LATINO COMMUNITIES
edited by A N T O I N E T T E SEDILLO LOPEZ U N I V E R S I T OYF N E WM E X I C O
A GARLAND SERIES...
10 downloads
364 Views
5MB 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
LATINO COMMUNITIES
edited by A N T O I N E T T E SEDILLO LOPEZ U N I V E R S I T OYF N E WM E X I C O
A GARLAND SERIES Copyright 2001 by Stephanie Bohon
LATINOS IN ETHNIC ENCLAVES
STEPHANIE BOHON
GARLAND PUBLISHING, INC.
N E WY O R K & L O N D O N/ 2 0 0 1 Copyright 2001 by Stephanie Bohon
Published in 2001 by Garland Publishing Inc. 29 West 35th Street New York, NY 10001 Garland is an imprint of the Taylor & Francis Group Copyright O 2001 by Stephanie Bohon All rights reserved. No part of this book may be reprinted or reproduced or utilized in any form or by any electronic, mechanical, or other means, now known or hereafter invented, including photocopying and recording, or in any information storage or retrieval system, without permission in writing from the publishers.
Library of Congress Cataloging-in-Publication Data Bohon, Stephanie. Latinos in ethnic enclaves : immigrant workers and the competition for jobs / Stephanie Bohon. p. cm. - (Latino communities) Includes bibliographical references and index. ISBN 0-8153-3765-5 (alk. paper) 1. Hispanic Americans-Employment. 2. Hispanic Americans. I. Title. 11. Series. HD8081.H7 B64 2000 331.6'238073-dc21 00-039314
Printed on acid-free, 250-year-life paper Manufactured in the United States of America Copyright 2001 by Stephanie Bohon
To Gordon
Copyright 2001 by Stephanie Bohon
Contents
List of Tables Acknowledgments 1. 2. 3. 4. 5. 6. 7. 8.
Latino Immigrants and Ethnic Enclaves Segmented Assimilation and Queuing Ethnic Enclaves and Enclave Theory Data and Methods Occupational Attainment within Latino Enclaves Discrimination and Competition among Long-Term Immigrants Discrimination and Competition among Recent Immigrants Conclusions
Appendices Notes Bibliography
Copyright 2001 by Stephanie Bohon
List of Tables
1. Ethnic Enclaves in the United States 2. Sample by Place of Origin by Place of Work in a Latino Enclave (Long-Term Immigrants) 3. Sample by Place of Origin by Place of Work in an Asian Enclave (Long-Term Immigrants) 4. Sample by Place of Origin by Place of Work in a Non-Enclave Location (Long-Term Immigrants) 5. Sample by Place of Origin by Place of Work in a Latino Enclave (Recent Immigrants) 6. Sample by Place of Origin by Place of Work in an Asian Enclave (Recent Immigrants) 7. Sample by Place of Origin by Place of Work in a Non-Enclave Location (Recent Immigrants) 8. Percent of Long-Term Immigrants in Commonly Held Occupations by Place of Origin 9. Percent of Recent Immigrants in Commonly Held Occupations by Place of Origin 10. Average Male Socioeconomic Index Scores (and Standard Deviations) by Place of Origin by Place of Work (Long-Term Immigrants) 11. Average Male Socioeconomic Index Scores (and Standard Deviations) by Place of Origin by Place of Work (Recent Immigrants) 12. Independent Variables Used in This Study 13. Descriptive Statistics for the Jersey City PMSA (Long-Term Immigrants)
Copyright 2001 by Stephanie Bohon
List of Tables 14. Descriptive Statistics for the Los Angeles-Long Beach PMSA (Long-Term Immigrants) 15. Descriptive Statistics for the Miami-Hialeah PMSA (Long-Term Immigrants) 16. Descriptive Statistics for the New York City PMSA (Long-Term Immigrants) 17. Impact of Country of Origin and Other Endowments on Occupational Attainment (MSEI) for Long-Term Latin American Immigrants in Latino Enclaves 18. Descriptive Statistics for the Miami-Hialeah PMSA (Recent Immigrants) 19. Descriptive Statistics for the New York City PMSA (Recent Immigrants) 20. Impact of Country of Origin and Other Factors on the Occupational Attainment (MSEI) of Recent Latin American Immigrants in Latino Enclaves 21. Summary of the Theoretical Implications of the Empirical Findings 22. Descriptive Statistics by Place of Origin for Long-Term Immigrant Workers in Latino Enclaves 23. Descriptive Statistics by Place of Origin for Long-Term Immigrant Workers in Asian Enclaves 24. Descriptive Statistics by Place of Origin for Long-Term Immigrant Workers in Non-enclave Areas 25. Parameter Estimates and Standard Errors of the OLS Model Regressing Occupational Attainment (MSEI) on Place of Work (Long-Term Immigrants) 26. Decomposition of Average Differences in MSEI between LongTerm Mexicans and Other Latin American Immigrants by Place of Work 27. Descriptive Statistics by Place of Origin for Recent Immigrants Working in Latino Enclave Areas 28. Descriptive Statistics by Place of Origin for Recent Immigrants Working in Asian Enclave Areas 29. Descriptive Statistics by Place of Origin for Recent Immigrants Working in Non-Enclave Areas 30. Impact of Place of Work and Other Factors on the Occupational Attainments (MSEI) of Recent Immigrants
Copyright 2001 by Stephanie Bohon
List of Tables Decomposition of Mean Differences in MSEI between Recent Mexican and Other Latin American Immigrants by Place of Work Sample Comparison by Place of Work and Place of Residence in Latino Enclaves (Long-Term Immigrants) Sample Comparison by Place of Work and Place of Residence in Asian Enclaves (Long-Term Immigrants) Sample Comparison by Place of Work and Place of Residence in a Non-Enclave Location (Long-Term Immigrants) Sample Comparison by Place of Work and Place of Residence in a Latino Enclave (Recent Immigrants) Sample Comparison by Place of Work and Place of Residence in an Asian Enclave (Recent Immigrants) Sample Comparison by Place of Work and Place of Residence in a Non-Enclave Location (Recent Immigrants) Regression Coefficients (and Standard Errors) by Place of Origin for Latino Enclave Workers Regression Coefficients (and Standard Errors) by Place of Origin for Asian Enclave Workers Regression Coefficients (and Standard Errors) by Place of Origin for Non-Enclave Workers Regression Coefficients (and Standard Errors) by Place of Origin for Latino Enclave Workers Regression Coefficients (and Standard Errors) by Place of Origin for Asian Enclave Workers Regression Coefficients (and Standard Errors) by Place of Origin for Non-Enclave Workers
Copyright 2001 by Stephanie Bohon
Acknowledgments
I want to thank several individuals and organizations who made this project possible. Foremost among them is Gordon De Jong to whom this work is dedicated. He read and commented extensively on each chapter and provided valuable advice on shaping the manuscript. Other scholars who reviewed this piece in its various stages, offered suggestions for reorganization, and helped to strengthen the methodology include Leif Jensen, Craig Humphrey, and Barry Lee, Penn State University; Dan Lichter, Russell Sage Foundation; and Michael Erickson, Albertson College. My colleagues who shared resources and offered statistical consultation include Melonie Heron, Florida State University; and Julie Kraut, Centers for Disease Control. Additionally, I received financial support from the Population Research Institute and the sociology departments at Penn State, Ohio University, and the University of Georgia. Several of the staff at the Population Research Institute at Penn State University also contributed to this work by providing technical assistance. I owe a special debt to the members of the Computer Core, especially Paul Riggs, Joe Broniszewski, Don Genismore, and Michael Zimmerman. I am also grateful for Lisa Broniszewski and the other members of the Information Core. Many friends also supported me throughout the research and writing phase of this project. Bridget Gorman read much of this work, offered advice, and provided counseling and moral support throughout the entirety of this endeavor. Rod Crossland, Caroline Kapinus, and Chris Calienes listened to my frustrations and offered encouragement as did my parents, Don and Marianne Bohon. Finally, I am especially grateful to my husband, Michael Jarrett, who never minded my late hours, moved me across the country (twice), and provided intellectual and spiritual guidance throughout the writing of this work.
Copyright 2001 by Stephanie Bohon
...
Xlll
Latinos in Ethnic Enclaves
Copyright 2001 by Stephanie Bohon
CHAPTER ONE
Latino Immigrants and Ethnic Enclaves
The changing composition of immigrant flows to the United States in the last half century has forced sociologists to reevaluate classical assimilation theory. Many researchers today doubt the ability of Latin American immigrants and their offspring to move through the same educational and occupational channels that Eastern European immigrants did in order to achieve economic success commensurate with average white Americans (Alba, 1990; Borjas, 1986; Denton & Massey, 1988; Glazer, 1993; Kao & Tienda, 1995; Portes & Bach, 1985; Portes & Zhou, 1995; Waldinger, 1993; Wallace, 1986). As a consequence, scholarly interest is increasing in adaptation barriers and alternative forms of economic attainment for these immigrants. With the exception of Cubans, Latin American immigrants in the U.S. face a life-long disadvantaged economic position relative to the native, non-Latino white population. While some would argue that the reduced success of these immigrant groups has not been demonstrated (Chiswick, 1980), more recent work refutes this notion (Bor~as,1991; Gurak, 1987). While unemployment is low among most Latin American immigrant groups, the jobs many take in the U.S. result in significant downward mobility (Stepick & Portes, 1986; Waldinger, 1997a). Some economists (e.g., Borjas, 1985, 1986, 1991), contend that the seeming inability of new immigrants to readily adapt to the U S . economy is directly related to changes in the quality of immigrant streams. Many scholars recognize that recent immigrants may have more difficulty gaining good jobs in the U S . market because they lack the job experience, education, and training necessary to compete with native workers (Aldrich & Waldinger, 1990; Borjas, 1985, 1986, 1991; Glazer, 1993; Light, 1984; Portes & Zhou, 1992; Tienda & Wilson, 1992; Waldinger, Ward, & Aldrich, 1985). It has been argued that these same
Copyright 2001 by Stephanie Bohon
4
Latinos in Ethnic Enclaves
limitations also restrict opportunities for self-employment by limiting access to capital (Borjas, 1986, 1990; Light, 1984). These arguments, which emphasize a human capital approach to adaptation, have been called into question by more recent research that brings new data to bear on the contention that individual characteristics can account for all of the difference between native and immigrant economic attainment (Castile, 1994; Duleep & Regets, 1996; Evans, 1989; Lindstrom & Massey, 1994; Massey, 1995; Min, 1987; Sanders & Nee, 1996; Waldinger, 1989). In particular, the general view that recent migrants are disproportionately drawn from an unskilled work force or that job skills acquired in the sending country wholly predict future mobility has been under fire lately. For example, two recent studies of the Latin American immigrant population show that it is surprisingly diverse and includes entrepreneurs, professionals, technicians and other skilled workers (Duleep & Regets, 1996; Portes & Grosfoguel, 1994). At the same time, studies such as Wallace's (1986) examination of Central Americans in California demonstrates that even immigrants with relatively high human capital may have difficulty achieving occupational success beyond that typically experienced by minority groups. As a consequence, more researchers are eschewing the human capital framework to explore structural factors such as race, housing opportunities, and industrial diversity that have already proven to be important in the adaptation process of Cuban and Mexican immigrants (e.g., Portes & Zhou, 1992; Wilson & Martin, 1982). One of the areas of scholarly interest in structural adaptation mechanisms is the ethnic enclave, seen by many researchers to offer an alternative route to immigrant adaptation around structural barriers. For the purposes of this research, an ethnic enclave is a metropolitan area characterized by a concentration of businesses owned and operated by immigrants from the same country of origin, or their direct descendants (Logan, Alba & McNulty, 1994).' Immigrants often fare well in enclaves because most jobs do not require good English language skills or high levels of education (Bailey & Waldinger, 1991; Logan, Alba & McNulty, 1994; Min, 1987; Waldinger, 1989; Wilson & Martin, 1982). Enclaves also offer many opportunities for self-employment, since there is local demand for ethnically defined goods, and immigrants have access to lending capital not readily available elsewhere (Aldrich & Waldinger, 1990; Evans, 1989; Kim, Huhr, & Fernandez, 1989; Light, 1984; Min, 1987; Portes & Bach, 1985; Portes & Stepick, 1993). Current enclave studies, however, focus on those ethnic groups that own or work in the largest proportion of minority businesses. Almost no attention has been paid to other enclave workers who share a native language and a similar culture, economic background, and immigration experience with the dominant
Copyright 2001 by Stephanie Bohon
Latino Immigrants and Ethnic Enclaves
5
enclave group but have emigrated from a different c o ~ n t r y .In~ short, the importance of ethnicity in an ethnic enclave has not been directly assessed. The current research that does focus on ascribed status relationships in the enclave has examined intra-group differences between immigrants from the same sending country (Kim, Huhr & Fernandez, 1989; Light, Sabagh, Bozorgmehr, and Der-Martirosian, 1993), racial differences among immigrants (Portes & Grosfoguel, 1994; Stepick & Portes, 1986), and differences between immigrants and native blacks (Portes & Zhou, 1992, 1993; Wilson & Martin, 1982). These studies have focused on racial discrimination (e.g., the attainment of black Caribbeans versus African-Americans) and class exploitation among immigrants who share the same country of origin (e.g., competition among different Iranian sub-groups). While these studies extend our knowledge of ethnic enclaves, sociologists have not examined the impact of inter-ethnic interactions on the economic well-being of non-coethnic workers (Model & Ladipo, 1996) .3 Another issue that has been left largely unexplored in the enclave literature is occupational attainment. As more and more immigrants enter the United States, job availability has become a primary concern of policy makers (Model & Eadipo, 1996). The issue of good jobs and who gets them becomes particularly important as many new unskilled workers enter labor markets making the transition from manufacturing to service economies (Waldinger, 1997b). Questions left unanswered by current ethnic enclave and occupational attainment studies include: what is the importance of enclave context on the occupational attainment of Latin American immigrants from various sending countries? Does the enclave promote economic adaptation for all immigrant groups or just for the dominant enclave group? What impact does time since immigration have on the labor market status of non-coethnic workers? Does time since immigration affect occupational attainment differently for workers who share a country of origin with the dominant enclave group and those who do not? Does the national origin of the dominant enclave group affect the occupational attainment of non-coethnic workers? Does the national origin of the non-coethnic group affect the economic adaptation of workers inside the enclave regardless of their human capital, household characteristics, and other endowments? Do the occupational attainment patterns of immigrants in same language, similar culture enclaves vary from the patterns of occupational attainment for immigrants in other types of ethnic enclaves? Finally, can noncoethnic immigrants who share traits with the dominant enclave group expect to acquire jobs similar to those of the dominant enclave group, or will they be retained in lower status occupations?
Copyright 2001 by Stephanie Bohon
6
Latinos in Ethnic Enclaves
The research proposed here will fill the gap in the existing literature on ethnic enclave economies by exploring the outcomes of coethnic versus noncoethnic workers. It will also explore occupational attainment within the enclave context.
BACKGROUND AND THEORY The formulation of hypotheses regarding the occupational attainment of noncoethnic workers in enclave and non-enclave economies necessitates an examination of both the development of enclave literature and an examination of two competing theories regarding immigrant economic adaptation: segmented assimilation and queuing theory. Segmented assimilation, as it was originally conceived by Portes and Zhou (1993), is based on the idea that different immigrants groups are likely to take on the characteristics of native ethnic and racial groups to which they are the most similar rather than mirroring nonLatino, white Americans. Haitian immigrants, for example, would become more like African-Americans as they adapt to the U S . economy over time. Several studies have demonstrated this segmented assimilation phenomenon for secondgeneration immigrants (Denton & Massey, 1988; Fernandez-Kelly & Schauffler, 1994; Gans, 1992; Kao & Tienda, 1995; Portes & Zhou, 1993). Segmented assimilation may pose the threat of downward mobility for some Latino groups (Rumbaut, 1997). Wallace (1986), for example, shows that Central American immigrants to California earn wages roughly equivalent to Mexicans, despite the fact that Central American immigrants, on average, have much higher human capital, including more white collar work experience and relatively higher levels of education. Under the theory of segmented assimilation, immigrants who share the same human capital characteristics as enclave workers should eventually demonstrate employment patterns similar to the dominant ethnic group from the same sending continent or broad geographic region. In practical terms that means that, in the case of Miami for example, all Latin American immigrants will become more like Cubans over time, and those groups with the highest levels of human capital (i.e., those most like the Cubans, such as Nicaraguans) will adapt at an accelerated pace.4 While segmented assimilation is one possible outcome for Latinos in an ethnic enclave, another possibility is queuing. In the simplest terms, queuing theory suggests that when a group is relatively large, it dominates the occupational hierarchy. Since there are a finite number of high prestige Qobs, when one of these jobs becomes available, it will be filled by another member of the dominant group. In short, good jobs are rationed according to a queuing mechanism, and where a person is in the queue depends largely on ascribed
Copyright 2001 by Stephanie Bohon
Latino Immigrants and Ethnic Enclaves
7
characteristics (Rich, 1995; Roos & Jones, 1993; Roos & Reskin, 1992; Sakamoto & Chen, 1991; Sakamoto & Powers, 1995).5 While queuing theory has generally been used to explain barriers to employment and occupational attainment for women and racial minorities (Roos & Jones, 1993; Rich, 1995; Sakamoto & Chen, 1991), there is no reason to think that queuing would not occur to immigrants in an enclave situation. Indeed, some argue that the enclave offers opportunities for exploitation unparalleled in other labor markets (Bonacich, 1987). Since, by definition, one ethnic group has supremacy in an enclave setting, there is a reasonable chance that members of that ethnic group will discriminate in favor of their fellow expatriates. It is possible that escaping the enclave offers few better options for immigrants, however, since queuing may also occur in the wider local labor market, as well. Queuing theory suggests that country of origin is the most important characteristic for determining labor market status in an ethnic enclave; therefore, there should be marked differences between the occupational attainment patterns of the dominant enclave group and other immigrant groups. Findings that support segmented assimilation, however, will show that non-coethnic workers in the enclave experience occupational attainment levels similar to the dominant enclave group with country of origin showing little or no effect once human capital, household characteristics, and other factors are taken into account.
HYPOTHESES Hypotheses regarding the effects of the enclave on the economic adaptation of non-coethnic immigrants are markedly different depending upon whether they are predicated on segmented assimilation or queuing theory. Segmented assimilation suggests inter-ethnic cooperation hypotheses, whereby immigrants who share a native language, similar culture, comparable economic backgrounds, and similar immigration experiences will hire and preferentially promote one another regardless of differing countries of origin. Cuban business owners in Miami, for example, will make hiring decisions based on human capital factors from among a pool of all Latin American immigrants, not just other Cubans. Queuing theory, on the other hand, suggests a set of inter-ethnic competition hypotheses, since country of origin has a profound effect on where immigrants are in the job queue. Members of the dominant enclave group, according to queuing theory, will always prefer to hire and promote immigrants from their own country of origin, giving little or no consideration to job experience or other human capital factors. Where other immigrants fit into the queue, then, depends largely upon how workers from their country of origin are viewed by members of the dominant group. Since there is no reason, a priori, to accept queuing Copyright 2001 by Stephanie Bohon
Latinos in Ethnic Enclaves
8
theory in favor of segmented assimilation, or vice-versa, a number of competing hypotheses are postulated. Adaptation Within Latino Enclaves The first set of hypotheses relevant to this study examine the question of whether or not national origin makes a difference to the economic adaptation of immigrants. That is, the first question relates to the occupational attainment of other immigrants relative to the dominant enclave group. Because assimilation is a lengthy process, the hypotheses pertain to immigrants who have been in the U S . for at least half a decade. Question: Will Latino immigrants working in metropolitan areas with Latino enclaves demonstrate lower than or the same occupational attainment as immigrants from the dominant enclave group after living in the United States for more than five years? Hypothesis H l a . After more than five years in the U S . , Latino immigrants working in metropolitan areas with Latino enclaves will demonstrate significantly lower occupational attainment, on average, than members of the dominant enclave group. Alternative Hypothesis H l b . After more than five years in the U.S., noncoethnic immigrants working in metropolitan areas with Latino enclaves will demonstrate average occupational attainment not significantly different from the dominant enclave group.
These alternative hypotheses suggest, on the one hand, that Latino enclaves create areas where Latin American immigrants can benefit from unique labor market opportunities, and the spirit of interethnic cooperation and other adaptation processes lead to segmented a ~ s i m i l a t i o n .This ~ possibility is captured in hypothesis Hlb. On the other hand, the presence of a large Latin American labor market controlled by members of one ethnic group may lead to interethnic competition where national origin differences are exacerbated, and members of the dominant enclave group may create a labor market queue giving preferential positions to members of their own group. This possibility is captured in hypothesis Hla. While enclave theory, segmented assimilation, and queuing allow for the clear development of testable hypotheses regarding immigrant adaptation for immigrants who have been in the labor market for some time, developing hypotheses about more recent immigrants is more difficult. It is possible that Copyright 2001 by Stephanie Bohon
Latino Immigrants and Ethnic Enclaves
9
since both coethnic and non-coethnic workers offer a cheap source of needed labor in the short-run, there is no difference between their employment outcomes. That is, employers may choose immigrant workers based on their human capital alone. On the other hand, employers may preferentially hire members of their own country of origin group either out of a sense of nationalistic loyalty or because these workers are most likely to be known or related to other workers at that place of employment. The first possibility suggests a segmented assimilation outcome, while the second suggests queuing. Given these possible divergent outcomes, alternative hypotheses are postulated concerning the occupational attainment of recent immigrants. Question: Will Latino immigrants working in metropolitan areas with Latino enclaves demonstrate lower than o r the same occupational attainment as immigrants from the dominant enclave group if they have resided in the U.S. for five years or less? Hypothesis H2a. Recent Latin American immigrants (those entering the country in 1985 or later) working in metropolitan areas with Latino enclaves will report occupational attainment significantly lower, on average, than that of recent Immigrants in the dominant enclave group. Alternative Hypothesis H2b. Recent Latin American immigrants working in metropolitan areas with Latino enclaves will report occupational attainment not significantly different from that of recent immigrants in the dominant enclave group.
Adaptation Across Labor Market Contexts The analysis undertaken here also lends itself to an examination of the impact of varying labor market contexts on the occupational attainment of immigrants. Latin American immigrants do not have to work in Latino enclave areas. They may, for example, choose to work in Chinese or Japanese enclave areas, or they may choose to work in areas without ethnic enclaves. The outcome of such a destination choice for some immigrant outcomes has been strenuously debated (e.g., see Jensen & Portes, 1992; Portes & Jensen, 1992; Sanders & Nee, 1992). Two sets of competing hypotheses about the economic outcome of both longterm and recent immigrant in these different types of contexts are postulated. Question: Will Latino immigrants working in metropolitan areas with Latino enclaves demonstrate higher or the same occupational attainment as Latino
Copyright 2001 by Stephanie Bohon
Latinos in Ethnic Enclaves immigrants working in Asian enclaves and non-enclave areas after living in the U.S. for more than five years? Hypothesis H3a. Comparing across economic enclave and non-enclave contexts, Latin American immigrants in metropolitan areas with Latino enclaves will have higher occupational attainment after five years in the U.S. than Latin American immigrants in areas with Asian enclaves or non-enclave areas, holding constant measurable endowment characteristics. Alternative Hypothesis H3b.Comparing across economic enclave and nonenclave contexts, Latin American immigrants in metropolitan areas with Latino enclaves will have occupational attainment after five years in the U S . not significantly different from Latin American immigrants in areas with Asian enclaves or non-enclave areas, holding constant measurable endowment characteristics. Question: Will recent Latino immigrants (residing in the U S . for five years or less) working in metropolitan areas with Latino enclaves demonstrate higher or the same occupational attainment as Latino immigrants working in Asian enclaves and non-enclave areas? Hypothesis H4a.Comparing across economic enclave and non-enclave contexts, recent Latin American immigrants in metropolitan areas with Latino enclaves will have higher occupational attainment than Latin American immigrants in areas with Asian enclaves or non-enclave areas, holding constant measurable endowment characteristics. Alternative Hypothesis H4b.Comparing across economic enclave and nonenclave contexts, recent Latin American immigrants in metropolitan areas with Latino enclaves will have occupational attainment not significantly different from Latin American immigrants in areas with Asian enclaves or non-enclave areas, holding constant measurable endowment characteristics.
Hypothesis H3a and H4a suggests that, for members of the dominant enclave group, the enclave offers advantages not found elsewhere. These immigrants can obtain jobs and job experience without a waiting period to learn English, establish employment networks, and learn the cultural skills necessary for a successful job search in the mainstream economy (Bailey & Waldinger, 1991; Logan, Alba & McNulty, 1994; Portes & Bach, 1995; Portes & Stepick, 1993; Waldinger, 1993). Non-coethnic workers, however, do not share these
Copyright 2001 by Stephanie Bohon
Latino Immigrants and Ethnic Enclaves
11
advantages. Hypotheses H3b and H4b suggests that the Latino enclave equally benefits all Latin American immigrant groups. This may be because discrimination in the larger labor market leads to the creation of a pan-ethnic identity, Latino, which supersedes national origin distinctions. The ethnic enclave may promote nationalism or a heightened awareness of intragroup differences. Members of the dominant enclave group may view their place and people as superior to that of other Latin Americans, so they may discriminate against them more than would otherwise be the case. The consequence of this would be increased interethnic competition (or greater queuing). The ethnic enclave may, however, create opportunities for all Latin American immigrants, and those immigrants may experience less barriers to attainment than in other labor market contexts. In order to test this, it is possible to compare the average amount of difference in occupational attainment due to discrimination across the various labor market contexts for both recent and longterm immigrant^.^ Question: Will longer-term Latino immigrants (i.e., those living in the United States for more than five years) working in metropolitan areas with Latino enclaves experience less or the same amount of discrimination in job hiring and promotion as Latino immigrants working in Asian enclave or non-enclave areas? Hypothesis H5a. Comparing across economic enclaves and non-enclave contexts, discrimination as a determinant of occupational attainment differences among longer term Latin American immigrant groups is lower in metropolitan areas with Latino economic enclaves and greater in areas with Asian enclaves and non-enclave labor markets. Alternative Hypothesis H5b.Comparing across economic enclaves and nonenclave contexts, discrimination as a determinant of occupational attainment differences among longer-term Latin American immigrant groups is no different in metropolitan areas with Latino economic enclaves than in areas with Asian enclaves and non-enclave labor markets. Question: Will recent Latino immigrants (i.e., those living in the United States for five years or less) working in metropolitan areas with Latino enclaves experience less or the same amount of discrimination in job hiring and promotion as recent Latino immigrants working in Asian enclave or nonenclave areas?
Copyright 2001 by Stephanie Bohon
Latinos in Ethnic Enclaves Hypothesis H6a. Comparing across economic enclaves and non-enclave contexts, discrimination as a determinant of occupational attainment differences among recent Latin American immigrants is lower in metropolitan areas with Latino economic enclaves and greater in areas with Asian enclaves and nonenclave labor markets. Alternative Hypothesis H6b.Comparing across economic enclaves and nonenclave contexts, discrimination as a determinant of occupational attainment differences among recent Latin American immigrants is no different in metropolitan areas with Latino economic enclaves than in areas with Asian enclaves and non-enclave labor markets.
CONTRIBUTIONS TO PRIOR RESEARCH AND THEORY This work examines occupational attainment as a form of economic adaptation among non-black Latin American immigrant workers in metropolitan areas with Latino and Asian enclaves and in selected metropolitan areas without enclaves. Different national origin groups are compared within and across labor markets. The importance of this examination and comparison is that it offers valuable contributions to the scientific and public policy debate concerning Latin American immigrants' economic adaptation in several ways. First, existing research and theory on the enclave as an adaptive mechanism is extended by exploring its effects on the occupational attainment of immigrant workers who do not share a national origin with the dominant enclave group. Second, a gap is filled in comparative attainment studies that examine the effects of ethnicity on occupational attainment but do not assess the impact of the enclave economy on these outcome^.^ Third, a comparison of segmented assimilation and queuing theory is offered as a way of determining the best model for exploring economic outcomes for Latin American immigrants in the United States. This study contributes to existing research on Latin American immigrant adaptation, in general, and the ethnic enclave, in particular. At the most general level, this study offers an examination of the importance of national origin on occupational attainment in metropolitan areas with ethnic enclaves among immigrants from the various sending areas in Latin America by examining, for the first time, the impact of inter-ethnic competition versus inter-ethnic cooperation on a vital area of immigrant adaptation. By comparing the occupational attainment patterns of immigrants from Mexico, Cuba, Nicaragua, the rest of the Central American isthmus, Colombia, and the rest of mainland South America valuable information can be gathered about competition and cooperation in labor force processes, irrespective of the Copyright 2001 by Stephanie Bohon
Latino Immigrants and Ethnic Enclaves
13
impact of human capital and household characteristics. Specifically, this study offers a unique opportunity to directly test the theory that the enclave eliminates the need for English fluency and high levels of education for economic success. This study also offers new insight into issues of discrimination and exploitation within the enclave context. Bonacich (1988) has argued that the enclave is a potential source of exploitation of both family members and other co-ethnic immigrants as workers become dependent on the enclave and fail to learn the skills needed for success in the mainstream (and presumably higher paying) labor market. At the same time, Portes & Zhou (1993) have argued that segmented assimilation should be the expected outcome for new immigrant groups. This study offers a better understanding of the potential for discrimination and exploitation of other immigrant groups within the enclave. More importantly, it tests the competing hypotheses that immigrants are more or less exploited in the non-enclave market as opposed to the enclave. Finally, this study contributes to the understanding of ethnic enclaves by offering a direct test of labor market effects on occupational attainment by examining occupational attainment outcomes in two different types of enclaves (Latino and Asian) as well as the non-enclave context. This is a unique contribution to the literature that, to date, has focused almost exclusively on wages and the transition to self-employment within a single enclave setting.
Copyright 2001 by Stephanie Bohon
CHAPTER TWO
Segmented Assimilation and Queuing
Recent studies of income, occupational mobility, occupational attainment, home ownership and unemployment suggest that some groups in the U.S. do not assimilate (i.e., demonstrate characteristics similar to white Americans) even after considerable time and many generations, while some groups assimilate quickly (Alba & Logan, 1992; Portes & Zhou, 1993; Rumbaut, 1997). Groups like the Chinese and the Cubans appear to be particularly advantaged on some measures of incorporation while Puerto Ricans and Vietnamese, for example, appear particularly disadvantaged even after socioeconomic and human capital factors are taken into consideration (Alba & Logan, 1992). Furthermore, the conditions necessary for assimilation--such as economic expansion and reduction in immigrant flows--are less apparent in the U S . in recent years (Lindstrom & Massey , 1994). Occupational outcomes are a particularly important indicator of immigrant adaptation (Alba & Nee, 1997). If the environment they enter is typified by interethnic competition, immigrants from certain countries of origin may have more difficulty obtaining good jobs in the marketplace. If, however, the environment is typified by interethnic cooperation, national origin will not matter, and all immigrants can expect to do equally well. These alternative outcomes are posited based on segmented assimilation and queuing theory. However, the use of segmented assimilation to describe the experience of first generation immigrants and application of queuing theory to immigrant labor markets is an atypical use of these theories. Consequently, a further examination of both schools of thought is necessary in order to create a better understanding of the ways in which these theories can be applied to first generation immigrants' occupational attainment patterns.
Copyright 2001 by Stephanie Bohon
16
Latinos in Ethnic Enclaves
SEGMENTED ASSIMILATION AND QUEUING THEORY The concept of segmented assimilation was first introduced by Alejandro Portes and Min Zhou in 1993, and it has been largely attributed to Zhou (DeWind & Kasinitz, 1997). Its development grew largely out of concern that new immigrants (i.e., Asians and Latin Americans) would not be incorporated into the U S . economic and social systems in the same way that Eastern European immigrants arriving at the turn of the century were (Alba & Logan, 1992). This fear was coupled with evidence that economic restructuring seemed to impact the assimilation experience of different groups in different ways (Alba & Nee, 1997; Portes & Zhou, 1993). This did not appear to bode well for newer immigrants. From these concerns stemmed fear of "second generation decline," or concern that the children and grandchildren of immigrants would fare worse economically than their parents due to the adoption of American attitudes coupled with barriers to minority attainment (Gans, 1992). Second generation decline is seen as inevitable by some because of the increased propensity of the first generation to maintain transnational business ties, the restructuring of the labor market, and widespread discrimination (Portes, 1997). This decline reflects the fact that some groups do not have the ability to shed ethnic identifiers over time, particularly if the identifiers are ascribed traits (Portes & Zhou, 1993). At the same time, as successive generations are exposed to American work ethics and American ideals, they are less willing to work in the low skilled, low paying industries that employed their parents or grandparents. An extreme typology of segmented assimilation was first offered by Portes and Borocz (1989) when they argued that the long-term outcomes for immigrants depended largely on conditions of exit, class origins, and reception contexts, and that the outcomes of the intersection of these three factors was placement into various market sectors including ghetto services, the mixed labor market, ethnic enclaves, mainstream small businesses and middleman minority markets. Four years later, Portes and Zhou (1993) published a more conservative typology of segmented assimilation where they argued that these same factors-exit conditions, class, and entry context--determined only the propensity for a group to experience straight-line assimilation, downward assimilation, or continued separation. In the simplest terms, the theory of segmented assimilation is an attempt to explain the differences between groups that eventually experience upward mobility (i.e., straight-line assimilation), groups that experience downward mobility over time, and groups that experience upward mobility, but only by maintaining strong ties to their cultural origins (Alba & Nee, 1997). Essentially, the theory indicates that immigrants tend to be incorporated into those native Copyright 2001 by Stephanie Bohon
Segmented Assimilation and Queuing
17
groups they are most similar to. If the group they are most similar to is disadvantaged in the marketplace, then downward assimilation occurs. However, some groups have shown themselves to be particularly resistant to this and are able to maintain separation from the native group by consciously isolating their group and maintaining strong ties to the language and cultural practices of their native land (DeWind & Kasinitz, 1997; Rumbaut, 1997). Segmented assimilation assumes that different groups are incorporated into U S . society at different rates, through different means, and to different degrees (DeWind & Kasinitiz, 1997), and these differences do not stem from differential levels of human capital between groups, as some economists would suggest (Borjas, 1990; Haberfield, 1993). Segmented assimilation, in short, explains how new groups enter the stratification system (Zhou, 1997). Since immigrants today are more geographically concentrated and linguistically isolated than ever before (Massey, 1995), segmented assimilation offers a plausible model of immigrant incorporation. Segmented assimilation, however, is not a sure bet, since it ignores the salience and persistence of discrimination in all levels of our society (DeWind & Kasinitz, 1997). One assumption of segmented assimilation is that similar groups cannot be differentiated. For example, Salvadoran or Guatemalan immigrants in California are indistinguishable from U.S.-born Mexican-Americans, consequently, the immigrants will be incorporated into an amorphous Latino group (as shown by Wallace, 1986) or all immigrants will be assumed to be members of the largest group (in this case, Mexicans). As a consequence of this assumption, segmented assimilation has tended to focus on the adaptation of pan-ethnic groups such as Latinos, Asians, and blacks, rather than their national origin subgroups. However, many second and third generation Latinos and Asians living in the U S . today believe that first generation immigrants negatively impact their personal well-being (Johnson, Farrell & Guinn, 1997). And, since migration streams tend to be selective of place, they can threaten to change the class and ethnic character of a community (Clark, 1982; Johnson, Farrell, & Guinn, 1997; Tienda & Wilson, 1992). This threat or perceived threat can lead to discriminatory behavior by the established minority group who wants to maintain their way of life (Waters & Eschbach, 1995). One problem with the theory of segmented assimilation is that it discounts the possibility that native minority groups and established immigrant groups may work hard to erect barriers to keep new immigrants from incorporating into their society. Incorporation into the dominant minority group may not occur if the immigrant group is looked upon unfavorably by the population that controls the economic resources.
