administrative ranks, in ascending order: clerks (who do not belong to the cadre system), nonranking cadres, “ke”-rank (head of section), “chu”-rank (head of division), “ju”-rank (head of bureau), and “bu”-rank (head of ministry), and ﬁnally senior governmental ofﬁcials (prime minister and heads of the state). To have a sense of the substantive meanings of these administrative ranks, consider the following standard practice: The head of a county in the Chinese administrative system is usually a cadre at the chu rank. The administrative rank of the head of a work unit depends on the administrative rank of that unit. For

The State and Life Chances in Urban China

158 1 2

Head of

3

the state

4 5

Head of

6

ministry,

7

province, Major city

8 9 10

Head of

11

bureau

12 13 14 15 16 17

“ju” Head of division, county (“chu”)

Head of section, town (“ke”)

18 19

Office

20

worker in

21

section

22

unranked

23 24 25

Other type of workers,

26

e.g. typist,

27

janitor.

28

Not in the

29

“cadre” rank

30

ﬁgure 6.1. The administrative rank system. Source: Based on Chen and Zhang (1992, p. 54, Table 2-28).

Climbing the Political Ladder: Bureaucratic Career Patterns

159

example, if a large city ofﬁcially belongs to a provincial rank (e.g., Shanghai), then the head of the city, like the head of provinces, belong to the bu rank. On the other hand, most large cities belong to the administrative ju rank; hence, the heads of these cities belong to the ju rank. The majority of cadres in the Chinese bureaucracy are at chu or lower levels. Cadres in our sample are mainly composed of bureaucrats at these levels. Second, within each rank, there are more reﬁned grades for differentiation and promotion, and the grades often overlap across adjacent ranks. For example, a ke-rank cadre may be in a grade between the fourteenth and eighteenth grade (in salary), and a chu-rank cadre between the twelfth and sixteenth grade rank (a lower number represents a higher rank). A centralized authority system has a tendency to replicate its governance structure in other spheres of social life. Soon, a professional rank system across all sectors was established that mirrored the national administrative system. This ranking system was designated for each and every professional specialty area – engineers, professors, teachers, artists, writers, coaches, and athletes, to name only a few. Figure 6.2 offers one example with regard to “engineer” track. It is worth noting that the Chinese professional system includes both professionals (e.g., engineers, librarians, etc.) and “technicians” and “assistant technicians,” who are typically “semiprofessionals” in industrialized societies. This is partly due to the historical legacy in the Chinese culture that emphasizes apprenticeship and internal promotion and partly due to the shortage of professional personnel in the early years. This system has been modiﬁed over the years, but the basic structure has remained intact. The most recent reform of the salary/rank system in 1985, for example, adopted a ranking system that was almost identical to the one in 1955 (see Chen and Zhang 1992). The composition of cadres at different levels evolves over time in our sample. In 1993, the distribution of administrative ranks was as follows: 329 were below ke level, 256 were at ke level, 53 and 55 were at the associate chu and chu level, respectively, and 12 were at the ju level. Of professionals, 191 were technicians, 189 were at the assistant engineer or comparable ranks, 190 were at the engineer level, and 72 were at the senior engineer rank. Policies and Practice of Recruitment and Promotion Both recruitment and promotion are heavily regulated by the central and local governments, communicated through pronounced policies and

The State and Life Chances in Urban China

160

Engineer 1

1st-rank

2

engineer

3

2nd rank

4

engineer

5

3rd rank

6

engineer

7 8

4th rank engineer

Technician 1 2 3

technician

4 5

Assistant Technician 1 2 3

Assistant technician

4

Apprentice 1 2

Apprentice

ﬁgure 6.2. The professional rank system. Source: Based on Chen and Zhang (1992, p. 36, Table 2-12).

Climbing the Political Ladder: Bureaucratic Career Patterns

161

administrative directives, and implemented by its personnel ofﬁces at various levels of the bureaucratic apparatus. Often times, state policies and administrative directives explicitly dictate the criteria of recruitment and promotion, quotas for promotion into certain ranks, and the amount of salary increase. As an illustration, the following is an excerpt from a directive on “Cadre Planning and Regulation” issued by the ministry of labor and personnel in 1987 (Xu and Zhang 1992, pp. 606–608): The additional demand for cadres in bureaus or workunits should be ﬁrst met through the internal allocation in the existing cadre pool. If the demand cannot be met using the existing pool, the quota given by the new cadre plan can be used to recruit, on a merit basis, from college and technical school graduates and cadres discharged from the military. In principle, recruitment directly from the society is no longer permitted . . . In principle, public organizations cannot recruit cadres directly from ordinary workers. For exceptional cases, the term-appointment system should be used; the hiring should be conﬁned within the quota structure for cadre and professional positions in that workunit.

Hundreds of such directives are issued each year from the ministries of the central government that regulate the recruitment, promotion, and salary changes of cadres and professionals. These directives generate further documents at lower levels of the corresponding local government agencies that interpret and implement these policies. Advancement in the Chinese bureaucracy takes place typically along either the administrative or the professional ladder in the national system. For the administrative ladder, one moves from unranked cadre to associate ke-rank, to ke, associate chu, chu, associate ju, ju, vice bu, and ﬁnally to bu levels. Similarly, a professional with a college degree usually is promoted along the ranks of assistant engineer, engineer, and senior engineer. These ranks are recognized nationwide, comparable across types of work organizations, economic sectors, and regions. In addition to the national bureaucratic system, another promotion track is promotion within work organizations, from low-level, to midlevel, to high-level managerial and professional positions. These positions are designated based on ﬁrm-speciﬁc structures. They are largely correlated, but do not necessarily coincide, with the national bureaucratic system. The ofﬁcial ranks of these organizational positions are contingent on the speciﬁc organization’s status, size, and property rights ownership. For example, in a large government agency, a chu rank cadre may be only a “mid-level” cadre in that work unit. On the other hand, in a small work organization, a ke rank cadre may be the head of that organization. Also,

162

The State and Life Chances in Urban China

one can be promoted into the position of the head of a work organization, but still does not belong to the ofﬁcial cadre system. Heads of work organizations may be in different ranks because of the administrative rank order of their work organizations in the national system. Within work organizations, there are two career tracks similar to the national bureaucratic system: One is administrative, such as the party secretary or the general manager in the workunit; the other is the technical line such as engineers and senior engineers of the company (usually a vice president in a factory). The implementation of state policies regarding recruitment and promotion relies heavily on bureaucrats in the work organization and those in the immediate supervising agencies. In most cases, promotion along the national administrative/professional system is carried out within work organizations (e.g., government agencies, ﬁrms, or universities) based on routinized bureaucratic policies and practice. In some cases, government policies would speed up or interrupt the routine recruitment and promotion practice, which will be discussed later in this chapter. In contrast, the appointment and promotion of managers within work organizations are more localized and often administered by the supervising government agencies. In what ways are bureaucrats recruited and promoted in the Chinese bureaucracy? Given the apparent stability in personnel management and political control, has the Chinese bureaucracy changed? If it has changed, in what ways and in which directions? These are the main issues to be addressed in the rest of the chapter.

explaining bureaucratic career patterns in urban china: two models To understand the making of the bureaucratic class, one needs to take a close look at bureaucratic career patterns, which consist of both recruitment into and promotion within the bureaucracy. Furthermore, given the importance of Communist Party membership as a political capital, the acquisition of party membership also sheds light on the political selection processes. To motivate our empirical examinations, we ﬁrst consider two theoretical models that address this set of issues. Walder’s Dual-Path Model Walder and his colleagues (Li and Walder 2001; Walder 1995a; Walder, Li, and Treiman 2000) developed a dual path model to explain

Climbing the Political Ladder: Bureaucratic Career Patterns

163

bureaucratic career patterns in state socialist societies. Walder’s model was aimed to address an apparent paradox in the political life chances in these societies. On the one hand, to exercise its political control and implement its policies, the state has to rely on a bureaucratic apparatus at various levels of the administrative system. As Walder (1995a, p. 309) observed: The power to offer career opportunities has long been recognized as a central pillar of Communist rule, either as a system of social control (rewards for loyalty) or as a means of fostering anticipatory socialization and (at least outward) ideological conformity.

The political logic of state socialism implies that political loyalty is rewarded to ensure the effectiveness in carrying out state policies. The bureaucracy as an organizational weapon of the state gives rise to the most salient characteristic of bureaucratic career patterns: entries into and promotions within the bureaucracy are tightly controlled by political processes and by the political criteria of selection (Burns 1987). The state uses privileges and status attached to bureaucratic positions as incentives to ensure political loyalty and compliance to state policies. Thus, political loyalty and the adherence to the party line are the most important criteria in recruitment and promotion in the Chinese bureaucracy, often at the expense of competence and efﬁciency; political screening, such as the recruitment of party membership, is an important gate-keeping device in political selection (Harding 1981; Lee 1991; Walder 1986). On the other hand, the “new class” theory originated in the East European ´ and Szel´enyi 1979) argued that the state socialist govcontext (Konrad ernments actively recruit intellectuals into their governance structures; as a result, educational credentials played a prominent role in one’s upward mobility. Walder’s (1995) dual-path model argues that political selection processes under state socialism involve both political loyalties and educational qualiﬁcations. He proposes a dual-career path model in which one path emphasized “expertise and education (but not politics) and led to professional occupations (with little decision-making authority),” and the other emphasized “political loyalty (but not higher education) and led to inﬂuential leadership positions” (Walder, Li, and Treiman 2001, p. 194). Although Walder’s model has broad implications for other state socialist societies, it has been mainly illustrated and examined in the Chinese context. Walder et al. (2001) presented the most comprehensive examination of the dual-path model using the life history data of a national

164

The State and Life Chances in Urban China

sample of respondents between 1949 and 1996. Their empirical ﬁndings demonstrate that there are two distinctive and segmented career lines for cadres and professionals. Moreover, the processes underlying position attainments along these two career lines are noticeably different. The former is affected by political screening devices such as party membership but unaffected by educational credential (e.g., college education), whereas the opposite is found for the attainment of professional positions (see also, Li and Walder 2001). A Model of Stratiﬁcation Dynamics My colleagues and I (Zhou 1995, 2001; Zhou, Tuma, and Moen 1996) proposed a model of stratiﬁcation dynamics that calls attention to the role of macro-political processes and shifting state policies that generate timevarying opportunities and allocative mechanisms underlying bureaucratic career patterns over time and over the stages of one’s life course. Bian, Shu, and Logan (2002) also developed a similar line of argument. Because this line of argument is also the main theme of this book, we elaborate these considerations in some detail. There are two underlying sources of such stratiﬁcation dynamics. The ﬁrst is the political dynamics reﬂected in state policy shifts. These dynamics affect bureaucratic career patterns through shifting selection criteria in bureaucratic recruitment and promotion. The second is the inherent contention between the centralized state and bureaucratic interests, which often resulted in frequent bureaucratic reforms and political campaigns that disrupted career patterns and reshaped the life chances of the bureaucrats. Political Dynamics and Bureaucratic Career Patterns A major source of changes in bureaucratic career patterns is the marked political dynamics, in the form of shifting state policies, which coexist with stable bureaucratic organizations. These political dynamics have enormous impacts on bureaucratic career patterns over time. First, state policy shifts lead to signiﬁcant shifts in the distribution of opportunities among bureaucrats in different economic sectors, localities, work organizations, and regions. For instance, when the top leaders emphasized political compliance, as in the Mao era, political bureaucrats such as party secretaries (Zhenggong ganbu) were given the highest authority in the work organizations. In contrast, when the post-Mao leaders shifted their agenda to

Climbing the Political Ladder: Bureaucratic Career Patterns

165

economic development, the new “manager-in-charge” policy (Changzhang buzezhi) gave managers more authority over political cadres. Policy shifts alter the life chances of different types of bureaucrats with regard to their status, promotion, and economic rewards. Second, the state relies on bureaucratic organizations, especially the personnel quality of the bureaucrats and their identiﬁcation with the top leaders’ political cause, to ensure effective policy implementation. As Lampton (1987, p. 9) observed: “A principal task of political leadership is to create an institutional and personnel context hospitable to the regime’s speciﬁc policy initiatives.” This characteristic is also captured in Walder’s neo-traditionalist model in China in which “the impersonal standards dictate preferential treatment in return for loyalty and ideological adherence, and standard party leadership practices require the cultivation of stable networks of such activists in all social settings” (1986, p. 6). Ironically, the importance of bureaucratic allegiance implies that, when state policies and priorities shift, the loyalty and competence of the bureaucrats identiﬁed with the “old cause” become a political liability. Along with state policy shift or leadership succession, the top leaders seek to recruit and promote those who are identiﬁed with the new policies by promulgating new selection and promotion criteria. Take the recruitment of Communist Party members as an example. In the earlier period of the People’s Republic of China, the Communist Party emphasized candidates’ family background and their “revolutionary experience” (Lee 1991). The criteria for recruiting Communist party members adopted in the party organization conference in 1962 were to “recruit those activists with good family background, strong ideological stand and qualiﬁed requirements” (Chen and Liu 1991, p. 864). But during the Cultural Revolution, the emphasis adopted by the radical Maoist leaders was “to recruit those proletarian activists with high revolutionary rebellious spirit, and actively engaged in the class struggles” (Chen and Liu 1991, p. 926). In the reform era, selection criteria have changed again, with more emphasis on competence as measured by educational credentials and political loyalty to the political course of reform. As a result, the life chances of the bureaucrats in different cohorts have been altered as a secondary consequence of the political dynamics. “Organizational Weapon” and Bureaucratic Interests: The Dilemma The second major source of changes in bureaucratic career patterns lies in the inherent dilemma of the bureaucracy as the “organizational weapon”

166

The State and Life Chances in Urban China

and the bureaucrats as a social class. Stable bureaucratic institutions foster institution-based bureaucratic interests and patterns of bureaucratic bargaining and factional conﬂicts in the political process (Harding 1981; Lieberthal and Lampton 1992; Shirk 1993; Shue 1988). These interests undermine the authority of the top leaders and threaten the effective implementation of state policies. As a result, there have been continuous struggles between the central government and bureaucratic organizations over policies, local initiatives, and resource allocation. More often than not, these bargains and compromises took place behind the closed doors. But occasionally, drastic policy shifts, such as the decentralization in the early 1960s and the Cultural Revolution period, signaled the undercurrents of major struggles. Over time, we observe frequent political purges that were intended to rectify these “bureaucratic problems” and tighten the political control from the top, which interrupted bureaucratic careers and generated varying life chances. The Cultural Revolution was the peak of this type of struggle, when bureaucratic apparatus at different levels were paralyzed and the bureaucrats were targeted and purged. The political dynamics induced by shifting state policies and by the efforts to combat bureaucratic interests in the communist political order led to shifting opportunities for bureaucratic careers. In this sense, the evolution of the Chinese bureaucracy is reﬂected in the changing composition of bureaucrats, regulated by shifting policies on entry into and promotion within the bureaucracy. To illustrate these considerations, we can highlight three distinctive periods associated with distinctive criteria of political selection in China. 1949–1965. In the early years, the Communist Party emphasized “following the party line” as the main criterion in recruitment of new members into the Communist Party and bureaucracy. For instance, after the antirightist campaign in 1957 that purged nearly thirty-three thousand party members (mostly intellectuals), the Communist Party also recruited a signiﬁcant number of the senior intellectuals who “had good performance during the campaign” (Zhao 1987). As Harding (1981, p. ix) pointed out: “a stiﬂing level of conformity among ofﬁcials, a reluctance to take independent initiatives, a low level of technical and managerial skills, and disillusionment and cynicism at the frequent shift in ofﬁcial line” characterized the Chinese bureaucrats in this political environment. 1966–1979. In the Cultural Revolution, all local party and administrative organizations were paralyzed and reorganized between 1967 and 1971 and a large number of “old cadres” were purged. At the same time, the Maoist leaders actively recruited and promoted their political

Climbing the Political Ladder: Bureaucratic Career Patterns

167

loyalists to ensure the implementation of the radical policies. Accordingly, the selection criteria were shifted to emphasize the recruitment and promotion of those who had “the rebellious spirits.” The party charter adopted in 1973 advocated, “All party members should have the rebellious revolutionary spirits.” As a result, “those who were actively involved in the ‘rebellions’ during the Cultural Revolution were recruited into the party, which became the sources of instability within the Communist Party” (Zhao 1987, p. 387). 1980–1994. In 1980, the new leadership initiated the “reform of the party and the state,” which signaled the beginning of major changes in the Chinese bureaucracy. The new selection criteria emphasized competence (educational credentials) and youth (age restriction). Political loyalty was also emphasized, but it now acquired a new meaning – adherence to the party’s reform policies. Since then, the central government has made a series of institutional reforms to rationalize the Chinese bureaucracy. Policies have been adopted since 1980 to recruit and promote young and educated bureaucrats. In 1982, the government formally established the cadre retirement system, which forced older cadres to retire from their positions. Implications for Empirical Examination Before we discuss the empirical implications of these theoretical models for bureaucratic career patterns, we should note that Communist Party membership has been an important political status in China, and it played an important role in bureaucratic careers. Recruitment into the Communist Party was often the precondition for being promoted in the Chinese bureaucracy. Therefore, we consider the acquisition of party membership as a political status and treat party membership as an indicator of political screening. Walder’s dual-path model and Zhou’s stratiﬁcation dynamics model share a common theme about the importance of the macro-political environment, political processes, and state policies, on bureaucratic career patterns. They differ in that Walder’s model emphasizes stable institutional bases that sustain the dual path of segmented career lines in the Chinese bureaucracy, whereas Zhou’s model highlights macro-political dynamics that are often inconsistent and disruptive of the institutionbased political processes. These two models are complementary in that they address different aspects of the Chinese bureaucracy and it is likely that we observe empirical patterns that are consistent with both

168

The State and Life Chances in Urban China

models. In some respects, these two models have distinctive implications with regard to bureaucratic career patterns. We now consider the empirical implications of these two models. Walder’s dual-path model yields explicit predictions with regard to the two bureaucratic career lines. As Walder (1995) argued, there are two segmented career lines associated with distinctive life chances in the Chinese bureaucracy: the administrative and the professional career lines. Following Walder’s arguments, we expect to ﬁnd that political screening is especially enforced in the recruitment and promotion of political career lines (e.g., administrators), whereas educational credentials play a larger role in the promotion of professional career lines. Thus, we expect to ﬁnd that political criterion (e.g., party membership) plays a more important role in administrative recruitment and promotion, whereas educational credentials are more important for recruitment and promotion in professional career lines. Moreover, the administrative career line is the fast track of the two, as reﬂected in the higher promotion rate. To examine the effects of stratiﬁcation dynamics, again we follow the strategy to use historical periods to compare and contrast the effects of the key covariates over time. In the following, we derive the empirical implications in this light. Shifts of State Policy and Variations in Opportunity Structures Over Time. Because of the tight control of personnel management in the Chinese bureaucracy, the impact of policy shifts on political opportunities is immediate and substantive. Shifts in state policies often lead to dramatic changes in both the size and the structure of opportunities, and inadvertently alter the life chances of bureaucrats in different historical periods. So, we expect that the shifts of state policy generate signiﬁcant variations in the availability of political opportunities across historical periods, giving rise to marked differences in bureaucrats’ life chances. Variations in Selection Criteria. One important implication of the preceding discussion is that, although the political selection processes and the institutional structure of personnel management have largely remained intact over time, speciﬁc criteria of political selection have changed in accordance with shifting state policies in distinctive periods. For example, education and seniority have been two salient and conﬂicting criteria of qualiﬁcation used to allocate political opportunities in the Chinese bureaucracy. In the Chinese politics, educational qualiﬁcations were often used as indicators of political loyalty and age as an indicator of cohorts

Climbing the Political Ladder: Bureaucratic Career Patterns

169

with different political allegiances. As a result, education and age are often used as explicit selection criteria to target particular groups of political followers in different historical periods. In the Mao era, individuals with high educational credentials were often treated with suspicion. The Maoist leaders during the Cultural Revolution adopted policies of promoting young cadres that were identiﬁed with the radical policies to replace old ones. On the other hand, in the Mao era, seniority was also emphasized in one’s career development under the egalitarian policy. But, in the reform era, the new cadre policies especially emphasize educational credentials and youthfulness in bureaucratic recruitment and promotion. Therefore we expect that the effects of education and seniority vary systematically with shifting state policies across historical periods. Speciﬁcally, we expect to ﬁnd that educational credentials were less important in gaining access to political opportunities in the Mao era than in the reform era and that seniority (e.g., age effect) had a more positive effect in the Mao era than in the reform era. Varying selection criteria should also be reﬂected in changes in gender roles across historical periods. Although the communist ideology has long advocated gender equality, the radical Maoist leaders forcefully implemented this policy during the Cultural Revolution, along with other destratiﬁcation policies. In contrast, the policy of promoting gender equality was considerably weakened along with the decline of the central authority in the reform era. Therefore, we expect to ﬁnd that there are signiﬁcant variations in gender inequality in access to political opportunities over time. In particular, gender inequality was least salient during the Cultural Revolution and most salient in the reform era. Cohort Effects in Gaining Political Status. Changes in recruitment criteria over time led to the selection of different cohorts of bureaucrats into the Chinese bureaucracy and accordingly different life chances for them. Because of the lifetime employment practice in redistributive economy, once they have entered the bureaucracy, most people stayed for the rest of their careers. Political purges that forced exits from the bureaucracy did occur, but they were less common in China. Preferential treatments often took the form of reallocating opportunities among different cohorts of bureaucrats. Those recruited in one period but lost favor in another period may have seen their careers stalled due to shifts in promotion criteria. For instance, the radical policies of the Cultural Revolution purged old cadres and rush-promoted young cadres with “rebellious spirits.” Similarly, in the post-Mao era, the new policies promoted young and

170

The State and Life Chances in Urban China

educated cadres and discriminated against the cohort of bureaucrats recruited during the Cultural Revolution (Zhou 1995). This consideration leads us to predict that shifting selective criteria over time led to different probabilities of acquisition of party membership and promotion for different cohorts of bureaucrats. Speciﬁcally, we expect that cadre policies of a particular period favored the cohort recruited in that period relative to those in earlier cohort(s). In particular, in the reform era, the cohort of cadres recruited in the Cultural Revolution is especially disfavored in their career development than other cohorts, because the radical policies in the Cultural Revolution are in stark contradiction with the reform policies. Intergenerational Status Inheritance. A central issue for our understanding of the making of the bureaucratic class is about the “class reproduction.” Class reproduction through intergenerational mobility has been a central issue in social stratiﬁcation studies in market societies and it is useful to take into consideration this potential causal process. For this purpose, we include measures of the father’s education and occupational status in our empirical analyses to examine whether intergenerational mobility provides a mechanism for class reproduction. Finally, as we argued earlier, one important source of political dynamics is the bureaucratic resistance to those state policies that threaten their own interests. The prevalence of paternalist-clientele relationships implies that the implementation of these state policies, such as shifting selection criteria, may have been resisted by the current bureaucrats whose rise to power was associated with different state policies in earlier periods. Therefore, the implementation of current policies may have been delayed, distorted, or rendered ineffective. One interesting comparison is between formal promotion patterns along the national-level bureaucratic ladders and those within work organizations. The former is based on the ofﬁcial bureaucratic rank system in the Chinese bureaucracy, and its promotion has been more centralized through personnel ofﬁces of the central government. In contrast, promotion events within work organizations tend to be more decentralized and to vary across organizations and places. Therefore these two types of promotion events may be regulated by different bureaucratic processes. One speculation is that, given the central control of formal bureaucratic promotions, we would expect such events to be more sensitive to changes in state policies. On the other hand, bureaucratic interests are organized along the ladder of formal bureaucratic administrations; thus,

Climbing the Political Ladder: Bureaucratic Career Patterns

171

bureaucratic resistance, if effective, is more likely to succeed in promotion patterns in the formal bureaucratic apparatus. Therefore, if state policies exert a strong effect, we expect to ﬁnd signiﬁcant changes in formal bureaucratic promotions relative to changes in workplaces across periods. On the other hand, if institution-based bureaucratic resistance is effective, we expect to ﬁnd fewer changes in promotion patterns along formal bureaucratic positions than those in workplaces. In the rest of this chapter, we examine three important aspects of the bureaucratic career: the acquisition of political capital – Communist Party membership, entry into and promotion in the Chinese bureaucracy. Our presentation follows the typical sequence of events over one’s life course: entry into the bureaucracy, entry into the Communist Party, and bureaucratic promotion.

descriptive patterns across historical periods We ﬁrst consider the distribution of bureaucratic career opportunities over time. Figure 6.3 displays the hazard rate of entry into the Communist Party and into the administrative and professional occupations. Before we discuss these patterns, some aspects of the data should be

0.05 0.045 0.04 Hazard rate

0.035 0.03 0.025 0.02 0.015 0.01 0.005 0 50

55

60

65

70

75

80

85

90

95

Year administrator

professional

party

ﬁgure 6.3. Hazard rates of entry into CCP, administrative, and professional occupations (1950–1993).

172

The State and Life Chances in Urban China

highlighted. First, the information used in this ﬁgure (and others in the following sections) is based on those employees still in labor force. Because our sample is conﬁned to respondents aged between twenty-ﬁve and sixty-ﬁve, older individuals have a higher attrition rate (i.e., a larger proportion of the early cohort may have not fallen in our sampling framework). Second, because of the nature of retrospective data collection, the size of the employees increased over time, as more people joined the labor force over time in our sample. The high ﬂuctuations in the earlier years are largely due to relatively fewer events and smaller sample size. These cautionary observations notwithstanding, the hazard rates in Figure 6.3 provide useful information about the distribution of opportunities over time. First, it appears that the three hazard rates more or less track each other, suggesting that they all respond to the same macro-political processes. In particular, the rate of entry into the party and that into the administrative occupation, in most cases, closely mirror each other, indicating that the same processes govern these two types of events. Second, the ﬂuctuations in the hazard rates indicate signiﬁcant variations in opportunities over time. For example, there appeared to be an expansion of opportunities in the 1950s due to the expansion of the bureaucracy, but there was a noticeable decrease in the early 1960s, in the period of economic contraction. The rates also were noticeably lower in the early phase of the Cultural Revolution (1966–1969). In the early years of the postMao era, 1978–1981, all three rates were also low, as the new leadership overhauled the recruitment process. Since then, we observed a signiﬁcant, higher rate in the 1980s. We emphasize that the rates reported here are somewhat underestimated because we included all employees in the labor force, which has increased substantially in the 1980s, due to the return of the sent-down youth. Figure 6.4 displays hazard rates of promotions along administrative and professional ladders in the national bureaucratic system over time. Here the risk set is deﬁned to be those already in each category at the time. Overall, the reform era witnessed the largest increase in promotion rates in both career tracks, and the rates were lowest in the Cultural Revolution period, when bureaucratic career patterns were severely disrupted. Similar to Figure 6.3, the two promotion trends closely resemble each other, in response to common macro-political processes across these historical periods. Variations across historical periods portrayed in these ﬁgures indicate noticeable changes in opportunities and in life chances among bureaucrats

Climbing the Political Ladder: Bureaucratic Career Patterns

173

0.1

Hazard rate

0.08

0.06

0.04

0.02

0 50

55

60

65

70

75 Year

administrative

80

85

90

95

professional

ﬁgure 6.4. Hazard rates of promotion in national bureaucratic system (1950– 1993).

over time. Given these variations, what are the selection processes that channel employees into these career lines and positions? Because entry into types of jobs is often the earliest event one encounters in one’s life course, we begin our analysis with the events of entry into cadre and professional careers.

entry into cadre and professional careers We now examine patterns of access to the Chinese bureaucracy – entry into the administrative and professional occupations – across periods. Both administrators and professionals are categorized as cadres. They are high-status occupations in China’s labor force and only a small proportion of the labor force participants have access to these occupations. In many instances, entry into a cadre career starts with the respondent’s ﬁrst job. Thus, our focus here is similar to our analysis of entry into ﬁrst jobs in Chapter 4. But, this set of analyses differs from those reported in Chapter 4 in several ways. First, the dependent variable here is the rate of experiencing the entry event (into the administrative or professional occupations) relative to no such event; in Chapter 4, the dependent variable is the rate of entering the administrative or professional occupations,

174

The State and Life Chances in Urban China

relative to entering the worker occupation.2 Second, in this chapter, we allow for the possibility of entering administrative and professional occupations from other occupations by those already in the labor force. Finally, in Chapter 4, we included both spouses in the analyses (to avoid too few cases in some job destinations), whereas we focus only on the respondents in this chapter. This set of analyses allows us to identify more broadly sources of the personnel supply to the Chinese bureaucracy and the underlying recruitment processes. Table 6.1 reports the logistic regression parameter estimates of the determinants of entry into the administrative occupation (the left panel) and the professional occupation (the right panel) across the three periods. We discuss the results by contrasting patterns across these two occupations. The entries in the table give the effects of the corresponding covariates on the log-odds (or “log-rate”) of experiencing the entry event. The Overall Trend. To have a more meaningful interpretation of the intercepts, we centered the age at twenty-ﬁve. Thus, the intercept indicates the overall log-rate for a male, twenty-ﬁve-year old worker with a middle school or lower education, who worked in a collective ﬁrm in a large city in that period. The intercepts across these periods show that there was a low rate of entry into these two occupations in the ﬁrst period, and there was no signiﬁcant trend in the second period. In the third period, there was a signiﬁcantly higher rate of entry into the administrative occupation, but not into the professional occupation. Gender. There are striking differences between men and women in their rates of entry into these two occupations. Female employees had a lower rate of entry into the administrative occupation but a higher rate of entering professional occupations in all three periods. It is important to note that, because we include semiprofessional jobs (e.g., nurse, technician) in the “professional” category, the higher rate of entering professional occupations for women does not necessarily mean that there is a gender advantage in favor of women in this regard. Rather, female employees are more likely to have concentrated in gendered jobs. Seniority and Education. Age had a signiﬁcant and negative (ﬁrstorder) effect in the ﬁrst period for recruitment into both occupations, 2

The reason for these different formulations is that, in Chapter 4, all those in the risk set eventually experience an entry event; thus, their entry events can be compared with the common category of workers. In the entry events in this chapter, a person may enter an administrative (or professional) occupation or he or she may stay in his or her current occupation, if he or she is already in the labor force. Thus, it is not feasible to use a common entry event as the reference category (such as entry into the worker occupation).

175

Intercept Female Age Age2 /100 education High school College occupation Professional Ofﬁce worker Service worker Military service In college Farm work organization Government Public organization State ﬁrm

Covariates

0.743∗∗ 0.676∗∗ −0.314 0.762∗∗ −0.156 0.741∗∗ 2.398∗∗ 0.538∗ 0.711∗∗ 0.174 0.141

−1.045∗∗ −0.104 −0.582† −0.953∗∗ 1.584∗∗ −1.255∗∗

0.959∗∗ 0.099 0.930∗∗

0.181 −0.538∗∗ 0.018 0.017

1966–1979

1.149∗∗ 0.890∗

−4.357 −0.623∗∗ −0.050∗∗ 0.414

∗∗

1949–1965

0.422∗ 0.093 0.080

−0.277 1.015∗∗ 0.066 1.715∗∗ 3.256∗∗ 1.236∗∗

0.789∗ 1.363∗∗

3.232 −0.774∗∗ 0.100∗∗ −0.441∗∗

∗∗

1980–1993

Entry Into Administrative Occupation

0.492† 1.311∗∗ −0.847∗∗

−3.079∗∗ −0.733∗ −0.768∗∗ −2.011∗∗ 2.157∗∗ −0.727∗∗

2.843∗∗ 2.415∗∗

−4.137 0.255∗ −0.163∗∗ 0.153

∗∗

1949–1965

0.436† 1.277∗∗ −0.079

−0.731∗∗ −0.436 −0.649∗ −0.508 2.777∗∗ −0.217

1.531∗∗ 1.830∗∗

1.049 0.405∗∗ −0.093∗∗ 0.060

1966–1979

(continued)

0.156 1.053∗∗ −0.148

−0.821∗∗ −0.894∗∗ −0.798∗∗ −0.657 2.952∗∗ 0.642∗

1.586∗∗ 2.195∗∗

1.479 0.338∗∗ 0.010∗∗ −0.258∗∗

1980–1993

Entry Into Professional Occupation

table 6.1. Parameter Estimates of Logistic Model for Entry Into the Cadre Occupation, by Period

176 0.017 0.059 269 −1,479.4 23

−0.124 −0.259† 473 −2,084.5 23

0.385∗∗ −0.300 0.283† 0.048

0.445∗ −0.074 −0.004 −0.255

0.504 0.176 0.545∗∗ 0.519∗

−0.042 −0.106 206 −923.8 23

0.267∗ 0.151 0.055

1980–1993

−0.036 0.027 0.288

1966–1979

0.562∗ −0.127 0.007

1949–1965

0.207 0.200 364 −956.2 23

0.022 0.157 0.326∗ 0.343∗

0.083 0.496∗∗ −0.257

1949–1965

0.061 0.176 356 −1,567.7 23

0.268 0.498∗ 0.337∗ 0.355∗

−0.064 0.140 0.044

1966–1979

−0.018 0.209 572 −1,932.9 23

0.178 0.651∗∗ 0.178 0.151

0.222† 0.341∗∗ 0.149

1980–1993

Entry Into Professional Occupation

∗

p < 0.01 p < 0.05 † p < 0.10 Note: Junior high school or below is the reference category for education; production worker, for occupation; collective ﬁrm, for work organization; and large city, for city location. This set of analyses also include indicators for those who had no prior job or work experience for statistical control (this gives two additional degrees of freedom to the estimated model).

∗∗

father’s education Middle school High school/college Missing value father’s occupation Cadre Professional Farmer Missing value location Medium city Small city Number of events Log-likelihood d.f.

Covariates

Entry Into Administrative Occupation

table 6.1 (continued)

Climbing the Political Ladder: Bureaucratic Career Patterns

177

2 1.8 1.6 1.4 rate

1.2 1 0.8 0.6 0.4 0.2 0 18

23

28

33

38

administrative

43 age

48

professional

53

58

63

party

ﬁgure 6.5. Age effects on entry into CCP, administrative, and professional occupations (1980–1994).

indicating that most recruitment events occurred at the early stage of one’s career, probably at their entry into the labor force. This trend continued into the second period for professional recruitment as well. For administrators, age played no signiﬁcant role in the second period. This may be due to the radical policies of the time that recruited cadres from ordinary workers already in the labor force. In the reform era, as Figure 6.5 shows, age had an inverted-U shape for entries into both occupations, but the age restriction is stronger for the professional occupations, which peaked in the late twenties, whereas age effect on entry into administrative jobs is around mid-thirties. That is, entry into professional occupations was more sensitive to age restriction. Education contributed signiﬁcantly to the rate of entry into both types of occupations. Those with a high school diploma or a college degree had a signiﬁcant and higher rate of entering the administrative occupation in all three periods. This pattern partly reﬂected the institutional practice in the Chinese bureaucracy, whereby college graduates were routinely placed into cadre jobs. Consistent with the dual-path model, education played a far more important role for entering professional careers than for entering administrative careers. But the impact of political dynamics was also evident: The effects of educational levels were noticeably smaller

178

The State and Life Chances in Urban China

in the Cultural Revolution and the effect of college education substantially increased in the post-Mao era, especially for administrative recruitment. These patterns were consistent with the shifting cadre policies that discriminated against intellectuals in the Cultural Revolution and those that emphasized educational credentials in the post-Mao era. Other Occupational Locations. There are several salient patterns with regard to the supply of administrators and professionals from other occupational origins. First, consistent with the dual path arguments, professionals and administrators had a lower rate of shifting to the other occupation, indicating that these two were segmented career patterns. On the other hand, the shifts from professional to administrative jobs in the second and third periods were not statistically different from the reference category of workers. In other words, there were some boundaries between cadres and professionals, but the barriers were not sharply drawn and, as often observed in practice, there were ﬂows from professional to administrative jobs, but not vice versa. Second, ofﬁce workers, military personnel, and school graduates were the largest supply to the administrative lines in the Chinese bureaucracy, especially in the postMao era. In contrast, college graduates have been the sole supply into professional occupations, among all occupational origins. Third, despite the ideological claim of the importance of the working class, except for the ﬁrst period, neither service workers nor production workers had any advantage in becoming a cadre, relative to other occupations.3 Not surprisingly, there was a higher rate of cadres in government agencies in all three periods; similarly, we ﬁnd a higher rate of professionals in public organizations, compared with those entering other types of work organizations. Intergenerational Inheritance. Finally, to examine the potential source of career patterns in light of intergenerational mobility, we examined the effects of the father’s education and occupational status. There is evidence of “inheritance” effects in that the children of cadres and professionals are more likely to follow their father’s footsteps and enter the same occupations. These effects emerged in the Cultural Revolution period, even

3

Although those from rural areas (farms) had higher rates of entering professional occupations in the ﬁrst two periods (and in the ﬁrst and the last period for entry into the cadre occupation), this was due to our sampling practice. Those living in rural areas could migrate into the city mainly by obtaining college education, which channeled them into the cadre occupation. Thus, this ﬁnding does not represent a general pattern for rural residents.

Climbing the Political Ladder: Bureaucratic Career Patterns

179

when destratiﬁcation policies were most stringent. On the other hand, the role of father’s education on career lines was negatively affected by the radical policies in the Cultural Revolution period, in which high educational credentials changed from resources to liabilities. That explains why the father’s high educational status contributed to one’s entry into professional jobs in the ﬁrst and third periods, but had no signiﬁcant effect in the second period.

recruitment into the communist party4 Membership in the Communist Party is an important political asset for social mobility in state socialist societies. As we noted before, there have been signiﬁcant changes in the selection criteria for party membership across historical periods. Thus, patterns of party membership recruitment over time capture the evolving processes of allocating political opportunities. Because eighteen is the minimum age at which one can join the party, we included only those respondents who were eighteen years old or above in this analysis. Table 6.2 reports the parameter estimates of the covariates in the logistic model of becoming a party member.5 Gender. Throughout the historical periods, females had lower rates than males. The gender gap was the smallest during the Cultural Revolution, and the largest in the reform era, suggesting signiﬁcant variations across historical periods, as consistent with changes in state policies. Seniority and Education. We speculated before that age effects are likely to vary with shifting state policies across historical periods. In the ﬁrst two periods, it appeared that younger people were more likely to be recruited 4

5

There is no information on party membership from the six cities in the two southern provinces of Guangdong and Jiangsu, which consist of about 20 percent of the total sample. The analysis of entry into the party excludes samples from these two provinces. To include all respondents of the representative sample in our study, we do not include party membership in the analyses of entry into and promotion in the bureaucracy. In our preliminary analysis, we estimated the same set of models with party membership using the sample of fourteen cities where party member information is available. The coefﬁcients for other covariates are similar to those reported in this chapter. The high percentage of party members in our sample is due to the fact that there is a larger proportion of residents in large cities in our sample, who have better access to party membership. It also reﬂects the fact that there has been a signiﬁcant increase in party membership in urban China. According to ofﬁcial statistics (Zhao 1987, p. 446), in 1981, 17.7 percent of the employees are party members in state-owned ﬁrms in manufacturing and transportation industries; 18.4 percent of the employees in state ﬁrms are in the commerce and ﬁnancial sectors.

180

The State and Life Chances in Urban China

table 6.2. Parameter Estimates of Logistic Model for Entry Into the Communist Party, by Period Covariates Intercept Female Age Age2 /100 education High school College occupation Cadre cohort 49–65 Cadre cohort 66–79 Cadre cohort 80–94 Prof. cohort 49–65 Prof. cohort 66–79 Prof. cohort 80–94 Ofﬁce worker Service worker Military service In college Farmer work organization Government Public organization State ﬁrm father’s education Middle school High school or above Missing value father’s occupation Cadre Professional Farmer Missing value city location Medium city Small city Number of events Log-likelihood ∗∗

1949–1965

1966–1979

1980–1994

∗

6.212 −0.234† −0.005 −0.157†

6.428∗∗ −0.653∗∗ 0.088∗∗ −0.324∗∗

0.003 0.050

0.299∗ 0.165

0.469∗∗ 1.031∗∗

1.781∗ — 0.273 —

1.348∗∗ 1.928∗∗ — 0.186 1.025∗∗

1.006∗ 0.653∗ 1.632∗∗ 0.313 0.428

0.691∗ 0.295 2.379∗∗ 0.669 0.485†

1.452∗∗ 0.991∗∗ 1.354∗∗ 0.871∗∗ 0.662∗∗ 0.921∗∗ 0.518† 0.093 2.270∗∗ 1.070∗∗ 0.051

1.818 −0.320† −0.036† −0.938∗

0.515† −0.052 0.180

1.063∗∗ 0.620∗ 0.472∗

0.490∗ −0.209 0.001

−0.393 −0.475 0.211

−0.220 −0.405 −0.058

0.111 −0.062 0.206

0.577† −0.238 0.239 −0.124

0.274 −0.375 0.525∗∗ 0.241

0.137 −0.214 0.348†

0.026 0.038 210 −912.1

0.28 0.36† 301 −1,395.6

0.370∗ 0.085 312 −1,632.6

p < 0.01 p < 0.05 † p < 0.10 Note: Middle school or below is the reference category for education; production worker, for occupation; collective ﬁrm, for work organization; and large city, for city location. ∗

Climbing the Political Ladder: Bureaucratic Career Patterns

181

into the party. In the third period, we observed an inverted-U shape, which peaks in the late thirties (see Figure 6.5). The fact that the age effects were similar for the rate of entry into the party and the rate into the cadre occupation indicate that the processes underlying these two types of events were very similar. The ﬁndings show that age restrictions in allocating political opportunities were most effective in the reform era. The negative effect of age in the early period reﬂected an era of expanding opportunities in the nation-building process when younger people were recruited into the party. Educational effects tell a similar story. In the ﬁrst period, education had no effect on entry into the party and only high school education had a signiﬁcant rate in the second period. In the reform era (the third period), both high school and college graduates had signiﬁcant, higher rates of being recruited into the Communist Party than those with middle school or lower education. These results are consistent with the state policies in the Mao era that discriminated against intellectuals. They also show that competence (educational qualiﬁcation) has become an important criterion in the reform era. Here we observe clear evidence of shifting selection criteria between the Mao and post-Mao regimes. Cohort Effects. To further explore the impacts of shifting state policies, we included a set of dummy variables to estimate varying life chances associated with cohorts of cadres and professionals, respectively, based on the historical periods in which they were recruited into the cadre occupation. The ﬁrst cohort consists of those cadres (professionals) recruited into their occupations between 1949 and 1965. The second cohort consists of those cadres (professionals) recruited into their occupations during the Cultural Revolution and its immediate aftermath, between 1966 and 1979. The third cohort includes cadres (professionals) recruited into their occupations after 1980. We speculated before that varying political selection criteria should be reﬂected in distinctive cohort effects for cadres recruited in different historical periods. The cadre cohort effects are consistent with this argument. Compared with production workers (the reference category), being a cadre signiﬁcantly improved one’s access to party membership, as indicated by the signiﬁcant, positive effects for all cadre cohorts over time. However, there were noticeable variations in the magnitude of effects among the three cohorts. During the Cultural Revolution, the cadres recruited in that period (1966–1979) had a higher probability of

182

The State and Life Chances in Urban China

entering the Communist Party than did those in the ﬁrst cohort. In contrast, in the reform era, the Cultural Revolution (1966–1979) cohort had the lowest probability of entering the Communist Party, among the three cadre cohorts. This is consistent with the new cadre policies that discriminated against those recruited under the radical policies during the Cultural Revolution. We observed parallel patterns for the professional cohorts across the three periods as well. Although educational credentials played no significant role in party recruitment during the Cultural Revolution, professionals in the second cohort had a signiﬁcant, higher rate of becoming party members. One plausible explanation is that, during this period, the Maoist leaders actively promoted their political followers, many of whom were “red guards” in college or in the professional-concentrated sectors (media, education). In contrast, “old” professionals in the ﬁrst cohort were not favored in the recruitment process. Other Occupational Locations. With respect to the effects of other occupational statuses, several ﬁndings are worth noting. First, ofﬁce workers, often the preparatory force for cadres, had a higher rate of entry into the party. Second, as is well known in China, military service offered better opportunities for acquiring party membership, compared with the reference category of production workers. Finally, it is also noticeable that educational qualiﬁcation was especially favored in the reform era: A college degree and those in college had the largest gains in the third period, relative to other occupations. Type of Work Organizations. Not surprisingly, employees in government agencies had a signiﬁcant, higher rate than the reference category of those in collective ﬁrms in all three periods. Employees in both public organizations and state ﬁrms showed a higher rate during the Cultural Revolution period. We speculate that this was caused by the radical policies of the time that actively recruited rebellious red guards (most of them concentrated in public organizations) and members of the working class (most of them concentrated in state ﬁrms). Social Origins. There is no evidence that social origins, as measured by the father’s education or occupational status, had any systematic effects on the acquisition of party membership.

promotion patterns in the chinese bureaucracy In state socialist societies, political power ultimately rests on one’s bureaucratic position. Promotion is especially important because positions

Climbing the Political Ladder: Bureaucratic Career Patterns

183

provide access to redistributive power and beneﬁts. The following analyses examine patterns of promotion in the Chinese bureaucracy. Speciﬁcally, we consider two types of promotions. First, we analyze patterns of formal promotion along the nationally established bureaucratic ladders for cadres and professionals, with comparable ranks across workplaces and localities. These promotions, although taking place within a workplace, are promulgated by the central government and regulated by state policies. For this set of analyses, we include those already in administrative or professional occupations in their respective risk set. Second, we examine patterns of promotion within workplaces, specifically promotion into low-, mid-, and high-ranking managerial positions in a work unit. Compared with formal bureaucratic promotion, promotion along organization-speciﬁc managerial positions within a workplace is more decentralized and more sensitive to local circumstances. For this analysis, we included all those who were employees in a work organization, because employees in the workplace may be promoted into managerial positions, regardless of whether they held nationally promulgated administrative ranks. Because bureaucratic promotions may occur more than once in one’s career, these repeated events are likely to be correlated with each other. For this reason, we adopt the generalized estimation equation (GEE) model that allows for this assumption in the analyses of promotion events (see Chapter 2 for details of the model). We also consider the role of party membership in promotion. Because of the missing information on party membership in six cities, we estimated the effects of party membership separately, using the subsample of fourteen cities where party membership information is available. For the convenience of presentation, we report the party membership effects in the same tables for the whole sample. Promotion Patterns in Formal Bureaucratic Ranks Table 6.3 reports the GEE parameter estimates from the analyses of promotion along formal administrative and professional hierarchies. The ﬁrst panel reports estimates for the promotion patterns for administrative career line and the second panel reports estimates for professional career line. Consistent with our theoretical interest, we estimated these patterns across the three periods. Because there were few professional promotion cases in the 1949–1965 period, we combined the ﬁrst two periods for professional promotion analysis.

184 0.168 0.902∗∗ 0.680∗∗ 0.032 −0.981∗∗ −0.636 −2.156∗ — 0.246 — −0.073 0.323 0.315

— — —

−0.152 0.862 0.607

0.106 −0.143 −0.295

— −0.767∗∗ −1.457∗∗

0.231 1.067∗∗ 1.072∗∗ −0.046∗∗ −1.159∗∗ −1.604∗∗ −2.531∗∗

10.376∗∗ −0.684∗∗ 0.132∗∗ −0.383∗∗

−7.447∗∗ −0.691∗ 0.092∗ −0.306†

−0.671 −0.435 −0.012 −0.974

Intercept Female Age Age2 /100 education High school College Party membershipa Duration in position Rank 2 Rank 3 Rank 4 cadre/professional 49–65 cohort 66–79 cohort 80–94 cohort father’s education Middle school High school or above Missing value

0.274 1.352∗∗ 1.287∗∗ 0.078† — — —

1980–1994

1966–1979

1949–1965

Covariates

Promotion in Formal Bureaucratic Rank

— −0.482∗∗ −0.659∗∗ −0.273† −0.102 −0.085

−0.944∗ −0.693† 0.495

0.776∗∗ 1.766∗∗ 0.460∗∗ −0.017∗ −0.123 −1.341∗∗ —

1.104∗ 2.386∗∗ 0.451† −0.027 −0.519† −2.739∗∗ — — −0.128 —

−4.210∗∗ −0.299∗∗ 0.046∗∗ 0.061

1980–1994

−7.139∗∗ −0.343 0.132∗∗ −0.232

1949–1979

Promotion in Formal Professional Rank

table 6.3. GEE Estimates of the Determinants of Promotion Patterns in the Chinese Bureaucracy

185

−0.467 −0.632† 108 706 −508.3 21

−1.106∗∗ −0.440∗ 248 901 −1,039.7 23

−0.329 −0.327 105 562 −487.3 22

0.355 −1.701 42 299 −191.5 18

−0.216 0.216 0.301

0.759∗∗ 0.035 0.009

0.456 0.254 0.190

0.516 1.104† 1.340∗

0.023 0.370 −0.609∗ −1.036

0.244 0.216 0.462∗ −0.227

0.080 −0.241 0.090 −0.387

−0.750 0.776 0.278 0.282

−0.225 −0.481∗∗ 427 1050 −1,650.0 22

0.489∗ 0.689∗∗ 0.530∗∗

−0.100 −0.114 −0.625∗∗ −0.530∗∗

∗

p < 0.01 p < 0.05 † p < 0.10 a The coefﬁcients associated with party membership are estimated using data from fourteen cities. Note: Middle school or below is the reference category for education; production worker, for occupation; collective ﬁrm, for work organization; and large city, for city location. 49–65 cadre cohort is the reference category for cadre cohorts in the ﬁrst panel, and 49–65 professional cohort is the reference category for professional cohorts in the second panel of analysis.

∗∗

father’s occupation Cadre Professional Farmer Missing value work organization Government Public organization State ﬁrm city location Medium city Small city Number of events Number of cases Log-likelihood d.f.

The State and Life Chances in Urban China

186 7 6

promotion rate

5 4 3 2 1 0 18

28

38

48

58

age administrative

professional

workplace

ﬁgure 6.6. Age effects on promotion rates, 1980–1994.

