Authority and Control in Modern Industry: Theoretical and Empirical Perspectives (Routledge Studies in Business Organization and Networks, 10.)

AUTHORITY AND CONTROL IN MODERN INDUSTRY

Authority and Control in Modern Industry examines processes by which technological change has been assimilated into capitalist workplaces since the Industrial Revolution. Contributors present theoretical propositions, drawing arguments from neo-classical, Marxist, evolutionary and transaction cost economics, the sociology of Anthony Giddens and Max Weber, and network analysis. These are coupled with historical case studies, including: • • • • •

the growth of the factory system during the Industrial Revolution in Britain the German chemical industry and German family firms before World War One the American automobile industry in the inter-war period agricultural labour in the American South after World War Two the recent relationship between US medical specialists and hospitals.

This volume demonstrates the value of analysing worker deployment in modern capitalist firms as an interactive and iterative process involving both management and labour. Authority and Control in Modern Industry offers a rare breadth of analysis through the juxtaposition of sociological, economic and historical approaches and in the variety of ideological viewpoints represented. This work will become essential reading for researchers in organisational behaviour, labour economics, economic history and labour history. Paul L.Robertson is Associate Professor of Economics and Management at the University of New South Wales, Australia. His previous publications include Firms, Markets and Economic Change: a Dynamic Theory of Business Institutions (Routledge, 1995), co-authored with Richard N.Langlois.

ROUTLEDGE STUDIES IN BUSINESS ORGANIZATIONS AND NETWORKS 1 DEMOCRACY AND EFFICIENCY IN THE ECONOMIC ENTERPRISE Edited by Ugo Pagano and Robert Rowthorn 2 TOWARDS A COMPETENCE THEORY OF THE FIRM Edited by Nicolai J.Foss and Christian Knudsen 3 UNCERTAINTY AND ECONOMIC EVOLUTION Essays in Honour of Armen A.Alchian Edited by John R.Lott jr 4 THE END OF THE PROFESSIONS? The Restructuring of Professional Work Edited by Jane Broadbent, Michael Dietrich and Jennifer Roberts 5 SHOPFLOOR MATTERS Labor–management relations in 20th-century American manufacturing David Fairris 6 THE ORGANISATION OF THE FIRM International Business Perspectives Edited by Ram Mudambi and Martin Ricketts 7 ORGANIZING INDUSTRIAL ACTIVITIES ACROSS FIRM BOUNDARIES Anna Dubois 8 ECONOMIC ORGANISATION, CAPABILITIES AND COORDINATION Edited by Nicolai Foss and Brian J.Loasby 9 THE CHANGING BOUNDARIES OF THE FIRM Explaining evolving inter–firm relations Edited by Massimo G.Colombo 10 AUTHORITY AND CONTROL IN MODERN INDUSTRY Theoretical and empirical perspectives Edited by Paul L.Robertson

AUTHORITY AND CONTROL IN MODERN INDUSTRY Theoretical and empirical perspectives

Edited by Paul L.Robertson

London and New York

TO JAMES, EMILY AND SARAH

First published 1999 by Routledge 11 New Fetter Lane, London EC4P 4EE This edition published in the Taylor & Francis e-Library, 2003. Simultaneously published in the USA and Canada by Routledge 29 West 35th Street, New York, NY 10001 Editorial material and selection © 1999 Paul L.Robertson Individual chapters © 1999 individual contributors All rights reserved. No part of this book may be reprinted or reproduced or utilised 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. British Library Cataloguing in Publication Data A catalogue record for this book is available from the British Library Library of Congress Cataloging-in-Publication Data Authority and control in modern industry: theoretical and empirical perspectives/edited by Paul L.Robertson. p. cm. Includes bibliographical references and index. 1. Industrial management—History. 2. Industrial management—Case studies. 3. Labor—History. 4. Labor—Case studies. 5. Industrial organization—History. 6. Industrial organization—Case studies. 7. Industrial revolution. 8. Authority—Case studies. Robertson, Paul L. HD30.5.A95 1999 338.6’09–dc21 98–25902 CIP ISBN 0-203-43540-0 Master e-book ISBN

ISBN 0-203-74364-4 (Adobe eReader Format) ISBN 0-415-13212-6 (Print Edition)

CONTENTS

ix x xi xii

List of figures List of tables List of contributors Acknowledgements Introduction

1

PAUL L.ROBERTSON

Authority and the management of labour 1 Summary and conclusions 13 1 The rise of the factory system in Britain: efficiency or exploitation?