Copyright 2001 by Stephanie Bohon
18
Latinos in Ethnic Enclaves
The concentration of immigrant populations in one place may also increase discrimination in other ways. Since an immigrant's destination decision is often based on networks rather than open labor market information, the influx of a sizeable new population into an area with a slow growing or stagnating economy creates greater competition for jobs. Such a climate breeds resentment among the native population who see the immigrants as a threat to their livelihood (Tienda & Lii, 1987). Support for such an effect can be seen in research such as Hansen and Cardenas' (1988) study that shows that people in non-border areas hold generally more favorable attitudes toward Mexican immigrant workers than people in the more immigrant-saturated border regions. If immigrants do not incorporate into the non-Latino white population and they also do not incorporate into the minority sub-cultures to which they are most alike, what is the outcome? Queuing theory offers a possible alternative to both straight line and segmented assimilation theory. Queuing theory is based on the idea that workers stand in an imaginary line to wait for good jobs. Place in the line is determined by some standard of desirability. The most desirable workers are at the front of the line, and the least desirable workers are at the end. When similar workers compete for the same jobs, groups that are viewed as more desirable gain preferential access to these jobs. For immigrants, the measure of desirability is usually race or national origin, and men and women "stand" in different queues (Model & Ladipo, 1996). In short, queuing theory explains how employers' prejudices can affect the marketplace (Model & Ladipo, 1996). This theory deviates from neoclassical economics by assuming that the supply and demand for good jobs does not necessarily determine the availability of these jobs or the wages paid. Under queuing, the number of available good jobs is relatively invariable (at least in the short run), and whether or not a person gets a job depends less on the supply of labor and more on their place in the queue (Sakamoto & Chen, 1991; Sakamoto & Powers, 1995). Under queuing, willingness to work for lower wages or under poorer conditions does not radically alter the probability of getting a good job (Sakamoto & Powers, 1995). Less desirable workers obtain good jobs only when the supply of more desirable workers has been exhausted (Model & Ladipo, 1 996). Queuing is also different from discrimination, since discrimination in the market place is generally exclusionary, while queuing is preferential. In one setting at one time, queuing cannot be distinguished from discrimination. However, when two labor markets are compared, the distinction between queuing and discrimination becomes clearer. Since workers are queued presumably according to group membership first and skills second, any groups'
Copyright 2001 by Stephanie Bohon
Segnzented Assimilation and Queuing
19
place in the queue depends largely on the number of competing groups in the marketplace. For example, if Cubans are at the front of the queue, Nicaraguans are next, and Mexicans are last, occupational attainment for Mexicans would be higher in those places where there are no Nicaraguans. This assumes, of course, that the order of groups in the queue is the same across labor market contexts. Queuing theory has been discussed extensively in the economics literature and has also been explored in sociological studies (Reimers, 1984; Reitz, 1990; Rich, 1995; Roos & Hennessy, 1987; Roos & Jones, 1993; Roos & Reskin, 1992; Sakmoto & Chen, 1991; Sakamoto & Powers, 1995; Woon, 1987). Particularly queuing has been used to explain differences in occupational attainment, labor market sector attainment, and wages between men and women (Rich, 1995), immigrants and natives (Woon, 1987), and different immigrant groups (Roos & Hennessy, 1987). Unlike segmented assimilation, where individual level characteristics are still recognized as important factors in immigrant adaptation (Portes & Zhou, 1993; Portes, 1997), queuing assumes that group characteristics are more important predictors of queuing than individual characteristics, since easily recognized characteristics of the group provide signals that this group should be treated differently (Alba & Logan, 1992). The study of queuing is also distinguished from research on segmented assimilation in that the former tends to focus more on differences within pan-ethnic groups (e.g., national origin differences and within ethnic group differences), while the latter tends to focus on differences between these pan-ethnic groups (e.g., Asians versus whites or Latinos versus blacks).
Empirical Applications Since segmented assimilation developed partially out of concern for a possible decline in the economic outcomes of immigrant children, most studies of segmented assimilation have focused on the second generation. For example, the educational performance, attainment, and aspirations of Asian and Latino second-generation children was examined by Kao and Tienda (1995), and the self-identification of school children was explored by Portes & Zhou (1993). These studies found that Haitian immigrant children, for example, were particularly vulnerable to downward assimilation due to identification as African-Americans. Other researchers have also suggested that other outcomes, such as the transition to self-employment by the children of some groups, offer evidence of segmented assimilation (Alba & Nee, 1997). Although segmented assimilation was conceived as a way of explaining outcomes for second and third generation immigrants (Alba & Nee, 1997; Zhou, 1997), the concept has been applied to the first generation as well. Kao and Copyright 2001 by Stephanie Bohon
20
Latinos in Ethnic Enclaves
Tienda (1995), for example, find that differences in the educational performance among school-age immigrants changes with increased years in the U S . This indicates that the trends toward incorporation begin with the first generation, and that an examination of first-generation outcomes can be indicative of things to come. Portes (1997) lends support for this contention. He notes that while full assimilation into a native group occurs with successive generations, "patterns of adaptation of the first generation set the stage for what is to come." Massey (1995) also reinforces this argument by noting that, while measures of assimilation such as English fluency and intermarriage take generations, other measures of assimilation such as income and occupational attainment status increase with additional years in the U S . Since increased economic opportunities generally lead to some upward mobility, first generation immigrants may benefit from current labor market conditions, and they may, in turn, pass on these benefits to the next generation (Alba & Nee, 1997). Consequently, the selection of a destination area by the first generation of immigrants and the resultant economic mobility associated with that choice are important areas for study. In cities like Miami, segmented assimilation suggests that Latin American immigrants, as well as their offspring, will do particularly well, since assimilation means incorporation into CubanAmerican society. Furthermore, those groups most like the Cubans should assimilate at the fastest pace. According to Portes, assimilation into the CubanAmerican experience should be particularly easy for Nicaraguans, since they share not only a language and culture, but also a similar refugee history. Furthermore, Nicaraguans were particularly welcomed into Miami's Cuban community, thus affording them a privileged reception context (Portes & Zhou, 1993; Portes & Stepick, 1993).9 Since the amount of incorporation that occurs with the first generation is a good barometer of trends in the second and successive generations, it is important to examine the occupational attainment patterns of first generation immigrants. From this examination and the application of segmented assimilation theory to the design and methodology, important information can be gathered regarding the potential experiences of future generations. Applying segmented assimilation theory to first generation immigrants in the manner used in this study allows greater insight into the early patterns of adjustment and the long-term effects of this adjustment on these immigrants and their children. While segmented assimilation is based on the idea of the incorporation and creation of pan-ethnic groups with similar characteristics, queuing is based on the concept of dominant labor market status. Dominant labor market status is gained when certain groups in the best position to exploit their available
Copyright 2001 by Stephanie Bohon
Segmented Assimilation and Queuing
21
resources enter an area. They maintain their status through any powerful and coercive means available that allows them to push other groups to the periphery (Alba & Nee, 1997). Under segmented assimilation, Latin American immigrants should become like other Latinos in the labor market. In labor markets where Latin American immigrants and their descendants dominate that labor market (i.e., in the Latino enclaves), occupational attainment among all Latin American immigrants should be high and there should be no significant difference in occupational attainment levels between the different immigrant groups. If queuing is occurring in the Latino enclaves, occupational attainment will be high for some groups (e.g., the dominant enclave group and other groups high in the queue) and significantly lower for others. In summary, under the segmented assimilation model, country of origin should not determine occupational attainment levels for Latin American immigrants. All Latin American immigrants should demonstrate similar levels of occupational attainment regardless of national origin once individual factors such as education are taken into consideration. Under queuing theory, however, national origin may be a strong predictor of occupational attainment, particularly in those places where one or two Latin American immigrant groups controls the labor market.
THEORIES OF ADAPTATION AND THE ETHNIC ENCLAVE ECONOMY This study looks at segmented assimilation, queuing, and enclave theory within three distinct labor markets: metropolitan areas with Latino enclaves, metropolitan areas with Asian enclaves, and non-enclave areas. The impact of place of work on the adaptation experience of immigrants is of paramount importance because successful adaptation is as dependent on immigrants' economic context as it is on the human capital and ethnic factors that are brought to the marketplace (Portes & Zhou, 1992; Tienda & Wilson, 1992; Wallace, 1986). Alba and Nee (1997) note: If at the community level the opportunities to express ethnicity are meager or socially inappropriate, the intent to maintain ethnicity, assuming it exists, may be thwarted or transformed. The desire to find ethnic modes of behavior and expression, then, is likely to succeed where the supply-side of ethnicity is fairly rich in possibility. Where individuals assimilate in large numbers and are not replaced by a continuing immigration stream, a pattern characterizing many European-ancestry groups, the supply-side of ethnicity is diminished as a whole as well as narrowed in specific respects. (p. 835)
Copyright 2001 by Stephanie Bohon
22
Latinos in Ethnic Enclaves
So place is important in the adaptation of immigrants, yet the relationship between adaptation and place is not always clear. Immigrants in some places certainly fare better than others; however, an immigrant's economic position may create a barrier to migration to a place that provides more opportunity for occupational mobility or higher wages (Barrera, 1979; Portes & Bach, 1985; Tienda & Wilson, 1991; Wallace, 1996). And while there is amble evidence that the ethnic composition of the labor market has considerable impact on economic success (Lieberson, 1980; Semyonov, Hoyt & Schott, 1984; Reimers, 1984; Tienda & Lii, 1987), with the exception of the ethnic enclave, the mechanisms responsible have not been clearly identified (Tienda & Wilson, 1992). For example, an immigrant may select a destination because that place maximizes the immigrant's proximity to friends and relatives regardless of the economic opportunities in that place (Nee, Sanders, & Sernau, 1994). On the other hand, if having connections in the destination area increases the likelihood of early, stable employment, an immigrant may benefit despite the selection of a poorer labor market (Tienda & Wilson, 1992). The interaction between ethnicity and place is also important because ethnicity can "modify the character of class relationships" within a locale (Portes & Bach, 1985). Nowhere is this more evident than in the ethnic enclave. In the native labor market, members of certain ethnic groups may be viewed disparagingly, relegating immigrants from this group to low paying, menial, often dangerous jobs that native workers are reluctant to take. The push into the secondary labor market limits the future economic success of the immigrant ensuring their continuation at the bottom of the class hierarchy.1° When immigrants control the local labor market, however, as they do in ethnic enclaves, the opportunities for both occupational and class mobility increase. One of the very interesting points about the ethnic enclave is that it tends to contradict straight-line assimilation theory (Nee, Sanders, & Sernau, 1994). Under the straight-line model, adaptation is aided by geographic dispersion. Segmented assimilation does not operate under such assumptions, since context retains its primacy. In ethnic enclaves, immigrants can achieve occupational attainment soon after resettlement and may benefit from education and work experience acquired in their native country. Immigrants outside of the enclave, however, are often forced to compete with other members of the underclass in a pool of poor jobs. To conclude, it should be reiterated that labor market context should affect the ability of immigrants to acquire additional human capital upon arrival in the host country. Furthermore, both destination area and country of origin may have an impact on economic adaptation separate from the education, work experience, language ability, and household characteristics of the individual immigrants. The
Copyright 2001 by Stephanie Bohon
Segmented Assinzilation and Queuing
23
intersection of immigrant adaptation and ethnic enclave theories will be more fully explored in the next chapter.
Copyright 2001 by Stephanie Bohon
CHAPTER THREE
Ethnic Enclaves and Enclave Theory
In the pioneering work on ethnic enclaves, Wilson and Portes (1980) defined the immigrant enclave as a "self enclosed inner-city minority community." This rudimentary conceptualization of the term enclave remained in use only until the second significant work on the ethnic enclave was published two years later as Wilson and Portes began to diverge on the important defining elements of the term, most notably geographic location and business ownership. Wilson and Martin (1982) expanded the original definition to argue that the economies of these minority communities were "composed of clusters of small businesses which [were] collectively vertically and horizontally integrated" and usually minority owned. Portes, however, dropped the inner-city element and reconceptualized the ethnic enclave more rigorously as a geographically bounded area with a preponderance of employers from the same ethnic group or country of origin and their co-ethnic or co-national workers (Portes & Bach, 1985; Portes & Jensen, 1987, 1989, 1992; Portes & Stepick, 1993; Portes & Rumbaut, 1996). While the differences between Wilson's and Portes' definitions of the enclave are minor, they are instructive because they underscore the difficulties that researchers have had in coming to a consensus regarding the exact nature of the ethnic enclave. This continuing controversy, dubbed the "enclave debate" (Waldinger, 1992), has resulted in a variety of both conceptual and operative uses of the term. According to Portes and his colleagues (e.g., Portes and Bach, 1985; Portes and Jensen, 1987, 1989, 1992; Portes and Stepick, 1993; Portes and Rumbaut, 1996), any analysis of the enclave must emphasize the importance of ethnically defined jobs rather than ethnic residential segregation as conceptualized by Sanders and Nee (1987). Not convinced by either argument, Zhou and Logan (1989) examined Chinese workers in New York defining the enclave
Copyright 2001 by Stephanie Bohon
26
Latinos in Ethnic Enclaves
alternatively as place of work, place of employment, and industrial sector. Other authors used other formulations (Bailey & Waldinger, 1991; Evans, 1989; Light, 1984). Today, the enclave debate is expanded to include arguments over the degree of the economy's vertical and horizontal integration necessary to create "enclave conditions" (Bailey & Waldinger, 1991), the degree of labor market isolation necessary to facilitate self-employment (Evans, 1989), the amount of geographic "boundedness" experienced by ethnic workers (Light, 1984), and the amount of "ethnic advantage" needed to allow immigrants to escape the secondary labor market (Logan, Al ba, & McNulty, 1994). Despite the fact that there is no widely accepted definition of the ethnic enclave, there remain three factors either stated or implied in all of the conceptualizations. First, the term enclave implies separation from the economic activity surrounding it (Waldinger, 1992). This implication is taken to its extreme by Portes and Manning (1986) who note that the enclave is such a complete institution that immigrants do not need to venture outside of it to provide for their needs. Second, the emphasis on boundedness is central to definitions of the ethnic enclave espoused by other researchers (e.g., Light, l984), although exactly what is bounded remains contentious.ll Returning to the debates between Portes and Jensen and Sanders and Nee (1992), for example, it appears that while the two sets of authors do not agree on whether it is more important to emphasize ethnically defined jobs or ethnic residential segregation, they do agree that the enclave exists within a small area (usually a portion of a county or metropolitan area). Finally, despite the various definitions, the term enclave has been used in the United States to indicate only a handful of metropolitan areas. That is, the same sets of enclave cities are identified again and again by researchers, despite the myriad definitions, as shown in Table 1. This finding suggests that the differing definitions of the enclave are not so different, and that there is some validity to each position. For the purposes of the research undertaken here, Logan, Alba, & McNulty's (1994) definition of the enclave is employed. These authors contend that any city characterized by "co-ethnic owners and employees, spatial concentration, and sectoral specialization" can be considered an enclave (p. 694). This definition is useful because it encompasses the three elements of the enclave concept: economic separation, geographic bounding, and identification of cities which generally match those identified by other.
Copyright 2001 by Stephanie Bohon
Ethnic Enclaves and Enclave Theory
27
Table 1. Ethnic Enclaves in the United States Metropolitan Area Anaheim Honolulu Houston Jersey City Los Angeles
Ethnic group Japanese Japanese Mexican Cuban Korean
Miami
Chinese Japanese Mexican Cuban Cuban
New York
Chinese
Korean Dominican
San Francisco
Colombian Chinese
San Jose
Japanese Japanese
Author(s) identifying place Logan, Alba, & McNulty (1994) Logan, Alba, & McNulty (1994) Logan, Alba, & McNulty (1994) Logan, Alba, & McNulty ( 1994) Bonacich, Light, & Wong (1977); Bonacich & Sung (1982); Light (1980, 1984); Logan,Alba, & McNulty ( 1994) Logan, Alba, & McNulty (1994) Logan, Alba, & McNulty ( 1994) Logan, Alba, & McNulty (1994) Logan, Alba, & McNulty (1994) Bailey & Waldinger (1991); Evans (1989); Forment (1987); Gilbertson (1 995); Light (1984); Light, Sabagh, Bozorgmehr, & Der-Martirosian (1994); Logan, Alba, & McNulty (1994); Model (1992); Portes & Bach, 1985; Portes & Jensen (1989); Portes & Stepick, 1993; Sanders & Nee (1987); Wilson & Portes (I 980); Wilson & Martin (1982) Bailey & Waldinger (1991); Gilbertson (1995); Logan, Alba, & McNulty (1994); Waldinger & Lapp (1993); Zhou & Logan (1989) Bailey & Waldinger (1991); Logan, Alba, & McNulty ( 1994) Bailey & Waldinger (1991); Gilbertson (1995); Gilbertson & Gurak (1993); Waldinger, 1997a; Waldinger & Lapp (1993) Gilbertson (1995); Gilbertson & Gurak (1993) Bailey & Waldinger (1991); Logan, Alba, & McNulty (1994); Model (1992); Sanders & Nee ( 1 987) Logan, Alba, & McNulty (1994) Logan, Alba, & McNulty ( 1994)
An examination of Table 1 shows nine metropolitan areas in the United States that can be considered an ethnic enclave using Logan, Alba and McNulty's (1994) and others' definition. These cities are Anaheim, Honolulu, Houston, Jersey City, Los Angeles, Miami, New York, San Francisco, and San Jose. Of these metropolitan areas, three have more than one enclave. They are
Copyright 2001 by Stephanie Bohon
28
Latinos in Ethnic Enclaves
New York (which has Chinese, Korean, Dominican, and Colombian enclaves) San Francisco (which has Chinese and Japanese enclaves) and Los Angeles (which has Korean, Chinese, Japanese, Mexican, and Cuban enclaves). Five of the nine areas in Table 1 have Latin American immigrant enclaves; however, Light, Sabagh, Bozoregmehr, and Der-Martirosian (1994) and Portes and Bach (1985) make compelling arguments against the existence of any Mexican enclaves. Furthermore, among all of the cities that Logan, Alba and McNulty (1994) argue are ethnic enclaves, they make the weakest case for the Mexican enclaves in Houston and Los Angeles. Eliminating Mexican enclaves, there are only four cities with Latin American immigrant enclaves. Of these, three are Cuban.
THE IMPACT OF THE ENCLAVE ON IMMIGRANT ECONOMIC ADAPTATION For nearly twenty years, scholars have debated the effects of the ethnic enclave on the income, occupational mobility, and self-employment of workers (Aldrich & Waldinger, 1990; Borjas, 1986; Bonacich, 1988; Evans, 1989; Light, 1984; Logan, Alba, & McNulty, 1994; McManus, 1990; Min, 1988; Portes & Bach, 1985; Portes & Jensen, 1987; 1989; Portes & Stepick, 1993; Robinson, 1988; Sanders & Nee, 1987). Much of this debate has focused on the importance of ethnic enclaves in reducing the human capital depreciation initially experienced by new immigrants and the ability of the enclave to improve the overall economic position of immigrants in the labor market in the long run (Bailey & Waldinger, 1991; Evans, 1989; Light, 1984; Light, Sabagh, Bozorgmehr, & DerMartirosian, 1993; Logan, Alba & McNulty, 1994; Min, 1987; Portes & Jensen, 1989; Portes & Stepick, 1993; Wilson & Martin, 1982). While there is considerable disagreement among scholars about the precise mechanisms which facilitate better outcomes for immigrants in enclaves, the theoretical justification and practical evidence for assuming that enclaves encourage at least some types of adaptation are widely noted (Aldrich & Waldinger, 1990; Bailey & Waldinger, 1991; Evans, 1989; Light, 1984; Light, Sabagh, Bozorgmehr, & DerMartirosian, 1994, 1995; Logan, Alba & McNulty, 1994; Min, 1987; Nee, Sanders, & Sernau, 1994; Portes & Bach, 1985; Portes & Grosfoguel, 1994; Portes & Jensen, 1987; 1989; 1992; Portes & Rumbaut, 1996; Portes & Stepick, 1993; Portes & Zhou, 1992; 1995; Reitz, 1990; Sanders & Nee, 1987; 1992; 1996; Waldinger, 1989; 1993; Wilson & Martin, 1982). It is generally assumed that immigrants initially experience significant downward mobility upon entry into the U.S. due to lack of experience and lower employment skills (Waldinger, 1997a; 1997b). In addition, immigrants may also encounter early difficulties due to language barriers, lack of networks, and Copyright 2001 by Stephanie Bohon
Ethnic Enclaves and Enclave Theory
29
cultural adjustment lags. Ethnic enclaves limit the amount of human capital depreciation that occurs at the time of entry because they offer a pool of jobs that do not require high skill levels or English fluency (Portes & Bach, 1985; Rockett, 1983). In the long run, immigrants in an enclave can also expect to fare better than their non-enclave counterparts because of preferential access to capital which can be used for self-employment (Aldrich & Waldinger, 1990; Evans, 1989; Light, 1984; Portes & Bach, 1985; Portes & Stepick, 1993; Sanders & Nee, 1996; Waldinger, 1989). While the mechanisms through which immigrants achieve higher economic attainment in an enclave have not been fully explored, there are four assumptions that underlie the reasons for the relatively greater success of enclave immigrants: 1. Ability to speak English is unimportant in the ethnic enclave, so language does not create a barrier to entry into the labor market, and ability to speak the native enclave language may actually improve enclave chances. 2. Enclaves are marked by "bounded solidarity" and "enforceable trust" which is created through the reaction to discrimination faced outside of the enclave and the close association of the enclave members (Portes, McLeod, & Parker, 1978; Portes & Bach, 1985). The end result of the "we" feeling created by these two factors is that employers feel obligated to employ other immigrants, while immigrants feel compelled to work hard for their new bosses; 3. Cultural differences in job acquisition processes and work ethics between the sending and receiving countries are muted in an ethnic enclave and may be much more like the immigrant's country of origin than those of the destination area; and 4. Insofar as immigrant enclaves trade in ethnically-defined goods, skills acquired in the sending country are in high demand which speeds the transition process. These four factors are the common threads that bind enclave theory. These assumptions imply that the enclave insulates workers from less favorable conditions in the secondary labor market, and it also provides some protection from unemployment or downward mobility immediately after arrival in the receiving country. These assumptions also imply that there are marked differences in what the enclave provides for the worker in the years immediately proceeding immigration (employment) and in the long run (self-employment and upward mobility).
Copyright 2001 by Stephanie Bohon
30
Latinos in Ethnic Enclaves
EXAMINING THE ENCLAVE: MIAMI, LOS ANGELES AND NEW YORK Given both the geographic and ethnic diversity of the nine cities in Table 1, it is prudent to consider whether or not the factors creating and perpetuating the enclaves are common to all of these metropolitan areas. Furthermore, it seems necessary to consider whether the outcomes for immigrants are also similar. In order to address these questions, the case study literature on the Miami, Los Angeles and New York enclaves is examined.
Miami The rise of Fidel Castro and the ensuing revolution in 1959 prompted a flood of emigration from Cuba to the United States. Given the proximity of Miami to this Caribbean island and the frequent choice of Miami as a vacation spot for affluent Cubanos, it is not surprising that Miami was the eventual destination of these migrants, since most of them assumed that they would soon be returning to their homeland (Portes & Bach, 1985). Like most refugees, the Cubans did not enter the U S . well prepared to adapt to U.S. society. Many lost much of their wealth during the revolution or did not have time to bring their assets with them, and despite the high levels of education and affluence of the first wave of immigrants, many did not speak English (Portes & Stepick, 1993). Consequently, the initial wave of Cuban refugees faced severe initial downward mobility and were forced to settle in Miami and Hialeah's inner city slums (Stepick & Portes, 1986). Although these early Cuban immigrants were skilled, their inability to speak English and the discrimination that they faced in the Miami labor market put them at a disadvantage vis-a-vis non-Latino white residents (Portes & Stepick, 1993). Despite these disadvantages, two factors worked in their favor. The first factor was the relatively advantaged reception context that these immigrants faced (Portes & Borotz, 1989). Given the hostile relations between the Castro regime and the U.S. government, the Cuban refugees were welcomed and provided full refugee benefits including medical care, access to welfare, and free English classes. Cuban refugees did not have to worry that they would be deported or that their legal immigrant status would come under scrutiny (Portes & Stepick, 1993). Given that fear of deportment has been shown to create a major obstacle to occupational mobility for many Latin American immigrants (Kossoudji & Cobb-Clark, 1996), the Cubans' favorable reception was an essential key to their later advancement. The other advantage that the Cubans had was that Miami was home to many Latin American-based financial institutions. Cubans who could not find work in
Copyright 2001 by Stephanie Bohon
Ethnic Enclaves and Enclave Theory
31
the mainstream economy often found employment among the various Central and South American banks located in the area. These jobs still resulted in downward mobility, as former Cuban bankers became loan officers for Latin American banks, for example; however, their positions were put to good use. Cuban loan officers convinced their employers to make rather unorthodox loans to other Cuban refugees. While the U S . banks often refused loans to immigrants because of their poor economic status and lack of credit history, the Latin American banks made small start-up loans to these refugees using family reputation as a main criteria for lending risk (Portes & Bach, 1985). As these small character loans were repaid, Cuban business owners were able to acquire larger loans for bigger business ventures. By the time the banks discontinued granting character loans in the 1970s, many Cuban-owned businesses had already been established (Portes & Stepick, 1993). Portes and Stepick (1993) contend that there are three factors that are essential for the establishment and perpetuation of an ethnic enclave. First, an area must have stable economic conditions coupled with the demand for ethnically defined goods. Second, employers must have access to a cheap pool of labor which is unmotivated or unable to work in another economic sector. Third, entrepreneurs and potential entrepreneurs must have access to capital. Miami provided (and continues to provide) all three of these prerequisites for the Cuban community. The presence of Latin American banks willing to make character loans in the early days, and the proliferation of Cuban-American owned and operated banks in Miami today ensures that capital will continue to be available for Cuban entrepreneurs and their children. In addition, the continuation and expansion of many Cuban-owned firms is aided by the fact that over 35 percent of the population of Dade County is Spanish-origin. Consequently, local industries producing goods such as tortillas, cigars and rum should continue to experience high demand for their products. Privileged access to labor also remains important in the Miami enclave. The second wave of refugees that arrived in Miami in the 1970s was disproportionately drawn from middle class, white collar workers. This new infusion of migrants took jobs in Cuban-owned shops and factories, bought Cuban goods from their countrymen, and acquired homes in Miami and Hialeah vacated by the entrepreneurial class as they moved to suburban Kendall. As these immigrants went on to start their own small businesses, the labor pool was replenished by a third wave of low-skilled, Cuban peasants. Although migration from Cuba has slowed considerably in the last ten years, available cheap labor is now provided by immigrants from other Central American countries, notably Haiti and Nicaragua.
Copyright 2001 by Stephanie Bohon
32
Latinos in Ethnic Enclaves
The Miami enclave has been able to sustain itself and grow due to its early advantages and the upward mobility that it offers coethnic workers who go on to establish businesses of their own. Portes estimates that one of every four Cubans in Miami is self-employed or employed by a co-ethnic, compared to one of every 8 4 employed blacks, and that the biggest predictor of self-employment for Cubans is previous employment by another Cuban (Portes & Bach, 198.5). This transition to self-employment is important, since it is the best means for immigrants to attain high occupational attainment (Borjas, 1986).
Los Angeles Despite the unique history of the Miami enclave, Portes and his colleagues have made a number of theoretical assertions about the creation of ethnic enclaves in general and the role they play in the lives of immigrants based entirely on this one case (Portes & Bach, 1985; Portes & Grosfoguel, 1994; Portes & Jensen, 1987, 1989; Portes & Manning, 1986; Portes & Rumbaut, 1996; Portes & Stepick, 1993; Portes & Zhou, 1995). The myriad works on the Miami enclave by other researchers (Bailey & Waldinger, 1991; Evans 1989; Forment, 1987, 1989; Gilbertson, 1995; Light, 1984; Light, Sabagh, Bozorgmehr, & DerMartirosian, 1994; Logan, Alba, & McNulty, 1994; Model 1992; Sanders & Nee, 1987; Wilson & Martin, 1982) also seem to indicate that the immigrant economy in this Florida city is typical of all ethnic enclaves. Some researchers, however, question the generalizability of the Cuban experience in Miami. Logan, Alba, and McNulty (1994) note: [Tlhe diversification found in Miami's Cuban community is not the norm for enclave economies. Rather, if we were to describe the "ideal type" of enclave we would point to the more common instances. The usual form of an enclave, we would propose, is based on low-wage production in a single productive sector with low levels of capitalization (and possibly high levels of female employment), in conjunction with ethnically stereotyped restaurants or food stores and some services oriented to group members. (p. 719)
In order to get a truer picture of the ethnic enclave, therefore, it is instructive to examine cities more closer to the "ideal type" that Logan, Alba, and McNulty (1994) describe. Los Angeles is one such city (Fernandez-Kelly & Garcia, 1980). Like Miami, Los Angeles is home one of the country's largest immigrant populations. In fact, Los Angeles is second only to Miami in proportion of the population that is foreign born (27 and 34 percent, respectively). Also like Miami, a majority of L.A.'s immigrant population (nearly 44 percent) comes Copyright 2001 by Stephanie Bohon
Ethnic Enclaves and Enclave Theory
33
from a single sending country. Unlike Miami, however, where the single largest immigrant ethnic group is comprised of the highly educated and highly skilled Cubans, Los Angeles is home to the largest (largely unskilled and uneducated) Mexican immigrant population in the U.S. (Waldinger, 1997a, 1997b; Waldinger & Bozoregmehr, 1996).12 In fact, with its nearly 4 million immigrants, Los Angeles boasts the United States' largest total immigrant population, which may also explain the extreme ethnic fragmentation of the Los Angeles economy (Waldinger, 1997b; Waldinger & Bozorgmehr, 1996). In L.A., most ethnic groups are highly concentrated as either owners or workers in a small number of industries (Waldinger, 1997b). This large immigrant population has ensured a continuing demand for ethnically defined goods, a necessary prerequisite for enclave creation according to Portes & Stepick (1993). On the other hand, unlike Miami where most of the Cuban enclave is located along Calle Ocho in Miami and in the neighboring city of Hialeah, industrial concentration does not equate with geographic clustering in Los Angeles. The three Asian enclaves in L.A. (Chinese, Japanese, and Korean) are widely dispersed throughout Los Angeles County (Light & Bonacich, 1991; Logan, Alba & McNulty, 1994). Furthermore, while the largest immigrant group in Miami is also the dominant enclave group, this is not the case with Los Angeles (Logan, Alba, & McNulty, 1994). The three largest immigrant groups in Los Angeles (Mexicans, Salvadorans, and Guatemalans), neither own many of their own businesses nor are employed in any numbers by co-ethnic employers (Waldinger, 1997b; Waldinger & Bozorgmehr, 1996). Nonetheless, these groups have had great success in finding jobs, and employment rates (often in low-paying, low skilled work) are high (Waldinger, 1997a). In fact, this large and relatively new immigrant population has sustained the second pre-condition for enclave development: a cheap pool of available labor (Portes & Stepick, 1993). While Mexicans, Salvadorans, and Guatemalans comprise more than half of Los Angeles' immigrant population, the Asian population in the metropolitan area is also sizeable. In fact, as of 1990, there were more Asians living in L.A. than African Americans. And unlike the larger Latin American immigrant groups in the city, the largest groups of Asian immigrants (Chinese, Filipino, Korean, Vietnamese, and Indian - in order of size), with the exception of the Vietnamese, were disproportionately selected from among the most skilled, educated, entrepreneurial, and wealthy (Waldinger, 1997b). The Chinese, Korean and Japanese immigrants--most of whom arrived after 1970--quickly established enclaves and were soon able to control large sectors of the produce
Copyright 2001 by Stephanie Bohon
34
Latinos in Ethnic Enclaves
farming, apparel manufacturing, printing, grocery, and restaurant markets (Jiobu, 1988; Logan, Alba & McNulty, 1994).13 Light (1972) has argued that the building of the Asian enclaves was the result of a "cultural proclivity toward business partnerships." However, more recent enclave theory suggests a broader explanation. The ease with which the Chinese, Japanese, and Koreans were able to build enclaves in Los Angeles, and the inability or unwillingness of the Mexican immigrants to do the same may be explained by the differences in job experience, education levels, enterpreneurial networks and capital access between the two groups (Portes & Rumbaut, 1996). In addition to the Asian immigrants' high levels of human capital, these entrepreneurs benefited from a system of rotating credit unions which allowed for the easy acquisition of small loans (Light, 1972; Logan, Alba, & McNulty, 1994). The Mexicans, however, had few financial resources to draw upon and little experience with urban enterprise. Again, this finding lends support to Portes and Stepick's (1993) contention that access to capital is necessary for enclave creation. It can also be argued that the relative ease with which Mexicans have been able to find employment in the Los Angeles area (Waldinger & Bozoregmehr, 1996) has removed some of the impetus for self-employment. Given the low education and skill levels of these immigrants, a job in the U.S. economy may be seen as upward mobility for Mexican workers, while the same job would be viewed poorly by the typical Cuban, Chinese or Korean immigrant. New York The Chinese and Korean enclaves in New York, the second largest immigrant receiving city in the U.S., are not markedly different from those in Los Angeles in either historical development or sectoral specialization. In both cities, the Chinese are highly concentrated in the garment industry, and, like Los Angeles, Manhattan and Queens are home to Chinatowns. And while the Korean population in New York is three times larger than that in Los Angeles, the Korean enclaves in both cities are geographically dispersed (Logan, Alba, & McNulty, 1994). Unlike Los Angeles, but similar to Miami, New York's largest immigrant group, Dominicans, also control an enclave economy (Portes & Grosfoguel, 1994; Waldinger, 1997b). Dominican immigrants (almost 90 percent of whom live in the New York area) own nearly 70 percent of the bodegas in New York, and control significant portions of the transportation and garment trade in that city. Much of this entrepreneurial success can be attributed to an influx of professional and white collar workers from the island in the late 1980s (Portes & Grosfoguel, 1994; Portes & Zhou, 1992). But unlike the Korean and Chinese
Copyright 2001 by Stephanie Bohon
Ethnic Enclaves and Enclave Theory
35
small business owners, who acquired start-up capital through rotating credit unions (Light, 1972; Logan, Alba, & McNulty, 1994), it is suspected that many of the Dominicans brought money with them or acquired it through money laundering and drug trafficking (Portes & Grosfoguel, 1994). More recently, Dominican businesses in the U S , have been funded through joint ventures between Dominican immigrants and firms in Santo Domingo and Santiago. In fact, more than 40 percent of the Dominican enclaves businesses in 1990 received some capital from their homeland, although the amounts were typically small (Portes & Zhou, 1992).