Gender. Female employees had a lower rate of being promoted in both formal administrative and professional ladders, and this pattern appears to have accelerated in the reform era (the third period). Interestingly, the gender gap over time was somewhat different from that observed in the previous analyses of entry into administrative and professional occupations. For instance, gender inequality in promotion (for administrative positions) seemed more salient during the Cultural Revolution than in the reform era, suggesting that the egalitarian policy during the Cultural Revolution did not help women gain access to positional power. Moreover, although there is a high concentration of female employees in the professional jobs, as shown in the previous analyses, females have been consistently disadvantaged in promotion, especially in the reform era. Seniority and Education. Age effects on promotion rates differed for the two career lines. Its effect on administrative promotion follows an inverted U-shape, whereas the effect on professional promotion showed a monotonic increasing trend. We plot the two patterns for the third period in Figure 6.6. The promotion rate for administrative ranks shows a bell-shape, increasing the rate between age eighteen and forty and then decreases with age. In contrast, the promotion rate along professional ranks increases at a slower rate in the earlier stages but accelerates along time over one’s entire career. It seems that the age effects on promotion

Climbing the Political Ladder: Bureaucratic Career Patterns

187

are opposite of their effects on the entry events: Age restriction is more salient in entry into professional occupations, whereas age restriction is more salient in promotions along the administrative career line. Clearly, the underlying processes for these two career lines are signiﬁcantly different, with more frequent replacements in the political line than in the professional one. With regard to educational qualiﬁcations, again the patterns are similar for both processes. Educational credentials (especially a college education) had a signiﬁcant and positive effect on promotion throughout the three periods. Similar to its effect on the entry rate into administrative and professional occupations, college education also contributed to a higher rate of promotion in all three periods. It is noticeable that, education is more important for promotion along the professional ranks than along the administrative ranks, consistent with Walder’s dual-path model. For promotions along the administrative ranks, college education had the smallest effect during the Cultural Revolution. Party membership contributed to the promotion rates along both tracks across the three periods. As expected, party membership played a more salient role for promotions in the administrative career line. But the party membership effect was smaller in the Cultural Revolution period. We suspect the reason is that during that period, many old cadres, who were old party members, were targeted and purged. Duration in the current position had varying effects on promotion over time. For administrative positions, it increased the rate in the ﬁrst period, had no signiﬁcant effect in the second period, and decreased the rate in the third. This pattern implies that seniority contributed to promotion in the early period, whereas it became a liability in the reform era. We observe a similar pattern for professional promotions. These patterns are consistent with state policies that promote younger, more educated cadres and professionals in the third period. Promotion rates associated with rank levels provided information on ﬂuidity and barriers in the bureaucratic system. We used a set of dummy variables to measure transition rates across ranks among administrative/ professional titles. The reference category is rank1 – transition from the lowest to the immediate next rank (i.e., from below ke to ke rank for the administrative line, and from technician to assistant engineer rank for the professional line). Rank 2 refers to transition from ke to associate chu for the administrative line, and from assistant engineer to engineer for the professional line. Rank 3 refers to transitions from associate chu to chu rank, and from engineer to senior engineer, respectively. Finally,

188

The State and Life Chances in Urban China

rank 4 refers to transition from chu to ju or higher positions for the administrative line. Here the coefﬁcients associated with rank levels indicate the promotion rate from that rank to the next higher rank, compared with the baseline – the promotion rate between the lowest rank and the second lowest rank (i.e., from ordinary cadre to ke-level for the administrative line, and from technician to assistant engineer for the professional line), net of the effects of other covariates in the model. The promotion rate decreased with the ascending order of administrative ranks, indicating a pyramid shape in which vacancies become fewer when one advances. The barrier was the highest between the chu level and the ju level (rank 4). In the professional line, there was no signiﬁcant barrier between assistant engineer and engineer level, but there is a signiﬁcant barrier for promotion from “engineer” level to “senior engineer” level (rank 3). Cadre/Professional Cohorts. The cohort effects for both administrative and professional promotions were similar across these periods. Compared with the ﬁrst cohort (the reference category), the second cohort had no signiﬁcant difference in promotion rate during the Cultural Revolution period. In the post-Mao era, there were signiﬁcant cohort effects, but the pattern was inconsistent with our expectation: The ﬁrst cohort has the highest rate of promotion, whereas those in the third cohort had the lowest rate of promotion, net of the effects of other covariates. This ﬁnding suggests the importance of the cohort-based seniority principle in promotion in the Chinese bureaucracy that is contradictory to the new policies in the post-Mao era. Social Origins. Finally, there were no systematic effects of father’s education and occupational status on one’s promotion along either career lines. Promotion Patterns in Work Organizations We now focus on another type of bureaucratic promotion – promotion events along the managerial ladder (low, middle, high) within work organizations. This type of promotion deserves attention because it is associated with substantive authority positions in an organization. We included covariates similar to those in previous analyses. But because promotion in workplaces may include not only cadres and professionals already in these occupations but also those ordinary workers, the latter should be included in the risk set. Furthermore, because both cadres and professionals may be promoted into managerial positions in the workplace,

Climbing the Political Ladder: Bureaucratic Career Patterns

189

we considered cohort effects for both cadres and professionals in the model.6 Table 6.4 reports the parameter estimates of the promotion patterns in the three periods. We organize our discussion of the results around the main research issues developed before. Macro-Political Processes Versus Organizational Attributes. We speculated that macro-political processes played a more important role than organizational attributes in bureaucratic promotions, and that the impacts of political selection should be reﬂected in differential cohort effects. Table 6.4 shows that there were signiﬁcant changes in the opportunity structures of bureaucratic promotions across the three periods. For instance, the intercepts give the log-rate of promotion for the reference group – a twenty-ﬁve-year-old male ordinary worker with a middle school or lower education, in a collective ﬁrm in a large city. The intercept was negative in the ﬁrst period and was not statistically signiﬁcant from zero in the second. In the reform era, this group had a signiﬁcant and higher probability of being promoted, as indicated by the larger and statistically signiﬁcant intercept. Changes in the intercepts over time indicated increasing promotion opportunities in work organizations in the reform era. Age effect varied across historical periods. In the third period, age restriction was more salient in promotion in work organizations than in the national system, as indicated by the lowest rate and earliest peak associated with promotions in workplaces in Figure 6.6. Education, especially a college degree, contributed to promotion throughout the three periods. Compared with the varying effects of education across periods in the national bureaucratic system (cf. Table 6.3), it appears that bureaucratic promotion in work organizations was largely shielded from the dramatic shifts in state policies. Again, we ﬁnd that party membership contributed signiﬁcantly to promotions in work organizations across the three periods. But similar to administrative promotions, the effect was smaller in the Cultural Revolution period. The other point of striking contrast is related to cohort effects. Recall that patterns of promotion in the national bureaucratic system showed that the promotion rate increases with the earlier cohort. In other words,

6

There are cases where promotion events and job-change events occur simultaneously. In such cases, we use the information on the newly entered organization in the statistical analyses, assuming that the attributes of the latter play a more important role in affecting the promotion opportunities.

table 6.4. GEE Estimates of the Determinants of Promotion Patterns in Work Organizations Covariates Intercept Female Age Age2 /100 education High school College Party membershipa Duration in position Low-rank Mid-rank cadre 49–65 cohort 66–79 cohort 80–94 cohort professional 49–65 cohort 66–79 cohort 80–94 cohort father’s education Middle school High school or above Missing value father’s occupation Cadre Professional Farmer Missing value work organization Government Public organization State ﬁrm city location Medium city Small city Number of events Number of cases Log-likelihood d.f. ∗∗

1949–1965

1966–1979

1980–1994

∗∗

−7.673 −0.872∗ 0.008 −0.926∗

−0.562 −0.731∗∗ 0.056∗ −0.109

3.263∗∗ −0.729∗∗ 0.097∗∗ −0.354∗∗

−0.132 0.690∗∗ 0.886∗∗ 0.034 −1.213∗∗ −2.099∗∗

0.316† 0.906∗∗ 0.710∗∗ 0.022 −0.956∗∗ −1.663∗∗

0.454∗∗ 0.914∗∗ 0.834∗∗ −0.006 −0.846∗∗ −1.368∗∗

3.387∗ — —

2.396∗∗ 3.621∗∗ —

2.547∗∗ 1.979∗∗ 2.985∗∗

0.654∗ — —

0.528∗ 1.414∗∗ —

1.330∗∗ 1.083∗∗ 1.643∗∗

0.194 −0.053 0.123

0.064 0.381 −0.268

0.090 −0.026 0.200

−0.798 −0.263 −0.442 −0.156

−0.049 −0.478 −0.311 0.148

0.052 0.017 −0.039 −0.313

−0.410 −0.314 0.070 −0.379 −0.765∗ 143 1380 −604.1 22

0.129 0.058 0.439∗ 0.144 −0.218 275 2984 −1,243.3 23

−0.067 −0.107 0.066 −0.611∗∗ −0.390∗∗ 634 4200 −2,805.8 26

p < 0.01 p < 0.05 † p < 0.10 a The coefﬁcients associated with party membership are estimated using data from fourteen cities. Note: Middle school or below is the reference category for education; production worker, for father’s occupation and administrative/professional cohorts; collective ﬁrm, for work organization; and large city for city location. ∗

Climbing the Political Ladder: Bureaucratic Career Patterns

191

the seniority rule was prevalent in the national bureaucratic system. In contrast, the evidence for promotion in work organizations shows distinctive and inconsistent life chances among different cohorts of cadres and professionals. In the second period, those who became cadre during the Cultural Revolution (the 1966–1979 cohort) had a higher rate of being promoted than those who became cadres before the Cultural Revolution (the 1949–1965 cohort). In the reform era, in contrast, for both cadres and professionals, the 1980–1994 cohort – those recruited into these two occupations in the reform era – had the highest probability of being promoted. In contrast, those who were recruited in the Cultural Revolution (the 1966–1979 cohort) had the lowest probability among the three cohorts. Variations in cohort effects indicated marked changes in political selection criteria across the two periods. These patterns imply that the composition of cadres in the bureaucratic positions differed between the national bureaucratic system and the workplace-centered system. Managerial positions in the formal bureaucratic system reﬂect more continuity with the past, whereas ordinary work organizations tended to be staffed by a younger cohort of cadres. These ﬁndings have important implications for future political changes in China, as we will discuss later. There is also evidence to support the arguments about the two segmented bureaucratic career lines. The magnitudes of cohort effects are larger for cadre cohorts than for professional cohorts, indicating that the cadre career line is more sensitive to political processes and leadership succession than the professional line. Overall, cadre cohorts had a higher rate of promotion than professional cohorts in both periods, indicating that the cadre occupation is the fast-track career line relative to professional occupations within organizations.7 With some minor exceptions, organizational attributes do not show systematic effects on promotion rates within organizations in any period. Government agencies, public organizations, as well as centralgovernment-owned and local-government-owned ﬁrms did not show patterns signiﬁcantly different from those for collective ﬁrms (the reference category). Duration in the present position did not have a signiﬁcant effect, 7

Notice that coefﬁcients for all these cohorts are large. This is because, although some ordinary workers (the reference group) were being promoted into managerial positions, these instances were relatively less common than promotions of those already in administrative or professional occupations. This pattern suggests the presence of a segmented workforce with the most formidable barrier between ordinary workers and administrators/ professionals.

192

The State and Life Chances in Urban China

indicating that there was no evidence of path dependency with respect to position duration, other things being equal. The promotion rate decreased along with the ascending managerial rank within organization, as indicated by the increasingly larger, negative coefﬁcients associated with the two managerial rank variables. Social Origins. Similar to the ﬁndings in the previous analyses, there was no evidence that social origins, as measured by the father’s education and occupational status, played a signiﬁcant role in promotion patterns in work organizations.

chapter summary This chapter reported a systematic examination of several aspects of bureaucratic career patterns. Our empirical analyses were guided by two theoretical models: one emphasizing stable institutional basis of segmented career lines and the other emphasizing the link between political dynamics and bureaucratic career patterns, and especially the impacts of varying state policies on bureaucratic careers. We also take into consideration the possible effects of intergenerational mobility in our analyses. We now summarize the main ﬁndings. First, there is evidence consistent with Walder’s dual-path model. As ﬁndings on the entry into administrative and professional occupations show, these are distinctive career lines with different boundaries and selective criteria. As Walder argued, educational credentials are much more important for entry into and promotion in the professional career line than for the administrative career line. Party membership plays a more salient role in administrative promotions than in professional promotions. Patterns of promotions also show that cadre cohorts are more sensitive to changes in the historical context than the professional cohorts, indicating different processes underlying these two career lines. Second, shifting state policies had signiﬁcant impacts on bureaucratic career patterns in terms of varying opportunities and selection criteria across historical periods. Political opportunities varied signiﬁcantly across historical periods, causing drastically different life chances for individuals who entered the labor force in different years. For instance, colleges were closed for many years during the Cultural Revolution, making it impossible for high school graduates of that period to access higher education for future career advancement. Moreover, there was also evidence of changes in selection criteria over time. The importance of educational

Climbing the Political Ladder: Bureaucratic Career Patterns

193

credentials varied signiﬁcantly in the recruitment of Communist Party members between the Mao and the post-Mao eras. Although education (e.g., a college degree) played a signiﬁcant and positive role in entry into and promotion in the Chinese bureaucracy, its magnitudes of effect varied across periods, in accordance with changes in state policies. Similar patterns were found in the varying gender roles over time. Finally, the impact of political dynamics was most evident in the cohort effects. In entry into the Communist Party and promotions within workplaces, we see a clear pattern of preferential treatment toward different cohorts of cadres in accordance with their presumed political allegiance. In particular, the relatively lower rate in the reform era of the second cohort’s promotion along administrative positions within organizations indicates the strong inﬂuence of discriminatory state policies and political selection processes. Perhaps the most surprising and interesting ﬁnding in this study is the two distinctive patterns of bureaucratic promotions with regard to different cohorts of cadres and professionals. In formal bureaucratic promotions in the national bureaucratic system, the seniority principle seemed to still play a major role even in the reform era, as indicated by higher rates of promotion for earlier cohorts. Promotions within the workplace, however, show a different pattern: The third cohort had the highest rate of promotion, whereas the second cohort had the lowest rate among the three cadre cohorts. How can we explain these two distinctive patterns? One speculation is that the formal bureaucratic apparatus was effectively organized by bureaucratic interests, and the implementation of the new cadre policies was resisted by those still in power in the bureaucracy. Because work organizations are more responsive to changing local environments, especially market competition introduced in the reform era, they are more likely to be active in promoting younger and talented personnel into managerial positions. Thus, the ﬁndings indicate that there are two distinctive selection processes in the two organizational contexts (for further analyses, see Zhao and Zhou 2004). With regard to the third line of arguments on intergenerational mobility, there is evidence that children of administrators and professionals are more likely to follow in their fathers’ footsteps into corresponding administrative or professional jobs. On the other hand, there is no evidence that the father’s occupation and education played a signiﬁcant role in the acquisition of Communist party membership or promotion. Thus, the absence of parental effects on promotion casts doubts on the effectiveness and stability of class reproduction.

194

The State and Life Chances in Urban China

Bureaucratic career patterns provide critical information about the operation of political processes and stratiﬁcation dynamics under state socialism; they also shed light on the continuity and changes in the recent transformation of state socialism. Therefore, this chapter provides a useful transition from the ﬁrst theme of this book, the interplay between redistribution and stratiﬁcation dynamics, to the second theme of assessing institutional changes in China in the subsequent chapters.

part two ASSESSING INSTITUTIONAL CHANGES IN THE POST-MAO ERA

S

ince the 1980s, China has embarked on a path of economic reform that has profoundly changed the landscape of the Chinese society. The impressive progress China has made since the 1980s and the distinctive characteristics of the Chinese experience, especially the coexistence of a strong state and striving market economies, have stimulated active social science research on China’s institutional transformation. In the next four chapters, we turn attention to the second theme of this study – the examination and assessment of the extent and sources of institutional changes in the era of economic transformations. These chapters are organized around particular topics of theoretical interest. Similar to the previous chapters, we present empirical evidence across the historical periods, from 1949 to 1994, and continue our discussion of social stratiﬁcation patterns. But the emphases of these discussions will be primarily on assessing changes and continuity in the reform era (1980– 1994), in contrast to the patterns in the Mao era (1949–1979). Chapter 7 focuses on changes in the determinants of manifest economic beneﬁts – personal income. Chapter 8 calls attention to, and presents evidence on, the importance of latent economic beneﬁts under redistribution, and reports a detailed analysis of one aspect of the latent economic beneﬁts – housing allocation in urban China. In Chapter 9, we assess the extent of institutional changes over time based on patterns of job shifts across types of organizations and economic sectors. Finally, Chapter 10 develops a life-course perspective to contrast cohort differences in the distribution of life chances in the era of economic transformation.

195

7 Economic Transformation and Changes in Income Inequality

This chapter begins a series of studies that assess changes and continuity in China’s institutional transformation by contrasting patterns of social stratiﬁcation before and in the reform era. As we discussed in Chapter 1, since the 1980s, there have been increasing interests among social scientists to study the Chinese experience in the transformation of state socialism (Guthrie 1997, 1998, 1999; Keister 1998, 2001; Oi 1999; Walder 1995c, 1996a). In Chapter 1, we discussed major theoretical models and highlighted ongoing debates among sociologists who study economic transformation of state socialism. On the one hand, Nee’s (1989, 1991, 1996) market transition theory emphasizes the importance of emerging markets in undermining state socialist redistributive institutions, whereas other scholars highlight the importance of the state and local governments in generating path-dependent institutional transformation. On this basis, we proposed a coevolutionary model of politics and markets to synthesize these theoretical ideas. The studies in the subsequent chapters are situated in this historical and intellectual context. Despite an increasing number of studies in this area, the accumulation of knowledge particularly suffers from a lack of a consensus on the operationalization of key concepts. For instance, the cadre status has been used as an indicator of the key concept of “positional power” in state socialist redistribution. But it was measured differently in different studies, even in similar urban or rural settings. As Parish and Michelson (1996) showed, different operationalizations of the cadre status may lead to very different

197

198

The State and Life Chances in Urban China

conclusions about the role of positional power even when analyzing the same data set. Although assessment of changes over time has been the main focus of research, most studies have used cross-sectional research designs. Such a design facilitates particular research purposes in comparing interregion or intergroup differences, but they are inadequate for studying changes over time. In addition, a cross-sectional research design is ill equipped to deal with unobserved individual attributes that may confound several key theoretical issues. For instance, the latent attributes of individual ability may affect an individual’s educational level, occupational status, and acquisition of party membership. In this chapter, we examine changes in income determinants among urban residents over time. We collected retrospective information on respondents’ income for selected years between 1955 and 1994. We begin with the premise that fundamental institutional changes should have affected, and are reﬂected in, changes in social stratiﬁcation processes. Similar to the previous chapters, we adopt a historical perspective and examine changes and continuity of the stratiﬁcation processes over time, contrasting the pre-reform era and the reform era. We believe that this approach is especially useful for understanding the extent and direction of institutional changes. The focus of our study on urban China, where China’s redistributive economy has reached its most elaborate form, provides a more appropriate examination of the changes in state socialist redistribution. The use of panel data rather than cross-sectional data allows us to assess changes over time in a more satisfactory way and helps alleviate some potential problems in measurements of key concepts. Income is a conventional indicator of social inequality in stratiﬁcation processes, and there are merits for focusing on income determinants: First, income distribution is sensitive to changes in economic institutions, and a study of income determinants can detect the most salient changes in the transformation processes. Second, most empirical studies so far have examined patterns of income inequality. Our study addresses the same set of empirical questions and thus can be directly compared with previous studies in the literature. However, scholars have called attention to the importance of latent and nonmonetary resources distributed through workplaces and other channels (Bian 1994; Oberschall 1996; Walder 1992; Zhou, Tuma, and Moen 1997). The next two chapters will address issues related to latent beneﬁts (housing, Chapter 8) and job shifts across organizations (Chapter 9).

Economic Transformation and Changes in Income Inequality 199

redistribution and marketization: implications for income inequality Empirically, the debate between market transition theory and those who emphasize the political economy of transformation has centered on the implications of their theoretical arguments for changes in the social stratiﬁcation order. We now recast these theoretical arguments and their implications for income inequality. We take the market processes emphasized in Nee’s market transition theory as our starting point and then consider how other processes may constrain or intertwine with market processes, and their implications for patterns of income determinants. Our emphasis here is less on adjudicating among competing theoretical arguments than on identifying multiple processes and their interactions in these theoretical propositions. Given the state of conceptual and empirical work in the literature and continuously evolving reform processes in China, attempts to adjudicate among competing theoretical arguments are premature and may not be conducive to advancing our knowledge of the ﬁeld. Because our research design focuses on comparing changes over time, we discuss empirical implications by focusing on a comparison between the pre-reform era and the reform era. Inequality between Redistributors and Producers. Building on Szel´enyi’s (1978) argument that inequalities “are expressions of the basic conﬂict of state socialist societies, the conﬂict between the ‘immediate producer’ and the ‘redistributor’” (p. 77), Nee predicts that market transitions produce fundamental changes in income inequality among social groups. As Nee put it (1996, p. 916): “The growth of market institutions (i.e., labor markets, subcontracting arrangements, capital markets, and business groups) causes a decline in the signiﬁcance of socialist redistributive power even in the absence of fundamental change in the political order.” The empirical implication is that the advantages of “redistributors” in the acquisition of economic beneﬁts decline relative to that of “producers” in the reform era, compared with the pre-reform era. To the extent that the emergence of markets creates new opportunities outside the redistributive system and entails principles of resource allocation different from the redistributive economy, Nee’s proposition is a plausible one. However, one’s market position depends on access to opportunities and resources, which, in the Chinese context, is signiﬁcantly affected by the existing institutional arrangements of redistribution. Those with positional power have advantages in access to both resources and market-induced opportunities, relative to other social groups: Previous

200

The State and Life Chances in Urban China

“redistributors” may now proﬁt from the economic arena in the role of “regulators” in market transactions or as parties to economic transactions (Bian and Logan 1996; Walder 1996b). Moreover, the weakening of government intervention into ﬁrm-level decisions may have strengthened local authority’s power in gaining a larger proportion of the surplus in proﬁt sharing on the shop ﬂoor. Therefore, with the advance of markets, some redistributors may have lost their beneﬁts, and some social groups may have gained their beneﬁts. But, once we take into consideration these intervening processes, it is likely that, on average, the advantage of redistributors in the acquisition of economic beneﬁts may not decline relative to that of producers in the reform era, compared with the pre-reform era. An important empirical issue is how to deﬁne redistributors and producers. Nee advocated to deﬁne “administrative elites” as redistributors and to deﬁne producers to include, in addition to ordinary workers, “entrepreneurs, managers, and technicians” (1996, p. 916). We develop two alternative measures of cadre (administrator) status in this chapter. The ﬁrst one uses the bureaucratic ranks of cadres promulgated by the central government. The second one distinguishes cadres in the public sector (government and public organizations) from those in the economic sector (in enterprises). The higher a cadre’s bureaucratic rank, the closer he or she is to the redistributive authority in state socialism. Similarly, cadres in the public sector are closer to the political authority than are cadres in the economic sector. Therefore, the former is closer to the concept of redistributors and the latter to that of managers. We measure professionals in a similar way to detect the possibility that economic beneﬁts vary with their positions (professional ranks) or their closeness to markets (public versus economic sectors). Political versus Human Capital. Institutional changes are reﬂected in changes in the mechanisms of resource allocation. In a market economy where factors of production are allocated through competitive market transactions, it is argued that human capital plays an important role in determining one’s economic rewards (Becker 1964). In contrast, it is often observed that the socialist state rewards political loyalty rather than competence. Following this logic, Nee (1989) argued that, as the political logic of redistribution is eroded by market mechanisms, returns to political capital decline and returns to human capital increase in the reform era, compared with the pre-reform era. In the literature, it is conventional to use Communist Party membership as an indicator of political capital and formal education and work experience as indicators of human capital.

Economic Transformation and Changes in Income Inequality 201 Now consider other coexisting processes in allocating and rewarding political and human capital in the course of China’s economic transformation. First, with regard to political capital, a direct implication can be drawn from the preceding discussion: So long as political authorities play a signiﬁcant role in bureaucratic recruitment and promotions, persistent and signiﬁcant returns to political status are logically plausible. Second, with respect to returns to human capital, we recognize multiple and competing processes of allocating human resources in urban China. Nee’s argument points to the rise of labor markets in allocating human capital as the source of increasing returns to education. But there is no intrinsic principle of state socialist redistribution that dictates low returns to education. Although the Chinese state emphasized political loyalty and state policies toward intellectual have varied over time, education has not been totally ignored. Instead, higher education accompanying with political loyalty was desirable in this political context, as epitomized in the ´ and Szel´enyi slogan “Red and Expert” (youhong youzhuan). As Konrad (1979) pointed out, socialist planning system also valued education and knowledge and incorporated intellectuals into political elites. Walder, Li, and Treiman (2001) developed a similar line of arguments in the Chinese context. A recent comparative study of redistributive patterns in the USSR and China reveals that returns to education were considerably higher in the USSR than in China (Zhou and Suhomlinova 2000). The state has been and is still actively involved in the allocation of human resources. Even in the reform era, a large proportion of individuals with high human capital (e.g., college education) entered the state sector, especially in government and public organizations; see Chapters 4 and 9 in this book. More important, since the early 1980s – before any substantive market reform took place in the urban areas – government policies have emphasized educational credential as one of the most important criteria in obtaining political and positional power (such as party membership and cadre promotion). In this light, the importance of educational credential may increase with changes in state policies, in directions that are consistent with market mechanisms. The Institutional Arrangements of Work Organizations. In state socialist China, scholars have called attention to work organizations as the major institutions of redistribution (Bian 1994; Lin and Bian 1992; Walder 1986, 1992; Whyte and Parish 1984; Zhou et al. 1996, 1997). These studies identiﬁed a hierarchical order among types of work organizations whose economic beneﬁts vary systematically with state policies and their property rights relationships to the state. Work organizations

202

The State and Life Chances in Urban China

in the state sector (government agencies, public organizations, and stateowned ﬁrms) beneﬁt more from redistribution and state policies than do those in the semistate and nonstate sectors (collective ﬁrms, hybrid, and private ﬁrms). How would the expansion of markets affect the institutional arrangements of work organizations? According to the logic of market transition theory (Nee 1992), one would argue that ﬁrms that are closer to market transactions are more likely to experience major changes in patterns of income determinants, relative to those organizations in the state sector. Because political control varies signiﬁcantly across types of work organizations, those who are less controlled by the state are more likely to break away from the redistribution system. Thus, with the advance of market processes, market transition theory expects that organizational hierarchies associated with the redistributive economy would change, so would the distribution of economic resources associated with types of work organizations. That is, in the reform era, patterns of economic beneﬁts associated with types of work organizations should favor those organizations (collective and hybrid/private ﬁrms) that are closer to market transactions and less controlled by the redistributive system. However, there are important competing processes that countervail the market-induced processes identiﬁed previously. Work organizations in the state sector provide legitimation for the political order and the interests of the central authority (Walder 1986). Relative to the new interests emerging from market expansion, the vested interests associated with existing institutions are better organized and have better access to political authorities. Institution-based bureaucratic bargaining often led to those policies and regulations that protect the vested interests associated with the state sector. Hence, these considerations point to state interventions in regulation or resource transfer in favor of those in the state sector, producing institutional persistence. In this view, the politics of vested interests embedded in existing institutions will lead to the persistence of income inequality based on existing organizational hierarchies in the reform era.

cross-sectional analyses of income determinants We now turn to empirical analyses. We use the total personal income (Chinese yuan/month, including basic wages, bonus, and income from other sources) as the dependent variable, inﬂation-adjusted using provinciallevel information (SSB 1990a). In the time span of our study, for the majority of the urban residents, earnings from workplaces are the major source of income.

Economic Transformation and Changes in Income Inequality 203 7000 6000

total income

5000 4000 3000 2000 1000 0 55

60

65

75

78

84 year

National statistics

87

91

92

93

94

Sample statistics

ﬁgure 7.1. Comparison of income in national and sample statistics.

In our sample, income information was collected for the following years: 1955, 1960, 1965, 1975, 1978, 1984, 1987, 1991, 1992, 1993, and 1994. Most of the selected years were associated with important historical events. For instance, 1960 was the year of a severe economic disaster; 1965 was the year of the eve of the Cultural Revolution. We selected these years for several reasons. First, income distribution in years of historical signiﬁcance helps us identify important characteristics of the redistributive system. Second, the salience of these years in one’s life experience may assist the respondent to recall information more accurately. The data yield an urban labor force sample whose composition varies over time. An individual enters our analyses when she or he entered the labor force and leaves the analyses when she or he exited the labor force (e.g., retires). We also exclude the income records for those were in the rural labor force in a particular year, because incomes in rural and urban areas were not comparable. To assess the quality of the income data, we compared the average income of our sample based on respondents’ self-reports in those selected years and the average income of our sample using ofﬁcial statistics of the average income in those provinces where the respondents were drawn.1 Figure 7.1 displays the patterns between the two. 1

It is desirable to use city-level information for such a comparison. However, there is no systematic information over time at the city level in ofﬁcial statistics.

204

The State and Life Chances in Urban China

As Figure 7.1 indicates, between 1955 and 1975, average income in our data is very close to that based on ofﬁcial statistics. The differences between the two increased between 1978 and 1994, with the average income in our sample being higher than the ofﬁcial statistics. But these differences are not substantial relative to the income level in these speciﬁc years. More important, the differences between the sample and the ofﬁcial statistics show a relatively stable trend. The higher average income in our sample in recent years is likely to have resulted from the fact that our sample has a larger proportion of residents from large cities (because of the inclusion of Beijing and Shanghai and capital cities of the provinces). This source of the observed discrepancies can be captured by modeling city locations in the statistical analysis. We ﬁrst examine patterns of income determinants in selected years, using a multivariate regression model. Our purpose is twofold: First, the results in this set of analyses help detect patterns of change in income determinants over time. Second, our analyses establish a basis for comparison with results in previous studies and in the panel data analysis in later sections. Table 7.1 reports the parameter estimates of the covariates in selected years. The estimated models also included a set of indicator variables for respondents’ city localities. We are interested in the patterns of resource allocation among social groups, while controlling for cross-city variations in overall income levels. As indicated at the bottom of Table 7.1, sample sizes vary across these selected years, reﬂecting changing compositions of the sample over time. The adjusted R2 indicates that the model speciﬁed for these selected years has a reasonably good explanatory power. This model accounts for 24 to 44 percent of the income variation in various years.2 We now discuss the effects of selected sets of the covariates and our focus is on the historical trend as reﬂected in variations in coefﬁcients in the same variables across the years. To best demonstrate the changes over time, we plotted changes in these coefﬁcients in selected years. The years of 1960 and 1965 were in the ﬁrst period – the pre-Cultural Revolution period; the years of 1970, 1975, and, to some extent, 1978, were the years of the Cultural Revolution; those years after 1978 were in the post-Mao era.

2

The high R2 is partly due to the inclusion of the city indicator variables, which captured substantial income variations across cities.

205

0.040 0.057 0.121∗∗

0.356∗∗ 0.075 0.253∗∗ 0.158∗∗ 0.034 0.077 0.074∗

0.147∗∗ 0.110∗∗ 0.118∗∗ 0.113∗∗ 0.067 0.38 1133

0.041 0.102∗ 0.178∗∗

0.408∗∗ 0.163∗∗ 0.249∗ 0.198∗∗ −0.004 0.038 0.061

0.187∗∗ 0.202∗∗ 0.221∗∗ 0.164∗∗ 0.183∗ 0.38 778

−0.245∗∗ 0.070∗∗ −0.078∗

2.258∗∗

1.983∗∗

−0.290∗∗ 0.088∗ −0.111

1965

1960

0.086∗ 0.102∗∗ 0.157∗∗ 0.113∗∗ 0.093† 0.34 2137

0.077∗ 0.085∗ 0.164∗∗ 0.108∗∗ 0.207∗∗ 0.29 2624

0.406∗∗ 0.166∗∗ 0.196∗∗ 0.075∗ 0.052 −0.023 0.058∗

0.038 0.028 0.062

0.064∗∗ 0.053† 0.149∗∗ 0.285∗∗ 0.074∗ 0.131∗ 0.048 0.005 −0.023 0.016

−0.162∗∗ 0.065∗∗ −0.068∗∗

2.421∗∗

1978

−0.179∗∗ 0.062∗∗ −0.061∗∗

2.289∗∗

1975

0.073† 0.045 0.121∗∗ 0.070∗∗ 0.319∗∗ 0.23 3257

0.266∗∗ 0.089∗ 0.138∗ 0.019 0.036 0.024 0.007

0.054† 0.023 0.039

−0.144∗∗ 0.016∗ −0.008

3.773∗∗

1984

0.066† 0.079∗ 0.109∗∗ 0.100∗∗ 0.537∗∗ 0.28 3477

0.209∗∗ 0.059† 0.141∗∗ 0.029 0.055† −0.008 −0.010

0.072∗ 0.096∗∗ 0.157∗∗

−0.188∗∗ 0.023∗∗ −0.022∗∗

3.754∗∗

1987

0.096∗∗ 0.123∗∗ 0.105∗∗ 0.098∗∗ 0.558∗∗ 0.39 3591

0.213∗∗ 0.103∗∗ 0.141∗∗ 0.071∗ 0.050 0.013 −0.007

0.089∗∗ 0.160∗∗ 0.157∗∗

−0.187∗∗ 0.040∗∗ −0.044∗∗

3.364∗∗

1991

0.166∗∗ 0.157∗∗ 0.172∗∗ 0.110∗∗ 0.557∗∗ 0.44 3627

0.342∗∗ 0.207∗∗ 0.202∗∗ 0.154∗∗ 0.070∗ 0.076∗ 0.029

0.106∗∗ 0.176∗∗ 0.167∗∗

3.345∗∗ −0.182∗∗ 0.040∗∗ −0.043∗∗

1993

∗

p < 0.01. p < 0.05. † p < 0.10. Note: Elementary school education or below is the reference category for education, unskilled manual worker for occupation, and collective ﬁrm for type of work organization. Nineteen dummy variables for each city are also included in the analysis.

∗∗

Intercept Female Age Age2 /100 education Middle high High school College occupation High-rank cadre Low-rank cadre High-rank professional Low-rank professional Service worker Ofﬁce worker Skilled worker work organization Government Public organization Central gov’t ﬁrm Local gov’t ﬁrm Private/hybrid ﬁrm Adjusted R2 N

Covariates

table 7.1. OLS Estimates of the Determinants of Log(Income), Selected Years

The State and Life Chances in Urban China

206 0.4

log-rate

0.2

0

-0.2

-0.4 1960

1965

1970

1975

1980

1985

1990

1995

year female

college

ﬁgure 7.2. Log-rates of income for female and college education (selected years).

Figure 7.2 displays changes in gender difference and the rate of returns to college education across the selected years. As one can see, gender differences were narrowing prior to the reform era, as indicated by the rising wage rate for women, relative to men, between 1960 and 1984. Compared with those who had elementary or no education (the reference category), returns to college education decreased in the Mao era after the 1960s, especially in the early years of the economic reform. After the mid1980s, when the economic reform in the urban area started, returns to education increased dramatically and bounced back to the pre-Cultural Revolution level.3 Figure 7.3 reports changes in income rates over time for cadres and professionals, as compared with unskilled workers (the reference category). First, all cadres and professionals followed a decreasing trend in the Mao era, reﬂecting the impact of the state destratiﬁcation policy. In the early phase of the reform (1978), there was a noticeable jump in income for these groups. This largely reﬂected the readjustment of compressed salaries/wages in the aftermath of the Cultural Revolution. Between 1978 3

It is worth pointing out that the inclusion of occupational status (e.g., cadre and professional) may dampen the magnitude of educational effects, as the latter may be mediated by one’s occupational status. But the general trend reported here remains the same even when occupational variables are removed from model estimation.

Economic Transformation and Changes in Income Inequality 207 0.6

log-rate

0.4

0.2

0 1960

1965

1970

1975

1980

1985

1990

1995

year high-rank cadre high-rank prof.

low-rank cadre low-rank prof.

ﬁgure 7.3. Log-rates of income for cadres and professionals (selected years).

and 1987, the rates narrowed relative to the reference group of ordinary workers. After the mid-1980s, the rates for these groups were on the rise again, propelled by the reform of the ofﬁcial salary system and the market processes. Finally, relative to employees in collective ﬁrms (the reference category), employees in organizations in the state sector experienced an initial decrease between 1960 and 1965, and then a slight increase during the Cultural Revolution period (see Figure 7.4). They experienced a decrease in the early phase of the reform; but their advantages have increased since the 1990s. It is noticeable that the three types of organizations in the state sector show similar patterns over time, reﬂecting the common processes that governed the distribution of economic beneﬁts in these organizations. In contrast, the most dramatic change is for those in hybrid ﬁrms (most are private entrepreneurs). In the reform era, the rate of income for this group took off and experienced the largest gain in the 1980s. Several major patterns emerge from these ﬁgures. First, there were noticeable variations in the distribution of economic resources over time and these variations were consistent with changes in state policies and macropolitical processes. Second, we can detect roughly two income regimes over time, and a transitional period in between. The ﬁrst income regime is associated with the Mao era, 1949–1977. In this period, there was a general trend of equalization in income, reﬂecting state destratiﬁcation

The State and Life Chances in Urban China

208 0.6 0.5

log-rate

0.4 0.3 0.2 0.1 0 1960

1965

1970

1975

1980

1985

1990

1995

year government

public org'n

state firm

hybrid firm

ﬁgure 7.4. Log-rates of income for type of organizations (selected years).

policies of the time. Across the three years under this regime (1960, 1965, and 1975), there was a trend of decreasing gender inequality as well as a narrowing of differences in occupational status (e.g., returns to cadres and professionals). There was also some narrowing of income differences among types of work organizations, although the patterns were less systematic. Patterns of income determinants in 1978 and 1984 reﬂected a transitional period between the pre-reform and reform eras. In 1978, as a rectiﬁcation of the radical policies of the Cultural Revolution, the postMao leadership adopted a series of wage increases, which were reﬂected in increasing returns to cadre and professional status in 1978 (see Table 7.1). In some aspects, patterns of income determinants in 1978 resembled a return to the pre-Cultural Revolution period, but there were few institutional changes introduced in the urban areas. In the mid-1980s, the state started the large-scale economic reform in the urban area. In the early 1980s, the central government also experimented with decentralization in allocation of resources in industrial ﬁrms. The most immediate effects of these policies on income were that managers in ﬁrms had the discretion to decide bonuses outside the stateset wage system. These changes were reﬂected in the patterns of income determinants in 1984, where employees in government and public organizations, on average, had no visible advantages, other things being equal.

Economic Transformation and Changes in Income Inequality 209 Patterns of income determinants in 1987, 1991, and 1993 showed evidence of a new income regime. Overall, the effects of the covariates indicate a consistent trend during this period. In particular, we observed increasing returns to education, as well as to cadre and professional positions. Types of work organizations remained important during this period and their effects appeared to be increasing. Private/hybrid ﬁrms had the largest gains in the reform era. To sum up, the results reported here are broadly consistent with both documented historical changes of the redistributive economy in urban China (Whyte and Parish 1984) and ﬁndings in other studies. The general patterns of income variations reﬂected two main distinctive regimes. In the Mao era, 1949–1977, variations in income showed a decreasing trend, with the declining importance of education and occupation, among others. This was clearly attributable to the “destratiﬁcation” state policies promulgated during this period (Parish 1984). However, in the era of urban reform, we observe a trend of increasing inequality, as indicated by the increasing importance of education, occupation, and work organizations. Changes in the transitional period between 1978 and 1984 were largely induced by state policies and had little to do with changes in the institutional arrangements of urban economies, with the noticeable exception of the signiﬁcant growth of the hybrid and private ﬁrms. For the purpose of examining institutional changes in urban China, these two years were closer to the pre-reform era than to the reform era.

comparing changes across two eras: a mixed model for panel data We now turn to panel data analysis. Based on the ﬁndings in the previous analyses, we used the year 1985 to divide the pre-reform and reform eras in urban China. The year 1985 marked the ofﬁcial beginning of the urban reform and of the second income regime. It also witnessed a stateinitiated reform of the salary system. The new salary system was based on principles of bureaucratic positions and ranks, which were substantially different from earlier salary systems. Given the speciﬁc years for which we collected income information, 1955–1984 and 1987–1994 are the time spans of the two eras covered in our data. To detect patterns of change across the two eras, we compare the “main effects” of the covariates and the “interaction effects” between these covariates and the period dummy variable, p ( p = 1 if year > 1985; p = 0, otherwise). Roughly speaking, in this formulation the main effects refer to the effects of the covariates in

210

The State and Life Chances in Urban China

the Mao era, and the interaction effects capture changes in these effects in the reform era. We adopted a mixed model to estimate the panel data of income.4 Table 7.2 reports the parameter estimates of two models, using the two alternative speciﬁcations of cadre/professional status.5 In Model 1, cadres and professionals were speciﬁed into “high-rank” and “low-rank” groups, respectively. In Model 2, cadres and professionals were differentiated between those in the public sector (government and public organizations) and those in the economic sector (enterprises). Columns 1 and 2 of Table 7.2 show the main effects and interaction effects of the covariates in Model 1. The intercept and the associated interaction effect indicate the “overall” income level and changes in the overall income level in the reform era. The positive and signiﬁcant effect of the interaction term shows that there was statistically signiﬁcant and substantial increase (exp[0.85] − 1 = 134%) in the overall income in the reform era. Gender. In the pre-reform era, a female employee earned 84 percent (exp[−0.176] = 0.84) that of a male employee’s income. In the reform era, there is a negative “female” effect, indicating that female employees’ earnings decreased during this period relative to male employees. But, this interaction effect is not statistically signiﬁcant. That is, there is a distinct possibility that the observed difference in gender effects across the two periods is due to sampling variations. Work Experience and Education. It is conventional to use formal education and work experience to measure human capital. Work experience, as measured by the ﬁrst- and second-order effects of age show signiﬁcant and varying effects across the two periods. In the pre-reform era, the effect of work experience shows a familiar inverted-U shape. In the second period, however, we observed changes in both the ﬁrst-order and the second-order effects of age. As it turns out, there is a sharp decline in returns to work experience in the reform era, resulting from the weakening of the traditional allocative mechanisms in workplaces. With regard to formal education, there were signiﬁcant returns to all educational levels in the pre-reform era, and these returns further increased in the reform era. As the main effects indicate, in the pre-reform

4 5

We refer interested readers to Zhou (2000a) for technical details in model speciﬁcation. The models also estimate coefﬁcients for the random effects of city and year indicator variables. Because of our interest in the ﬁxed effects of the theoretically related covariates, we do not report the parameter estimates for random effects.

211

Intercept Female Age Age2 education Middle school High school College occupation High-rank cadre Low-rank cadre High-rank professional Low-rank professional Cadre in public sector Cadre in economic sector Professional in public sector Professional in economic sector Ofﬁce worker Service worker Skilled worker work organization Government agency Public organization

Covariates

0.026 0.044†

0.100∗∗ 0.083∗∗

0.046∗ 0.083∗∗ 0.102∗∗

0.038∗ 0.073∗∗ 0.106∗∗ 0.013 0.007 0.043 0.027 — — — — 0.011 0.017 −0.016

0.845∗∗ −0.020 −0.024∗∗ 0.028∗∗

3.042∗∗ −0.176∗∗ 0.064∗∗ −0.071∗∗

0.231∗∗ 0.112∗∗ 0.118∗∗ 0.058∗∗ — — — — 0.028 0.020 0.018

Interaction Effects

a

Main Effects

Model 1

0.070∗ 0.061∗

— — — — 0.183∗∗ 0.100∗∗ 0.085∗ 0.062∗∗ 0.035 0.021 0.017

0.042∗ 0.078∗∗ 0.117∗∗

3.027∗∗ −0.177∗∗ 0.064∗∗ −0.069∗∗

Main Effects

(continued)

0.047 0.052†

— — — — −0.035 0.022 0.034 0.028 0.006 0.017 −0.017

0.049∗ 0.088∗∗ 0.111∗∗

0.828∗∗ −0.021 −0.024∗∗ 0.028∗∗

Interaction Effectsa

Model 2

table 7.2. Parameter Estimates of the Determinants of Log(Income) in the Full Mixed Model

212 — — —

0.017∗∗ —

0.042∗∗ 0.016∗

0.212∗∗ −0.00007∗∗ 0.024∗∗ 4,731 27,417

0.006 0.024 0.226∗∗∗

Interaction Effectsa

0.144∗∗ 0.075∗∗ 0.235∗∗

Main Effects

0.212∗∗ 0.00008∗∗ 0.024∗∗ 4,731 27,392

0.042∗∗ 0.016∗

0.143∗∗ 0.076∗∗ 0.233∗∗

Main Effects

— — —

0.017∗∗ —

0.006 0.024 0.228∗∗

Interaction Effectsa

Model 2

∗

p < 0.01. p < 0.05. † p < 0.10. a The interaction effects indicate changes in the effects of the particular covariates in the 1987–1994 period. Note: Elementary school education or below is the reference category for education, unskilled manual worker for occupation, and collective ﬁrm for work organization. Both models also estimate a set of coefﬁcients for the random effects of the cities and the selected years.

∗∗

Central govt ﬁrm Local-govt ﬁrm Private/hybrid ﬁrms G-covariance structure ν city µyear R-covariance structure σ ρ ε N N× T

Covariates

Model 1

table 7.2 (continued)

Economic Transformation and Changes in Income Inequality 213 era, a college education increased one’s income by 11 percent, and a high school education by 8 percent, compared with those who had elementary or no education (the reference category). The estimates of the interaction effects show that there were signiﬁcant increases in returns to these educational levels in the reform era. A college degree in the reform era had 23 percent higher income (exp[0.106 + 0.102] − 1 = 23%), and a senior high school graduate had 17 percent higher income, compared with those who had elementary or no education. As noted before, this ﬁnding may reﬂect confounded effects of market forces and state policies.6 Overall, returns to education in China were noticeably low, compared with those in the industrialized societies (Psachargopoulos 1994). Occupation. The main issues of contention in the debate are related to the role of positional power in the transformation processes. The parameter estimates for occupational groups show returns to various occupational statuses. The main effects indicate that, in the pre-reform era, high-rank cadres had the highest income – 26 percent higher than unskilled manual workers (the reference category), other things being equal. Professionals (both high-rank and low-rank) and low-rank cadres also had signiﬁcantly higher income, but the magnitudes appeared to be substantively smaller than that for high-rank cadres. There were no statistically discernible differences among other occupational groups (ofﬁce workers, service workers, and skilled manual workers), as compared with unskilled manual workers. Market transition theory predicts that returns to the cadre status decline relative to that of producers in the marketization processes. The interaction effects for occupational groups show that there were no statistically discernible changes for these occupational groups in the reform era. The magnitudes for the professionals appeared to be larger than those for others. But, statistically speaking, we cannot rule out the possibility that these differences are due to statistical variability rather than to substantive changes. Type of Work Organization. Organizational hierarchy is the center of the redistributive economy in urban China. As the results of the main effects show, in the pre-reform era, such an organizational hierarchy in income distribution was indeed present. Compared with the reference 6

As Mincer (1974) showed, the estimates of educational effects (school years) may be biased downward when age is included in the wage model. We need to be cautious in interpreting the magnitudes of returns to education in the analysis. Because of the use of panel data in our study, the comparison of changes in returns to education between the two eras is less likely to be affected by this problem.

214

The State and Life Chances in Urban China

category of collective ﬁrms, employees in central-government-owned ﬁrms enjoyed the highest income. Those in governmental agencies and public organizations as well as those in local-government-owned ﬁrms also had advantages. This pattern is consistent with the organizational hierarchy of the redistributive economy and the prevalent inﬂuence of state policies documented in the literature. The main exception is the signiﬁcant and higher income for employees in the private/hybrid ﬁrms. This resulted from the rise of a private sector in urban economies since 1978, as our cross-sectional analyses (cf. Table 7.1) indicated. To what extent has this organizational hierarchy changed in the reform era? According to the results from the “interaction effects,” not much. Except for the signiﬁcant and positive effect for private/hybrid ﬁrms, the economic beneﬁts associated with types of work organizations have not shown statistically discernible changes. Moreover, collective ﬁrms had not beneﬁted from market activities, relative to other types of work organizations, although they were more market-oriented than those workplaces in the state sector. In other words, there is no strong evidence that the institutional arrangements of work organizations and the associated patterns of income distribution in urban China had changed signiﬁcantly until the early 1990s. Party Membership. In a separate analysis similar to that reported in Table 7.2, excluding samples with no information on party membership (samples from Guangdong and Jiangsu Provinces), we estimated that the net return to party membership is about 6 percent (β = .058, p < .001), and there is no signiﬁcant change in the reform era. That is, there is no evidence of “declining in signiﬁcance” of political capital across the two periods. We note that, because a large proportion of employees in the urban labor force are party members (about 20 percent at the national average), this political status is widely diffused, raising questions about whether it is still a useful indicator of political capital in urban areas. Findings Based on the Second Measure of Cadre Status. In the second measure of cadre/professional status, we distinguish cadres in government and public organizations (the public sector) from those in production/ service organizations (the economic sector). Similarly, we distinguish professionals in public and economic sectors. Because of the small number of cases, we do not differentiate ranks among cadres and professionals in this speciﬁcation. Other occupational categories remain the same. According to the logic of market transition theory, managers and professionals in economic sectors would beneﬁt more from market transactions relative to those in the public sector who were more regulated by

Economic Transformation and Changes in Income Inequality 215 state policies. As the parameter estimates in Model 2 show, in the prereform era, cadres in the public sector had the highest income, cadres in the economic sector and professionals in both public and economic sectors also had signiﬁcantly higher income than unskilled workers (the reference category), but their advantages were less salient. There were no statistically discernible changes in these patterns in the reform era, as indicated by their corresponding interaction effects. The interaction effect did point to a decrease in income for cadres in the public sector. However, this coefﬁcient was neither substantial (in magnitude) nor statistically signiﬁcant enough to warrant a different conclusion. The effects of government and public organizations were smaller in Model 2 than in Model 1. This was because the speciﬁcations of cadres and professionals were based on their locations in types of work organizations. Clearly, a large proportion of cadres and professionals were concentrated in government and public organizations. In the reform era, public organizations had higher income than before and the difference is marginally signiﬁcant (p < .10). Other types of work organizations had no signiﬁcant changes, except for income increase for employees in hybrid/private ﬁrms. Summary. The main evidence for the impacts of market expansion, as consistent with market transition theory, is the signiﬁcant and higher returns to those working in the private and hybrid ﬁrms, whose economic activities were more governed by market transactions than by the redistributive economy. The increasing returns to education are also consistent with market transition arguments. On the other hand, there is no evidence of decline in signiﬁcance in returns to positional power, as measured by cadre/professional status in two alternative speciﬁcations. Besides the signiﬁcant role of private/hybrid ﬁrms, there is no evidence of substantial changes in the organizational hierarchy in urban economies at least until the early 1990s. A natural question is whether the ﬁndings reported here are sensitive to the choice of periodization. We explored several alternative periodization schemes to check the robustness of our ﬁndings. In our preliminary analyses, we explored several alternative speciﬁcations of the two eras. The ﬁrst one was to divide the periods between 1955–1978 and 1984–1994. The results based on this speciﬁcation showed less signiﬁcant changes in parameter estimates across the two eras than those reported in the text. In the second speciﬁcation, we omitted information in 1984 and divided the periods between 1955–1978 and 1987–1994 on the grounds that income determinants in 1984 appeared to be an anomaly. The results showed

216

The State and Life Chances in Urban China

similar patterns as those reported in the text. Alternatively, one may argue that, in the 1990s, economic reform in urban China has accelerated. Therefore, to detect more recent changes in the processes of marketization, one may use a third, alternative baseline to compare changes in patterns of income determinants between 1955–1987 and 1991–1994. We also estimated this model. Overall, more dramatic changes in the 1990s, with both the expansion of markets and the weakening of the central government, did not lead to a signiﬁcant alteration of the patterns of income determinants as compared with the ﬁndings in Table 7.2. These ﬁndings reinforce the view that institutional changes involve multifaceted processes that produce institutional changes in some areas but institutional persistence in others (see Zhou 2000a for more detail).

exploration of regional variations An implication derived from market transition theory is that, given the variations in marketization across regions in China, one should expect to observe more changes in patterns of income determinants in more marketized regions than in less marketized regions (Nee 1996; see also Xie and Hannum 1996). We now compare variations in income determinants across regions and over time. Because local governments play an increasingly important role in the reform era, one would expect signiﬁcant changes at provincial levels. We grouped samples into four regions. Guangdong and Jiangsu were grouped into the “coastal region” that experienced the greatest market reform. Hebei, Heilongjiang, Gansu, and Sichuan were grouped into the “inland region” where marketization was relatively slow. In addition, we treated Beijing, the political center, and Shanghai, the largest industrial city in China, as two distinct regions. The model speciﬁcation is the same as that in Table 7.2, except that the variance components for city location are not estimated for Beijing and Shanghai, because only a single city is involved in each analysis. In addition, we included the “party membership” variable in these models, except for the “coastal region,” where the party membership information was not available for the two provinces in this region. Table 7.3 reports the parameter estimates for these regions, using the ﬁrst speciﬁcation of cadre/professional status. We organize our discussion around the main research issues. Returns to Party Membership. In the pre-reform era, party membership showed a signiﬁcant and positive effect in Beijing and the inland region in the pre-reform era. It contributed to about 10 percent higher income

217

Intercept Party membership Female Age Age2 /100 education Middle school High school College occupation High-rank cadre Low-rank cadre High-rank professional Low-rank professional Ofﬁce worker Service worker Skilled worker

Covariates

−0.027 −0.015 −0.054 0.069 −0.034 0.031 −0.032 0.046 −0.021 −0.076†

0.149∗ 0.106∗ 0.153∗ 0.111∗ 0.037 0.039 −0.012

1.096∗∗ −0.033 0.021 −0.028∗ 0.028∗

3.519∗∗ 0.090∗∗ −0.167∗∗ 0.061∗∗ −0.065∗∗∗

0.80∗ 0.138∗∗ 0.175∗∗

Interaction Effectsa

Main Effects

Beijing

0.222 0.148∗∗ 0.169∗ 0.084† 0.055 −0.005 0.072∗

0.058 0.065 0.039

3.722∗∗ 0.052 −0.180∗∗ 0.076∗∗ −0.086∗∗

Main Effects

0.021 −0.127∗ −0.128 −0.099† −0.077 −0.027 −0.135∗∗

−0.068 −0.010 0.113

1.502∗∗ 0.038 −0.014 −0.036∗∗ 0.039∗∗

Interaction Effectsa

Shanghai

0.222† 0.118∗ 0.044 0.016 0.041 0.067 0.011

0.062 0.142∗∗ 0.147∗

3.141∗∗ — −0.122∗∗ 0.070∗∗ −0.083∗∗

Main Effects

−0.098 −0.011 0.021 0.051 0.024 −0.032 −0.039

0.025 0.026 0.113

0.981∗∗ — −0.062∗ −0.023∗ 0.029∗

Interaction Effectsa

Coastal Provinces

0.214∗∗ 0.042 0.037 0.006 −0.042 −0.002 −0.008

0.038 0.040 0.107∗

2.936∗∗ 0.050∗ −0.198∗∗ 0.055∗∗ −0.055∗∗

Main Effects

(continued)

0.055 0.086∗ 0.154∗∗ 0.083∗ 0.018 0.034 0.076∗∗

0.082∗∗ 0.126∗∗ 0.108∗

0.549∗∗ 0.008 0.003 −0.021∗∗ 0.025∗∗

Interaction Effectsa

Inland Provinces

table 7.3. Parameter Estimates of the Determinants of Log(Income) in the Mixed Model, by Region

218

Main Effects

— — —

— —

0.035 0.026 0.011 −0.002 0.404∗∗

Interaction Effectsa

0.191∗∗ −0.00004 0.032∗∗ 921 5,145

— 0.024∗

0.210∗∗ 0.239∗∗ 0.242∗∗ 0.243∗∗ 0.265∗∗

Main Effects

— — —

— —

0.034 0.029 0.004 0.001 0.447∗∗

Interaction Effectsa

Shanghai

0.262∗∗ −0.00003† 0.022∗∗ 1,068 6,151

0.027 0.027∗

0.045 0.045 0.065 0.002 0.187∗∗

Main Effects

— — —

0.012 —

−0.019 0.041 −0.030 0.063† 0.165∗∗

Interaction Effectsa

Coastal Provinces

0.200∗∗ 0.00008∗∗ 0.016∗∗∗ 1,782 10,549

0.036∗ 0.007∗

0.101∗∗ 0.073∗ 0.181∗∗ 0.056∗ 0.234∗∗

Main Effects

— — —

0.019∗ —

0.065† 0.132∗∗ 0.028 0.030 0.127∗

Interaction Effectsa

Inland Provinces

∗

p < 0.01. p < 0.05. † p < 0.10. a The interaction effects indicate changes in the effects of the particular covariates in the 1987–1994 period. Note: Elementary school education or below is the reference category for education; unskilled manual worker, for occupation; and collective ﬁrm, for work organization. Both models also estimate a set of coefﬁcients for the random effects of the cities and the selected years.