Similarities and Differences in the Latino Enclaves Returning to the question of whether or not there are commonalities in the Latino enclaves, the examination of Miami, Los Angeles, and New York reveals a number of trends. First, the initial means of acquiring venture capital differed markedly between dominant enclave groups in New York, Miami, and Los Angeles prior to the creation of the enclaves. However, each group had access to capital from sources other than traditional U.S. lending institutions. Furthermore, each group was able to acquire funds from sources that other immmigrants groups in the same city did not have access to. Finally, each group has created a system whereby capital will be readily accessible (at least to members of the dominant enclave group) in the future. A second factor common to the three cities is the relatively stable economic conditions within the enclave sector. Unlike Los Angeles, where low-skilled manufacturing jobs remained abundant throughout the 1980s despite the postindustrialization that plagued other cities (Waldinger, 1997b), New York has seen marked declines in the number of available manufacturing jobs. Interestingly, while the garment industry was particularly hard hit, experiencing declines from the 1950s throughout the 1980s, employment opportunities in immigrant apparel firms actually grew during this same period (Waldinger, 1997b). Many of the hundreds of garment shops found in New York today are owned by Chinese, Korean, and Dominican immigrants who employ co-ethnic workers (Bailey & Waldinger, 1991; Portes & Grosfoguel, 1994; Portes & Zhou, 1992), and the number of available jobs in these sectors continues to rise (Waldinger, 1997b). One reason for the success of the immigrant apparel industry has been the ability to recruit semi-skilled workers from among other immigrants, a process which Bailey and Waldinger (1991) refer to as the training systems approach. Since the needle trades are often marked by demand for experienced workers, but the pay is typically low, entrepreneurs can rely on their workers to refer their immigrant relatives and friends for open positions within the firm. Another Copyright 2001 by Stephanie Bohon
36
Latinos in Ethnic Enclaves
reason for the success of the immigrant garment industry, particularly in New York's Chinatown, has been an influx of Asian immigrants in recent years who still tend to settle in the traditionally Chinese and Korean areas (Portes & Zhou, 1992). Clearly the enclaves in New York, like Miami and Los Angeles, benefit from a plentiful supply of cheap immigrant labor. In summary, there are three factors common to the three enclave cases examined here. First, the dominant enclave groups have preferential access to capital for self-employment. Second, the enclaves appear to be insulated from conditions that may depress the larger economy. Finally, there are high levels of employment in each of the enclave areas, suggesting that the enclaves provides ample jobs for immigrant workers. These three cities' enclaves are not entirely similar, however. Unlike Miami, where second generation Cubans tend to stay and work in businesses owned by their immigrant parents, enclave workers in New York tend to be disproportionately first-generation immigrants. More precisely, given that much of enclave employment in New York is in sweatshops, children of immigrants generally prefer less grueling work in the service sector (Bailey & Waldinger, 1991). In New York, therefore, employment by the first generation in the immigrant garment industry is a pathway to the primary labor market for the second generation. A second notable difference between the three enclaves is in the numerical prevalence of the dominant enclave group. In New York and Miami, the largest immigrant groups were able to establish enclave economies, while the largest Asian groups also established enclaves in Los Angeles, but the largest Latin American groups did not. A final difference between the three areas is the diversity among the types of industries that comprise the ethnic enclave economy. In Los Angeles and New York, the dominant enclave groups have particularly concentrated their efforts in the textile and apparel industries (Logan, Alba, & McNulty, 1994). New York's enclave is slightly more diversified, since the concentration of Dominican immigrants in the poor, predominantly Puerto Rican section of Washington Heights has provided opportunities for self-employment not available elsewhere. As mainstream supermarket, drug store, and nightclub chains retreated from this impoverished sector, and cab companies increasingly excluded this area from their service, cash-rich Dominicans were not only able to purchase the left-over infrastructure, they were able to reorient these businesses to suit the tastes of the largely Latino clientele (Portes & Grosfoguel, 1994; Portes & Zhou, 1992). Despite these additional self-employment opportunities, however, New York's Dominican enclave remains much more
Copyright 2001 by Stephanie Bohon
Ethnic Enclaves and Enclave Theory
37
concentrated in a few industries than Miami, where the businesses in the enclave economy encompass nearly every business sector (Portes & Bach, 1985). Overall, what is particularly striking about the enclaves in these three cities is the similarity in the conditions which led to their creation and success. In each case, heavy concentrations of immigrants created a stable demand for ethnic goods. This immigrant concentration also ensured a ready supply of cheap labor. Finally, each dominant enclave group benefitted from privileged access to capital for self-employment. Clearly, the most successful immigrant groups in these three cities were able to capitalize on these three factors to create enclaves that have enjoyed economic stability despite changing conditions in the larger economy.
THE IMPACT OF THE ENCLAVE ON NON-COETHNIC WORKERS The preceding sections indicate that members of the dominant enclave group tend to fare well in enclaves, either through self-employment or through preferential access to jobs. While the evidence that immigrants fare well in enclave environments is abundant (Aldrich & Waldinger, 1990; Bar-jas, 1986; Evans, 1989; Min, 1987; Portes & Bach, 1985; Portes & Jensen, 1989; Portes & Zhou, 1992), it is also limited in ethnic group coverage. As indicated in the preceding sections, much of the work on this topic has been conducted by Portes and his colleagues examining the Cuban enclave in Miami (e.g., Portes & Bach, 1985; Portes & Jensen, 1989; Portes & Stepick, 1993; Portes & Rumbaut, 1996; Portes & Zhou, 1996; Wilson & Portes, 1980). The other works on enclave effects include studies of the economic outcomes of Chinese, Korean, and Filipino immigrants in Los Angeles (Sanders & Nee, 1987; 1992), and Latino and Asian groups in New York (Kim, Huhr, & Fernandez, 1989; Waldinger, 1989). These studies focus on the earnings, self-employment potential, and occupational mobility of ethnic employers and their co-ethnic workers, while virtually ignoring other immigrants who live and work in these economies. The previous section outlines the four assumptions common to enclave theories based on studies focusing on co-ethnic economic interaction in metropolitan areas. However, if these assumptions hold true, there is also reason to suspect that all immigrants who share common languages and similar cultures should fare well in ethnic enclaves, even if they do not share a country of origin. Immigrants from various sending countries should enjoy the enclave advantages insofar as they provide a continued source of labor with the needed language and trade skills of their homelands. This should be particularly true of Latin American immigrants since they are relatively homogeneous with respect to language and religion. In short, while Cubans appear to do better in Miami than
Copyright 2001 by Stephanie Bohon
38
Latinos in Ethnic Enclaves
in other areas of the U.S., it could also be the case that Salvadorans, for example, also fare better in the Cuban enclave than elsewhere. Owners of enclave businesses may also appreciate the plight of the recently arrived immigrants, even if they are from a different country of origin. Some may be refugees who can identify with the turmoil caused by sudden uprooting or forced immigration. Others may merely remember their own difficulties in reestablishing themselves in a new country. This empathy may motivate these employers to hire new arrivals and provide them with the economic assistance essential to rapid economic mobility. Portes and Stepick (1993), for example, predict that in the long run Nicaraguan immigrants in Miami will enjoy many of the same advantages that Cubans and Cuban-Americans receive in that area because of the two groups' similar class and occupational backgrounds and historical refugee experiences. On the other hand, it may be the case that immigrants with needed language and trade skills are an easily exploited source of cheap labor for ethnic employers who feel no social obligations to these immigrants (Bonacich, 1988). Consequently, while immigrants will find jobs in the enclave in the short run, their ability to obtain better jobs or capital to start their own businesses in the long run may be hindered. Furthermore, the jobs that they may find in the enclave in the short run may be no better than those available in the secondary labor market. This research explores these possible divergent outcomes by contrasting segmented assimilation theory (that suggests that non-coethnic immigrants will fare well in the enclave) and queuing theory (that suggests that non-coethnic immigrants may face as much or more discrimination in the enclave economy). By also examining the outcome of immigrants who share a language, culturallydefined job skills, and discrimination in the non-Latino white or Asian economic sector, the underlying assumptions of enclave theory can also be tested. If it is the case that Latin American immigrants do not fare well in those metropolitan areas with Latino enclaves in which they are not the dominant group, researchers must re-examine those elements of enclave theory that currently explain why the enclave benefits only dominant enclave group members.
Copyright 2001 by Stephanie Bohon
CHAEER FOUR
Data and Methods
DATA AND SAMPLE SELECTION The data used for this study are taken from the U.S. Census 1990 Public Use Microdata (A) Sample (PUMS-A). The PUMS-A is a sample of census records for 5 percent of the housing units in the United States and the people in them. The PUMS contains records of both people and housing units, although only person records will be used in this analysis. This data set was selected because it contains relatively detailed immigration and labor force data and is sufficiently large to study small sub-populations in each of the enclave and non-enclave areas of interest.
Sample Selection Criteria Non-black Latino men between the ages of 23 through 6 4 in 1990 who immigrated prior to 1985 from Latin America and who are currently in the labor force and not currently enrolled in school or the military were included in the models of long-term immigrant occupational attainment. The age 23 was selected as the starting point because it is assumes that those men who immigrated before 1985 will have spent at least five years in the U S . labor force, or will have had the opportunity to have done so. As a consequence, the long-run economic effects of the enclave can be assessed for these immigrants. The recent immigrant sample includes non-black, Latino males aged 18-64 in 1990 who immigrated from a Latin American country between 1985 and 1990 and are currently in the labor force.14 Women are excluded from this study for theoretical reasons. Foremost among these reasons is that the advantages that the enclave offers to women may not be reflected in occupational attainment (Bonacich, 1988; Perez, 1986).
Copyright 2001 by Stephanie Bohon
40
Latinos in Ethnic Enclaves
While most enclave studies focus on men (and, as a consequence, enclave theory is a reflection of men's experiences), the few studies that have looked at women indicate that the economic advantages that women receive from the enclave are generally through self-employment in a joint-owned venture with a spouse (Bonacich, 1988; Perez, 1986). Another advantage of the enclave, at least for Latina women, is the ability to bring children with them to work. This allows them to work outside of the home while still maintaining the cultural imperative that women should provide primary care for their children (Perez, 1986). Such a situation, however, does not necessarily allow for high occupational attainment. While a more in-depth examination of women's mobility in the enclave is an important direction for future research, at present, their role lies outside of the theoretical dimensions of this study. Non-black Latino men were selected for this study to avoid confusing ethnicity issues with problems of racial discrimination. While there are some problems with the self-reporting of race among Latinos, it is assumed that few white Latinos would classify themselves as black. In addition to those immigrants reporting themselves as black, Jamaicans, Dominicans, Haitians and "Other Hispanic" immigrants indicating Caribbean regional descent (except Cubans) were also eliminated from the sample.15 Puerto Ricans were also excluded from the analysis, since they are not subject to the same immigration barriers as the other groups in this study.
Sample Selection Issues Two major difficulties arise in using Census data to study Latino immigrants. One difficulty arises from the large number of undocumented Latinos working in the United States. The other difficulty arises from the racial and ethnic selfassessment. Both of these difficulties relate to the enumeration of Latinos in the decennial counts. In the 1990 Census, about 1.4 percent of the total population went uncounted; however, blacks and Latinos are much more likely to be undercounted than other racial and ethnic groups (Schirm, 1991). One of the reasons that Latino undercount is problematic stems from the size of the undocumented population; workers in the country illegally are not likely to fill out and return Census questionnaires to the government. Estimating the size of the undocumented population, separating out the Latino portion of that population, and estimating the size of the documented but isolated component of that population (Latinos in the country legally but unable to speak English) presents one of the largest challenges that demographers face when attempting to estimate the number of Latinos in the U.S. (Passell, 1993). Another potential difficulty with this study stems from the design of the Census itself and answers to questions regarding Hispanic origin and Hispanic Copyright 2001 by Stephanie Bohon
Data and Methods
41
ancestry. Beginning in 1970, respondents to the Census were asked to identify themselves and members of their household by race (black, white, AsianIPacific Islander, American IndianIAleutian, or other race) and ancestry. Additionally, they were asked whether or not they or the members of their household were of Hispanic origin. In 1980 Census, the question of Hispanic origin yielded the highest non-response rate of all the questions on the long and short forms. Additionally, there has been a sizeable mismatch between race, ancestry, and Hispanic origin responses. Numerous blacks living in the southern United States identified themselves as South Americans, and other respondents reported Hispanic origin but no Hispanic ancestry (Martin, Demaio, & Campanelli, 1990). In short, the Census fails to provide data on many Latinos because they are remain uncounted or they failed to respond to the Hispanic origin question. Additionally, some non-Latinos may be included in the Hispanic counts. These difficulties suggest that the Latino immigrants in this sample are selected from primarily documented workers who have adapted the most readily to the U S . political and social system.
SAMPLING AREAS Enclave Selection The methodology employed in this study relies on a three stage sampling area encompassing twelve metropolitan places. This sampling scheme assumes that the urban destination of Latin Americans immigrants in the U S . can be divided into three types of receiving sites: metropolitan areas with Latino enclaves, metropolitan areas with Asian enclaves, or non-enclave areas.16 The Latino enclave sites used in this study are the Miami-Hialeah Primary Metropolitan Statistical Area (PMSA), the New York City PMSA, the Los Angeles-Long Beach PMSA, and the Jersey City PMSA. The Asian enclave sites are the San Jose PMSA, the Honolulu Metropolitan Statistical Area (MSA), the San Francisco PMSA, and the Anaheim-Santa Ana PMSA. These eight metropolitan areas encompass all of the enclaves listed in Table 1, with the exception of Houston, whose enclave status remains contentious. PMSAs are subsets of metropolitan areas with one million or more people that consist of a large urban county or counties that demonstrate strong social and economic ties. MSAs are metropolitan areas (MA) that are not closely connected to other metropolitan areas and are generally surrounded by nonmetropolitan counties (Myers, 1992). The choice of PMSAs and MSAs is useful for this study because it offers a larger sampling area than central cities. At the same time, since PMSAs and MSAs are generally a smaller geographic area than MAS, they are more appropriate for enclave designation, since the enclave
Copyright 2001 by Stephanie Bohon
42
Latinos in Ethnic Enclaves
concept implies tight boundedness. Finally, many MAS, such as Los Angeles, contain more than one enclave area (e.g., Los Angeles and Anaheim), so a smaller geographic unit is preferred.
Non-enclave Selection Because one of the purposes of this study is to compare enclave and non-enclave areas, a four-point criteria was devised in order to select the non-enclave MSAs and PMSAs. For this study, a PMSA is considered a non-enclave area if there are no compelling arguments in the enclave literature for designating the PMSA as an ethnic enclave (such as is the case for Houston), and there are no other enclaves in the same MA. Furthermore, the PMSA cannot fit any of the current, competing definitions of an ethnic enclave, as discussed in Chapter Three. Finally, the area must have a sizeable non-black Latino population comprised of immigrants from at least three different places of origin. There are eight cities that fit the first three criteria: Philadelphia, Detroit, Washington, D.C., San Diego, San Antonio, Chicago, Newark, and El Paso. Of these, four were selected through a process of elimination. Philadelphia and Detroit were eliminated because their Latin American immigrant populations were too small. El Paso was also eliminated because, although it has 1134 respondents in the PUMS sample that fit the sample selection criteria for this study (as discussed above), 98 percent of the immigrants in the El Paso MSA are from Mexico. Finally, Washington D.C. was eliminated because the Latin American immigrants in the PUMS sample primarily identify themselves as black, leaving too few non-black respondents to analyze. The remaining non-enclave areas selected for this study are the Chicago PMSA, the Newark PMSA, the San Antonio MSA, and the San Diego MSA. Each of these areas has a non-black Latino immigrant population of at least 12,000 men over the age of 23, providing a sample of at least 900 total respondents per PMSA.
Place of Work and Place of Residence Considerations The selection of the twelve metropolitan areas examined in this study is fairly straightforward; however, the proper means of allocating respondents to these areas remains under debate. According to Portes and his colleagues (e.g., Portes and Bach, 1985; Portes and Jensen, 1987,1989, 1992; Portes and Stepick, 1993; Portes and Rumbaut, 1996), an ethnic enclave is a geographically bounded area with a preponderance of ethnic employers and their co-ethnic workers. This definition emphasizes the importance of ethnically defined jobs rather than ethnic residential segregation as conceptualized by Sanders and Nee (1987).
Copyright 2001 by Stephanie Bohon
Data and Methods
43
The question then becomes, should respondents be included in the sample if they live in an enclave area or if they work there? For those respondents in the labor force but not working, place of residence and place of work are the same (i.e., unemployed respondents are selected if they fit the sample criteria and they live in the PMSA or MSA under study). For those respondents who live in one PMSA and work in another, the appropriate selection criteria is less straightforward; however, Portes and Jensen (1992) make a compelling argument for place of work as the appropriate context when they contend that the factors that determine economic well-being occur in the marketplace, not at home. Furthermore, as workers (particularly entrepreneurs) become more affluent, they are likely to move away from the enclave to upscale, suburban areas while maintaining the enclave as their place of work. Selection of respondents as enclave place of residence, therefore, biases outcomes toward the least well off workers. While the theoretical argument over sample selection is interesting, it may not pose much of a practical problem. For example, Sanders and Nee (1987) note that, for the Miami-Hialeah PMSA, results comparing income attainment do not vary significantly when place of residence is used as a sample selection criteria rather than place of work. However, their contention is not without criticism (Portes & Jensen, 1992). T o test this, total sample size, sample distribution, and average Male Socioeconomic Index (MSEI) scores were compared by place of work and place of residence. These results are reported in Appendix A. The results show that using place of work, rather than place of residence does not significantly effect average MSEI even when there is a substantial difference in sample size, as is the case for San Francisco. Consequently, place of work is used in this research as the sample selection criteria, because it is theoretically more compelling, but it is assumed that either method would yield similar results.
Problems in the Allocation of the Place of Work Variable in the 1990 PUMS While there are compelling arguments for using the place of work variable in the 1990 PUMS to select respondents, there are also some difficulties with this procedure inherent in the data set. As previously mentioned, the PUMS utilizes unique subareas designated as Public Use Microdata Areas (PUMAS). However, according to personal communication with Celia Bortlein at the Census Bureau (February 13, 1998), errors were made in the allocation of location data from the Census Bureau's master sample detail file to the microdata areas in the PUMS for some metropolitan places. The consequence of this error is that respondents' place of residence, place of residence five years earlier, and place of work (the variable used in this study), cannot be examined at the PUMA level for the Copyright 2001 by Stephanie Bohon
44
Latinos in Ethnic Enclaves
Miami-Fort Lauderdale, Chicago-Gary-Lake County, and Los AngelesAnaheim-Riverside consolidated metropolitan statistical areas. This problem with the 1990 PUMS has also been documented in the literature (Portes & Jensen, 1992). However, since PMSAs (the geographic subareas used in this analysis) are collections of PUMAS, the problem is less significant, because the improper allocations usually did not occur across county lines. That is, although some respondents were not allocated to the proper PUMA (e.g., the PUMS data show no Cubans working in Miami City), they are generally allocated to a neighboring PUMA within the same county. Consequently, the decision to use the PMSA as the geographic unit of analysis rather than some smaller area resulted in capturing most of the respondents. Verification of this fact comes from the Census Bureau (personal communication, February 16, 1998). Although the Census Bureau could not confirm the numbers for the sample of men used in this study, they were able to confirm that the number of workers reported in the PUMS data for the Los Angeles-Long Beach, Anaheim-Santa Ana, Miami-Hialeah, and Chicago PMSAs were within one-percent of the numbers reported in the Bureau's master sample detail file.
Sample Distributions Sample distributions by place of origin for the 12 PMSAs and MSAs under study are shown in Tables 2-7. Table 2 shows the sample distributions for the long term immigrants working in the Latino enclave PMSAs. These samples range in size from the Los Angeles-Long Beach PMSA which has 16958 respondents in the sample to the Jersey City PMSA, which has slightly less than 600 respondents. There is also considerable variation in the internal composition of these enclaves. The Miami-Hialeah PMSA, for example, is predominantly Cuban, with 78 percent of that sample selecting Cuba as their country of origin. The Jersey City PMSA is also predominantly Cuban, although less than half of the total sample in that area is Cuban. On the other hand, the Los Angeles-Long Beach PMSA is predominantly Mexican (75.2 percent), while the New York PMSA is mainly South American (37.3 percent). Table 3 shows the sample distribution by place of origin for the longer term immigrants working in the Asian enclave PMSAs. Again there is considerable variation in the sample sizes across the PMSAs. In each of the Asian enclave areas, however, Mexicans are the dominant Latin American immigrant group, accounting for at least 46 percent of the total sample. Table 4 shows the sample distribution by national origin for the long-term immigrants working in the non-enclave PMSAs. Again, Mexicans dominate the sample, except in the Newark PMSA where Cubans are the largest group, Copyright 2001 by Stephanie Bohon
45
Data and Methods
accounting for more than a third of the total sample, followed by South Americans (30.9 percent) Tables 5 through 7 show the sample distributions by place of work in the PMSAs for the recent immigrants. In every case, the total sample size is smaller than in the longer term samples, and there are differences in the internal distributions, as well. In Jersey City, where Cubans account for about half of all longer term Latin American immigrants in this study, they count for only about seven percent of the recent immigrants. The percent of Cubans is also smaller in the recent immigrant sample in the Miami-Hialeah PMSA and in the New York PMSA. In Los Angeles, the percent of recent Mexican immigrants is still high, but it is slightly smaller than in the longer term sample, primarily due to an increase in Central American migrants. In the Asian enclave and non-enclave PMSAs, the percent Mexican is even larger in the recent immigrant sample than in the longer term sample, except in Honolulu and San Antonio, as seen in Tables 6 and 7. Furthermore, the number of Cubans drops considerably in both of these contexts. The San Francisco, Honolulu, San Antonio, and San Jose metropolitan areas report no Cuban immigrants arriving in 1985 or later that fit the selection criteria, and there are only 17 Cubans in all of the other Asian and non-enclave areas put together. These two economic contexts also report few recent Nicaraguan immigrants, and, in most cases, few Colombians, as well. Table 2. Sample by Place of Origin by Place of Work in a Latino Enclave (LongTerm Immigrants)"
Place of Origin
Primary Metropolitan Statistical Areas Los Angeles- Miami-Hialeah, Jersey City, NJ Long Beach, CA FL New York, NY % n % n % n % n 280 47.9 446 2.6 5282 78.0 604 16.8 5 0.9 160 0.9 258 3.8 47 1.3 71 12.1 2744 16.2 250 3.7 550 15.3
Cuba Nicaragua Central America (except Nicaragua) Mexico 10 1.7 1275275.2 96 1.4 327 9.1 Colombia 71 12.1 158 0.9 429 6.3 729 20.2 South America 148 25.3 698 4.1 455 6.7 1344 37.3 (except Colombia) Total 585 100.0 16958 100.0 67,70 100.0 3601 100.0 *The terms Latino and Asian enclave refer to PMSAs containing these enclaves and do not imply that all workers in the sample are participating in the enclave economy.
Copyright 2001 by Stephanie Bohon
Latinos in Ethnic Enclaves
46
Table 3. Sample by Place of Origin by Place of Work in an Asian Enclave (LongTerm Immigrants)
Place of Origin
Primary Metropolitan Statistical Areas and Metropolitan Statistical Areas Anaheim-Santa Honolulu, HI San Francisco, San Jose, CA Ana, CA CA % n Yo n % n % n 67 2.0 0 0.0 39 1.9 12 1.2 13 0.4 0 0.0 128 6.3 24 2.4
Cuba Nicaragua Central America (except Nicaragua) 209 6.3 Mexico 2893 86.5 Colombia 39 1.2 South America (except Colombia) 122 3.6 Total 3343 100.0
1 6 1
5 13
7.7 46.2 7.7
399 19.8 1248 61.8 33 1.6
75 7.5 822 82.4 11 1.1
38.5 100.0
173 2020
53 997
8.6 100.0
5.3 100.0
Table 4. Sample by Place of Origin by Place of Work in a Non-Enclave Location (Long-Term Immigrants)
Place of Origin
Primary Metropolitan Statistical Areas and Metropolitan Statistical Areas Newark, NJ San Antonio, TXChicago, IL n % n % n YO n 130 4.0 235 35.2 8 0.9 25 3 0.1 6 0.9 8 0.9 3 149 4.6 71 10.6 15 1.7 30
Cuba Nicaragua Central America (except Nicaragua) Mexico 2769 Colombia 64 South America 149 (except Colombia) 3264 Total
9% 1.4 0.2 1.7
84.8 2.0 4.6
21 128 206
3.1 19.2 30.9
823 5 10
94.7 0.6 1.2
1641 10 42
100.0
667
100.0
869
100.0
1751 100.0
Copyright 2001 by Stephanie Bohon
93.7 0.6 2.4
47
Data and Methods
Table 5. Sample by Place of Origin by Place of Work in a Latino Enclave (Recent Immigrants)
Place o f Origin Cuba Nicaragua Central America (except Nicaragua) Mexico Colombia South America (except Colombia) Total
Primary Metropolitan Statistical Areas Los Angeles- Miami-Hialeah, FL New York, NY Jersey City, NJ Long Beach, C A n % n % n % n % 12 7.2 17 0.3 416 29.5 21 1.5 9 5.4 110 1.6 469 33.2 19 1.4 45 26.9 1396 20.9 166 11.8 301 21.6 15 27 59
9.0 16.2 35.3
4961 38 154
74.3 0.6 2.3
65 113 183
4.6 8.0 13.0
357 250 445
25.6 17.9 31.9
167
100.0
6676
100.0
1412
100.0
1393
100.0
Table 6. Sample by Place of Origin by Place of Work in an Asian Enclave (Recent Immigrants)
Place o f Origin
Primary Metropolitan Statistical Areas and Metropolitan Statistical Areas Anaheim-Santa San Francisco, CA San Jose, C A Ana, C A Honolulu, HI n % n % n % n % 4 0.2 0 0.0 0 0.0 0 0.0 14 0.7 0 0.0 52 7.2 6 1.3 146 7.1 2 40.0 135 18.7 37 8.3
Cuba Nicaragua Central America (except Nicaragua) Mexico 1863 Colombia 10 South America 30 (except Colombia) 2067 Total
90.1 0.5 1.5
2 0 1
40.0 0.0 20.0
485 9 41
67.2 1.2 5.7
386 2 16
86.4 0.4 3.6
100.0
5
100.0
722
100.0
447
100.0
Copyright 2001 by Stephanie Bohon
48
Latinos in Ethnic Enclaves
Table 7. Sample by Place of Origin by Place of Work in a Non-Enclave Location (Recent Immigrants) Place of Origin
Chicago, IL n % 2 0.2 7 0.7 70 7.0
Cuba Nicaragua Central America (except Nicaragua) 879 Mexico Colombia 15 South America 29 (except Colombia) 1002 Total
Primary Metropolitan Statistical Areas Newark, NJ San Antonio, TX San Diego, CA n % n % n % 11 4.2 0 0.0 0 0.0 5 1.9 2 1.6 0 0.0 60 23.2 5 3.9 30 3.9
87.7 1.5 2.9 100.0
44 53 86
17.0 20.5 33.2
116 3 1
91.3 2.4 0.8
731 2 6
95.1 0.3 0.8
259
100.0
127
100.0
769
100.0
VARIABLES
The Male Socioeconomic Index Most of the work on ethnic enclaves has focused on immigrant income and selfemployment, while very little attention has been paid to the issue of jobs, despite the fact that the acquisition of good jobs is an important indicator of economic success. This study extends work on ethnic enclaves by examining occupational attainment among immigrant workers. Tables 8 and 9 show the top ten occupations held by at least one-quarter of each of the immigrant groups in this study. Table 8 shows that the largest portion of Latin American immigrant men who arrived in the U.S. before 1985 are employed overwhelmingly in blue collar jobs, although a sizeable proportion of Cubans, Nicaraguans, Colombians, and other South Americans work as managers and administrators and as supervisors and proprietors of sales occupations. The occupations that seem to particularly attract long-term immigrants are truck drivers and janitors and cleaners, in that order. However, there is considerable variability in the distribution of these jobs across ethnic groups. For example, the largest portion of Cubans and South Americans are employed in management and administrative jobs, while the largest portion of Nicaraguans and Central Americans are truck drivers. Nearly five percent of the Mexicans in this study are grounds keepers and gardeners, while over five percent of Colombians are janitors and cleaners. Cubans, Colombians, South Americans, and, to some extent, Nicaraguans, are attracted to similar jobs, and these jobs
Copyright 2001 by Stephanie Bohon
Data and Methods
49
tend to be slightly higher on the Male Socioeconomic Index (MSEI) scale than those held in the greatest numbers by Mexicans and Central Americans. Table 8. Percent of Long-Term Immigrants in Commonly Held Occupations by Place of Origin
; Occupation (Census Code) Cuba Managers and Administrators 6.0 (022) Supervisors and Proprietors, 5.4 Sales Occupations (243) Cooks, except short order 0.5 (436) Janitors and Cleaners (453) 3.1 Groundskeepers and 1.0 Gardeners (486) Automobile Mechanics (505) 2.1 Machine Operators (779) 1.0 Assemblers (785) 1.5 Truck Drivers (804) 5.6 Helpers, Extractive 2.1 Occupations (869)
Central South Nicaragua America Mexico Colombia America 5.6 2.1 1.8 5.1 5.4
4.7
1.8
1.4
2.4
3.4
2.1
3.3
4.7
1.4
2.3
4.3 1.2
5.6 1.8
4.6 4.9
5.4 0.6
4.6 0.4
2.7 1.4 2.4 6.6 1.8
3.4 2.3 2.4 5.9 2.4
1.9 3.4 3.7 4.5 4.5
1.9 1.9 1.5 4.5 1.1
2.1 1.5 1.1 3.5 1.7
In Table 9 there is a slightly different picture. This table shows that the top ten jobs that account for at least 30 percent of all recent immigrants in the study are somewhat different than those reported in Table 8. Managers and administrators, supervisors and proprietors, automobile mechanics, and machine operators drop out of the list. They are replaced by food preparers, carpenters, painters, construction and maintenance workers, and non-construction laborers. In short, recent immigrants tend to be more concentrated in those jobs that require less English fluency or skills that are specific to the U S . labor market. It is interesting to note that there is somewhat more heterogeneity across place of origin for the occupations commonly held by recent versus long-term immigrants. Recent immigrants appear in larger proportions as janitors and cleaners, truck drivers, extractive occupation helpers, and cooks. There remains the pattern, however, of larger proportions in occupations shared by Cubans, Colombians and South Americans as one group, and Mexicans and Central Americans as another.
Copyright 2001 by Stephanie Bohon
Latinos in Ethnic Enclaves
50
Table 9. Percent of Recent Immigrants in Commonly Held Occupations by Place of Origin
Occupation (Census Code) Cuba Cooks, except short order 0.6 (436) Miscellaneous Food 1.7 Preparation Occupations (44.4) Janitors and Cleaners (453) 5.8 Groundskeepers and 1.4 Gardners (486) Carpenters (567) 4.3 Painters, Construction, and 1.0 Maintenance (579) Assemblers (785) 1.4 Truck Drivers (804) 5.2 Helpers, Extractive 4.8 Occupations (869) Laborers, except construction 3.3 (889)
Place of Origin Central South Nicaragua America Mexico Colombia America 2.7 4.9 7.6 2.9 4.5 2.9
3.4
4.3
1.9
3.3
5.6 1.6
8.1 3.9
5.5 6.5
5.7 1.4
5.2 1.4
2.7 2.2
3.3 4.3
2.8 2.0
1.5 2.9
2.5 3.7
4.8 6.2 5.6
2.8 3.0 5.6
3.6 2.2 5.9
2.7 5.9 3.3
1.8 5.4 2.9
3.0
3.2
3.3
1.5
1.O
Given the variation in types of jobs held by place of origin and the importance of good jobs in economic adaptation, the need to examine immigrant occupational attainment more fully seems obvious. In this study, occupational attainment is measured using Hauser and Warren's (1997) Socioeconomic Index for male workers. MSEI is a continuous scale similar to Stevens and Featherman's (1981) SEI formulations. Based on 1990 three-digit census occupations, the MSEI is calculated by regressing occupation on education, occupational earnings, and job prestige. The job prestige measure used in the creation of the MSEI was taken from work by Nakao and Treas (1994) using the 1989 General Social Survey. A male-specific scale was chosen for this study, because it seemed more appropriate for the restricted sample. Tables 10 and 11 shows the average MSEI scores by place of origin for the respondents included in this study. The results of this table show that there is considerable variation in MSEI across place and between ethnic groups. The results also show that longer-term immigrants have considerably higher occupational attainment (MSEI) scores than recent immigrants. These differences will be discussed in greater detail in Chapters 5 and 6.
Copyright 2001 by Stephanie Bohon
Data and Methods
51
While SEI scores are routinely used in occupational attainment research, there are two drawbacks in using the MSEI to study immigrant attainment. First, the use of the MSEI assumes that immigrants are interested in maximizing occupational prestige as well as earnings. This may be untrue, particularly for those immigrants engaging in circular migration (Kossoudji & Cobb-Clark, 1996). Second, use of the MSEI assumes that similar occupations require the same educational prerequisites and result in the same earnings for immigrant males as for other males in the population. Despite these difficulties, there are currently no immigrant-specific occupational attainment scales available based on 1990 Census occupational classifications. Consequently, MSEI is used as a good, alternative indicator of immigrant attainment. Table 10. Average Male Socioeconomic Index Scores (and Standard Deviations) by Place of Origin by Place of Work (Long-Term Immigrants). Place of Origin Central
Primary Metropolitan Statistical
(15.30) (12.85) Honolulu, Hawaii San Francisco, California San Jose, California
Copyright 2001 by Stephanie Bohon
(9.17) 12.92 (0.00) 39.86 28.97 27.29 (14.92) (12.55) (12.77) 48.37 40.86 28.84 ---
---
(9.09) 43.65 (21.O 1) 24.08 (9.81) 24.08
South
(17.17) (14.94) 21.91 51.72 (0.00) (22.49) 36.50 35.97 (14.30) (14.74) 44.24 39.94
52
Latinos in Ethnic Enclaves
Table 11. Average Male Socioeconomic Index Scores (and Standard Deviations) by Place of Origin by Place of Work (Recent Immigrants). Place of Origin Central
Primary Metropolitan Statistical
(8.98)
(8.04)
Honolulu, Hawaii
---
---
San Francisco, California
---
San Jose, California
---
19.63 (9.40) 20.17
(10.14) 22.42 (3.30) 20.59 (9.78) 18.14
(8.23) 23.77 (0.00) 19.38 (7.46) 19.46
South
(8.80) ---
22.79 (10.57) 40.09
(14.72) 26.46 (0.OO) 23.12 (10.31) 31.82
The Independent Variables The predictor variables used for this study are shown in Table 12. These measures include time of entry into the U.S., place of origin, self-employment status, educational attainment, age, potential work experience, English speaking ability, disability status, marital status, and household headship. Many of these variables have been commonly used to predict occupational attainment in other studies. Age, education, English fluency, and work experience are widely accepted measures of human capital (Borjas, 1991; Model & Ladipo, 1996). It is expected that older, more educated, more experienced immigrants with good English skills will have higher occupational attainment than other immigrants, ceterus
Copyright 2001 by Stephanie Bohon
Data and Methods
53
paribus. Work disability is also included since disabilities may lower occupational attainment through both physical limitations and discrimination (Tienda and Wilson, 1992). These human capital variables are included in the analysis, despite the fact that their effects may be unimportant within the enclave context (e.g., English language fluency should not predict occupational attainment where the market is dominated by Spanish speakers). Household characteristics including marital status and household headship are also included as control variables in the models. The role of these variables in predicting self-employment among enclave workers is well documented (Portes and Zhou, 1996; Sanders and Nee, 1996), while the relationship to occupational status attainment is less clear. It is expected, however, that married men, particularly those who are the household heads, face more social pressure to acquire good jobs than single men (Sanders and Nee, 1996). Self-employment is gaining increased attention as an important means of economic security for immigrant workers (Borjas, 1986; Evans, 1989; Min, 1988; Sanders & Nee, 1996; Waldinger, Ward, & Aldrich, 1985; Werbner, 1984). While some authors argue that the ability to enter into entrepreneurial ventures is a function of human capital (Borjas, 1986), others argue that factors such as age, education, and English ability explain only part of the transition to self-employment. Evans (1989), for example, contends that the size of the immigrant population and the "linguistic isolation" of the immigrant workers is also important, while Sanders and Nee (1996) demonstrate that family-related social capital factors also have an influence on immigrant entrepreneurship. While it is clear from the enclave literature that self-employment is at least one key to upward mobility, the relationship between self-employment and occupational attainment is less clear than the relationship between entrepreneurship and earnings. Nonetheless, self-employment is included in this study because of the possibility of this variable's interaction with other variables, particularly education. Time spent in the U.S. increases an immigrant's ability to acquire education and English skills. Additionally, each year spent in an enclave should be time spent building bounded solidarity and enforceable trust. Additional years of U S . residence should also bring an immigrant higher occupational attainment either through promotion or through more opportunities for self-employment. Consequently, year of entry is included in this analysis with categories determined by Census coding.
Copyright 2001 by Stephanie Bohon
Latinos in Ethnic Enclaves Table 12. Independent Variables Used in this Study Independent variable place of origin
year of entry
self-employed
educational attainment
age experience2 ability to speak english disabled
married
head of household
Description Place of birth. Categories are Cuba, Nicaragua, Central America (isthmus countries except Nicaragua), Mexico, Colombia and South America (mainland except Colombia). Time of entry into the the U.S. Categories are 1987 to 1990,1985 to 1986,1980 to 1984,1970 to 1979,1960 to 1969 and 1959 or earlier. Class of worker is self-employed in own business, professional practice, or farm. This is a dichotomous variable coded 1 if self-employed and 0 otherwise. Years of schooling completed. Categories are less than high school (0 to 8 years of school completed), some high school (9 to 12 years of school completed, but no diploma), high school graduate (diploma or equivalent), some college (associates degree in an occupational or academic program or other college attendance with no degree completed), and college graduate (Bachelor's Master's, professional, or Doctorate degree). Years of age. This variable is measured continuously. Potential years of job experience. This is a continuous variable calculated as (age - years of school - 612. Self-assessment of English speaking ability. Categories are very well, well, not well and not at all. Physically or mentally limited in the amount or kind of work undertaken. This is a dichotomous variable coded 1 if respondent is disabled and 0 otherwise. Now manied and living with spouse. This is a dichotomous variable coded 1 if the respondent is married and living with his spouse and 0 otherwise. Relationship of person to other members of the household is reported as householder. This is a dichotomous variable coded 1 if the respondent is the householder and 0 otherwise.
Finally, place of origin is the predictor variable of most interest in this study. In order to determine if queuing or segmented assimilation are the most likely mechanisms controlling immigrant adaptation, it is necessary to separate origin effects. That is, if Cubans in the Miami enclave, for example, engage in preferential hiring practices that create greater opportunities for other Cubans,
Copyright 2001 by Stephanie Bohon
Data and Methods
55
this can be determined only by controlling for place of origin effects. Place of origin in this study is reported place of birth categorized as Cuba, Nicaragua, Central America, Mexico, Colombia, and South America17. This categorization scheme assumes that immigrants will identify themselves and be identified by their birth place, despite their immigration history. The scheme also assumes a certain amount of internal homogeneity within broad geographic regions. Nicaraguans are separated from other Central Americans because of the unique circumstances surrounding the exodus from their homeland. Colombians are also separated from other South Americans because there is a Colombian enclave in New York City.
METHODOLOGY Within Area Models In order to test the hypotheses that immigrants will achieve occupational attainment that is less than or no different from the dominant enclave groups, Latino enclave PMSAs were analyzed separately using ordinary least-squares regression (OLS). OLS regression is a statistical technique that allows for the examination of the simultaneous impact of several variables on a single dependent variable-in this case, occupational attainment. The two samples (recent and longer-term immigrants) were analyzed separately in these models. Occupational attainment was regressed on place of origin, year of entry, self-employment, human capital, and household characteristics. Place of origin variables were dichotomized with the dominant enclave group omitted in each enclave's models, since the omitted-or reference-group is the group to which other groups are compared. Cubans are the reference group in the MiamiHialeah, Los Angeles-Long Beach, and Jersey City PMSA models, while Colombians serve as the omitted group in the New York City model. OLS models were not constructed for those areas with sample sizes less than 50, since small sample sizes may distort the findings. This is only a problem with the recent immigrant samples. The OLS models allow for an examination of the effects of place of origin on occupational attainment within each enclave controlling for other factors.
Across Area Models A comparison of the results of the OLS analysis should offer valuable information on the occupational attainment of different Latino immigrant groups relative to the dominant enclave ethnic group. However, such an analysis does not allow for the analysis of enclave effects separate from individual place effects. In other words, it may be the (hypothetical) case that Mexicans do better Copyright 2001 by Stephanie Bohon
56
Latinos in Ethnic Enclaves
in Los Angeles than in Miami because of features unique to the Los Angeles economy that have nothing to do with exploitation by Cuban-American or Cuban immigrants. Another procedure is needed to determine if the enclave phenomena has an effect on the attainment of immigrants. In order to examine context effects, regression models were run for both the recent and the longer-term immigrants in pooled samples from the various Latino, Asian, and non-enclave places. For the longer-term immigrant model, occupational attainment of all the long-term immigrants in this study are regressed against the variables used in the within area models. A place of work variable is also included as a dummy variable indicating place of work or unemployment in an Latino enclave, an Asian enclave, or a non-enclave area. Work in an Latino enclave is the omitted category. This analysis allows for the examination of place of work effects once the other variables are controlled. Another pooled OLS model was also constructed using the recent immigrant samples. In each of these models, Mexicans were selected as the omitted category because they appear in sizeable numbers across the three contexts and they represent one country of origin group. While this methodology allows for the testing of the question whether Latin American immigrants fare better versus no better in a Latino enclave area than in other areas, the extent to which discrimination may be occurring cannot be fully examined with an OLS model. In order to examine the average difference in MSEI attributable to discrimination across areas and between groups, OLS regression models were constructed for each place of origin group within each economic context where the samples sizes were sufficiently large (n 2 50). The average occupational attainment differences between Mexican workers and other workers was then decomposed into portions attributable to discrimination, endowments, and the interaction between the two using the method specified by Jones and Kelley (1984). The Jones and Kelley method was selected because it has a precedent in the literature on immigrant occupational attainment (Model & Ladipo, 1996). The formula allows for average differences between groups to be decomposed into three portions. The decomposition formula is as follows:
Copyright 2001 by Stephanie Bohon
Data and Methods
where the average difference between the occupational attainment (yM-Y') between Mexican immigrants (M) and another ethnic group (I) is the sum of (A) discrimination, (B) endowments, and (C) an interaction term. Regression decomposition analysis results in coefficients that indicate the amount of shortfall occurring between the two groups (i.e., negative signs represent advantages over Mexicans). In other words, differences in average MSEI between Mexicans and other groups that can be explained by differences in education, English fluency, marital status, etc., comprise equation (B). The remaining difference is unexplained and is assumed to be at least partially the result of discrimination.18 While not amenable to parametric testing, the results of the decomposition analysis offer a major advantage over OLS regression alone. The decomposition analysis allows for an examination of how the discrimination component varies between enclave types by examining both the size and range of the discrimination components. By using decomposition, we can actually make rank types of labor markets in terms of the amount of discrimination that appears to be occurring in those places. The analysis of data are presented in Chapters Five, Six, and Seven. Chapter Five examines differences between dominant enclave groups and other enclave workers within each of the four Latino enclave areas using the technique outlined above. This analysis allows for a direct examination of queuing and segmented assimilation. Chapter Six compares occupational attainment for longterm immigrants across the different labor markets using both pooled OLS models and regression decomposition. Finally, Chapter Seven replicates the analysis in Chapter Six using the recent immigrant sample. The summary of the findings of these chapters and their implications for assimilation theory, enclave theory, and immigration policy are discussed in Chapter Eight.
Copyright 2001 by Stephanie Bohon
CHAPTER FIVE
Occupational Attainment within Latino Enclaves
Segmented assimilation and queuing theory lend themselves to competing hypotheses regarding the effects of country of origin on immigrants' occupational attainment, as discussed in Chapter One. TWOof these sets of competing hypotheses pertain to the occupational attainment outcomes of different place of origin groups within the different Latino enclave areas. Question: Will Latino immigrants working in cities with Latino enclaves demonstrate lower or the same occupational attainment as the dominant enclave group after living in the United States for more than five years? Hypothesis H l a . Within cities with Latino economic enclaves, after more than five years in the U S . , Latino immigrant workers will demonstrate significantly lower occupational attainment, on average, than members of the dominant enclave group. Alternative Hypothesis H l b . After more than five years in the U S . , noncoethnic immigrants working in metropolitan areas with Latino enclaves will demonstrate average occupational attainment not significantly different from the dominant enclave group. Question: Will recent Latino immigrants working in cities with Latino enclaves demonstrate lower or the same occupational attainment as recent immigrants from the dominant enclave group ?
Copyright 2001 by Stephanie Bohon
Latinos in Ethnic Enclaves
60
Hypothesis H2a. Recent Latin American immigrants (those entering the country in 1985 or later) working in metropolitan areas with Latino enclaves will report occupational attainment significantly lower, on average, than that of recent immigrants in the dominant enclave group. Alternative Hypothesis H2b. Recent Latin American immigrants working in metropolitan areas with Latino enclaves will report occupational attainment not significantly different from that of recent immigrants in the dominant enclave group. The interethnic competition hypotheses (Hla and H2a), based on queuing theory suggest that members of the dominant enclave group will exhibit higher levels of occupational attainment than other immigrants. The interethnic cooperation hypotheses (HI b and H2b), based on segmented assimilation theory suggests that occupational attainment will not vary significantly between different country of origin groups, once year of entry, self-employment, household characteristics and human capital are controlled for. In order to test these hypotheses, models of occupational attainment for immigrants residing in the Latino enclave PMSAs of Jersey City, Los-AngelesLong Beach, Miami-Hialeah, and New York City were constructed using first the long-term then the recent immigrant samples. It seems reasonable to begin the analysis with the test of these hypothesis, because processes within areas will have an effect on the analysis of across-area differences. An examination of the within-area models is also important because enclave theory has, for the most part, been predicated on work conducted within (rather than across) enclaves.
LONG-TERM RESIDENTS Descriptive Statistics Tables 13 through 16 show the averages and proportions of the dependent and independent variables for the long-term immigrant sample by place of origin. Each table presents data for one of the four Latino enclave cities. The results in Table 13 through 16 show that there is variation in average immigrant group MSEI scores in all four enclaves. Occupational attainment rankings range from the mid-thirties for Cubans to the low- to mid-twenties for Mexicans. Cuban immigrants have higher MSEI scores than the other national origin groups in Jersey City, Los Angeles, and New York. It is interesting to note, however, that South Americans and Colombians display higher average MSEI scores in Miami, where Cubans are found in the greatest numbers. Mexicans have the lowest MSEI scores across all of the enclaves where they
Copyright 2001 by Stephanie Bohon
Occupational Attainment within Latino Enclaves
61
work in large numbers, with MSEI scores ranging from 25.10 in Los Angeles to 21.69 in New York. Table 13. Descriptive Statistics for the Jersey City PMSA (Long-Term Immigrants)
Cuba
Place of Origin* Central America Colombia
South America
MSEI Year of entry 1980 to 1984 1970 to 1979 1960 to 1969 1959 or earlier Self-employed Educational Attainment < high school Some high school High school grad Some college College graduate Age
Ability to Speak English Not at all Not well Well Very well Disabled Married Head of Household .79 .73 .68 .76 n 280 71 71 14 8 *Place of origin groups with sample sizes under 50 were omitted from the analysis. (Standard deviations in parenthesis)
Copyright 2001 by Stephanie Bohon
Latinos in Ethnic Enclaves
62
Table 14. Descriptive Statistics for the Los Angeles-Long Beach PMSA (Long-Term Immigrants)
MSEI
Place of Origin Central South Cuba Nicaragua America Mexico ColombiaAmerica 35.67 31.96 26.03 25.10 35.58 34.76
Year of entry 1980 to 1984 1970 to 1979 1960 to 1969 1959 or earlier Self-employed Educational Attainment
Ability to Speak English Not at all Not well Well Very well Disabled Married .80 Head of Household n 446 (Standard deviations in parenthesis)
Copyright 2001 by Stephanie Bohon
.76 160
.69 2744
.70 12752
.79 158
.79 698
63
Occupational Attainment within Latino Enclaves
Table 15. Descriptive Statistics for the Miami-Hialeah PMSA (Long-Term Immigrants)
Cuba
Place of Origin Central South Nicaragua America Mexico ColombiaAmerica
MSEI Year of entry 1980 to 1984 1970 to 1979 1960 to 1969 1959 or earlier Self-employed Educational Attainment < high school Some high school High school grad Some college College graduate Age
Ability to Speak English Not at all Not well Well Very well Disabled Married Head of Household .78 n 5282 (Standard deviations in parenthesis)
Copyright 2001 by Stephanie Bohon
.78 258
.74 250
.66 96
.81 429
.82 455
64
Latinos in Ethnic Enclaves
Table 16. Descriptive Statistics for the New York City PMSA (Long-Term Immigrants) Place of Origin* Cuba
Central America
Mexico
South Colombia America
MSEI Year of entry 1980 to 1984 1970 to 1979 1960 to 1969 1959 or earlier Self-employed Educational Attainment < high school Some high school High school grad Some college College graduate Age
Ability to Speak English Not at all Not well Well Very well Disabled Married .76 .70 .64 .72 Head of Household n 604 550 327 729 (Standard deviations in parenthesis) "Nicaraguans omitted from the analysis due to small sample size.
.74 1344
There is also variation in the MSEI scores within national origin groups across the enclaves. Cubans have the highest average MSEI scores in New York, the only one of the four cities without a Cuban enclave. Since prior research has suggested that most Cuban immigrants prefer to stay in Miami, or return to Miami after migrating to other cities (Portes & Bach, 1985), it is possible that Cuban immigrants who choose to settle in New York may do so because they
Copyright 2001 by Stephanie Bohon
Occupational Attainment within Latino Enclaves
65
are successful, while the less successful immigrants return to the Cuban enclaves. Central and South Americans have the highest MSEI scores in Miami (30.74 and 35.79, respectively), and Nicaraguans, Mexicans and Colombians have the highest average occupational attainment in Los Angeles. Each of the groups displays their lowest average MSEI scores in Jersey City, except those groups not examined in that labor market. Those groups are Mexicans, who have the lowest occupational attainment in New York (MSEI = 21.69), and Nicaraguans, who have the lowest occupational attainment in Miami (MSEI = 30.71). An examination of the other variables shows a consistent pattern. With a few exceptions, those groups with the highest average occupational attainment scores also have the highest levels of educational attainment, the highest average ages, the most potential work experience, and the highest proportion married and heads of households. The high MSEI scores are also correlated to high proportions of self-employment and greater English fluency. Overall, the advantaged groups appear to be Cubans, Colombians, and South Americans, while the disadvantaged groups are Mexicans and Central Americans. The story for Nicaraguans is less clear. Their MSEI scores lay somewhere in between the two groups, but they are more advantaged than some of the high occupational attainment groups on some of the independent variables. It should be expected from these preliminary findings, therefore, that MSEI is at least partially related to year of entry, self-employment, human capital and household characteristics. However, the extent that country of origin plays or does not play a role in predicting occupational attainment cannot be seen from these tables. In order to test for place of origin effects, then, OLS regression models were constructed for each enclave in order to test the competing hypotheses regarding segmented assimilation and queuing. Regression Analysis Ordinary least squares regression analysis was conducted for each Latino enclave examining the impact of place of origin on occupational attainment while controlling for other endowments. Table 17 shows the parameter estimates and associated standard errors for the OLS regression models for the Jersey City, Los Angeles-Long Beach, Miami-Hialeah, and New York PMSAs using the long-term resident immigrant sample. Parameter estimates are used to show the size of the impact of the independent variables on occupational attainment. Standard errors show the value of the variable in predicting MSEI, with smaller standard errors generally indicating better predictors. Model 1 shows country of Copyright 2001 by Stephanie Bohon
66
Latinos in Ethnic Enclaves
origin effects only. Model 2 shows the full model which also examines year of entry, self-employment, household characteristics, and human capital. For ease of comparison across areas, the models have also been labeled with the first letter of the enclave area (Model MI for Miami, Model J1 for Jersey City, etc.). Models 1 in Table 17 show the effects of country of origin on occupational attainment relative to Cubans in Jersey City, Miami, and Los Angeles, and relative to Colombians in New York. Nicaraguans were eliminated from the New York and Jersey City analysis and Mexicans were eliminated from the Jersey City analysis because the sample sizes for these groups were too small.19 The adjusted coefficients of determination (indicated as R2) for Models 1 range from .O1 in Miami to .07 in Los Angeles and New York. This shows that place of origin alone accounts for at least one percent-and up to seven percentof the variance in occupational attainment in the various samples. While these R2 values may be low by some standards, they are larger than expected, since there is no prior research suggesting that national origin within the same sending region is a significant predictor of occupational attainment for immigrants. The coefficient of determination for Los Angeles and New York are particularly large. An examination of the R2 value alone lends support for segmented assimilation in Miami, since national origin does not appear to have much effect; however, the larger R2 values in the other three enclaves suggest that queuing may be occurring there. These findings are interesting because much of the work on segmented assimilation can be attributed to Portes and Zhou (e.g., 1993) using data from the Miami area. In most of the areas, Central Americans show a significant disadvantage in occupational attainment, relative to Cubans, as do Nicaraguans and Mexicans. Furthermore, the occupational attainment of the Cuban, South American and Colombian groups are not significantly different from the dominant enclave group, on average. Some exceptions to this can be found in Miami, where South Americans have more than a two-point advantage in MSEI over the dominant group, and Jersey City, where Colombians and South Americans are significantly disadvantaged.
Copyright 2001 by Stephanie Bohon
Table 17. Impact of Country of Origin and Other Endowments on Occupational Attainment (MSEI) for Long-Term Latin American Immigrants in Hispanic Enclaves'r Primary Metropolitan Statistical Areas Jersey City, NJ Los Angela-Lung Beach, C A Model J1 Model 52 Model L1 Model L2 COUNTRY OF ORIGIN Cuba Nicaragua Central America Mexico Colombia South America
REFERENCE REFERENCE REFERENCE REFERENCE -3.71*** NOT USED NOT USED -2.34** (0.82) -6.43*** -4.25*** -2.96* (1 .SO) (1.62) (0.47) NOT USED NOT USED -4.87*** (0.44) -6.74*** -3.80** -0.93 (1.62) (1.47) (0.82) -1.41 -2.35* -1.59** (1.13) (1.24) (0.54)
YEAR OF ENTRY 1980 to 1984 1970 to 1979 1960 to 1969 1959 or earlier SELF-EIMPLOYED EDUCATIONAL ATTAINMENT Less than high school Some high school High school graduate Some college College graduate AGE
EXPERIENCE^
-2.46*** (0.24) -1.45*** (0.20) REFERENCE 1.07*** (0.35) 2.43*** (0.24)
-3.45* (1.65) -2.48* (1.3 1) REFERENCE 1.75 ( 1.46) 11.oo*** ( 1.55) 0.07
-1.92*** (0.25) -1 .so*** (0.24) REFERENCE 3.66*** (0.31) 11.02*** (0.34) 0.09*** (0.02) -o,oo*** (0.00)
-0.00 (0.00)
ABILITY TO SPEAK ENGLISH Not at all Not well Well Very well DISABLED MARRIED HEAD OF HOUSEHOLD ADJUSTED Rz n
-4.40*** (1.31) -2.57* (1.18) REFERENCE -0.66 (2.01) 5.48*** (1.26)
.0415 570
REFERENCE 1.87 (1.79) 1.11 (1.91) 4.02* (2.06) -0.89 (3.20) 2.07* (1.11) 0.27 ( 1.27) ,3061 570
Copyright 2001 by Stephanie ?Parameter estimates shownBohon (standard errors in parenthesis). *p<.o5; **p<.o1; ***p<.oo1
,0689 16958
REFERENCE 0.24 (0.25) 1.!lo*** (0.26) 3.07*** (0.28) - 1.29*** (0.42) 0.32* (0.18) 1.39*** (0.19) ,2604 16958 (continued)
Table 17 (continued). Impact of Country of Origin and Other Endowments on Occupational Attainment (MSEI) for Long-Term Latin American Immigrants in Latino Enclaves Primary Metropolitan Statistical Areas Miami-Hialeah New York Model MI Model M2 Model N1 Model N2
COUNTRY OF ORIGIN Cuba Nicaragua
REFERENCE REFERENCE 7.04*** (0.74) NOT USED
- 1.97** (0.76) -7.63*** (0.89)
Central America
-0.33 (0.64) -3.73** (0.77) REFERENCE REFERENCE
Mexico Colombia South America
YEAR OF ENTRY 1980 to 1984 1970 to 1979 1960 to 1969 1959 or earlier
SELF-EMPLOYED EDUCATIONAL ATTAINMENT Less than high school Some high school High school graduate Some college College graduate
AGE
EXPERIENCE^ ABILITY TO SPEAK ENGLISH Not at all Not well Well Very well
DISABLED MARRIED HEAD OF HOUSEHOLD ADJUSTED R~ Copyright 2001 by Stephanie Bohon
3.47*** (0.64) NOT USED
0.96 (0.62)
0.33 (0.51)
-2.57*** (0.37) -0.81* (0.36) REFERENCE -1.10* (0.36) 3.00*** (0.32)
-3.87*** (0.58) -1.96*** (0.51) REFERENCE 0.76 (0.89) 3.08*** (0.59)
-0.48 (0.54 - 1.29** (0.42) REFERENCE 3.99*** (0.44) 11.45*** (0.43) 0.36*** (0.04) -0.01*** (0.00)
-0.97 (0.74) -1.20* (0.56) REFERENCE 3.27*** (0.60) 13.58*** (0.64) 0.18** (0.06) -o.oo*** (0.00)
REFERENCE -0.02 (.50) 2.95*** (0.53) 5.55*** (0.57) - 1.75* (0.76) 1.24*** (0.34) 1.56***
REFERENCE - 1.33 (0.93) 0.50 (0.94) 4.06*** (0.97) -2.95** (1 .l6) 0.76* (0.43) 1.55*** (0.48) .3718
Occupational Attainment within Latino Enclaves
69
Adding the variables that have been shown to explain much of the variance in occupational attainment both reduces the effect of place of origin, in most places, and significantly increases the R2 value, as shown in Models 2. The effect of adding year of entry, self-employment, human capital, and household characteristics to the base model increases the coefficient of determination at least twenty percent, with the greatest increases seen in Miami, where R2 increases from .O1 to .37 with the addition of the control variables. In other words, in Miami, country of origin, year of entry, self-employment, human capital, and household characteristics account for 37 percent of the variance in occupational attainment among respondents in this sample. Adding the other variables renders the impact of South American origin, relative to Cuban origin, insignificant in Miami. The same is true for Central Americans in New York, relative to Colombians. In most of the other cases, adding self-employment, year of entry, human capital, and household characteristics to the model does not result in non-significant place of origin coefficients. However, it reduces the impact of national origin on occupational attainment by about half, relative to the dominant enclave group. There are two notable exceptions. Models MI and M2 shows that the negative impact of Nicaraguan origin on occupational attainment in the MiamiHialeah PMSA actually increases slightly (from -2.71 to -2.95) once the control variables are added. Descriptive statistics in Table 15 suggest that Nicaraguans may actually be better poised than Cubans to fare well in the Miami economy, since they have higher levels of education and comparable English skills. Model M2 shows, however, that once endowments are held constant, Nicaraguans still fare poorly relative to Cubans. The other exception is found in Models L1 and L2, where in the impact of Colombian origin on occupational attainment changes from a non-significant regression coefficient of -0.91 in Model L1 to a significant regression coefficient (p < .01) of -1.59, once the control variables are added. 20 Overall, the findings suggest mixed evidence for segmented assimilation and queuing. Mexicans, Nicaraguans, and other Central Americans are disadvantaged in most of the Latino enclaves, relative to the dominant enclave group, while Colombians and South Americans generally display occupational attainment similar to Cubans in the places where they are the dominant enclave group. Impact of the Control Variables While the variable of interest for the purpose of hypothesis testing is place of origin, many researchers contend that year of entry, self-employment, human capital, and household characteristics are important factors in labor market adaptation (Borjas, 1986; Duleeg & Regets, 1996; Kossoudji & Cobb-Clark, Copyright 2001 by Stephanie Bohon
70
Latinos in Ethnic Enclaves
1996; Model & Ladipo, 1996; Sanders & Nee, 1996). These contentions are underscored in the findings presented here. In each of the four PMSAs, year of entry after 1969 is significantly associated with lower occupational attainment, relative to the entering the U.S. between 1960 and 1969. The effects of entering the country prior to 1960 is less consistent across the metropolitan areas. In Jersey City and New York, there is no significant difference in occupational attainment between the pre-1960 and 1960-1969 cohorts. In Los Angeles-Long Beach, however, the earliest entrants have significantly higher occupational attainment than those who entered in the sixties, while in Miami-Hialeah, the pre-1960 cohort demonstrates somewhat lower occupational attainment, on average, than the reference group, even after controlling for place of origin. Self-employment is a significant and strong predictor of occupational attainment in each of the four Latino enclaves. In each area, MSEI scores are, on average, at least two points higher for those who work for themselves than those who work for others. In Jersey City, however, the magnitude of the impact of self-employment is particularly strong, with a regression coefficient of 5.48 (p<.001), indicating that immigrants who are self-employed report occupational attainment that is almost five and a half points higher on average than the occupational attainment of immigrants who work for others. The parameter estimate associated with self-employment is the smallest in Los Angeles-Long Beach at 2.43, and Miami-Hialeah and New York are slightly higher at 3.00 and 3.08, respectively (p<.001 for all PMSAs). The impact of the human capital variables on the model is mixed. Immigrants who reported some high school education had significantly lower MSEI scores than those with a high school education in all four PMSAs. Immigrants with less than nine years of education had significantly lower occupational attainment than high school graduates in Jersey City and Los Angeles-Long Beach, but the differences between the groups are not significant in Miami-Hialeah and New York. Likewise, having some college education is significantly associated with higher occupational attainment in Los AngelesLong Beach, Miami-Hialeah, and New York, but not in Jersey City. Finally, graduating with at least a Bachelor's degree is significantly related to occupational attainment, with college graduates demonstrating MSEI scores at least 11 points higher than high school graduates in each of the cities (p<.001). For the human capital variables age, experience squared, and disabled, the effect is again significant everywhere except Jersey City. Age has a significant, positive effect on occupational attainment in Los Angeles-Long Beach, MiamiHialeah, and New York City, although the effect is modest. The square of experience also has a significant impact in every area except Jersey City.
Copyright 2001 by Stephanie Bohon
Occupational Attainment within Latino Enclaves
71
Despite the high significance level, however, the actual magnitude of the impact is minimal. Disabilities that limit the amount or kind of work a person can do is also negatively associated with occupational attainment for three of the four areas, but few immigrants fall into this category as can be seen in the descriptive statistics. Obviously, the lack of significant association between these three human capital factors and occupational attainment in Jersey City is interesting and offers an important area for future research. The final human capital variable in these models, ability to speak English, also shows some inconsistencies across the four enclaves. First, there is no statistically significant difference between the average occupational attainment of immigrants who speak little English and those who speak none at all. In Jersey City and New York, there is also no significant difference between those who speak English well and those who speak no English, while in Los AngelesLong Beach and Miami-Hialeah, the impact is positive and highly significant. This difference may be a factor of the relatively larger Latin American immigrant and Latino populations in Los Angeles and Miami relative to New York and Jersey City. In the cities where there are higher proportions of Latinos, employers may be more tolerant of workers who speak good, albeit imperfect English, and may hire or promote them more often than employers in other cities. Finally, relative to those who do not speak English at all, the MSEI scores for those who speak English fluently is at least three points higher, on average. The association between occupational attainment and household characteristics is also tested in Model 2 for each of the four cities. Immigrants who are married and living with their spouse report significantly higher occupational attainment in all four areas, but the impact is much smaller in Los Angeles-Long Beach and New York. Heading a household is strongly associated with higher occupational attainment in Los Angeles-Long Beach, MiamiHialeah, and New York, but not in Jersey City. It is interesting to note that while most of the control variables are significant in Los Angeles-Long Beach, Miami-Hialeah, and New York, and few of the variables are significant in Jersey City relative to the other Latino enclaves, the coefficient of determination (R2)is large in each of the four areas. Furthermore, there is little difference in the R2 values associated with the full models in each of the PMSAs. In fact, the lowest R2 value is found in the model for Los Angeles-Long Beach (.2604),not Jersey City (.3061). To summarize these findings, while place of origin does have an important impact on occupational attainment for most groups, human capital, year of entry, self-employment, and, to some extent, household characteristics still play an important role in determining this type of economic success. The continued ability of these variables to predict occupational attainment suggests that
Copyright 2001 by Stephanie Bohon
72
Latinos in Ethnic Enclaves
achieved characteristics help immigrants to overcome some of the structural barriers that they may face in the marketplace. However, the extent to which these endowments can be translated into good jobs varies from enclave to enclave, with some enclaves (such as Jersey City) offering fewer opportunities for immigrants to improve their occupational standing by improving their marketable characteristics.
RECENT IMMIGRANTS The expected theoretical impact of country of origin on occupational attainment for recent immigrants is less clear. Segmented assimilation cannot be usefully applied to the experience of recent immigrants, since the theory assumes an over-time adjustment process. Nonetheless, the basic tenets of ethnic enclave theory (as outlined in Chapter Three) suggest that all recent Latin American immigrants should fare equally well in Latino enclaves since they provide, at best, a steady supply of labor skilled in the production of ethnically defined goods, or, at worst, a cheap source of exploitable labor. Queuing theory also does not offer a clear hypothesis regarding the occupational attainment of recent immigrants. The actual allocation of jobs may favor one group over another, however, that impact should be more fully realized in outcomes such as unemployment than in occupational attainment.21 The occupational attainment experienced by recent immigrants under a queuing theory model may depend largely on the size of the available labor pool and the quality and number of entry level jobs. Where good jobs are scarce, queuing may occur. Where they are plentiful, there may be less reason to discriminate in hiring and promotion. However, in keeping with the general tenor of Hypotheses H l a and H l b , alternative hypotheses regarding the occupational attainment of recent immigrants are postulated in Hypotheses H2a and H2b reported at the beginning of this chapter. It is hypothesized that recent immigrants will have occupational attainment significantly lower than that of the dominant enclave group, or that recent immigrants will display occupational attainment not significantly different from the dominant enclave group.
Descriptive Statistics Because the sample size of the dominant enclave groups is less than 50 in both Jersey City and Los Angeles-Long Beach, hypotheses H2a and H2b were tested for Miami-Hialeah and New York only. In the New York enclave, Cubans and Nicaraguans were omitted from the analysis due to small sample size (see Table
Copyright 2001 by Stephanie Bohon
Occupational Attainment within Latino Enclaves
73
5). Descriptive statistics for the remaining immigrants are shown in Tables 18 and 19. Tables 18 and 19 show that recent immigrants in Miami and New York have lower average occupational attainment, less education, and less ability to speak English well than longer term immigrants (Tables 15 and 16). They are also younger and less likely to be married or heads of households. The portion of recent immigrants self-employed is also much smaller than for the longer term immigrant samples. Table 18. Descriptive Statistics for the Miami-Hialeah PMSA (Recent Immigrants) Place of Origin Central South Cuba Nicaragua America Mexico Colombia America MSEI Year of entry 1987 to 1990 1985 to 1986 Self-employed Educational Attainment < high school Some high school High school grad Some college College graduate Age
Ability to Speak English Not at all Not well Well Very well Disabled Married Head of Household .62 n 416 (Standard deviations in parenthesis)
Copyright 2001 by Stephanie Bohon
.54 469
.47 166
34 65
58 113
.68 183
Latinos in Ethnic Enclaves
74
There is, however, considerable variation across the place of origin groups and across the two enclaves used in examining the recent immigrants. Among recent arrivals in Miami, both South Americans and Colombians display higher average MSEI scores than Cubans, despite the large Cuban population in the enclave. This trend may be partially due to self-employment, which is higher among the two South American groups, as well as higher levels of college education and greater English fluency. Table 19. Descriptive Statistics for the New York City PMSA (Recent Immigrants) Place of Origin*
MSEI Year of entry 1987 to 1990 1985 to 1986 Self-employed Educational Attainment < high school Some high school High school grad Some college College graduate Age ~x~erience~
Central America 22.91 (9.10)
Mexico 18.88 (8.54)
Colombia 25.54 ( 10.77)
South America 25.47 (13.02)
.57 .43 .06
.63 .37
.47 .53 .10
.55 .45 .06
.45 .23 .18 .07 .07 29.39 (8.37) 523.43 (498.29)
.45 .25 .20 .06
.17 .22 3.5 .13 .13 3 1.04 (9.22) 497.12 (550.29)
.21 .17 30 .15 .17 3 1.71 (8.59) 498.92 (493.70)
.04
.04
25.82 (7.01) 350.25 (371.61)
Ability to Speak English Not at all .29 .22 .15 Not well .36 .41 .44 Well .18 .14 .26 Very well .18 .22 .15 Disabled .01 .03 .O1 Married .22 .I5 .35 Head of Household .32 .23 .42 n 301 357 250 (Standard deviations in parenthesis) "Nicaraguans omitted from the analysis due to small sample size.