∗∗

work organization Government agency 0.027 Public organization 0.011 Central gov’t ﬁrm 0.074∗ Local gov’t ﬁrm 0.065 Private/hybrid ﬁrm 0.154† G-covariance structure ν city — µyear 0.021∗ R-covariance structure σ 0.167∗∗ ρ −0.0002† ε 0.033∗∗ N 960 N× T 5,572

Covariates

Beijing

table 7.3 (continued)

Economic Transformation and Changes in Income Inequality 219 in Beijing, and to a 5 percent higher income in the inland region. In Shanghai, it had no signiﬁcant effect on income before and after the reform processes started. In the reform era, the magnitude of party membership appeared to decline in Beijing and increase in the inland region, but neither change was statistically signiﬁcant. Due to data limitation, there is no information on the role of party membership in the coastal provinces where marketization processes were more advanced. However, insofar as Beijing and inland provinces had experienced various degrees of market reforms, the results show no evidence that the role of party membership has changed in signiﬁcance in the reform era. Returns to Human Capital. Changes in work experiences (as measured by the ﬁrst and second-order effects of age) show similar patterns across regions, as consistent with the trend reported in Table 7.2. But there were considerable regional variations in the effects of education, especially in the reform era. In the pre-reform era, there were signiﬁcant and positive returns to college education across regions, and returns to high school education were less systematic. The main exception is that there were no signiﬁcant returns to education in Shanghai. In the reform era, there were no signiﬁcant changes in returns to education in Beijing, Shanghai, or the coastal region. In contrast, returns to education increased more systematically (at all three levels) in the inland region where market reforms were less advanced. Returns to Cadre and Professional Status. In the pre-reform era, highrank cadres had advantages in income in Beijing and the inland region, compared with unskilled workers. They also had a marginally signiﬁcant and higher income in the coastal region. In Shanghai, the magnitude of high-rank cadres was large but not statistically signiﬁcant. In both Shanghai and the coastal region, low-rank cadres had signiﬁcant, higher income compared with unskilled workers. Returns to high-rank professional status were signiﬁcant in Beijing and Shanghai but not signiﬁcant in other regions. This is probably due to the fact that high-rank professionals were concentrated in these two cities. In the reform era, low-rank cadres and low-rank professionals in Shanghai had a signiﬁcant and lower income than in the previous period. In the inland region, professionals and low-rank cadres all had signiﬁcant and higher income, compared with unskilled workers. Cadres did show a lower income in the coastal region, but this decline was not statistically discernible. Overall, there was no evidence that returns to positional power declined substantially.

220

The State and Life Chances in Urban China

Organizational Hierarchy. There were noticeable variations in the effects of types of work organizations across regions in the pre-reform era. Shanghai and the inland region appeared to be most consistent with the conventional models of organizational hierarchy in state socialist China. In Beijing, only employees in central-government-owned ﬁrms had significantly higher income, compared with those in collective ﬁrms. In coastal provinces, there were no statistically discernible differences among types of work organizations. These patterns suggest that there were signiﬁcant differences in the institutional arrangements of urban economies prior to the reform. These institutional differences in the past may be an important factor to explain the trajectories of reform across regions. In the reform era, except for a signiﬁcant increase in income for public organizations (and a marginally signiﬁcant increase for government) in the inland region, there were no statistically discernible changes in other regions. Overall, there is no evidence that patterns of income distribution associated with types of work organizations have changed signiﬁcantly in the early 1990s. The main exception are the signiﬁcant and positive gains by private and hybrid ﬁrms. Findings are Based on the Second Measure of Cadre Status. We proposed a second measure of cadre/professional status that distinguishes those in the economic sector and those in the public sector. Using the second measure, we estimated the same regional models (not reported here). The results were similar to those in Table 7.3. The main difference is that, for inland provinces, cadres and professionals in the economic sector had a signiﬁcant and higher income in the reform era. These results appeared to be consistent with market transition theory. However, government and public organizations in this region also had a signiﬁcant and higher income in the reform era, whose magnitudes were even larger than the gains by cadres and professionals in the economic sector. Because cadres and professionals in the public sector were exclusively located in government and public organizations (because of the coding decision), the results indicate that, if we take into consideration the contribution from their workplaces, cadres and professionals in the public sector did not have income decline. In fact, their income may have increased in the reform era. But sources of income may have changed, with the role of workplaces becoming more important for those in the public sector. Summary. Overall, there were noticeable regional variations in the effects of the covariates in this set of analyses. Some are consistent with the market transition argument, especially with regard to increase in returns to private/hybrid ﬁrms. But, in general, there were few statistically

Economic Transformation and Changes in Income Inequality 221 discernible changes in parameter estimates for the key variables across the two periods. Even in the coastal region, returns to cadre status (in both measures) have not changed in signiﬁcance over time. These results are largely consistent with the patterns reported in Table 7.1.

comparison with the ols results and other studies Comparing the parameter estimates based on cross-sectional and panel data (Tables 7.1 and 7.2), we ﬁnd that the directions of effects for key covariates are largely consistent in these two sets of analyses. This is comforting evidence that ﬁndings based on cross-sectional data analysis in previous studies parallel those based on the panel data analysis. The ﬁndings based on the panel data analysis are also similar to other published studies in the urban setting. For instance, although model speciﬁcations and data vary across these studies, the estimated returns to party membership and/or positional power in this study are similar to those reported in Bian and Logan (1996), Xie and Hannum (1996), and Walder (1995a). The effects of organizational hierarchy in Table 7.1 are also consistent with the ﬁndings on “work-unit characteristic” in Bian and Logan (1996). The estimated returns to education in the reform era (in Table 7.2) appear to be higher than those reported in other studies. This may be due to several reasons: The panel data analysis may be better able to assess changes over time and to disentangle the covariation among education, occupation, and work organizations; our data also captured more recent changes in the course of economic transformation. The considerable variations in the effects of the covariates as indicated in Table 7.1 may reﬂect the confounding effects of substantive changes in the reform era, temporal changes in response to shifts in state policies, and sample variability. As we see in Table 7.1, choices of different years may lead to different conclusions about changes and stability over time. Panel data and the mixed model speciﬁed here have advantages in addressing the theoretical issues while controlling for random ﬂuctuations and unobservable individual attributes over time.

chapter summary This chapter is the ﬁrst in a series of studies of the economic transformation in China – the focus of our second theme in this book. Building on several threads of theoretical ideas in previous studies, we proposed to conceptualize institutional changes as processes of coevolution between

222

The State and Life Chances in Urban China

politics and markets and to focus on the interactions among the multiple processes to understand changes in patterns of income determinants in China’s economic transformation. On this basis, we examined changes in and sources of income inequality over time, especially in view of recent changes in the transformation of state socialism. The use of panel data analyses and the operationalization of key theoretical variables have improved over the research designs in the previous studies. It is worth pointing out that the estimated patterns of income determinants for key theoretical variables (returns to positional power and returns to types of organizations) in the pre-reform era were consistent with the main theoretical arguments of the state socialist stratiﬁcation order in the literature. This is important evidence that the operationalization of key concepts in this study is consistent with the theoretical speciﬁcations, and that the empirical ﬁndings (the main effects) establish a baseline for comparing changes in the effects of these covariates across the periods. The empirical evidence reveals some noticeable changes between the pre-reform era and the reform era. The strongest evidence consistent with market transition theory is the signiﬁcant role of private/hybrid ﬁrms in the new-income regime. These new organizational forms reﬂect institutional changes that break away from the traditional redistributive economy. Given that this type of work organization is most closely associated with market transactions, we infer that marketization processes have indeed altered channels of resource distribution in urban China. Also, increasing returns to education may be partly attributed to emerging labor markets that better realize values of human capital than before. On the other hand, the evidence also shows strong institutional persistence. First, regarding returns to political capital and positional power, there is no evidence that returns to party membership or cadre status (in both measures) “decline in signiﬁcance.” That is, we did not observe signiﬁcant changes in returns to the “redistributors” (high-rank cadres) relative to “producers” (unskilled workers). Second, except for the salient beneﬁts of employees in private/hybrid ﬁrms, there were no signiﬁcant changes in the organizational hierarchy across the two periods. Employees in collective ﬁrms, which were closer to market activities than those in the state sector, had not improved their income relative to those in the state sector (see also Wu and Xie 2003). To put these ﬁndings in a broad context, we note that changes in income determinants were most sensitive to changes in economic activities and

Economic Transformation and Changes in Income Inequality 223 captured the most salient aspects of economic transformation in urban China. In addition, our sampling scheme has a larger proportion of urban residents in large cities, which tend to change more rapidly than other types of cities in China’s economic reform. All these factors should help capture more recent and more dramatic changes due to the expansion of markets and, in this sense, they should favor hypotheses derived from market transition theory. However, we should point out that our observation span ended in 1994. Since then, tremendous changes took place in work organizations and in China’s economy, especially in the direction of market mechanisms of resource allocation.7 Finally, some comparative observations are helpful to put our ﬁndings in perspective. Returns to education in China were markedly low, even in the reform era, compared with those in industrialized societies. For instance, using the 1960 census data, Mincer (1974) estimated a return of 10 percent for each additional year of schooling in the United States (see also Psachargopoulos 1994; Zhao and Zhou 2002). In contrast, with women earning about 85% of men’s wages, gender inequality in urban China was also marked low, compared with industrialized societies. For example, until in the 1990s women earn less than 70 percent of men’s wage in the United States (Blau and Ferber 1992). Social inequality in China was markedly low even compared with that in other state socialist societies. Zhou and Suhomlinova (2000) compared income distribution between urban China and USSR on the eves of their respective transformation of state socialism (1978 in urban China and 1991 in USSR). The ﬁndings show many similarities between the two largest state socialist societies, with high-rank ofﬁcials and professionals having signiﬁcantly higher salaries. However, there are also noticeable differences. Gender inequality is much more salient in the USSR, with women earning less then 70 percent of men’s wage. Returns to education and to bureaucratic and professional status were much more salient in the USSR than in China, indicating that allocative processes are more hierarchically structured in the USSR than in China. Measured by the Gini coefﬁcient, a conventional measure of inequality in the literature, the USSR witnessed the largest 7

The publication of the original study in American Journal of Sociology (Zhou 2000a) triggered a commentary by Cao and Nee (2000) and a rejoinder by Zhou (2000b). Interested readers can ﬁnd more about the intellectual debates therein. One unsettling issue from the previous analyses is about changes in the rate of returns to education across historical periods for different social groups. Zhao and Zhou (2002) explored this set of issues in greater detail.

224

The State and Life Chances in Urban China

income inequality among state socialist societies in the former Soviet bloc in the mid-1980s (Gottschalk, Gustafsson, and Palmer 1997). In contrast, urban China had the lowest income inequality among state socialist societies in the mid-1980s (see Chapter 11 for further discussions on this issue).8 8

The Gini coefﬁcient on urban China does not take into consideration disparities between urban and rural areas, which are considerably higher than the urban areas alone. Unfortunately, we did not ﬁnd reliable Gini coefﬁcient for China as a whole for the mid-1980s.

8 Redistribution and Latent Economic Beneﬁts

In the last chapter, we examined changes in the distribution of economic beneﬁts over time, focusing on the determinants of personal income. This chapter continues this line of inquiry but looks into a distinct and critical aspect of economic beneﬁts in state socialist societies – latent economic beneﬁts in the form of fringe beneﬁts and welfare programs such as housing, health care, and retirement beneﬁts. We call them “latent economic beneﬁts” because they are not always directly associated with one’s personal income. Often times, one’s membership in a type of work organizations or in an economic sector provides the most important entitlement to these latent beneﬁts, regardless of his or her personal income. In many ways, the distribution of latent economic beneﬁts is more important than that of personal income: Theoretically, the allocation of latent economic beneﬁts is closely associated with the organization of redistributive institutions under state socialism; substantively, differences in latent beneﬁts across workplaces often far exceed, and compensate for, overt differences in personal income. This chapter is organized as follows: We ﬁrst outline the conceptual issues about latent economic beneﬁts for understanding the redistributive phenomena under state socialism. On this basis, we provide an overview of the evolution of the welfare programs in China’s state socialist economy. Finally, we analyze patterns of housing allocation as an empirical example of changes and continuity in the distribution of latent economic beneﬁts in urban China.

225

226

The State and Life Chances in Urban China

redistribution and latent economic beneﬁts Why should we pay attention to latent beneﬁts under state socialism? Before we look into the speciﬁc issues, let us ﬁrst consider an example that is, by now, familiar in contemporary, industrialized societies – policies on maternity leave for female employees in the workplace. Policy on Maternity Leave: An Illustration In urban China, the ﬁrst government regulation on “labor insurance and welfare” explicitly set up the policy on maternity leave as follows (Bureau of Policies and Regulations 1990, Vol. 2, p. 6): In her pregnancy and delivery period, a female employee is entitled to 56 days of maternity leave, with full pay; If a female employee has miscarriage (i.e., less than 7 months of pregnancy), she is entitled to up to 30 days of leave with full pay, based on doctor’s recommendation; If a female employee experiences difﬁculty in delivery or has a twin, she is entitled to 14 extra-day maternity leave with full pay; A female employee’s pregnancy medical check and delivery cost is to be paid by the factory administration; If a female employee has used up her maternity leave but still cannot work, her extended leave may be treated as sick leave according to related regulations; When a female employee or the wife of a male employee gives birth, the employee is entitled to ¥4 from the labor insurance fund . . .

By now, these stipulations sound familiar to citizens in industrialized market societies. But, this policy was established in China more than half a century ago, in 1951, and they were applied to all organizations with more than 100 employees! These policies were very generous, even according to today’s standard and practice in the welfare states of the industrialized societies. Indeed, as one goes through the evolution of these regulations on maternity leave, one cannot help but marvel at the extent of the “welfare state” in socialist China in the early 1950s!1 Since the early 1950s, this policy has been elaborated and modiﬁed over time, but the basic framework has remained intact. In 1965, the Compensation Bureau of the Ministry of Labor issued a directive instructing that: If a female employee gives birth out-of-wedlock, she is not entitled to the maternity beneﬁts as stipulated in the labor insurance and welfare regulation. She is not 1

For comparisons with similar policies on maternity leave in the United States, see Kelly and Dobbin (1999) and Guthrie and Roth (1999).

Redistribution and Latent Economic Beneﬁts

227

entitled to paid maternity leave. For those with economic difﬁculties, the ﬁrm administration can provide subsidy as appropriate. (p. 78)

The maternity policy was revised in 1982 – for those whose pregnancy was in compliance with the government’s family planning regulation: A female employee’s maternity leave has 90 days, including 15 days of leave prior to the delivery; she is entitled to additional 15 days if she experiences delivery difﬁculty, or has twins; If a female employee has miscarriage, she is entitled to a leave depending on doctor’s recommendation; When a female employee has a baby less one year old, she is entitled to breastfeed (including other feeding methods) twice in her daily work hours, 30 minutes each. The breastfeed time and travel time for this purpose are counted as part of work time. (p. 79)

The Issues The dimension of latent redistributive beneﬁts has not been carefully studied in the social science literature on China. Most studies of changes in social stratiﬁcation processes in China’s economic transformation have focused on income differences and used changes in the determinants of income to evaluate changes in stratiﬁcation mechanisms. As the ﬁndings in the last chapter show, in the era of economic transformation, the distribution of income has changed in important ways, especially with increasing returns to education and to employees in the private and hybrid ﬁrms. But, to assess patterns of social inequality under state socialism, an exclusive focus on personal income is not entirely satisfactory, because personal income is only a small proportion of the economic beneﬁts redistributed under state socialism. To understand stratiﬁcation in these societies, one needs to distinguish conceptually between manifest and latent economic rewards. Manifest, direct monetary rewards (especially income) are far less useful measures of economic position in China than in a market economy for at least two reasons. First, due to deliberate “destratiﬁcation” policies, income differences among social groups and occupations have been compressed, especially in the pre-reform era. Second, many economic rewards take the form of redistributive beneﬁts (e.g., health care, welfare programs, housing) that depend on the workplace and its property-rights relationship to the state. For a large proportion of China’s urban population, differences in subsidies across workplaces greatly exceed differences in personal incomes within a particular workplace. As a result, focusing on changes in income disparities without considering

228

The State and Life Chances in Urban China

the signiﬁcant, latent economic beneﬁts distributed through workplaces ignores key characteristics and components of China’s redistributive economy. Indeed, the prominence of latent economic beneﬁts is consistent with the principle and economic operation of the redistributive economy. The effectiveness of the redistributive system under state socialism is contingent on the capacity to mobilize resources through nonmarket, organizational channels. That is, the central planning system operates through the administrative allocation of a large proportion of resources across economic arenas and through the direct administrative interventions into economic activities. We can draw several general observations from the preceding discussion: First, under state socialism, as Szel´enyi (1978) pointed out, “the state redistributes surplus . . . [which] was never accumulated in personal income, but [which] was directly centralized in the State budget and reallocated according to centrally deﬁned goals” (p. 75). A signiﬁcant proportion of resources is allocated in the form of latent beneﬁts through redistributive channels by the central planning authority. In this sense, a focus on personal income may be misleading to understand the principles of resource distribution under state socialism. Second, these latent beneﬁts are not distributed equally to all citizens, or among urban residents. Rather, these latent beneﬁts are distributed in accordance with the political logic and reﬂect the priorities in state policies. In the Soviet Union, for example, in order to direct the labor ﬂows to the sectors instrumental to accomplishing political and economic goals, the state manipulated economic rewards to such extent that the “branches of industry exerted much greater inﬂuence on the level of income than any other factor” (Wesolowski and Wnuk-Lipinski 1992, p. 84). Researchers also observed that the intersector and interbranch differentials cut across occupational categories, with the effect that a “director of a plant in the consumer sector will generally earn less than his counterpart in a complex metallurgical plant” (Connor 1979, p. 226) or that “a messenger or a cleaning woman can earn more than an engineer; a highly skilled worker more than a managing director . . . , if the former happens to be employed in an enterprise with very high average pay and the latter in an enterprise with very low average pay” (Flakierski 1993, p. 21). Walder (1992) developed a similar line of arguments in the Chinese context and found signiﬁcant variations in redistributive beneﬁts across economic sectors. In the Chinese context, the distribution of economic beneﬁts is mainly through work organization and the extent of latent beneﬁts depends on the

Redistribution and Latent Economic Beneﬁts

229

property rights relationships of work organizations to the state. In urban China, these patterns are closely associated with the ownership of these organizations. Employees in the state sector enjoy better welfare programs – better medical coverage, more generous sick leave, and better housing facilities – than those who are in collective or nonstate ﬁrms. In recent years, Chinese sociologists have examined the “workunit” (danwei) in detail (see Li and Li 2000; Li and Zhang 2000). Patterns of latent beneﬁt allocation provide information on the principles and channels through which economic beneﬁts are allocated in state socialist societies.2 Third, as the episodes of changes in maternity leave policy indicate, these welfare beneﬁts are also used as positive incentives for compliance with the morality of marriage and family (discouraging out-of-wedlock birth), and with the state’s family planning policy. Thus, the distribution of latent beneﬁts is an integral part of the redistributive system and reﬂects the principles of state socialism.

distribution of latent economic beneﬁts: historical overview In this section, we provide a historical overview of the allocation of latent beneﬁts in China’s redistributive economy. Because rural China has been largely excluded from the central planning and redistribution, the redistribution of latent beneﬁts in reality is conﬁned to urban residents, especially to those working in the state sector, and to a considerably lesser extent, to semistate ﬁrms (e.g., collective ﬁrms). In as early as 1951, the state council issued its directive to establish the regulation of labor insurance for the purpose of “protecting workers’ health and lessening their living difﬁculties” (Bureau of Policies and Regulations 1990, Vol. 2, p. 3). This regulation covered on-the-job injuries, sick leave, retirement fund, maternity leave, among other items. It applied to all ﬁrms with more than one hundred employees, including state ﬁrms, collective ﬁrms, hybrid ﬁrms, as well as private ﬁrms in major industries. For each ﬁrm, the labor insurance funds were equivalent to 3 percent of the total wages for its employees. The labor insurance funds were “entirely paid by the ﬁrm owner covered by this regulation. Part of the fund is to be directly paid by the ﬁrm administration; the rest through the payment 2

Recent research shows that latent beneﬁts are important even in industrialized market societies, as exempliﬁed in the social policies of the welfare states and hidden payrolls (O’Rand 1986).

230

The State and Life Chances in Urban China

table 8.1. Distribution of Welfare Expenditure, by Type of Organization Type of Organization

(1) Number in labor force (in millions) (2) Number of retirees (in millions) (3) Total number of workers/retirees (row 1 plus row 2) (4) Average annual income (in RMB) (5) Total welfare (in billions of RMB) (6) Average annual welfare per capita (in RMB)(row 5 divided by row 3) (7) Average welfare as percentage of average annual income (row 6 as a percentage of row 4)

Othera

State

Collective

112.1 22.5 134.6

32.9 6.2 39.0

7.6 0.6 8.2

4,797 164.6 1,222.8

3,245 24.8 635.5

6,303 6.4 779.0

25.5%

19.6%

12.4%

a

“Other” organizations include work units categorized as “joint-owned,” “share-holding,” “foreign-funded,” “overseas Chinese funded,” and “other ownership,” but excludes private entrepreneurs. See SSB (1995, pp. 84–85). Source: Information is drawn from SSB (1995). See pp. 84–85 for row 1; p. 688, for row 2; p. 112 for row 4; and p. 685, for row 5. Row 5 includes “labor insurance and welfare funds” (see detailed deﬁnitions on p. 710 of SSB 1995).

of labor insurance premiums by the owners, to be managed by the labor unions” (pp. 3–4). This regulation was greatly elaborated and formalized in 1953. Over the years, this regulation has been revised, modiﬁed, and at times unattended to, but, generally speaking, it has provided the basic framework for labor insurance and welfare policies in urban China over the last few decades. The welfare programs and policies were seriously undermined in the era of economic reform. As state ﬁrms gained more autonomy in resource allocation, many reduced or eliminated the welfare beneﬁts. According to a directive jointly issued by several government agencies in 1983 (p. 24): In the [decentralization] processes, some state-owned retail and service workunits have inappropriately lowered or even eliminated some labor insurance and welfare beneﬁts that workers are entitled to. For example, some ﬁrms reduced or eliminated payment in sick-leave payment, maternity leave, retirement pension; some reduced the beneﬁts for on-the-job injuries and medical coverage . . . some eliminated child care, food service, and clinics, and subsidies such as funds for home visit travel, heating, and female employees’ baby-feeding, etc.

Despite these noticeable changes under way, during the period of our study, latent beneﬁts were still substantive. Evidence on the manifest and

Redistribution and Latent Economic Beneﬁts

231

table 8.2. Distribution of Expenditure on Housing Construction, by Type of Organization Type of Organization State (1) Number in labor force (in millions) (2) Average annual income (in RMB) (3) Total annual residential housing investments (in millions of square meters) (4) Construction costs (RMB in square meter) (5) Annual residential housing investment per capita (in RMB; row 3 times row 4 divided by row 1) (6) Annual residential housing investments per capita as a percentage of average annual income

Collective

112.1 4,797 169.8

32.9 3,245 13.6

797.0 1,206.9

797.0 330.1

25.2%

10.2%

Source: Information is drawn from SSB (1995). See pp. 84–85 for row 1; p. 112, for row 2; pp. 149 and 179, for row 3; and p. 188, for row 4.

latent economic rewards distributed by different types of organizations illustrates this point. Table 8.1 displays the ofﬁcial 1994 statistics on “labor insurance and welfare funds” (laobao) “paid by enterprises, organizations, and institutions to their employees and retired workers and staff in addition to their wages and salaries” according to the State Statistical Bureau (SSB 1995, p. 710, italics added). These funds are not part of personal income and are either allocated to workplaces by direct administrative redistribution (via policies on proﬁt retention) or dictated indirectly by state policies. Table 8.1 omits private ﬁrms and family businesses because their employees do not receive such funds. According to the State Statistics Bureau, average annual income (row 4) varies substantially across the three types of organizations, being noticeably highest in “other” types of organizations (which include mainly nonstate ﬁrms). But the apparent advantage of “other” types of organizations looks much smaller when latent beneﬁts are considered: Compare the average levels of labor insurance and welfare beneﬁts for employees in the three types of organizations (row 7). The average beneﬁt in state organizations is twice the average for collective ﬁrms and 1.5 times the average for “other” types of organizations. Housing is another latent beneﬁt that varies with type of organization. Table 8.2 reports capital investments in housing construction in 1994 for state organizations and for collective ﬁrms. Capital investment in

232

The State and Life Chances in Urban China

residential housing construction per employee is 25 percent of the average annual income in state organizations, but only about 9 percent in collective ﬁrms. “Other” types of organizations (i.e., nonstate ﬁrms) rarely invest in or provide housing for their employees. Over many years, the cumulative beneﬁts to employees from housing investments made by state organizations are enormous. Clearly, latent beneﬁts redistributed through workplaces are a large proportion of total beneﬁts in state organizations and are not strongly associated with personal income. To contrast the latent beneﬁts provided by workplaces with the individual-level determinants of income, consider this fact: The percentage increase in personal income attributable to key individual characteristics (see Bian and Logan 1996; Peng 1992; Walder 1995a; Xie and Hannum 1996) shows that education (measured in years or by completion of the next higher level of education) and political position (measured by being a Communist Party member or a cadre) increase personal income by 4 to 16 percent. It is striking that total beneﬁts vary more with latent beneﬁts associated with workplaces (see Tables 8.1 and 8.2) than with the income differences associated with key individual attributes like education and political position. The fact that nonmonetary transfers (e.g., welfare funds) based on state policies continue to be so important in 1994, ﬁfteen years after China’s economic reform, suggests that redistributive mechanisms still have important impacts on life chances and that workplaces remain a major vehicle for the allocation of economic beneﬁts in urban China. Chinese sociologists Peilin Li and Yi Zhang (2000) provided a comprehensive assessment of the welfare programs in state ﬁrms. Based on a survey of 508 ﬁrms, they found that, in as late as 1995, welfare programs in state ﬁrms were still prevalent compared with other types of ﬁrms. For example, in state ﬁrms, the personnel involved in welfare functions comprised of nearly 5 percent of the total employees in the ﬁrms, whereas this statistic was less than 1 percent for collective ﬁrms, private ﬁrms, or joint venture ﬁrms. The most striking fact was that the expenditure by these welfare functions was over 450 percent of the ﬁrms’ proﬁts for state ﬁrms, but less than 1 percent for private ﬁrms and joint venture ﬁrms (no data for collective ﬁrms). That is, the state ﬁrms’ expenditure on welfare programs greatly exceeded their proﬁts. According to an estimate, between 1980 and 1994, expenditures on welfare beneﬁts for employees in the state ﬁrms increased at a rate that was 26.4 percent higher than the increase in total wages. By 1994, welfare beneﬁts were 31.8 percent of the total wage for state ﬁrms, 24.3 percent for collective ﬁrms, and 14 percent for other types of ﬁrms (Li and Zhang 2000, p. 212).

Redistribution and Latent Economic Beneﬁts

233

We will return to the recent trend of changes in the distribution of latent beneﬁts at the end of this chapter, but in the following section we focus on a case study of latent beneﬁts – the allocation of housing space in urban China.

a case study: determinants of housing distribution To examine patterns of the distribution of latent economic beneﬁts, we focus on housing allocation in urban China. This choice is partly dictated by the availability of data for such analyses: Other forms of latent beneﬁts – subsidies, health care, and retirement funds – are not well documented, are subtle to grasp, and are difﬁcult to measure, especially at the individual level. In contrast, information on housing accommodations is easier to collect through retrospective surveys. More important, housing allocation is an extremely important aspect of social inequality in urban areas. In his study of housing allocation in state socialist Hungary, Szel´enyi (1983) pointed out that, given the spatial scarcity in urban areas, housing space has been one of the most valuable resources in most socialist societies. This is especially true in China’s highly populated urban areas. Therefore, housing distribution is likely to reﬂect the fundamental principles on which scarce resources are allocated. Khan and Riskin (2001) also showed that housing subsidies took a large proportion of income of urban residents despite its recent decline in the reform era. Moreover, rent value of owned housing and housing subsidies are the most important source of inequality for urban residents in China, accounting for about 37 percent of the urban Gini ration (p. 36). An examination of housing distribution may address some fundamental issues about redistribution under state socialism. First, until recent years, housing construction and distribution were ﬁrmly controlled by the central government. Investment in housing, as part of “basic construction” fund, was an important category of ﬁnancial allocation through central planning. The main source of housing supply came from governmental budgetary allocation in housing construction. The absence of a housing market until the 1990s means that the distribution of housing space was exclusively based on redistribution controlled by governmental agencies. Therefore, an examination of housing distribution sheds light on the processes in the distribution of economic resources. Second, housing construction and allocation also capture multiple bureaucratic channels in the redistribution processes. Except for a small

234

The State and Life Chances in Urban China

proportion of urban residents who own their private housing prior to the founding of the PRC, housing space has been allocated through multiple bureaucratic channels such as local governments (the housing authority of the municipal governments) or one’s work organization. Because a large proportion of housing construction and allocation depends on one’s workplace, the distribution of housing space reﬂects patterns of redistribution based on the hierarchy of work organizations. Before we report patterns of determinants of housing space, a brief description of the housing administration in urban China is in order. In the urban areas, there are several sources of housing supply. One source is private housing ownership inherited from the pre-1949 era. Due to the limited housing construction in the Mao era, a signiﬁcant proportion of urban residents, especially those with low political status, stayed in their pre-PRC housing projects till the late 1980s. In general, housing conditions for this group were poor, lacking basic facilities such as running waters or toilet. The second major source of housing supply is from the housing authority under the metropolitan government, which allocates apartments or other dwelling accommodations to urban residents within their jurisdiction. The third source is from one’s work organizations, whose housing advantages closely depend on the hierarchical orders of type of work organizations. From a historical point of view, we should note two important constraints in housing allocation in the urban areas. First, little new housing construction took place between the mid-1960s and 1978 in urban China. As a result, a signiﬁcant number of urban residents were still living in housing projects inherited from the pre-revolution era at the time of our data collection in 1994. Second, in large cities where urban space is scarce (e.g., Shanghai), housing construction has been tightly controlled by the metropolitan government; the role of work organizations was thus limited in the direct supply of housing to their employees. We need to keep these issues in mind in interpreting the empirical ﬁndings reported in the following. In the rest of the chapter, we focus on analyses of patterns of the allocation of housing space in urban China based on our sample. Determinants of Housing Space We use living space (in square meters) in a respondent’s residential apartment/house as an indicator of housing space.3 If the respondent is 3

Because of housing scarcity, younger employees tended not to have their own apartments and lived in dormitories provided by the workplaces. Some employees may have no

Redistribution and Latent Economic Beneﬁts

235

married, housing allocation may come from either side of the couple. If respondent’s spouse received the housing allocation, we use the attributes of the spouse in the analyses. Following the conventional practice, we use the logarithm of housing space in the statistical analysis to minimize the effects of the cases with extreme values. Compared with income, changes in housing conditions (i.e., move to a different apartment) occur less frequently for an urban resident and his or her family. Therefore, we analyze the determinants of housing space and housing quality in selected years. To facilitate contrast with the determinants of income over time, we use the same selected years as those in income analysis (1960, 1965, 1975, 1978, 1984, 1987, 1991, and 1993). Table 8.3 reports the parameter estimates of the determinants of housing space in the selected years. These selected years give us snapshots of the distribution of housing space across distinct historical periods. As indicated at the bottom of the table, sample sizes varied over the years. Given the small sample sizes, we need to interpret the coefﬁcients in the earlier years with caution. The adjusted R2 shows that the proposed model has reasonably good explanatory power, accounting for about 23 to 36 percent of the variations in living space. However, it should be noted that a large proportion of the variations are explained by the set of indicator variables for city localities (estimated as part of the model, but not reported in the table). Although gender had signiﬁcant effects in most of the years, one should not take these results seriously, because, in the Chinese urban housing allocation system, either side of a married couple can get housing allocation from his or her workplaces depending on housing availability in the workplaces, or from the housing authority. Some municipal governments have regulations to stipulate that housing allocation should be mainly responsible from the female or male side of the workplace. The general rule of thumb is that the couple would try to get housing from the workplace of that spouse that could provide better housing accommodations. Household size, as measured by the number of residents living in the apartment, had a consistent and positive effect on living space. Its increasing effect in recent years reﬂected partly larger housing space and partly smaller family size. This ﬁnding captured an important characteristic; that is, housing allocation was developed on a “need-based” principle.

housing space at all. We excluded these respondents from the analyses. Some stayed with their parents, relatives, or friends. We included these respondents as a separate category. We also excluded the records of those living in rural areas or military residence from the analysis since the allocation of housing for these groups is qualitatively different.

236

Intercept Female Number of residents Age Age2 /100 education High School College occupation High-rank cadre Low-rank cadre High-rank prof Low-rank prof Ofﬁce worker Service worker Skilled manual worker

Covariates

∗∗

∗∗

∗∗

0.003 0.021 0.443∗∗ 0.093∗ 0.175∗ 0.028 0.061 −0.040 −0.035

0.257∗ 0.061 0.120 −0.010 −0.005 −0.072 −0.062

0.065 0.040 −0.028 −0.147† −0.208† −0.120 −0.115∗

2.963 0.031 0.089∗∗ −0.026† 0.041†

1975

0.018 0.182∗

2.149 0.085∗ 0.116∗∗ 0.019 −0.029

1965

0.042 0.121

2.937 0.088∗ 0.098∗∗ −0.032 0.063

1960 ∗∗

0.357∗∗ 0.087∗ 0.114† 0.072 0.020 −0.022 −0.031

0.021 0.007

2.687 0.037† 0.097∗∗ −0.012 0.022

1978

0.338∗∗ 0.049 0.176∗∗ 0.064† 0.066 0.056† −0.023

0.314∗∗ 0.087∗∗ 0.117∗∗ 0.050 0.039 0.039 −0.037

0.038† 0.050

∗∗

0.054∗∗ 0.008

1987 2.759 0.032† 0.108∗∗ −0.018∗∗ 0.028∗∗

∗∗

2.889 0.046∗ 0.098∗∗ −0.023∗∗ 0.036∗∗

1984

∗∗

0.270∗∗ 0.113∗∗ 0.125∗ 0.061∗ 0.079∗ 0.030 −0.020

0.018 0.075∗

2.625 0.032† 0.117∗∗ −0.012∗ 0.021∗∗

1991

table 8.3. OLS Estimates of the Determinants of Living Space (Log[Square Meter]), Selected Years

0.243∗∗ 0.107∗∗ 0.067† 0.047 0.078∗ 0.002 −0.044†

0.027 0.098∗∗

2.482∗∗ 0.039∗ 0.123∗∗ −0.006 0.016∗

1993

237

0.059 −0.002 −0.029 0.028 0.155

−0.034 0.219∗∗ −0.418∗∗ 0.30 994

0.120 0.205∗ −0.012 0.061 0.017

−0.040 0.205∗∗ −0.250∗∗ 0.25 701 0.012 0.265∗∗ −0.133∗ 0.23 1,979

0.028 0.005 0.007 0.043 0.070 0.004 0.244∗∗ −0.107∗ 0.23 2,363

0.058 0.034 −0.009 0.042 0.012 0.015 0.226∗∗ −0.117∗∗ 0.29 3,113

0.099∗∗ 0.024 0.038 0.048 −0.032

0.134∗∗ 0.054† 0.021 0.026 0.080∗ 0.092∗∗ 0.229∗∗ −0.045 0.35 3,368

0.082∗ 0.062† 0.039 0.046† −0.008 0.076∗∗ 0.240∗∗ −0.023 0.31 3,257

0.088∗∗ 0.246∗∗ −0.018 0.36 3,301

0.125∗∗ 0.057† 0.037 0.032 0.023

∗

p < 0.01 p < 0.05 † p < 0.01 Note: Middle school is the reference category for education, unskilled manual worker for occupation, collective ﬁrm for work organizations, and housing authority for housing ownership. The models estimated here also include nineteen dummy variables for each city included in the analysis.

∗∗

work organization Government Public Central govt ﬁrm Local govt ﬁrm Private/hybrid ﬁrms housing ownership Workplace Private Other Adjusted R2 N

238

The State and Life Chances in Urban China

This ﬁnding is consistent with the patterns found in the USSR (Zhou and Suhomlinova 2000). Work experience, as indicated by the ﬁrst- and second-order effects of age, was not signiﬁcant in the Mao era (1960–1978), probably because of little changes in housing construction. Another possibility is that housing allocation was need-based and less affected by seniority. In the reform era, seniority became an important factor but follows a U-shape. Seniority increased housing space after one’s mid-thirties. The age effect increased over the selected years, partly reﬂecting overall improvement of housing situations in urban areas. In contrast to the ﬁnding that educational levels contributed consistently to income, there had no systematic and positive effects of education until recently. Since the early 1990s, a college education contributed signiﬁcantly to housing space, but the magnitude of contribution was relatively minor – only about 10 percent, compared to those with middle school or lower education. This ﬁnding indicates that the allocation of housing space followed principles that were signiﬁcantly different from those for personal income: Housing allocation was more consistent with the traditional egalitarian policy, at least in the early stage of the economic reform. Returns to occupational status are closely related to the redistribution of a scarce resource among social groups. The ﬁndings showed that high-rank cadres indeed had better access to housing resources than unskilled manual workers, the reference group. The differences between these two groups were between 30 and 40 percent across the years. Since 1984, returns to high-rank cadre status appeared to decline, indicated by the successively smaller coefﬁcients in the last three years. However, note the increasing importance of government and workplace ownership in recent years in Table 8.3. Since high-rank cadres were mostly in government agencies and receive housing from their workplaces, the joint effects of being a high-rank cadre in fact increased in recent years. Since the mid-1970s, low-rank cadres and high-rank professionals also had signiﬁcant and larger housing space than unskilled workers. But the magnitudes of advantages were much smaller. Clearly, these two groups were distinctive from the group of high-rank cadres. Interestingly, “ofﬁce workers” had signiﬁcantly larger living space in the last two years. Recall the ﬁnding in Chapter 6 that “ofﬁce workers” were a major supply of the “cadre” occupation. The proximity of this group to the redistributive authority appeared to have beneﬁted them in important ways. In

Redistribution and Latent Economic Beneﬁts

239

contrast, there were no signiﬁcant differences in living space among skilled, unskilled, and service workers. It is somewhat surprising that type of work organizations did not have systematic effects on housing allocation. This was mainly because the housing ownership variables have captured their effects. Since 1987, those who received their housing from their workplaces had signiﬁcant and larger space than those allocated through ofﬁcial housing authority (the reference category). Because most work organizations that were able to allocate housing were in the state sector (government, public organization, and state ﬁrms), this means that employees in the state sector beneﬁted, especially in housing allocation in the reform era. The ﬁndings show that work organizations did not have substantive difference in the Mao era, when housing allocation as a type of welfare program was more subject to egalitarian policies and was “need-based.” It is also noticeable that those in government agencies and, to a lesser extent, public organizations, had signiﬁcantly larger housing spaces than those in collective ﬁrms in the reform era. Inequality in housing distribution has accelerated in recent years, even when overall housing conditions have improved greatly during this period. In the reform era, with the decline of welfare beneﬁts such as housing allocation, the redistributive beneﬁts have especially favored core state organizations. Government agencies and public organizations took advantage of the shifts of state policies (e.g., the abolishment of the egalitarian principles) to obtain larger beneﬁts for themselves. Private housing owners had consistently larger living space throughout the entire period. However, it is worth pointing out that, in the period of our study, there were virtually no housing markets in urban China. Private housing was largely inherited from generations before 1949. These houses were typically old, not well maintained, and should not be construed as an indication of housing advantages. This fact also points to the limitation of using living space as the indicator of advantages in housing conditions. For this reason, we now examine another set of indicators of housing quality: the number of facilities in the residential apartment (house). Determinants of Housing Quality We use the number of facilities present in an apartment to indicate the housing quality. The facilities include (1) running water, (2) kitchen, and (3) restroom. Under the low living standard in urban China during most of the last few decades, these basic housing facilities were important

240

The State and Life Chances in Urban China

indicators of an individual’s or a family’s well-being. The dependent variable is a count of the presence of these items, ranging from 0 to 3. Table 8.4 reports the estimates based on a Poisson model. Compared with the previous ﬁndings, family and individual attributes contributed to important differences in the distribution of housing facilities. Household size had no signiﬁcant effects throughout the years. Nor did work experience over time. In the era of economic transformation, education appeared to have signiﬁcant and positive contribution to better quality housing. In contrast, the effects of occupational status resembled those in the previous ﬁndings. Since the mid-1970s, high-rank cadres have had better quality housing, with respect to the number of housing facilities; so have low-rank cadres and high-rank professionals. Similar to the previous ﬁndings, ofﬁce workers also beneﬁted in the 1990s. In contrast, there were no substantial differences among other occupational groups. Although the type of work organizations had no systematic effects on housing facilities, the signiﬁcant role of workplace ownership implies that employees in work organizations in the state sector enjoyed better quality housing than those in nonstate sectors. As we indicated before, private housing often was of low quality, despite large living spaces and more rooms. This view is conﬁrmed in the ﬁnding that private ownership had no signiﬁcant contribution to housing facilities during most years of our study.

chapter summary Patterns of the determinants of housing allocation show interesting contrasts with those of income determinants revealed in Chapter 7. As we noted before, housing allocation is indicative of latent economic beneﬁts, whereas income distribution is associated with manifest economic beneﬁts. The processes allocating these two types of economic rewards are different in important ways. We now discuss the implications. Implications for Understanding Redistribution and Stratiﬁcation Dynamics The analyses of housing allocation revealed interesting patterns in the redistribution of latent beneﬁts. On the one hand, housing allocation as a welfare program in socialist China was more subject to egalitarian principles, especially in the Mao era. Education, age, occupational status, and

241

Intercept Female Number of residents Age Age2 /100 education Senior high College occupation High-rank cadre Low-rank cadre High-rank prof Low-rank prof Ofﬁce worker Service worker Skilled worker

Covariates

∗∗

0.008 0.158

0.150 0.202† 0.399∗ 0.105 −0.093 −0.010 −0.006

−0.131 0.163 0.346 0.011 0.044 −0.078 −0.017

−0.841 0.203∗∗ 0.018 0.060 −0.090

1965

0.033 0.345∗

−18.859 0.183∗ 0.013 −0.092 0.179

1960

0.053 0.089 0.200† 0.161∗ 0.207∗ 0.094 0.174† 0.051 0.023

0.265∗ 0.209∗∗ 0.349∗∗ 0.064 0.185† 0.004 0.003

0.848 0.058 0.001 0.020 −0.018

1978

0.041 0.110

1.377 0.085∗ −0.004 −0.008 0.018

1975 ∗∗

0.253∗∗ 0.119∗ 0.192∗∗ 0.084† 0.085 0.053 0.006

0.064∗ 0.094†

2.614 0.055∗ 0.005 −0.008 0.014

1984 ∗∗

0.272∗∗ 0.127∗∗ 0.221∗∗ 0.085† 0.045 0.051 0.039

0.063∗ 0.092∗

3.111 0.024 0.006 0.006 −0.006

1987

∗∗

0.263∗∗ 0.130∗∗ 0.204∗∗ 0.056 0.108∗ 0.055 0.054†

0.040† 0.093∗∗

3.729 0.037† 0.007 0.009 −0.009

1991

table 8.4. Parameter Estimates of the Poisson Model for Number of Facilities,a Selected Years

(continued)

0.182∗∗ 0.116∗∗ 0.151∗∗ 0.024 0.091∗ 0.042 0.038

0.044∗ 0.099∗∗

3.644∗∗ 0.034† 0.007 0.009 −0.008

1993

242

1960

1965

1975

1978

1984

1987

1991

1993

∗

p < .01 p < .05 † p < .01 a Facilities include three items: running water, kitchen, and restroom. Note: Middle school is the reference category for education; unskilled worker, for occupation; collective ﬁrm, for work organizations; and housing authority, for housing ownership. The models estimated above also include nineteen dummy variables for each city included in the analysis.

∗∗

work organization Government 0.317∗ 0.254∗ 0.058 0.069 0.121∗ 0.108∗ 0.055 0.043 Public 0.297† 0.187 0.038 0.092 0.094† 0.096∗ 0.080∗ 0.066 Central govt ﬁrm 0.234† 0.114 0.013 0.017 0.066† 0.074∗ 0.038 0.044 Local govt ﬁrm 0.088 0.084 0.100† 0.070 0.071† 0.070∗ 0.035 0.012 Private/hybrid ﬁrms 0.087 0.046 0.075 0.005 0.022 −0.018 0.029 −0.007 housing ownership Workplace 0.134 0.176† 0.229∗∗ 0.201∗∗ 0.225∗∗ 0.207∗∗ 0.200∗∗ 0.208∗∗ ∗ ∗ Private 0.129 0.095 0.159 0.099 0.057 0.043 0.045 0.058† † ∗∗ † ∗ Other −0.319 −0.534 −0.088 −0.046 0.128 0.065 0.122 0.153∗∗ Log-likelihood −546.9 −792.5 −1623.9 −1909.0 −2606.4 −2696.1 −2657.2 −2489.0 d.f. 40 40 40 40 40 40 40 40 N 702 995 1977 2359 3112 3252 3364 3297

Covariates

table 8.4 (continued)

Redistribution and Latent Economic Beneﬁts

243

organizational type did not have systematic effects on having a larger living space or better facilities in the Mao era. This is in sharp contrast with income inequality over the last few decades. On the other hand, we ﬁnd that, with the downplaying and abolishing of the egalitarian policies, inequality in housing allocation became more salient in the reform era. Similarly, since the 1970s, high-rank cadres have enjoyed signiﬁcant beneﬁts in larger living space and facilities. Low-rank cadres and highrank professionals also appeared to have beneﬁted from redistribution. But the magnitudes of differences among social groups and the importance of household size suggest that the extent of inequality based on positional power was kept under control. The determinants of housing allocation highlight the critical importance of work organizations in China’s redistributive processes, as indicated by the role of housing ownership. This is not surprising. In fact, almost all latent economic beneﬁts, such as welfare programs, labor insurance, health care, and other fringe beneﬁts, are distributed through work organizations one is afﬁliated with. Differential latent beneﬁts associated with type of workplaces are important sources of social inequality in urban China. What is more revealing is the fact that, as far as housing is concerned, the role of work organizations became important only in the era of economic reform. For instance, with regard to living space, government agencies and workplace ownership had signiﬁcant advantages since the late 1980s (cf. Table 8.3). These results imply that one’s location in the social stratiﬁcation structure, as indicated by type of work organizations, does not necessarily confer economic beneﬁts. Rather, returns to membership in these organizations vary with state policies. When the state promoted an egalitarian policy of destratiﬁcation, as in the Mao era, even work organizations in the state sector did not have signiﬁcant advantages in housing conditions. But in the reform era when social inequality was legitimated, the disparity based on workplaces had increased considerably, as indicated in the case of housing allocation. These patterns have rendered further support to the importance of stratiﬁcation dynamics that we emphasized throughout these studies. Implications for Understanding Recent Economic Transformation Changes in the determinants of income and housing also inform us of the processes of institutional change in the era of economic transformations

244

The State and Life Chances in Urban China

since the 1980s. As the ﬁndings in the last chapter indicated, there were indeed important changes in income determinants, noticeably the increasing importance of private/hybrid ﬁrms and increasing returns to education since the 1980s. In contrast, patterns of housing allocation portrayed a different picture. Although there is some evidence of increasing returns to education, major changes since the 1980s indicate the increasing importance of type of work organizations and that those in the state sector beneﬁted most in the redistribution of housing allocation. Since 1992, the central government has experimented with the privatization of housing ownership in urban areas. By the end of the 1990s, a large proportion of housing projects were sold to urban dwellers. The privatization of housing signals the emergence of housing markets. Indeed, there are a large number of “commercial apartments” that can be purchased through real estate markets (Davis 1995). However, the implications of housing privatization are by no means straightforward. First, because the current housing residents purchase their apartments in heavily subsidized prices, often in accordance with their positions and work tenure, housing privatization has in fact legitimized permanent housing inequality. Second, there are still a large number of urban residents who are unable to afford housing purchase, because they are not afﬁliated with high-status work organizations that provide generous housing subsidies. There is also anecdotal evidence that when a cadre is promoted, he or she is able to change to better apartments through subsidized processes, continuing to reap the beneﬁts provided by the workplace. Therefore, work organizations continue to play a critical role in the emergence of housing markets (Bian et al. 1997; Logan, Bian, and Bian 1999). According to one statistic, among all of apartment purchases in Beijing in 1996, about 76 percent were purchased by government agencies, large corporations, and military headquarters, which were resold at a subsidized price to their employees. Eighty-eight percent of the purchases by work organizations were located in the best residential areas in Beijing. In contrast, among those purchases by individuals, 73 percent were in the unfavorable, remote suburb areas. Among those housing purchases by work organizations, state ﬁrms and collective ﬁrms comprise 38 and 33 percent, respectively; and government agencies and public organizations, 29 percent (Li and Li 2000, p. 39). Similarly, Khan and Riskin (2001) showed that the highly de-equalizing system of net subsidies and the promotion of housing reform noticeably increased the inequality in distribution of housing assets (p. 50). Inequality related to housing

Redistribution and Latent Economic Beneﬁts

245

(including rent value and housing subsidies) has become one of the most important elements of urban inequality in recent years. These contrasting patterns portrayed a picture of multifaceted processes in large-scale social changes, as we found in the last chapter. Relative to changes in personal income, inequality in housing space and other fringe beneﬁts has been more stable in the early stage of the reform era. The redistributive institutions played an important role in allocation of housing spaces, as reﬂected in persistent or even increasing beneﬁts for those closer to the redistributive authorities.

9 Institutional Changes and Patterns of Job Shifts

Since the 1980s, an increasing number of sociological studies have examined processes and consequences of transitions from state socialism, and the experience of economic reform in China has drawn particular attention. Central to these recent studies are assessments of large-scale institutional transformation and implications of these changes for mechanisms of social stratiﬁcation. As we noted before, an interesting observation is that most studies in this literature have used determinants of direct economic beneﬁts (e.g., income) to infer changes in institutions and mechanisms of stratiﬁcation (e.g., Bian and Logan 1996; Nee 1989, 1991, 1996; Peng 1992; Walder 1995a; Zhou 2000a). There is an implicit assumption behind the almost exclusive focus on personal income. In market societies where resource changes take place in the marketplace, personal income serves as the most important medium for resource acquisition and a good measure of one’s economic beneﬁts. Thus, to a large extent the primary focus on personal income to study differences in economic beneﬁts and social status is justiﬁed (but see O’Rand 1986). Many studies of the industrialized market societies and analytical models and research designs accumulated in the sociological and economic literature. As sociologists trained in these traditions began to examine social inequality in state socialist societies, there was a natural tendency to adopt similar analytical models and apply similar research experience to study personal income. But, as we argued and showed in the last chapter, because of the nature of latent beneﬁts distributed through work organizations in nonmonetary forms, an exclusive focus on personal income in state socialist societies is limited and may be misleading. 246

Institutional Changes and Patterns of Job Shifts

247

In this chapter we focus on access to opportunities within and across workplaces as an analytical strategy to address this question: To what extent have the redistributive institutions changed in the era of economic transformation? A salient fact emerging from our discussions in the preceding chapters is the centrality of work organizations in the stratiﬁcation order in China. Type of work organizations in their property rights relationship to the state is the cornerstone of the state socialist redistribution in urban China and is a major channel of stratiﬁcation associated with differential income, latent economic beneﬁts, social status, and access to political opportunities. In contrast to promotion patterns that typically take place within the same organization and along the same career line, job shift patterns are movements of employees from one workplace to another, which often involves major changes in one’s work environment and career. The extent of institutional changes, then, should be reﬂected in patterns of job shifts across work organizations. We compare job-shift patterns among types of organizations and economic sectors in urban China across historical periods, from 1949 to 1994, thereby establishing a baseline for examining changes in China’s redistributive economy. As we noted before, assessing changes in stratiﬁcation mechanisms over time on the basis of information for a single reference year is problematic because access to different types of organizations has varied considerably across historical periods with frequent shifts in state policies (Zhou, Tuma, and Moen 1996). Thus, an analytical focus on historical changes over time is especially appropriate for understanding the transformation of state socialism in China.

job-shift patterns as a focus of study Why Study Job Shifts? Our second theme in this book is to assess the extent and direction of institutional changes in China in the reform era. At the center of these changes are the transformation of institutions and mechanisms of resource allocation, especially the relative strength of the redistributive and market economies. In this light, the extent to which resources are allocated through market transactions rather than through bureaucratic channels provides a useful indicator of institutional changes. As we have seen in the last two chapters, traditional redistributive mechanisms and emerging market mechanisms coexisted and interacted with each other in the transformation processes during the period of our study.

248

The State and Life Chances in Urban China

A focus on job shifts among different types of organizations and among economic sectors provides an excellent vantage point to assess institutional changes in urban China. Patterns of job shifts in China reﬂect not only individual preferences and the desirability of workplaces but also, and more important, the relative strength of competing economic institutions. Because the hierarchy of workplaces and the structure of political authority in China are so closely linked, job-shift patterns reﬂect the arrangements of China’s economic and political institutions. Therefore, changes in job-shift patterns should signal signiﬁcant changes in these institutions (Zhou et al. 1997). Our main analytical focus is on jobs shifts among types of organizations. In China, type of organization refers to the ofﬁcial classiﬁcation of workplaces on the basis of their property-rights relationship to the state. This relationship is a key feature of the redistributive economy and is the basis of socialist redistribution: Ownership of an organization to a large extent determines employment relationships in a workplace, the extent of administrative interventions as well as governmental protection. Different types of organizations vary systematically with different kinds of economic institutions, with state organizations being part of the redistributive economy and collective and private ﬁrms close to markets. Therefore, job shifts among types of organizations provide valuable information about the relative strength of types of organizations. Among state organizations, we distinguish government agencies, public organizations, and state-owned ﬁrms. Collective ﬁrms are semistate organizations; private and “hybrid” ﬁrms (i.e., ﬁrms with mixed property rights ownership or foreign ﬁrms) are nonstate organizations. Although state organizations receive the most redistributive beneﬁts, they are also subject to the tightest state control. We also examine job shifts among types of economic sectors in China as another way to assess institutional change. Type of economic sector refers to industries/arenas with different technologies and production processes. In China, they also have different ties to the redistributive economy and market conditions. We distinguish heavy industry, light industry, the service industry, the public sector, and the government sector. The sectoral structure also reﬂects the relative dominance of command and market economies. Among these sectors, the government and public sectors are at the center of the redistributive system. Among the industries, heavy industry was especially favored in the Mao era when rushed growth was emphasized, whereas light industry and service industry were largely ignored or downplayed. In the reform era, various economic sectors have grown at different rates and experienced different degrees of marketization.

Institutional Changes and Patterns of Job Shifts

249

Light industry and the service industry have expanded enormously and become more market-oriented, compared with state-supported heavy industry. The service sector is the most decentralized and privatized. For instance, by 1994, 62.5 percent of employees and entrepreneurs working in private enterprises were in the “wholesale, retail trade, and food service” sector (SSB 1995, p. 100). Clearly, job shifts among economic sectors also provide information about the expansion and declines of different economic institutions. Given the organizational hierarchy in urban China, one’s location in a given type of workplace also implies his or her location in the stratiﬁcation system. Hence, job shifts can reveal both the changing status among types of organizations (and of economic sectors) and the information about institutional changes and evolving mechanisms of stratiﬁcation. If the redistributive economy declines and a market economy arises, it would become more difﬁcult for the government to sustain substantial redistributive beneﬁts based on the hierarchy of organizational types supported by the redistributive economy and by state policies. Employment in state organizations then would become less attractive, and employment in semistate and nonstate organizations would become more attractive than in the past. If there is effective market competition for human resources, in the reform era employees with greater human capital and more social resources are likely to leave jobs in state organizations to enter jobs in the more proﬁtable nonstate ﬁrms. However, the persistence of redistributive power and beneﬁts may still provide incentives for employees to stay in the state sector. In sum, job shifts across types of workplaces provide an excellent arena to assess changes in both institutions and mechanisms of stratiﬁcation. Several previous studies have pointed in this direction (Davis 1990, 1992a, 1992b; Lin and Bian 1991; Rona-Tas 1994; Zhou et al. 1996). These previous studies, however, are limited in scope (analyzing samples from only one or two cities) or time span (analyzing limited life-course information). The study reported in this chapter overcomes these limitations to a considerable extent. Explaining Job Shifts: Some Considerations To guide our analyses, we draw on discussions about different mechanisms of institutional changes in the recent literature. Given the continuing dominance of the state and state bureaucracies in China’s economic transformation, it is not surprising that the most significant constraints on job shifts are the existing redistributive institutions.

250

The State and Life Chances in Urban China

As we observed in the previous chapters, over the years the Chinese state developed an elaborate system of organizations to mobilize and transfer resources and to tie people to a web of workplaces controlled by the state. In the Mao era, entry into and exit from workplaces were tightly controlled by ofﬁcials in local governments and workplaces in accordance with directives from the central government. Most jobs were assigned by government agencies, and transfers between workplaces must be approved by the workplaces involved, and often also by local government agencies. In the post-Mao era, more and more economic transactions occurred in the marketplace, but the role of political authority – central and local governments – in economic processes has not fundamentally changed (Naughton 1996; Shirk 1993; Walder 1995b). State policies and administrative authorities also had more direct and immediate impacts on job shifts. In the traditional redistributive system, state exerted tight control of labor mobility and there were formidable institutional barriers for job shifts across different organizations and sectors. For instance, in the early 1980s, the central government still put many constraints on employees of state organizations (especially cadres and professionals) who wanted to leave their jobs and move to nonstate organizations. Davis (1990; 1992a) found only limited job mobility among those working in state-owned ﬁrms in the early reform era, which she attributed to restrictive state labor policies and greater beneﬁts available in state organizations. As we have shown before, in the organizational hierarchy under the redistributive economy, those in state organizations beneﬁted from the redistributive economy more than those in semistate and nonstate organizations. Because of persistent institutional barriers and state policies, this structural inequality between state and semi-/nonstate organizations has not changed substantially by 1994, the end year of our data collection. Several testable hypotheses can be drawn from these considerations. First, if these arguments are valid, rates of job shifts from workplaces in state organizations to ones in collective or nonstate ﬁrms should remain low throughout the history of the People’s Republic of China, and they should not differ substantially between the reform era (1980–1994) and the Mao era (1949–1979). Second, stratiﬁcation mechanisms should not have substantially changed. Speciﬁcally, individuals with more positional power (e.g., cadres) and more human capital (e.g., higher education) should be more likely to stay in state organizations than to leave them. On the other hand, the emergence of market economies since the 1980s was also evolving into an increasing force competing with state organizations for resources. As Nee (1989, 1991) argued, market mechanisms

Institutional Changes and Patterns of Job Shifts

251

provide incentives and opportunities that differ from those in the redistributive economy. Moreover, the emergence of a market economy raises the power of producers relative to that of redistributors. Accordingly, as a market economy emerges and the redistributive economy declines, beneﬁts formerly associated with political power and favorable locations in state organizations diminish relative to the economic returns offered to more market-oriented nonstate ﬁrms. Therefore, the rise of a market economy in China increases opportunities for those who have been most disadvantaged in the command economy (“producers”). Following the market transition logic, then, resource ﬂows and economic exchanges should be increasingly governed by the market, whose efﬁciency exceeds that of the redistributive economy. Accordingly, higher economic returns in the market economy should induce individuals to leave workplaces in the state redistributive system and to engage in market activities. This argument leads us to hypothesize that individuals will shift jobs in the reform era – from redistribution-based state organizations to market-oriented, semistate or nonstate organizations. Further, within state organizations, we should expect an increase in job shifts from those organizations close to the redistributive economy (government agencies and public organizations) to those more involved in market activities (state-owned ﬁrms). Moreover, the mechanisms of job shifts should also change. If market mechanisms prevail in the allocation of resources, we expect to ﬁnd that, in the reform era, individuals with human capital characteristics (e.g., a college education) that position them better in a competitive market should be especially likely to move from organizations in the redistributive economy to ones more closely linked to market activities. In contrast, the role of political power on job shifts (e.g., being a cadre or working in a government agency) should be less important during the reform era than before. The recognition of these competing mechanisms and processes underlying the ongoing institutional transformation implies that, similar to the patterns we observed in income determinants, we are likely to observe mixed ﬁndings that capture both continuity and changes in job shift patterns over time. Analytical Strategies in Modeling Job Shifts Our analysis focuses on job shifts after a person has obtained his or her ﬁrst job. Until recently, ﬁrst jobs were usually assigned by government administrators and rarely reﬂected individual choice (see Chapter 4). We

252

The State and Life Chances in Urban China

assume here that shift away from the ﬁrst job may reﬂect the desirability of a new workplace as compared to the original one, especially in the reform era. In the most time covered by this study (1949–1994), a person can stay in the same workplace for life. Therefore, this assumption is reasonable.1 We distinguish two types of job shifts. External job shifts are shifts to a different type of organization or economic sector. Internal job shifts are shifts to a different workplace but within the same type of organization or economic sector. Job shifts differ from promotions (analyzed in Chapter 6) in that the former involve changes in workplaces but not in one’s positions or bureaucratic/professional rank. Patterns of external shifts are especially informative about institutional change because they reﬂect the rise and decline of economic sectors and the types of organizations with different property-rights relationships to the redistributive economy. In contrast, internal shifts may reﬂect individuals’ efforts to relocate to workplaces closer to market activities while remaining in the same property-rights regime. So far, our discussion focused on changes in relative incentives due to the emergence of market mechanisms. But, job shifts also involve risks. In the early years of the reform, state policies and institutional constraints made moving from the state sector to the nonstate sector especially risky. In this uncertain environment, individuals at different stages of the life course are associated with differential risks. We include two life-course factors as statistical controls in our empirical analyses: age and era of labor force entry (i.e., cohort effects). Older workers tend both to beneﬁt most from the redistributive system and to be more risk averse. In addition, those who entered the labor force during periods of political and economic crises (e.g., during the Cultural Revolution) were more likely to be assigned initially to less desirable workplaces and were therefore more likely to beneﬁt from shifts to better workplaces. In the reform era, there had been profound changes in economic structures, with new opportunities and greater avenues for labor mobility. We expect that younger employees beneﬁt especially from these new opportunities (see Chapter 10 for further analyses of cohort differences in the reform era). As in our analyses of entry into ﬁrst jobs and promotion events, we adopt an event history approach to examine the rate of job shifts 1

Certain state policies (e.g., nationalization and collectivization in the late 1950s, the return of sent-down youths during and after the Cultural Revolution) caused huge numbers of job shifts. Our research design controls for the net effects of these massive shifts by including variables on “organizational origins” and historical periods in model estimation.

Institutional Changes and Patterns of Job Shifts

253

across types of organizations and economic sectors. The advantages of the event history approach for understanding labor force dynamics have been demonstrated by previous research on job shifts in industrialized societies (e.g., Carroll and Mayer 1986; Sørensen and Tuma 1981). Given the dramatic changes in China over the forty-ﬁve years of state socialist governance, one must consider changes in China’s labor force structure as well as in its state policies. For this reason, we adopt statistical models that allow period-speciﬁc changes in both labor force structure and state policies. We also control for structural changes by including measures of the relative strength of collective and nonstate organizations.

historical trends of job shift patterns To provide a general view of the job shift patterns over time in urban China, we ﬁrst take a look at the estimated hazard rates of job shifts (both external and internal) among different types of organizations and among different economic sectors for the entire labor force (i.e., everyone in the labor force in a given year). The hazard rate measures the instantaneous rate of change given the risk set. Figure 9.1 displays the rates of job shifts to government agencies, public organizations, state-owned ﬁrms, and collective ﬁrms; Figure 9.2 shows the rates of job shifts to private and hybrid ﬁrms. (Note that the scale on the y-axis has been enlarged to improve readability.)2 The trends in Figures 9.1 and 9.2 reveal the impacts of the twists and turns of state policies on the labor force over time. The noticeable increase in the rate of shifts to state-owned ﬁrms in the late 1950s reﬂects the state’s policies of nationalization and “rushed-growth” (adopted during the Great-Leap-Forward period). The decline thereafter resulted from the economic contraction in the early 1960s. The beginning of the Cultural Revolution in 1966 was marked by another noticeable decline in the rate of job shifts to state organizations. The sharp increase in the late 1970s was due largely to the accommodation into urban economies of the millions of the urban youth who had been sent to the rural areas during the Cultural Revolution. It is interesting to note that, in most years, the patterns of hazard rates of shifts to state-owned and collective ﬁrms, and to a lesser extent to hybrid ﬁrms, roughly track one another. That is, shifts to these different 2

Due to the small number of events, in some years the hazard rates for shifts to private and hybrid ﬁrms cannot be estimated, resulting in the jumpiness for the plots in Figure 9.2.

The State and Life Chances in Urban China

254 0.08 0.07

hazard rate

0.06 0.05 0.04 0.03 0.02 0.01 0 50

55

60

65

70

75

80

85

90

95

90

95

year government

public

state firm

ﬁgure 9.1. Hazard rates of job shifts to the state sector.

hazard rate

0.03

0.02

0.01

0 50

55

60

65

70

75

80

85

year collective

private

hybrid

ﬁgure 9.2. Hazard rates of job shifts to collective and nonstate ﬁrms.

Institutional Changes and Patterns of Job Shifts

255

types of organizations responded similarly to the macro-political conditions. These drastic ﬂuctuations over time provide evidence that the rates of job shifts among organizational types can be directly traced to shifts in state policies and development strategies. In the post-Mao era (1980–1994) the hazard rate of job shifts to nonstate organizations (see Figure 9.1b) increased substantially, whereas the rate of shifts to state organizations declined. These patterns point to a broad trend of the declining of the state sector and the expansion of the nonstate sector. But the absolute magnitude of the rate of job shifts to nonstate organizations was still far smaller than the rate of shifts to state and collective organizations because nonstate ﬁrms still had a relative small proportion in the national economy. The sharp drop in the rate of shifts to private ﬁrms between 1990 and 1992 almost surely reﬂected the adverse state policies following the political turmoil in China in 1989, whereas the sharp increase afterwards reﬂected the resumption of economic reform and market expansion by Deng Xiaoping’s new initiative in 1992.

explaining job shift patterns across historical periods There are two key issues concerning the impacts of changing institutions on job-shift patterns across historical time. First, to what extent has a market economy emerged and eroded the redistributive economy in the reform era? Empirically, to what extent were job-shift patterns associated with a workplace’s location in the redistributive hierarchy? To address this issue, we examine the effects of the organizational type and the economic sector characterizing a person’s current workplace (i.e., his or her job origin) on his or her propensity to move to a different type of organization or to a different economic sector (his or her job destination). Second, to what extent have stratiﬁcation mechanisms changed across historical periods? To answer this question, we compare the effects of theoretically relevant covariates across periods. External Job Shifts: Moves Across Types of Organizations We estimated a set of logistic models for discrete-event history to study shifts to jobs in different types of organizations in three periods: 1949– 1965, 1966–1979, and 1980–1994. Table 9.1 reports the results. The columns indicate the types of organizations as job destinations, whereas

256

Intercept Female Age education High School College occupation Cadre Professional Military service Service worker job origin Government Public org’n State-ﬁrm Hybrid/private ﬁrm Farm Cohort 1966–1979 1980–1994 % industrial output Collective Hybrid/private ﬁrm χ2 d.f. N of events

Public Organization

0.368 —

— —

−0.027 0.334∗∗ 194.8 24 155

0.112 −0.117 −0.226 1.519∗∗

−0.683 −0.626∗ −0.234 0.551†

−0.001 0.002 101.1 23 94

0.161 0.084 −0.551 −0.378

0.003 −0.033∗ 204.5 25 168

0.349 0.169

0.014 0.190 0.060 1.944∗∗

0.053 −0.284 −1.065 0.067

0.588∗∗ 2.093∗∗

0.798∗∗ 0.465

0.520† 1.267∗∗

−0.321 −0.672† −1.175∗ 0.601†

−5.491 −0.170 −0.040∗∗

−4.811∗∗ −0.637∗∗ −0.022

−2.129∗∗ −1.494∗∗ −1.107∗∗

0.033 0.014∗ 89.9 23 67

— —

0.043 0.279† 189.3 22 127

−0.034 −0.032† 190.7 25 128

0.391 −0.063

1.114∗∗ —

0.238 — 0.071∗∗ 0.216∗∗ 782.4 23 476

0.049∗∗ 0.017∗∗ 143.8 23 196

— 2.493∗∗ — —

0.545 1.280∗∗

0.016 −0.042∗∗ 437.8 25 297

−0.081 −0.204

−0.549 2.599∗∗

0.710∗∗ 0.235

1.292∗∗ 0.727∗∗ 0.299 0.179

0.517† 0.242 −1.420 2.339∗∗

0.113 −0.230 1.642∗∗ 0.137

−0.120 −0.337† 1.292∗∗ −0.098

−0.717∗ −0.424 0.974∗∗ −0.524

0.267† 0.134

−3.544∗∗ −0.381∗∗ −0.060∗∗

−0.269 −0.066 1.293∗∗ −0.231

−6.366∗∗ −0.075 −0.020∗ 0.395∗∗ 0.085

−5.976∗∗ −0.480∗∗ 0.016

1966–1979 1980–1994

State Firm

0.210 −0.069

0.964∗∗ 1.913∗∗

−4.886∗∗ −0.134 −0.029†

−0.640∗ — 1.346∗∗

0.191

0.582 −0.678† −0.295 −0.083

0.036 0.040 1.113∗ —

0.550∗ 1.583∗∗

−7.672∗∗ 0.272 0.011

−1.244 0.899∗ 1.235∗ −0.190

0.222 0.721†

−5.487∗∗ 0.298 −0.031

1949–1965 1966–1979 1980–1994 1949–1965 1966–1979 1980–1994 1949–1965

Government

table 9.1. Parameter Estimates of Logistic Model for Job Shifts Outside Type of Work Organization, by Period

257

Private Firm

−0.404 −0.723∗ 1.492∗∗ −0.060 0.085 0.387 0.237 2.806∗∗ 0.330 0.678† 0.011 −0.042∗ 304.0 25 179

−0.345 −0.361 0.084 −0.774∗

0.754∗∗ −0.174

−1.608 1.969∗∗

0.397∗ —

0.072∗∗ 0.110 362.4 24 290

−0.598 −0.418 −1.101 0.263

0.207 0.376

−0.382 1.025∗∗

— —

0.012 0.005 60.1 23 111

0.153 0.331

0.051 −0.204

−0.095 −0.597

−5.990∗∗ 0.199 −0.027∗

−7.064∗∗ 0.152 −0.006

−6.250∗∗ 0.122 0.005

26

46

0.023 0.009 121.2 24 106

0.294 0.891†

1.302∗∗

−0.066 −1.024∗ −0.336

−0.204 −0.381 0.379 −0.113

0.355 0.836∗

−6.197∗∗ −0.550∗∗ −0.031†

12

25

0.061∗ 0.002 202.5 24 166

0.558∗ 0.803∗

−1.450∗∗ −1.458∗∗ −0.795∗∗ — 1.404∗∗

−0.229 −0.888∗∗ — 0.101

0.552∗ 0.787∗

−6.511∗∗ −0.606∗∗ −0.031∗

p < 0.01 ∗ p < 0.05 † p < 0.10 Note: Job shifts refer to job changes across organizations after one’s ﬁrst job. Middle school or below is the reference category for education, 1949–1965 cohort for cohort, production worker for occupation, and collective ﬁrm for work organization. All model estimations also included a set of dummy variables indicating respondents’ province and city size.

∗∗

Intercept Female Age education High School College occupation Cadre Professional Military service Service worker job origin Government Public org’n State-owned ﬁrm Hybrid/private ﬁrm Farm cohort 1966–1979 1980–1994 % industrial output Collective Hybrid/private ﬁrm χ2 d.f. N of events

Hybrid Firm

1949–1965 1966–1979 1980–1994 1949–1965 1966–1979 1980–1994 1949–1965 1966–1979 1980–1994

Collective Firm

258

The State and Life Chances in Urban China

job origins are incorporated in the analyses as independent variables.3 The parameter estimates indicate the effects of the corresponding variables on the log-rate of leaving one’s current job to the new job destination indicated in that column. Let us begin with the overall trends over time. The intercepts associated with different destinations across periods provided a general picture of historical trends in job shifts to various types of organizations. Estimated rates of job shifts to private and hybrid ﬁrms were included only for the 1980–1994 period because these rates were negligible before 1980. The negative coefﬁcients associated with the intercepts indicate that rates of job shifts were very low in urban China. The hazard rates of jobshifts were the lowest in the Cultural Revolution period, reﬂecting the shrinking opportunities in that period. In contrast, the rates were highest (as reﬂected in the smaller negative intercept terms) in the reform eras, suggesting the rise of labor markets and that China’s emerging market economy has indeed induced structural changes over time. We next consider the role of human capital in job-shift decisions. In the pre-reform and reform eras, the rates of job shifts to government agencies and public organizations were, in most instances, signiﬁcantly higher for those with a high school or college education than for those with less education. Moreover, college education was associated with a higher rate of job shifts to government agencies, even in the earliest period, indicating the prominent role of education in the state building in China. But the effects of a college education on moving to these two types of organizations were somewhat larger in the reform era than in earlier periods. These patterns suggest that government agencies and public organizations continued to be the main destinations for the job shifts for college-educated individuals over time. In the reform era, college graduates did have signiﬁcantly higher rates of shifts to both high-status organizations in redistributive system (e.g., governmental agencies and public organizations) and most market-oriented organizations (e.g., hybrid ﬁrms and private ﬁrms). That is, human capital was playing a key role in channeling shifts to jobs offering greater incentives, including those market-oriented ﬁrms. But the effects of education were smaller in magnitude for shifts to hybrid ﬁrms and private ﬁrms than 3

In the analyses, we combine private and hybrid ﬁrms as one indicator of job origin due to small size of observations in these two types of organizations in the pre-reform era in the data.

Institutional Changes and Patterns of Job Shifts

259

for shifts to government agencies and public organizations. These patterns lead us to infer that state organizations offered incentives to those with high human capital that were at least as attractive as those offered by hybrid ﬁrms and private ﬁrms. Overall, there is mixed evidence in the reform era: Employees with the most human capital were more likely to move to jobs in the types of organizations closest to the market (private/hybrid) and to jobs in the types most dependent on the redistributive economy (government agencies and public organizations). In contrast, those organizations in between such as state ﬁrms and collective ﬁrms became less desirable. On the one hand, these ﬁrms were pushed to engage in market competition and with declining redistributive beneﬁts from the government. On the other hand, their burden in tax and retirement support and governmental interventions often put them in disadvantageous positions in market competition for human resources. Occupations are important indicators of redistributive beneﬁts and political power. For example, cadres receive the greatest beneﬁts and have the most power under state socialism. Therefore, we expect that job-shift rates vary with occupational origin depending on their relationship to the redistributive institutions. The estimated effects of occupational origin show that occupational origins did not have systematic effects on job shifts across types of organizations and their effects were not appreciably different in the reform era than in earlier periods. To address our main question about sources of job-shift patterns, we need to consider not only job-shift destinations but also workplace origins. Two types of job shifts are especially important for understanding the extent of institutional change: (1) those away from state organizations (government, public organizations, and state-owned ﬁrms) to semistate (collective) and nonstate organizations (private/hybrid ﬁrms) and (2) those away from government agencies and public organizations to state-owned ﬁrms (where market activities are increasingly important). Both would indicate moves away from the core redistributive institutions and toward market-oriented institutions. The most striking effects of a person’s organizational origin occurred for the variable “farm.” The statistically signiﬁcant effect of farm origin in the 1949–1965 period reﬂected mainly high levels of rural–urban migration during this period of rapid economic development. In the other two periods, the effects of “village” probably reﬂected the return to urban areas of the youth previously sent down to work in rural areas. All types of urban organizations expanded after state policies permitted sent-down

260

The State and Life Chances in Urban China

youth to return to their original area in 1979.4 The strong effect of farm on shifts to private and hybrid ﬁrms, as well as to collective and state-owned ﬁrms, suggests that the expansion of these types of nonstate organizations was partly a response to saturated employment in state organizations in the early 1980s. In contrast, except for the effect of farm, there was no systematic evidence that shifts to jobs in government agencies, public organizations, collective ﬁrms, or hybrid ﬁrms depended on organizational origin in the reform era.5 However, the job-shift rates to private ﬁrms during the reform era were signiﬁcantly higher for job origins in collective ﬁrms (the reference category) than for origins in state-owned organizations. The rates of shifts to state-owned ﬁrms were signiﬁcantly higher from government agencies than from collective ﬁrms (the reference category) in the last two periods, and also were signiﬁcant for moves from public organizations in the 1966–1979 period.6 This may indicate fewer institutional barriers for movements within the state-owned organizations. But, notice that the rates of shifts from government agencies and public organizations to state-owned ﬁrms were smaller in magnitude in the reform era than in 1966–1979. In other words, workers employed in the organizational types closest to the redistributive economy (government agencies and public organizations) did not tend to move to state-owned ﬁrms in the reform era than before. This may result from the fact that, except for the early years of the reform, the privileged welfare and income beneﬁts associated with state ﬁrms were seriously weakened in the reform era. Overall, we do not ﬁnd clear evidence that there were signiﬁcant shifts from state organizations to nonstate ﬁrms during the years of our study. In the reform era, shifts to the market-oriented private organizations were signiﬁcantly more likely for those working in collective ﬁrms, which beneﬁted least from the state redistributive economy. These patterns underscore our arguments that job-shift patterns reﬂect a workplace’s net

4

5

6

As shown in Chapter 5, sent-down youth returned to cities individually throughout the Cultural Revolution, but the largest inﬂux was in 1979 when the send-down policy was formally abandoned. We suspect that the high rate of job shifts from collective ﬁrm to government agencies and public organizations in the 1949–1965 period reﬂect the implementation of nationalization policies at that time. Because Table 9.1 pertains to external job shifts – job shifts from other types of organizations – “collective ﬁrm” cannot be used as the reference category for organizational origin when shifts to a collective ﬁrm are the destination. Hence, in this particular instance, we use “state ﬁrm” as the reference category for organizational origin.

Institutional Changes and Patterns of Job Shifts

261

attractiveness and that state organizations still offered considerable advantages by 1994. Before we leave this set of analyses, we brieﬂy assess the role of life course factors in job shift decisions. We use the historical era when individuals ﬁrst entered the labor force to assess how historical context shaped their subsequent job shifts. Age gauges both biological age and seniority (i.e., labor force experience), with which age correlates highly. Not surprisingly, as has been found in other countries, the rates of job shifts tend to fall as age increases, reﬂecting the cumulative advantages of keeping the current job and the rising costs of leaving it. As for cohort effects, we ﬁnd that the younger generation that had less cost in job shifts tended to shift their jobs to hybrid ﬁrms and private ﬁrms and, to a less extent, to collective ﬁrms. In summary, these results suggest that, by 1994 – the end of the observation span in this study, changes in institutions and in stratiﬁcation mechanisms had only a modest impact on job shifts in urban China since the reform began in the 1980s. A college education signiﬁcantly increased job-shift rates to government agencies before as well as during the reform era. Effects of workplace origins suggest that state organizations still offered strong incentives that operate to retain their employees. Employment in private and hybrid ﬁrms had grown, but this growth seems to be due to job shifts from workplaces with the least redistributive beneﬁts (collective ﬁrms and villages) more than departures from state organizations. Overall, we witnessed two trends: One the one hand, state organizations, especially government agencies and public organizations, still enjoyed advantages inherited from the past; on the other hand, nonstate ﬁrms (private and hybrid ﬁrms) signiﬁcantly increased their status along with emerging market economies. Those organizations in between such as state ﬁrms and collective ﬁrms experienced both the declining assistance from the government and increasing competitive pressures in the marketplace; they became the biggest losers in this transformation process. Internal Job Shifts: Moves Within Types of Organizations In addition to job shifts across types of organizations, there are also events of job shifts between organizations of the same ownership type. For example, a cadre may move from one government agency to another, or an employee may move from one state ﬁrm to another. We call these events “internal job shifts,” because they do not involve a change in the

262

The State and Life Chances in Urban China

workplace’s property-rights relationship with the state. Nevertheless, such internal job shifts are informative, especially for the reform era, because rewards (e.g., bonuses, opportunities) varied considerably within as well as across organizational types. Table 9.2 reports the maximum likelihood estimates for models of internal job shifts. We excluded private and hybrid ﬁrms from this analysis because internal shifts within these types of organizations occurred too rarely in our data to be analyzed. As the intercepts indicates, the overall job-shift rates were much lower within government agencies and public organizations than within stateowned or collective ﬁrms. In the reform era, there were relatively higher job-shift rates within government and pubic organizations (as indicated by the relatively smaller negative coefﬁcients) in the reform era than before, but the rate of job shifts within state ﬁrms or within collective ﬁrms was lower in the reform era than before. It appears that different types of organizations experienced different internal dynamics in the reform era. Human capital again played a signiﬁcant but uneven role for internal job shifts in different types of organizations. Except for the Cultural Revolution period, the job-shift rates within government agencies and public organizations were signiﬁcantly higher for college graduates than for those with a middle school or less education. During the two most recent periods, rates of shifts within state-owned ﬁrms were signiﬁcantly higher for high school graduates but not for college graduates in either period. With respect to occupational origin, we observed a similar pattern as before. Overall, there was no systematic evidence of job shifts among occupational groups within types of work organizations. Nor was there evidence that the effects of occupation origin differed substantially between the reform era and earlier periods. In other words, decisions about job shifts were not signiﬁcantly affected by their location in current jobs (occupation). We observed different effects of life-course factors. Age did not significantly affect the rate of internal job shifts except for a small effect on shifts within state-owned ﬁrms. That is, age had somewhat smaller effects on internal job shifts than on external job shifts (cf. Table 9.1). This ﬁnding is not difﬁcult to understand. State policies explicitly recognize one’s work experience (seniority) within state organizations and it can be transferred to a new workplace within the state sector. As a result, one’s age had fewer adverse effects on job-shift decisions within the same type of organizations.

263

Intercept Female Age education High School College occupation Cadre Professional Service worker cohort 1966–1979 1980–1994 % industrial output Collective Hybrid ﬁrm χ2 d.f. N of events −0.153 −0.346 0.475 0.719† 0.505 −0.030 −0.086∗∗ 66.8 20 109

0.127 0.280 0.101 −0.043 — 0.082∗ 0.091 30.9 19 94

— —

0.007 0.022∗∗ 29.4 18 73

−0.187 0.574∗

−0.056 0.267

0.360 0.875∗∗

−0.295 −0.098 1.043

−3.140∗∗ −0.775∗∗ −0.002

−5.126∗∗ 0.110 −0.006

−4.118∗∗ −0.232 0.006

1980–1994

1966–1979

1949–1965

Government

0.056 0.028 32.0 18 37

— —

−0.820 0.584 1.773∗

0.465 0.976†

−6.181∗∗ −0.303 0.003

1949–1965

0.033 0.113 35.5 19 76

0.395 —

(continued)

0.029 −0.045† 30.2 20 71

0.155 0.035

0.545 0.203 −0.193

0.281 0.107

1.081∗∗ 0.836∗ 0.163 −0.011 −0.618

−4.623∗∗ 0.390 −0.035

1980–1994

−5.383∗∗ 0.133 0.002

1966–1979

Public Organization

table 9.2. Parameter Estimates of Logistic Model for Job Shifts Within Type of Organization, by Period

264 0.235 —

— —

0.025 −0.002 64.2 18 190

0.020 0.127 56.2 19 245

−0.575† −0.188 −0.230

−0.506 0.183 −0.483

0.008 −0.044∗∗ 105.7 20 395

0.571∗∗ 0.434

−0.005 −0.004 0.230

0.288∗ 0.487∗

0.421∗∗ 0.612∗

0.202 0.588†

−4.137 0.090 −0.022†

−3.992 −0.073 −0.019†

−3.079 0.021 −0.036∗

∗∗

1980–1994

−0.197 — 0.022 0.346∗ 33.0 19 151

−0.015 −0.004 36.6 18 85

−0.446 −0.097 0.040

0.476∗ 0.486

−3.359 −0.186 −0.028∗

∗∗

1966–1979

— —

−0.228 0.567 −0.040

−0.340 −0.052

−3.354 −0.516∗ −0.011

∗∗

1949–1965

Collective Firm

0.003 −0.035∗ 49.8 20 197

0.134 0.524

0.296 −0.098 −0.090

0.262 0.445

−3.689∗∗ −0.098 −0.020

1980–1994

∗

p < 0.01 p < 0.05 † p < 0.10 Note: Job shifts refer to job changes across organizations after one’s ﬁrst job. Middle school or below is the reference category for education, the 1949–1965 cohort for cohort, production worker for occupation, and collective ﬁrm for work organization. All model estimations also included a set of dummy variables indicating respondents’ province and city size.

∗∗

Intercept Female Age education High School College occupation Cadre Professional Service worker cohort 1966–1979 1980–1994 % industrial output Collective Hybrid ﬁrm χ2 d.f. N of events

1966–1979 ∗∗

∗∗

1949–1965

State Firm

table 9.2 (continued)

Institutional Changes and Patterns of Job Shifts

265

Overall, employees with a high school or a college education tended to have higher rates of internal job shifts than those with little education in most historical periods. These ﬁndings suggest that these groups can improve their work situations through internal job shifts. It is noteworthy that these patterns have not changed substantially in the reform era. Therefore, we ﬁnd no evidence that stratiﬁcation mechanisms underlying job-shift patterns changed substantively in the period of our study. External Job Shifts: Moves Across Economic Sectors In the reform era, the pace and extent of economic reform and marketization have varied greatly across China’s economic sectors. Hence, job shifts across economic sectors provide additional evidence on the relative incentives offered by redistributive and market institutions across these sectors. Table 9.3 reports maximum likelihood estimates for the model of job shifts across economic sectors. The results parallel the results in Table 9.1 for job shifts across types of organizations, with each column indicating speciﬁc type of organizational destination. Our analyses included the government and public sectors, and, as noted before, these two sectors overlap with, but are not the same as, the similarly named types of organizations. Economic sector is deﬁned by the product or service provided by a workplace, whereas type of organization is deﬁned by the workplace’s property-rights relationship to the state. The results for government and public organizations are similar to those reported in Table 9.1 and therefore were omitted from Table 9.3. We ﬁrst consider the role of human capital. Paralleling the results for job shifts across types of organizations (see Table 9.1), the rates of shifts to the government and public sectors (not reported in Table 9.3) were signiﬁcantly higher in the reform era for employees with a high school or college education than for those with less education. We also observed a similar pattern for heavy industry in the reform era. Moreover, the effects of college education on job shifts to these three sectors (the ones most closely tied to the redistributive system) were generally largest in the reform era. In contrast, education had little effect on the job-shift rates to the light manufacturing and service industries (the sectors where market incentives were largest in the reform period). That is, those sectors where redistributive economy still dominated were able to attract people with better human capital.

266

Intercept Female Age education High School College occupation Cadre Professional Military service Service worker Farmer

Light Industry

Service Industry

−5.464∗∗ −0.196† −0.031∗∗

0.236† 0.159

−0.604∗ −0.720∗∗ 1.078∗∗ 0.020 1.917∗∗

−5.575∗∗ −0.818∗∗ 0.002

−0.533† −0.380

−1.021∗ 0.103 0.053 −0.429 1.121∗∗

0.476 −0.014 0.532 0.373 1.117∗∗

−0.228 −0.870

0.353∗ 0.581∗ −0.586∗ −0.768∗∗ 1.183∗∗ −0.280 2.349∗∗

−5.113∗∗ 0.229 −0.032

−3.289∗∗ −0.298∗ −0.055∗∗

−0.816∗ −0.499 0.341 −0.639 2.121∗∗

0.326∗ −0.449

−8.129∗∗ 0.223 −0.003

−0.182 −0.428 0.753 −0.284 2.145∗∗

−0.041 −0.082

−4.724∗∗ −0.168 −0.062∗∗

−0.812† −0.956∗ 0.244 −1.569∗ 0.239

0.269 0.631

−3.851∗∗ 0.114 −0.050∗

−0.779∗∗ −0.611∗ −0.739∗ −0.611† 1.557∗∗

0.110 −0.234

−6.323∗∗ −0.015 −0.020†

−0.065 −0.370† 0.922∗∗ −0.324† 2.395∗∗

−0.020 −0.070

−5.132∗∗ −0.019 −0.036∗∗

1949–1965 1966–1979 1980–1994 1949–1965 1966–1979 1980–1994 1949–1965 1966–1979 1980–1994

Heavy Industry

table 9.3. Parameter Estimates of Logistic Model for Job Shifts Outside Economic Sectors, by Period

267

0.042∗ 0.127 532.8.1 24 387 −0.020 −0.032∗ 318.6 24 220 0.010 0.001 34.5 23 69

— —

−0.388† −0.180

0.300† —

— —

0.043∗ 0.010∗ 104.8 23 148

0.033

0.179

0.180

0.159 −0.209

0.077

0.895∗∗ 0.262

0.854∗∗ 0.376

0.847∗ 0.147

0.045† −0.072∗∗ 198.7 25 152

0.023 0.147

0.570∗∗ — 0.121∗∗ 0.050 337.8 24 214

−0.362

0.236 −0.706†

0.196

1.326∗∗ −0.007

−0.001 −0.001 37.6 23 112

— —

0.335 0.146 −0.311

0.098∗∗ 0.356∗∗ 362.2 24 307

0.266 —

1.503 −0.540 −0.545

0.027† −0.015 370.4 25 374

0.665∗∗ 0.943∗∗

0.672∗∗ −0.184 0.438∗∗

∗

p < 0.01 p < 0.05 † p < 0.10 Note: Job shifts refer to job changes across organizations after one’s ﬁrst job. Middle school or below is the reference category for education, 1949–1965 cohort for cohort, production worker for occupation, and collective ﬁrm for work organization. All model estimations also included a set of dummy variables indicating respondents’ province and city size.

∗∗

job origin (sector) Gov’t sector Public sector Light industry Service industry cohort 1966–1979 1980–1994 % industrial output Collective Hybrid ﬁrm χ2 d.f. N of events

268

The State and Life Chances in Urban China

As for the role of occupational status, both cadres and professionals tended to have lower rates of external job shifts, compared with production workers (the reference group). That is, cadres and professionals seemed to have strong incentives to remain in their current economic sector, despite growing variation among sectors in job rewards. There were no systematic or noteworthy differences between service workers and production workers within or across the three periods. Next, we consider the sector of origin for job-shift decisions. Compared with the reference category (heavy industry), those in the government sector in the reform era had signiﬁcantly higher rates of shifts to two of the three goods-producing sectors (heavy industry and the service industry, but not light industry). But note that this pattern started in the ﬁrst or second period, before economic reform began. The relative size of the effects in the two most recent periods did not suggest that signiﬁcant changes occurred in the reform era. Farmers (mostly the sent-down youth) from villages had consistently higher rates of entering all economic sectors in all three periods, resulting from urbanization and the return of sent-down youth (in the last two periods). There were no other systematic effects of sector of origin on rates of external job shifts, except that workers in light industry (where redistributive beneﬁts are low) had a signiﬁcantly higher rate of job shifts to the service sector in the reform era. As for those life-course factors, age had a consistently negative effect on job-shift rates across sectors in the reform era. The underlying message is obvious: job moves across economic sectors were more costly, less available, and/or less beneﬁcial for older or senior individuals. The cohort effects were mixed. Those who entered the labor force in the 1966–1979 period were more likely to move to the public sector during this period, but they exhibited a lower rate of entering heavy industry and a higher rate of entering the service industry in the reform era. These observed patterns were due to a large extent the effects of state policies on job opportunities across these sectors, especially for the sent-down youth (who entered the labor force in 1966–1979). Since the late 1970s, the government opened up the service sector to deal with unemployment problems in the urban area. Those entering the labor force in 1980–1994 also had signiﬁcantly higher rates of shifts to the service sector, again reﬂecting the saturation of employment in state organizations that forced new labor force entrants to work in the service sector where collective and nonstate ﬁrms were prevalent.

Institutional Changes and Patterns of Job Shifts

269

chapter summary This chapter has focused on two interrelated issues: First, we examined patterns of job attainment, as measured by the destination of subsequent job shifts. Job-shift patterns shed light on mechanisms and processes of social stratiﬁcation regarding the allocation of individuals to locations in the social stratiﬁcation structure. Second, the focus on individuals’ access to workplaces as revealed in job-shift patterns across types of organizations and economic sectors also sheds light on the direction and extent of institutional changes in China over time. As we argued at the beginning of this chapter, the extent and nature of change in China’s economic institutions are better measured by such job-shift patterns than by manifest differences in economic beneﬁts, such as income. The hazard rates of job shifts in three periods (1949–1965, 1966–1979, and 1980–1994) signaled the relative rise and decline of the type of organizations that are associated with competing mechanisms for allocating human resources, thus providing clues to the extent of change in China’s economic institutions and in its stratiﬁcation mechanisms. Regarding our previous discussion of stratiﬁcation dynamics, job-shift patterns reported in this chapter revealed marked variations over time, indicating that there have been considerable changes in opportunities in the labor force over time as a consequence of shifts in state policies. The effects of education and positions (occupational locations) also varied across historical periods, in response to changes in state policies. There is also evidence of barriers in job movements across types of work organizations and sectors. These ﬁndings are consistent with our previous arguments about the stratiﬁcation processes in China’s redistributive economy. Our analyses of job-shift patterns suggest that, by 1994, the changes in China’s economic institutions were still limited. On the one hand, patterns of shifts among types of state-owned organizations were similar prior to and during the reform era. Rates of leaving the economic sectors closest to the redistributive economy (government agencies, public organizations, and state-owned ﬁrms) were no greater in the reform era than in the past. These patterns suggest that governmental agencies and public organizations largely kept their advantages and were still desirable in the reform era. The growth of a market economy (as indicated by the growth of private ﬁrms, hybrid ﬁrms, and the service sector) did not seem to result from people leaving the state sector. Rather, shifts to private ﬁrms mainly originated in semistate and nonstate organizations.

270

The State and Life Chances in Urban China

Both continuity and change were evident in stratiﬁcation mechanisms since the 1980s. Human capital played a somewhat larger role on jobshift decisions to government agencies and public organizations, as well as on shifts to hybrid ﬁrms. But the effects of education on shifts to government agencies and public organizations were not much less in earlier periods. Those with greater human capital (e.g., a college degree) and in occupations with more political power (i.e., cadres) tended to stay in the economic sectors closest to the redistributive economy (i.e., government sector, public sector, and heavy industry) and not to move to sectors closer to the market economy. The most salient change in the reform era was the rise in employment in private and hybrid ﬁrms, which grew hand in hand with the emergence of market economies. Over time, private ﬁrms and hybrid ﬁrms have increased their status in the national economy. More educated employees and the younger generation tended to shift their jobs to these most market-oriented ﬁrms. However, it is worth noting that this occurred mainly through shifts from workplaces lower in the hierarchy of types of organizations (hybrid ﬁrms and collective ﬁrms). Employees in state organizations were not signiﬁcantly more likely to move to nonstate organizations, indicating continuing incentives and constraints provided by the redistributive economy. Nor have moves from government agencies and public organizations to state-owned ﬁrms become signiﬁcantly more likely in the reform era. The historical approach we have used allows us to assess the extent of changes by comparing job shift patterns over time and helps to avoid some potential misinterpretations. For instance, the signiﬁcant effect of working in a government agency on the rate of job shifts to state-owned ﬁrms in the reform era (Table 9.1) could be construed as reﬂecting the market incentives offered by state-owned ﬁrms. However, because the same pattern occurs before the reform era, doubt is cast on this interpretation. The signiﬁcant and lower rate of cadres’ job shifts to heavy industry in the reform era could also be misinterpreted as reﬂecting changing stratiﬁcation mechanisms, if one were unaware that this rate has been consistently low over time (see Table 9.3). The historical evidence offered by our study is more systematic and robust than that of studies focusing on the reform era alone. The observed job-shift patterns reﬂect an institutional transformation that has been administered by the Chinese state and increasingly by local governments. Rona-Tas (1994) found that “redistributors” in Hungary in the 1990s enjoyed continuing beneﬁts by moving to more proﬁtable and

Institutional Changes and Patterns of Job Shifts

271

more market-oriented workplaces and by establishing their own companies. In contrast, those who formerly beneﬁted most from China’s redistributive economy (cadres and employees in government agencies) have tended to maintain jobs in the types of organizations and economic sectors closest to the redistributive economy. Although average income is lower in state than in nonstate organizations (see Table 7.1), the considerable redistributive beneﬁts (e.g., housing) offered by state organizations, as we saw in Chapter 8, help explain why these employees are not prone to leave jobs in public organizations during the reform era. The continuing redistributive beneﬁts and the state’s ongoing interventions in labor force practices create incentives for those in state organizations to stay and also structural barriers for those outside to enter. In addition, state policies discriminating against private entrepreneurs create uncertainties about jobs in nonstate ﬁrms, thereby discouraging employees from leaving state organizations. Job-shift patterns help us understand institutional changes in China, but the two are by no means equivalent. There have been fundamental changes in central-local government relationships, worker–management relationships, organizational boundaries, and market activities, even within the state sector. For instance, certain public organizations (e.g., hospitals, research institutions) have become actively engaged in market activities, which help to provide incentives for their employees to remain in their current workplace. Thus, the attractiveness of jobs in some state organizations probably reﬂects both continuing redistributive beneﬁts and beneﬁts from newly developed market activities of such workplaces. Finally, we should note that, since the mid-1990s, there were major developments in the transformation of the state sector and in the privatization of small and midsized state ﬁrms, forcing a large number of employees in state organizations to move to the nonstate sector. These more recent changes and the underlying mechanisms are beyond the scope of our study and remain to be further explored. As we noted in the analyses in this chapter, life-course factors play an important role in generating variations in job-shift patterns. Indeed, there are signiﬁcant cohort-based differences in individual life chances in the era of large-scale social changes. We explore this topic in the next chapter.

10 Economic Transformation and Life Chances: A Life-Course Perspective

In the previous chapters, we have followed a sample of the urban residents and their life experiences over stages of their life course. They grew and aged and, as we have shown, their life chances have changed over time, along with drastic political and economic events in the history of the People’s Republic of China. Implicitly, and at times explicitly as in Chapter 4, we have adopted a life-course perspective to trace changes in life chances not only over historical time but also over one’s life course. Now it is time to explicitly address this question: Do stages in the life course matter in one’s life chances, especially in the era of large-scale social changes and economic transformation? This is the central issue for this chapter. Large-scale societal transformations inevitably have fundamental impacts on individuals’ life chances. These are often caused by drastic changes in the processes and mechanisms of resource allocation and social stratiﬁcation, including the reallocation of opportunities and risks among social groups. What is less apparent, however, is that these impacts are often shaped by important contextual considerations that are not an integral part of sociologists’ traditional theoretical explanations. In this chapter, we propose and demonstrate that life-course timing is such a contextual construct, and that the life-course location of particular cohorts at the time of major socioeconomic upheavals plays a critical role in mediating the impacts of opportunities and risks. “Life course” refers to the biographical patterning of events, resources, and roles over the life span (e.g., Giele and Elder 1998; Kohli 1986; Mayer and Tuma 1990; Settersten and Mayer 1997). Thus a life-course framing provides an important lens through which we can 272

Economic Transformation and Life Chances

273

understand and explain the links between institutional change and individual life chances. As we reviewed in the previous chapters, many studies have documented the impacts of economic transformations on China’s system of social stratiﬁcation. These studies have primarily focused on the key mechanisms of social stratiﬁcation identiﬁed in the literature, such as human capital, political capital, and occupational status, to assess the impacts of institutional changes. Little attention has been given to the contingent processes that shape the links between the macro-institutional changes and individual life chances, including an understanding of the timing of social upheavals in individuals’ biographies. For example, in the last chapter, we analyzed job-shift patterns across distinct historical periods to assess institutional changes and found noticeable changes as well as continuity in China’s reform era. But our analytical model there did not take into consideration factors that are related to an individual’s stage in the life-course. Indeed, to simplify the analysis in the last chapter, we assumed that the effects of the covariates in any period would be uniform for all individuals in the sample, regardless of their life-course stage. Most recent studies of the impacts of institutional changes in transitional societies published in major sociological journals did not incorporate life-course location. Research designs in these studies focus on intergroup or interindividual differences and do not address a basic and obvious question: In assessing the relationship between changing life chances and large-scale social changes, do individuals’ stages in the life course matter? If the answer is yes, then we need to develop theoretical arguments and adopt appropriate research designs to explicitly address these issues. In this chapter we develop a life-course argument for assessing the impacts on life chances of the large-scale social changes in China. Our premise is that drastic institutional changes do not affect people’s lives in the same way in terms of their career pathways and economic well-being; rather, life-course factors shape the possibilities and payoffs for workers in various cohorts. The economic transformations in China present differential opportunities and risks for individuals in different cohorts precisely because they experience these transformations at different stages of their life course. Thus, the impacts of the economic reform play out through individual life biographies, generating distinctive patterns of life chances for different cohorts of individuals. In the following sections, we ﬁrst develop theoretical arguments as to how life-course factors affect life chances

274

The State and Life Chances in Urban China

in the form of career pathways and access to resources among different cohorts of labor force participants in urban China. Second, we empirically examine the role of life-course factors in job-shift patterns and the distribution of economic beneﬁts in the reform era, from 1980 to 1994.

institutional changes and life-course implications Our starting point is the recognition that life-course factors play a critical role in moderating the impacts of social changes on individual life chances; that is, dramatic institutional changes have differential impacts on individuals at different life stages (e.g., Elder 1974, 1985a, 1995). In contemporary societies, the life course itself has been increasingly institutionalized by the state (Mayer and Muller 1986; Mayer and Schoepﬂin 1989). Major social and economic upheavals, often induced by shifting state policies, restructure institutional resources, constraints, and options associated with different life stages. Thus, the “timing” of historical events in terms of life stage (or age) is critical in producing cohort differences in both exposure and response to social change, given cohort members’ distinctive biographies as well as their embeddedness in the current social structure. Cohort (based on age at birth or age at the time of a major role shift such as ending school, marrying, or entering the workforce) captures two related temporal dimensions: historical experience and biographical life stage (in addition to biological age). Looking at particular transitions (such as job shifts) requires consideration of both institutional options and constraints and workers’ strategies of action (Kohli 1986; Mayer and ¨ Muller 1986; Moen 1985; Moen, Elder, and Luscher 1995). An underlying theme in life-course analysis is the notion of cumulative advantage (or disadvantage), with individuals privileged in their earlier life stage experiencing subsequent beneﬁts from their social location (e.g., Elder 1995; Merton 1968). As Elder (1974) showed in his classic study of the Great Depression, impacts of dramatic social upheavals may result in both cross-cohort and within-cohort variations. We believe that both variations are present in the Chinese case. People in different cohorts occupy different positions in career development and in the stratiﬁcation structure, and, consequently, differ in their vulnerability to social dislocations accompanied by large-scale social changes. Hence, individuals in different cohorts experience unique impacts of historical transformations at least in part because of their advantaged (or disadvantaged) position in the social structure. Thus, life stage serves as a marker of both (1) distinctive age-graded risks and options and (2) distinctive, cohort-related social

Economic Transformation and Life Chances

275

locations in the opportunity structure. Gold (1991) and Davis (1992b) developed explicit arguments in the Chinese context. These ideas are especially useful for our understanding of variations in life chances in light of China’s economic transformation. In times of social upheaval, two broad historical and organizational processes affect individual life chances. The ﬁrst is the changing opportunity structure. Along with large-scale socioeconomic transformations, old institutional arrangements are undermined and new institutional arrangements emerge. These changes alter existing opportunity structures for individuals in different locations of the social structure. The second process reﬂects the impacts of societal transformations on changing career pathways and mechanisms of attainment. In the Chinese context, the expansion of markets provides new opportunities outside the redistributive economy and alters the mechanisms of social stratiﬁcation (Nee 1989, 1991, 1996). However, these changes appear to affect different social groups in different ways, as shown in previous studies (Bian and Logan 1996; Wu and Xie 2003; Xie and Hannum 1996; Zhou 2000a; Zhou, Tuma, and Moen 1997). That is, those who hold positional power in the state bureaucracy beneﬁt from both their political authority and market opportunities, reﬂecting the persistence of cumulative advantage in the existing system of stratiﬁcation. How can we explain these patterns and anomalies? We believe that a life-course perspective can provide insight into these issues. To illustrate our main arguments, we consider the interrelationships among institutional changes, individual life chances, and the moderating role of the life course in China’s reform processes. First, to highlight the critical role of the life course, we need to make a distinction between the emergence of new opportunities and access to these opportunities. Social changes are those processes that alter old opportunity structures and create new opportunities. However, it is important to recognize that access to these new opportunities is not equally distributed among members of a society. We argue that access to new opportunities is moderated in important ways by one’s location in the life course. In this respect, it is useful to note that social changes produce both opportunities and risks for the individuals experiencing them. Consider the emergence of market opportunities in China’s economic transformation. Job opportunities in foreign ﬁrms and other nonstate ﬁrms (e.g., private ﬁrms) have indeed substantially altered previous opportunity structures under state socialism. But, these new types of jobs present both opportunities and risks. Foreign ﬁrms and private ﬁrms, for instance, tend

276

The State and Life Chances in Urban China

to offer better monetary payoffs based on their competitive advantages in the marketplace relative to those ﬁrms in the state sector. But without government’s subsidies, they are also more exposed to the higher risks associated with market ﬂuctuations and business cycles. More important, persistent state interventions into the marketplace mean that ﬂuctuations in state policies make those relying on market opportunities especially vulnerable. In this light, we expect to ﬁnd noticeable cohort-based differences in response to these opportunities. Behavioral psychologists have shown that individual judgments under uncertainty reﬂect heuristics and bias rather than rational calculation (Kahneman, Slovic, and Tversky 1982). Given the uncertainty of market opportunities and of policy stability, individuals’ decisions related to career progression depend heavily on their interpretations of their past experiences. Thus, one’s own biography provides the vantage point from which to assess the potential risks and beneﬁts attached to new options. Those who have experienced extreme ﬂuctuations in the past are likely to weigh heavily the potential costs of (future) adverse state policies, and adopt a risk-averse attitude. This consideration is especially relevant given the ﬂuctuation and turmoil in China’s political history and their impacts on life chances, as we demonstrated in previous chapters. With the cumulative beneﬁts in the state-owned organizations, older individuals, for example, may be especially risk averse in response to new types of jobs outside of the state sector and see them more as risks than opportunities (Li 1999). But it is also possible that ﬂuctuating life experiences cultivated more risk-taking attitudes, as we witnessed in the sent-down generation; see Chapter 5. Second, the effects of allocative mechanisms are also closely associated with, and moderated by, life-course factors such as cohort effects. This is especially true in a rapidly changing society like China. Because of evolving opportunity structures over time, different cohorts of individuals have been entrenched into different types of work organizations in the organizational hierarchies in China. Most of those who entered the labor force before 1966 worked in state and collective sectors, as a result of a state push toward nationalization and collectivization in the 1950s. In contrast, a large proportion of “the Children of the Cultural Revolution” – those who entered the work force between 1966 and 1979 – were forced to work in rural areas (see Chapter 5). Even when they returned to urban areas in the 1980s, the urban labor market structure had drastically changed; as a result, a large number of this cohort began their careers in the nonstate sector. Finally, those who entered the work force in the reform era (since 1980) faced a very different range of career alternatives, as a consequence