Copyright 2001 by Stephanie Bohon
.I6 37 .24 .23 .O1 32 .40
445
Occupational Attainment within Latino Enclaves
75
Recent Central American immigrants, on average, have higher MSEI scores than both Nicaraguans and Mexicans. The Central American-Mexican difference can be attributed at least partially to higher levels of self-employment, more high school and college graduates, older ages, and marginally better English skills, as well as more favorable household characteristics. The Nicaraguan deficit is less easy to explain with the descriptive statistics. Nicaraguans have higher levels of education and more favorable household characteristics than Central Americans, despite the fact that they have lower average occupational attainment. They are roughly equal to their Central American counterparts in English fluency, age, and job experience. Average occupational attainment in New York is considerably lower for each of the recent Latin American immigrant groups than in Miami. Each of the immigrant groups examined in the New York PMSA show occupational attainment at least two points lower than in Miami. Standard deviations are lower in New York, as well. This finding persists despite the fact that a greater portion of the New York immigrants are in the earliest year of entry category used here. On the other hand, smaller proportions of the recent immigrants in New York are selfemployed or well educated, and larger portions report speaking English not well or not at all in this enclave. Immigrants in New York also have considerably lower proportions married or heading households than in Miami, and New York immigrants are younger and have less potential work experience, on average. As in the other areas, New York's Colombians and other South Americans appear to be slightly advantaged in nearly all of the human capital, selfemployment, year of entry, and household characteristic measures. Insofar as these endowments contribute to occupational attainment, it should be expected that Colombians and South Americans will have significantly higher occupational attainment, on average. Regression Analysis In order to examine the effects of these variables on the occupational attainment of recent immigrants, MSEI was regressed on place of origin and the control variables for the Miami-Hialeah and New York PMSAs. The results of the OLS regression models, including parameter estimates and standard errors, are shown in Table 20. As in the models of the long-term resident immigrants, Models 1 show the impact of regressing occupational attainment on country of origin only. Models 2 show the results for the full models. Model M I shows the impact of country of origin on MSEI for recent immigrants in Miami. Relative to recent Cuban immigrants, Nicaraguans, Mexicans and other Central Americans have significantly lower occupational Copyright 2001 by Stephanie Bohon
76
Latinos in Ethnic Enclaves
attainment. South Americans, on the other hand, have relatively higher MSEI scores, on average, than Cubans. These findings are similar to those found for long-term immigrants, although the coefficient of determination is considerably larger in the recent immigrant model, which suggests that country of origin is more important in the short-run. In Miami, country of origin explains more than five percent of the variance in occupational attainment among recent immigrants, while the same variable explains only one percent of the variance for long-term workers. The results of the regression models for the New York City PMSA are shown in Model N1. Like the long-term immigrant model results presented in Table 17, Central Americans and Mexicans have significantly lower occupational attainment than Colombians. These findings are also similar to those found for recent immigrants in Miami, although the reference group is different. Also similar to the long-term model is the finding that there is no significant difference between Colombians and other South Americans. Adding the control variables improves both models' fit, as seen in the larger R2 value (.22 for Miami and .26 for New York), although the improvement is not as large as in the long-term models (Table 17).In Miami, adding the control variables reduces the impact of Nicaraguan origin by about one point, on average. Again, as with the long-term models, it appears that Nicaraguans are significantly disadvantaged in the Miami-Hialeah enclave, relative to Cubans. Somewhat different from the long-term model, however, is the reduction of the impact of place of origin to non-significant levels for each of the other place of origin groups. In New York, adding the control variables in Model N2 reduces the effect of place of origin on Mexicans and Central Americans, but the effects remain significant for the two immigrants groups. This is different from the Miami model. In fact, the effects of national origin on occupational attainment in the full model for recent immigrants are stronger in the New York enclave for Mexicans and Central Americans than for long-term residents. This finding may reflect differences in the two labor markets and discrimination that may be occurring outside of the enclave context. In Miami where the enclave is larger and more diverse, there may be more opportunities for enclave employment. In New York, however, where the enclave sector is more specialized, there may be less opportunities for recent immigrants to find work within the enclave, and they may be working in the larger labor market area where they can face greater discrimination. This possibility will be explored in Chapter Seven.
Copyright 2001 by Stephanie Bohon
Table 20. Impact of Country of Origin and Other Factors on the Occupational Attainment (MSEI) of Recent Latin American Immigrants in Latino Enclaves? Primary Metropolitan Statistical Areas Miami-Hialeah, FL New York, NY Model 1 Model 2 Model 1 Model 2 COUNTRY OF ORIGIN Cuba Nicaragua
Central America Mexico Colombia South America YEAR OF ENTRY 1987 to 1989
REFERENCE REFERENCE NOT USED NOT USED -2.91*** - 1.95** NOT USED NOT USED (0.77) (0.73) -1.48* -2.05* -0.58 -3.29*** (0.81) (1.04) (0.99) (0.88) -3.87*** -6.02*** -2.27 -6.32*** (0.81) (1.52) (1.45) (0.85) - 1.17 REFERENCE REFERENCE 1.75 (1.21) (1.14) 4.85*** 1.37 0.27 -0.49 (1.01) (0.97) (0.81) (0.73) -0.30 (0.59) REFERENCE 4.28*** (0.94)
0.30 (0.54) REFERENCE 2.12* (1 .05)
-1.78 (1 .O7) -0.93 (0.86) REFERENCE 1.16 (1 .O6) 6.44 (1 .OO) 0.22** (0.09) -o.oo* (0.00)
-1.56* (0.87) - 1.99* (0.77) REFERENCE 0.98 (0.95) 11.34*** (1.01) 0.17* (0.09) -o.oo* (0.00)
REFERENCE 0.05 (0.74) 2.54** Well (0.92) 5.1 I*** Very well (0.92) DISABLED -0.66 (2.48) 1MARRIED 0.90 (0.73) 2.20** HEAD OF HOUSEHOLD (0.74) ADJUSTED ~2 ,0533 ,2177 n 1412 1412 ?Parameter estimates shown (standard errors in parenthesis). *p<.05; **p<.01; ***p<.001 Copyright 2001 by Stephanie Bohon
REFERENCE -0.29 (0.72) 1.26 (0.85) 3.56*** (0 84) 1.42 (1 .92) 0.47 (0.70) 1.76** (0.65) ,2607 1393
1985 to 1986 SELF-EMPLOYED EDUCATIONAL ATTAINMENT Less than high school
Some high school High school graduate Some college College graduate AGE
EXPERIENCE^ ABILITY TO SPEAK ENGLISH Not at all Not well
,0597 1393
78
Latinos in Ethnic Enclaves
An examination of the control variables shows that, unlike the long-term model, year of entry has no impact on occupational attainment in the Miami enclave, nor does disability or marriage status. Educational attainment is also non-significant, except for one category. Immigrants with a Bachelor's degree or better have significantly higher occupational attainment than those with only a high school education. The impact of a college education, however, relative to a high school education for recent immigrants is nearly half the size of the impact for long-term immigrants. Like with the long-term sample, age and experience-squared are significant in the recent immigrant model in Miami, although the effects are minor. Once again the proxy for experience is negative, indicated that the increase in occupational attainment declines at later ages. Also like the long-term model for Miami, speaking English well or very well is a significant predictor of occupational attainment, relative to speaking no English, while there is no difference in the occupational attainment between poor and non-speakers. Finally, heading a household is a strong and significant predictor of occupational attainment in Miami with householders reporting MSEI scores two points higher, on average, than those who are not. The significance level associated with this variable is lower in the recent model than in the long-term model, however, the effect in the former appears to be greater. The impact of the control variables in New York is similar to Miami, with a few notable exceptions. Immigrants without high school diplomas have significantly lower MSEI scores than immigrants with a high school degrees, on average. Also, immigrants who speak English well do not have occupational attainment levels significantly different from those who speak no English at all. This finding, however, is similar to that found in the long-term immigrant model for the New York PMSA. Overall, it can be concluded that, except for Nicaraguans, segmented assimilation appears to be a better explanation of occupational attainment for recent immigrants in Miami-Hialeah than the queuing hypothesis. The findings for New York, however, offer a less straightforward interpretation. For Mexicans and Central Americans, place of origin appears to affect occupational attainment for recent immigrants. For all other groups there is no difference between their occupational attainment, on average, and that of the dominant enclave group. For South Americans, therefore, segmented assimilation offers a better explanation of the early labor force experience of these immigrants than queuing, although this is not true for other groups.
Copyright 2001 by Stephanie Bohon
Occupational Attainment within Latino Enclaves
79
CONCLUSIONS In this chapter, competing hypotheses regarding the effects of country of origin on occupational attainment are tested for both recent and long-term immigrants. The results show a mixed picture that varies from group to group and place to place as well as across the sample. In order to facilitate a better understanding of these findings, the results are summarized in Table 21. In hypotheses H l a and H2a it is proposed that recent and long-term immigrants in Latino enclave cities will achieve occupational attainment significantly lower than that of the dominant enclave group. These hypotheses are predicated on an interethnic competition model derived from queuing theory. Insofar as the dominant enclave group is able to control the labor market, they will reserve better jobs for members of their own group. Significant differences in the regression coefficients associated with place of origin using the dominant enclave group as a reference are consistent with these hypotheses and are labeled as supportive of queuing theory in Table 21. The competing hypotheses (Hlb and H2b) suggest no difference between groups, once endowments are controlled. Lack of significant differences are consistent with segmented assimilation and are indicated as such on Table 21. Overall, the findings suggest that there is both support for segmented assimilation and for queuing, depending on the enclave and the particular group under discussion. Support for the queuing or segmented assimilation hypotheses varies from enclave area to enclave area. Segmented assimilation appears to be much more likely in Miami than in the other PMSAs with enclaves for either sample, although some queuing does seem to be occurring there, as well, particularly for Nicaraguan immigrants. One possible explanation for this may lie in the fact that there is less sectoral specialization in the Miami-Hialeah labor market than in the other areas. Consequently, while there are not necessarily more jobs in this market, there may be a wider variety of jobs, and employers are freer to match employees with jobs more carefully. A more plausible explanation is that, given the high levels of self-employment among Colombians and other South Americans in the Miami enclave, what appears to be evidence of segmented assimilation for some groups is just more queuing, albeit not perpetuated by the dominant enclave group.
Copyright 2001 by Stephanie Bohon
Latinos in Ethnic Enclaves Table 21. Summary Findings
S ANGELES-LONG
There are also differences in the findings across the two immigrant sample groups. That is, there is more evidence of segmented assimilation among recent versus long-term workers in Miami and more evidence of queuing among recent vs. long-term workers in New York. As mentioned previously, lack of a queuing effect may signal an abundance of entry level jobs and less ability to discriminate. This may be the case in Miami. Another explanation for the greater discrimination among longer term immigrants may be due to social class. Insofar as recent Cuban immigrants are viewed as lower class, they may be viewed by Cuban employers as no different from other immigrants. If this is the case, then segmented outcomes are occurring, not because other groups are moving up, but because the recent Cuban entrants are experiencing downward assimilation.
Copyright 2001 by Stephanie Bohon
Occupational Attainment within Latino Enclaves
8I
The fact that more queuing appears to be occurring among recent, rather than long-term immigrants in New York may be due to geographic factors. The Miami-Hialeah enclave is rather isolated. On the other hand, workers facing poor treatment in New York can easily commute outside of the enclave for work. However, these workers must be aware of other opportunities. Consequently, employers are freer to discriminate against recent immigrants and may stack the queues in favor of their own group. Groups that appear to be more subject to queuing (or, alternatively, less able to assimilate into the dominant group) are Mexicans, Central Americans, and to some extent, Nicaraguans. On the other hand, South Americans and Colombians usually display levels of occupational attainment commensurate with Cubans in the Cuban enclaves. The South American groups appear to be more assimilated within the enclaves and have the same opportunities as Cubans in most areas. At the same time, the Central American groups appear to face more structural barriers that limit their ability to fully realize their endowment potential. This is seen in the fact that controlling for endowments reduces the effects of place of origin on occupational attainment for most immigrants but not for all. Mexican immigrants who came to the U S . before 1985 are disadvantaged in each of the Latino enclaves where they work in large numbers. Long-term resident Central Americans are disadvantaged in Jersey City, Los Angeles-Long Beach, and Miami-Hialeah, while Colombians are disadvantaged in Jersey City only, and Nicaraguans are disadvantaged in Miami-Hialeah and Los AngelesLong Beach. Recent Nicaraguan immigrants have lower occupational attainment in Miami, while recent Central American and Mexican immigrants are disadvantaged in New York but not in Miami, relative to the dominant enclave group. These findings have important implications for adaptation theory. First of all, they suggest that assimilation, even into a minority group, is easier for some groups than for others. In fact, assimilation into another minority group may be no easier for some groups than assimilation into white America. A second implication of these findings is that the extent to which a group experiences segmented assimilation or queuing may depend largely on labor market area factors that determine the ability of the dominant enclave group to discriminate. It can be hypothesized that in those areas where immigrants have opportunities outside of the enclave (either in the larger labor market or in another enclave), segmented assimilation will be more likely to occur. Furthermore, in areas where unemployment rates are low and there is a great deal of industrial complexity, employers may have less reason to queue and may be more likely to match jobs with appropriately skilled employees.
Copyright 2001 by Stephanie Bohon
82
Latinos in Ethnic Enclaves
A final implication of the findings presented here is that time in the U S . , self-employment status, human capital and household characteristics are all important predictors of occupational attainment, despite enclave theory that suggests that these variables may be of minimal importance. It can be noted, however, that endowments appear to be less important in the first five years after immigration, suggesting that Latin American immigrants must "pay their dues" in lower levels jobs after first arriving in the enclave, despite the skills that they bring to the market. As a consequence, the enclave may provide more or less economic opportunities for recent immigrants, depending on the endowments that they bring with them.
Copyright 2001 by Stephanie Bohon
CHAPTER SIX
Discrimination and Competition among Long -Term Immigrants
An examination of occupational attainment within the four Latino enclave areas shows that Mexicans, Nicaraguans, other Central Americans, and even Colombians are subject to queuing in some labor markets. Compared to the dominant enclave group, the ability of groups to achieve occupational attainment commensurate with the endowments that they bring to the marketplace may be limited by factors related to place of origin. Despite this disadvantage, however, it is possible that Latin American immigrants are still better off in these areas than in other types of enclaves or non-enclave cities. That is, while Mexicans and Central Americans, for example, have significantly lower occupational attainment, on average, than Cubans (or Colombians in New York), these low levels may still be higher than those found in other places. The possibility of this outcome is discussed in Chapter One and the following alternative hypotheses are postulated: Question: Will Latino immigrants working in metropolitan areas with Latino enclaves demonstrate higher or the same occupational attainment as Latino immigrants working in Asian enclaves and non-enclave areas after living in the US.for more than five years? Hypothesis H3a.Comparing across economic enclave and non-enclave contexts, Latin American immigrants in metropolitan areas with Latino enclaves will have higher occupational attainment after more than five years in the U S . than Latin American immigrants in Asian enclave areas or non-enclave places, holding measurable endowments constant.
Copyright 2001 by Stephanie Bohon
Latinos in Ethnic Enclaves Alternative Hypothesis H3b.Comparing across economic enclave and nonenclave contexts, Latin American immigrants in metropolitan areas with Latino enclaves will have occupational attainment after more than five years in the U.S. not significantly different from Latin American immigrants in Asian enclave areas or non-enclave places, holding constant measurable endowment characteristics. Question: Will longer-term Latino immigrants (i.e., those living in the United States for more than five years) working in metropolitan areas with Latino enclaves experience less or the same amount of discrimination in job hiring and promotion as Latino immigrants working in Asian enclave or non-enclave areas? Hypothesis H5a.Comparing across economic enclaves and non-enclave contexts, discrimination as a determinant of occupational attainment differences among longer term Latin American immigrant groups is lower in metropolitan areas with Latino enclaves and greater in areas with Asian enclaves and nonenclave labor markets. Alternative Hypothesis HSbComparing across economic enclave and nonenclave contexts, discrimination as a determinant of occupational attainment differences among longer-term Latin American immigrant groups is no different in metropolitan areas with Latino economic enclaves than in areas with Asian enclaves and non-enclave labor market contexts.
A test of these alternative hypotheses necessitates a shift in analysis from a within context to an across context analysis where the impact of labor market type can be assessed and other variables can be held constant. Accordingly, each of the twelve PMSAs in the three economic contexts (areas with Latino enclaves, areas with Asian enclaves, and non-enclave areas) were combined into three samples and descriptive statistics for each of the three labor markets by place of origin were calculated. Then the samples were pooled, and MSEI was regressed on the independent variables controlling for labor market context. This method allowed for a test of the competing hypotheses regarding the impact of labor market area on attainment. Finally, average differences in MSEI between the control group and the other immigrant groups were decomposed using the methodology discussed in Chapter Four in order to fully explore the extent to which discrimination across national origin groups (as determined by the unexplained differences in average MSEI) takes place within the different types of labor markets. These findings are discussed below.
Copyright 2001 by Stephanie Bohon
Discrimination and Competition among Long-Term Immigrants
85
RESULTS Descriptive Statistics Tables 22,23 and 24 show the averages, standard deviations, and proportions for the variables analyzed in each labor market context. Table 22 shows the descriptive statistics for long-term immigrants in Latino enclave areas (Jersey City, Los Angeles, Miami, and New York combined), while Tables 23 and 2 4 show statistics for the Asian enclave (Anaheim, Honolulu, San Francisco and San Jose) and non-enclave areas (Chicago, Newark, San Antonio, and San Diego) respectively. Place of origin groups with sample sizes of less than fifty were eliminated from the analysis. Table 22 shows that, in the Latino enclaves, Cubans have the highest levels of occupational attainment followed by South Americans and Nicaraguans. This is also true for the Asian enclaves (Table 23). In the non-enclave areas, again Cubans have the highest average MSEI scores at 34.96, as seen in Table 24. This score is much closer to that found in the Latino enclaves (33.78) than in the Asian enclaves (40.05). Nicaraguans are omitted from the non-enclave analysis due to sample size; however, the patterns for the other groups reflect rankings consistent with both the Asian and Latino enclaves and numbers similar to those found in the Latino enclaves. This finding probably reflects to some extent the fact that more than half of all Cuban men in the sample arrived in the U.S. prior to 1970, while more than half of each of the other national origin groups entered the U S . between 1970 and 1984. In general, the distribution of the sample by year of entry by immigrant group does not vary much between the different labor market types.
Copyright 2001 by Stephanie Bohon
Latinos in Ethnic Enclaves
86
Table 22. Descriptive Statistics by Place of Origin for Long-Term Immigrants Workers in Latino Enclaves Cuba
Central South Nicaragua America Mexico Colombia America
MSEI Year of entry I980 to 1984 1970 to 1979 1960 to 1969 1959 or earlier Self-employed Educational Attainment Less than high school Some high school High school graduate Some college College graduate Age
Ability to speak English Not at all Not well Well Very well Disabled Married Head of household .782 ,764 n 6612 470 (standard deviations are in parentheses)
Copyright 2001 by Stephanie Bohon
,698 3615
,699 13181
,757 1387
.770 2643
Discrimination and Competition among Long-Term Immigrants
87
Table 23. Descriptive Statistics by Place of Origin for Long-Term Immigrant Workers in Asian Enclaves Cuba
Central South Nicaragua America Mexico Colombia America
MSEI Year of entry 1980 to 1984 1970 to 1979 1960 to 1969 1959 or earlier Self-employed Educational Attainment Less than high school Some high school High school graduate Some college College graduate Age
Ability to speak English Not at all Not well Well Very well Disabled Married .788 .697 Head of household n 118 165 (standard deviations are in parentheses)
Copyright 2001 by Stephanie Bohon
.697 684
.658 4966
.738 84
.816 353
Latinos in Ethnic Enclaves
88
Table 24. Descriptive Statistics by Place of Origin for Long-Term Immigrant Workers in Non-enclave Areas
MSEI
Cuba 34.96
Central America 27.94
Mexico 24.84
South Colombia America 32.33 32.94
.770 265
.745 5254
.761 205
Year of entry 1980 to 1 984 1970 to 1979 1 960 to 1969 1959 or earlier Self-employed Educational Attainment Less than high school Some high school High school graduate Some college College graduate Age
Ability to speak English Not at all Not well Well Very well Disabled Married ,816 Head of household n 397 (standard deviations are in parentheses)
.788 406
Cubans also have high rates of self-employment. In the Latino enclave, over 20 percent of the Cuban men in the sample are self-employed, compared to less than 10 percent of the Mexicans and Central Americans, and 12 to 15 percent for the other groups. Patterns of self-employment by country of origin are considerably different in the Asian enclave than in the Latino enclave areas. First, the proportion of self-employment for Latin American immigrants, except for South Americans, is markedly smaller in the Asian enclave areas. Cuban immigrants, for example, have only slightly higher than 12 percent
Copyright 2001 by Stephanie Bohon
Discrimination and Competition among Long-Term Immigrants
89
entrepreneurship rates in the Asian enclaves compared to 21 percent in the Latino enclaves. Also, about twice as many Colombians are self-employed in the Latino enclaves compared to the Asian. In the non-enclave areas, selfemployment is also highest among the Cuban and South American immigrants, however, the proportion self-employed is generally lower for most of the groups in the non-enclave than in the enclave areas. Interestingly, the proportion of South Americans self-employed is fairly stable across each of the labor market types, while the proportions for the other groups varies markedly from place to place. Cubans are also advantaged over most of the other immigrants in both age and job experience. The average age for the Cuban immigrants in Latino enclaves, 44.86 years, is considerably higher than the average for the lowest group (Central Americans at 25.90 years), and is also somewhat higher than the next highest group, South Americans. This pattern is also seen in the nonenclave places. In the Asian enclaves, the average age and experience of the Cuban men is higher than for most of the groups, but it is not the highest. South Americans are the oldest group, with an average age just slightly above that of the Cubans. Mexicans, however, show the highest average level of potential work experience. However, the impact of age and experience on MSEI was shown to be inconsistent in the preceding chapter. The impact of having a work disability is also unclear, although there is considerable variance in the proportions disabled among the immigrant groups. One interesting finding in the Latino enclaves (Table 22) is that Nicaraguans, not Cubans or South Americans, have the highest proportion of college graduates. In fact, almost half of the Nicaraguans in the Latino enclave sample have some schooling beyond high school. South Americans and Colombians also have high levels of education, with 45 and 41 percent of those groups, respectively, reporting at least some college. Cubans ranked fourth at 39 percent. In the Asian enclaves, education levels are much higher for all of the immigrant groups (except Nicaraguans) than in the Latino enclaves. In that context, Cuban men have highest proportion of college educations, with 6 4 percent reporting at least some college. More than half of the South American and Colombian men in the sample also reported acquiring at least some college education. The proportion of Nicaraguan men with at least some college education remained about the same in the two enclave labor markets, but Mexicans and Central Americans reported much higher proportions of college education at 13 and 32 percent, respectively. In the non-enclave areas, South Americans have the highest proportions of college educated men at 45 percent, which is somewhat lower than the
Copyright 2001 by Stephanie Bohon
90
Latinos in Ethnic Enclaves
proportion reported among those living in the Asian enclaves, but nearly identical to the proportion in the Latino enclaves for that group. The proportion of Cubans reporting some education (.43) was among the highest of the groups in the non-enclave area, but more than 20 percent lower than the level for that group in the Asian enclaves. The proportion college educated among the Central Americans and Mexicans again was the lowest of the immigrant groups in the non-enclave area, but not markedly different from the educational attainment levels found in the Latino enclaves. English speaking ability was also mixed across the economic contexts, with more than 70 percent of Colombians and almost 80 percent of South Americans in the Latino enclaves reporting that they spoke English well or very well. In fact, South Americans in all three labor market contexts have the largest proportion of good English speakers. In the Latino enclave, 68 percent of Cubans reported speaking at least well, as did 66 percent of Nicaraguans. The smallest proportions come from Central America and Mexico, with 63 and 56 percent reported speaking English well or very well, respectively. Like education, English speaking ability is considerably different in the Asian enclave compared to the Latino enclaves. In every immigrant group, the proportion speaking English well or very well was larger in the Asian enclave than in the Latino enclaves. This is not surprising, since the need to speak English should be greater for Latin Americans in the Asian and non-enclave immigrant labor market. South Americans fluent English speakers (92 percent) are followed by Cubans and Colombians. Also like the Latino enclaves, Mexicans and Central Americans had the smallest proportion of good English speakers in their groups at 58 and 74 percent, respectively, with Nicaraguans falling in the middle range with 82 percent speaking English well. In the non-enclave areas, the trends in English fluency are similar, but the proportion of South Americans speaking English well or very well is lower than in the other two areas. At the same time, Mexicans, who show the smallest proportion of good English speakers in each of the areas, tend to be more fluent in English in non-enclave areas than in the enclave areas. A smaller proportion of Central Americans report speaking English well in the non-enclave area than in the Asian enclave, but the percent is higher than in the Latino enclaves. This same pattern holds true for Cubans, as well. For most of the groups, therefore, the percent of persons speaking English well or very well is higher in the nonenclave areas than in the Latino enclaves, but highest in the Asian enclaves. Finally, there is little variance in the household composition variables either between groups or between enclave contexts. Nearly all the groups show proportions married and living with their spouse in the 60 percent range, while household headship tended to fall in the 70 percent range.
Copyright 2001 by Stephanie Bohon
Discrimination and Competition among Long-Term Immigrants
91
Taken together, these preliminary findings call into question recent studies that suggest that declining immigrant endowments, particularly human capital, have been instrumental in limiting the economic success of Latin American immigrants (Borjas, 1985; 1991; Hansen & Cardenas, 1988). These studies suggest that groups that are high on endowments will have the highest levels of economic success. However, some groups (e.g., Nicaraguans) that are high on many of the control measures demonstrate rather modest occupational attainment. These findings also lend support to the contention that labor market contexts may have much do with the economic attainment of immigrant workers (Kossoudji, 1989; Lindstrom & Massey, 1994; Nee, Sanders & Seranau, 1994; Orcutt-Duleep & Regets, 1996; Rich, 1995; Roos & Hennessy, 1987). However, in order to test this contention, it is necessary to conduct an empirical analysis that examines the impact of labor market context on occupational attainment while holding other variables constant.
Regression Analysis In order to test the alternative hypotheses H3a and H3b the samples for the three labor market areas were combined and an OLS regression analysis was conducted examining the impact of labor market context on occupational attainment. The model is similar to those used in Chapter Five with a few notable exceptions. First, the reference group for the place of origin models was changed from the dominant enclave group to Mexicans. Mexicans provide a better reference group in these models because they appear in large numbers across each of the areas and their average MSEI scores are remarkably similar across the labor markets, as seen in tables 22 through 24. The other adaptation to the model was the addition of a new variable, place of work. The place of work variable indicates whether the respondent works in a Latino enclave, Asian enclave or non-enclave PMSA. Working in a Latino enclave was selected as the reference group, since this construction allows for a direct test of the hypotheses under question. The results the OLS analysis are in Table 25 that shows that place of work is a significant predictor of occupational attainment for long term Latin American immigrants. Model 1 shows just the impact of place of work on MSEI. Working in an Latino enclave results in two additional points in occupational attainment, on average, than working in another labor market. Labor market context alone, however, accounts for less than one percent of the total variance in MSEI, indicating the need to control for other factors. Model 2 shows the effect of place of work on occupational attainment controlling for endowments. Including the other variables markedly improves Copyright 2001 by Stephanie Bohon
92
Latinos in Ethnic Enclaves
the predictive power of the model, with the coefficient of determination increasing from .O1 to .36. This indicates that country of origin, year of entry, self-employment, human capital, and household characteristics, along with place of work, account for more than one-third of the total variance in MSEI in this sample. Adding the control variables substantially decreases the impact of place of work on the model, but the coefficients remain significant and negative. The effect of working in an Asian enclave on occupational attainment is reduced by more than one point, and the reduction is nearly two points for the workers in the non-enclave areas. Note the cross-over that occurs when the other variables are added to the model. In Model 1 it appears that the highest levels of occupational attainment occur in the Latino enclave areas followed by Asian enclave areas followed by non-enclave areas. Controlling for endowments shows that occupational attainment is the highest in Latino enclave areas and the lowest in the Asian enclaves, with non-enclave areas falling somewhere between. Nearly all of the control variables in are significant in Model 2 at p<.001 level. This is not surprising, given the large sample size (n=40830). There is no significant difference, however, between speaking English poorly and speaking no English at all, consistent with the findings in the within area models. While the effect of a Colombian or other South American place of origin increases occupational attainment by more than 2.3 points relative to Mexican origin, the impact of place of origin for Nicaraguans and Central Americans is much smaller (1.33 and .80, respectively). Cubans have a considerable advantage in MSEI relative to Mexicans, with occupational attainment levels nearly four points greater than the reference group, all other endowments being equal. The findings presented here support Hypothesis H3a. Long term immigrants do achieve the greatest occupational attainment in Latino enclaves, controlling for place of origin and other factors. On the other hand, immigrants in nonenclave areas fare better than immigrants in Asian enclaves, once endowments are controlled for. This suggests that Latin American immigrants may face greater discrimination in areas where other (non-Latino) immigrant groups control the marketplace. These findings have implications for the enclave debate, and these implications will be discussed more fully below.