Economic Transformation and Life Chances

277

of a contracting state sector and an expanding nonstate sector (Li 1999). Thus, different social institutions offer competing incentives for different cohorts in the processes of China’s economic transformation. For instance, in the hierarchy of work organizations in China, state-owned ﬁrms, until the mid-1990s, offered better job security and extensive welfare programs compared to collective and private ﬁrms (Bian 1994; Walder 1986, 1992; Zhou, Tuma, and Moen 1997). That is, the earliest cohort who launched their career paths prior to 1966 and typically were concentrated in state ﬁrms have more accumulative beneﬁts from the redistributive institutions. At the same time, they were especially vulnerable in the private-sector marketplace, given their lack of education and training compared with younger cohorts. As a result, changes in the allocative mechanisms between redistribution and markets have had distinctive impacts on individuals in different locations in the social structure, which in turn is closely associated with one’s stage in the life course at the time of these changes.

empirical implications We now consider the empirical implications of these ideas for understanding the role of life-course timing on the emerging patterns of social stratiﬁcation in urban China in the reform era of the 1980s and 1990s. Our discussion focuses on two sets of empirical patterns. First, to capture distinctive career pathways, we focus on patterns of job shifts to assess cohort variations in response to changes in the opportunity structure and to shed light on life-course-based redistribution of opportunities and risks. Speciﬁcally, we examine the determinants of job shifts across different types of work organizations in the era of economic transformation (1980–1994). In our analyses of job shifts in Chapter 9, cohort effects are conspicuous and deserve further exploration here. Second, we examine cohort variations in economic beneﬁts (e.g., personal income). Our purpose is to examine cohort-based differences in access to different sources of economic beneﬁts and how competing allocative mechanisms affect cohorts in different ways in the time of major social changes. Job-Shift Patterns: Cohort Difference in Response to Opportunities Recent economic transformations are having profound impacts on the lives of urban residents in China. With the rise of the private sector, including the presence of joint ventures and foreign ﬁrms, job opportunities

278

The State and Life Chances in Urban China

have greatly diversiﬁed beyond the traditional state sector. Consistent with these changes in opportunity structures there has been corresponding shifts in the distribution of resources such as income. As we showed in Chapter 7, job opportunities in emerging sectors tend to offer better economic payoffs. However, these transformations have also produced greater risks in terms of job and economic security; welfare and other beneﬁts associated with prior social and economic institutions can no longer be taken for granted. Although a job in the private sector may offer higher wages, the lack of job security, fringe beneﬁts, housing, health care, and other welfare beneﬁts may offset its attractiveness to an employee holding a secure job in the state sector. The coexistence of competing institutions and incentives has different implications for different cohorts of workers. Older cohorts tend to be well placed in existing redistributive institutions, whereas younger cohorts enjoy fewer beneﬁts even when they are located in the state sector, because of their brief time in the labor force and the recent reform of the welfare system. By contrast, market-induced opportunities tend to favor those in the earlier stages of their life course for several reasons. First, younger workers tend to have better educational preparation. This is especially true in China, because access to higher education and vocational training expanded greatly since the beginning of the economic transformation (see Chapter 3). Second, younger cohorts are likely to have less biographical experience of loss; thus, they are more likely to be risk taking and mobile in response to opportunities. This line of argument points to a general proposition: for older cohorts, more risks (either perceived or real) are attached to emerging economic opportunities; consequently the less likely that members of earlier cohorts will take advantage of these opportunities. Therefore, we expect to ﬁnd signiﬁcant cross-cohort differences in responses to emerging opportunities and risks: In the era of market reform in China, there is a declining rate of job shifts across cohorts, net of the effects of other factors. In particular, younger cohorts tend to have higher rates of job shifts, especially from the state sector to the nonstate sector, which is closer to market opportunities. We also expect signiﬁcant variations in the effects of allocative mechanisms across cohorts. Individual responses to opportunities are moderated by their social locations; social location, in turn, is shaped by both historical experience and life stage, both of which are reﬂected in cohort membership. For instance, those with positional power or who work in high-status organizations tend to beneﬁt more from their current

Economic Transformation and Life Chances

279

jobs than those in peripheral positions or organizations. Given that members of earlier cohorts are more entrenched (compared with later cohorts) into the redistributive system of state socialism, we expect signiﬁcant cohort differences in the effects of these allocative mechanisms. That is, the behaviors of older workers (located in earlier cohorts) are more likely to be governed by the allocative mechanisms of the redistributive institutions; on the other hand, younger cohorts may exploit market-oriented advantages (e.g., education) and be less bounded by their locations in the redistributive economy. Behaviorally, then, the younger cohorts would be more sensitive to market-related mechanisms of allocation. Thus, we expect to ﬁnd that, in the era of market reform, in each cohort those with positional power (e.g., cadres) and those located in work organizations closer to the redistributive authority (e.g., governmental agencies) tend to have lower rates of job shifts to the nonstate sector. This pattern is signiﬁcantly stronger for workers in older cohorts, compared to those in younger cohorts. In contrast, younger cohorts tend to be more sensitive to market mechanisms such as returns to formal education. Determinants of Personal Income: Cohort Variations in Economic Resources If social stratiﬁcation processes are moderated by life-course factors, we should also expect to ﬁnd signiﬁcant variations in the association between stage of the life course and sources of economic resources. Given our recognition that large-scale economic transformations in China involve competing institutions and incentives for individuals and that there are signiﬁcant variations among cohorts in their response to these opportunities and risks, logically we expect that patterns of determinants of economic resources also reveal signiﬁcant cross-cohort variations. Two considerations motivate our hypotheses. First, as previously noted, there is a close link between cohort membership and the extent to which workers are entrenched into the state socialist redistributive system. Each cohort entered the work force in distinctive historical contexts. Large-scale nation building and nationalization in the evolution of state socialism in China point to the fact that the older cohorts were more likely to be incorporated into the existing institutional structure of redistribution. By contrast, as shown in previous chapters, signiﬁcant institutional changes in the last two decades (and dramatic state policy shifts during the Cultural Revolution) produced social dislocations and much more heterogeneous

280

The State and Life Chances in Urban China

organizational locations for the younger cohorts. Locations in different organizations and occupations signiﬁcantly affect their economic wellbeing. Second, seniority was used as a main principle in the redistribution of economic beneﬁts and promotion in pre-reform China (Zhou and Suhomlinova 2000). As a result, older cohorts, through their seniority in work organizations, tend to enjoy more economic beneﬁts than younger cohorts. Thus, the beneﬁts of positional power associated with the redistributive system should be especially salient for older cohorts, but less so for younger workers. Similarly, beneﬁts from one’s location in certain types of work organizations should also be signiﬁcantly different across cohorts. These structural characteristics are coupled with the fact that individuals in the youngest cohort who are in less desirable jobs in the state sector may have a stronger incentive and more opportunities to seek alternative jobs in the nonstate sector, leading to more equalizing of beneﬁts within the younger cohort. These considerations suggest that the youngest cohort beneﬁts more from returns to education than do older cohorts in the era of economic reform. Moreover, positional power and employment in the state sector tend to beneﬁt older cohorts more than younger cohorts, whereas younger cohorts are more likely to beneﬁt from employment in occupations and organizations that are closer to markets (e.g., private or hybrid ﬁrms).

historical context, cohort attributes, and evolving opportunity structures In the analyses reported in this chapter, we delineate three “historical cohorts” to capture respondents’ location in the life course at a particular point in time. By historical cohort we refer to those individuals who entered the labor force in the same historical period, rather than those of the same age. Because of the political ﬂuctuations in China, individuals entered the labor force at signiﬁcantly different ages. For instance, the closure of colleges during the Cultural Revolution forced millions of high school graduates into the labor force. Periodic changes in educational curriculums led to variations in age of entry into the workforce. More important, historical cohorts based on workforce entry capture each cohort’s distinctive experiences of the macro-political and economic changes in China in the forty-ﬁve years of our study. For our purposes we specify three cohorts: (1) those who entered the labor force between 1949 and 1965; (2) those who entered the labor force during the Cultural Revolution (1966–1979); (3) those who entered the

Economic Transformation and Life Chances

281

50

percent

40

30

20

10

hy

br

id

te iv a pr

rm fa

co

lle c

tiv

e

m st

at

e

fir

bl ic pu

go

v't

0

cohort 1

cohort 2

cohort 3

ﬁgure 10.1. Distribution of ﬁrst-job destinations, by cohort.

labor force during the era of economic transformation (1980–1994). We also consider within-cohort variations in the timing of labor force entry by incorporating age into our analysis. To understand the impacts of these changes for the life chances of individuals at different stages of the life course, we ﬁrst take a close look at the distinctive historical contexts in which different cohorts were situated when they entered the labor force. Our thesis is that, given the dramatic changes of macro-political environments and frequent shifts of state policies, cohorts who entered the work force in different historical periods experienced very different opportunity structures. Indeed, as Figure 10.1 shows, there are distinctive patterns of ﬁrst-job destinations across cohorts, reﬂecting variations in opportunity structures across historical periods. The earliest cohort had the highest proportion entering government agencies. It also had noticeable advantages over the second cohort in entering public organizations and state ﬁrms. However, Figure 10.1 also suggests that changes in the opportunity structure did not always evolve in a linear fashion. The youngest cohort has had the largest proportion entering public organizations and state ﬁrms. By contrast, the second cohort – who entered the labor force during the Cultural Revolution – had dramatically different job opportunities. They had the

The State and Life Chances in Urban China

282 50

percent

40

30

20

10

cohort 1

cohort 2

hy br id

pr iv at e

rm fa

co lle ct iv e

m fir e at st

pu bl ic

go v't

0

cohort 3

ﬁgure 10.2. Distribution of job destinations in 1993, by cohort.

smallest proportion entering the state sector and a large proportion sent to rural areas. But the advantages or disadvantages associated with ﬁrst-job entry may or may not be perpetuated over the subsequent life course. By 1993 (the most recent year in which we have complete information for all respondents in our sample) there remained noticeable differences as well as similarities among the three cohorts, as shown in Figure 10.2. Although the proportion of those in government jobs continued to be similar for all three cohorts, a larger proportion of the earliest cohort was most apt to still be in public organizations (compared with the other two cohorts). Overall, the earliest cohort retained the largest proportion in the state sector (government, public, and state-owned ﬁrms). Note, however, that the second cohort had a lower proportion in government and public organizations than did those in the third cohort. This was mainly due to the social dislocations of the Cultural Revolution when a large number of urban youth were sent to rural areas, most of them returning to urban areas in the late 1970s (see Chapter 5; see also Davis 1990, 1992a, 1992b). The distribution of those in private ﬁrms (including self-employed) and hybrid ﬁrms reﬂects cohort effects. The youngest (third) cohort has the largest proportion in these two categories, whereas the ﬁrst cohort has the lowest. Clearly, workers in the most recent cohort have especially

Economic Transformation and Life Chances

283

beneﬁted from the emerging market opportunities in China. Note that, for the third cohort, it is likely that very little time elapsed between ﬁrst and current jobs due to their relative short career duration. Overall, these patterns are consistent with our argument that the earliest cohort remains more entrenched in the institutions of state socialist redistribution (the state sector), whereas the most recent cohort is more likely to be located outside the state sector. However, this proposition needs to be qualiﬁed: the cross-cohort variations do not follow a linear trend. In particular, the organizational locations of the second cohort appear to be an anomaly. These variations reﬂect drastic state policy shifts in the past that produced severe disruptions in the normative life course, as we have discussed in previous chapters. Table 10.1 compares the characteristics of the three cohorts. There is more than a ten-year difference between adjacent cohorts. Some differences across cohorts are due to sampling variations. For instance, the smaller proportion of women in the earlier (1949–1965) cohort reﬂects the fact that female workers retire early; hence, they are no longer in our sample. Other variations, however, reﬂect substantive differences in opportunity structures experienced by members of the three cohorts. Note the signiﬁcant variations in formal educational attainment across the three cohorts. More than a quarter of the employees in the earliest cohort only have an elementary education, compared to only about 4 percent of the latest cohort, reﬂecting the dramatically increasing educational opportunities over time (Zhou, Moen, and Tuma 1998). Similarly, the proportion of those with a college education varies signiﬁcantly across cohorts. By 1993, 18 percent of individuals in the ﬁrst cohort have a college degree, whereas 26 percent of those in the third cohort have a college degree. In contrast, only 14 percent of those in the second cohort have had college education. This anomaly can be explained by the closing of colleges during the Cultural Revolution period; this dramatic change severely limited the second cohort’s access to higher education.1 There is also evidence that earlier cohort was in better social positions, given their career progression, than the younger cohorts. For example, in 1993, a much larger proportion of the ﬁrst cohort, rather than the third cohort, was in highrank cadre and professional positions.

1

Note that the percentage of college education in Table 10.1 is higher than the national average. This is partly because a larger proportion of our sample is drawn from large cities, where average educational levels are high. Also employees with lower educational levels are likely to retire early, thus leaving our sample of analysis. Another factor is that we included “dazhuan” (college granting associate degree) into the “college” category.

284

The State and Life Chances in Urban China

table 10.1. Descriptive Statistics of Three Cohortsa of Workers in Urban China, 1993 Cohort 1 (1949–1965 cohort) Female Age education Elementary Middle school High school College occupation High-rank cadre Low-rank cadre High-rank professional Low-rank professional Ofﬁce worker Service worker Skilled worker Unskilled worker Private entrepreneur type of work organization Government Public organization Central gov’t-owned ﬁrm Local gov’t-owned ﬁrm Collective ﬁrm Private/hybrid ﬁrm Total N a

Cohort 2 (1966–1979 cohort)

Cohort 3 (1980–1994 cohort)

0.33 52.50

0.44 39.52

0.44 28.73

0.28 0.29 0.25 0.18

0.10 0.40 0.36 0.14

0.04 0.24 0.46 0.26

0.08 0.18 0.17 0.10 0.05 0.09 0.21 0.10 0.02

0.02 0.13 0.06 0.13 0.06 0.13 0.28 0.16 0.03

0.005 0.08 0.04 0.20 0.09 0.15 0.26 0.13 0.05

0.11 0.17 0.28 0.18 0.21 0.05 847

0.09 0.10 0.26 0.23 0.23 0.08 1746

0.11 0.13 0.24 0.20 0.21 0.12 1078

Cohorts based on the historical period of entry into the labor force.

Given these observed cohort-speciﬁc differences in both human capital and location in the occupational and organizational structures, we now examine cohort variations in career patterns (with respect to job shifts) and in economic well-being (personal income).

response to opportunities and risks: cohort variations in job-shift patterns One of the most signiﬁcant structural changes associated with China’s economic transformation is the rise and expansion of the nonstate sector.

Economic Transformation and Life Chances

285

Because economic beneﬁts have historically been tied to one’s work organization, patterns of job-shifts provide important information about individuals’ changing career opportunities and risks. For instance, the last chapter shows that, despite large-scale reform, job-shift patterns still reﬂected the importance of the state sector, as evidenced by higher rates of shifting to the state sector relative to the nonstate sector. What is not clear from the last chapter is (1) whether there are signiﬁcant variations across cohorts in their response to job opportunities across different types of work organizations and (2) how these observed patterns reﬂect the moderating role of life-course factors in social stratiﬁcation processes. To address these issues, we estimated the hazard rates of job shifts, by the three cohorts deﬁned earlier, toward a distinctive set of work organization destinations between 1980 and 1994.2 We included only those job shift events that took place across different types of work organizations. For instance, in the analysis of job shifts to government agencies, only those events that originated from nongovernmental organizations (e.g., public organizations, state ﬁrms, collective ﬁrms, and nonstate ﬁrms) were included. We also took into consideration their previous organizational location, which is captured by the set of “work organization” variables in the model. By doing so, we can assess directions of career patterns among different types of work organizations. Table 10.2 reports the parameter estimates in this set of analyses. We have speculated that the moderating effects of life-course factors should be reﬂected in differential rates of job shifts across cohorts and, in particular, there should be a decreasing rate of job shifts across cohorts. We examined the magnitudes of intercept coefﬁcients associated with the three cohorts. To allow a more meaningful interpretation of these intercepts, we centered the age variable at the mean age of each cohort. Thus,

2

For this set of analyses, we included both respondent and his or her spouse (if being married) in our sample who were in the labor force during this period. This is because job shift events took place infrequently and it was not feasible to estimate patterns of job shift to some types of work organizations when only respondents were included. The use of both spouses in the analysis raises the issue of statistical independence of the cases in the sample. One might argue that spouses are interrelated in their job shift patterns. Zhou and Moen (2001b) compared spouses’ job shift patterns and found that the effects of individual attributes are not affected signiﬁcantly by the characteristics of the other spouse. We suspect that this is because husband and wife in urban China participate in the labor force throughout one’s own life course, and their life events evolve in a parallel, rather than in a dependent, manner. The ﬁndings reported in the text are similar to those based on the analysis of respondents only. However, several job destinations are not estimable due to small number of cases when only respondents are included.

286

−0.558 −0.805∗∗ — −0.470 −0.876∗∗ −0.233 1.233∗ −1.243 −1.154 104 81.1/12 −5.090∗∗ −0.012 −0.001

−0.270 0.495†

— 0.063 0.082 0.560 1.342∗∗ 0.481 1.041 140 161.2/12 −6.949∗∗ 0.141 −0.054∗∗

−5.176 −0.444∗ −0.044† 0.627∗ 2.116∗∗

−7.486 −0.346† −0.080∗∗ 0.594∗∗ 2.592∗∗

Intercept −10.142 Female −0.365 Age −0.067∗∗ education High school 0.443 College 1.270∗∗ occupation Cadre 1.751∗∗ Professional 0.577 work organization Government — Public organization 0.941† State ﬁrm 1.141∗∗ Private/hybrid ﬁrm 0.566 Farm 3.189∗∗ % Collective sector output 3.202∗ % Hybrid sector output −0.318 Number of events 44 χ 2 /d.f. 71.8/12 Intercept −7.735∗∗ Female 0.253 Age −0.010

∗∗

∗∗

∗∗

Cohort 3

Cohort 2

Cohort 1

Government

−0.445 −0.232 0.331 — 0.033 −0.689 1.848∗∗ −0.128 −2.709∗ 116 115.6/12 −5.758∗∗ −0.704∗∗ −0.051∗∗

0.971† — 0.526 — 2.627∗∗ −0.385 −2.196 33 56.4/11 −7.326∗∗ −1.519∗∗ −0.028†

0.980∗∗ 2.286∗∗

−6.876 −0.305 −0.027∗

∗∗

Cohort 2

0.329 −0.248

1.976∗∗ 2.652∗∗

−9.118 −0.270 −0.046∗

∗∗

Cohort 1

0.732∗∗ 0.853∗∗ — −1.009† 2.436∗∗ −0.936† −1.900∗ 250 255.1/12 −7.126∗∗ −0.074 −0.019

2.062∗∗ 0.300 — −0.175 2.920∗∗ −0.578 0.329 85 124.1/12 −9.203∗∗ −0.648 −0.046∗ 0.009 — −0.764∗∗ — 1.681∗∗ −1.884 −3.501∗ 69 115.9/11 −4.830∗∗ −0.678∗∗ −0.043∗

−0.415† −0.594∗∗ 0.139 0.689∗

−1.498∗ −1.345∗∗

0.277† 0.432†

−5.548 −0.384∗∗ −0.043∗∗

∗∗

Cohort 2

0.249 −0.217

−7.337 −0.502∗ −0.054∗∗

∗∗

Cohort 1

State Firm

1.756∗∗ 3.760∗∗

−6.759 −0.113 −0.081∗

∗∗

Cohort 3

Public Organization

−0.743 −1.170∗ — −0.851† 0.889∗ −2.058∗∗ −2.653∗∗ 169 177.2/12 −6.493∗∗ −0.395 −0.069∗

−1.280∗ −2.255∗∗

0.716∗∗ 1.739∗∗

−5.106∗∗ 0.344∗ −0.102∗∗

Cohort 3

table 10.2. Parameter Estimates of Logistic Model for Job Shifts Across Type of Organization in the Reform Era (1980–1994), by Cohorta

287

0.042 0.027

−0.181 −0.191

0.295 0.068 — 0.457 2.835∗∗ 0.793 −0.916 150 225.4/12

0.842∗ 0.390

−0.303 −0.936†

−0.403 0.476 — −0.078 3.309∗∗ 1.635 −2.446 59 104.5/12

Cohort 2

−0.205 0.039 — −1.719† 2.083∗∗ 0.516 −4.984∗∗ 102 57.2/12

−0.878 −0.292

−0.065 0.214

Cohort 3

−1.001 — −0.475 — 2.403∗∗ 1.337 4.381∗∗ 40 57.6/11

−0.934 −1.209†

−0.330 0.426

Cohort 1

0.172 −1.033∗ −1.238∗ −0.842† −0.692∗∗ — 1.504∗∗ −0.340 2.741∗∗ 100 71.4/12

−1.376∗ −1.174∗ −0.957∗∗ — 1.390∗∗ 0.220 3.751∗∗ 155 178.7/12

−0.505∗ 0.197

Cohort 3

−0.640 −1.404∗∗

−0.072 −1.922†

Cohort 2

Private Firm

1.059 — 0.950† — 3.119∗∗ 1.748 3.509† 27 34.3/11

0.312 −1.804†

0.232 0.755

Cohort 1

−0.785 −1.207∗ −0.616∗ — 0.441 1.172 2.598∗ 80 35.8/12

−0.683 −0.492

0.565∗ 1.575∗∗

Cohort 2

Hybrid Firm

−0.824 −0.702 −0.453 — 1.467∗∗ 1.258 2.360∗ 70 28.7/12

−0.394 0.077

0.212 0.817∗

Cohort 3

∗

p < .01. p < .05. † p < .10. a Cohorts based on the period of entry into the labor force: cohort 1, 1949–1965; cohort 2, 1966–1979; cohort 3, 1980–1994. Note: Job shifts refer to job changes across organizations after one’s ﬁrst job. Middle school or below is the reference category for education; worker, for occupation; collective ﬁrm, for work organization. All models also included a set of dummy variables to indicate the respondents’ locations by province and city size.

∗∗

education High school College occupation Cadre Professional work organization Government Public organization State ﬁrm Private/hybrid ﬁrm Farm % Collective sector output % Hybrid sector output Number of events χ 2 /d.f.

Cohort 1

Collective Firm

288

The State and Life Chances in Urban China

the intercept for the ﬁrst cohort referred to the “overall” log-rate of job shift for a male respondent with the average age of that cohort, a middle school or lower education, working as an ordinary worker in a collective ﬁrm and entering the labor force before 1966. Similarly, the intercept for the second cohort refers to the log-rate of a worker with the same attributes except that he or she had the mean age of his/her cohort and entered the labor force between 1966 and 1979. The intercept for the third cohort refers to the log-rate of a respondent of the same attributes but with the mean age of that cohort and entering the workforce after 1980. In Table 10.2, the negative signs for all these intercepts reﬂected the fact that job shifts were rare events among those already in the workforce. Looking at job shifts toward all organizational destinations, there was a decreasing rate of job shift across the cohorts, net of the effects of other covariates in the model, with the third cohort having the highest and the ﬁrst cohort the lowest rates. For instance, the “overall” rate of shifts to “government” destination was 0.006 (exp [−5.176] = 0.006) for the third cohort, but it was 0.00004 and 0.0006 for the ﬁrst and second cohorts, respectively. This is not surprising, given the cost of such shifts, the absence of better opportunities, and the general process of duration dependence in career paths. These cohort differences in rates of job shifts may have reﬂected advantages of cumulative advantages in social positions for earlier cohorts. However, in an era of dramatic institutional change, the lower rate of job changes among the earlier cohort may also reﬂect its effective exclusion from emerging opportunities, leading to cohort-based reallocation of opportunities. We also speculated that, because of the variations in the extent to which different cohorts were entrenched into existing institutional arrangements, there should be signiﬁcant variations across cohorts with respect to the effects of positional power and market values on their job shift patterns. This can be tested by examining the effects of the covariates in Table 10.2. We limited our attention to those variables that directly measure social position.3 The Role of Education. Overall, educational credentials, especially a college education, increased the rate of job shifts to work organizations in the state sector (government agencies, pubic organizations, and state

3

Note that the insigniﬁcant effects of the covariates associated with the ﬁrst cohort in some of these destinations (e.g., private entrepreneurs and hybrid ﬁrms) may be partly due to the smaller number of events and sample size for that cohort. Thus, we should be cautious in interpreting these ﬁndings.

Economic Transformation and Life Chances

289

ﬁrms). But educational level was less salient for shift toward the nonstate sector, except for the signiﬁcant, positive effect of college education on the rate to hybrid ﬁrms for the second and third cohorts. Our theoretical interests are in the patterns of cohort variations. The overall patterns suggest that a college education was beneﬁcial for the second and third cohorts, as indicated by its positive, signiﬁcant effects for these two cohorts, especially in job shifts to the state sector and to hybrid ﬁrms. These ﬁndings are consistent with our expectation that there are signiﬁcant variations across cohorts in the effects of education on workers’ response to emerging opportunities. The Role of Occupational Positions. Effects of occupational status showed a more complex pattern. Overall, there was strong inertia associated with cadre and professional positions. That is, cadres and professionals tended to remain in their work organizations (as indicated by the negative signs associated with these covariates in most of the job shift analysis). Although there were cohort variations, the patterns were opposite to our expectations. It appeared that the younger cohorts (the second and third cohorts), rather than the oldest, tended to experience the most inertia, especially in job shifts to public organizations, state ﬁrms as well as to government agencies. One speculation is that the young generation of cadres and professionals may have been well placed in their work organizations, because of the new cadre policies since the 1980s that promoted younger, educated cadres and professionals. As a result, cadres and professionals in the third cohort may have had especially desirable jobs, discouraging their propensity for job shifts. The role of Organizational Locations. The set of covariates for “work organization” indicates the origins of job shifts. The effects of organizational origins allow us to identify the speciﬁc directions and paths of job shifts in the reform era. Consistent with evidence in Chapter 9, patterns of effects by organizational location indicate some evidence of segmented labor markets in urban China. Job shifts among types of organizations within the state sector (government agencies, public organizations, and state ﬁrms) were more ﬂuid than across the boundaries of state and nonstate sectors. This pattern was revealed in the relatively permeable barriers in shifts across government, public organizations, and state ﬁrms. By contrast, those working in the state sector tended to be less likely to shift to the nonstate sector (collective, hybrid, and private ﬁrms). There were also noticeable cohort variations among different types of organizations. For example, there was a higher rate of shift from government agencies to state ﬁrms for the ﬁrst two cohorts. In the early stages of the reform era, state ﬁrms tended to offer higher income than did government

290

The State and Life Chances in Urban China

agencies and public organizations. These patterns suggest that workers in earlier cohorts may be in a better position to exploit these opportunities within the state sector than are their younger colleagues. The high rates of job shifts from “farm” origins to all organizational destinations reﬂected the returns of the sent-down youth to urban areas after 1979. Because of this unique historical context, we included the farm origin mainly as a statistical control and therefore do not interpret the effects of farm origin. Overall, we found signiﬁcant cohort variations in the effects of social positions on job shift patterns. First, it is evident that those with positional power and those in high-status organizations (e.g., government) tended to shift within the state sector, where the beneﬁts attached to their positions were presumably protected or enhanced. Second, a college degree had differential effects for individuals in different cohorts: It had the largest effect on the third cohort, especially in facilitating their shifts to the state sector (government, public organizations, and state ﬁrm as job destinations). Note also that college education increased the rate of shifting to the hybrid ﬁrms for the second and the third cohorts, indicating that the younger cohorts were more likely to exploit the emergence of market opportunities in allocating human resources.

cohort variations in economic rewards We next consider cohort variations in economic beneﬁts during the economic growth period in China in the 1980s and early 1990s. Table 10.3 reports the OLS estimates of the covariates for the 1993 income. To control for location-based variations in income, we included a set of dummy variables indicating the speciﬁc city in which the sample was drawn (with Beijing as the reference category). The parameter estimates and the associated signiﬁcance levels indicate within-cohort variations. We discuss both within-cohort and cross-cohort variations. Intercept and Age. The intercept for each cohort indicates the overall log-income for a male worker in a collective ﬁrm in Beijing with middle school or lower education and with the mean age of that cohort. The magnitudes of the three cohorts were comparable. Although women had lower incomes in all three cohorts, age effects varied signiﬁcantly across cohorts.4 4

Recall that the age variable is constructed by subtracting mean age in that cohort from the age of each respondent. These within-cohort age effects indicate returns to work

Economic Transformation and Life Chances

291

table 10.3. OLS Parameter Estimates of Determinants of 1993 Income for Three Cohortsa of Workers in Urban China Cohort 1 Intercept Female Age Age2 /100 education High school College occupation High-rank cadre Low-rank cadre High-rank professional Low-rank professional Ofﬁce worker Service worker Skilled worker Private type of work organization Government Public organization Central gov’t-owned ﬁrm Local gov’t-owned ﬁrm Private/hybrid ﬁrm Total N Adjusted R2

∗∗

Cohort 2

Cohort 3

∗∗

4.280 −0.172∗∗ 0.003 −0.095∗∗

4.184∗∗ −0.124∗∗ 0.009† −0.101∗∗

0.038 0.044

0.070∗∗ 0.092∗

0.166∗∗ 0.162∗∗

0.362∗∗ 0.251∗∗ 0.249∗∗ 0.348∗∗ 0.082 −0.023 0.125† 0.338∗

0.313∗∗ 0.253∗∗ 0.249∗∗ 0.161∗∗ 0.138∗ 0.096∗ 0.043 0.154†

— 0.163∗ 0.259∗∗ 0.143∗ 0.138∗ 0.181∗∗ 0.061 0.347∗∗

0.308∗∗ 0.141∗ 0.172∗∗ 0.126∗ 0.245∗∗ 841 0.44

0.127∗∗ 0.189∗∗ 0.195∗∗ 0.105∗∗ 0.476∗∗ 1715 0.43

0.099 0.085 0.108∗ 0.085† 0.496∗∗ 1057 0.46

4.313 −0.255∗∗ −0.015∗∗ −0.127∗

∗∗

p < 0.01. p < 0.05. † p < 0.10. a Cohorts based on the period of entry into the labor force: cohort 1, 1949–1965; cohort 2, 1966–1979; cohort 3, 1980–1994. Note: Middle school or below is the reference category for education, worker for occupation, collective ﬁrm for work organization. All models also include a set of dummy variables to indicate the respondents’ city locations. ∗

Returns to Education. Education had no signiﬁcant effect on the 1993 income of members of the ﬁrst cohort. By contrast, income increased with educational level for both the second and third cohorts. Returns to education were especially pronounced for those in the third cohort (although

experience at different stages of one’s occupational life course. They are not directly comparable across cohorts.

292

The State and Life Chances in Urban China

having a college education did not appear to be more advantageous than a high school education). Those with a college education in the third cohort also had a signiﬁcantly higher income than their counterparts in the other two cohorts. These ﬁndings indicate that members of the third cohort beneﬁted more economically from their human capital, compared with the earlier cohorts. There are several plausible explanations for the observed patterns. First, it is likely that those in the third cohort, because of their more recent formal education, had the most updated knowledge, an especially attractive attribute in the labor market. Second, it is also likely that economic returns to education for the earlier cohorts were largely transmitted through their present social positions (cadre and professional ranks). When we estimated models with only educational variables included, education contributed signiﬁcantly and in large magnitude for the ﬁrst cohort as well. This is evidence that, for the ﬁrst cohort, the effects of education are mediated through their occupational positions. Returns to Occupational Status. We use more detailed occupational status categories for the analysis of income determinants, given the feasibility allowed by the sample size. Because of the few cases of highrank cadres in the third cohort, we combined this category with low-rank cadres for this cohort. As we hypothesized, income returns to positional power (high-rank cadre) were highest for the ﬁrst cohort. Recall that the ﬁrst cohort had the highest proportion of high-rank cadres (8 percent). This ﬁnding means that association with political authority was especially beneﬁcial for the ﬁrst cohort. Returns to low-rank cadre and high-rank professional status were similar within the ﬁrst cohort and the second cohort. However, low-rank professionals in the ﬁrst cohort enjoyed considerably higher income than later cohorts. Also note that private entrepreneurs in the ﬁrst and third cohort had signiﬁcantly higher incomes. Returns to Organizational Locations. For the ﬁrst cohort, government employees reported the highest incomes (other things being equal) and employees in the state sector also had signiﬁcantly higher incomes than those in collective ﬁrms. This was also the case for the second cohort, with those in the state sector having signiﬁcantly higher incomes than those in collective ﬁrms. But the income advantages of being in the state sector were less obvious for the third cohort. Employees in private/hybrid ﬁrms had higher incomes than those in collective ﬁrms (the reference group), and this was especially the case for those in the youngest cohort. Consistent with our expectation, we ﬁnd distinctive cohort differences in the effects of organizational location on income, with members of the ﬁrst cohort working in government agencies having the highest

Economic Transformation and Life Chances

293

income across cohorts. Conversely, those in the last two cohorts working in private or hybrid ﬁrms had signiﬁcantly higher incomes than those in the ﬁrst cohort working in such ﬁrms. These two polarized cases show the strongest evidence that market and state socialist institutions have distinctive incentives for different cohorts. There were no signiﬁcant differences among the three cohorts in other types of organizational location. Overall, the strongest evidence of cohort differences is found in returns to both education and organizational location. These patterns are consistent with our ﬁndings related to job shift patterns. Speciﬁcally, the ﬁrst (oldest) cohort tends to beneﬁt from working in organizations closer to redistributive authorities, whereas youngest cohorts beneﬁt more from working in organizations closer to market activities.

chapter summary A major task of sociological research is to explain how social processes produce variations in individual life chances. This task becomes especially critical in times of dramatic social upheavals, such as those currently still underway in former state socialist societies. Social changes incorporate two sets of processes inﬂuencing the life chances of individuals: (1) by changing the opportunity structure and/or (2) by changing the mechanisms affecting opportunity. A life-course approach to the links between history and biography considers the signiﬁcance of cohort membership in shaping both processes. Thus, a life-course formulation offers a unique vantage point from which to examine how the timing of large-scale changes in individual biographies moderates their impacts on individuals’ life chances through differential exposure to opportunities and risks, as well as their differential locations in the social structure. The main theme of this chapter is to explicitly recognize and develop arguments about the importance of life-course factors in shaping individual life chances and to demonstrate these life-course effects in a context of fundamental social changes. What have we gained by adopting a lifecourse perspective and the corresponding research design? By uncovering distinctive cohort variations in access to the opportunity structure, we are able to identify the role of life-course factors in moderating and mediating the links between institutional changes and life chances. For example, in Chapter 9, we found that individuals with a college education had a higher rate of moving to both government agencies and hybrid ﬁrms in the reform era (1980–1994). But our research design there implicitly

294

The State and Life Chances in Urban China

assumed this to be uniform for all those in the workforce, controlling for the effects of other covariates in the model. In contrast, the examination of the cohort-speciﬁc effects reported in this chapter allows us to assess how different cohorts have, in fact, distinctive opportunities. We ﬁnd that, whereas college graduates in all three cohorts experienced signiﬁcant and higher rates of shifting to government agencies, there are noticeable cohort differences in their shifts to hybrid ﬁrms, with those in the older cohort (those who entered the labor force between 1949 and 1965) having no advantage at all. Moreover, our income analysis further reveals the rationale behind this ﬁnding: Those in the older cohort who stayed in government agencies beneﬁted more from the redistributive beneﬁts than those who were in hybrid ﬁrms, whereas the opposite is true for the younger cohorts. Educational credentials especially facilitated the ability of members of younger workers (the second and the third cohort) to capitalize on the job opportunities in emerging markets and to reap greater returns to education in the distribution of income. Thus, a life-course perspective addresses a distinctive set of sociological issues about the relationship between social changes and life chances. As we shift our attention from aggregate conventional categories (e.g., occupational groups) to take into consideration particular cohorts’ stage of life course, we detect a more dynamic picture of variations and changes, identifying differential impacts of social changes on members of a society, as moderated by life-course factors. The life-course approach also facilitates assessment of the differential effects of allocative mechanisms across cohorts. We found signiﬁcant differences across cohorts along several dimensions of theoretical interest. Most striking is the fact that cross-cohort variations mainly relate to the impacts of organizational location, as evidenced in the signiﬁcant variations between the ﬁrst cohort on the one hand and the second and third cohorts on the other, especially in terms of job shifts to government and state ﬁrms and in the determinants of income. There are several processes that lead to this pattern. First, the oldest cohort was more entrenched in the existing social structure and closer to the redistributive economy (with a larger proportion of high-rank cadres and of those in the state sector). As a result, they tended to beneﬁt more from redistributive processes and were less sensitive to the market opportunities emerging in the 1980s and early 1990s. Second, the signiﬁcant disruption of the life course for the second cohort (as a result of the Cultural Revolution) may have also made them more prone to embrace new opportunities in the market

Economic Transformation and Life Chances

295

process, rendering characteristics of this cohort more similar to those of the youngest (third) cohort. However, there are some unexpected ﬁndings. For instance, in most cases, cadres and professionals from the second and third cohorts have a lower rate (as indicated by the negative sign of the coefﬁcients) of moving into state-owned organizations (e.g., government agencies, public organizations, and state ﬁrms), relative to ordinary workers. To interpret this pattern, note that the third cohort has had a relatively higher rate of job changes, as indicated by their intercepts and that they had a wider range of choices when they entered (or re-entered) the workforce. Their inertia may thus be partly attributed to the favorable positions they already occupy in the labor force. More important, the life-course factors we emphasize in this chapter are only part – albeit we believe an important part – of the contextual constraints on the social stratiﬁcation processes. There are many other “push” and “pull” factors that generate these observed patterns. Greater monetary rewards in the marketplace have induced younger cohorts to shift to the nonstate sector. On the other hand, shrinking opportunities in the state sector also have pushed employees to leave it. For instance, we ﬁnd that the ﬁrst and second cohort had a higher rate of shifts from government agencies to state ﬁrms. To explain this pattern, we note that, in the early stages of the reform, state ﬁrms had signiﬁcant improvement in economic beneﬁts (in the form of bonuses), relative to those in governments and public organizations; see Chapter 7. Thus, the pattern of shifts may indicate that those in governments try to move to more desirable organizations to exploit these beneﬁts. Another contributing factor is the periodical downsizing of government agencies in the reform era, which may have forced some of their employees to move to state ﬁrms. We believe that a life-course perspective also helps highlight some important theoretical issues about social changes (Hogan 1981). First, our focus on cohort differences provides clues to the unique path of China’s transition from state socialism. Unlike the “shock therapy” in the former Soviet Union and Eastern Europe, in China we observed the coexistence and dual structure of state and nonstate sectors. Our analyses suggest that many aspects of the institutional changes in China take place through generational replacement: the oldest cohort (in their forties and ﬁfties by the time of the economic transformation) experienced fewer job changes than did those in cohort 2 (in their thirties during the reform period) and those in cohort 3 (in their twenties) just entering the workforce when the

296

The State and Life Chances in Urban China

market economy had begun to open up.5 There are also competing incentives among different institutions for different cohorts of employees. Whereas younger cohorts beneﬁt greatly (in terms of income) from human capital, older cohorts gain signiﬁcantly from working in the state sector. Consideration of the life stage and social location of different cohorts has permitted a more reﬁned assessment of the impacts of economic transformations on, and the reallocation of opportunities and risks among, social groups. The life-course bridging of the divide between the study of macro-level events and micro-level biographies shed new light on the links between institutional changes and variations in life chances among social groups, suggesting as well the future trajectories of change in the social stratiﬁcation processes. 5

In our sample, 7.5 percent in the ﬁrst cohort experience job changes in the reform era, and 20.3 and 25.7 percent for the second and third cohorts, respectively.

part three SUMMARY AND CONCLUSION

W

e conclude this book with the next two chapters. In Chapter 11, we summarize the major ﬁndings regarding the ﬁrst theme of this book – redistribution and stratiﬁcation dynamics under state socialism – and discuss the implications of these ﬁndings for theoretical models of social stratiﬁcation in state socialist societies. In Chapter 12, we turn to issues related to the second theme of our study – sources and direction of institutional changes in urban China and consider the consequences of stratiﬁcation dynamics for the transformation of state socialism in China. We conclude this study with a discussion of the implications of the Chinese experience for understanding relationships between the state and society in a comparative perspective.

297

11 Social Stratiﬁcation Under State Socialism: Summary and Assessment

It is probably too soon to revisit state socialism and the possibilities that were never allowed to mature. Nonetheless, in time there will be a revisionist history that, while not denying communist horrors, will recognize socialism’s potentialities. Burawoy (2001, p. 1118)

We have taken the reader on a journey of examining the state and life chances in the history of the People’s Republic of China. Focusing on a sample of urban residents, we followed their educational experiences, their entries into the labor force, their life experiences in career promotion, and job shift across organizations, and the economic beneﬁts associated with these life chances. We “observed,” so to speak, dramatic changes in their life chances over the historical periods and over the stages of their life course. We have organized our observations, analyses, and discussions around two major themes: ﬁrst, the impacts of state socialist redistribution and stratiﬁcation dynamics on individual life chances; second, changes and continuity in the stratiﬁcation processes in China’s economic transformation since the 1980s. How do these empirical ﬁndings inform us of the theoretical models of state socialist redistribution and social stratiﬁcation? What lessons can we draw from the Chinese experience to understand institutional transformation of state socialism? In this chapter, we take stock by summarizing the main ﬁndings related to the ﬁrst theme – redistribution and stratiﬁcation dynamics under state socialism in China. We organize the discussion around the main theoretical issues outlined in Chapter 1, namely, the bureaucratic class thesis, the political capital thesis, the redistributive institution thesis, and 299

300

The State and Life Chances in Urban China

the thesis of stratiﬁcation dynamics. We will discuss the second theme – the assessment of institutional transformation in China – in the next, concluding chapter.

the bureaucratic class thesis revisited As we pointed out before, no concept is as central as that of the bureaucratic class in studies of social stratiﬁcation under state socialism. Since the 1950s, the notion of a new bureaucratic class has been the single most important theme to guide research on social stratiﬁcation in state socialist societies. Indeed, insofar as we are interested in the stratiﬁcation processes through which social positions are hierarchically structured and resources and opportunities are allocated across social groups, the central role of the socialist state and its bureaucratic organizations in redistribution logically lead us to the bureaucratic class as the focus of theoretical and empirical research. In Chapter 1, we proposed a working deﬁnition of the bureaucratic class in state socialist societies as a group whose members (1) enjoy signiﬁcantly better socioeconomic beneﬁts relative to other social groups, (2) exercise power and control over organizational assets and in “authority relationships” in the work organizations, and (3) can reproduce their class through status inheritance across generations. As we pointed out in Chapter 1, the bureaucratic class thesis was developed largely on the basis of anecdotal evidence and has not been subject to careful empirical scrutiny. Today, with the availability of data from large-scale social surveys and of ofﬁcial documents and statistics in the public domain, we are in a position to revisit and assess these issues on a rigorous basis. Below we navigate a series of issues, both theoretical and empirical, to reach this simple conclusion: In the Chinese context, the imagery of a bureaucratic class enjoying privileges and with almighty power is a theoretical construct supported by only weak empirical evidence. But this statement needs careful clariﬁcation and qualiﬁcation.

Socioeconomic Status and Bureaucratic Privileges Did the bureaucrats enjoy socioeconomic beneﬁts and other life chances that are signiﬁcantly different from other social groups in urban China under state socialism? Based on the empirical evidence revealed in this book, the short answer is no. The differences in income, housing, and

Social Stratiﬁcation Under State Socialism

301

table 11.1. Summary of Returns to Bureaucratic Positions

Beneﬁts/Status

Return to Cadre Positiona

Fluctuating Over Time?

Income Housing space Housing facilities Party membership

Yes. 24%–50% Yes. 31%–56% Yes. 20%–31% Positive (cohort)

Promotion in organization Job shifts

Positive (cohort)

Yes, weak Yes, weak Yes, weak Yes (among cadre cohorts) Yes (among cadre cohorts) —

a

No

Reference Tables 7.1 and 7.2 Table 8.3 Table 8.4 Table 6.2 Table 6.4 Tables 9.1 and 9.2

Unless otherwise indicated, the cadre category refers to high-rank cadres and the reference category is refers to unskilled workers.

other welfare beneﬁts among different social groups in urban China have been strikingly small and (theoretically) insigniﬁcant. In our study, socioeconomic beneﬁts are measured in several ways: personal income and housing beneﬁts, career advancement in organizations (promotion), and the acquisition of political status (party membership). In most cases, the category of “high-rank cadres” (administrative rank of “chu” or above) was used in these analyses as a measure of the bureaucratic class. In the ofﬁcial administrative system, the “chu” rank is equivalent to the rank of the head of the county government, although the substantive authority varies greatly depending on speciﬁc positions. In some cases where this category was inappropriate, we used cadre cohorts to infer socioeconomic beneﬁts associated with bureaucratic positions. Table 11.1 summarizes the main ﬁndings related to the various aspects of returns to bureaucratic positions in the previous chapters. Does the bureaucratic class, as measured by “high-rank” cadre status, confer signiﬁcant advantages in socioeconomic beneﬁts? We can answer this question in two parts: (1) In most cases, there are indeed statistically discernible differences among occupational groups, with high-rank cadres having better socioeconomic beneﬁts; but (2) such differences among occupational groups are relatively small, and they can hardly be construed as convincing evidence for the bureaucratic class thesis. First, consider the empirical evidence on personal income. As Chapter 7 shows, other things being equal, a high-rank bureaucrat on average earned about 24 to 50 percent higher income than the income of an unskilled worker – the lowest status group in the urban labor force – throughout the 1949–1994 years, net of the effects of other covariates in the

302

The State and Life Chances in Urban China

model. There were also positive returns to high-rank bureaucratic status in housing allocation. But the differences were also small: A high-rank bureaucrat had about 30 to 50 percent larger housing space than an ordinary worker. Second, in terms of upward mobility in acquiring party membership and promotions, there were positive returns to bureaucratic positions but we also found signiﬁcant variations across cadre cohorts, indicating that there was considerable heterogeneity in life chances among these bureaucrats and that it is questionable to treat them as belonging to the same class with similar, stable life chances. Of course, any assessment of social inequality depends on the choice of reference framework. In a subsistence economy, a 50 percent difference in income and housing space is substantial indeed. However, this is far from what the imagery of a privileged bureaucratic class has portrayed. A functional theory of stratiﬁcation or a typical economic argument on incentives would suggest that such small differences in socioeconomic beneﬁts between top managers and ordinary workers do not even provide strong enough incentives for the former to act responsibly. Indeed, if there are only about 50 percent differences in personal income and housing conditions between a county-level ofﬁcial and an unskilled manual worker, one can hardly treat this as evidence of substantive social inequality. One may be tempted to argue that the bureaucratic beneﬁts were often latent and not captured in personal income or even housing beneﬁts. However, for those who are familiar with everyday lives in Mao’s China, it is difﬁcult to sustain the claim that economic disparity between the bureaucrats (say, the head of a workplace) and the ordinary workers was substantially different from the empirical patterns revealed in our studies. In the Mao era, most cadres lived in the same residential housing projects with the ordinary workers in their ﬁrms, and their apartments were as plain as their subordinates’. The ﬁndings reported in this book are consistent with Whyte and Parish’s (1984) observation about the pursuit of an egalitarian policy by the Chinese state in the Mao era. Indeed, China was alone in this regard. Students of state socialism have already taken note of this fact in other state socialist societies. Szel´enyi (1978) summarized his ﬁndings in the Hungarian context as follows: In the case of cash allowances, a signiﬁcant discrepancy can be observed among different social strata: in white collar families, per capita allowances were about 20% higher than in families of industrial workers; agricultural labourers received less than half as much cash allowances as the white collars . . . White collars received almost 50% more per capita noncash beneﬁts than industrial workers,

Social Stratiﬁcation Under State Socialism

303

high managers and professional more than twice as much as agricultural labourers. (p. 69)

As Szel´enyi (1978: 77) further pointed out: “I am inclined to believe that even if we can speak about a new ruling class in East European state socialism then it is the most miserable privileged class in the history of mankind.” After enumerating the privileges and material beneﬁts associated with bureaucratic positions, Kornai (1992, p. 43) put it bluntly: “Let there be no misunderstanding: in spite of these material privileges, a functionary’s standard of living falls short of the level of the wealthy in a capitalist society. Nevertheless, the privileges are great enough for the withdrawal of them to come as a blow.” A comparative perspective may also shed light on this issue. Table 11.2 reports the Gini coefﬁcients for income inequality across selected countries. The Gini coefﬁcient indicates the expected absolute difference in incomes, relative to the mean, between any two persons drawn at random from the population. A larger Gini coefﬁcient implies greater income inequality. Table 11.2 shows that, compared with developed countries and East European countries, urban China had the lowest income inequality in the mid-1980s (the Gini coefﬁcient for China as a whole is higher, around .38 in 1988). Clearly, measured by income inequality, our ﬁndings and conclusions are consistent from a comparative perspective. In the reform era, economic inequality among social groups in China has increased substantially. But the processes cross-cut social groups. Among the bureaucrats, some gathered enormous wealth using their positional power; others were worse off (relatively) when certain functions of the central government and the Communist Party were weakened. Similarly, some nonbureaucratic groups were better off than others because of their entrepreneurial endeavor or association with speciﬁc economic sectors or activities. The conventional bureaucratic class model is too simplistic to account for these multifaceted changes in the processes of large-scale social transformations. Moreover, the increasing disparity in the reform era is less derived from state socialist redistribution than from the interplay between political power and markets.

Authority Relationships in Work Organizations Wright (1989) argued that the authority relationship in the work environment is an important dimension for identifying social classes, especially in state socialist societies. Despite the importance of authority relationships

304

The State and Life Chances in Urban China table 11.2. Gini Coefﬁcients of Income Inequality Across Countries Country Urban China eastern europe Czechoslovakia Hungary Poland USSR developed countries Finland Sweden Luxembourg West Germany Netherlands Canada United Kingdom France New Zealand Italy Australia United States

GINI Coefﬁcient

Data Year

.19 .28

1985 1995

.20 .22 .24 .28

1987 1986 1986 1986

.20 .21 .23 .25 .26 .28 .29 .30 .30 .31 .31 .34

1985 1987 1985 1984 1987 1987 1986 1984 1985/6 1986 1885/6 1986

Sources: Data on urban China are from Riskin, Zhao, and Li (2001, p. 28); data on “East European countries” are drawn from Atkinson and Micklewright (1992, p. 81); data on “developed countries” are from Gottschalk, Gustafsson and Palmer (1997, p. 39), with permissions from the publishers.

in these societies, there were surprisingly few studies of employment relationships in these societies (for exceptions, see Burawoy and Krotov 1992; Burawoy and Lukacs 1985; Stark 1986). In the Chinese context, Walder (1986) is the best work available in documenting and theorizing about authority relationships in China’s work organizations. Arguing against the totalitarian imagery of the Communist regimes, Walder emphasized the pervasiveness of traditional clientele relationships that were used to develop dependent employment relations between workers and the socialist managers. Moreover, positive incentives, rather than political repression, were often used to induce political compliance and adherence to the party line. In this “neotraditionalism” model, the socialist managers at the local level (e.g., heads of the factory, or in other types of workunits) had considerable power

Social Stratiﬁcation Under State Socialism

305

in exercising authority in political control in the communist political order. Does this mean that the authority relationship in the workplace reﬂected the almighty power of the bureaucratic class? On the one hand, the power wielded by these local bureaucrats was enormous and often uncontested by the rank and ﬁle workers. Because there was no exit option for ordinary workers, it was indeed difﬁcult for ordinary workers to resist political discrimination in the workplace. In this sense, the socialist managers were powerful and could well be repressive. On the other hand, the socialist managers in China, like those in other command economies, were highly bound in their authority by the political system through regulations and close bureaucratic supervision. In the Mao era, moreover, the socialist manager had few resources to reward friends and punish enemies. The life employment policies, which began to change only in the 1990s, and the tight state control of wages and welfare programs meant that the employment relationship has been largely externalized – determined by macro-political processes and state policies. Until in the late 1980s, the head of the state ﬁrm (or other workunits) had little authority to lay off workers. Over time workers in state ﬁrms grew a strong sense of entitlement, which provides a legitimate basis for their resistance to managers’ efforts to take these entitlements away (Cai 1996, 1998). In many cases, even reallocating workers from one section of the ﬁrm to a different section could be extremely difﬁcult. Thus, although the worker had no exit option to escape the authority relationship, nor did the manager have a free hand to push out those whom they disliked. For example, in the wave of reform in 1984, managers in 55 factories in Fujian Province cosigned an appeal to the provincial government for authority to appoint their own managerial assistants and to decide on bonuses in their workplaces (Renmin Ribao, March 30, 1984). In that year, the State Council issued a directive to increase the managerial autonomy of the industrial organizations which, in part, said: “Under the premise of national wage standards, regional classiﬁcations and subsidies, the ﬁrm can decide on the form of wage payment. The general manager had the authority to increase the wage grade for those who made special contribution to the ﬁrm; the proportion of those being increased cannot exceed 3% annually.” This well-publicized story showed the depth of the state penetration into the workplace and the limited authority the socialist managers possessed at that time. Stark’s (1986, p. 503) study of employment relations in a Hungarian ﬁrm portrayed a similar picture:

306

The State and Life Chances in Urban China

To protect themselves from uncertainties in the production process and from rigidities in the central regulations, workers use resources at hand, starting with their knowledge of the peculiarities of the particular machines on which they work and including their contacts and connections for obtaining parts, materials, and information. They use the threat of exit as a means to be shifted to another position, perhaps one with no immediate promise of a pay increase but with more opportunity to build credit among other workers and with shop management, yielding higher rewards in the longer term. Managers, for their part, enter into the informal networks of the shop ﬂoor because only through cooperation with worker groups can production goals be realized. Above all, these groups are not classes of persons – aggregations of individuals who fall into the same category based on shared abstract features; they are interacting groups, whose members share tools, information, and particular experiences, and are connected to each other (including shop management) through debts and credits, obligations and expectations.

In these descriptions and imageries, it is difﬁcult to ﬁnd convincing evidence of a powerful bureaucratic class in authority relationships in organizations. Rather, it appears that the socialist managers and workers lived in a symbiotic relationship, with the managerial power constrained by both formal authority structures from the above and the egalitarian culture from below. In this picture, the bureaucrat had the double identity as the representative of the state on the one hand and as the patron of the workers under his or her supervision on the other. The authority relationship in the workplace was much more complicated than the kind of “exploitative” relationship commonly observed in class analysis.

Intergenerational Status Inheritance Class reproduction is central to the making of a social class. Conventional studies of intergenerational status inheritance typically use mobility tables to detect and analyze patterns of social mobility between fathers and sons. At the empirical level, patterns of intergenerational-resource transfer and mobility are descriptions of the patterns of social stratiﬁcation. Similar patterns of intergenerational mobility in two societies may be caused by very different stratiﬁcation mechanisms. In a market society, a positive relationship between social origin and status attainment may be due to family-based resources that allow children of the privileged class to be situated in advantageous market positions. On the other hand, such a positive relationship in a state socialist society may be caused by state policies or redistributive authority, and it is not necessarily related to the resources one can provide for his or her children. For example, toward

Social Stratiﬁcation Under State Socialism

307

table 11.3. Summary of Findings on Intergenerational Status Inheritance

Beneﬁts/status Educational attainment Entry into the labor force Sent-down experience Party membership Promotion in organization

Return to Father’s Cadre Status

Return to Father’s Education

Fluctuating Over Time?

Reference

Positive

Positive

Yes

Table 3.2

Positive, weak

Positive, weak

Yes

Positive

Negative

Yes

No No

No No

— —

Tables 4.3 and 4.4 Tables 5.1 and 5.2 Table 6.2 Tables 6.3 and 6.4

the end of the Cultural Revolution, the government adopted the “replacement” (dingti) policy, by which the children of the urban residents could enter the same workplaces once their parents retired from their jobs. This policy was originally designed to encourage retirement and to deal with unemployment problems. But it had the inadvertent effect of reinforcing intergenerational status transfer (in terms of inheritance of work organizations). It is obvious that such status transfer was not endogenously determined by family-based mechanisms. Instead, one needs to seek answers in the institutional arrangements whose logic led to such “replacement” policies. We need to keep these issues in mind when we interpret evidence of intergenerational mobility. Although we did not use mobility table models in the studies reported here, several analyses have examined the role of intergenerational resource transfer and status inheritance, especially with regard to educational attainment (Chapter 3), entry into the labor force (Chapter 4), promotion in the Chinese bureaucracy (Chapter 6), and sheltering of children from adverse effects of state policies (Chapter 5). Table 11.3 summarizes these ﬁndings. Overall, there is evidence that social origin (e.g., father’s educational levels, occupational status, or family class background) had positive effects on educational attainment and entry into the labor force. For instance, the father’s high-status occupations facilitated one’s educational attainment and location in the labor force (e.g., acquiring more desirable occupations or workplaces). Parents with high political status also had some ability to reduce the negative impacts of state policies on their children, as shown in the case of channeling their sent-down children back to

308

The State and Life Chances in Urban China

urban areas in the Cultural Revolution. There is also evidence that children of cadres and professionals were likely to follow their fathers’ footsteps in their occupational destinations. In these instances, we observed the association between social origins and children’s status attainment, suggesting that parental resources contributed positively to children’s upward mobility. However, this statement must be qualiﬁed in two ways: First, although parental resources showed positive effects on children’s status attainment in some instances, the processes through which these two were linked varied greatly over time. In particular, the fact that this link at times was signiﬁcantly weakened (e.g., in the Cultural Revolution) cautions us that the association between social origin and social mobility is exogenously determined by macro-political processes. More important, as we have seen, shifting state policies also had direct negative impacts on the children of the bureaucratic class. As the analyses of educational attainment (Chapter 3) indicate, the reduction of inequality under radical state policies disproportionally affected the children of cadres and professionals. During the send-down episode, high-rank cadres were unable to protect their children from being sent to rural areas. Second, as Table 11.3 shows, most evidence of intergenerational status inheritance was observed at or before one’s entry into the labor force (e.g., educational advancement). Once a person has entered the workforce, parental resources did not play a signiﬁcant role in one’s status attainment in such areas as acquisition of party membership or promotion along bureaucratic ladders, or economic beneﬁts. For instance, the analyses of promotion patterns (Chapter 6) found no systematic effects of the father’s educational or occupational status on the life chances in organizations. In our preliminary analyses, we also examined the role of parental resources in personal income and housing and found no significant effect. This evidence suggests that, once a person formally enters the redistributive system, the impacts of parental resources and social origins attenuate over one’s life course, giving way to the organizational apparatus and other stratiﬁcation processes. A Summing Up Has there been a bureaucratic class in China’s social stratiﬁcation system? We have considered this issue from three dimensions: socioeconomic beneﬁts associated with their class situations, their positions in authority

Social Stratiﬁcation Under State Socialism

309

relationships, and intergenerational status inheritance. As we went through these areas, we found that the bureaucrats (cadres) in state socialist redistributions did have positional powers, and often enjoyed better economic beneﬁts, but their life chances were as ﬂuctuating, as vulnerable, as haphazard, as other social groups. In the Chinese context, the evidence is not convincing for an imagery of a bureaucratic class who has consistently better life chances, compared with other classes, or independent sources of authority in the workplace, or can effectively transfer status across generations. Insofar as the mid- and lower-level bureaucrats are concerned, the main focus of almost all empirical studies on this topic in the literature, it is questionable that they should be treated as a stable, autonomous social class with distinctive boundaries whose interests are protected through social closure. Before more convincing evidence is uncovered, we should treat the bureaucratic class thesis as a working hypothesis, to be further examined and empirically tested, rather than as a foregone conclusion.

human capital versus political capital Scholars have long argued that an important characteristic of the state socialist redistribution is that the allocation of opportunities and economic beneﬁts favors those with political capital rather than those with human capital. The logic behind this argument is well established: Because of the political logic of redistribution, political status is closely associated with the redistributive authority (Nee 1989). And the state used “positive incentives” to induce compliance (Walder 1986). Political status is a reward for political loyalty, whereas human capital (i.e., educational qualiﬁcation) is usually acquired through merit-based competition. In the analyses reported here, we studied the role of education extensively as one of the social stratiﬁcation processes. In most cases, we did not examine “party membership” in our statistical analyses because of missing information on this variable associated with a portion of our sample. However, it is logical that returns to political capital resemble returns to administrative status discussed in the preceding section. Table 11.4 summarizes returns to education in major areas of stratiﬁcation reported in previous chapters. The empirical evidence reported in this study showed suppressed returns to education, especially in the Mao era. For instance, educational credentials had little effect on bureaucratic promotions until the postMao era; returns to education were also dubious with regard to party

The State and Life Chances in Urban China

310

table 11.4. Summary of Findings on Returns to Education

Beneﬁts/Status Entry into the labor force Sent-down experience Party membership Promotion in organization Job shifts Personal income Housing space/facilities

Return to Education Positive Negative Yes in the reform era Yes Yes Yes Yes, weak

Fluctuating Over Time? Yes Yes — Yes, weak Yes Yes Yes, weak

Reference Tables 4.3 and 4.4 Tables 5.1 and 5.2 Table 6.2 Tables 6.3 and 6.4 Table 9.1 Table 7.1 Tables 8.3 and 8.4

membership or economic beneﬁts in certain periods. On the other hand, party membership does not always yield positive returns, as indicated by the determinants of personal income in the regional analyses (Chapter 7). The evidence (see Table 11.4) suggests that, although returns to education were positive in many areas of social mobility, they were also governed by similar stratiﬁcation dynamics and ﬂuctuated over time. As we noted in the preceding chapters, although the role of party membership appeared to be stable over time, the selection criteria for party membership varied considerably across distinctive historical periods. Indeed, the very notion of political loyalty varied over time with changes in state policies and leadership succession. These patterns need to be further elaborated and qualiﬁed. First, the processes that select individuals into political status (party membership) and into an educational slot (college enrollment) may be at times the same. For instance, during the Cultural Revolution, state policies that favored a certain group of individuals in their entry into the Communist Party also facilitated their selection into colleges through political recommendation. Similarly, parental resources may facilitate one’s enrollment into college, which in turn increases his or her chances to acquire party membership. In other words, political status and human capital may not be necessarily incongruent with each other. An important ﬁnding in Walder’s (1995a) dual-path model is that the “rationalization” of the state socialist planning led to an increasing emphasis on educational qualiﬁcations as well as to political loyalty. Second, the marked low rate of returns to education is more speciﬁc to the Chinese experience than reﬂects the principle of state socialist redistribution. In a comparative study of the redistribution patterns in the

Social Stratiﬁcation Under State Socialism

311

USSR and China, Zhou and Suhomlinova (2000) found that there were considerable variations in the roles of education in the patterns of redistribution across the two countries. Education had a much stronger effect on income and housing in the USSR than in China. The ﬁnding that education played a signiﬁcant role in the USSR under state socialism suggests that the theoretical claim that the political logic of socialist redistribution rewards political loyalty (party membership) over human capital (education) needs careful qualiﬁcations. Even in the Chinese context, returns to education also varied greatly across historical periods (Zhao and Zhou 2002). One needs to go beyond the statistical association and examine the speciﬁc processes through which human resources are allocated, in particular whether human resources are allocated through market processes or bureaucratic mechanisms (e.g., recruitment of college graduates into governmental agencies and ﬁrms), and to what extent income/housing distribution is dictated by market competition or state policies. In sum, although a distinction between political and human capital is useful in understanding mechanisms of state socialist redistribution, this distinction may have become blurred in the course of state socialist development because of complicated interactions among these selection processes. As a result, it may not be feasible to disentangle these processes empirically.

redistributive institutions in social stratiﬁcation State socialist redistribution is carried out through bureaucratic organizations. We have emphasized that these redistributive institutions played a central role in social stratiﬁcation in both creating hierarchical structure of positions and in allocating opportunities and resources to different positions. Because of this recognition, we have examined the role of work organizations in our analyses throughout this book. The ﬁndings in the previous chapters support this line of arguments (see Table 11.5). Evidence shows that, until the recent economic reform, state ﬁrms offered the best economic beneﬁts for their employees, with signiﬁcantly higher latent beneﬁts than those in the nonstate sector who were left out of the redistributive beneﬁts. Employees in state ﬁrms had higher personal income and better housing conditions. Moreover, as we noted in Chapter 8, employees in the state sector also enjoyed better latent economic beneﬁts. Indirect evidence also comes from patterns of job attainment in different types of work organizations. For example, there was a higher probability of acquiring party membership in the state sector, where

The State and Life Chances in Urban China

312

table 11.5. Summary of Findings on Returns to Organizations (the State Sector)

Beneﬁts/status Party membership Promotion in organization Job shifts Personal income Housing space/ facilities

Return to Organizational Status

Fluctuating Over Time?

Reference

Yes Yes, weak

Yes Yes

Table 6.2 Tables 6.3 and 6.4

No Yes Yes

— Yes Not clear

Table 9.1 Table 7.1 Tables 8.3 and 8.4

favorable positions (cadre) and better human capital (education) also tended to be concentrated. Promotions along the bureaucratic ladders also favored those employees in state-owned organizations. These ﬁndings are consistent with the patterns revealed in other studies (Bian 1994; Lin and Bian 1991; Walder 1986, 1992). The role of work organizations in China also highlights the importance of institutional arrangements in shaping particular stratiﬁcation processes and generating cross-cultural variations in the social stratiﬁcation system. In the USSR, almost all work organizations (including farms) were state owned; work organizations with varying property rights like those in China were virtually nonexistent. As a consequence, economic sectors such as type of industries played a more important role. In contrast, the less developed state socialism in China gave rise to a variety of organizational forms, thereby providing differential bases of redistribution and stratiﬁcation.

stratiﬁcation dynamics under state socialism: theoretical implications A central theme of this book is our emphasis on the stratiﬁcation dynamics to explain patterns and consequences of social stratiﬁcation processes in urban China. As we have summarized in the preceding discussions, although these structural categories (e.g., locations in occupational and organizational hierarchies) are meaningful, the effects of these structural locations varied considerably over historical periods, in accordance with shifting state policies. The ﬁndings in substantive areas of stratiﬁcation reported in the previous chapters – educational attainment, job attainment, the distribution of economic beneﬁts, bureaucratic promotions – all

Social Stratiﬁcation Under State Socialism

313

show that attention to the stratiﬁcation dynamics is critical to explain these observed patterns. For example, as revealed in Chapter 4, there have been signiﬁcant variations in both opportunities and selection criteria in urban China in the last ﬁve decades. Similarly, parental political and human resources also have had time-varying effects on one’s educational attainment. More important, all social groups were affected by macro-political and economic processes induced by shifting state policies, as demonstrated by patterns of job attainment (Chapters 4 and 5) and promotion (Chapter 6). The emphasis on stratiﬁcation dynamics makes several theoretical contributions to our understanding of state socialist redistribution and social stratiﬁcation processes. First of all, it sheds lights on the dynamic processes of social stratiﬁcation under state socialism that were neglected in the literature. Students of state socialism have long recognized the importance of macro-political processes and state policies in these societies. But the study of social stratiﬁcation in this area has been heavily inﬂuenced by the scholarship in the studies of industrialized market societies that emphasize stable, family-based structures of social mobility. As a result, the ﬂuctuating processes and their implications have not been carefully theorized or empirically studied in the Chinese context. In developing theoretical arguments on stratiﬁcation dynamics in this book, we insist that stratiﬁcation dynamics cannot be dismissed as transient and local disequilibria. Rather, they are rooted in the very institutional foundations of redistribution and macropolitical processes, and they have lasting impacts on individuals over their life course. Stratiﬁcation dynamics is an integral part of the social stratiﬁcation patterns in those major areas of our study. Moreover, as we have argued before, these stratiﬁcation dynamics are often secondary consequences of state policies in the pursuit of political and economic goals, rather than results from conscious designs of social stratiﬁcation processes on stable group bases. For example, the send-down episode in the Cultural Revolution was largely a result of the top leaders’ response to unemployment pressures and stagnant urban economies, not because of conscious efforts in social stratiﬁcation and mobility. Similarly, the Great-Leap-Forward disaster of the early 1960s had severe consequences on individual life chances and economic well-being with shrinking educational and job opportunities and signiﬁcant decline of living standards. But, these effects were indirectly caused by the top leaders’ pursuit of rushed growth. As Li and Zhang (2000) demonstrated, the development of the employment system in state socialist China was

314

The State and Life Chances in Urban China

largely an organizational response to economic crises and unemployment pressures in the past ﬁve decades. An important implication of this recognition is that, because the stratiﬁcation dynamics are triggered and channeled by other macro-political processes, social groups are ill-equipped to resist or alter these dynamics, at least in the short run. For example, when the state imposed the senddown policy on the urban residents, neither cadres, nor professionals, nor any other social groups could effectively protect their children from the adverse effects of the policy. Therefore, to explain patterns of social stratiﬁcation in state socialist societies, we need to go beyond the immediate stratiﬁcation realms and examine the interplay between broader channels of redistribution and political processes. Finally, considerations of the stratiﬁcation dynamics in the Chinese context point to several implications of the political logic of redistribution: 1. The political logic of redistribution imposes a limitation of social inequality. In industrialized market societies, welfare states have a tendency to stabilize transitions across stages in the life course and redistribute life chances so as to ensure the legitimacy of the state and governability. In state socialist political logic, the state adopted policies to suppress economic inequality so as to pursue its egalitarian ideology and rushed growth. Overall, economic inequality, as measured by income and housing inequality, as well as by patterns of status attainment, was less salient than it was in capitalist societies. These macro-political processes sustained an egalitarian policy throughout the Mao era (Whyte and Parish 1984). 2. The political logic of redistribution also imposes a limitation on the bureaucratic power. The top-down macropolitical processes impose severe institutional limitations on processes of resource transfer or status inheritance across generations. As a result, it is difﬁcult to establish and maintain stable class boundaries. Our ﬁndings lead us to draw two main conclusions on the bureaucratic class thesis. First, we question the stability of the bureaucratic class, especially with regard to class reproduction. The impacts of shifting state policies and varying patterns of bureaucratic promotions imply that the bureaucratic class is unable to protect itself. The evidence of class reproduction through intergenerational mobility is weak at best, at least for the majority of the cadres (and professionals) at the lower levels of the Chinese bureaucracy. Second, given the impacts of shifting state policies as embodied in varying cohort effects

Social Stratiﬁcation Under State Socialism

315

in party membership recruitment and bureaucratic promotion, we also question the uniformity of the bureaucratic class over time. It is likely that the bureaucrats are highly differentiated with divided loyalties toward distinctive state policies across historical periods. Moreover, as we have discussed in the ﬁrst part of this chapter, there were strong institutional constraints on the scope of authority for administrators in the ﬁrm or a work unit such that the power he or she could exercise was rather limited under the traditional redistributive system.

further issues Profound political, economic, and social changes in China in the second half of the twentieth century drew worldwide attention. State socialist China was once acclaimed as a model for developing countries struggling for economic growth and social equality, with good reasons: In a short period of time since 1949, China experienced rapid economic growth and industrialization. In this process, millions of Chinese came out of poverty, were employed in the workforce, and began to enjoy basic living standards and health care. The marked low social inequality in the Mao era was impressive and had great appeal for those who advocated social engineering. But one needs to keep in mind that the low social inequality in China was achieved through considerable cost. As revealed in this book, the low social disparity in opportunities and in the distribution of economic beneﬁts was often generated as a consequence of overall shrinking of opportunities and declining economic beneﬁts for the entire population. It is interesting to observe that the periods of great achievement of equalizing opportunities (e.g., in access to education and in income) also coincided with those of economic and political crises, such as the aftermath of the Great-Leap-Forward period and of the Cultural Revolution. Moreover, the stratiﬁcation dynamics that played an important role in weakening and, at times, breaking up permanent boundaries of social groups also created highly ﬂuctuating political and social environments for individuals in all social groups and interrupted individuals’ life courses such as the timing of education, marriage, child rearing, and living conditions. The low disparity in economic beneﬁts may also be a major factor that contributed to the stagnation of state socialist economy in the later years. Although nonmonetary rewards (e.g., political awards) played an important role in motivating people’s involvement in production and political

316

The State and Life Chances in Urban China

activities, the marked low incentives for productivity contributed to the deep crises in state socialist economy. To appreciate this point, one needs only to look at the Chinese agriculture in the early reform era, to see the enormous energies and remarkable improvement in productivity released from the Chinese peasants, once the traditional collectivist communes were discarded. How do we evaluate these different and often inconsistent aspects of the Chinese experience under state socialism? Burawoy’s cautionary observation quoted at the beginning of this chapter offers a somber reminder that, although state socialist redistribution as an institutional phenomenon has receded into history, both theoretical and empirical issues regarding state socialism and its social stratiﬁcation processes remain to be fully understood.

12 Stratiﬁcation Dynamics and Institutional Transformation

Bad fortune will promote the good; good fortunate, too, gives rise to the bad. Lao Zi

In this book, we have examined the forty-ﬁve-year history of redistribution and social stratiﬁcation processes in the People’s Republic of China, from 1949 to 1994. We have emphasized the profound impacts of the state and state policies on individual life chances, especially in the form of stratiﬁcation dynamics. In this concluding chapter, we hope to accomplish two goals. We ﬁrst consider the implications of stratiﬁcation dynamics for the crisis of state socialism and for the societal basis of institutional transformation in the Chinese context. On this basis, our second task is to discuss the issues related to the second theme of this book – sources and directions of China’s institutional transformation in the post-Mao era. We conclude this chapter and this book with some reﬂections on the implications of the Chinese experience for understanding social stratiﬁcation and institutional changes in a comparative perspective.

rethinking state–society relationship under state socialism The demise of state socialism is one of the major historical events of the twentieth century. Interestingly, few social science theories have anticipated either the large-scale social protests in the late 1980s or the sudden collapse of the state socialist regimes in the USSR and Eastern Europe. With much hindsight, we can say that an important, missing piece to this 317

318

The State and Life Chances in Urban China

puzzle is the role of society in state socialist countries. Indeed, our emphasis on the role of the state on individual life chances does not imply that individuals and social groups are merely passive recipients of these political turbulences. On the contrary, they have been active participants of the political processes. Because of the distinctive features of the centralized authority and the political structure under state socialism, the forms and impacts of their participation differ in important ways from those in other types of societies. It is useful to reconsider state–society relationships and especially the role of society in the Chinese experience. Imageries of Society under State Socialism Social scientists have developed different conceptualizations of the role of society – ordinary citizens and social groups – in the political processes under state socialism. Several theoretical models are prevalent in the literature. In the following, we highlight those issues that are related to our theoretical concerns. Arendt (1951) delivered the earliest assault on the Soviet-style totalitarian regime. A salient characteristic of the totalitarian model is an asymmetric relationship between the state and the society. On the one hand, the party-state wielded enormous power, monopolized all resources based on an elaborate bureaucratic apparatus, and penetrated every corner of the society. All citizens were organized into the redistributive institutions – state ﬁrms in urban areas and state farms in the rural areas. Moreover, it used repressive measures to silence any dissenting voices and systematically got rid of organized resistance to its power (Brzezinski 1989). On the other hand, the society under totalitarianism was characteristic of undifferentiated masses without distinctive class bases. In China, from the metropolis to rural areas, social life has been organized by the state bureaucratic system (Parish and Whyte 1978; Whyte and Parish 1984). Accompanying the state organizational apparatus were political sanctions, political labeling, monitoring, and campaigns that systematically repressed and deterred opposition to the state (White 1989). Thus, the centralized state has effectively monopolized the resources for social mobilization and denied the legitimacy of organized interests outside its control. Since the 1970s, closer inquiries of state socialist societies led many scholars to question the imagery of a monolithic state in these societies. In the studies of Soviet politics, political scientist H. Gordon Skilling (1966) ﬁrst applied the pluralist model of interest politics to emphasize factional conﬂicts and bureaucratic interests. In China studies, scholars also

Stratiﬁcation Dynamics and Institutional Transformation

319

observed power struggles among the top leaders as well as the persistent bureaucratic interests (see Lieberthal and Lampton 1992). These studies recognized the importance of bureaucratic apparatus and bureaucratic interests that may be at odds with the top leaders and considerable deviations in policy implementation. But the role of society, again, was conspicuously absent in these conceptualizations. Walder’s (1986) study of authority relationships in Chinese organizations sheds new light on the state–society relationships in Maoist China. Based on detailed interviews with immigrants from China to Hong Kong, Walder developed systematic descriptions of everyday life in workplaces and in-depth analyses of authority relationships and mechanisms of political control at the grassroots level. Walder’s model of communist new traditionalism recognizes that it was not merely political suppressions that were used to sustain the state. Rather, elaborate authority relationships were developed to exercise political control based on clientele social relations. In particular, “the positive incentives offered for compliance” (Walder 1986, p. 6) in the workplaces tied workers to their work units and peasants to their villages. Work units functioned not only as the state apparatus of political control but also as redistributive agencies in which rewards and opportunities were linked to individuals’ political attitudes and loyalty. One consequence of this institutional structure is the prevalence of the clientelist system “in which individual members of subordinate social groups pursue their interests not by banding together for coordinated group action, but by cultivating ties based on the exchange of loyalty and advantage with individuals of higher status and power” (Walder 1986, p. 47). In this light, interests from society are organized around stable institutional entities that are controlled by the state and its bureaucratic apparatus. Differential economic beneﬁts across types of work organizations also produced fragmented interests among urban residents. Unlike the scenario of totalitarianism, there are factions, groups, and classes in state socialist society, but their interests are unorganized and fragmented outside the redistributive institutions. As a result, they can hardly act as an autonomous force to exert their political demands. The totalitarian/authoritarian model, models of plural bureaucratic interests, and the neo-traditionalist model have captured different aspects of the state–society relationships in state socialist societies. But these models are unable to provide satisfactory explanations of collective actions in state socialist societies. As Tilly (1978, p. 228) pointed out, “the general effect of sustained repression is not to build up tensions to the point of a great explosion, but to reduce the overall level of collective action.” Thus,

320

The State and Life Chances in Urban China

persistent repressions as portrayed in the totalitarian regimes would lead to inaction rather than to revolts and open protests. Similarly, fragmented and clientele-based interests articulation would not cultivate the kinds of large-scale social protests and waves of political participation by citizens as witnessed in the political history of the People’s Republic of China. To understand these collective action and social movements, we need to search for alternative explanations and to rethink state–society relationships under state socialism. An Institutional Model of Unorganized Interests and Collective Action Zhou (1993b) developed an institutional model of state–society relationships to address these issues. The central theme of this model is to explain how the institutional structure of state–society relationships in China has systematically transformed unorganized interests into collective action. The proposed institutional model recognizes the premise of the state dominance of society in the totalitarian model; it also draws on Walder’s arguments that organizational control and the “positive incentives” for compliance in the Chinese institutional arrangements have successfully prevented collective action based on organized interests. The institutional model develops theoretical arguments to identify the mechanisms embedded in the institutional arrangements of state socialism that generate collective actions based on unorganized interests. We begin with the premise that institutional structure, as embodied in stable patterns of interactions based on formal authority relationships, speciﬁes the legitimacy of forms of political participation, thus circumscribes both the solution space and the channels of political input and the nature of collective action. In studying changing forms of social movements in France, Tilly (1986) argued that the rise of a capitalist economy and the expansion of the modern state in the nineteenth century produced the “proletarianization” of society, engendered widespread discontent, and transformed local conﬂicts and revolts to the national level. Our study of redistribution of social stratiﬁcation under state socialism also point to the decisive role of the socialist state in generating unique patterns of stratiﬁcation dynamics. The socialist state claims a monopoly of the public goods and denies the legitimacy of interests at the individual level. An important consequence is that any behavior outside state control is seen as a challenge to the state. As Havel (1985) observed: “Anything which leads people to overstep their predetermined rules is regarded by

Stratiﬁcation Dynamics and Institutional Transformation

321

the system as an attack upon itself. And in this respect it is correct: every instance of such transgression is a genuine denial of the system” (p. 30). When these behaviors appear in large numbers, they constitute collective deﬁance against the state. On this basis, Zhou (1993b) developed three propositions: First, the large numbers phenomenon. The very institutional structure of state socialism that prevents organized interests has facilitated collective action based on unorganized interests. The institutional structure eliminated intermediate groups between the state and the individuals, and links each and every citizen and social group directly with the state. As a result, these individuals and social groups tend to share similar life experiences, cultivate similar grievances, adopt similar behavioral patterns, and respond to similar political environments. The institutional arrangements reduce the barriers to collective action by producing “large numbers” of individuals with similar life experiences and similar demands, thus providing the basis for social mobilization on a broader scale that cuts across conventional boundaries of groups, workplaces, and localities. Second, the institutional structure of state socialism also provides a direct link between the workplace and the state, hence inﬂuences the direction of the local demands. Under state socialism, local bureaucrats act as the agents of the state; as a result, conﬂicts between social groups and local bureaucracies must be directed to the central government for solutions. Even conﬂicts between social groups or within workplaces tend to be directed toward the political center. Therefore, once the opportunity is given, large numbers of discontented individuals tend to adopt similar behaviors that converge in the same direction – toward the state. These instances of discontents may not be based on common interests, nor are they necessarily consistent with each other, but they often take a “collective” form because of their similar patterns and targets. Third, opportunities for collective action are embedded in this distinctive state–society relationship. The use of political campaigns and mass mobilizations by the state to deal with its bureaucratic problems and in response to economic crises enables individuals to articulate their interests through their responses to changes in state policies. Therefore, state policy shifts have often unwittingly induced spontaneous individual behaviors across workplaces and localities at the same time, leading to the outbreak of collective action. Another aspect of the collection action based on unorganized interests is the critical role of “collective inaction” – spontaneous behaviors in passive resistance, evasion of public duties, and so forth. When the state

322

The State and Life Chances in Urban China

policies exert repressive measures or take away incentives in production processes (e.g., collectivization of farm or private enterprises), individuals could not openly express their grievances; instead, they adopted “weapons of the weak” such as passive resistance or noncooperation to resist adverse policies (Scott 1985). Because individuals of all walks are likely to adopt such behaviors in a similar manner, these behaviors acquire a collective character. As a result, spontaneous individual behaviors based on unorganized interests often generate enormous political pressures forcing the government to abandon unpopular policies and to change the course of action. Therefore, the institutional links impose structure and organization on these otherwise unorganized interests, allowing them to “act together” and converge into collective action or collective inaction.

Stratiﬁcation Dynamics and State–Society Relationship: Further Considerations The institutional model described previously primarily focuses on the institutional arrangements and their consequences for generating political opportunities. We now discuss how this model can be enriched by considering the implications of the stratiﬁcation dynamics revealed in this book. Stratiﬁcation Dynamics and the Large-Numbers Phenomenon. At the core of the “large-numbers” phenomenon is that state–society relationships induce similar behavioral patterns among a large number of individuals of all walks. Here we see an immediate link with the stratiﬁcation dynamics discussed in this book. First, one important consequence of the stratiﬁcation dynamics revealed in the previous chapters is that they have produced ﬂuid group boundaries and undermined the vested interests in the current political system. Frequent political campaigns led to the dramatic “reversal of fortunes” among individuals and social groups. As we pointed out before, political activists in a certain period may become victims of political purge in another period once state policies shifted. Frequent political campaigns prior to and during the Cultural Revolution also led to dramatic changes in individual life chances. In other words, both the political status and the membership associated with these positions have varied greatly over time, making group boundaries more permeable than in a stable society.

Stratiﬁcation Dynamics and Institutional Transformation

323

Second, the ﬂuidity of group boundaries is further reinforced by the fact that all social groups tend to have similar life experiences, even when the relative status of their social positions remains stable. The implementation of state policies often leads to dramatic changes in the political and economic environments whose effects transcend group boundaries. For instance, during the Cultural Revolution, almost all bureaucrats were thrown out of their ofﬁces and many were sent to the rural areas. Their family members and children suffered in a similar manner, as the senddown episode (Chapter 5) demonstrated. The Great-Leap-Forward Campaign and the ensuing economic disaster led to the extreme economic conditions that were imposed on all social groups. Therefore, the stratiﬁcation dynamics provide an important mechanism to cultivate the “large numbers” phenomenon by linking different social groups together through their similar life experiences and by inducing similar behavioral patterns among them. Stratiﬁcation Dynamics and Opportunities of Political Input from Society. As we noted earlier, stratiﬁcation dynamics are often secondary consequences of changes in policy priorities and in political conﬂicts at the top. Once the state adopts certain policies, citizens of all walks must respond, thereby providing political opportunities for social groups to input their views into the political process. Moreover, the severe consequences of the stratiﬁcation dynamics also induce individuals to respond simultaneously and in similar behavioral patterns. These responses are by no means just “following the party line.” More often than not, they tended to capitalize on opportunities to pursue or protect their interests. As a result, stratiﬁcation dynamics often provide opportunities for the otherwise unorganized interests to articulate their interests through spontaneous behaviors in response to policy oscillation. Fore example, in 1987, when the government launched the antibourgeois campaign to repress student demonstrations, individuals sensed the impending political repression and policy shifts. In response, peasants cut the trees they were growing and slaughtered pigs they were raising in fear that they would be conﬁscated by the government. Enterprises withheld production and investment because of the possibility that the autonomy granted to them by the government might be revoked (Zhengming March 1987, pp. 19–21). Although similar responses may occur in a market economy, the difference is that the Chinese were responding to political uncertainty rather than to market uncertainty, and their behaviors were directly translated into political pressure. A speech by then Acting General Secretary Zhao Ziyang (Renmin Ribao,

324

The State and Life Chances in Urban China

October 10, 1987) clearly reﬂected such political pressures upon the government: If the current [reform] policy is interpreted as the result of bourgeois liberalism, it will create great uncertainty among the people. If production is unattended, commercial activities disrupted, forest destroyed and pigs killed, who can shoulder all these responsibilities?

Stratiﬁcation Dynamics and Collective Inaction. Collective action is not the only form of political input from the society. Frequent and conﬂicting policy shifts, drastic changes in one’s life chances, and inability of individuals or groups to shelter themselves from the adverse effects of state policies inevitably cultivate corresponding individual responses: disillusionment, cynicism, and passive inaction in resistance to state-initiated mobilization. Here, past experiences in stratiﬁcation dynamics – opportunities expanded and shrank, life chances gained and lost, careers advanced and disrupted – cultivated high sensitivities toward state policy shifts. Individuals were likely to adopt behaviors in the form of collective inaction especially when they experienced political pressures and the tightening of political (or regulatory) control by the state. Therefore, changes in state policies provided the opportunity for society to respond and to articulate their behaviors. In so doing, their behaviors also became an important input into the political processes, inducing future changes in state policies. In this light, state policy shifts, hence the stratiﬁcation dynamics, are not merely the state imposing its will on the society; rather, it is the interplay between the state and society that shapes the paths of political processes and perpetuates the observed stratiﬁcation dynamics.

the social basis of institutional transformation in china Schurmann (1968) argued that “organization and ideology” are two main pillars of Communist rule. Over time, however, the stratiﬁcation dynamics have gradually weakened the organizational basis of redistribution and led to ideological crises, bred popular disillusionment with socialist idealism and with the redistributive state, and cultivated political cynicism and withdrawal from the political process, as embodied in the refusal to participate in state-initiated political campaigns. It was these undercurrents that subtly but surely undermined the capacity of the socialist state, led to the breakthrough in rural economic reform in the late 1970s, and

Stratiﬁcation Dynamics and Institutional Transformation

325

set the stage for the great institutional transformation in China since the 1980s. Erosion of the Organizational Basis of Redistribution Organizational failures in the political institutions and the command economy were intertwined with the dynamics of state–society relationships, as in the form of stratiﬁcation dynamics. First, stratiﬁcation dynamics undermined the organizational basis of redistribution. The variety of property rights in Chinese organizations and the presence of organizational hierarchies have played important roles in providing legitimacy to the central authority by offering positive incentives for selected social groups (Walder 1986). However, the stratiﬁcation dynamics have greatly undermined these institutional bases by subjecting all social groups wherever located in the social structure to similar ﬂuctuating political environments and similar life experiences. As a result, preferential treatments based on redistributions were often much less effective in life chances in comparison with ﬂuctuations in state policies, providing the basis for the largenumbers phenomenon. Second, the presence of a variety of organizational forms in their property rights relationship to the state implies that the organizational control of the state has never been uniform. Rather, there were signiﬁcant variations among urban residents in their manifest and latent beneﬁts, as we examined in the previous chapters. In particular, the limited capacity in redistribution by the Chinese state led to a signiﬁcant presence of the collective sector that has never been an integral part of the central planning. It is not surprising that, as China embarked on the course of institutional transformation, it was those disadvantageous organizations in the nonstate and semistate sectors that had strong incentives to break away from the old command economy and to open up new avenues for growth. Even within the traditional state sector, redistributive mechanisms could no longer provide stable institutional basis for political control. The drastic episode of Cultural Revolution left a vast vacuum in China’s political arena. Toward the end of 1970s, at the initial stage of the economic reform, everyone in urban China appeared to be in a process of relocating himself or herself: “old” cadres were returning to their previous positions after the political purge of the Cultural Revolution; the sent-down youth were rushing back to their home cities in search for jobs; more than half a million rehabilitated “rightists” were relocating their sociopolitical

326

The State and Life Chances in Urban China

status; “rebellious” leaders in the Cultural Revolution were being forced to leave their positions in humiliation. The stratiﬁcation dynamics over the years have succeeded in breaking up any permanent group boundaries of the privileged with vested interests. In contrast to the reforms in the USSR and East European societies in the late 1980s, in China there were virtually no identiﬁable social groups, nor an organized bureaucratic class, to oppose the reform policies initiated by the post-Mao leadership. In retrospect, the major breakthrough in China’s economic transformation was the shift of state policies that opened the private sector in the early 1980s. Since the collectivization and nationalization period of the 1950s, governmental policies had been oriented to constrain and eliminate the private sector. There were many factors that contributed to this fundamental policy shift in the early 1980s. But the most immediate, direct cause is easy to identify: this new policy initiative resulted from organizational failures in redistribution; it was an organizational response to the mounting pressures of unemployment and relocation of various social groups in the aftermath of the Cultural Revolution. In light of our discussions, we can clearly see the role of stratiﬁcation dynamics in initiating this breakthrough. The triggering event was the wave of the return of the sent-down youth in 1979. As discussed in Chapter 5, in 1979 the post-Mao leaders abandoned the send-down policy and allowed all sent-down youth to return to urban areas. When millions of the sent-down youth rushed back to their home cities, what awaited them were stagnant urban economies and scarce job opportunities after a decade of political turmoil, with millions more recent high school graduates waiting for job assignments. In July 1979, the head of the State Labor Bureau reported that, by the end of that year, the urban economy needed to create jobs for 7.6 million sent-down youth alone. During this period, the returned youth organized numerous demonstrations in various urban areas demanding the government to provide jobs. For the Chinese government who had long assumed the responsibility of full employment and job assignments, the return of the sent-down youth generated enormous political pressures and led to the major shift in state labor policies. By 1979, the mounting pressure for job creation forced the government to allow and encourage employment in nonstate sectors. In July 1979, the central government issued a directive to instruct local governments to “open all channels of employment” to create jobs (Renmin Ribao July 24, 1979). In 1980, the central government adopted a new employment guideline that instituted the fundamental shift. The new

Stratiﬁcation Dynamics and Institutional Transformation

327

guideline stated that, “under the planning and guidance of the state,” the government allows three parallel employment channels: (1) employment allocated by the governmental administration, (2) job opportunities created by collective organizations on a voluntary basis, and (3) selfemployment. The last category speciﬁcally refers to the private sector. This was the ﬁrst time since the 1950s that the socialist state ofﬁcially legitimized and encouraged the existence and expansion of a private sector. In October 1981, the central government announced its most comprehensive directive on dealing with unemployment problems in the urban areas. The directive observed that, “at the present, the private economy has an initial recovery and development, but its current size is still not in accordance with the need of the national economic development.” The directive instructed local government administrations to develop a broad recognition that an urban private economy plays an important role to “increase production, revitalize the economy, meet the need of customers, and expand employment opportunities.” It pointed out “the need to reduce the weariness among the cadres, the masses, and the wait-for-job youth” in their attitudes toward the development of a private economy in the urban areas (Bureau of Policies and Regulations 1990). This began an era of the great expansion of the private sector that has now become the most dynamic and fast growing economic force in China’s economy and institutional change. Varied life experiences also cultivated attitudes toward risk taking and opportunistic behaviors that helped the rise of the private sector outside the state sector. A large proportion of those in the private sector in the 1980s were the unemployed or those who were left out of the redistributive system (idle residents, etc.), whose tea drink sales, street vendors, and long-distance arbitrage fueled the emergence of market economies. The very organizational failures in redistribution and the stratiﬁcation dynamics weakened the grips of the planning economy and paved the way for China’s economic transformation since the 1980s. Life Experiences and Ideological Crises Stratiﬁcation dynamics have also cultivated political disillusions among the Chinese and led to ideological crises in the post-Mao era (Gold 1985). For a long time, mass mobilization provided an important means for the socialist state to transfer resources and to exercise political control. In this light, the ideological crises have had signiﬁcant ramiﬁcations. First, insofar as the Chinese state relied on political monitoring mechanisms,

328

The State and Life Chances in Urban China

the ideological crises greatly weakened its capacity in political control. Second, political disillusionment and the corresponding collective inaction greatly constrained the organizing capacity of the state to pursue its political and economic goals. Finally, over time the ideological crises and disillusionment with the state socialist system have encouraged the ﬂow of human capital from the state sector to the nonstate sector. The kind of stratiﬁcation dynamics we discussed in this book has played a critical role in this process. Large-scale social changes inevitably shape and alter individuals’ life course, producing lasting effects on their perceptions and behaviors (Alwin, Cohen, and Newcomb 1991; Elder and Clipp 1988; McAdam 1989). Take the sent-down experience as an example. The historical impacts and personal reﬂections varied greatly among those who suffered during this period. As one sent-down youth commented: “I want to thank life. Without the 10-year hardship of the sentdown life, our generation probably could never have been as mature as we are today.” Others were bitter, as one woman put it: “The lives of our generation were totally wasted. When we were in school, our learning was interrupted. We had to participate in the Cultural Revolution. Later we were sent to the rural areas. When we left the city, many girls did not even understand their bodies. When we returned to the city, we all became ‘old women’” (Deng 1993, pp. 354–55). These life experiences have profoundly changed the worldviews of the new generation of the Chinese. As an editor of a book on the sent-down experience observed: If the sent-down youth were born herdsmen, peasants, or rural workers, they might not have keenly felt the hardship and misfortune in the rural areas. But, on the contrary, they had so-called education and civil experience. They went to those primitive areas stricken with poverty. Many of them left cities with wounded hearts and with personal or family losses. As a result, they felt much more deeply the meaning of hardship and misery, the roughness and ﬂuctuation in life, and the coldness of youth. (Li 1993, pp. 4–5)

Stratiﬁcation dynamics and varying life chances in China’s political history have had profound impacts on individuals who went through these life experiences. An ordinary family letter written in 1967, the early phase of the Cultural Revolution, told the following story: In an effort to join the youth league, an important political status for the Chinese youth in the Mao era, a high school student actively participated in the “Red Guard rebellions” and led the red guards to her relative’s family in search of “counter-revolutionary” evidence. But she still failed to join the youth league. “She tried to act as a political activist, but in the end she

Stratiﬁcation Dynamics and Institutional Transformation

329

inﬂicted suffering on herself as well as on others. Because her father was also purged, everyone now looked at her and her siblings in a different way” (Xu 2000, pp. 29–31). In a highly ﬂuctuating political environment, one’s life chances often changed overnight, thus leading to political disillusionment and passive resistance to the political system. As we noted earlier, an important consequence of the ideological crises is that they undermined another weapon of the state – mass mobilization. Several political campaigns in the 1980s – the “anti-spiritual pollution” campaign, the “anti-westernization” campaign, political crackdown in the aftermath of the 1989 student protests – all fell through because of the resistance from lower-level bureaucrats in the workplaces and the refusal of participation by the ordinary workers, college professors, and even the party cadres. This largely contributed to the infrequencies of resorting to political campaigns in the post-Mao era, as we documented in Chapter 1 (cf. Table 1.1). This, then, is the irony of the institutional logic of redistribution. The very strong presence of the state and state policies in individual life chances and the suffering that the stratiﬁcation dynamics inﬂicted on its people led to those organizational failures and ideological crises that greatly constrained the state’s mobilization capacities, forcing it to resort to reform, to adopt new policies in response to economic and political crises, thus setting the stage for the great transformation in China in the last twenty some years. In this light, I (1992b, p. ii) observed: The state has repressed the society and the intellectuals have ignored society. But this does not mean that the role of ordinary citizens is trivial on China’s political stage. On the contrary, at the opposite side of the state, the society has never ceased to resist the constraints and repressions imposed by the state. If we reﬂect on the forty-year history of the People’s Republic of China, we can discover that, despite the fact that the top leaders used “mass mobilizations” to achieve their political goals, the outcomes of these mobilizations often ran counter to the top leaders’ intended goals: ordinary citizens often took actions that exceeded the boundaries designated by state policies, or they took actions that were not in accordance to the ofﬁcial plan, or they became immobilized and refused to participate. Consider those instances in the Great-Leap-Forward period, the Hundred-Flower episode, the Anti-Rightist campaign, the Cultural Revolution, the April 5th episode, student demonstrations and other forms of passive resistance, the ideological crises, purchasing behaviors, instances of resistance to governmental procurements, evasions of family planning, strategies of exit to foreign countries, etc. In all these instances, the Chinese society has been stubbornly expressing its own will, in its distinctive, even distorted ways. These conscious and unconscious collective behaviors interrupted formal political processes and accelerated the tensions among the top leaders. They have played an important role in undermining the Party’s

330

The State and Life Chances in Urban China

governing basis, inducing changes in the state apparatus, and in triggering the instability of state policies . . . In this sense, the Chinese society is, in its own ways, shaping the characteristics of the Chinese politics.

from social stratiﬁcation to institutional transformation In The Contentious French, Charles Tilly (1986: 9) observed: “We shall know that a new era has begun not when a new elite holds power or a new constitution appears, but when ordinary people begin contending for their interests in new ways.” Given that social stratiﬁcation processes reﬂect fundamental institutional arrangements of a society, we can make an analogous claim. That is, we should know that a new era has begun when patterns of social stratiﬁcation begin to show signiﬁcant changes from its past. The second theme of this book was to assess the extent and direction of institutional transformation in urban China in the last two decades by contrasting changes in social stratiﬁcation patterns in the pre-reform and reform era. We designed several empirical analyses to examine changes in social stratiﬁcation patterns in several areas such as income (Chapter 7), housing (Chapter 8), job-shift patterns (Chapter 9), and cohort differences in labor force experience (Chapter 10). Our analyses of stratiﬁcation dynamics in the earlier chapters (Chapters 3–6) also provide evidence on changes and continuity of social stratiﬁcation processes over time. In the following, we summarize the main ﬁndings in this regard. The markedly low inequality in the Mao era, as revealed in this book, raised questions about the baselines for understanding the transformation of state socialism in China and the theoretical issues in the recent studies of the transformation of state socialism in the literature. For example, Nee’s market transition theory is based on the premise that there was substantive social inequality under state socialism and the introduction of market mechanisms would alter these patterns. A main starting point of his theoretical model is an imagery of the bureaucratic class based on redistribution. The image of privileges associated with cadre status is often used as the baseline for comparison of changes under institutional transformations. If, as we have demonstrated throughout this book, social inequality across occupational groups was in fact very limited in the pre-reform era, then it is more likely that the loosening of state control and the weakening of redistributive institutions would increase social inequality and enhance returns to education as well as to authority positions.

Stratiﬁcation Dynamics and Institutional Transformation

331

Indeed, there were salient, signiﬁcant changes in social stratiﬁcation processes in several major areas in the reform era. The most important aspect is the signiﬁcant presence of private ﬁrms, foreign ﬁrms, and hybrid ﬁrms in China’s economy. They are now playing an increasingly important role in resource allocation and are reshaping the distribution of economic beneﬁts among social groups. By 1994, those working in private ﬁrms earn 75 percent more than those working in collective ﬁrms and also signiﬁcantly higher than those working in the state sector (see Table 7.1), indicating markedly new patterns resulting from emerging markets. Another noticeable evidence is that, in the reform era, college graduates, especially those in the recent cohort, have a higher probability of moving to the private sector than before (see Chapter 9), again indicating the increasing strength of competition from the emerging market institutions for human resources that undermines the traditional state redistributive institutions. Similarly, we observed that returns to education have increased signiﬁcantly in recent years. All these changes signaled signiﬁcant changes in social stratiﬁcation orders in urban China. A major force that has driven these changes was the emerging market economies and market mechanisms in resource allocation, as Nee’s market transition theory has emphasized. On the other hand, judging from the patterns revealed in our study, institutional changes were both more limited and uneven by 1994 – the last year of our observation. Let us carefully qualify and unpack this statement. The rapid economic growth in China in the last decade has been remarkable and historically unprecedented. Relative to these changes in the economic arena, we ﬁnd both changes and continuity in social stratiﬁcation processes. For example, compared with noticeable changes in patterns of income determinants, changes in housing allocation by 1994 were limited and more recent changes in housing allocation appeared to reinforce the existing social inequality rather than altering it (see our discussions in Chapter 8). By 1994, job-shift patterns were still dominated by movements into or within the state sector. Advantages enjoyed by cadres and work organizations in the state sector have not signiﬁcantly changed, based on the series of empirical assessments we reported in this book. These ﬁndings and our interpretations by no means imply that there were no signiﬁcant changes in urban China in the last two decades. On the contrary, changes have been both fundamental and widespread. Economic transformations since the 1980s have had profound impacts on China’s economy and redistributive institutions. But it is important to

332

The State and Life Chances in Urban China

recognize that the state and its bureaucratic apparatus still signiﬁcantly shape institutional changes, and hence social stratiﬁcation patterns. Take job-shift patterns as an example. Although many aspects of the state sector have changed, redistributive beneﬁts (e.g., housing, welfare programs) are still substantial enough to motivate employees to remain in the state sector. Moreover, state organizations, even when actively engaged in market activities, are both more protected by central and local authorities (through bank loans, resource allocation) and more regulated by governmental policies (e.g., in hiring, layoffs, distribution of welfare funds) than are nonstate organizations. That is, the redistributive power and beneﬁts associated with political authorities are still inﬂuential, and jobshift patterns in the reform era continue to be affected by state policies and by bureaucratic interventions. Therefore, the state continues to play a critical role in shaping the direction of institutional change in China’s transitional economy. Furthermore, along with the weakening of central planning, the role of local governments has become increasingly important. As Walder (1995) argued, institutional arrangements in China involve more direct administrative intervention, in a manner similar to “corporate hierarchies,” by local governments who have both ﬁnancial and nonﬁnancial interests in ﬁrms in their jurisdiction (see also Oi 1989, 1999). As a result, the observed institutional changes were multifaceted, as were the processes that generated them. For instance, increasing returns to education in the reform era clearly reﬂect the increasing importance of human capital in market transactions. But it is also consistent with state policies in personnel management in the reform era. Similarly, even when we observe similar beneﬁts associated with governmental agencies and positional power of cadres, the sources of these beneﬁts may be partly related to market transactions, as seldom does any work organization or individual in the public sector nowadays solely depend on redistribution from the state. In this sense, although the observed patterns of income determinants may appear to have not changed in signiﬁcance over time, the processes that generate the apparent continuity may have changed substantively. Our empirical studies reported in this book ended in 1994 – the last year of our data collection. After Deng Xiaoping’s advocacy in 1992, major changes in China took place at a scale and in a scope that were much more far-reaching than before. Here, we brieﬂy summarize the major changes since the mid-1990s and discuss their implications along with the major ﬁndings in this book. The most signiﬁcant changes took place within the

Stratiﬁcation Dynamics and Institutional Transformation

333

state sector, with the privatization of a large number of small- and midsized state ﬁrms. A large number of employees in state ﬁrms and other state organizations were pushed out of permanent employment. At the same time, the welfare system, such as health care and retirement beneﬁts, has withered considerably. As a result, there is a considerable further weakening of the traditional state sector, along with the advancement of market mechanisms in economic transactions. On the other hand, the emerging patterns of social inequality are still heavily marked by the active role of the state and the state sector. In particular, the reform era witnessed a series of bureaucratic reforms in the Chinese bureaucracy. As we reported in the earlier chapters, the recruitment and promotion of bureaucrats became more merit based, and the career lines are more institutionalized and stabilized in the reform era. Since the mid-1990s, there have been considerable improvements in salaries and fringe beneﬁts as well as professionalization for those cadres and professionals in the Chinese bureaucracy. What are the implications of these emerging institutional changes? In a comparative study of state structures and economic growth, Evans and Rauch (1999) argued and showed that those countries whose state bureaucracies exhibited “Weberian” characteristics – merit-based selection and stable careers – also experience faster economic growth. In this light, the Chinese state and its bureaucratic organizations have provided a major source for the rapid economic growth in China since the 1980s. The emergence of new institutions often takes the form of a recombination of existing routines, authority relationships, and available solutions – such as networks, local corporatism, clientele relationship – within the current institutional arrangements (Lin 1995; Oi 1999; Stark 1996; Walder 1995b).

beyond china and beyond state socialism The forty-ﬁve-year episode we studied in this book is but a ripple in the long Chinese history. But history is embodied in individual life experiences and activities through time. Through the lens of social stratiﬁcation patterns, we observed many different, often contradicting, images of state socialist China. On the one hand, there were impressive improvements in different aspects of the living standards for the Chinese people and equalizing opportunities among social groups in this period; on the other hand, we witnessed the ups and downs of individual career trajectories, the losses and regains of opportunities, the break-ups and reunions of

334

The State and Life Chances in Urban China

families, and the political purges and rehabilitations of cadres and intellectuals. After going through these zigzag experiences in the last ﬁve decades, China is now embarking on a path of societal transformation that has led to profound and still ongoing institutional changes. At this historical conjecture, it is natural to ask: What is the implication of this episode for China’s future? What is the historical signiﬁcance of the Chinese experience for other parts of the world? An immediate implication is that the stratiﬁcation dynamics and life experiences are likely to affect the behavioral patterns of the Chinese people. For example, as the children of the Cultural Revolution now become the main component of the labor force in urban China, as many of these cohorts are now well-known writers, intellectuals, or currently occupying leadership positions at different levels of the government and other types of work organizations, their life experiences are inevitably reﬂected in state policies, managerial practice, and individual behaviors. The original proposal for rural reform was developed by a group of young scholars who had been sent to the rural areas during the Cultural Revolution. The adoption of their proposal into state policies initiated the highly successful rural reform in the late 1970s that eventually led to the urban reform in the mid-1980s. The disrupted life course and harsh life experience in the rural areas may have also led to the disillusion about the state socialist political system and fostered a more materialistic orientation that better ﬁt a market economy than the socialist political order. The varied life experiences of the Chinese people are likely to cultivate a more pragmatic attitude toward the future development in China. Moreover, the historical legacies also provide a rich repertoire from which individuals draw strategies and behavioral responses in the new era (Lee 1998; Shi 1997; Tang and Parish 2000). On the other hand, as China has moved away from the Soviet-style state socialist model and embarked on a new course of development, its brief ﬂirtation with utopian socialism is likely to exert long-lasting impacts. Ironically, perhaps the more subtle but profound legacy of the state socialist experience in China is immersed in a new worldview. The past forty-ﬁve-year experience of a strong state with enormous organizational capacities and a sovereign nation – subsistent but unbowed to great powers – has infused a strong sense of national identity and a new world perspective among the Chinese. As China embarked on an uncharted territory in international arenas of trade, diplomacy, and competition for national interests, the legacy of a strong, centralized state that had effectively mobilized resources for political and economic goals will always be

Stratiﬁcation Dynamics and Institutional Transformation

335

a mental framework against which future leaders are measured. Moreover, the institutional legacies of state socialist redistribution are likely to linger on, in various forms, in new institutional forms of organizations, and in the mentalities of the Chinese. In this light, knowledge of China’s past helps us understand its present and foresee its future. In this book we took a close look at the consequences of redistribution under state socialism and the resulting stratiﬁcation dynamics in the People’s Republic of China. As we observed at the beginning of this book, although state socialism as a political institution is receding into history, redistribution of resources through centralized authorities is likely to remain and even to increase in contemporary societies. The rise of the welfare states has also enhanced the active role of political authorities in the stratiﬁcation processes in industrialized market societies (EspingAndersen 1990; Gottschalk, Gustafsson, and Palmer 1997; Mayer and Muller 1986). Even in the United States, where the role of the state was traditionally weak, the impacts of governmental policies have been both substantive and far-reaching. DiPrete and Grusky (1990), for instance, found that governmental social policies in the United States have played a signiﬁcant role in equalizing opportunities. As they pointed out (p. 109): “The most important ‘independent forces’ in advanced industrial systems may be the political policies and programs undertaken by the state itself.” In a way, the historical path of societal changes is a trend of transcending boundaries. The boundaries of villages and communities gave way to those of towns and municipals, which in turn to the modern states, and now to international capital and global economies. As the scale of transactions across societies and regions increases, the role of central authorities and the magnitude of redistribution, in one form or another, also increase. As the modern state and other forms of central authorities expand and penetrate civil life, we expect channels of resource transfer to be altered, and the effects of state policies on individual life chances to increase. In this context, the lessons from the Chinese experience on the state and life chances have implications beyond China and beyond former state socialist societies.