Copyright 2001 by Stephanie Bohon
Discrimination and Competition among Long-Term Irnmigrants
93
Table 25. Parameter Estimates and Standard Errors of the OLS Model Regressing Occupational Attainment (MSEI) on Place of Work (Long-Term Immigrants)
Place of work Latino enclave Asian enclave Non-enclave Country of origin Cuba Nicaragua Central America Mexico Colombia South America Year of entry 1980 to 1984 I 970 to 1979 1960 to 1969 1959 or earlier Self-employed Educational attainment Less than high school Some high school High school graduate Some college College graduate Age ~xperience~ Ability to speak English Not at all Not well Well Very well Disabled Married, spouse present Head of household Adjusted R2 n
Model 1 Model 2 Regression Regression Coefficient Standard Error Coefficient Standard Error REFERENCE REFERENCE -2.02*** 0.16 -0.60""" 0.13 -2.39""" 0.16 -0.46""" 0.14
Copyright 2001 by Stephanie Bohon
3.94""" 0.15 1.33""" 0.38 0.80""" 0.16 REFERENCE 2.37""" 0.25 2.41""" 0.19 -2.91 *** 0.15 -1.63""" 0.13 REFERENCE 0.72""" 0.22 2.84""" 0.15 -1.04""" 0.17 -1.30""" 0.16 REFERENCE 3.63""" 0.19 12.63""" 0.19 0.17""" 0.02 O.OO*** 0.00
.0075 40830
REFERENCE 0.17 0.18 1.41""" 0.19 3.69""" 0.20 -1.52""" 0.29 0.61""" 0.12 1.36""" 0.13 3568 40830
94
Latinos in Ethnic Enclaves
Decomposition Analysis While OLS regression allows for hypothesis testing regarding occupational attainment outcomes, it does not directly test discrimination, which has implications for segmented assimilation and queuing theory. Since the extent to which inter-ethnic competition and cooperation occur across labor markets is a major theme of this work, a question remains regarding the magnitude of discrimination faced by different Latin American immigrants in the three labor market contexts. Answers to this question can be explored using regression decomposition analysis. As mentioned in Chapter Four, regression decomposition allows for the separation of average differences in occupational attainment between two groups to be separated into three parts: the amount of average MSEI differences due to discrimination, the amount due to endowments, and the amount due to the interaction between discrimination and endowments. Regression decomposition has been used by other researchers to compare economic outcomes for groups across labor markets (McLaughlin & Perman, 1991; Model & Lapido, 1996; Semyonov, 1988). Recently, Model and Lapido (1996), for example, used regression decomposition to examine differences in queuing among immigrants in New York City and London. In this work, they noted the need for similar analyses that take into account the effects of ethnic labor markets. This research fills that need. The basic decomposition model recommended by Jones and Kelly (1984) takes the form
The use of this three-part decomposition allows for a comparison of the discrimination effects to be examined by place of origin and by place of work. An issue, however, in using decomposition analysis is the selection of the reference group (Jones & Kelly, 1984). Queuing theory suggests that the reference should be that group which controls the economic resources in the labor market. For the Latino enclaves, that group is Cubans. Segmented assimilation suggests that the reference group should be the most prominent group in the area. Despite the fact that Mexicans are the most sizeable group in all three contexts, Cubans are the reasonable reference group in the Latino enclave areas, since they remain the most economically prominent group. The picture is less clear for the other two contexts. Decomposition analysis necessitates that the same reference group should be used across areas. While Cubans are the reasonable reference group in the Latino enclave, they are not Copyright 2001 by Stephanie Bohon
Discrimination and Competition among Long-Term Immigrants
95
elsewhere. There is no reason to expect that immigrants in Asian enclaves or non-enclave areas will become like Cubans over time, as segmented assimilation suggests, and, since Cubans do not control the economies in Asian and nonenclave labor markets, they are a poor reference group choice under queuing theory as well. Nicaraguans and Colombians cannot be used because the former group has an insufficient sample size in the non-enclave areas and the latter group has a marginal sample size in the Asian enclaves. Central and South Americans also offer a poor choice because their internal composition shifts from place to place. More specifically, the countries of origin that comprise the Central American group are different in the Latino enclaves than in the Asian enclaves. The same is true for the South Americans. By default this leaves Mexicans as the decomposition reference Table 26 shows the decomposition of the average differences in MSEI between Mexican and other Latin American immigrants in the three labor market contexts. The regression models used in the decomposition can be found in Appendix B. The coefficients reflect actual portions of the average difference in MSEI between Mexicans and other other groups listed in the table. Negative discrimination and endowment terms and the endowment terms reflect an advantage to the other groups. The coefficients of interest are the discrimination terms, which reflect the magnitude and extent of queuing within and between contexts (Model & Ladipo, 1996). The use of Mexicans as a control group is particularly useful for this analysis, since Mexican average MSEI is fairly constant across areas (see Tables 22 through 24), as are the associated standard deviations. While some of the queuing effect is captured in the interaction term, attention is focused on the discrimination coefficient, since methods to separate the interactions effects have questionable validity (Model & Lapido, 1996). The total column shows the total difference in average MSEI scores between Mexicans and the other groups (yM-Y'). In the Latino enclave, for example, Cubans have occupational attainment 8.38 points higher, on average, than Mexicans in the same area. Of this average difference, 3.9 points are attributable to discrimination ([(aM + ZbMxl)-(a1+2b1x')1), and 5.22 points are attributable to differences in human capital, household characteristics, and other endowments (2b1(xMx1)),while -.74 points are attributable to the occupational attainment returns inherent in being a Mexican immigrant (2(bM-bl)(p- x ' ) ) . ~ ~
Copyright 2001 by Stephanie Bohon
Table 26. Decomposition of Average Differences in MSEI between Long-Term Mexican and Other Latin American Immigrants by Place of Work Discrimination 6.19 - 14.65 - 12.23
Copyright 2001 by Stephanie Bohon
Asian Enclave Areas EndowInterments action -6.82 0.3 1 -4.5 1 1.82 -2.76 0.71
Total
Discrimination and Competition among Long-Term Immigrants
97
The results of Table 26 show, first, that the discrimination components are smaller in size and narrower in range in the Latino enclave areas, discrimination components are the largest in the Asian enclaves, and the non-enclave area discrimination components fall somewhere in between. This suggests that the least discrimination is occurring in the Latino enclaves and the most is occurring in the Asian enclave areas, which is consistent with the finding from the regression analysis. Second, the findings suggest that discrimination is occurring in all three labor markets. The smallest difference between the most advantaged and least advantaged group (relative to Mexicans) is 2.79 in the Latino enclaves. Again, this suggests that many immigrant origin groups continue to be faced with interethnic competition rather than inter-ethnic cooperation, as suggested in Chapter Five. Using the Mexican average MSEI scores as a baseline, comparisons can be made across groups, as well. Ignoring the discrimination coefficient for Cubans in the Asian enclave for the reasons stated above, it can be seen that a substantial portion of the difference between the average occupational attainment of each group and Mexicans is attributable to discrimination. With one exception, these figures consistently show that the difference between Mexicans and each group is smallest in the Latino enclave and largest in the Asian enclave. The exception is that Central Americans in non-enclave areas actually lose 3.27 MSEI points, on average, relative to Mexicans, despite the fact that they have higher levels of endowments. From both the decomposition and the regression analysis presented in Table 25, then, it can be concluded that long-term Latin American immigrants achieve higher occupational attainment in the Latino enclaves and the least occupational attainment in the Asian enclaves. Furthermore, immigrants in the Latin American enclaves experience the least discrimination in the Latino enclave and the most discrimination in the Asian enclaves.24
CONCLUSIONS The findings from this analysis are striking and have significant implications for enclave theory. First, the findings presented allow for a rejection of the hypotheses that immigrants will have similar occupational attainment regardless of labor market context. While most of the enclave literature has focused on other economic outcomes (e.g., income and self-employment), the results presented here support the contention by Portes and his colleagues (Portes & Bach, 1985; Portes & Jensen, 1989; Portes & Stepick, 1993) that Latin American immigrants receive greater returns to human capital in those metropolitan areas with Latino enclaves. These findings may also call into Copyright 2001 by Stephanie Bohon
98
Latinos in Ethnic Enclaves
question the contention of Sanders and Nee that immigrants have reasonable access to opportunities in the broader labor market, and that these opportunities are similar to those found by native workers (Nee, Sanders, & Sernau, 1994; Sanders & Nee, 1987; 1992). While this cannot be tested directly with the data here, it is reasonable to assume that if this were true, there would not be significant negative effects on occupational attainment for immigrants in nonenclave places. The findings from the regression analysis also suggest that the metropolitan areas with Asian enclaves offers the poorest occupational opportunities for Latin American immigrants. One possible explanation for this finding is that racial queuing is particularly prevalent in the Asian enclave, and that good jobs are reserved for members of the dominant enclave groups (usually Chinese or Japanese). This finding may also suggest that the Asian enclave creates language and other cultural barriers that are more difficult to overcome than those found in the non-enclave context. It is important to note, however, that while the hypotheses regarding discrimination across labor market contexts are based on theoretical expectations derived from queuing theory and segmented assimilation theory, the analysis undertaken in this chapter, unlike the preceding chapter, does not offer a clear means of rejecting one theory in favor of the other. That is, while more or less discrimination may be evidence of a great propensity toward segmented assimilation into an amorphous Latino group, it does not suggest that queuing is not a factor. Differences in discrimination across labor market contexts may merely be evidence of a larger pool of good jobs for Latin American immigrants, in general, rather than the absence of queuing per se. Nonetheless, the findings from the decomposition analysis are interesting with respect to enclave theory. The findings from the decomposition analysis rejects the hypothesis that differences in average occupational attainment due to discrimination will be similar across contexts. Decomposing average MSEI scores shows that the least discrimination occurs to longer-term immigrants in Latin American enclave areas. While there remain considerable unexplained differences in average MSEI scores between groups, the smaller magnitude of these differences due to discrimination in the Latino enclave areas suggest that the presence of a large and economically powerful Latin American group attenuates, rather than magnifies, differences between Latino groups, at least for immigrants who have been in the U S . for more than five years. Labor markets with Latino enclaves, therefore, encourage more inter-ethnic cooperation than would normally be found in the labor market.
Copyright 2001 by Stephanie Bohon
Discrimination and Competition among Long-Term Immigrants
99
Consequently, while Mexicans, Central Americans and Nicaraguans have considerable difficulty gaining occupational returns to their endowments commensurate with that of other Latin American immigrants in all three labor market contexts, working in an Latino enclave city does improve the attainment possibilities for long term immigrants relative to other location choices.
Copyright 2001 by Stephanie Bohon
CHAPTER SEVEN
Discrimination and Competition among Recent Immigrants
The results of the previous chapter suggest that enclave and non-enclave places of work effect occupational attainment, and that national origin partially determines both the amount of discrimination faced in the labor market and the expected returns to human capital for immigrants who have spent more than five years in the labor market. However, as immigration has increased precipitously in the last decade and immigrant flows from Latin America have become sizeable, scholars have recognized the importance of examining short term immigrant economic adaptation outcomes as well as longer term outcomes. This chapter explores occupational attainment among recent Latin American immigrants in Latino enclaves, Asian enclaves, and non-enclave places. The introductory chapter notes that segmented assimilation and queuing theory do not lend themselves easily to clear hypotheses regarding the economic outcome of recent immigrants. Segmented assimilation is based on processes that occur over time and across generations while queuing theory generally describes selective allocation in the types of jobs for which most recent immigrants are not qualified. Nonetheless, there is no reason to think that certain types of economic contexts offer better opportunities for recent immigrants than others. Latino enclaves, for example, may provide an environment where new immigrants can capitalize on education and skills brought with them from their home countries. On the other hand, business owners in these same types of enclaves may be more sensitive to country of origin differences which may lead to greater discrimination. Given these alternative outcomes, competing hypotheses regarding the occupational attainment of recent immigrants were formulated based on the competing ideas inherent in segmented assimilation and queuing theory.
Copyright 2001 by Stephanie Bohon
102
Latinos in Ethnic Enclaves Question: Will recent Latino immigrants (residing in the U.S. for five years or less) working in metropolitan areas with Latino enclaves demonstrate higher or the same occupational attainment as Latino immigrant working in Asian enclaves and non-enclave areas? Hypothesis H4a. Comparing across economic enclave and non-enclave contexts, recent Latin American immigrants in metropolitan areas with Latino enclaves will have higher occupational attainment than recent Latin American immigrants in areas with Asian enclaves or non-enclave places. Alternative Hypothesis H4b. Comparing across economic enclave and nonenclave contexts, recent Latin American immigrants in metropolitan areas with Latino enclaves will have occupational attainment not significantly different from recent Latin American immigrants in areas with Asian enclaves or nonenclave places, holding constant measurable endowment characteristics. Question: Will recent Latino immigrants (i.e., those living in the United States for five years or less) working in metropolitan areas with Latino enclaves experience less or the same amount of discrimination in job hiring and promotion as recent Latino immigrant working in Asian enclave or non-enclave areas? Hypothesis H6a. Comparing across economic enclaves and non-enclave contexts, discrimination as a determinant of occupational attainment differences among recent Latin American immigrants is lower in metropolitan areas with Latino economic enclaves and greater in areas with Asian enclaves and nonenclave labor markets. Alternative Hypothesis H6b. Comparing across economic enclaves and nonenclave contexts, discrimination as a determinant of occupational attainment differences among recent Latin American immigrants is no different in metropolitan areas with Latino economic enclaves than in areas with Asian enclavse and non-enclave labor markets.
Hypotheses H4a and H6a suggest that there are fewer barriers to occupational attainment for immigrants within the metropolitan areas with Latino enclaves. This is consistent with enclave theory. Hypotheses H4b and H6b suggest that there is no difference in opportunities for recent immigrants across the three contexts or that the queuing processes in the Latino enclave
Copyright 2001 by Stephanie Bohon
Discrimination and Competition among Recent Immigrants
103
cities offsets the gains that could be otherwise achieved by some immigrant groups. In order to test these hypotheses, descriptive statistics are presented using samples representing recent immigrants in each of the three economic contexts (metropolitan areas with Latino enclaves, Asian enclaves, and non-enclave places). Averages and proportions are compared across groups to explore the differences in average occupational attainment by place of origin by place of work. This analysis also allows for the exploration of differences in endowments (i.e, year of entry, self-employment, human capital, and household characteristics) across national origin and economic context for immigrants who have been in the U S . five years or less. Then, the results of a pooled OLS regression model are presented that combine all of the respondents in the recent immigrant sample using both place of work and economic context as variables of interest. By using work in an Latino enclave PMSA as the omitted category, Hypotheses H4a and H4b are tested in order to determine the effect of location choice on occupational attainment. Finally, the results of separate OLS regressions conducted for each adequately sized national origin group within each economic context are decomposed in order to test Hypotheses H6a and H6b.
RESULTS Descriptive Statistics Tables 27,28 and 29 show the descriptive statistics for occupational attainment, year of entry, self-employment, human capital, and household characteristics by country of origin for the recent immigrants working in the three labor market types. Results were generated if the country of origin group had 50 or more respondents in the sample. This selection criteria resulted in the elimination of Colombians and Cubans from this stage of the analysis for the Asian enclaves, and it also necessitated the removal of Cubans and Nicaraguans from the analysis of the non-enclave areas. An examination of Table 27 shows that average MSEI in the Latino enclaves range from 20.91 to 27.50 with South Americans demonstrating the highest average occupational attainment and Mexicans demonstrating the lowest. These trends are consistent with those found in the Asian enclaves where Mexican attainment is 19.72 and South American attainment is 27.66. Average MSEI is nearly the same for Central Americans in both contexts, but Nicaraguans working in Asian enclaves have an average occupational attainment score more than four points lower than in the Latino enclave areas. Average Nicaraguan attainment in both contexts, however, is higher than both the
Copyright 2001 by Stephanie Bohon
104
Latinos in Ethnic Enclaves
Mexican and the Central American averages and lower than the South American scores. Mexican average occupational attainment in the non-enclave areas is, again, the lowest of all the groups and remarkably consistent across the three economic contexts at about 20. These findings are consistent with the findings from the long-term sample in the previous chapter where average Mexican MSEI was also nearly invariable across labor markets. These preliminary findings seem to indicate that occupational attainment for Mexicans is less determined by location choice than by other factors such as human capital or that these immigrants' choice of work areas is not highly selective. Finally, the average MSEI scores for the recent immigrants are markedly lower than the averages for the long term immigrants in the same areas. This is not surprising, since occupational attainment should increase with more time in the labor market. Furthermore, longer periods in the U.S. should increase the probability that immigrants are able to gain endowments that are valued in the U.S. labor market. In all three contexts immigrants were more likely to have arrived in the U.S. between 1987 and 1990 than between 1985 and 1986. The notable exception are Colombians. Fifty-two percent of Colombian immigrants in both the nonenclave and the Latino enclaves sample arrived in the U S . prior to 1987. The preponderance of new arrivals, however, is consistent with the steady increase in migration streams from Latin America to the United States throughout the 1980s. Tables 27,28, and 29 also show the proportion self-employed by country of origin. It seems logical to expect, a priori, that few recent Latin American immigrants would be self-employed, given the generally impoverished economies of the sending countries and the difficulties in acquiring capital without long-term credit histories. The results of these tables show, however, that nearly ten percent of all recent South American immigrants are selfemployed and almost 11 percent of the Cubans working in Latino enclaves are, as well. Furthermore, nearly 12 percent of recent Colombian immigrants operate their own businesses in Latino enclave areas.
Copyright 2001 by Stephanie Bohon
Discrimination and Competition among Recent Immigrants
105
Table 27. Descriptive Statistics by Place of Origin for Recent Immigrants Working in Latino Enclave Areas.
MSEI
Central South Cuba Nicaragua America Mexico Colombia America 26.85 24.48 21.82 20.91 26.60 27.50
Year of entry 1987 to 1990 1985 to 1 986 Self-employed Educational Attainment Less than high school Some high school High school graduate Some college College graduate Age
Ability to speak English Not at all Not well Well Very well Disabled Married Head of household .599 n 466 (standard deviations are in parentheses)
Copyright 2001 by Stephanie Bohon
S19 607
,302 1908
.241 5397
,467 428
,498 841
Latinos in Ethnic Enclaves
106
Table 28. Descriptive Statistics by Place of Origin for Recent Immigrants Working in Asian Enclave Areas. Nicaragua
Central America
Mexico
South America
.23 1 320
,220 2735
,534 88
MSEI Year of entry 1987 to 1990 1985 to I 986 Self-employed Educational Attainment Less than high school Some high school High school graduate Some college College graduate Age
Ability to speak English Not at all Not well Well Very well Disabled Married Head of household n 72 (standard deviations are in parentheses)
Copyright 2001 by Stephanie Bohon
107
Discrimination and Competition among Recent Immigrants
Table 29. Descriptive Statistics by Place of Origin for Recent Immigrants Working in Non-Enclave Areas.
MSEI Year of entry 1987 to 1990 1 985 to 1986 Self-employed Educational Attainment Less than high school Some high school High school graduate Some college College graduate Age
Central America 22.35 (9.33)
Mexico 20.09 (8.84)
Colombia 24.20 (13.15)
South America 23.69 (10.21)
.539 .461 ,036
.559 .44 1 .037
.479 ,521 .055
.516 ,484 .098
.333 .267 .236 .097 .067 28.65 (8.86) 463.36 (494.15)
,458 ,248 .I83 .058 .053 27.07 (8.46) 419.54 (5 18.35)
.I64 .288 315 .I37 .096 30.92 (9.66) 504.64 (2.53)
.230 .I80 ,369 .057 ,164 32.01 (8.76) 552.30 (505.10)
,285 .4l2 .I60 ,143 ,016 ,245 ,300 1770
.I92 397 .288 .I23 .027 ,342 .479 73
.I56 .426 .254 .I64 .025 352 377 122
Ability to speak English .291 Not at all Not well .358 Well ,176 ,176 Very well Disabled ,018 Married .303 Head of household ,376 n 165 (standard deviations are in parentheses)
Not surprisingly, the measures of human capital (educational attainment, age, potential work experience, English fluency, and disabilities) are lower in the recent immigrant sample than in the long-term models. Furthermore, Mexicans and Central Americans appear to be the most disadvantaged across the three labor markets on these variables. Some similarities with the longer-term immigrants sample should be noted. First, Nicaraguans have higher educational attainment than Cubans in the Latino enclaves, and their education levels are only exceeded by those of the Colombians and the South Americans working in the same area. Another point is that the Latino enclave areas have the largest proportions of immigrants who speak no English at all. Again, this is consistent with enclave theory which
Copyright 2001 by Stephanie Bohon
108
Latinos in Ethnic Enclaves
suggests that Latino enclaves offer work in a labor market where ability to speak English is unimportant. Finally, an examination of the household characteristics of the different groups shows that recent immigrants are less likely to be married (15 to 58 percent versus 59 to 75 percent married, respectively) and less likely to be heads of households (22 to 60 percent versus 66 to 82 percent, respectively) than the longer term immigrants in this study. This is consistent with their lower average ages. The disadvantaged groups (i.e., Mexicans and Central Americans) have the highest proportions of men married and living with their spouses and the highest proportions of household headship in the non-enclave areas and the lowest levels in the Asian enclaves. By comparison, the advantaged groups (i.e., Cubans, Colombians, and South Americans) have the highest proportions in the Latino enclaves compared to the other economic settings. With the exceptions of Cubans in the Latino enclaves and South Americans in the Asian enclaves, however, none of the groups show married or household headship figures greater than 50 percent. An examination of these preliminary findings shows that low levels of endowments (e.g., recent years of entry, young ages, low levels of educational attainment, and low levels of family commitments) are consistent with the lower levels of MSEI for Central American and Mexican immigrants in this sample. At the same time, those groups with the highest average MSEI scores (i.e., Cubans, Colombians, and South Americans) have the highest endowments. The story is less clear for the Nicaraguans who display the highest average occupational attainment in the Latino enclaves where they also report the lowest levels of endowments. While the missing groups make it difficult to gain a complete picture, these results indicate that the hypothesis that immigrants will receive the highest occupational attainment in the Latino enclaves is supported by the reported outcomes for Nicaraguans and Colombians. The hypothesis that there is no difference is supported by the outcomes for Central Americans and Mexicans. Finally, the results for the Cubans and South Americans offer no clear insight into the alternative hypotheses posed in H4a and H4b. Regression Analysis A clearer picture of the impact of economic context and endowments is given in Table 30, which provides the results of the pooled OLS regression analysis. Model 1 shows the effects of economic context. Although place of work accounts for only one percent of the total variance in MSEI in the model, an examination of the regression coefficients shows that Latin American immigrants have significantly lower occupational attainment in Asian enclaves Copyright 2001 by Stephanie Bohon
Discrimination and Competition among Recent Immigrants
109
and non-enclave areas than those working in Latino enclaves. Immigrants working in Latino enclaves have occupational attainment that is 1.77 points higher, on average, than those working in Asian enclaves, and 2.33 points higher than those working in non-enclave areas. These effects are highly significant (p<.OO 1). Model 2 shows the effects of the full model. While controlling for place of origin and endowments decreases the effect of economic context on MSEI, the results show that Latino enclave workers still fare better than other workers, net of other factors. This finding supports hypothesis H4a that recent Latin American immigrants will have higher occupational attainment in the Latino enclaves than in the Asian or non-enclave areas. Model 2 also shows the effects of the other factors on the pooled model. As in Chapter Six, Mexicans were selected as the control group because of their stable average MSEI across the different economic contexts and because that group is comprised of a single country of origin. The place of origin variables are all statistically significant and positive. This finding is consistent with the findings from the descriptive analyses in Tables 27, 28 and 29 and the findings in the previous chapters that suggest that recent Mexican immigrants have lower levels of occupational attainment in every context. Cubans, Colombians and South Americans fare well, even after controlling for labor market type, with occupational attainment about two and a half points higher than Mexicans, on average. Nicaraguans and Central Americans have a much smaller advantage over Mexicans with average MSEI scores only about one-half point greater, net of other factors. These findings suggest that queuing may still offer the most likely model of immigrant adaptation, at least in the first five years after immigration. Most endowment variables in Model 2 are significant and display the anticipated sign. The exceptions is disability status, which is not significant. Consistent with the findings in Chapter Five, self-employment continues to be a strong and highly significant predictor of occupational attainment for immigrants. Recent immigrant entrepreneurs have occupational attainment nearly four points higher, on average, than those who work for others. Given the fairly sizeable proportions of self-employed recent immigrants, this finding has significant implications for attainment studies. It suggests that loosening the barriers to capital access for new immigrants may significantly improve their economic well-being in the early adjustment period.
Copyright 2001 by Stephanie Bohon
Latinos in Ethnic Enclaves
110
Table 30. Impact of Place of Work and Other Factors on the Occupational Attainment (MSEI) of Recent Immigrants
Place of work Latino enclave Asian enclave Non-enclave Country of origin Cuba Nicaragua Central America Mexico Colombia South America Year of entry 1987 to 1990 1985 to 1986 Self-employed Educational attainment Less than high school Some high school High school graduate Some college College graduate
Model 1 Model 2 Standard Regression Standard Error Regression Coefficient Coefficient Error REFERENCE REFERENCE -1.77*** 0.23 -1.33*** 0.21 -2.33*** -0.19 -1.15*** 0.18
Age Experience2 Ability to speak English Not at all Not well Well Very well Disabled Married, spouse present Head of household Adjusted R~ N *p<.o5, **p<.o1, ***p<.oo1
Copyright 2001 by Stephanie Bohon
2.42""" 0.43 0.35 0.61* 0.56"" 0.20 REFERENCE 2.23""" 0.41 2.50*** 0.3 1 -0.51*** 0.15 REFERENCE 3.96*** 0.33 0.24 -0.51 * -0.64** 0.23 REFERENCE 1.60*** 0.33 6.97*** 0.34 0.22*** 0.02 -o.oo*** 0.00 REFERENCE 0.17 0.18 1.81*** 0.24 2.21""" 0.23 -0.27 0.59 1.08*** 0.21
,0114 15046
,1669 15046
Discrimination and Competition among Recent Immigrants
111
The educational effects shown in Table 30 are also significant across the categories. Relative to immigrants with a high school diploma, recent immigrants who did not graduate from high school have lower occupational attainment losing at least one-half point of MSEI for their educational deficiency. At the same time, immigrants with some college education display 1.6 points greater MSEI, on average, than high school graduates, and recent immigrants with a college degree have nearly a seven point advantage. The two remaining human capital measures, age and English fluency, are also highly significant predictors of MSEI. A one year increase in age, for recent immigrants, translates into nearly a quarter point advantage on the MSEI scale, on average, while immigrants who speak English well or very well have nearly two points greater occupational attainment than those who do not speak English. The latter finding is significant even after controlling for economic context. Finally, the household characteristic variables are also both significant with married immigrants reporting a more than one point greater occupational attainment score, on average than those not living with a spouse. Householders also demonstrate higher levels of occupational attainment than other members of a household. In summary, economic context, place of origin, and personal endowments account for about 17 percent of the total variance in occupational attainment for recent immigrants in this study. Year of entry, self-employment, human capital and household characteristics are all important predictors of occupational attainment. However, place of origin is also significant, which suggests that queuing is a reality for recent arrivals, as well as longer term immigrants (as shown in the previous chapter). Cubans, Colombians, and South Americans maintain a dominant position in the Latin American labor market that cannot be fully accounted for by their higher than average endowments. At the same time, labor market choice also has an impact on the occupational attainment of recent immigrants, as Latin American immigrants in Latino enclaves are able to attain occupational levels significantly higher than their counterparts in Asian enclaves and non-enclave areas. This suggests that, despite queuing, there are better occupational opportunities for recent immigrants working in labor markets with a Latino enclave (i.e., Miami, Jersey City, New York, and Los Angeles) than in the other areas studied.
The Decomposition Model While the pooled OLS regression model allows for the testing of the competing hypotheses concerning the context of recent immigrant occupational attainment, it does not indicate the amount of discrimination occurring between groups and across labor market contexts. As suggested in the previous section, while Copyright 2001 by Stephanie Bohon
112
Latinos in Ethnic Enclaves
immigrants fare significantly better in the Latino enclave areas, the continuing significance of place of origin suggests that queuing still occurs in the three contexts. However, if the discrimination effects are less within the Latino enclaves than within the other two contexts, such evidence would support those aspects of enclave theory that suggest that most Latin American immigrants will benefit from the presence of Latino enclaves. T o explore this possibility, the results of separate regression models by place of origin and labor market contexts were decomposed using Mexicans as the control group. Again, Mexicans offer a good group for comparison because recent immigrants are found in sizeable numbers in all three labor market area contexts, and their average attainment levels are stable across these contexts. The results of the decomposition analysis are shown in Table 31 .25 The original OLS regression models used in the decomposition analysis are shown in Appendix B. As noted in Chapter Six, when interpreting decomposition coefficients, it is good to remember that negative signs indicate favorable outcomes relative to the reference group. Since Mexicans provide the comparison group, it should be expected that the signs for both the discrimination and the endowment terms will be negative, if the findings here are consistent with the analysis of the longerterm immigrants. Also like in Chapter Six, the discrimination term refers to the portion of the difference in average MSEI between Mexicans and other groups that is not attributable to either endowment differences (e.g., differences in income, education, etc.) or the interaction between endowments and discrimination. In other words, a negative discrimination term is a measure of the amount of difference between a group's average MSEI and Mexican average MSEI that is attributable to membership in the group rather than some measurable, marketable characteristics that the group might possess. Obviously, some of discrimination effect is captured in the interaction coefficient, but it is impossible to decompose the interaction term with reasonable accuracy (Jones & Kelly, 1984).As a consequence, only the discrimination and endowment terms will be discussed here. Looking at the coefficients for the Latino enclave areas presented in Table 31, it can be seen that the greatest differences in endowments and in discrimination occur between Cubans, Colombian, and other South Americans versus Mexican immigrants. Almost three points of the average difference between these three groups and the Mexican control group is due to factors in the marketplace that are more favorable to those groups, net of the endowments that they bring to that marketplace.
Copyright 2001 by Stephanie Bohon
Table 31. Decomposition of Mean Differences in MSEI between Recent Mexican and Other Latin American Immigrants by Place of Work
Asian Enclave Areas Discrimination N/A 1.53 -1.11
Copyright 2001 by Stephanie Bohon
Endowments N/A -0.34 -0.83
Interaction N/A -1.47 -0.63
Total N/A -0.28 -2.57
114
Latinos in Ethnic Enclaves
As with the longer-term immigrants, the most queuing (as determined by the discrimination coefficient) occurs in the Asian enclaves where South Americans gain four and a half points of MSEI, on average, over Mexicans due to factors unrelated to the control variables. Nicaraguans, on the other hand, have a one and a half point average MSEI deficit due to discrimination, relative to Mexicans. One possible explanation for the Mexican-Nicaraguan differences in the Asian enclaves may be related to data issues. As previously mentioned, decomposition analysis assumes that returns to endowments should be the same for all groups. Given that Nicaragua in the 1980s had the highest rates of literacy and education in Latin America, what appears to be a disadvantage for Nicaraguans due to queuing may reflect differences in returns to Latin American versus U S . acquired education. In the Latino enclave, negative attitudes toward education received outside of the U.S. should be negligible. This may explain why the discrimination coefficient for Nicaraguans in the Latino enclaves are more reasonable. Unlike the longer term immigrant findings, the least discrimination among recent immigrants is found in the non-enclave, rather than the Latino enclave areas. In non-enclave places, the average MSEI advantage of Colombians and South Americans over Mexicans is much less than that found in the Latino enclaves, net of endowments and interactions. The most striking finding is that, within the Latino enclave, those groups with the highest levels of endowments (Cubans, Colombians, and South Americans), also have the biggest discrimination advantage. In the non-enclave areas, however, a large proportion of the average MSEI difference between Mexicans and Central Americans cannot be explained by education, time of entry, English fluency, or other factors.
CONCLUSIONS The findings of the regression and decomposition analysis call into question the hypotheses that there is no difference in occupational attainment among recent immigrants in different labor market contexts and that discrimination levels will be similar in enclave and non-enclave areas. Labor market context has a strong, significant association with occupational attainment for immigrants who have been in the U S . for five years or less. Recently arriving immigrants will achieve the most occupational attainment in the PMSAs with Latino enclaves and the least occupational attainment in the Asian enclave areas, all else being equal. This finding lends credence to the contention of enclave theorists that Latino enclaves create opportunities for recent immigrants not found in other U.S. labor markets (Portes & Stepick, Copyright 2001 by Stephanie Bohon
Discrimination and Competition among Recent Immigrants
115
1993).This argument is further supported by the finding that the impact of most of the endowment factors, while significant, are less sizeable for recent versus longer term immigrants. Consequently, the enclave offers labor market choices not available elsewhere to those who are able to take advantage of them. Latin American immigrants who choose Asian enclave or non-enclave locations as their early destination choice limit their opportunities to acquire good jobs in the first five years after migration. Although the metropolitan areas with Latino enclaves offers the most opportunities for occupational attainment, Latin American immigrants are subject to more discrimination there than in non-enclave places, however. More specifically, while immigrants achieve better occupational attainment, overall, in the Latino enclave cities, they have less chance of receiving opportunities for attainment comparable to other Latin American immigrant groups in the U S . For groups like Cubans, Colombians, and other South Americans, the occupational advantage gained from choosing to work in an Latino enclave locale is further bolstered by a disproportionate advantage in attainment relative to other immigrants there. The most significant fact that can be concluded from this chapter, then, is that recent Latin American immigrants to the U S . benefit from working in an Latino enclave area, but that they are less likely to experience discrimination in non-enclave labor markets.
Copyright 2001 by Stephanie Bohon
CHAPTER EIGHT
Conclusions
Latin American immigrants began coming to the United States in great numbers in the mid- 1950s. Today, immigrants from Latin America comprise the largest percentage of the United States' foreign born population. Of the more than 21 million known immigrants living in the United States today, 47 percent are Latino and 43 percent are from Latin America (Portes & Zhou, 1993; Waters & Eschbach, 1995). However, there is considerable variation between migrants from the many Latin American countries, and these variations determine both the composition and destination of the immigration streams. Both factors are important, because they effect the economic adaptation of these immigrants, as the results presented here have shown. This study examines the impact of ethnic enclaves on Latin American immigrant occupational attainment and evaluates segmented assimilation and queuing theory as alternative explanations for the immigrant adaptation experience. This study helps to bridge gaps in the ethnic enclave literature by exploring, for the first time, the impact of national origin on immigrant outcomes vis-a-vis the dominant enclave group. This study also expands the literature on immigrant adaptation by focusing on occupational attainment rather than income or self-employment. Finally, this study extends the work on occupational attainment by comparing attainment across three distinct labor market types: the Latino enclave metropolitan areas, the Asian enclave metropolitan areas, and the non-enclave place. Toward this end, a number of general research questions were posed in the introductory chapter. These questions, and the results of the data analysis pertaining to them are presented below.
What is the importance of enclave context on the occupational attainment of Latin American immigrants from various sending countries?
Copyright 2001 by Stephanie Bohon
118
Latinos in Ethnic Enclaves
The selection of a Latino enclave, an Asian enclave, or a non-enclave area as an immigrant's destination has a statistically significant impact on occupational attainment. For longer-term immigrants, Cubans, Colombians and South Americans have the highest average MSEI scores in Asian enclave PMSAs (MSEI = 40.05, 37.16 and 37.05, respectively), Mexicans have the highest scores in Latino enclave PMSAs (MSEI = 25.01), and Nicaraguan and Central American MSEI scores are about the same regardless of context (MSEI = 30 and MSEI = 27, respectively). Controlling for year of entry, national origin, human capital, and household characteristics, however, reveals that workers who have been in the United States for more than five years can expect to achieve occupational attainment at least one-half point higher by choosing to work in Jersey City, Miami, New York or Los Angeles instead of a non-enclave place. Furthermore, workers in Latino enclave areas can expect occupational attainment levels six-tenths of a point higher, on average, than what they could achieve in Asian enclave areas. For recent immigrants, the relationship between place of origin and occupational attainment is even greater. Latin American immigrants working in metropolitan areas with Latino enclaves achieve occupational attainment one and a third points higher, on average, than workers in Asian enclaves areas. The impact of working in a non-enclave place is less than the Latino-Asian enclave differential, but it is still considerably greater for recent immigrants than for those who have been in the country for some time. These findings suggest not only that areas with a preponderance of Latin American immigrant-owned businesses offer the best occupational advantages for Latin American immigrants, but that Asian immigrant-dominated areas severely restrict opportunities for Latin American immigrant men to translate their endowments into high occupational attainment. At the same time, the possibility of a selection effect cannot be ignored. It is possible that areas with Latino enclaves offer occupational opportunities to immigrants which attract them to these places. These factors may have led to the creation of ethnic enclaves. Consequently, it is possible that it is not the enclave itself, but, rather, some other characteristics of places which attract Latino enclaves that draw Latin American immigrants and create opportunities for high occupational attainment, relative to other places.
Does the enclave promote economic adaptation for all immigrant groups or just for the dominant enclave group? Those cities with Latino enclaves appear to promote greater economic adaptation in the form of occupational attainment for all Latin American immigrants compared to the Asian and non-enclave places. Nonetheless, some Copyright 2001 by Stephanie Bohon
Conclusions
119
groups appear to benefit more from the Latino enclave context than others, and some Latino enclave cities offer greater possibilities than others. Cubans, for example, benefit from the presence of Latino enclaves more than most other Latin American immigrants, even when they are not the dominant enclave group, as is the case in the New York enclave. Overall, Cubans, Colombians, and other South Americans appear to do particularly well in Latino enclave places, achieving high average levels of occupational attainment, relative to other immigrants. One possible exception to this is Los Angeles, where South American immigrants appear to be somewhat disadvantaged. At the same time, Mexicans, Nicaraguans and Central Americans face national origin related obstacles to occupational attainment in most areas, although the disadvantages are lessened somewhat in the Latino enclave areas. While the Mexicans and Central Americans seem to face disadvantages in each of the labor market contexts, the finding that Nicaraguans are disadvantaged in metropolitan areas with Latino enclaves is surprising, given the speculation in the literature that this group would benefit greatly from their high levels of human capital and the sympathy that they received from the Cuban community in Miami as refugees from Communism (Portes & Stepick, 1993; Portes & Zhou, 1993). One possible explanation for this finding is offered by Stanley (1993), who contends that Nicaraguan refugees in the U S . are perceived by Cubans and Cuban-Americans as a greater economic threat than other Latin American immigrant groups, since they have high levels of human capital and widespread U.S. government support. Furthermore, they tend to settle in Miami. As a consequence, the Cuban community in Miami has lobbied extensively to block Nicaraguan refugee asylum efforts. This contention, however, is disputed by Portes and Stepick (1993) who argue that Miami's Cuban community was instrumental in acquiring resources for Nicaraguan resettlement in Florida. The seemingly disproportionate advantage in occupational attainment displayed by the South American immigrants in this study (including Colombians) can also be partially explained by historical factors. South American immigrants have a long history of migration to the United States and were well established in areas like Miami long before the Cubans and Central Americans arrived. In fact, it has been argued by Portes and others (Portes & Bach, 1985; Portes & Stepick, 1993) that the success of the Cubans is partially attributable to assistance by established South Americans. At the same time, the low levels of occupational attainment demonstrated by the Mexicans and Central Americans in this study may also be partially attributed to factors inherent in their migration processes. A great deal of Mexican migration is circular, clandestine, and motivated by economic desperation (Massey, 1989; Tienda & Wilson, 1992). Family links in the U.S.