References

Allison, Paul D. 1982. “Discrete-Time Methods for the Analysis of Event Histories.” pp. 61–98 in Sociological Methodology, edited by S. Leinhardt. San Francisco, CA: Jossey-Bass. 1995. Survival Analysis Using the SAS System. Cary, NC: SAS Institute, Inc. Alonso, William and Paul Starr (eds.). 1987. The Politics of Numbers. New York: Russell Sage Foundation. Althauser, Robert P. 1989. “Internal Labor Market.” Annual Review of Sociology 15:143–61. Althauser, Robert P. and Arne L. Kalleberg. 1981. “Firms, Occupations, and the Structure of Labor Markets: A Conceptual Analysis and Research Agenda.” pp. 119–49 in Sociological Perspectives on Labor Markets, edited by I. Berg. New York: Academic Press. Alwin, Duana F., Ronald L. Cohen, and Theodore M. Newcomb. 1991. Political Attitudes over the Life Span. Madison: University of Wisconsin Press. Anderson, Margo J. 1988. The American Census: A Social History. New Haven: Yale University Press. Arendt, Hannah. 1951. The Origins of Totalitarianism. New York: World Publishing. Asselain, Jean-Charles. 1987. “The Distribution of Incomes in East-Central Europe.” pp. 21–62 in Equality and Inequality in Eastern Europe, edited by P. Kende and Z. Stmiska. New York: St. Martin’s Press. Atkinson, A. B., and John Micklewright. 1992. Economic Transformation in Eastern Europe and the Distribution of Income. London: Cambridge University Press. Banister, Judith. 1987. China’s Changing Population. Stanford: Stanford University Press. Barnett, A. Doak. 1969. Chinese Communist Politics in Action. Seattle: University of Washington Press. Barnett, A. Doak, and Ezra F. Vogel. 1967. Cadres, Bureaucracy, and Political Power in Communist China. New York: Columbia University Press. 337

338

References

Baron, James N. 1984. “Organizational Perspectives on Stratiﬁcation.” Annual Review of Sociology 10:37–69. Baron, James N., and William T. Bielby. 1980. “Bringing the Firm Back In: Stratiﬁcation, Segmentation, and the Organization of Work.” American Sociological Review 45:737–65. 1986. “The Proliferation of Job Titles in Organizations.” Administrative Science Quarterly 31:561–76. Bauman, Zygmunt. 1974. “Ofﬁcialdom and Class: Bases of Inequality in Socialist Society.” pp. 129–48 in The Social Analysis of Class Structure, edited by F. Parkin. London: Tavistock Publications. Becker, Gary Stanley. 1964. Human Capital: A Theoretical and Empirical Analysis, with Special Reference to Education. New York: National Bureau of Economic Research; distributed by Columbia University Press. Bendix, Reinhard. 1964. Nation-Building and Citizenship: Studies of Our Changing Social Order. New York: Wiley. Bernstein, Thomas P. 1977. Up to the Mountains and Down to the Villages: The Transfer of Youth from Urban to Rural China. New Haven: Yale University Press. Bian, Yanjie. 1994. Work and Inequality in Urban China. Albany: State University of New York Press. 1997. “Bringing Strong Ties Back In: Indirect Connection, Bridges, and Job Search in China.” American Sociological Review 62:266–85. 2002. “Chinese Social Stratiﬁcation and Social Mobility.” Annual Review of Sociology 28:91–116. Bian, Yanjie, John R. Logan, Hanlong Lu, Yunkang Pan, and Ying Guan. 1997. “‘Work Units’ and the Commodiﬁcation of Housing: Observations on the Transition to a Market Economy with Chinese Characteristics.” Social Science in China XVIII:28–35. Bian, Yanjie, and John R. Logan. 1996. “Market Transition and the Persistence of Power: The Changing Stratiﬁcation System in Urban China.” American Sociological Review 61:739–58. Bian, Yanjie, Xiaoling Shu, and John R. Logan. 2002. “Communist Party Membership and Regime Dynamics in China.” Social Forces 79:805–41. Bielby, William T., and James N. Baron. 1984. “A Woman’s Place Is with Other Women: Sex Segregation within Organizations.” pp. 27–55 in Sex Segregation in the Workplace, edited by B. F. Reskin. Washington, DC: National Academy Press. 1986. “Men and Women at Work: Sex Segregation and Statistical Discrimination.” American Journal of Sociology 91:1031–59. Blau, Francine D., and Marianne A. Ferber. 1992. The Economics of Women, Men, and Work. Englewood Cliffs, NJ: Prentice-Hall. Blau, Peter M., and Otis Dudley Duncan. 1967. The American Occupational Structure. New York: Wiley. Blau, Peter M., and Danching Ruan. 1990. “Inequality of Opportunity in Urban China and America.” Research in Social Stratiﬁcation and Mobility 9:3– 32.

References

339

Boisot, Max, and John Child. 1996. “From Fiefs to Clans and Network Capitalism: Explaining China’s Emerging Economic Order.” Administrative Science Quarterly 41:600–28. Bourdieu, Pierre, and Jean Claude Passeron. 1977. Reproduction in Education, Society and Culture. Beverly Hills: Sage. Brzezinski, Zbigniew. 1989. The Grand Failure: The Birth and Death of Communism in the Twentieth Century. New York: Collier Books. Bunce, Valerie Jane. 1980. “The Succession Connection: Policy Cycles and Political Change in the Soviet Union and Eastern Europe.” American Political Science Review 74:966–77. Burawoy, Michael. 2001. “Neoclassical Sociology: From the End of Communism to the End of Classes.” American Journal of Sociology 106:1099–120. Burawoy, Michael, and Pavel Krotov. 1992. “The Soviet Transition from Socialism to Capitalism: Worker Control and Economic Bargaining in the Wood Industry.” American Sociological Review 57:16–38. Burawoy, Michael, and Janos Lukacs. 1985. “Mythologies of Work: A Comparison of Firms in State Socialism and Advanced Capitalism.” American Sociological Review 50:723–37. Bureau of Policies and Regulations, Ministry of Labor and Bureau of Labor of Jilin Provincial Government. 1990. The Encyclopedia of Labor Policies and Regulations in the PRC. Changchun: Jilinsheng Kexue Jishu Chubanshe. Burns, John F. 1987. “China’s Nomenklatura System.” Problems of Communism 36:36–51. Cai, He. 1996. “The Role of Authority in State Firms (in Chinese) [Lun Guoyou Qiye De Quanwei Wenti].” Shehuixue Yanjiu 6:17–24. 1998. “Employee Rights Awareness and Informal Constraints across Ownership Types (in Chinese) [Cong Zhigong De Quanli Yishi Kan Feizhengshi Yuesu De Suoyouzhi Chayi Jiqi Yingxiang].” Guanli Shijie 5:191–6. Campbell, John L., J. Rogers Hollingsworth, and Leon N. Lindberg. 1991. Governance of the American Economy. New York: Cambridge University Press. Cao, Yang. 2001. “Careers inside Organizations: A Comparative Study of Promotion Determination in Reforming China.” Social Forces 80:683–712. Cao, Yang, and Victor Nee. 2000. “Comment: Controversies and Evidence in the Market Transition Debate.” American Journal of Sociology 105:1175–89. Chen, Danyan. 2000. A Woman’s Story in Shanghai (in Chinese) [Shanghai De Hongyan Yishi]. Beijing: Zuojia Chubanshe. Chen, Shaoping and Huiping Zhang. 1992. The History of the Salary System in Government and Public Organizations (in Chinese) [Guojia Jiguan He Shiye Danwei Gongzi Zhidu Biangu]. Beijing: Zhonguo Renshi Chubanshe. Chen, Zhili, and Ji Liu. 1991. The History of the Chinese Communist Party (in Chinese) [Zhongguo Gongchandong Jiansheshi]. Shanghai: Shanghai Renmin Chubanshe. Chirot, Daniel. 1978. “Social Change in Communist Romania.” Social Forces 57:457–99. Coleman, James S. 1988. “Social Capital in the Creation of Human Capital.” American Journal of Sociology 94:S94–S120.

340

References

Connor, Walter D. 1979. Socialism, Politics, and Equality: Hierarchy and Change in Eastern Europe and the USSR. New York: Columbia University Press. Davis, Deborah. 1990. “Urban Job Mobility.” in Chinese Society on the Eve of Tiananmen, Contemporary China Series, No. 7, edited by D. Davis and E. Vogel. Cambridge, MA: Council on East Asian Studies Publications, Harvard University. 1992a. “Job Mobility in Post-Mao Cities.” China Quarterly 84:1062–85. 1992b. “Skidding: Downward Mobility among Children of the Maoist Middle Class.” Modern China 18:410–37. 1995. Urban Spaces in Contemporary China: The Potential for Autonomy and Community in Post-Mao China. New York: Cambridge University Press. Deng, Xian. 1993. The Dreams of the Sent-Down Youth (in Chinese [Zhongguo Zhiqingmeng]. Beijing: Renmin Wenxue Chubanshe. Deng, Zhong, and Donald J. Treiman. 1997. “The Impact of the Cultural Revolution on Trends in Educational Attainment in the People’s Republic of China.” American Journal of Sociology 103:391–428. Diggle, Peter, Kung-Yee Liang, and Scott L. Zeger. 1994. Analysis of Longitudinal Data. Oxford, UK: Clarendon Press. DiPrete, Thomas, A. and David B. Grusky. 1990. “Structure and Trend in the Process of Stratiﬁcation for American Men and Women.” American Journal of Sociology 96:107–43. Djilas, Milovan. 1957. The New Class: An Analysis of the Communist System. New York: Praeger. Doeringer, Peter B., and Michael J. Piore. 1971. Internal Labor Markets and Manpower Analysis. Lexington, MA: Heath. Douglas, Mary. 1986. How Institutions Think. Syracuse: Syracuse University Press. Elder, Glen H. Jr. 1974. Children of the Great Depression: Social Change in Life Experience. Chicago: University of Chicago Press. 1985a. Life Course Dynamics: Trajectories and Transitions, 1968–1980. Ithaca, NY: Cornell University Press. 1985b. “Perspectives on the Life Course.” pp. 23–49 in Life Course Dynamics: Trajectories and Transitions, 1968–1980, edited by Glen H. Elder Jr. Ithaca, NY: Cornell University Press. 1995. “The Life Course Paradigm: Social Change and Individual Development.” pp. 101–39 in Examining Lives in Context, edited by P. Moen, and K. Luscher. Washington, DC: American Psychological Association. Elder, Glen H., Jr., and Elizabeth C. Clipp. 1988. “Wartime Losses and Social Bonding: Inﬂuences across 40 Years in Men’s Lives.” Psychiatry 51:177–98. Erikson, Robert, and John H. Goldthorpe. 1992. The Constant Flux: A Study of Class Mobility in Industrial Societies. Oxford: Clarendon Press. Esping-Andersen, Gøsta. 1990. The Three Worlds of Welfare Capitalism. Cambridge: Polity Press. Evans, Peter B., Dietrich Rueschemeyer, and Theda Skocpol. 1985. Bringing the States Back In. New York: Cambridge University Press.

References

341

Evans, Peter, and James E. Rauch. 1999. “Bureaucracy and Growth: A CrossNational Analysis of the Effects of ‘Weberian’ State Structures on Economic Growth.” American Sociological Review 64:748–65. Featherman, David L., and Robert Mason Hauser. 1978. Opportunity and Change. New York: Academic Press. Featherman, David L., F. Lancaster Jones, and Robert M. Hauser. 1975. “Assumptions of Mobility Research in the United States: The Case of Occupational Status.” Social Science Research 4:329–60. Feng, Denggong, and Lufeng Liu. 1992. Major Events in New China (in Chinese) [Xin Zhongguo Dashi Jiyao]. Jinan: Shandong Renmin Chubanshe. Flakierski, Henryk. 1993. Income Inequalities in the Former Soviet Union and Its Republics. Armonk, NY: M. E. Sharpe. Fligstein, Neil. 1996. “The Economic Sociology of the Transitions from Socialism.” American Journal of Sociology 101:1074–81. Freidson, Eliot. 1986. Professional Powers: A Study of the Institutionalization of Formal Knowledge. Chicago: University of Chicago Press. Ganzeboom, Harry B. G., Donald J. Treiman, and Wout C. Ultee. 1991. “Comparative Intergenerational Stratiﬁcation Research: Three Generations and Beyond.” Annual Review of Sociology 17:277–302. Gardiner, Karen. 1997. “A Survey of Income Inequality over the Last Twenty Years – How Does the UK Compare.” pp. 36–59 in Changing Patterns in the Distribution of Economic Welfare: A International Perspective, edited by P. Gottschalk, B. Gustafsson, and E. Palmer. London: Cambridge University Press. George, Linda. 1993. “Sociological Perspectives on Life Transitions.” Annual Review of Sociology 19:353–73. Gerber, Theodore P. 2000. “Membership Beneﬁts or Selection Effects? Why Former Communist Party Members Do Better in Post-Soviet Russia.” Social Science Research 29:25–50. 2002. “Structural Change and Post-Socialist Stratiﬁcation: Labor Market Transitions in Contemporary Russia.” American Sociological Review 67:629–59. Gerber, Theodore P., and Michael Hout. 1995. “Education Stratiﬁcation in Russia during the Soviet Period.” American Journal of Sociology 101:105–51. 1998. “More Shock Than Therapy: Market Transition, Employment, and Income in Russia, 1991–1995.” American Journal of Sociology 104:1–50. Giele, Janet Zollinger, and Glen H. Elder (eds.). 1998. Methods of Life Course Research: Qualitative and Quantitative Approaches. Thousand Oaks, CA: Sage. Gold, Thomas B. 1985. “After Comradeship: Personal Relations in China since the Cultural Revolution.” China Quarterly No. 104: 657–75. Gold, Thomas B. 1991. “Youth and the State.” China Quarterly No. 127: 594– 612. Goldthorpe, John H. 1987. Social Mobility and Class Structure in Modern Britain. Oxford, UK: Clarendon Press. Gottschalk, Peter, Bjorn Gustafsson, and Edward Palmer (eds.). 1997. Changing Patterns in the Distribution of Economic Welfare: An International Perspective. Cambridge, UK: Cambridge University Press.

342

References

Granovetter, Mark. 1985. “Economic Action and Social Structure: The Problem of Embeddedness.” American Journal of Sociology 91:481–510. Grusky, David B., and Robert M. Hauser. 1984. “Comparative Social Mobility Revisited: Models of Convergence and Divergence in Sixteen Countries.” American Sociological Review 49:19–38. Grusky, David B., and Jesper Sørensen. 1998. “Can Class Analysis Be Salvaged?” American Journal of Sociology 49:19–38. Guthrie, Doug. 1997. “Between Markets and Politics: Organizational Response to Reform in China.” American Journal of Sociology 102:1258–303. 1998. “The Declining Signiﬁcance of Guanxi in China’s Economic Transition.” China Quarterly 154:31–62. 1999. Dragon in a Three-Piece Suit: The Emergence of Capitalism in China. Princeton, NJ: Princeton University Press. Guthrie, Doug, and Louise Marie Roth. 1999. “The State, Courts, and Maternity Policies in U.S. Organizations: Specifying Institutional Mechanisms.” American Sociological Review 64:41–63. Hallinan, Maureen T. 1988. “Equality of Educational Opportunity.” Annual Review of Sociology 14:249–68. Hamilton, Gary G., and Nicole Woolsey Biggart. 1988. “Market, Culture, and Authority: A Comparative Analysis of Management and Organization in the Far East.” American Journal of Sociology 94:S52–S94. Hankiss, Elemer. 1989. “Demobilization, Self-Mobilization and QuasiMobilization in Hungary, 1948–1987.” Eastern European Politics and Societies 3:105–51. Hanley, Eric, and Matthew McKeever. 1997. “The Persistence of Educational Inequalities in State-Socialist Hungary: Trajectory-Maintenance versus Counterselection.” Sociology of Education 70:1–18. Hannum, Emily, and Yu Xie. 1994. “Trends in Educational Gender Inequality in China: 1949–1985.” Research in Social Stratiﬁcation and Mobility 13:73–98. Harding, Harry. 1981. Organizing China: The Problem of Bureaucracy, 1949– 1976. Stanford, CA: Stanford University Press. Hauser, Robert M., and David L. Featherman. 1977. The Process of Stratiﬁcation: Trends and Analyses. New York: Academic Press. Hauser, Robert M., and David B. Grusky. 1988. “Cross-National Variation in Occupational Distributions, Relative Mobility Chances, and Intergenerational Shifts in Occupational Distributions.” American Sociological Review 53:723–41. Havel, Vaclav. 1985. The Power of the Powerless: Citizens against the State in Central-Eastern Europe. Armonk, NY: M. E. Sharpe. Heyns, Barbara, and Ireneusz Bialecki. 1993. “Educational Inequalities in Postwar Poland.” pp. 303–35 in Persistent Inequality, edited by Y. Shavit and H. P. Blossfeld. Boulder, CO: Westview. Hodgson, Geoffrey M. 1994. “The Return of Institutional Economics.” pp. 58– 76 in The Handbook of Economic Sociology, edited by N. J. Smelser and R. Swedberg. Princeton, NJ: Princeton University Press. Hodson, Randy, and Robert L. Kaufman. 1982. “Economic Dualism: A Critical Review.” American Sociological Review 47:727–39.

References

343

Hogan, Dennis P. 1981. Transitions and Social Change: The Early Lives of American Men. New York: Academic Press. Hout, Michael. 1988. “More Universalism, Less Structural Mobility: The American Occupational Structure in the 1980s.” American Journal of Sociology 93:1358–400. Jacobs, Jerry A. 1989. Revolving Doors. Palo Alto, CA: Stanford University Press. Jermakowicz, Wladyslaw, and Jane Thompson Follis. 1988. Reform Cycles in Eastern Europe, 1944–1987. Berlin: Duncker & Humblot. Jiang, Kun, Xiaohua Li, and Xiaoyan Shi. 1996. Recollections on the Sentdown Experience (3 Volumes, in Chinese) [Zhongguo Zhiqing Huiyilu]. Cangchun: Jilinsheng Kexuejishu Chubanshe. Kahneman, Daniel, Paul Slovic, and Amos Tversky, (eds.). 1982. Judgment under Uncertainty: Heuristics and Biases. New York: Cambridge University Press. Katz, Zev. 1972. Patterns of Social Stratiﬁcation in the USSR. Cambridge, MA: MIT, Center for International Studies. Keister, Lisa. A. 1998. “Engineering Growth: Business Group Structure and Firm Performance in China’s Transition Economy.” American Journal of Sociology 104:404–40. 2001. “Exchange Structures in Transition: Lending and Trade Relations in Chinese Business Groups.” American Sociological Review 66:326–60. Kelly, Erin, and Frank Dobbin. 1999. “Civil Rights Law at Work: Sex Discrimination and the Rise of Maternity Leave Policies.” American Journal of Sociology 105:455–92. Khan, Azizur Rahman, and Carl Riskin. 2001. Inequality and Poverty in China in the Age of Globalization. New York: Oxford University Press. Kohli, Martin. 1986. “Social Organization and the Subjective Construction of the Life Course.” in Human Development and the Life Course: Multidisciplinary Perspectives, edited by A. B. Sorensen, F. E. Weinert, and L. R. Sherrod. Hillsdale, NJ: Erlbaum. ´ Gyorgy, ¨ ´ Szel´enyi. 1979. The Intellectuals on the Road to Class Konrad, and Ivan Power. New York: Harcourt Brace Jovanovich. ´ Kornai, Janos. 1959. Overcentralization in Economic Administration: A Critical Analysis Based on Experience in Hungarian Light Industry. London: Oxford University Press. 1972. Rush versus Harmonic Growth. Meditation on the Theory and on the Policies of Economic Growth. Amsterdam: North-Holland Pub. Co. 1992. The Socialist System: The Political Economy of Communism. Princeton, NJ: Princeton University Press. Kraus, Richard Curt. 1981. Class Conﬂict in Chinese Socialism. New York: Columbia University Press. Lampton, David M. 1987. Policy Implementation in Post-Mao China. Berkeley, CA: University of California Press. Lardy, Nicholas R. 1975. “Centralization and Decentralization in China’s Fiscal Management.” The China Quarterly 61:25–60. Larson, Magali Sarfatti. 1977. The Rise of Professionalism: A Sociological Analysis. Berkeley, CA: University of California Press.

344

References

Lavely, William, Zhenyu Xiao, Bohua Li, and Ronald Freedman. 1990. “The Rise in Female Education in China: National and Regional Patterns.” China Quarterly 12:61–93. Lee, Ching Kwan. 1995. “Engendering the World of Labor.” American Sociological Review 60:378–97. Lee, Ching Kwan. 1998. “The Labor Politics of Market Socialism: Collective Inaction and Class Experiences among State Workers in Guangzhou.” Modern China 24:3–33. Lee, Hong Yung. 1991. From Revolutionary Cadres to Party Technocrats in Socialist China. Berkeley, CA: University of California Press. Li, Bobai, and Andrew G. Walder. 2001. “Career Advancement as Party Patronage: Sponsored Mobility into the Chinese Administrative Elite, 1949–1996.” American Journal of Sociology 106:1371–408. Li, Cheng, and David Bachman. 1989. “Localism, Elitism, and Immobilism: Elite Formation and Social Change in Post-Mao China.” World Politics 42:64–94. Li, Guangping. 1993. The Life Experience of the Sentdown Youth (in Chinese) [Zhongguo Zhiqing Beihuanlu]. Guangzhou: Huacheng Chubanshe. Li, Lulu. 1993. “The Danwei Phenomena and Institutional Reform.” Zhongguo Shehui Kexue Jikan 16. Li, Lulu, and Hanlin Li. 2000. Danwei Organizations: Resources, Power, and Exchange (in Chinese) [Zhongguo De Danwei Zuzhi: Ziyuan, Quanli Yu Jiaohuan]. Hongzhou: Zhejiang Renmin Chubanshe. Li, Meng, Feizhou Zhou, and Kang Li. 1996. “Danwei: Internal Mechanisms in the Institutionalized Organizations (in Chinese) [Danwei: Zhiduhua Zuzhi De Neibu Jizhi].” Zhongguo Shehui Kexue Jikan 16. Li, Peilin (ed.). 1992. Chinese Enterprises in Transition (in Chinese) [Zhuanxing Zhong De Zhongguo Qiye]. Jinan: Shandong Renmin Chubanshe. Li, Peilin, and Yi Zhang. 2000. The Social Costs of State-Owned Firms (in Chinese) [Guoyou Qiye Shehui Chengben Fenxi]. Beijing: Beijing Shehui Kexue Wenxian Chubanshe. Li, Qiang. 1999. The Life Course (in Chinese) [Shengming de Licheng]. Hongzhou: Zhejiang Renmin Chubanshe. Lieberthal, Kenneth, and David M. Lampton, (eds.). 1992. Bureaucracy, Politics and Decision Making in Post-Mao China. Berkeley, CA: University of California Press. Lin, Nan. 1995. “Local Market Socialism: Local Corporatism in Action in Rural China.” Theory and Society 24:301–54. Lin, Nan, and Yanjie Bian. 1991. “Getting Ahead in Urban China.” American Journal of Sociology 97:657–88. Lindblom, Charles Edward. 1977. Politics and Markets: The World’s PoliticalEconomic Systems. New York: Basic Books. Logan, John R., Yanjie Bian, and Fuqin Bian. 1999. “Housing Inequality in Urban China in the 1990s.” International Journal of Urban and Regional Research 23:7–25. Lu, Feng. 1989. “Danwei: A Unique Form of Social Organization (in Chinese) [Danwei: Yizhong Teshu De Shehui Zuzhi Xingshi].” Chinese Social Science 1:71–88.

References

345

MacFarquhar, Roderick. 1960. The Hundred Flower Campaign and the Chinese Intellectuals. New York: Praeger. Mao, Zedong. 1938. “The Status of the CCP in the Anti-Japanese War.” pp. 519– 36 in Selected Work of Mao Zedong (Vol. 2, in Chinese) [Mao Zedong Xuanji], edited by Z. Mao. Beijing: Renmin Chubanshe. Mare, Robert D. 1981. “Change and Stability in Educational Stratiﬁcation.” American Sociological Review 46:72–87. Marshall, T. H. 1950. Citizenship and Social Class, and Other Essays. Cambridge, UK: Cambridge University Press. Mateju, Peter. 1993. “Who Won and Who Lost in a Socialist Redistribution in Czechoslovakia.” pp. 251–71 in Persistent Inequality, edited by Y. Shavit and H. Blossfeld. Boulder, CO: Westview. Matthews, Mervyn. 1978. Privilege in the Soviet Union. London: Allen & Unwin. Mayer, Karl Ulrich, and Walter Muller. 1986. “The State and the Structure of the Life Course.” pp. 217–45 in Human Development and the Life Course: Multidisciplinary Perspectives, edited by A. Sorensen, F. E. Weinert, and L. R. Sherod. Hillsdale, NJ: Erlbaum. Mayer, Karl Ulrich, and Urs Schoepﬂin. 1989. “The State and the Life Course.” Annual Review of Sociology 15:187–209. Mayer, Karl Ulrich, and Nancy Brandon Tuma. 1990. Event History Analysis in Life Course Research. Madison, WI: University of Wisconsin Press. McAdam, Doug. 1989. “The Biographical Consequences of Activism.” American Sociological Review 54:744–60. McAuley, Alastair. 1979. Economic Welfare in the Soviet Union: Poverty, Living Standards, and Inequality. Madison, WI: The University of Wisconsin Press. Merton, Robert. K. 1968. “The Matthew Effect in Science: The Reward and Communication System on Science.” Science 199:55–63. Mincer, Jacob. 1974. Schooling, Experience, and Earnings. New York: National Bureau of Economic Research; distributed by Columbia University Press. Moen, Phyllis. 1985. “Continuities and Discontinuities in Women’s Labor Force Activity.” in Life Course Dynamics: Trajectories and Transitions, 1968–1980, edited by G. H. Elder. Ithaca, NY: Cornell University Press. ¨ Moen, Phyllis, Glen H. Elder, and Kurt Luscher. 1995. Examining Lives in Context: Perspectives on the Ecology of Human Development. Washington, DC: American Psychological Association. Moore, Barrington. 1987. Authority and Inequality under Capitalism and Socialism. Oxford, UK: Clarendon Press. Nathan, Andrew J. 1976. “Policy Oscillations in the People’s Republic of China.” The China Quarterly 68:720–33. Naughton, Barry. 1996. Growing out of the Plan. New York: Cambridge University Press. Nee, Victor. 1989. “A Theory of Market Transition: From Redistribution to Markets in State Socialism.” American Sociological Review 54:663–81. 1991. “Social Inequality in Reforming State Socialism: Between Redistribution and Markets in State Socialism.” American Sociological Review 54:663– 81.

346

References

1992. “Organizational Dynamics of Market Transition: Hybrid Forms, Property Rights, and Mixed Economy in China.” Administrative Science Quarterly 37:1–27. 1996. “The Emergence of a Market Society: Changing Mechanisms of Stratiﬁcation in China.” American Journal of Sociology 101:908–49. Nee, Victor, and Yang Cao. 1999. “Path Dependent Societal Transformation: Stratiﬁcation in Hybrid Mixed Economies.” Theory and Society 28:799– 834. 2002. “Postsocialist Inequalities: The Causes of Continuity and Discontinuity.” Research in Social Stratiﬁcation and Mobility 19:3–39. Nee, Victor and Peng Lian. 1994. “Sleeping with the Enemy: A Dynamic Model of Declining Political Commitment in State Socialism.” Theory and Society 23:253–96. Nee, Victor, and Rebecca Matthews. 1996. “Market Transition and Societal Transformation in Reforming State Socialism.” Annual Review of Sociology 22:401–35. North, Douglass C., and Robert Thomas. 1973. The Rise of the Western World. Cambridge, UK: Cambridge University Press. Oberschall, Anthony. 1996. “The Great Transition: China, Hungary, and Sociology Exit Socialism into the Market.” American Journal of Sociology 101:1028–41. Ofﬁce of Organization Committee, CCP Central Committee. 1993. The Grand Trend of Administrative Reform (in Chinese) [Xingzheng Gaige Daqushi]. Beijing: Jingji Kexue Chubanshe. Oi, Jean C. 1989. State and Peasant in Contemporary China: The Political Economy of Village Government. Berkeley, CA: University of California Press. 1992. “Fiscal Reform and the Economic Foundations of Local State Corporatism in China.” World Politics 45:99–126. 1999. Rural China Takes Off: Institutional Foundations of Economic Reform. Berkeley, CA: University of California Press. Oksenberg, Michel. 1968. “The Institutionalization of the Chinese Communist Revolution: The Ladder of Success on the Eve of the Cultural Revolution.” China Quarterly 36:61–92. O’Rand, Angela M. 1986. “The Hidden Payroll: Employee Beneﬁts and the Structure of Workplace Inequality.” Sociological Forum 1:657–83. Parish, William L. 1984. “Destratiﬁcation in China.” pp. 84–120 in Class and Social Stratiﬁcation in Post-Revolution China, edited by J. Watson. New York: Cambridge University Press. Parish, William L., and Ethan Michelson. 1996. “Politics and Markets: Dual Transformation.” American Journal of Sociology 101:1042–59. Parish, William L., and Martin King Whyte. 1978. Village and Family in Contemporary China. Chicago: University of Chicago Press. Parkin, Frank. 1979. Marxism and Class Theory: A Bourgeois Critique. New York: Columbia University Press. Peng, Yusheng. 1992. “Wage Determination in Rural and Urban Industrial China.” American Sociological Review 57:198–213.

References

347

Pepper, Suzanne. 1996. Radicalism and Education Reform in 20th-Century China: The Search for an Ideal Development Model. Cambridge UK: Cambridge University Press. Polanyi, Karl. 1965. The Great Transformation. Boston: Beacon Press. Psacharopoulos, George. 1994. “Returns to Investment in Education: A Global Update.” World Development 22:1325–43. Raftery, Adrian E., and Michael Hout. 1993. “Maximally Maintained Inequality: Expansion, Reform, and Opportunity in Irish Education, 1921–1975.” Sociology of Education 66:41–62. Research Ofﬁce, CCP Central Committee. 1991. Major Events in the CCP History (in Chinese) [Zhongguo Gongchandang Lishi Dashiji]. Beijing: Renmin Chubanshe. Reskin, Barbara F. and Patricia A. Roos. 1990. Job Queues, Gender Queues: Explaining Women’s Inroads into Male Occupations. Philadelphia: Temple University Press. Riskin, Carl, Zhao Renwei, and Li Shi (eds.) 2001. China’s Retreat from Equality. Armonk, NY: M. E. Sharpe. Rona-Tas, Akos. 1994. “The First Shall Be Last? Entrepreneurship and Communist Cadres in the Transition from Socialism.” American Journal of Sociology 100:40–69. Rona-Tas, Akos, and Alya Guseva. 2001. “The Privileges of Past Communist Party Membership in Russia and Endogenous Switching Regression.” Social Science Research 30:641–52. Rosen, Stanley. 1982. Red Guard Factionalism and the Cultural Revolution in Guangzhou (Canton). Boulder, CO: Westview Press. Rosenbaum, James E. 1984. Career Mobility in a Corporate Hierarchy. New York: Academic Press. Ryder, Norman B. 1965. “The Cohort as a Concept in the Study of Social Change.” American Sociological Review 30:843–61. Scalapino, Robert A., and Gordon A., Bennett. 1972. Elites in the People’s Republic of China. Seattle: University of Washington Press. Schurmann, Franz. 1968. Ideology and Organization in Communist China. Berkeley, CA: University of California Press. Scott, James C. 1985. Weapons of the Weak: Everyday Forms of Peasant Resistance. New Haven, CT: Yale University Press. Selznick, Philip. 1952. The Organizational Weapon: A Study of Bolshevik Strategy and Tactics. Santa Monica, CA: Rand Corp. Settersten, Richard A., Jr., and Karl Ulrich Mayer. 1997. “The Measurement of Age, Age Structuring, and the Life Course.” Annual Review of Sociology 23:233–61. Shao, Yanxiang. 1998. “Preface.” in The Summer in 1957: From Hundred Flowers to Struggle between Two Sides (in Chinese) [1957 Nian De Xiaji: Cong Baijia Zhengming Dao Liangjia Zhengming], edited by Z. Zhu. Zhengzhou: Henan Renmin Chubanshe. Shavit, Yossi and Hans-Peter Blossfeld. 1993. Persistent Inequality: Changing Educational Attainment in Thirteen Countries. Boulder, CO: Westview Press.

348

References

Shi, Tianjian. 1997. Political Participation in Beijing. Cambridge, Mass: Harvard University Press. Shi, Weimin (ed.). 1996. Selected Correspondence by Sentdown Youth (in Chinese) [Zhiqing Shuxin Xuanbian]. Beijing: Zhongguo Shehui Kexue Chubanshe. Shirk, Susan L. 1982. Competitive Comrades: Career Incentives and Student Strategies in China. Berkeley, CA: University of California Press. 1993. The Political Logic of Economic Reform in China. Berkeley, CA: University of California Press. Shue, Vivienne. 1988. The Reach of the State: Sketches of the Chinese Body Politic. Stanford, CA: Stanford University Press. Simkus, Albert, and Rudolf Andorka. 1982. “Inequalities in Educational Attainment in Hungary, 1923–1973.” American Sociological Review 47:740–51. Skilling, H. Gordon. 1966. “Interest Groups and Communist Politics.” World Politics 18:435–51. Sørensen, Aage B. 1977. “The Structure of Inequality and the Process of Attainment.” American Sociological Review 42:965–78. 1986. “Theory and Methodology in Social Stratiﬁcation.” pp. 69–95 in The Sociology of Structure and Action, edited by U. Himmelstrand. London: Sage. Sørensen, Jesper. 1992. “Locating Class Cleavages in Intergenerational Mobility: Cross-National Commonalities and Variations in Mobility Patterns.” European Sociological Review 8:267–81. Sørensen, Aage B., and Arne L. Kalleberg. 1981. “An Outline of a Theory of the Matching of Persons to Jobs.” pp. 49–74 in Sociological Perspectives on Labor Markets, edited by I. Berg. New York: Academic Press. SSB (State Statistics Bureau). 1990a. Historical Statistics of Provinces, Autonomous Regions, and Metropolises, 1949–1989 (Quanguo Gesheng, Zizhiqu, Zhixiashi Lishi Tongji Ziliao Huibian (1949–1989). Beijing: Zhongguo Tongji Chubanshe. 1990b. Statistical Yearbook of the Chinese Cities 1990 (Zhongguo Chengshi Tongji Nianjian 1990). Beijing: Zhongguo Tongji Chubanshe. 1995. China Statistical Yearbook, 1995. Beijing: Zhongguo Tongji Chubanshe. 1999. Comprehensive Statistical Data and Materials on 50 Years of New China (in Chinese) [Xin Zhongguo Wushinianlai Tongji Ziliao Huibian]. Beijing: Zhongguo Tongji Chubanshe. Stark, David. 1986. “Rethinking Internal Labor Markets: New Insights from a Comparative Perspective.” American Sociological Review 51:492–504. 1996. “Recombinant Property in East European Capitalism.” American Journal of Sociology 101:993–1027. Stark, David, and Laszlo Bruszt. 1998. Postsocialist Pathways: Transforming Politics and Property in East Central Europe. New York: Cambridge University Press. Stone, Katherine. 1973. “The Origins of Job Structures in the Steel Industry.” Radical America 7:19–64. Strang, David, and James N. Baron. 1990. “Categorical Imperatives: Structure of Job Titles in California State Agencies.” American Sociological Review 55:479–85.

References

349

Szel´enyi, Ivan. 1978. “Social Inequalities in State Socialist Redistributive Economies.” International Journal of Comparative Sociology 16:63–87. 1983. Urban Social Inequalities under State Socialism. New York: Oxford University Press. Szel´enyi, Ivan, Katherine Beckett, and Lawrence P. King. 1994. “The Socialist Economic System.” pp. 234–51 in The Handbook of Economic Sociology, edited by N. J. Smelser and R. Swedberg. Princeton, NJ: Princeton University Press. Szel´enyi, Ivan, and Eric Kostello. 1996. “The Market Transition Debate: Toward a Synthesis?” American Journal of Sociology 101:1082–96. Szel´enyi, Ivan, and Robert Manchin. 1987. “Interrupted Embourgeoisement: Social Background and Life History of Family Agricultural Entrepreneurs in Socialist Hungary.” Research in Social Stratiﬁcation and Mobility 8:253– 78. Szel´enyi, Szonja. 1998. Equality by Design: The Grand Experiment in Destratiﬁcation in Socialist Hungary. Stanford, CA: Stanford University Press. Szel´enyi, Szonja, and Karen Aschaffenburg. 1993. “Inequalities in Educational Opportunity in Hungary.” pp. 273–302 in Persistent Inequality, edited by Y. Shavit and H. Blossfeld. Boulder, CO: Westview. Tang, Wenfang, and William Parish. 2000. Chinese Urban Life under Reform: The Changing Social Contract. New York: Cambridge University Press. Tilly, Charles. 1978. From Mobilization to Revolution. Reading, MA: AddisonWesley. 1986. The Contentious French. Cambridge, MA: Belknap Press of Harvard University Press. 1998. Durable Inequality. Berkeley, CA: University of California Press. Titma, Mikk, and Nancy Brandon Tuma. 1993. “Stratiﬁcation Research in a Changing World.” pp. 225–54 in Eastern European Societies at the Threshold of Change, edited by J. Szmatka, Z. Mach, and J. Mucha: Edwin Mellen Press. Treiman, Donald J. 1970. “Industrialization and Social Stratiﬁcation.” pp. 207– 34 in Social Stratiﬁcation: Research and Theory for the 1970s, edited by E. O. Laumann: Bobbs-Merrill. Treiman, Donald J., and Kam-Bor Yip. 1989. “Educational and Occupational Attainment in 21 Countries.” pp. 373–94 in Cross-National Research in Sociology, edited by M. L. Kohn. Newbury Park: Sage. Tsou, Tang. 1986. The Cultural Revolution and Post-Mao Reforms: A Historical Perspective. Chicago: University of Chicago Press. Tuma, Nancy Brandon, and Michael T. Hannan. 1984. Social Dynamics: Models and Methods. Orlando: Academic Press. Unger, Jonathan. 1982. Education under Mao: Class and Competition in Canton Schools, 1960–1980. New York: Columbia University Press. Viola, Lynne. 1987. The Best Sons of the Fatherland: Workers in the Vanguard of Soviet Collectivization. New York: Oxford University Press. Vogel, Ezra. 1967. “The “Regulation” of Cadres.” China Quarterly 29:36–60. Voslensky, Michael. 1984. Nomenklatura: The Soviet Ruling Class. Garden City, NY: Doubleday & Co.

350

References

Wade, Robert. 1990. Governing the Market: Economic Theory and the Role of Government in East Asian Industrialization. Princeton, NJ: Princeton University Press. Walder, Andrew G. 1986. Communist Neo-Traditionalism: Work and Authority in Chinese Industry. Berkeley, CA: University of California Press. 1992. “Property Rights and Stratiﬁcation in Socialist Redistributive Economies.” American Sociological Review 57:524–39. 1995a. “Career Mobility and the Communist Political Order.” American Sociological Review 60:309–28. 1995b. “Local Governments as Industrial Firms.” American Journal of Sociology 101:263–301. (ed.). 1995c. The Waning of the Communist State: Economic Origins of Political Decline in China and Hungary. Berkeley, CA: University of California Press. (ed.). 1996a. China’s Transitional Economy. New York: Oxford University Press. 1996b. “Markets and Inequality in Transitional Economies: Toward Testable Theories.” American Journal of Sociology 101:1060–73. 2002. “Markets and Income Inequality in Rural China: Political Advantage in an Expanding Economy.” American Sociological Review 67:231–53. Walder, Andrew G., Bobai Li, and Donald J. Treiman. 2000. “Politics and Life Chances in a State Socialist Regime: Dual Career Paths into the Urban Chinese Elite, 1949 to 1996.” American Sociological Review 65:191–209. Weber, Max. 1946. From Max Weber: Essays in Sociology. Translated by H. H. Gerth and C. W. Mills. New York: Oxford University Press. 1978. Economy and Society: An Outline of Interpretive Sociology. Edited by Guenth Roth and Claus Wittich. Berkeley, CA: University of California Press. Wesolowski, Wlodzimierz, and Edmund Wnuk-Lipinski. 1992. “Transformation of Social Order and Legitimization of Inequalities.” pp. 83–95 in The Polish Road from Socialism: The Economics, Sociology, and Politics of Transition, edited by W. D. Connor and P. Ploszajski. Armonk, NY: M. E. Sharpe. White, Lynn T. III. 1989. Politics of Chaos: The Organizational Causes of Violence in China’s Cultural Revolution. Princeton, NJ: Princeton University Press. Whyte, Martin King. 1973. “Bureaucracy and Modernization in China: The Maoist Critique.” American Sociological Review 38:149–63. 1984. “Sexual Inequality under Socialism: The Chinese Case in Perspective.” pp. 198–238 in Class and Social Stratiﬁcation in Post-Revolutionary China, edited by J. Watson. New York: Cambridge University Press. 1985. “The Politics of Life Chances in the People’s Republic of China.” pp. 244– 65 in Power and Policy in the PRC, edited by Y. M. Shaw. Boulder, CO: Westview Press. 1995. City versus Countryside in China’s Development. Canberra: Australian National University. Whyte, Martin King, and William L. Parish. 1984. Urban Life in Contemporary China. Chicago: University of Chicago Press. Wolf, Margery. 1985. Revolution Postponed: Women in Contemporary China. Stanford, CA: Stanford University Press.

References

351

Wright, Erik O. 1984. “A General Framework for the Analysis of Class Structure.” Politics and Society 13:383–423. 1989. The Debate on Classes. New York: Verso. Wu, Xiaogang, and Yu Xie. 2003. “Does the Market Pay Off? Earnings Returns to Education in Urban China.” American Sociological Review 68:425– 42. Xie, Yu, and Emily Hannum. 1996. “Regional Variation in Earnings Inequality in Reform Era Urban China.” American Journal of Sociology 101:950–92. Xu, Songtao, and Linyang Zhang (eds.). 1992. Handbook of Personnel Management Practice (in Chinese) [Zhongguo Renshi Guanli Gongzuo Shiyong Shouce]. Beijing: Zhongguo Caizheng Jingji Chubanshe. Xu, Xiao. 2000. Correspondence in Ordinary Lives (in Chinese) [Minjian Shuxin]. Hefei: Anhui Wenyi Chubanshe. Yamaguchi, Kazuo. 1983. “The Structure of Intergenerational Occupational Mobility: Generality and Speciﬁcity in Resources, Channels, and Barriers.” American Journal of Sociology 88:718–45. Yang, Dali L. 1996. Calamity and Reform in China. Palo Alto, CA: Stanford University Press. Yang, Zhiyun (ed.). 1992. Sentdown Youth Archive (in Chinese) [Zhiqing Dangan]. Chengdu: Sichuan Wenyi Chubanshe. Yao, Ruobing. 1984. Education in China (in Chinese) [Zhongguo Jiaoyu] Hong Kong: Huafeng Shuju. Yu, Fu, and Weihua Wang. 1993. Tragic Youth (in Chinese) [Beichuang Qingchun]. Beijing: Tuanjie Chubanshe. Zhao, Shenghui. 1987. The Organizational History of the CCP (in Chinese) [Zhongguo Gongchandang Zuzhishi Gongyao]. Hefei: Anhui Renmin Chubanshe. Zhao, Wei, and Xueguang Zhou. 2002. “Institutional Changes and Returns to Education in Urban China: An Empirical Assessment.” Research in Social Stratiﬁcation and Mobility 19:337–73. 2004. “Chinese Organizations in Transition: Changing Promotion Patterns in the Post-Mao Era.” Organization Science forthcoming. Zhou, Xueguang. 1992a. “On Political Cycles in China” (in Chinese) [Lun “Yiguan Jiusi, Yifong Jiuluan.]” pp. 149–69 in State and Society in Contemporary China (in Chinese), edited by X. Zhou. Taipei: Guiguan Press. (ed.). 1992b. State and Society in Contemporary China (in Chinese) [Dangdai Zhongguo De Guojia Yu Shehui Guanxi]. Taipei: Guiguan Press. 1993a. “Occupational Power, State Capacities and the Diffusion of Licensing in the American States: 1890–1950.” American Sociological Review 58:536– 52. 1993b. “Unorganized Interests and Collective Action in Communist China.” American Sociological Review 58:54–73. 1995. “Partial Reform and the Chinese Bureaucracy in the Post-Mao Era.” Comparative Political Studies 28:440–68. 2000a. “Economic Transformation and Income Inequality in Urban China: Evidence from Panel Data.” American Journal of Sociology 105:1135– 74.

352

References

2000b. “Reply: Beyond the Debate and toward Substantive Institutional Analysis.” American Journal of Sociology 105:1190–95. 2001. “Political Dynamics and Bureaucratic Career Patterns in the People’s Republic of China, 1949–1994.” Comparative Political Studies 34:1036–62. Zhou, Xueguang, and Liren Hou. 1999. “Children of the Cultural Revolution: The State and the Life Course in the People’s Republic of China.” American Sociological Review 54:12–36. Zhou, Xueguang, and Phyllis Moen. 2001a. “Explaining Life Chances in China’s Economic Transformation: A Life Course Approach.” Social Science Research 30:552–77. 2001b. “Job Shift Patterns of Husbands and Wives in Urban China.” pp. 333– 67 in Careers of Couples in Contemporary Society, edited by H.-P. Blossfeld and S. Drobnic. Oxford, UK: Oxford University Press. Zhou, Xueguang, Phyllis Moen, and Nancy B. Tuma. 1998. “Educational Stratiﬁcation in Urban China: 1949–1994.” Sociology of Education 71:199–222. Zhou, Xueguang, and Olga Suhomlinova. 2000. “Redistribution under State Socialism: A USSR and PRC Comparison.” Research in Social Stratiﬁcation and Mobility 18:163–204. Zhou, Xueguang, Nancy B. Tuma, and Phyllis Moen. 1996. “Stratiﬁcation Dynamics under State Socialism: The Case of Urban China, 1949–1993.” Social Forces 74:759–96. 1997. “Institutional Change and Patterns of Job Shifts in Urban China: 1949– 1994.” American Sociological Review 62:339–65. Zukin, Sharon. 1978. “The Problem of Social Class under Socialism.” Theory and society 6:391–427.

Index

administrative authorities, job shift impact of, 250 administrative cadres, distinguished from professionals, 13 administrative career line, 168, 183 administrative (salary) grades, 157 administrative hierarchy of the cadre system, 156 administrative ladder. See professional ladder administrative occupations, 171, 173–179 administrative ranks age-related promotion rate for, 186 distribution of in 1993, 159 age. See also seniority effects on 1993 income across cohorts, 290 effects on formal promotion rates, 186 effects on internal job shifts, 262 effects on job-shift rates across sectors, 268 of entry into the workforce, 280 as an indicator of political allegiance, 168 rates of job shifts and, 261 restrictions in allocating political opportunities, 181 Allison, Paul D., 62 allocative mechanisms across cohorts, 278 changes in, 81 effects moderated by life course factors, 276

Alonso, William, 99 Althauser, Robert P., 15, 101, 102 Alwin, Duane F., 127, 328 Anderson, Margo J., 99 Andorka, Rudolf, 73 Arendt, Hannah, 318 Aschaffenburg, Karen, 59 Asselain, Jean-Charles, 13 Atkinson, Anthony B., 304 authorities, access to, 31 authority relationships in organizations, 319 under state socialism, 8 in work environments, 10, 303–306 Bachman, David, 155 Barnett, A. Doak, 155 Baron, James N., 15, 99, 102 Bauman, Zygmunt, 8, 154 Becker, Gary Stanley, 8, 100, 200 Beckett, Katherine, 8 Beijing, 216 Bendix, Reinhard, 5 Bennett, Gordon A., 155 Bernstein, Thomas P., 125 Bialecki, Ireneusz, 73 Bian, Fuqin, 244 Bian, Yanjie, 11, 14, 15, 27, 28, 29, 30, 41, 64, 102, 164, 198, 200, 201, 221, 227, 232, 244, 246, 249, 275, 277, 312 Bielby, William T., 15, 99 Biggart, Nicole Woolsey, 31 biographical life stage, 274

353

354

Index

Blau, Peter M., 5, 15, 23, 71, 100, 223 Blossfeld, Hans-Peter, 72, 96 Boisot, Max, 32 Bourdieu, Pierre, 72 Bruszt, Laszlo, 25, 31 Brzezinski, Zbigniew, 318 Bunce, Valerie Jane, 19 Burawoy, Michael, 8, 299, 304, 316 Bureau of Policies and Regulations, Ministry of Labor, 131, 149, 156, 226, 229, 327 bureaucracy, 156 advancement in, 156–162, 173–179, 182–192 beneﬁting from political authority and market opportunities, 275 versus bureaucrats as a social class, 165 career tracks in, 157 identifying sources of personnel supply to, 174 as an organizational weapon of the state, 163 bureaucratic allegiance, 165 bureaucratic career, 171 lines, 13, 191 opportunities, 171–173 bureaucratic career patterns, 162–171 dual path model for, 162–164 historical evolution of, 155 macro-political processes and shifting state policies, 164 theoretical models for, 167–171 bureaucratic class, 154, 300 deﬁning, 13, 300 dismantlement during the Cultural Revolution, 127 in state socialist societies, 10 bureaucratic class thesis, 9–16, 300–309 conclusions on, 314 criteria to evaluate, 10 bureaucratic hierarchy, levels of, 55 bureaucratic privileges, socio-economic status and, 300–303 bureaucratic promotions distinctive patterns of, 193 institutional channels of, 12 opportunity structures of, 189 bureaucratic reforms, 333 bureaucratic resistance, 170, 171

bureaucratic system, evolution of, 157–159 bureaucrats, 13, 156. See also cadres versus all other social groups, 9 authority of, 9 mid- and low-level, 11 reallocating opportunities among different cohorts of, 169 in the redistributive economy, 101 Burns, John F., 163 cadre background, 60 cadre career line, 191. See also administrative occupations cadre class background label, 120 cadre cohort effects, 181 cadre retirement system, 167 cadre status alternative measures of, 200 alternative speciﬁcations of, 210 different operationalizations of, 197 measures of, 214 regional ﬁndings, 220 regionalized returns to, 219 “cadre” system, 156 cadres, 9, 55, 157. See also bureaucrats; high-rank cadres career lines for, 164 composition of, 45, 159, 191 differentiating between public sector and economic sector, 210 entry into, 112, 173–179 “high-rank” and “low-rank” groups of, 210 rate of becoming, 105 rates of income for, 206 sending on-the-job to colleges, 47 socioeconomic conditions of, 11 Cai, He, 305 campaigns. See political campaigns Campbell, John L., 30 Cao, Yang, 27, 29, 223 capitalist societies. See market societies capitals, impacts on life chances over time, 23 career tracks, 157, 162 Carroll, Glenn R., 253 CCP rectiﬁcation, 17 central government ﬁrms, 43, 57 central planning authority, allocating resources as latent beneﬁts, 228

Index centralization during the 1949–1965 historical period, 36 gradual decline of, 39 impact of economic disaster on, 37 centralized state capacities for resource allocation and transfer, 20 contention with bureaucratic interests, 164 monopolizing resources for social mobilization, 318 Chen, Danyan, 75 Chen, Shaoping, 101, 156, 159 Chen, Zhili, 38, 165 Child, John, 32 children of the Cultural Revolution, 38 job allocations for, 133 sent-down experience consequences, 127 Chirot, Daniel, 12 cities. See sampling class background. See family class background class identiﬁcation in Communist ideology, 24 class reproduction, 10, 170 absence of parental effects on promotion, 193 central to the making of a social class, 306 principles of, 10 processes reﬂected by educational stratiﬁcation, 71 classes. See also social classes; social groups constructed in terms of occupations, 22 stable boundaries, 314 state policy shifts, sensitivity to, 22 clerks, 44, 45, 56. See also ofﬁce workers clientelist system, 319 Clipp, Elizabeth C., 127, 328 coevolution assessing the outcomes of, 33 mechanisms governing, 31 of politics and markets, 30–33, 197 Cohen, Ronald L., 127, 328 cohort effects, 252 for administrative and professional promotions, 188 on Communist Party membership, 181–182 in gaining political status, 169

355

on job-shift rates across sectors, 268 on work organization promotions, 189 cohort variations in access to the opportunity structure, 293 among different types of organizations, 289 in economic beneﬁts, 277, 290–293 in the effects of social positions on job shift patterns, 290 cohort-based seniority principle, 188 cohorts, 274. See also historical cohorts; professional cohort results allocative mechanisms across, 278 analytical importance of, 125 characteristics of historical, 283 competing incentives for different, 277 differences in response to opportunities, 276 distinctive historical contexts, 281 distinctive opportunities, 294 distribution of job destinations, 282 entering the work force, 279 exposure and response to social change, 274 variations in job shift patterns, 284–290 Coleman, James S., 5 collective action opportunities in the state-society relationship, 321 transforming unorganized interests into, 320 collective ﬁrms (jiti qiye), 14, 57, 65 job shifts to, 253 under local governments, 37 proportions employed in, 41 rates of ﬁrst entry into, 106 collective inaction critical role of, 321 stratiﬁcation dynamics and, 324 collective sector, 39, 40, 44 expansion of, 44 industrial output by, 47 presence of, 325 rates of ﬁrst entry into, 105 collectivization of the craft industry, 36 college education effects in different cohorts, 290 effects on entering high-status organizations, 116 effects on job shifts, 265, 288

356

Index

college education (cont.) rate of job shifts to government agencies, 258 rate of returns across selected years, 206 college graduates downward mobility experienced by, 114 entering the administrative system, 103 internal job shift rates, 262 moving to the private sector, 331 “college granting associate degree” (dazhuan), 55 colleges admission closed, 38 enrollment rate into, 47, 48 entry into, 83, 84 number of, 77, 78 rates of entry into, 76, 77, 80 reopening of, 77, 129 transition rates to, 95 commune and collectivization movement, 17 communist new traditionalism model, 319 Communist Party authority exercised by, 8 interwoven with the administrative apparatus, 57 recruitment into, 165, 179–182 Communist Party membership, 179 acquisition of as a political status, 167 effects on formal promotion rates, 187 effects on promotions in work organizations, 189 income advantages of, 214 as an indicator of political capital, 55, 162, 200 rates of entry into, 171 regionalized income returns to, 216 role in administrative recruitment and promotion, 168, 183 community college education, 49 competence. See also educational credentials emphasized as a selection criteria, 167 versus political loyalty, 12 Connor, Walter D., 8, 13, 228 cultural capital, 72 Cultural Revolution (1962–1968), 17. See also children of the Cultural Revolution disruption of educational progression, 82

effects on bureaucrats, 323 elimination of gender equality, 90 impact on high school and college rates of entry, 76 “send-down” policies, 115 targeting and purging of bureaucrats, 166 treatment of intellectuals, 23 cumulative advantage (or disadvantage), 274 data collection for social science research, 50 Davis, Deborah, 15, 19, 23, 126, 244, 249, 250, 282 Deng, Xian, 129, 130, 328 Deng, Xiaoping, 255, 332 Deng, Zhong, 73 “destratiﬁcation” egalitarian policy of, 65 policies during the 1960–1965 historical period, 82 state policies during the Mao era, 209 Diggle, Peter, 63 DiPrete, Thomas, 5, 15, 335 directives, on recruitment, promotion and salary changes, 161 discrete-time event history model, 62, 108 Djilas, Milovan, 9, 10, 12, 73, 154 Dobbin, Frank, 226 Doeringer, Peter B., 102 Douglas, Mary, 99 dual path model, 155 compared to the stratiﬁcation dynamics model, 167 evidence consistent with, 192 explaining bureaucratic career patterns, 162–164 explaining cadre and professional trajectories, 12 predictions regarding bureaucratic career lines, 168 dual structure of workplace and occupation, 103 Duncan, Otis Dudley, 5, 23, 71, 100 durable inequality, based on social relations and social closures, 6 dynamic aspects of state socialist redistribution, 19 dynamic modeling, 60, 62

Index “economic adjustment”, policies of, 37 economic agents, rent-seeking behavior of, 32 economic beneﬁts cohort variations in, 277 distribution of, 54, 228 low disparity in, 315 economic resources cross-cohort variations, 279 variations in distribution over time, 207 economic rewards cohort variations in, 290–293 in the form of redistributive beneﬁts, 227 economic sectors, 15, 58, 265. See also industries external job shifts moving across, 265–268 importance in the redistribution of resources, 13 tracked by ofﬁcial statistics, 99 types of, 248 variations in redistributive beneﬁts across, 228 education. See also “technical school” (jixiao); college education; high schools; middle schools composition in the sample, 48 credentials emphasized in the reform era, 169 credentials important for professional career lines, 168 effect on 1993 income by cohort, 291 effect on Communist Party membership, 179 effect on entering cadre and professional occupations, 115 effect on entering high-status organizations, 116 effect on formal promotion rates, 186 effect on housing allocation, 238 effect on individuals’ life chances, 23 effect on job shifts to work organizations, 288 effect on upward mobility, 163 effect on work organization promotions, 189 impact entry into jobs, 177 impact on entry into a professional occupation, 114 important for promotion along the professional ranks, 187

357

low returns compared to industrialized societies, 223 regionalized returns to, 219 role in ﬁrst-job destinations, 122 role in social mobility, 70 as a sorting mechanism for class reproduction, 73 summary of ﬁndings on returns to, 309 suppressed returns to, 309 value to a socialist planning system, 201 educational attainment, 69 direct role of the state and state policies on, 74 distinctive pattern of, 96 industrialization hypothesis, 88 mediating effects of stratiﬁcation dynamics, 123 role of social origins in, 85 educational credentials, 201 educational examination system, 69 educational levels advanced, 48 composition of, 48 distribution in the sample, 49 probability of entering, 86–87 processes reﬂecting the attainment of, 81 transition rates of entry into, 81 educational opportunities, 47–49 dramatic ﬂuctuation over time, 75 expansion during the early period, 77 historical evolution of, 75–80 historical trend of, 75 increasing over time, 283 in socialist societies, 73 educational qualiﬁcations as indicators of political loyalty, 168 treating as political liabilities, 23 educational stratiﬁcation, 53, 69–97 mechanisms of, 73 models of, 72 patterns of, 70 stability over time and across societies, 71 egalitarian policy in the Mao era, 302 Elder, Glen H., Jr., 20, 126, 127, 128, 272, 274, 328 elementary or no education, 48, 54 employees, distribution across types of organizations, 41 equalizing opportunities, periods of great achievement, 315

358

Index

Erikson, Robert, 5, 101 Esping-Andersen, Gøsta, 6, 30, 335 Evans, Peter B., 30, 333 event history analysis, 252 of continuous-time and discrete-state outcomes, 61–64 logistic regression model, 62 “exploiting class”, 60, 120 children experiencing a declining rate, 95 children sent to rural areas, 137 children systematically discriminated against, 120 external job shifts, 252 across economic sectors, 265–268 across types of organizations, 255–261 family class background (jiating chushen), 82 basis of, 93 class label of, 38 effects of, 93–95, 141 labels, 59 ofﬁcial label of, 59, 120 variations with entries into ﬁrst jobs, 120 family resources, individual-based, 72 famine (1959–61), 1 farms, 58 entry into, 107 occupation, 56, 114 origin, 259 sector, 44 fathers education, 59, 90, 110 occupation, 59, 138, 141 socio-political status, 59 Featherman, David L., 5, 6, 101 female employees entry into administrative and professional occupations, 174 lower promotion rate of, 186 maternity leave policy, 226–227 Feng, Denggong, 37 Ferber, Marianne A., 223 ﬁrst jobs association with social origins, 108 distribution by cohort, 281 entry into, 104, 108, 112–116, 131 patterns of attainment, 98–111, 123 role of education in, 122

shifts away from, 252 for urban youth during the Cultural Revolution, 133 Flakierski, Henryk, 228 Follis, Jane Thompson, 19 foreign-owned ﬁrms, 58 nationalization of, 36 signiﬁcant presence of, 331 Freedman, Ronald, 88 Freidson, Eliot, 6 Ganzeboom, Harry B. G., 6 Gardiner, Karen, 304 GEE (generalized estimation equation) model, 63, 183 gender educational attainment transition rates by, 89 effects on formal promotion patterns, 186 effects on transition rates to educational levels, 85–92 rates of income differences over selected years, 206 gender gap in Communist Party membership, 179 gender inequality in educational attainment over time, 88 patterns over time, 88 in the reform era, 169 gendered jobs, 112, 174 generalized estimation equation (GEE) model. See GEE model George, Linda, 126 Gerber, Theodore P., 27, 29, 32, 59, 73, 88 Giele, Janet Zollinger, 272 Gini coefﬁcients, 303 in income inequality across countries, 303 measuring the USSR’s income inequality, 223 for urban and rural China, 224 Goldthorpe, John H., 5, 101 Gottschalk, Peter, 224, 335 government agencies (zhengfu bumen), 57 analysis of cohorts entering, 281 changing from redistributors to regulators, 32 downsizing in the reform era, 295

Index ﬁrst entry into, 106 higher rate of entry into the Communist Party, 182 job shifts to, 253 government sector, 248 grade system for salaries, 157 Granovetter, Mark, 31 Great Depression in the United States, 126 Great-Leap-Forward, 17, 313 extreme economic conditions, 323 high college enrollment rate, 48 impact on high school and college rates of entry, 76 group boundaries, 323, 326 Grusky, David B., 5, 6, 15, 335 Guan, Ying, 244 Guseva, Alya, 29 Gustafsson, Bjorn, 224, 335 Guthrie, Doug, 30, 197, 226 Hallinan, Maureen T., 95 Hamilton, Gary G., 31 Hankiss, Elemer, 19 Hanley, Eric, 73 Hannan, Michael T., 60, 62 Hannum, Emily, 27, 73, 88, 216, 221, 232, 275 Harding, Harry, 9, 12, 163, 166 Hauser, Robert Mason, 5, 6, 101 Havel, Vaclav, 320 hazard rates, 131, 253 capturing variations in, 62 of entry into advantageous positions, 171 of entry into different types of work organizations, 105 of entry into ﬁrst jobs, 104, 131 of job entries, 103 of job shifts, 253 of promotion, 172 of returning to urban areas (1967–1987), 131 heavy industry, 58, 248 favored by state growth strategies, 14 favored over consumer goods, 7 Heyns, Barbara, 73 hierarchical order among types of work organizations, 202 of jobs and workplaces, 100 hierarchical system for salary grades, 157

359

hierarchies related to the social construction of jobs, 122 of work organizations, 14 high schools enrollment rate into, 48 entry into, 76, 77, 79, 84 transition rates to, 93 higher education institutions, 77, 78. See also colleges high-rank cadres (chu or above), 13, 44, 45, 55, 301 access to housing resources, 238 channeling children back earlier from rural areas, 152 children entering high-status organizations, 119 children returning to urban areas, 138 Cultural Revolution’s impact on the children of, 134 housing quality, 240 income in the pre-reform era, 213 living space and facilities, 243 high-rank professionals, 44, 45, 56 high-status groups, protecting children from state policies, 78 historical approach to analyses of job shift patterns, 270 to patterns of social stratiﬁcation, 25 historical cohorts, 280. See also cohorts 1949–1965, 280, 281, 294 1966–1979, 281, 283, 289, 294 1980–1994, 281, 289 life course location, 280 rates of job shifts, 285, 288 historical periods 1949–1965, 36–37 1966–1979, 38–39 1980–1994, 39 for analyses of educational stratiﬁcation, 82 characterizing changes in policies and political processes, 36–39 criteria of political selection, 166 effects of key covariates over time, 168 impacts of state policies on individual life chances, 61 job shift patterns across, 255–268 Hodgson, Geoffrey M., 31 Hodson, Randy, 102 Hogan, Dennis P., 126, 295

360

Index

Hollingsworth, J. Rogers, 30 household size, effect on living space, 235 housing administration, 234 authority under the metropolitan government, 234 construction, 231 distribution, 233–240 facilities, 240 inequality related to, 244 as a latent beneﬁt, 231 privatization of ownership, 244 quality, 239 space, 234–239 housing allocation, 233–240 “need-based” principle, 235 constraints in urban areas, 234 in Hungary, 233 housing markets absence of, 233 emergence of, 244 Hout, Michael, 5, 6, 23, 27, 32, 59, 72, 73 human capital allocating and rewarding, 201 internal job shifts in different types of organizations, 262 in a market economy, 8 measures of, 54, 210 versus political, 8, 200–201, 309–311 regionalized returns to, 219 role in job shift decisions, 258 role in job shifts across economic sectors, 265 hybrid ﬁrms, 44, 58 higher returns to those working in, 215 job shifts to, 253 as nonstate organizations, 248 presence of, 331 rate of income for, 207 rates of entry into, 107 rise in employment in the reform era, 270 ideological crises, life experiences and, 327–330 income. See also personal income assessing the quality of data, 203 distribution in years of historical signiﬁcance, 203 effects of the sent-down experience on, 148

as an indicator of social inequality, 198 information collected in the sample, 203 income determinants changes among urban residents over time, 198 cross-sectional analyses of, 202–209 for the sent-down experience, 146–151 income inequality fundamental changes among social groups, 199 Gini coefﬁcients for, 303 redistribution and marketization implications for, 199–202 income regimes over time, 207 individual mobility, 65 individual-based family resources, 72 individual-based skills, channeling into occupational careers, 100 industrial output, 46, 47 industrial workers, high political status of, 24 industrialization thesis, 96 “ideal” scenario consistent with, 88 of increasing educational opportunities, 72 industrialized market societies. See market societies industrialized societies, work organizations in social stratiﬁcation processes, 102 industries prioritized on political goals, 7 redistribution of resources and, 13 institutional arrangements of urban economies, 220 of work organizations, 201–202 institutional changes assessments of large-scale, 246 explaining outcomes by the advance of market economies, 28 impacts on individuals at different life stages, 274 implications of, 333 life course implications and, 274–277 multifaceted, 216, 332 path dependent characteristics of, 25 in the post-Mao era, 195 processes of, 30, 39 as processes of coevolution, 30–33 in the reform era, 26–33

Index shaped by the state, 332 social basis of, 324 from social stratiﬁcation to, 330–333 through generational replacement, 295 in the transformation of state socialism, 3, 25 institutional economics, 31 institutional model of unorganized interests and collective action, 320–322 institutional persistence, 202, 222 institutions distribution of attention in, 99 links between the state and individual life chances, 41 role in the processes of social changes, 28 intellectuals, 60. See also professionals incorporating into the political elites, 11 as initiators of political disturbances, 70 intelligentsia, “contradictory location” in the class structure, 12 interactions between existing and emerging institutions, 28 between stratiﬁcation dynamics and life course factors, 54 interest politics in the political arena, 31 intergenerational inheritance, 10, 58, 170, 306 education as a critical mechanism in, 71 ﬁndings regarding, 307 literature based on occupational categories, 101 over time, 5 as a potential source of career patterns, 178 reinforced by the “replacement” policy, 307 suppressed during the Cultural Revolution, 115 internal job shifts, 252, 261–265 Jacobs, Jerry A., 15 Jermakowicz, Wladyslaw, 19 Jiang, Kun, 130 job shift patterns, 247 across different types of work organizations, 277 across historical cohorts, 288 across historical periods, 255–268 across types of organizations, 285–290 among types of economic sectors, 248

361

among types of organizations, 247, 248 cohort difference in response to opportunities, 277–279 comparing over time, 270 as a focus of study, 247–253 historical trends of, 253 outside economic sectors by period, 265–268 outside type of work organization by period, 255–261 overall trends over time, 258 patterns capturing distinctive career pathways, 277 rates to private ﬁrms, 260 response to opportunities and risks, 284–290 to the state sector, 253 within type of organization by period, 262–265 jobs, 99–104, 255. See also occupations destinations by cohort, 282 entry pattern into, 107–111 mobility, 53, 98 patterns of attainment, 100, 111–120, 269 patterns of entry, 104 social origins and, 107 state policies allocating, 103 joint venture ﬁrms (sanzi qiye), 44, 58 Jones, F. Lancaster, 5 junior high. See middle schools Kahneman, Daniel, 276 Kalleberg, Arne L., 15, 102 Katz, Zev, 12 Kaufman, Robert L., 102 Keister, Lisa A., 197 Kelly, Erin, 226 Khan, Azizur Rahman, 233, 244 King, Lawrence P., 8 kitchen as a housing facility, 239 Kohli, Martin, 272, 274 ´ Gyorgy, ¨ Konrad, 8, 11, 12, 163, 201 ´ Kornai, Janos, 8, 37, 40, 303 Kostello, Eric, 30 Kraus, Richard Curt, 73 Krotov, Pavel, 304 labor force composition of, 40–46 duration, 55

362

Index

labor force (cont.) entry into, 98–107, 123 participation, 38 labor insurance, 229, 231 labor statistics, 99 Lampton, David M., 31, 165, 166, 319 Lardy, Nicholas R., 20 large numbers phenomenon, 321 basis for, 325 stratiﬁcation dynamics and, 322–323 Larson, Magali Sarfatti, 6 latent beneﬁts, 225 allocated to industries or work organizations, 14 compared to effects of education and political position, 232 distributed through work organizations, 243 historical overview of the distribution of, 229–233 Lavely, William, 88 Lee, Ching Kwan, 27 Lee, Hong Yung, 9, 12, 22, 23, 24, 155, 163, 165 Li, Bobai, 11, 12, 13, 155, 162, 163, 164, 201 Li, Bohua, 88 Li, Cheng, 155 Li, Guangping, 328 Li, Hanlin, 15, 229, 244 Li, Kang, 15 Li, Lulu, 15, 229, 244 Li, Meng, 15 Li, Peilin, 15, 229, 232, 313 Li, Qiang, 276, 277 Li, Xiaohua, 130 Lian, Peng, 29 Liang, Kung-Yee, 63 Lieberthal, Kenneth, 31, 166, 319 life chances affected by political dynamics and shifts in state policies, 16 consequences of state socialist redistribution on individual, 16 effect of education on, 23 effects of individual or family-based attributes and capital on, 23 impacts of state and state policies on, 20, 40 impacts on individuals’, 16

processes inﬂuencing, 293 re-valuing social resources for, 16 life course, 272 contextual constraints on the social stratiﬁcation processes, 295 effects of the sent-down experience, 145 effects on internal job shifts, 262 effects on job-shift rates across sectors, 268 implications of institutional changes, 274 moderating the impacts of social changes, 274 perspective, 272, 295 relationship to social change, 126 role in job shift decisions, 261 shaping individual life chances, 293 transformations experienced at different stages of, 273 life course approach, 293 bridging macro-level events and micro-level biographies, 296 facilitating assessment of allocative mechanisms across cohorts, 294 life experiences ideological crises and, 327–330 recording over time, 50 life history data from “the State and Life Chances” project, 50–52 lifetime employment practice, 169. See also permanent employment light industry, 14, 58, 248, 249 Lin, Nan, 14, 15, 102, 201, 249, 312, 333 Lindberg, Leon N., 30 Lindblom, Charles Edward, 7 Liu, Ji, 38, 165 Liu, Lufeng, 37 local governments authorities over state ﬁrms, 37 decisions on individual life chances of workers, 103 ﬁrms owned by, 43, 57 increasing role of, 28 strengthening the power of, 200 Logan, John R., 11, 14, 27, 28, 29, 30, 164, 200, 221, 227, 232, 246, 275 logistic regression model for event history analysis, 62 low social inequality, costs of, 315 Lu, Feng, 15

Index Luckacs, Janos, 8, 304 ¨ Luscher, Kurt, 274 MacFarquhar, Roderick, 23 macro-institutional changes, links with individual life chances, 273 macro-political processes affecting life chances of all social groups, 23 affecting the very notion of “capital”, 74 association between social origin and social mobility determined by, 308 highlighted by the stratiﬁcation dynamics model, 167 impacting educational attainment, 74 versus organizational attributes in regard to work organization promotions, 189 reshaping micro-level relationships between social origins and status attainment, 54 for social stratiﬁcation, 19 sustaining an egalitarian policy, 314 managerial autonomy, increasing in industrial organizations, 305 managerial positions promotion within a workplace, 183 promotions into, 183 “manager-in-charge” policy (Changzhang buzezhi), 165 managers. See also political managerial positions given more authority over political cadres, 165 holding dual track titles, 157 included with cadres, 55 limited authority of, 305 Manchin, Robert, 19, 126 manual and non-manual workers, 13 manufacturing industries (chanye gongren), 101 Mao era. See also post-Mao era “destratiﬁcation” state policies, 209 egalitarian policy, 302 income regime associated with, 207 seniority emphasized in, 169 Mao Zedong, 124, 154 “send-down” policy, 125 implementing his vision of socialism, 38 send-down 1968 instruction, 128 suspicion and mistrust toward intellectuals, 70

363

Mare, Robert D., 81, 95 market economies convergence of, 39 emergence of, 27, 42, 131, 250 market expansion, encouraged by the state, 32 market logic, reﬂected by market societies, 7 market mechanisms allocation of resources and, 251 convergence of, 39 market societies, 5 consistent patterns of social mobility, 5 detailed classiﬁcation of occupations in, 99 family social origins as an indicator, 72 job concept referring to occupational status, 100 labor markets role in social stratiﬁcation processes, 15 role of the state in shaping life course, 126 social mobility and status attainment, 5 social origins as channels of social stratiﬁcation, 54 market transition theory, 27, 197, 330 competing explanations, 28 importance of emerging market economies, 27 institutional logic of markets, 29 market uncertainty versus political uncertainty, 323 “marketable skills”, acquisition of, 71 markets, 30 emergence creating new opportunities, 199 expansion affecting different social groups, 275 expansion not self-evolving, 30 marriage age of the sent-down cohort, 142 Marshall, T. H., 5 Marxian approach to social class, 9 mass mobilization. See also political mobilizations consequences of, 20, 329 state-initiated to carry out new policy initiatives, 16 undermined by ideological crises, 329 Mateju, Peter, 73 maternity leave, 226–227 Matthews, Mervyn, 12

364

Index

Matthews, Rebecca, 29, 227 “maximally maintained inequality” (MMI) hypothesis, 72 Mayer, Karl Ulrich, 15, 126, 253, 272, 274, 335 McAdam, Doug, 127, 328 McAuley, Alastair, 13 McKeever, Matthew, 73 membership. See Communist Party membership merit-based selection criteria for education, 78 Merton, Robert K., 274 Michelson, Ethan, 11, 28, 29, 30, 31, 32, 197, 202 Micklewright, John, 304 middle class, 60, 71, 120 middle schools. See also education; high schools enrollment rate into, 48 improvement in model ﬁt for entry into, 84 rates of entry into, 76 military personnel, 56, 178, 182 Mincer, Jacob, 8, 213, 223 mixed model, estimating panel data of income, 210 mobility tables, 306 mobilization. See mass mobilization Moen, Phyllis, 15, 27, 73, 83, 102, 126, 155, 164, 198, 247, 248, 249, 274, 275, 277, 283, 285 Moore, Barrington, 12 Muller, Walter, 126, 274, 335 multi-stage sampling scheme, 51 Nathan, Andrew J., 20 Naughton, Barry, 29, 39, 250 Nee, Victor, 9, 11, 27, 28, 29, 39, 197, 200, 202, 216, 223, 227, 246, 250, 275, 309, 330, 331 negative consequences of the sent-down experience, 128 neo-traditionalist model, 165, 304, 319 “new class” theory, 12, 163 Newcomb, Theodore M., 127, 328 nonmonetary transfers, 232 nonproﬁt organizations in the public domain, 57 nonstate ﬁrms increasing status of, 261

job shifts to, 253 proportions employed in, 41 versus the state sector, 249 nonstate sector, 39 employment encouraged in, 326 types of organizations in, 44 work organizations in, 57 North, Douglass C., 32 Oberschall, Anthony, 29, 198, 202 occupational groups, 55 constructed through deliberate state policies, 22 effect on housing allocation, 238 inter-sector and inter-branch differentials, 228 in ofﬁcial statistics, 99 selected by the sent-down cohort, 145 social status of, 24 occupational hierarchies based on state redistributive institutions, 122 constructed by political processes, 24 occupational status, 55, 100–102 effect on 1993 income by cohort, 292 effect on job shifts, 268, 289 occupations. See also jobs broadly deﬁned, 101 composition of, 44 high-status, 110, 173 as indicators of political power, 259 low boundaries across within a workplace, 101 socioeconomic status of, 8, 101 Ofﬁce of Organization Committee, CCP Central Committee, 57 ofﬁce workers, 56. See also clerks composition of, 45 in government agencies, 56 higher rate of entry into the Communist Party, 182 as a supply to administrative occupations, 178 Oi, Jean C., 29, 40, 197, 332, 333 Oksenberg, Michel, 155 “old sent-down youth”, 133 on-the-job training. See work experience opportunities access to, 275 changing, 293 emergence of, 275

Index mechanisms affecting, 293 state policy shifts expanding or reducing, 21 O’Rand, Angela M., 15, 229, 246 organizational hierarchy based on state redistributive institutions, 122 constructed by political processes, 24 legitimating the state socialist system, 14 of the redistributive economy, 214, 250 in urban economies, 215 organizational locations effect on 1993 income, 292 effect on job shifts to different work organizations, 289 organizational positions, ofﬁcial ranks of positions, 161 “organizational weapons” of the communist state, 154 organizations. See also work organizations entering high-status, 110, 116 erosion of basis of redistribution, 325–327 external job shifts moving across types, 255–261 failures generated by inherent tensions, 19 internal job shifts moving within types, 261–265 ownership of, 248 promotion rates within, 191 stratiﬁcation in market societies, 15 summary of ﬁndings on returns to, 311 types of, 248 organizing capacity of vested interests, 31 Palmer, Edward, 224, 335 panel data analysis, 209, 221 versus cross-sectional data, 198 parental economic resources versus parental political status, 74 parental status, 58 contributing to children’s location in the labor force, 110 impacts on life chances over time, 23 mediating the effects of the send-down policy, 134 Parish, William L., 11, 13, 14, 19, 22, 27, 28, 29, 30, 31, 32, 36, 64, 65, 73,

365

126, 127, 197, 201, 202, 209, 302, 314, 318 Parkin, Frank, 10, 71 party secretaries (Zhenggong ganbu), 164 Passeron, Jean Claude, 72 peasants. See also workers children’s rate of entry into high school or college, 79 Peng, Yusheng, 11, 27, 227, 232, 246 People’s Republic of China (PRC), 2 changes since the mid-1990s, 332 distinctive historical periods of, 35 founding of, 35 future of, 334 labor policies and regulations in, 156 political/economic campaigns, 16, 17–18 Pepper, Suzanne, 124 periodization, 61, 215 period-speciﬁc job attainment patterns, 111–120 period-speciﬁc models, 112 permanent employment, 333. See also lifetime employment practice persistent inequality, 96 for children with “cadre” background, 95 in processes of educational attainment, 72 regarding educational attainment in industrialized societies, 95 personal income. See also income cohort variations in economic resources, 279–280 effects of the sent-down experience, 146 empirical evidence, 301 increasing disparity for sent-down youth, 151 as a small proportion of the redistributed economic beneﬁts, 227 personnel system in the Chinese bureaucracy, 156–162 Piore, Michael J., 102 pluralist model of interest politics, 318 Poisson model for housing facilities, 240 Polanyi, Karl, 7 policy implementation, 165 policy shifts. See state policy shifts political activists different cohorts of, 25 effects of shifts in state policies on, 24

366

Index

political cadres in the workplace, 155 political campaigns “4-Clean” campaign, 17 “5-Anti” movement, 17 during the 1949–1965 historical period, 37 anti-bourgeois campaign, 323 anti-criminal activities campaign, 17 Anti-Gao-Rao Clique Campaign, 17 Anti-Hu Feng Clique Campaign, 17 Anti-Peng Clique Campaign, 17 Anti-Rightist Campaign, 17–166 anti-spiritual pollution campaign, 17–329 anti-westernization campaign, 329 in the history of PRC, 16, 17–18 as an integral part of political and economic organization, 18 production enhancement and thrift campaign, 17 “san fan” movement, 17 thought reform movement, 17 “wu fan” movement, 17 political capital coexisting processes in allocating and rewarding, 201 versus human capital, 8, 200–201, 309–311 measure of, 55 political control exercising through elaborate authority relationships, 319 weakened by ideological crises, 327 political criterion in administrative recruitment and promotion, 168 political dynamics bureaucratic career patterns and, 164 in the former Soviet Union and Eastern Europe, 19 taking the form of policy ﬂuctuations or oscillations, 20 underlying sources of, 164 political input from society, stratiﬁcation dynamics and, 323 political liability, education as, 70 political logic of redistribution, 7–9, 314 political loyalty versus competence, 12 emphasized in the leadership career path, 163 as a mechanism for upward mobility, 25

premium placed on, 8 redeﬁned as adherence to reform policies, 167 rewarding, 163, 200 political managerial positions, 8 political mobilizations, 18. See also mass mobilization political opportunities of political/economic campaigns, 18 varying signiﬁcantly across historical periods, 192 political persecution during the 1966–1979 historical period, 38 political pressures enforcing the send-down policy, 129 generated by the return of the sent-down youth, 326 initiated by responses to political uncertainty, 324 political processes deciding allocative as well as redistributive priorities, 7 inducing unexpected shifts in state policies, 20 “political recommendation”, replacing entrance examination for college, 38 political screening of bureaucrats, 163 party membership as an indicator, 167 in the recruitment and promotion of political career lines, 168 political selection criteria of, 168 for educational progression, 78 historical periods associated with criteria of, 166 processes under state socialism, 13 political status cohort effects in gaining, 169 contrast with human capital, 310 importance of parental, 74 political uncertainty versus market uncertainty, 323 political/economic campaigns. See political campaigns positional power associated with the redistributive system, 280 income returns, 292 role in the second send-down period, 141 role in the transformation processes, 213

Index post-Mao era, 39. See also Mao era; reform era cadre policies, 114 changes in income determinants, 150 education as a factor for upward mobility, 24, 70 ideological crises in, 327 institutional changes, 27, 197 job shifts to nonstate organizations, 255 resumption of the college entrance examination, 82 sent-down experience effects, 146 PRC. See People’s Republic of China private entrepreneurs. See self-employment private ﬁrms, 14, 40, 44, 57 disguised as “collective” ﬁrms, 32 expansion in response to employment pressures, 153 higher returns to those working in, 215 income for employees in, 214 job shifts to, 253 largest income gains in the reform era, 209 legitimized and encouraged in 1980, 327 as nonstate organizations, 248 presence of, 331 rise in employment in the reform era, 270 role in the new income regime, 222 state policies opening in the early 1980s, 326 private housing, 234, 239, 240 private property rights in market societies, 5 privatization of housing, 244 producers conﬂict with redistributors, 199 deﬁning, 200 power relative to “redistributors”, 251 production processes, property right relations in, 9 production workers, 56 professional cohort results, 182 professional groups, advancing common interests, 6 professional hierarchy of the cadre system, 156 professional jobs, shifts to administrative jobs, 178 professional ladder, 159, 168, 183 advancement in, 161 rates of promotion, 172

367

professional occupations, entry into, 112, 171, 173–179 professionals, 56. See also intellectuals children’s rate of entry into high school or college, 79 class position of, 12 composition of, 45 concentrated in public organizations, 110 differentiating between public sector and economic sector, 210 distinctive and segmented career lines for, 164 distinguishing high-rank from rank-and-ﬁle, 12 high-rank, 44, 45, 56 integral to the Chinese “cadre” system, 156 rate of becoming, 105 rates of income for, 206 specifying into “high-rank” and “low-rank” groups, 210 “proletarianization” of society, 320 promotion patterns, 53 in the Chinese bureaucracy, 182–192 comparing bureaucratic ladders with work organizations, 170 in formal bureaucratic ranks, 183–188 in work organizations, 170, 188–192 promotions compared to job shifts, 252 policies and practice of, 159–162 types of, 183 within work organizations, 161 property rights ﬁrms closer to the state, 14 in market societies, 5 provinces, selected for data collection, 51 Psachargopoulos, George, 71, 213, 223 public organizations (shiye danwei), 43, 57 job shifts to, 253 rates of ﬁrst entry into, 106 public sector, 248 “push” and “pull” factors, 295 radical policies, 19, 167 Raftery, Adrian E., 72 rational choice explanation for maximally maintained inequality, 72 Rauch, James E., 333

368

Index

‘rebellions’ during the Cultural Revolution, 167 “the rebellious spirits”, recruitment and promotion of, 167 recruitment into the Communist Party, 179–182 in early career stages, 177 policies and practice of, 159–162 rectiﬁcation movement, 17 in literary areas, 17 in rural areas, 17 “Red and Expert” (youhong youzhuan) slogan, 90, 201 “Red Guard Rebellion”, 124, 328 redistribution, 7 bureaucratic channels in, 233 erosion of the organizational basis of, 325–327 favoring political capital, 309 institutional arrangements based on, 4 institutional logic of, 329 political logic of, 7–9, 314 under state socialism, 299–316 stratiﬁcation dynamics and, 67 through centralized authorities, 335 work organizations as the major institutions of, 201 redistributive economy beneﬁts offered by state organizations, 271 lifetime employment practice in, 169 as the main employer of the urban labor force, 44 measures of the extent of, 42 organizational hierarchy of, 213, 250 redistributive institutions, 3, 36 as constraints on job shifts, 249 evolution of, 40–49 in social stratiﬁcation, 311 thesis, 13–16 redistributive system government and public sectors at the center of, 248 mobilizing resources through nonmarket channels, 228 redistributors, 9, 12 conﬂict with producers, 199 deﬁning, 200 reﬁned periods return to urban areas of sent-down youth, 141 for the send-down analyses, 137

reform era, 39. See also post-Mao era changes and continuity in, 195 changes revealed by empirical evidence, 222 comparing with pre-reform, 209 income regime, 208 increase in economic inequality among social groups, 303 institutional transformation and stratiﬁcation processes in, 26–33 negative effect of education on entering hybrid ﬁrms, 119 series of bureaucratic reforms, 333 substantial increase in overall income, 210 “reform of the party and state”, 167 regional variations in the effects of education, 219 exploration of, 216–221 regulations and laws, affecting life course, 126 Renwei, Zhao, 304 “replacement” (dingti) policy, 307 research design, general considerations, 52–60 Research Ofﬁce, CCP Central Committee, 37 residential blocks (juweihui), 51 residential housing construction. See housing Reskin, Barbara F., 15 resource allocation historical patterns, 3 role of the Communist Party, 7 through market transactions, 247 resource transfers in state socialist societies, 22 resources allocation through a centralized authority, 7 redistributing across sectors, 22 state policies changing the meaning of, 23 restroom as a housing facility, 239 retrospective data collection, 50 reward system, favoring political status and loyalty, 8 “rightist”, 49 “risk sets”, 104 concept of, 60 constructing, 83 risk-averse attitudes, 276

Index Riskin, Carl, 233, 244, 304 risks involved in job shifts, 252 of new types of jobs, 275 Rona-Tas, Akos, 25, 27, 28, 29, 30, 32, 249, 270 Roos, Patricia A., 15 Rosen, Stanley, 18, 73 Rosenbaum, James E., 102 Roth, Louise Marie, 226 Ruan, Danching, 15 Rueschemeyer, Dietrich, 30 ruling class of the communist state, 154 running water as a housing facility, 239 rural residents versus urban, 13 rural township enterprises, 46 “rushed growth” during the 1949–1965 historical period, 36 expansion of educational institutions, 77 in industrialization, 37 resulting in economic disaster in the early 1960s, 37 Ryder, Norman B., 125 salary system formally established in 1955, 157 state-initiated reform of, 209 sampling, 51–52, 54 cities selected for, 51 distribution in selected provinces and cities, 52, 53 households selected for, 52 missing values, 64 spouses, 51 Scalapino, Robert A., 155 Schoepﬂin, Urs, 15, 126, 274 Schurmann, Franz, 1, 154, 324 Scott, James C., 322 secondary schools. See high schools sectoral structure, 248 segmented labor markets, evidence of, 289 selection criteria, reﬂecting changes in gender roles, 169 self-employment, 44, 45. See also private ﬁrms encouraged in 1980, 327 legitimizing of, 131 Selznick, Philip, 154 semi-professional jobs, 114, 159, 174 semi-state organizations, 248 semi-state sector. See nonstate sector

369

send-down episode, 131, 152, 313 send-down experience. See sent-down experience send-down policies abandonment of, 326 adverse effects on all social groups, 151 affecting the life chances of youth, 115 during the Cultural Revolution, 38 denunciation of, 130 forcing urban youth to rural areas, 107 implementation of, 127, 129, 131 institutionalization of over time, 137 of Mao Zedong, 125 targeted social groups, 129 senior governmental ofﬁcials, 157 senior high. See high schools seniority. See also age effects on Communist Party membership, 179 effects on formal promotion rates, 186 effects on work organization promotions, 189 emphasized in the Mao era, 169 factor on housing allocation in the reform era, 238 main principle in pre-reform China, 280 rates of entry into administrative professional jobs, 174 sent-down experience. See also send-down episode as cohort-speciﬁc, 125 consequence of, 142–151 effects on life course, 145 effects on personal income in the post-Mao era, 146 effects on the life course of individuals, 328 in historical context, 128–131 pivotal role in China’s economic transformation, 153 positive consequences of, 153 sent-down youth altered life courses of, 128 duration in rural areas counting as work experience, 150 economic well-being affected in the reform era, 151 opportunities to return to urban areas, 130 periods of time spent in rural areas, 130 rates of entry into ﬁrst jobs, 131

370

Index

sent-down youth (cont.) rates of returning to urban areas, 131 return of, 326 service industry, 58, 248, 249, 268 service workers, 44, 45, 56 Settersten, Richard A. Jr., 272 Shao, Yanxiang, 18 Shavit, Yossi, 72, 96 Shi, Li, 304 Shi, Weimin, 124, 130 Shi, Xiaoyan, 130 Shirk, Susan L., 18, 29, 31, 39, 166, 250 Shu, Xiaoling, 164 Shue, Vivienne, 166 Simkus, Albert, 73 skilled manual workers, 44, 45 Skilling, H. Gordon, 318 Skocpol, Theda, 30 Slovic, Paul, 276 social basis of institutional transformation, 324 social boundaries, ﬂuidity of, 22 social changes emphases given to different processes, 29 relationship to life course, 126 as a trend of transcending boundaries, 335 social classes. See also classes; social groups Marxian approach to, 9 promoting “collective projects” of class reproduction, 72 Weberian approach to, 9 social groups. See also classes advancing own interests in market societies, 6 boundaries of, 22 discriminating among by occupational category, 101 distribution of political resources among, 130 eliminating autonomous intermediate layers, 22 impacts of state policies on, 133–142 shifts in status related to state policies, 22 social inequality assessment of, 302 legitimated in the reform era, 243 social mobility channels of, 5 critical role of education in, 70 in market societies, 5

social origins alternative speciﬁcation of, 119–120 association with educational attainment, 70, 85, 307 association with job destination, 107 effects on Communist Party membership, 182 effects on promotion, 188 effects on promotion patterns, 192 effects on returning to urban areas, 152 effects on transition rates to educational levels, 85–92 impact of education entry rates, 79 impact of radical social policies, 79 measures of, 58, 82 role in children’s educational attainment, 74, 83 social policies affecting life course in market societies, 126 role in equalizing opportunities in the United States, 335 social positions ability to resist macro-political processes, 25 cohort variations in effects on job shift patterns, 290 scarcity of individual-level information on, 15 variables directly measuring, 288 social stratiﬁcation, 2 bureaucratic class as the starting point for, 154 comparative, 4, 5–6, 24 descriptions of the patterns of, 306 dynamic processes under state socialism, 313 historical approach to, 25 including macro political processes and larger historical contexts, 21 interaction between institutions affecting patterns of, 28 mechanisms of, 130 models of, 2 patterns reﬂecting institutional arrangements of a society, 3 patterns showing signiﬁcant changes, 330 redistributive institutions in, 311 under state socialism, 299–316

Index social stratiﬁcation processes changes in response to state policies, 39 changing across historical periods, 122 distinctive patterns shaped during the Cultural Revolution, 125 mediating role, 152 methodological issues in studying, 25 socialist education movement, 17 socialist managers. See managers socialist redistribution. See redistribution socialist redistributive power, 199 socialist state. See state societal changes. See social changes societal transformations, impacts on careers, 275 socioeconomic beneﬁts, measuring, 301 socioeconomic changes, shaping the urban economy, 130 socioeconomic status bureaucratic privileges and, 300–303 as a function of income and education, 101 improvements in market societies, 5 Sørensen, Aage B., 5, 6, 102, 253 Sørensen, Jesper, 5, 6 Soviet model, 7, 36, 40, 334 spouses, job shift patterns of, 285 SSB (State Statistics Bureau), 39, 41, 46, 77, 99, 202, 230, 231, 249 stable institutional bases, emphasized by the dual path model, 167 Stark, David, 8, 25, 27, 28, 30, 31, 32, 304, 305, 333 Starr, Paul, 99 state actions affecting individual life course, 126 claiming a monopoly of public goods, 320 in the coevolution of politics and markets, 31 collective deﬁance against, 321 critical role in setting up institutional rules, 30 encompassing role of, 7 impact on the opportunity structure, 73 life course institutionalized by, 274 organizational control, 325 reliance on bureaucratic organizations, 165 role in shaping life course, 126 wage policies, 101

371

State and Life Chances project, 50–52 state building process, 36 state bureaucracy. See bureaucracy “state cadres” [guojia ganbu], 103 state ﬁrms rates of ﬁrst entry into, 106 sense of entitlement of workers in, 305 types of, 43 welfare programs in, 232 state organizations, 41, 248 compared to collective ﬁrms and “other” types, 231 enjoying advantages inherited from the past, 261 evidence of signiﬁcant shifts to nonstate ﬁrms, 260 incentives to those with high human capital, 259 newly-developed market activities of, 271 proportions employed in, 41 protected by central and local authorities, 332 redistributive beneﬁts offered by, 271 web of, 40 state policies affecting all social groups, 21 altering opportunity structures and individual life chances, 20 bureaucratic resistance to, 170 changing the mechanisms of social stratiﬁcation, 123 creating incentives to stay in state organizations, 271 effect on bureaucratic promotions, 171 expanding or reducing opportunities, 123 identifying historical periods associated with, 61 impact on individual life chances, 19 impacts on job shifts, 250 impacts on the labor force over time, 253 implementing in regard to recruitment and promotion, 162 opening the private sector in the early 1980s, 326 promotion in work organizations shielded from, 189 returns to work organizations varying with, 243 sensitivity of the meanings and status of jobs to, 103

372

Index

state policies (cont.) shaping educational stratiﬁcation, 77 signiﬁcant impacts on bureaucratic career patterns, 192 on social groups, 133–142 state policy shifts, 18 consequences for individuals, 20 dynamics of, 20 effects on opportunities among bureaucrats, 164 generating variations in political opportunities, 168 impact on political opportunities, 168 political dynamics reﬂecting, 164 redrawing symbolic boundaries, 23 signaling struggles within the bureaucracy, 166 state policy-induced social changes, 152 state redistributive institutions, organizational hierarchy based on, 122 state sector, 14, 40 coexistence with nonstate, 295 composition of type of organizations in, 43 employees enjoying better welfare programs, 229 gradual shrinking of, 39 industrial outputs by, 46, 47 versus nonstate ﬁrms, 249 privatization, 333 rates of ﬁrst entry into, 105 rates of income for organizations, 207 rates of job shifts to, 253 shifts less likely to nonstate sector, 289 types of organizations in, 43 work organizations in, 57 state socialism characteristics of, 20 compared to the USSR, 312 demise of, 317 educational stratiﬁcation under, 73 evolution and decline as a political institution, 3 institutional changes in the transformation of, 3 interplay between redistribution and stratiﬁcation dynamics, 2 link between the state and life chances under, 3 mechanisms in resource allocation under, 8

mechanisms leading to collective actions, 320 political turbulence rooted in, 19 redistributing economic beneﬁts, 102 redistributing surplus centralized in the state budget, 228 redistribution and stratiﬁcation dynamics under, 7–26, 299–316 rethinking state-society relationships under, 317–330 role of society under, 318–320 social stratiﬁcation under, 2, 7, 15, 16–26, 73, 299–315, 316 stratiﬁcation as a top-down process, 22 transformation involving multifaceted processes, 29 transition from, 295 state socialist China as a model for developing countries, 315 state socialist countries, role of society in, 318 state socialist economy, stagnation of, 315 state socialist institutional structure, direct citizen linkage to the state, 22 state socialist model, movement away from, 334 state socialist planning, increasing emphasis on educational qualiﬁcations, 310 state socialist redistribution evolution of, 35–65 fundamental break away from, 39 institutional legacies of, 335 social stratiﬁcation processes and, 313 state socialist societies authority relationship in the work environment, 303 bureaucrats and bureaucratic organizations as the cornerstone of, 154 distinctive stratiﬁcation patterns in, 8 education opportunities and labor “markets” state, 73 explanations of collective actions in, 319 factions, groups, and classes in, 319 focusing exclusively on personal income, 246 greater and more uniform policy impacts, 21 political life chances in, 12

Index resource transfers, 22 stratiﬁcation in, 6 transformation of, 26 state-owned ﬁrms (guoyou qiye), 57 job shifts to, 253 pushing into market competition, 32 shifts to from government agencies, 260 state-society relationships institutional structures of, 4 stratiﬁcation dynamics and, 322–324 statistical analysis, models and methods, 60–64 status, modeling changes in, 61 status attainment model, 71, 100 status inheritance, evidence of, 114 Stone, Katherine, 99 Strang, David, 99 stratiﬁcation empirical patterns in major areas of, 40 mechanisms, 270 order, 13 outcomes, 55 in state socialist societies, 6 as a top-down process under state socialism, 22 stratiﬁcation dynamics, 3 collective inaction and, 324 concept of, 16 examining the effects of, 168 institutional transformation and, 335 large numbers phenomenon and, 322–323 lasting impacts on individuals, 313 model of, 164–167 opportunities of political input from society, 323 producing ﬂuid group boundaries, 322 as secondary consequences of state policies, 313 under state socialism, 16–26, 312–315 state-society relationships and, 322–324 undermining the organizational basis of redistribution, 325 stratiﬁed random sample, 51 structural aspects of state socialist redistribution, 19 Suhomlinova, Olga, 15, 65, 201, 223, 238, 280, 311 survey, large-scale, 50

373

symbiotic relationship of socialist managers and workers, 306 symbolic boundaries, redrawing in a politicized society, 23 ´ 1, 7, 8, 9, 11, 12, 19, 30, Szel´enyi, Ivan, 126, 154, 163, 199, 201, 228, 233, 302, 303 Szel´enyi, Szonja, 59, 73, 88 Tang, Wenfang, 27 targets of political campaigns, 18 “technical school” (jixiao), 48, 54 tensions, inherent within state socialist institutions, 20 theoretical models for analyses of bureaucratic career patterns, 192 Thomas, Robert, 32 Tilly, Charles, 6, 319, 320, 330 timing of historical events, 274 Titma, Mikk, 15 totalitarian model, 318 transformation processes, importance of incumbents and vested interests in, 29 transition rates, 77 to college at different levels of father’s education, 91 to college by period and class label, 95 for educational attainment by period and gender, 89 of entry into different educational levels, 81 ﬂuctuations associated with high school and college levels, 77 to high school at different levels of father’s education, 90 to high school by period and class label, 94 into lower educational levels, 77 measuring, 85 transitional economies interplay between politics and markets, 30 studying the unique Chinese experience, 26 Treiman, Donald J., 5, 6, 11, 12, 13, 72, 73, 96, 155, 162, 163, 201 Tsou, Tang, 23, 38 Tuma, Nancy Brandon, 15, 27, 60, 62, 73, 83, 102, 126, 155, 164, 198, 247, 248, 249, 253, 272, 275, 277, 283 Tversky, Amos, 276

374

Index

Ultee, Wout C., 6 unemployment, send-down policy as a response to, 133 unexpected ﬁndings, 295 Unger, Jonathan, 73, 124 United States Great Depression, 126 impacts of governmental policies, 335 unorganized interests, transforming into collective action, 320 unranked cadres, 56 unskilled manual workers, 44, 45 upward mobility educational attainment as a channel of, 69 in occupational categories, 108 provided by educational opportunities, 47 urban employment opportunities increased the probability of returning to urban areas, 141 increasing the return rate in the second send down period, 141 percentage change in, 133 urban job opportunities, measure of, 133 urban labor force composition across sectors, 41, 42 waves of expansion of, 107 USSR, redistribution patterns compared to China, 310 utopian socialism, impacts of ﬂirtation with, 334 vested interests advantage of, 31 inﬂuence in the remaking of institutional rules, 32 institution-based bureaucratic bargaining protecting, 202 in the transformation process, 29 victims of political campaigns, 18 “village” origin, 259 Viola, Lynne, 19 Vogel, Ezra F., 155 Voslensky, Michael, 12 Wade, Robert, 30 Walder, Andrew G., 8, 9, 11, 12, 13, 14, 15, 18, 24, 27, 29, 30, 40, 41, 64, 101, 102, 103, 120, 149, 155, 162, 163, 164, 165, 167, 168, 192, 197,

198, 200, 201, 202, 221, 227, 228, 232, 246, 250, 277, 304, 309, 310, 312, 319, 320, 325, 332, 333 Walder’s Dual Path Model. See dual path model Wang, Weihua, 130 “weapons of the weak”, 322 Weber, Max, 5, 71, 154 Weberian approach to social classes, 9 Weberian characteristics of state bureaucracies, 333 welfare beneﬁts for employees in the state ﬁrms, 232 as positive incentives for compliance, 229 welfare expenditure by type of organization, 231 welfare programs enjoyed by employees of the state sector, 229 undermined in the era of economic reform, 230 welfare state in the early 1950s, 226 in market societies, 126 rise of, 335 withering of, 333 Wesolowski, Wlodzimierz, 228 White, Lynn T. III, 22, 318 white-collar occupations, 44, 45, 102 Whyte, Martin King, 13, 14, 19, 21, 36, 38, 64, 73, 88, 126, 201, 209, 302, 314, 318 within-cohort variations, 274, 281 Wnuk-Lipinski, Edmund, 228 Wolf, Arthur, 1 Wolf, Margery, 88 work experience as a factor in the reform era, 238 as an indicator of human capital, 54, 200 measuring, 55 work organizations. See also organizations authority relationships in, 303–306 career tracks within, 162 centrality of in the stratiﬁcation order in China, 247 effect on Communist Party membership, 182 effects on housing facilities, 240 empirical studies on, 14 heads of, 162 hierarchy of, 14, 102–103

Index housing ownership and, 239 importance in China’s redistributive processes, 243 importance in the redistribution of resources, 13 increasing promotion opportunities in the reform era, 189 institutional arrangements of, 201–202 job shift events across different, 285 low job mobility across, 98 moving away from the command economy, 151 overall patterns of entry into types of, 110 promotion patterns in, 188–192 promotion within, 161, 162, 170 purchases of apartments, 244 rate of ﬁrst entry, 105 role important only in the era of economic reform, 243 role of, 102, 311 sectors of, 14 as a source of housing supply, 234 tracked by ofﬁcial statistics, 100 types of, 40, 57–58 variations in the effects of types across regions, 220 “worker/peasant” class background label, 120 workers, 60. See also peasants concentrated in state or collective ﬁrms, 110 rate of becoming, 105 rate of entry into high school or college for children of, 79 workforce, age of entry into, 280 workplace origins, 259 workplaces access to, 269 direct link with the state, 321 focusing on access to opportunities within and across, 247

375

patterns of promotion within, 183 promotion patterns in regard to seniority, 193 “workunit” (danwei), 229, 319 worldview, new, 334 Wright, Erik O., 10, 12, 303 Wu, Xiaogang, 222, 275 “Wuxunzhuan” movie, 17 Xiao, Zhenyu, 88 Xie, Yu, 27, 73, 88, 216, 221, 222, 232, 275 Xu, Songtao, 156, 161 Xu, Xiao, 329 Yamaguchi, Kazuo, 5 Yang, Dali L., 2 Yang, Zhiyun, 75, 130 Yao, Ruobing, 74, 90, 91 Yip, Kam-Bor, 96 youth (age restriction) as a selection criteria, 167 youthfulness, emphasized in the reform era, 169 Yu, Fu, 130 Zeger, Scott L., 63 Zhang, Huiping, 101, 156, 159 Zhang, Linyang, 156, 161 Zhang, Yi, 15, 229, 232, 313 Zhao, Shenghui, 37, 166, 167, 179 Zhao, Wei, 193, 223, 311 Zhao, Ziyang, 323 Zhou, M. L., 15 Zhou, Xueguang, 6, 9, 11, 15, 20, 22, 27, 28, 65, 73, 83, 102, 126, 155, 164, 167, 170, 193, 198, 201, 202, 210, 216, 223, 227, 238, 246, 247, 248, 249, 275, 277, 280, 283, 285, 311, 320, 321, 329 Zhou, Zeizhou, 15 Zukin, Sharon, 12