Copyright 2001 by Stephanie Bohon
120
Latinos in Ethnic Enclaves
(i.e, migration networks), also play a large role in Mexico-U.S. migration (Massey, 1987; Portes & BoroCz, 1989). Mexicans, therefore, may have a smaller range of destination choices and may be more likely to view their migration as temporary (Kossoudji & Cobb-Clark, 1996). Hence, they may be less motivated or less able to achieve high levels of occupational attainment, irrespective of endowments. Central American immigrants are somewhat similar to Mexicans in that many, particularly Salvadorans, Guatemalans, and Hondurans, tend to be refugees (Jacobsen, 1996). As a consequence, they may also be likely to view their migration to the U S . as temporary. In fact, there is evidence that many Central Americans are returning to their homeland as the political situations in these countries have become more stable (Russell, 1995). Whatever the explanation, the overall finding is that groups with high levels of endowments (with the exception of Nicaraguans) receive greater returns to these endowments in Latino enclave PMSAs than do groups with lower levels of human capital, later years of entry, and lower household composition scores. At the same time, Latino enclave areas still offer greater returns in the form of occupational attainment to these latter Latin American immigrants than do other labor market contexts.
What impact does time since immigration have on the labor market status of non-coethnic workers, and does time since immigration affect labor market status differently for workers who share a country of origin with the dominant enclave group and those who do not? Time of entry has an impact on both the amount of occupational attainment an immigrant can expect to achieve and the amount of queuing that they face in the labor market, regardless of national origin. Using the 1960s as a starting point, occupational attainment increases with longer residence in the U.S. The less years of potential exposure to the U S . labor market, the lower the occupational attainment. Consequently, occupational attainment among recent Latin American immigrants is relatively low and attainment among longer-term immigrants is relatively high. Some of this effect may be due to changes in immigration laws over time that have altered the criteria under which immigrants were allowed entry into the U S . However, national origin effects on occupational attainment are greater for immigrants who have been in the U S . for more than five years than for recent immigrants. For most Latin American immigrants arriving in the U.S. after 1984, there is no difference in occupational attainment by place of origin. This is not true for longer-term workers. Furthermore, even when there are significant
Copyright 2001 by Stephanie Bohon
Conclusions
121
differences by national origin, these between-group differences are smaller for nearly all recent immigrants than for longer-term migrants. Looking at the impact of place of origin on occupational attainment, however, it can be seen that there are marked differences between groups. Compared to Mexicans, Cubans tend to benefit the most from more years in the labor market. That is, the impact of national origin on average occupational attainment is 3.94 points higher for longer term Cuban immigrants than Mexicans compared to 2.94 points higher for recent immigrants. The impact of being Nicaraguan or other Central American on occupational attainment is also greater for longer-term than recent immigrants. However, the impact of Colombian or other South American origin does not change much for the two groups. That is, these two groups can expect to gain about two and a third points on the MSEI scale, on average, compared to Mexicans due to their national origin alone, and this impact does not change much regardless of whether the immigrants are longer-term or recent. Does the national origin of the dominant enclave group affect the occupational attainment of non-coethnic workers? No, but the results are limited. Comparing New York--where the dominant enclave group is Colombian--to the other Latino enclave places where the dominant enclave group is Cuban, it appears that there is less difference between the dominant enclave group and other groups in New York. Furthermore, for places where the differences between the dominant and subordinate groups are significant, the magnitude of the impact is much smaller in New York. It cannot be concluded that the results have to do with characteristics of the dominant enclave group rather than particular factors that may be unique to New York City. The other problem is that even in the Latino enclave where Cubans are not the dominant enclave group, Cubans still reported significantly higher occupational attainment. This may suggest that Cubans, for whatever reason, receive greater returns to their endowments than other Latin American immigrants regardless of enclave conditions. Does the national origin of the non-coethnic group affect the economic adaptation of workers inside the enclave regardless of their human capital, household characteristics, and other endowments? Yes, national origin usually impacts occupational attainment, however, the findings of this study suggest that human capital, as well as year of entry, selfemployment, and household characteristics are more important predictors of occupational attainment for Latin American immigrants who have been in the
Copyright 2001 by Stephanie Bohon
122
Latinos in Ethnic Enclaves
U.S. for more than five years. These endowments are less important for recent immigrants, but they still explain a good deal more of the variance in average occupational attainment among these workers than national origin and labor market context alone. Three factors have particularly large, significant, and consistent effects across the different labor market contexts and across the various Latino enclave cities. The first of these is self-employment. Self-employed workers have occupational attainment two to four points higher than immigrants who work for others. This underscores the contention by a number of scholars that entry into entrepreneurial ventures may be the best means for immigrants to achieve high levels of economic success (Borjas, 1987; Evans, 1989; Min, 1988; Sanders & Nee, 1996; Waldinger, Ward & Aldrich, 1985; Werbner, 1984). The other two important factors are the human capital measures of English fluency and educational attainment. Ability to speak English very well is a large and significant predictor of occupational attainment. Variation in the proportion of fluent English speakers between groups is considerable, with 25-30 percent of Mexican and 40-65 percent of Cuban long term immigrants reporting speaking English very well. Of course, far fewer recent immigrants speak fluent English and there is much less variance between groups in the proportions reporting speaking English very well. Finally, most Latinos in the U S . have low levels of educational attainment (Fernandez-Kelley & Schauffler, 1994), and Latin American immigrants are no exception. However, long term immigrants with a college degree can expect to receive occupational attainment almost 13 points higher than immigrant with a high school diploma. The effects for recent immigrants are also significant, but the magnitude of the effect is considerably less. It can be speculated from this finding that not all education receives equal rewards and that college degrees attained in Latin America have considerably less value than those earned in the United States.
Do the occupational attainment patterns of immigrants in same language, similar culture metropolitan areas with enclaves vary from the patterns of occupational attainment for immigrants in other ethnic enclave areas? Yes, comparing the occupational attainment outcomes of Latin American immigrants in Latino enclave PMSAs to those in Asian enclave areas reveals both significant differences in average occupational attainment and significant across-group differences within the two labor market contexts. Most of the longer-term immigrant groups have higher average occupational attainment in the Asian enclave areas, but the returns to their endowments are significantly lower. That is, occupational attainment is significantly lower in Asian enclave Copyright 2001 by Stephanie Bohon
Conclusions
123
places once human capital and other factors are held constant. This suggests that the immigrants who choose to live in Asian enclave areas, particularly the Cubans, Colombians, and other South Americans, are selected from among those who have done considerably well. Either unique advancement opportunities have brought them to these places, the unsuccessful Latin American immigrants have migrated elsewhere, or both processes are at work. For recent Latin American immigrants, metropolitan areas with Latino enclaves offer better opportunities for occupational attainment overall than Asian enclave places. That is, not only do recent immigrants have higher returns to their human capital in Latin American enclave PMSAs, but the average occupational attainment scores alone are higher in Latino enclave places than in Asian enclave areas. Furthermore, the between group differences in the metropolitan areas with Asian enclaves are considerably smaller in the Asian enclave places for recent immigrants than for longer-term immigrants. Can non-coethnic immigrants who share traits with the dominant enclave group expect to acquire jobs similar to those of the dominant enclave group, or will they be retained in lower status occupations? It depends on the enclave. For longer term immigrants in Los Angeles and New York, groups with endowments equal to or greater than those of the dominant enclave group have similar occupational attainment levels. In Miami, this is true for all but Nicaraguans who exceed Cubans in educational attainment and are similar in other areas, but who are extremely disadvantaged in occupational attainment. This finding holds true for both longer-term and recent immigrants. This finding may indicate that either Nicaraguan human capital was accrued primarily in Nicaragua and is rewarded at a lower rate, or that Nicaraguans pose a greater threat to Cubans in Miami and are more actively discriminated against. There is not enough evidence to assess Jersey City, since Cubans (the dominant enclave group) are significantly more highly endowed than the other national origin groups. Overall, the findings suggest that groups with endowments similar to the dominant enclave group can also expect occupational attainment that is somewhat similar. However, national origin still matters, and most groups will face some occupational attainment barriers (relative to the dominant group) due to nationality. This is particularly true for Nicaraguans in Miami, who tend to have occupational attainment levels only slightly higher than Mexicans and other Central Americans, despite the fact that they have high levels of education and considerable English fluency.
Copyright 2001 by Stephanie Bohon
124
Latinos in Ethnic Enclaves
IMPLICATIONS OF THIS STUDY Assimilation and Queuing Theory Segmented assimilation suggests that Latin American immigrants should become more like other Latinos in the labor market. In labor markets where Latin American immigrants and their descendants dominate that labor market (i.e., in the Latino enclave PMSAs), occupational attainment among Latin American immigrants should be high and there should be no significant difference in occupational attainment levels between different immigrant groups holding endowments constant. Also, in labor markets where Latinos are disadvantaged (i.e., the Asian enclave PMSAs and non-enclave PMSAs), no difference in occupational attainment between groups should be observed, but immigrants should have lower levels of occupational attainment there than those found in Latino enclave areas. If queuing is occurring in the Latino enclave PMSAs, occupational attainment will be high for some groups (e.g., the dominant enclave group and other groups high in the queue) and significantly lower for others. Overall, the findings suggest support for both queuing and segmented assimilation. Looking across the economic contexts, it is clear that queuing, in the form of discrimination is occurring in every labor market examined here. Furthermore, the across context OLS analysis shows that national origin matters even in Latino enclave PMSAs. Nonetheless, while these factors indicate a queuing effect, the across context analysis does not allow for a direct labor market comparative test of queuing versus segmented assimilation (see below). The within context analysis, however, allows for a direct examination of queuing and segmented assimilation. Hypotheses H l a and H l b pose alternative explanations for occupational attainment differences between longer-term coethnic and non-coethnic immigrants working within metropolitan areas with Latino enclaves. The findings show that certain groups face greater barriers to occupational attainment, due to national origin alone, than other groups. While Colombians and South Americans in Cuban enclave PMSAs have achieved outcomes similar to the dominant group, other immigrants have not. The findings suggest that queuing and segmented assimilation may both be occurring in the metropolitan areas. Hypotheses H2a and H2b parallel H l a and H l b but relate to recent immigrants. H2a suggests that non-coethnic workers will have lower occupational attainment than coethnic workers, while hypothesis H2b asserts that there will be no difference between the two groups. Overall, the findings suggest that occupational attainment outcomes across groups are more similar among recent immigrants. This suggests that most recent Latin American
Copyright 2001 by Stephanie Bohon
Conclusions
125
immigrants can expect to experience segmented assimilation within the Latino enclave area. In other words, it appears that there is less distinction between national origin groups in the hiring and promotion of recent immigrants. One possible explanation for this is that recent immigrants, regardless of national origin, are allotted high MSEI jobs, while, at the same time, there is an abundance of low MSEI jobs. Consequently, there is no need to reserve certain jobs for newly arrived members of the dominant enclave group. While the results of the analysis of occupational attainment among longerterm immigrants within the various Latino enclave PMSAs suggest queuing, and the results of the same analysis using recent immigrants suggest segmented assimilation, there is considerable variability from enclave place to enclave place. Segmented assimilation, on the whole, appears to be much more likely in Miami than in the other enclave areas. This may be an enclave-specific phenomenon whereby the increased diversity of industries in the Latino enclave in the Miami-Hialeah PMSA creates a situation where less immigrants, overall, are vying for any particular type of job, assuming that most Latino immigrants are, indeed, working in the Cuban enclave. There is also variation in the assimilation levels (relative to the dominant enclave group) from group to group. Cubans, Colombians, and other South Americans tend to do about equally well, while Nicaraguans, Mexicans, and other Central Americans are disadvantaged. This findings may be related to selfemployment. Colombians and South Americans have high levels of selfemployment in most of the enclave cities. Consequently, they may also have considerable control of the queuing mechanism, even when they are not the dominant enclave group. While queuing theory and segmented assimilation have been presented here as competing theoretical perspectives on the likely adaptation experience of immigrants, the findings suggest another possibility. Queuing and segmented assimilation can be simultaneous occurrences. In fact, one reasonable possibility is that segmented assimilation occurs over time but is slowed by queuing. In other words, segmented assimilation will eventually occur, but the speed at which it takes place depends on the ability of pan-ethnic groups to differentiate between groups in a way that impacts economic well-being. Another possibility is that queuing leads to segmented assimilation. If disadvantaged Latin American immigrant groups sense that they are facing barriers not only in the larger, mostly white, labor market, but also in the Latino labor market, as well, they may begin to ameliorate this by attempting to take on the characteristics of the dominant Latino group. This process would be similar to that undertaken by Eastern European immigrants in the U.S. after the turn of the century.
Copyright 2001 by Stephanie Bohon
Latinos in Ethnic Enclaves
Enclave Theory In addition to the relevance of these findings for segmented assimilation and queuing theory, this examination of occupational attainment in different labor market contexts also extends the literature on the ethnic enclave. First, it examines occupational attainment rather than income or self employment as an element of immigrant adaptation. Second, it compares outcomes in Latino, Asian, and non-enclave areas. Finally, it examines the impact of national origin on adaptation outcomes and contrasts those outcomes for coethnic and noncoethnic workers. These extensions are particularly useful because they clarify some contentious elements of the enclave debate insofar as they relate to occupational attainment. Portes and Jensen (1987, 1989, 1992) argue that the ethnic enclave protects immigrants from the human capital depreciation they are likely to experience in the secondary labor market. They also argue that the secondary labor market is the likely alternative labor market for most workers. While this study does not test the primary versus secondary labor market hypothesis, it does offer evidence to support the contention that the presence of an enclave creates better opportunities for economic advancement than those found elsewhere. At the same time, the findings here call into question the argument by Nee, Sanders, and Sernau (1994) that mixed economies create better economic outcomes for immigrants. This study not only supports Portes and Jensen's (1992) side of the enclave debate, it also supports other contentions by Portes and Stepick (1993) regarding bounded solidarity among Cubans in the ethnic enclave. They note: Language and a common culture provided Cubans with a basis for solidarity but by themselves were not enough to create a level of mutual support stronger than that typical of many other immigrant communities. Rather, the common circumstances of exile and the common experience of successive political defeats had cemented a strong sense of "we-ness" among these refugees . . . Bounded solidarity was the outcome, a mechanism that led the Cubans--despite diverse class origins and views---to patronize other Cuban-owned businesses and to prefer co-nationals as business associates. (pp. 135)
The findings here reiterate the contention that nationalism plays an important role in the advancement of immigrant occupations. Cuban immigrants can expect better access to higher status jobs than Mexicans, Nicaraguans, or Central Americans. Of course, this is true outside of the enclave, as well, but the results here underscore the fact that ethnicity remains an important consideration in
Copyright 2001 by Stephanie Bohon
Conclusions
127
hiring and promotion practices in the Miami enclave area, as well as Los Angeles, Jersey City, and New York. While the results of this study lend support for some elements of Portes' analysis of ethnic enclaves, they also call for some reevaluation of Bailey and Waldinger's (1991) training systems approach. The training systems approach suggests that business owners will hire disproportionately from one group for low paying, skilled blue-collar jobs, since network hiring ultimately lowers training costs. Since the findings from this study show that opportunities, overall, appear to be greater in the Latino enclave PMSAs for nearly all Latin American immigrant workers, this either suggests that jobs networks drastically cut across national boundaries, or that there are gains to be had from hiring Latin American immigrants, even if they are not recruited through migrant labor networks. Finally, the work here should force researchers to question some of the underlying arguments regarding why enclave areas offer advantages to immigrants. Despite the inference throughout the enclave literature that native language and native skills are in high demand, the findings here indicate that ability to speak English well is very important to economic adaptation, regardless of the labor market context. Work experience, while statistically significant, has a negligible impact on occupational attainment. So while this study reiterates the value of the ethnic enclave as a means of promoting immigrant adaptation, the results also call into question some of the mechanisms through which adaptation is assumed to be accomplished.
Implications for Public Policy The results of this study show that both labor market and national origin are important in the adaptation of Latin American immigrants. Immigrants experience the least occupational attainment if they are from Mexico, Nicaragua, or other Central American countries. Occupational attainment is also limited if the receiving area for these immigrants is an Asian or non-enclave area. This creates two distinct problems for public policy. The first problem has to do with the fact that the immigrants who receive the lowest returns to their endowments (i.e., they are discriminated against the most in every labor market), are those who are coming into the U.S. in the greatest numbers. For the last fifty years, more Mexicans have legally entered the U S . than any other national group. By 1990, immigrants from Mexico were outnumbered only by Filipinos, and almost 55,000 Mexicans were legally admitted across the U S . border that year (Russell, 1995). In fact, of the entire population of foreign-born Latin American immigrants living in the United States today, more than sixty percent are Mexican (Portes & Zhou, 1993). The Copyright 2001 by Stephanie Bohon
Latinos in Ethnic Enclaves findings of this study suggest that greater efforts need to be made to promote the economic adaptation of Mexican workers, and that more work should be done to explore the occupational attainment barriers faced by immigrants from Mexico, Nicaragua, and other Central American countries. The second public policy problem arises from the fact that most Latin American immigration is network-driven, and one of the side-effects of network-driven migration is the selection of a handful of U.S. states as major immigrant receiving areas (Tienda & Lii, 1987). The more than two-million US.-born Mexican-Americans in California attracts the overwhelming majority of Mexican immigrants (Wallace, 1986). California is also home to half of all Central American immigrants living in the United States and the largest population of Salvadorans, Guatemalans, Nicaraguans, and Costa Ricans (Wallace, 1986; Zohlberg, 1995). Of the other Central American sending countries, only New York has more Hondurans and Panamanians (Wallace, 1986). The result of this geographic concentration is a backlash by California residents, including many Asian immigrants, who are pushing for increasingly restrictive immigration legislation (Johnson, Farrell, & Guinn, 1997). Insofar as reception context is a factor in economic adaptation, these increasingly negative attitudes by natives in the receiving area can impact the labor market opportunities for Latin American immigrants (Portes & Borocz, 1989). T o summarize, not only are the least adaptive immigrants most likely to cross U S . borders, but they are also the most likely to choose California, including one of the four Asian enclaves, as their receiving area. The findings from this study suggest three ways that government can minimize these difficulties and improve the economic adaptation opportunities for Latin American immigrants. First of all, the creation of programs to promote access to education, particularly at the post-secondary level should have a marked impact on the occupational attainment of Latin American immigrants, according to the findings presented here. Second, the proliferation of English language classes, both in school or through adult community education programs should increase the opportunity for immigrants to establish fluency in English. This fluency should improve occupational attainment, on average. Finally, the establishment of low interest loan programs which make capital available to immigrants for entrepreneurial ventures should also improve the ability of immigrants to increase their occupational attainment, while, at the same time, providing more jobs for the continuing inflow of new immigrants from Latin America.
Copyright 2001 by Stephanie Bohon
Conclusions
RECOMMENDATIONS FOR FUTURE STUDY This study offers new insight into theories regarding immigrant adaptation and the role of the ethnic enclave. The findings from this research also suggest that there are many areas that still must be explored in order to fully understand economic adaptation barriers. This study examines segmented assimilation and queuing among firstgeneration immigrants, while most segmented assimilation studies examine second- or later-generations. While this study may offer a window into the future, the examination of occupational attainment outcomes within and across labor market contexts for the second generation is important. However, given the recent entry dates of many of the immigrant groups in this study, it will be a few years before the second generation is old enough enter the labor market. Returning to the first generation, the results presented here also suggest a need for longitudinal data. The results of the decomposition analysis show less discrimination for recent immigrants and more discrimination for longer-term immigrants. Differences in the decomposition coefficients cannot be tested across the two samples. However, if it is truly the case that greater discrimination (or more queuing) occurs after greater time in the labor market, segmented assimilation is called into question. That is, while the lack of differences between Latin American immigrant groups normally suggests segmented assimilation, group differences that become more pronounced, rather than muted, over time, would suggest queuing. Two areas for further research are also suggested by the findings regarding Nicaraguan adaptation, particularly in Miami. Given that Nicaraguans are strongly disadvantaged in the Miami enclave area, relative to Cubans, and that this disadvantage becomes greater once endowments are controlled for suggests that more can be learned from examining this group and others like them. One area for future study suggested by this finding pertains to the possible differential adaptation experience of refugees versus economic migrants. Does the sudden influx of a group on a labor market (generally due to widespread displacement) significantly effect long-term adaptation? A second area for future study suggested by the Nicaraguan findings entails the disaggregation of human capital factors to examine endowments accrued at the place of origin separate from those accrued after immigration. That is, one possible explanation for the Nicaraguan paradox is that their college educations, for example, were acquired in Nicaragua and are, therefore, rewarded differently than other groups' college educations that may have been acquired in the U.S. Available national data on immigrants rarely allows for the exploration of these kinds of distinctions.
Copyright 2001 by Stephanie Bohon
Latinos in Ethnic Enclaves Finally, this study focuses on the occupational attainment of Latin American immigrants men, while ignoring the role of women in the market. Enclave theory suggests that adaptation for immigrant women is in areas less easily observed than occupational attainment (Perez, 1986). Nonetheless, this contention has not been tested. Furthermore, the advantages (or disadvantages) of enclave work for women may be changing. In order to explore these two possibilities, the occupational attainment patterns for Latin American immigrant women must also be examined.
Copyright 2001 by Stephanie Bohon
Appendixes
Copyright 2001 by Stephanie Bohon
Appendix A
THE ALLOCATION OF IMMIGRANTS TO THE SAMPLE According to Portes and Jensen (1987, 1989, 1992), it is important to identify enclave immigrants by place of work rather than place of residence, since it is usually the case that those immigrants who have achieved the highest levels of economic attainment will live in affluent areas outside of the enclave, although they may continue to work and own businesses there. Sanders and Nee (1987), however, argue that selection of immigrants by place of work or place of residence makes little real difference. Place of work was used as the selection criteria for this study, but the following tables lend support to Sanders and Nee's assertion that either means of allocating respondents to enclaves would yield similar results.
Copyright 2001 by Stephanie Bohon
Table A.1. Sample Comparisons by Place of Work and Place of Residence in Latino Enclaves (Long-Term Immigrants)
%Central American % Mexican % Colombian
Copyright 2001 by Stephanie Bohon
Table A 2 . Sample Comparisons by Place of Work and Place of Residence in Asian Enclaves (Long-Term Immigrants) Primary Metropolitan Statistical Areas and Metropolitan Statistical Areas
% Nicaraguan
%Central American % Mexican
Copyright 2001 by Stephanie Bohon
Table A 3 . Sample Comparisons by Place of Work and Place of Residence in a Non-Enclave Location (Long-Term Immigrants)
Copyright 2001 by Stephanie Bohon
Table A.4. Sample Comparisons by Place of Work and Place of Residence in a Latino Enclave (Recent Immigrants)
Copyright 2001 by Stephanie Bohon
Table A 5 . Sample Comparisons by Place of Work and Place of Residence in an Asian Enclave (Recent Immigrants)
Copyright 2001 by Stephanie Bohon
Table A.6. Sample Comparisons by Place of Work and Place of Residence in a Non-Enclave Location (Recent Immigrants) Primary Metropolitan Statistical Areas and Metropolitan Statistical Areas
Copyright 2001 by Stephanie Bohon
I
Appendix B
Regression Coefficients Used in the Decomposition Analysis in Chapter Six (Tables B. 1-B.3) Regression Coefficients Used in the Decomposition Analysis in Chapter Seven (Tables B.4-B .6)
Copyright 2001 by Stephanie Bohon
142
Appendix B
Table B.1. Regression Coefficients (and Standard Errors) by Place of Origin for Latino Enclave Workers (Long Term) Cuba Year of entry 1980 to 1984 1970 to 1979 1960 to 1969 1959 or earlier Self-employed Educational Attainment Less than high school Some high school High school graduate Some college College graduate Age ~x~erience2 Ability to speak English Not at all Not well Well Very well Disabled Married Head of household Adjusted R~ n
-3.20*** (0.39) -0.88** (0.37) REF -0.96* (0.53) 3.27*** (0.33)
Central Nicaragua America 0.05 (1 .SO) 1.14 (1.82) REF 4.18 (3.12) 3.70** (158)
South Mexico Colombia America
-3.01*** -2.68*** -2.80** -2.20*** (0.60) (0.68) (0.27) (0.92) -2.15*** -1.40*** -3.09*** -0.75 (0.57) (0.22) (0.58) (0.80) REF REF REF REF 1.37 1.09** 1.41 2.36 (1.17) (0.37) (1 35) (1 54) 1.81*** 2.31*** 3.82*** 3.13*** (0.63) (0 S8) (0.28) (0.82)
-0.68 -0.74 -l.7O** -1.86*** -0.87 (0.56) (0.28) (2.14) (1.39) (0.54) -1.24** -1.63*** -1.17*** - 1.89* -1.94 (1.60) (0.43) (0.49) (0.27) (0.88) REF REF REF REF REF 4.15*** 4.23*** 3.54*** 2.58 1.64* (0.92) (0.45) (1.65) (0.62) (0.37) 12.75*** 9.45*** 9.67*** 11.04*** 9.98*** (0.96) (0.45) (1.66) (0.67) (0.43) O.lO*** 0.37*** 0.28*** 0.23 0.06 (0.04) (0.03) (0.17) (0 .O5) (0.11) -o.oo*** -0.00 -o.oo*** -0.01*** -0.00 (0.00) (0.00) (0.00) (0.00) (0.00) REF REF -0.11 0.50 (0.51) (2.17) 2.53*** 3.31 (0.55) (2.22) 5.002*** 7.86*** (0.60) (2.28) -2.54*** 8.53** (0.75) (3.02) 1.18*** 2.25* (1.28) (0.34) 1.58*** 0.36 (0.39) (1.36) 3835 ,3045 6612 470
Copyright 2001 by Stephanie Bohon
REF REF REF 1.27* 0.65 0.09 (1.47) (0.62) (0.27) 1.69 2.74*** 1.25*** ( 1.47) (0.63) (0.28) 4.95*** 2.50*** 5.65*** (0.66) (0.30) (1.55) 0.42 -2.32* - 1 .oo** (2.27) (0.45) (1.07) 1.47* 0.42* 0.61 (0.72) (0.38) (0.20) 1.22 1.87*** 1.21*** (0.40) (0.21) (0.83) .2339 ,1856 ,2935 3615 13181 1387
- 1.94"
(1.02) -2.06** (0.70) REF 3.63*** (0.67) 12.39*** (0.72) 0.23** (0.08) -o.oo** (0.00) REF -2.14 (1.41) 0.59 (1.39) 4.36** ( 1.42) -4.26** (1.41) -0..30 (0.54) 1.65** (0.61) ,3237 2643
143
Appendix B
Table B.2. Regression Coefficients (and Standard Errors) by Place of Origin for Asian Enclave Workers (Long Term) Cuba Year of entry 1980 to 1984
1970 to 1979 1960 to 1969 1959 or earlier Self-employed Educational Attainment Less than high school Some high school High school graduate Some college College graduate Age ~xprience' Ability to speak English Not at all Not well Well Very well Disabled Married Head of household Adjusted R~ n
Central Nicaragua America
South Mexico Colombia America
- 1.33 (1.32) -1.40 (1.27) REF 7.01 (1.78) 0.32 (1.44)
-3.23 (0.43) -1.89 (0.35) REF 0.61 (0.57) 2.40 (0.51)
0.67 (4.08) 1.73 (3 30) REF 16.57 (6.74) -3.59 (5.21)
-2.59 (1.99) 0.50 (1.76) REF - 1.48 (2.59) 131 (1.77)
4.00 -1 .I9 (6.48) (4.53) -1.17 5.22* (4.63) (2.73) REF REF 3.22 5.70* (4.25) (2.62) 14.45*** 18.63*** (4.32) (3.30) 0.71 0.14 (0.49) (0.33) -0.01 -0.00 (0.01) (0.01)
-0.0 1 (1.45) -1.61 (1.18) REF 4.79 (1.29) 10.90 (1.33) 0.09 (0.14) -0.00 (0.00)
-1 .O3 (0.44) -1.12 (0.43) REF 2.76 (0.55) 10.92 (0.64) 0.16 (0 .O4) -0.00 (0.00)
-5.64 (8.13) 3.10 (5.73) REF 0.81 (4.41) 15.78 (4.78) 0.19 (0.63) -0 .oo (0.01)
3.81 (4.09) -3.88 (2.66) REF 0.98 (1.94) 13.62 (2.14) 0.74 (0.25) -0.01 (0.00)
REF -19.99* (10.08) -17.91* (10.40) -10.90 (10.99) -7.41 (7.35) 0.29 (2.77) -1.52 (3.27) ,3712 118
REF -0.78 (2.10) 1.43 (2.12) 4.56 (2.16) -5.59 (2.44) 0.34 (0.95) 2.26 (1.03) ,3147 684
REF 0.10 (0.43) 0.70 (0.45) 2.61 (0.48) -0.58 (0.74) 0.41 (0.31) 1.40 (0.32) ,2111 4966
REF -27.32 (13.30) -28.15 (12.46) -16.90 (12.17) -6.52 (8.58) 4.65 (3.17) 2 .oo (3.56) SO36 84
REF -1.37 (2.77) -1.70 (1.58) NIA
5.46 (4.99) 1.57 (3.17) REF 36 7 (4.08) 3.22 (3.62)
Copyright 2001 by Stephanie Bohon
4.68 (2.86) 2.44 (2.89) REF 5.60* (2.91) 035 (3.16)
REF -4.51 (6.23) 1.56 (6.27) 3.48 (6.35) -7.80 (6.94) 4.75* (2.51) 1.62 (2.63) ,3373 165
-9.60 (4.33) -0.67 (1.71) 1.82 (1.96) ,3288 353
Appendix B
144
Table B 3 . Regression Coefficients (and Standard Errors) by Place of Origin for Non-Enclave Workers (Long Term)
Year of entry 1980 to 1984
1970 to 1979 1960 to 1969 1959 or earlier Self-employed Educational Attainment Less than high school Some high school High school graduate Some college College graduate Age ~x~erience~ Ability to speak English Not at all Not well Well Very well Disabled Married Head of household Adjusted R' n
Cuba
Central America
Mexico
Colombia
South America
-4.60** (130) -2.03 (1.41) REF 2.33 (2.21) 7.49*** (1.70)
3.47 (2.15) 3.26 (1.92) REF 1.47 (3 33) 0.93 (2.42)
-2.83 (0.39) -1.92 (0.32) REF 0.77 (0.48) 2.69 (0.43)
-3.72 (2.32) -1.20 (2.02) REF 1.61 (4.61) 6.02 (3.27)
-4.45 (1.82) -3.83 (1.41) REF -0.87 (2.34) 5.96 (1.57)
0.75 (2.47) -4.55** (1 233) REF 0.30 (1.87) 15.22*** (1.98) 0.37* (0.20) -0.01* (0.00)
0.65 (2.03) -0.20 (1.85) REF 6.66 (2 .OO) 15.33 (2.04) 0.41 (0.21) -0.00 (0.00)
-0.94 (0.40) -1.17 (0.40) REF 2.94 (0.50) 12.16 (0.57) 0.19 (0.O4) -0.00 (0.00)
-0.37 (3.41) -6.03 (3.27) REF 0.36 (3.41) 21.02 (2.20) 0.43 (2.30) -0.01 (2.58)
2.50 (2SO) 2.11 ( 1.SO) REF 6.73 (1.71) 12.40 (1.85) 0.57 (0.23) -0.01 (0.00)
REF 1.14 (2.70) -034 (2.83) 5.49* (2.93) -4.00 (3.37) -0.38 (1.59) 0.87 (1.77) ,4747 397
REF -0.97 (3.20) -1.12 (2.19) 3.64 (3.35) -2.65 (4.05) 1.64 (1.38) 3.99 (1.58) ,3676 265
REF 0.77 (0.46) 1.44 (0.47) 3.23 (0.49) -0.23 (0.68) 0.53 (0.30) 1.26 (0.32) ,2513 5254
REF -1.20 (4.21) 0.19 (.14) 0.14 (4.2 1) 3.71 (4.10) 5.20 (1.86) -2.88 (2.13) ,5477 205
REF 3.87 (4.11) 7 .O9 (4.09) 11.80 (4.16) 1.93 (3.29) -0.15 (1.33) 2.67 ( 1.47) ,4404 406
Copyright 2001 by Stephanie Bohon
Appendix B Table B.4. Regression Coefficients (and Standard Errors) by Place of Origin for Latino Enclave Workers (Recent Immigrants) Cuba Year of entry 1987 to 1990 1985 to 1986 Self-employed Educational Attainment Less than high school Some high school High school graduate Some college College graduate Age Experience2 Ability to speak English Not at all Not well Well Very well Disabled Married Head of household Adjusted R~ n
- 1.90* (1.02) REF 3.88** (1.55)
Central Nicaragua America -0.96 (0.88) REF 5.22** (1.75)
-0.18 (0.43) REF 3.72*** (0.91)
South Mexico Colombia America -0.39 (0.24) REF 4.66*** (0.53)
-2.69 (1.89) -2.69* (1.43) REF 1.37 (1.97) 8.12*** (1.83) 0.12 (0.16) -0.00 (0.00)
1.35 (1.23) REF 1.38 (1.60) 4.85*** (1.48) -0.05 (0.13) 0.00 (0.00)
-0.68* -2.05** (0.39) (0.70) -0.77* - 1.45* (0.38) (.70) REF REF 1.48" 1.42 (1.06) (0.61) 4.40*** 5.57*** (0.66) (1 .O3) 0.29*** 0.15*** (0.04) (0.06) -o.oo*** -o.oo** (0.00) (0.00)
REF 0.83 (1.22) 2.15 (1.59) 2.24 (1.63) -4.05 (3.55) 1.48 (1.31) 1.25 (1.35) ,2019 466
REF 0.90 (0.97) 3.56** (1.31) 1.05 (1.32) 1.60 (3.66) 1.54 (1.02) 2.61** (1.02) ,1310 607
REF 0.09 (0.48) 0.79 (0.67) 2.55*** (0.63) -0.98 ( 1.64 0.99 (0.61) 1.21* (0.53) ,1227 1908
Copyright 2001 by Stephanie Bohon
-0.41 ( 1.44)
REF 0.17 (0.27) 1.93*** (0.40) 1.29*** (.035) 0.84 (0.92) 1.10*** (0.35) 1.30*** (0.32) ,0824 5397
-0.50 ( 1.04)
REF 3.39* (1 S8) -0.72 (2.22) -0.69 (1.43) REF 0.66 (1.65) 9.88*** (1.65) ,017 (0.18) -0.00 (0.00) REF - 1.93
(1 33) -0.84 (1 .72) 1.95 (1.86) 1.47 (4.37) 1 .I9 (1.31) 1.62 (1.29) ,1694 428
1.40 (0.79) REF 1.58 (1.3 1) -0.69 ( 1.47)
-0.21 (1.18) REF 2.36 (1.16) 9.41 (1.28) 0.40 (0.12) -0.00 (0.00) REF 0.33 ( 1.24) 3.43 (1.35) 8.36 (1.41) 4.10 (4.58) -0.09 (0.97) 4.55 (0.97) ,3062 84
Appendix B
146
Table B.5. Regression Coefficients (and Standard Errors) by Place of Origin for Asian Enclave Workers (Recent Immigrants) Nicaragua Year of entry 1987 to 1990 1985 to 1986 Self-employed Educational Attainment Less than high school Some high school High school graduate Some college College graduate Age ~x~erience~ Ability to speak English Not at all Not well Well Very well Disabled Married Head of household Adjusted R~ n
Central America
Mexico
South America
-1.62 -1.66*** -0.94 1.49 (2.32) (1.10) (0.32) (2.83) REFERENCE REFERENCE REFERENCE REFERENCE 5.43 3.34 3.05** -2.25 (5.30) (3.OO) (1.OO) (5.14) 7.78 -0.32 4.23 -0.21 (7.03) (0.50) (4.27) (1.79) -4.42 -0.23 0.47 -3.88** (3.24) (5.03) (0.49) (1.73) REFERENCE REFERENCE REFERENCE REFERENCE -18 0 0.93 -0.88 6.75** (3.63) (4.28) (0.80) (2.79) 2.92*** 8.13* 3.55 6.47 (5.02) (2.53) (4.69) (0.84) 0.21*** -0.08 0.51** 0.62 (0.35) (0.62) (0.05) (0.18) -0.00 -0.01** -0.01 -o.oo** (0.01) (0.01) (0.00) (0.00) REFERENCE REFERENCE REFERENCE REFERENCE 5.89* 2.46* 0.40 0.83 (3.04) (1.29) (0.36) (4.99) 12.80*** 3.79* 1.79*** 3.32 (3.67) (1.87) (0.52) (4.95) 6.05 3.39* 1.32** 4.51 (4.58) (1.63) (0.49) (5.14) -6.14 -8.23* -0.72 3.23 (4.68) (4.63) (1.20) (13.71) 1.30** 7.30* 1.59 -0.38 (3.18) (1.76) (0.50) (3.89) 1.20 1.47 0.40 0.40 (3.43) ( 1.47) (0.47) (3 32) 3372 ,1931 ,0652 ,3305 72 320 2735 88
Copyright 2001 by Stephanie Bohon
147
Appendix B
Table B.6. Regression Coefficients (and Standard Errors) by Place of Origin for Non-Enclave Workers (Recent Immigrants) Central America Year of entry 1987 to 1990
Mexico
Colombia
South America
-1.57 (3.26) REFERENCE 9.06 (6.60)
3.42* (1.76) REFERENCE 2.85 (2.88)
2.04 (2.15) 2.59 (1.97) REFERENCE 7.79** (2.68) 11.47*** (3.30) 0.38* (0.22) -0.00 (0.00)
-0.37 037 (0.61) (6.43) - 1.23 -1.11* (0.61) (4.23) REFERENCE REFERENCE 2.77** 0.86 (0.93) (5.3 1) 21.77*** 7.40*** (1.01) (5.89) 0.10 0.30*** (0.51) (0.07) -0.00 -o.oo*** (0.01) (0.00)
-0.43 (3.02) -2.35 (2.44) REFERENCE 0.26 (3.SO) 7.05** (2.66) 0.45 (0.28)-0 .o 1 (0.01)
REFERENCE -235 (1.78) -1.36 (2.28) -1.77 (2.14) -2.65 (5.13) -2.73 (1.88) 4.71** (1.72) ,1785 165
REFERENCE 034 (.049) 1.74** (0.64)** 1.74 (0.65) 1.76 (1.57) 1.26* (0.57) 0.70 (0.53) ,1466 1770
REFERENCE 1.27 (2.58) 7.80** (2.90)* 5.86 (3.10) 0.30 (5.56) -1.17 (2.16) 2.43 (2.08) ,2096 122
-1.17
0.08 (0.41) REFERENCE REFERENCE -2.94 6.23*** (3.66) (1.05) ( 1.47)
I 985 to 1986 Self-employed Educational Attainment Less than high school Some high school High school graduate Some college College graduate Age ~x~erience~ Ability to speak English Not at all Not well Well Very well Disabled Married Head of household Adjusted R~ n
Copyright 2001 by Stephanie Bohon
REFERENCE -1.36 (4.12) 4.02 (4.64) -4.82 (6.40) -10.62 (9.38) 0.08 (3.74) 1.25 (3.58) ,3648 73
Notes
1. While this definition of ethnic enclaves is appropriate for the research undertaken here, there is considerable debate surrounding the appropriate definition of ethnic enclaves. A detailed discussion of this debate can be found in Chapter Three. 2. The phrase "dominant enclave group" refers to immigrants in a metropolitan area from the same country of origin who control the majority of immigrant-owned businesses in that area and their coethnic employees. 3. T o avoid confusion over distinctions between and within Latin American groups, for the purposes of this study, the term coethnic refers to immigrants from the same country of origin, while non-coethnic refers to immigrants who do not share a country of origin with the dominant enclave group, unless otherwise specified. Phrases such as interethnic competition and inter-ethnic cooperation refer to relationships between groups from different sending countries within Latin America. 4. This assertion is based on the application of segmented assimilation theory to first generation immigrants. The justification for such an application is discussed at length in Chapter Two. 5. There is no precedent in the literature to posit queuing theory as the alternative to segmented assimilation theory; however, this may be because the theories developed in different academic disciplines. In sociology, segmented assimilation is generally viewed as an alternative to straight-line assimilation, while in economics, queuing is seen as an argument against neoclasssical labor market theory. Nonetheless, these theories have oppositional value for the study of immigrant adaptation. 6. These hypotheses suggest that workers in metropolitan areas with enclaves work for enclave firms. Current enclave literature supports the contention that co-ethnic workers in areas with enclaves are likely to be enclave participants (Logan, Alba & McNulty, 1994; Portes & Jensen, 1989). Nonetheless, this assumption cannot be tested with available data (Portes & Jensen, 1992).The research presented here is predicated on the notion that whether workers choose to work in the enclave or not, the presence of an
Copyright 2001 by Stephanie Bohon
150
Notes
enclave may create occupational opportunities for immigrants not found in other economic contexts. For simplicity, however, terms such as "enclave worker" are used in this study to refer to workers in primary metropolitan statistical areas that have Latino or Asian enclaves. 7. Discrimination refers to the differences in average occupational attainment not explained by measured factors such as education and English fluency. The discrimination term used in this study will be explained more fully in Chapter Four. 8. The need for such research was recently noted in Model and Ladipo (1996) who argue that, in order to properly assess the mechanisms through which different ethnic groups achieve occupational attainment, it is necessary to separate the ethnic economy from other sectors of the labor market. (pp. 505-506) 9. It should be noted that Portes' contention regarding the favorable Nicaraguan reception is questioned by Stanley (1993) who argues that the Cuban community in Miami displayed open hostility toward the resettling Nicaraguans and pressured the U.S. government to relocate these refugees away from Southern Florida. 10. The concept of primary and secondary labor markets is based on dual labor market theory which suggests that there are separate markets for good and bad jobs. It is generally assumed that Latin American immigrants are barred from competing in the primary labor market due to discrimination. The position of the enclave vis-a-vis the dual labor market remains under debate. The enclave does provide better outcomes for immigrants through self-employment opportunities, however, regardless of whether the enclave is seen as a part of the secondary labor market or as a conduit between the two markets. 11. Light, Sabagh, Bozoregmehr, and Der-Martirosian (1993) argue that the term "ethnic enclave" should be discarded altogether in favor of an ethnic economy concept which does not entail geographic boundedness, however, they continue to distinguish between the ethnic enclave (bounded) and the ethnic economy (unbounded). Waldinger (1992) also supports this position, contending that scholars should move beyond the enclave debate and focus on the ethnic economy in order to make meaningful new assertions about immigrant adaptation. 12. Los Angeles is, however, home to nearly 50,000 Cubans who have created an enclave with industrial specialization in textiles, apparel, ethnic trade, and services (Logan, Alba & McNulty, 1994). Unfortunately, very little has been written about the Cuban enclaves outside of Miami, however, there is some indication that the Cuban enclaves in Jersey City, New Jersey and Los Angeles exhibit patterns more similar to the Chinese economies than to the Miami enclave (Logan, Alba, & McNulty, 1994; Portes & Stepick, 1993). 13. The Japanese immigrant population numbered only 117,200 in 1990, making it one of the smallest Asian populations in Los Angeles. Nonetheless, they have been able to create an enclave through control of much of the local agricultural market. The larger
Copyright 2001 by Stephanie Bohon
Notes
151
Asian immigrant groups--notably the Filipinos and the Indians--have also been able to own and operate businesses in concentrated economic sectors; however, neither the rates of concentration or the levels of co-ethnic employment for these two groups are pronounced enough for inclusion as ethnic enclaves (Logan, Alba & McNulty, 1994). 14. The limitation of the sample to men currently in the civilian labor force and not In school necessitates that another portion of the population be excluded from the analysis. Such limitations may create a selection bias that can impact the validity of inferential statistics, especially regression coefficients (Heckman, 1979; Winship & Mare, 1992). Since much of the data analysis for this study involves regression, there is some concern about this potential selection bias problem. Unfortunately, despite the myriad techniques currently available to adjust for selection bias (Winship & Mare, 1992), a serious problem with these techniques is that they are likely to distort the fit of the model. That is, the use of a correction factor can sometimes produce estimates that are less accurate than the parameters associated with an uncorrected regression analysis (Hartman, 1991), and the resultant outcome can be "catastrophic" (Stolzenberg & Relles, 1985). It has also been noted that the probability of this undesirable outcome is not small (Lillard, Smith, & Welch, 1986), and the standard method of calculating an inverse Mills Ratio on which the dependent variable can be regressed (along with the other independent variables) does not provide an intuitive means for assessing the impact of selection bias on the model (Stolzenberg & Relles, 1990, 1997). In other words, adjusting for bias may create worse results than not adjusting for bias. For these reasons, selection bias has not been adjusted for in the analysis presented here. 15. While the term Latino is used throughout this paper, the term Hispanic is also used in cases where I directly refer to the U S . Census and U S . Census data, since the Bureau uses the term Hispanic to refer to persons of Central or South American origin. Latino immigrants were selected for this study if it was indicated in the 1990 Census that they were of Hispanic origin. 16. There remains the theoretical possibility of a non-Latino, non-Asian enclave. However, while there are many diverse ethnic neighborhoods in the U S . , and some Middle Eastern groups have established ethnic economies, no non-Asian, non-Latino enclave has been documented consistently in the literature. 17. The category, Central America, includes only persons born on the Central American isthmus and does not include Mexicans or Nicaraguans. The category, South America, includes only those persons born in continental South America and does not include Colombians. 18. It should be noted that the use of the term discrimination to describe average differences not accounted for by either endowments or interaction factors is a sociological convention that may be misleading. Jones and Kelley (1984) used the term in their methodological treatise on the subject of decomposition, and it has been adapted by other researcher (e.g., Model & Ladipo, 1996) in explaining occupational attainment
Copyright 2001 by Stephanie Bohon
152
Notes
differences. Nonetheless, the term discrimination, in this application, refers only to unexplained differences in occupational attainment. In other words, it is the portion of the average differences in MSEI not explained by measured endowments. What is called discrimination, therefore, may include some discrimination but also many other uncontrolled factors. 19. See Table 2. 20. This finding suggests an interaction between place of origin and other variables in the analysis, such as education and year of entry. These interaction effects were modeled separately, and the interaction between education and ethnicity and year of entry and ethnicity were significant for South Americans; however, the discussion of the coefficients of interaction effects precludes the discussion of the main effects of place or origin which is the focus of this analysis. The results of the interaction models are, however, available from the author upon request. 21. The impact of country of origin for both long-term and recent immigrants was tested in logit models which regressed unemployment and self-employment on place of origin and the control variables. The models showed that unemployment is not significantly affected by national origin for long-term immigrants in both the models with or without the control variables. The exception is in New York, where there is a significant interaction between country of origin and year of entry. For recent immigrants, Mexican and Central American origin resulted in a significant likelihood of unemployment in Miami, but not in New York. Odds of self-employment for long-term immigrants are significant and negative for national origin in each of the areas for Mexicans and Central Americans and for Colombians in Jersey City. The picture is less clear for recent immigrants. In Miami, Nicaraguans are significantly less likely to be selfemployed, relative to Cubans, while Colombians and South Americans are significantly more likely to be self-employed. In New York, Central and South Americans are significantly less likely, relative to Colombians, to be self-employed. 22. Preliminary analyses were conducted using Cubans as the reference group. The results showed that all groups were disadvantaged, relative to Cubans, in the Latino and non-enclave areas (except Colombians in non-enclaves). Those same results suggested that Cubans in Asian enclaves are highly selected from among the Cuban immigrant population, and that the selection bias would distort the findings for the Asian enclaves. For example, the decomposition analysis shows a discrimination component of -20.01 for Nicaraguans, relative to Cubans. Interpreting this finding as a 20 point advantage for Nicaraguans over Cubans, net of endowments, is counter-intuitive and is artificially created by the poor model fit for Cubans in the Asian enclave context. It is likely that the few Cubans who choose to live in the Asian enclave areas do so because they are unusually successful there for reasons that cannot be controlled for in the model. The results of the OLS regressions and this decomposition analysis are available from the author upon request.
Copyright 2001 by Stephanie Bohon
Notes
153
23. The coefficients resulting from the decomposition analysis are not amenable to parametric testing, and some caution should be taken in interpreting coefficients for subsample groups smaller than 100. 24. Some factors that may bias the results of the decomposition analysis should be noted. The discrimination term includes all average differences not attributable to endowments or the interaction effect. Poorly specified models, then, will have larger discrimination effects, since the endowment variables will be unable to capture much of the variance. In this analysis, there may be similarities or differences in the labor markets that are not reflected in the model specification. Another important point is that decomposition assumes that endowments are rewarded at the same rate for each group. This creates a potential source of difficulty here, since some groups may have been more likely to acquire factors such as education at their country of origin than other groups. Insofar as different types of education are rewarded differently, this can inflate the discrimination coefficient and distort the interaction term in the decomposition analysis. 25. The results of the decomposition analysis are not as reliable as those found in Chapter Six for a number of reasons. First, the sample sizes of recent immigrants are much smaller for all of the groups. Second, the presence of very small sample sizes forced the elimination of selected groups from the analysis which may work to lower the observed discrimination effect. Furthermore, the results for Nicaraguans and South Americans in Asian enclave areas and the results for the Colombians in the non-enclave areas are based on sample sizes of less than 100 (see Appendix C). Finally, the individual regression models that are the basis of the decomposition analysis are less robust than those using the long-term immigrant samples. More specifically, the recent immigrant OLS models (see Appendix C) have lower R 2 values and fewer significant regression coefficients than those used in Chapter Six. Again, coefficients in the decomposition analysis are not amenable to parametric testing, so there is less certainty surrounding the significance of within group differences.
Copyright 2001 by Stephanie Bohon
Bibliography
Alba, R. D. (1990). Ethnic identity: The transformation of white America. New Haven, CT: Yale University Press. Alba, R. D., & Logan, J. R. (1992). Assimilation and stratification in the home ownership patterns of racial and ethnic groups. International Migration Review, 26, 1314-1341. Alba, R. D., & Nee, V. (1997). Rethinking assimilation theory for a new era of immigration. International Migration Review, 31,826-874. Aldrich, H., & Waldinger, R. (1990). Ethnicity and entrepreneurship. Annual Review of Sociology, 16, 11 1-35. Bailey, T., & Waldinger, R. (1991). Primary, secondary, and enclave labor markets: A training system approach. American Sociological Review, 5 6 , 4 3 2 4 5 . Barrera, M . (1979). Race and class in the Southwest. South Bend, IN: University of Notre Dame Press. Bonacich, E. (1987). "Making it" in America: A social evaluation of the ethics of immigrant entrepreneurship. Sociological Perspectives, 3 0 , 4 4 6 4 3 6 . Bonacich, E. (1988). The social costs of immigrant entrepreneurship. Amerasia, 14, 119128. Bonacich. R., & Sung, T. H. (1982). A portrait of Korean small business in Los Angeles: 1977. In E. Y u, E. H. Phillips, & E. S. Y ang (Eds.), Koreans in Los Angeles. Los Angeles: Koryo Research Institute and Center for Korean- American and Korean Studies, California State University. Bonacich, E., Light, I. H., & Wong, C. C. (1977). Koreans in small business. Society, 14, 54-59. Borjas, G . J. (1985). Assimilation, changes in cohort quality, and the earnings of immigrants. Journal of k b o r Economics, 3,463-489. Borjas, G. J. (1986). The self-employment experience of immigrants. Journal of Human Resources, 21,485-506.
Copyright 2001 by Stephanie Bohon
156
Bibliography
Borjas, G. J. (1987). Immigrants, minorities, and labor market competition, Industrial and Labor Relations Review, 40, 382-392. Borjas, G. J. (1990). Friends or strangers: The impact of immigrants on the U S . economy. New York: Basic Books. Borjas, G. J. (1991). Immigrants in the U S . labor market: 1940-1980. American Economic Review, 81,287-91. Castile, M. A. (1994). A preliminary analysis of emigration determinants in Mexico, Central America, northern South America and the Caribbean. International Migration, 32,269-306. Chiswick, B. R. (1980). An analysis of the economic progress and impact of immigrants (Final Report to the U.S. Department of Labor). Washington, DC: U.S. Department of Labor. Clark, W. A. V. (1982). Recent research on migration and mobility: A review and interpretation. Progress in Planning, 18, 1-56. Denton, N. A , , & Massey, D. S. (1988). Residential segregation of blacks, Hispanics, and Asians by socioeconomic status and generation. Social Science Quarterly, 69, 797817. DeWind, J., & Kasinitz, P. (1997). Everything old is new again? Processes and theories of immigrant incorporation. International Migration Review, 31, 1096-1 1 11. Duleep, H. O., & Regets, M. C. (1996). The elusive concept of immigrant quality: Evidence from 1970-1990 (Working Paper PRIP-UI-41). Washington, DC: The Urban Institute. Evans, M. D. R. (1989). Immigrant entrepreneurship: Effects of ethnic market size and isolated labor pool. American Sociological Review, 54, 950-962. FernAndez-Kelly ,M. P., & Garcia, A. M. (1990). Power surrendered, power restored: The politics of work and family among Hispanic garment workers in California and Florida. In L. Tilly & P. Guerin (Eds.), Women and politics in America (pp. 130149). New Y ork: Russell Sage Foundation. Fernindez-Kelly, M. P., & Schauffler, R. (1994). Divided fates: Immigrant children in a restructured U S . economy. International Migration Review, 28, 6 6 2 4 8 9 . Forment, C . (1987). The formation of the Cuban American enclave, 1959-1979: The primacy of politics. Unpublished manuscript, Harvard University, Department of Sociology, Cambridge. Forment, C. (1989). Political practice and the rise of the ethnic enclave: The CubanAmerican case, 1959-1979. Theory and Society, 18,47-81. Gans, H. J. (1992). Second-generation decline: Scenarios for the economic and ethnic futures of the post-1965 American immigrants. Ethnic and Racial Studies, 15, 172192.
Copyright 2001 by Stephanie Bohon
Bibliography Gilbertson, G. A. (1995). Women's labor and enclave employment: The case of Colombian and Dominican women in New York City. International Migration Review, 29, 657-670. Gilbertson, G., & Gurak, D. T. (1992). Household transitions in the migration of Dominicans and Colombians to New York. International Migration Review, 26, 2245. Glazer, N. (1993). Is assimilation dead? Annals of the American Academy of Political and Social Science, 530, 122-136. Gurak, D. T. (1987). Family formation and marital selectivity among Colombian and Dominican immigrants in New York City. International Migration Review, 21, 275297. Haberfield, Y. (1993). Immigration and ethnic origin: The effect of demographic attributes on earnings of Israeli men and women. International Migration Review, 2 7,286305. Hansen, N., & Cardenas, G. (1988). Immigrant and native ethnic enterprises in Mexican American neighborhoods: Differing perceptions of Mexican immigrant workers. International Migration Review, 22,226-242. Hartman, R. (1991). A Monte Carlo analysis of alternative estimators in models involving selectivity. Journal of Business and Economic Statistics, 9 , 4 1 4 9 . Hauser, R. M., & Warren, J. R. (1997). Socioeconomic indexes for occupations: A review, update, and critique. Sociological Methodology, 27, 177-298. Heckman, J. (1979). Sample selection bias as a specification error. Econometrika, 47, 153-161. Jacobsen, K. (1996). Policy responses of host governments to mass refugee influxes in the Third World. International Migration Review, 30,655-678 Johnson, 9. H., Farrell, W. C., & Guinn, C. (1997). Immigration reform and the browning of America: Tensions, conflicts and community instability in metropolitan Los Angeles. International Migration Review, 31, 1055-1095. Jensen, L., & Portes, A. (1992). Correction to Portes and Jensen, ASR, December, 1989. American Sociological Review, 57,411414. Jiobu, R. M. (1988). Ethnic hegemony and the Japanese of California. American Sociological Review, 53,353-367. Jones, F . L., & Kelley, J. (1984). Decomposition of differences between groups: A cautionary note on measuring discrimination. Sociological Methods and Research, 12,323-343. Kao, G., & Tienda, M. (1995). Optimism and achievement: The educational performances of immigrant youth. Social Science Quarterly, 76, 1-19. Kim, K. C., Huhr, W. M., & Femandez, M. ( I 989). Intra-group differences in business participation: Three Asian immigrant groups. International Migration Review, 19, 82-1 11.
Copyright 2001 by Stephanie Bohon
158
Bibliography
Kossoudji, S. A. (1989). Immigrant worker assimilation: Is it a labor market phenomenon? Journal of Human Resources, 24,494427. Kossoudji, S. A., & Cobb-Clark, D. A. (1996). Finding good opportunities within unauthorized markets: U.S. occupational mobility for male Hispanic workers. International Migration Review, 30,901-924. Lieberson, S. (1980). A piece of the pie. Berkeley, CA: University of California Press. Light, I. H. (1972). Ethnic enterprise in America: Business and welfare among Chinese, Japanese, and blacks. Berkeley: University of California Press. Light, I. H. (1980). Asian enterprise in America: Chinese, Japanese and Koreans in small business. In S. Cummings (Ed.), Self-help in urban America (pp. 35- 57). Port Washington, NY: Kennikat Press. Light, I. H. (1984). Immigrant and ethnic enterprise in North America. Ethnic and Racial Studies, 7, 195-216. Light, I., & Bonacich, E. (1988). Immigrant entrepreneurs: Koreans in Los Angeles, 1965-1982. Berkeley, CA: University of California Press. Light, I., Sabagh, G., Bozorgmehr, M., & Der-Martirosian, C. (1993). Internal ethnicity in the ethnic economy. Ethnic and Racial Studies, 16,581-597. Light, I., Sabagh, G., Bozorgmehr, M., & Der-Martirosian, C. (1994). Beyond the ethnic enclave economy. Social Problems, 41,65-80. Lillard, L., Smith, J. P., & Welch, F. (1986). What do we really know about wages? The importance of nonreporting and census imputation. Journal of Political Economy, 94,489-506. Lindstrom, D. P., & Massey, D. S. (1994). Selective emigration, cohort quality, and models of immigrant assimilation. Social Science Research, 23,3 15-349. Logan, J., Alba, R., & McNulty, T. (1994). Ethnic economies in metropolitan regions: Miami and beyond. Social Forces, 72,691-724. Martin, E., Demaio, T. J., & Campanelli, P. C. (1990). Context effects for census measures of race and Hispanic origin. Public Opinion Quarterly, 54, 551-566. Massey, D. S. (1987). Understanding Mexican migration to the United States. American Journal of Sociology, 92, 1372-1403. Massey, D. S. (1995). The new immigration and ethnicity in the United States. Population and Development Review, 21,631652. McLaughlin, D. K., & Perman, L. (1991). Returns versus endowments in the earning attainment process for metropolitan and nonmetropolitan men and women. Rural Sociology, 56, 339-365. McManus, W. (1990). Labor market effects of language enclaves: Hispanic men in the United States. Journal of Human Resources, 25, 228-252. Min, P. G. (1987). Factors contributing to ethnic business: A comprehensive synthesis. International Journal of Comparative Sociology, 28(34), 173-1 93.
Copyright 2001 by Stephanie Bohon
Bibliography
159
Min, P. G. (1988). Ethnic small business enterprise. Staten Island, NY: Center for Migration Studies. Model, S. (1992). The ethnic economy: Cubans and Chinese reconsidered. Sociological Quarterly, 33, 63-82. Model, S., & Ladipo, D. (1996). Context and opportunity: Minorities in London and New York. Social Forces, 75,485-5 10. Myers, D. (1992). Analysis with local census data: Portraits of change. Boston: Academic Press. Nakao, K., & Treas, J. (1994). Updating occupational prestige and socioeconomic scores: How the new measures measure up. Sociological Methodology, 24, 1-72. Nee, V., Sanders, J. M., & Sernau, S. (1994). Job transitions in an immigrant metropolis: Ethnic boundaries and the mixed economy. American Sociological Review, 59, 849872. Passel, J. S. (1993). Estimation of population coverage in the 1990 United States Census based on demographic analysis: Comment. Journal of the American Statistical Association, 88, 1074-1077. Perez, L. (1986). Immigrant economic adjustment and family organization: The Cuban success story revisited. International Migration Review, 20,4-20. Portes, A. (1997). Immigration theory for a new century: Some problems and opportunities. International Migration Review, 31,799-825. Portes, A., & Bach, R. (1985). Latin journey: Cuban and Mexican immigrants in the United States. Berkeley, CA: University of California Press. Portes, A., & BoroEz, J. (1989). Contemporary immigration: Theoretical perspectives on its determinants and modes of incorporation. International Migration Review, 23, 606-630. Portes, A,, & Grosfoguel, R. (1994). Caribbean diaspora: Migration and ethnic communities. Annuls of the American Academy of Political and Social Science, 533, 48-69. Portes, A., & Jensen, L. (1987). What's an ethnic enclave? The case for conceptual clarity. American Sociological Review, 52, 768-770. Portes, A., & Jensen, L. (1989). The enclave and the entrants: Patterns of ethnic enterprise in Miami before and after Mariel. American Sociological Review, 54, 929-949. Portes, A., & Jensen, L. (1992). Reply to Sanders and Nee: Disproving the enclave hypothesis. American Sociological Review, 57,418-420. Portes, A., & Manning, R. (1986). The immigrant enclave: theory and empirical examples. In J. Nagel & S. Olzak, (Eds.), Competitive ethnic relations (pp. ). Orlando, FL: Academic Press. Portes, A., McLeod, S. A., Jr., & Parker, R. N. (1978). Immigrant aspirations. Sociology of Education, 51,241-260.
Copyright 2001 by Stephanie Bohon
Bibliography Portes, A., & Rumbaut, R. (1996). Immigrant America: A portrait. Berkeley, CA: University of California Press. Portes, A., & Stepick, A. (1993). City on the edge: The transformation of Miami. Berkeley, CA: University of California Press. Portes, A., & Zhou, M. (1992). Gaining the upper hand: Economic mobility among immigrant and domestic minorities. Ethnic and Racial Studies, 15,491- 522. Portes, A., & Zhou, M. (1993). The new second generation: Segmented assimilation and its variants. Annals of the American Academy of Political and Social Science, 530, 74-96. Portes, A., & Zhou, M. (1995). Divergent destinies: Immigration, poverty and entrepreneurship. In K. McFate, R. Lawson, & W. J. Wilson (Eds.) Poverty, inequality and thefuture of socialpolicy (pp. ). New York: Russell Sage. Portes, A., & Zhou, M. (1996). Self-employment and the earnings of immigrants. American Sociological Review, 61,219-230, Reimers, C . (1984). The wage structure of Hispanic men: Implications for policy. Social Science Quarterly, 65,40 1 4 6 . Reitz, J. (1990). Ethnic concentrations in labor markets and their implications for ethnic inequality. In R. Breton, W. Isajiw, W. Kalbach, & J. Reitz (Eds.), Ethnic identity and inequality: Varieties of experience in a Canadian City (pp. 135-195). Toronto, Canada: University of Toronto Press. Rich, B. L. (1995). Explaining feminization in the U S . banking industry, 1940-1980: Human capital, dual labor markets, or gender queuing? Sociological Perspectives, 38, 357-380. Robinson, C. (1988). Language choice: The distribution of language skills and earnings in a dual-language economy. Research in Labor Economics, 9,53-90. Roos, P. A., & Hennessy, J. F. (1987). Assimilation or exclusion? Japanese and Mexican Americans in California. Sociological Forum, 2,278-304. Roos, P. A., & Jones, K. W. (1993). Shifting gender boundaries: Women's inroads into academic sociology. Work and Occupations, 20,395-428. Roos, P. A., & Reskin, B. F. (1992). Occupational desegregation in the 1970s: Integration and economic equity? Sociological Perspectives, 35, 69-91. Rumbaut, R. G. (1997). Assimilation and its discontents: Between rhetoric and reality. International Migration Review, 31,923-960. Russell, S. S. (1995). Migration patterns of U S . foreign policy interest. In M. S. Teitelbaum & M. Weiner (Eds.), Threatened peoples, threatened borders: World migration and US.policy (pp. 39-87). New York: W. W. Norton. Sakamoto, A,, & Chen, M. D. (1991). Inequality and attainment in a dual labor market. American Sociological Review, 56,295-308. Sakamoto, A., & Powers, D. A. (1995). Education and the dual labor market for Japanese men. American Sociological Review, 60,222-246.
Copyright 2001 by Stephanie Bohon
Bibliography Sanders, J., & Nee, V. (1987). Limits of ethic solidarity in the enclave economy. American Sociological Review, 52,745-773. Sanders, J., & Nee, V. (1992). Comment on Portes and Jensen, ASR, December, 1989: Problems in resolving the enclave economy debate. American Sociological Review, 57,415418. Sanders, J. M., & Nee, V. (1996). Immigrant self-employment: The family as social capital and the value of human capital. American Sociological Review, 61, 23 1- 249. Schirm, A. L. (1991). The effects of census undercount adjustment on congressional apportionment. Journal of the American Statistical Association, 86,526-541. Semyonov, M . (1988). Bi-ethnic labor markets, mono-ethnic labor markets, and socioeconomic inequality. American Sociological Review, 53, 256-266. Semyonov, M., Hoyt, D. R., & Scott, R. I. (1984). The place of odds ratios in the study of place, race and differential occupational opportunities. Demography, 21,667- 671. Stanley, W. D. (1993). Blessing or menace: Security implications of Central American migration. In M. Weiner (Ed.), International migration and security (pp. 229- 260). Boulder, CO: Westview. Stepick, A., & Portes, A . (1986). Flight into despair: A profile of recent Haitian refugees in south Florida. International Migration Review, 20, 329-350. Stevens, G., & Featherman, D. L. (1981). A revised socioeconomic index of occupational status. Social Science Research, 10, 364-395. Stolzenberg, R. M., & Relles, D. A. (1985). Calculation and practical application oj GMAT predictive validity measures. Santa Monica, CA: Graduate Management Admissions Council. Stolzenberg, R. M., & Relles, D. A. (1990). Theory testing in a world of constrained research design: The significance of Heckman's censored sampling bias correction for nonexperimental research. Sociological Methods and Research, 18,395415. Stolzenberg, R. M., & Relles, D. A. (1997). Tools for intuition about sample selection bias and its correction. American Sociological Review, 62,494-507. Tienda, M., & Lii, D. (1987). Minority concentration and earnings inequality: Blacks, Hispanics, and Asians compared. American Journal of Sociology, 93, 141-165. Tienda, M., & Wilson, F. D. (1992). Migration and the earnings of Hispanic men. American Sociological Review, 57, 661-678. Waldinger, R. (1989). Structural opportunity or ethnic advantage? Immigrant business development in New York. International Migration Review, 23,48-72. Waldinger, R. (1992, February). The ethnic enclave debate revisited. Paper presented at the meetings of the American Sociological Association, Pittsburgh, PA. Waldinger, R. (1993). The ethnic enclave debate revisited. Journal of Industrial and Labor Relations, 4 4 4 4 5 1 .
Copyright 2001 by Stephanie Bohon
Bibliography Waldinger, R. (1997a, March). Which way L.A.? Immigration and ethnic change in southern California. Paper presented at the California Policy Seminar, Los Angeles, CA. Waldinger, R. (1997b, September). Los Angeles and its immigrants. Paper presented at the Second International Metropolis Conference, Copenhagen, Denmark. Waldinger, R., & Bozoregmehr, M. (Eds.). (1996). Ethnic Los Angeles. New York: Russell Sage Foundation. Waldinger, R., Ward, R., & Aldrich, H. (1985). Ethnic business and occupational mobility in advanced societies. Sociology, 19, 586-597. Wallace, S. P. (1986). Central American and Mexican immigrant characteristics and economic incorporation in California. International Migration Review, 20, 657671. Waters, M. C., & Eschbach, K. (1995). Immigration and ethnic and racial inequality in the United States. Annual Review of Sociology, 21,419-446. Werbner, P. (1984). Business on trust: Pakistani entrepreneurship in the Manchester garment trade. In R. F. Ward and R. Jenkins (Eds.), Ethnic communities in business (pp. 166-188). Cambridge, England: Cambridge University Press. Wilson, K. L., & Martin, W. A. (1982). Ethnic enclaves: A comparison of the Cuban and black economies in Miami. American Journal of Sociology, 88, 135-160. Wilson, K. L., & Portes,A. (1980). Immigrant enclaves: An analysis of the labor market experience of Cubans in Miami. American Journal of Sociology, 86,295- 3 19. Winship, C., & Mare, R. B. (1992). Models of sample selection. Annual Review of Sociology, 18, 327-350. Woon, Y. F. (1987). The mode of refugee sponsorship and the socioeconomic adaptation of Vietnamese in Victoria: A three year perspective. In K. Chan & D. Indra (Eds). Uprooting, loss and adaptation: The resettlement of Indochinese rehgees in Canada. Ottawa, Canada: Canadian Public Health Association. Pp. 132-146. Zhou, M. (1997). Segmented assimilation: Issues, controversies, and recent research on the new second generation. International Migration Review, 31,975-1008. Zhou, M., and Logan, J. (1989). Returns on human capital in ethnic enclaves: New York City's Chinatown. American Sociological Review, 54,809-820.
Copyright 2001 by Stephanie Bohon