17

S.R.H.JONES

Introduction 17 Factories before the Industrial Revolution 21 Technology and the factory 23 Traditional explanations for the rise of the factory system 26 Factories, discipline and the ‘new left’ 29 New technology and exploitation 34 Transaction costs and the factory system 35 Transport improvements, transaction costs and technology 39 Conclusion 41 2 The coevolution of technology and organisation in the transition to the factory system RICHARD N.LANGLOIS

Introduction 45 What is a factory? 47 v

45

CONTENTS

Division of labour, routine and technology 50 ‘Durability’, fixed costs and supervision 55 Evolution, explanation and the inevitable 63 3 Class structures and the firm: the interplay of workplace and industrial relations in large capitalist enterprises

73

THOMAS WELSKOPP

The problem and the argument in outline 73 The limitations of conventional theory 75 ‘Micropolitics’ as macroreductionism 79 The concept of the ‘firm’ and critical social theory 82 ‘Class relations’ and ‘class structures’ as axes of the structuration of ‘employment relations’ 85 ‘Class relations’ in social systems of production: the ‘enterprise’ context 89 ‘Class relations’ as system-specific power resources in the configuration of workplace and intrafirm industrial relations 102 Concluding remarks: the ‘firm’ as an incompletely organised ‘political coalition’; and consequences for historical research 110 4 Knowledge, information and organisational structures

120

P.P.SAVIOTTI

Introduction 120 Knowledge: some basic considerations 121 Knowledge and information 125 Knowledge and the external environment of firms 128 The division of labour and coordination 130 Organisations, organisational boundaries and environmental change 133 Summary and conclusions 136 5 Technological change, transaction costs, and the industrial organisation of cotton production in the US South, 1950–1970

140

LEE J.ALSTON

The organisation of cotton production before the cotton picker, circa 1950 140 The adoption of the mechanical cotton picker, and changes in the industrial organisation of cotton production 143 Concluding remarks 151

vi

CONTENTS

6 The maintenance of professional authority: the case of physicians and hospitals in the United States 155 DEBORAH A.SAVAGE AND PAUL L.ROBERTSON

Introduction 155 The professional mode of production 157 Professionals and complementary institutions 163 The Joint Commission Model 165 Conclusion 169 7 Men and monotony: fraternalism as a managerial strategy at the Ford Motor Company

173

WAYNE A.LEWCHUK

The automobile industry as a male domain 175 Masculinity in the period before mass production 181 Masculinity and mass production at the Ford Motor Company 184 Conclusions 196 8 Management and labour in German chemical companies before World War One 203 SACHIO KAKU

Introduction 203 Development of the German coal-tar dyes tuffs industry 205 The managers of the German coal-tar dyestuffs companies 207 Terms of labour 209 Company welfare facilities 214 Conclusion 216 9 Buddenbrooks revisited: the firm and the entrepreneurial family in Germany during the nineteenth and early twentieth centuries DIRK SCHUMANN

The debate about the family firm 221 ‘Family firm’: a definition 224 Bavaria as a point of reference 225 The family firm in the early stage of industrialisation 226 The family firm in the advanced stage of industrialisation 228

vii

221

CONTENTS

Heirs 231 The entrepreneurial family and the middle classes 232 Summary 234 240

Index

viii

FIGURES

2.1 2.2 2.3 2.4 2.5 2.6 2.7 4.1 5.1 6.1 8.1

Explanatory alternatives 46 Organisational alternatives 49 Throughput and fixed costs 56 Wage and effort when marginal products observable 57 An exploitation explanation 59 Backward-bending effort supply 60 Discontinuous indifference curve 61 Overlap of knowledge base 135 Increase in the relative use of wage contracts with mechanisation 146 The internal and external relationships of a profession 160 Managerial hierarchy within the coal-tar dyestuffs factory 210

ix

TABLES

1.1 3.1 5.1 5.2 5.3 5.4 7.1 7.2 7.3 8.1 8.2 8.3

Simple efficiency properties of the putting out and factory systems Class structures and the ‘firm’ as a hierarchical and contextualist concept Percentage of upland cotton mechanically harvested, 1949–1972 Tenants, sharecroppers and real wage expenditure Percentage of gins with cleaners and dryers, 1945 Percentage of gins with daily capacity greater than 36 bales, 1945 Employment of women in Detroit automobile plants, 1941 and 1942 Female employment at Ford Motor Company, 1912–1941 Hourly earnings in the American motor vehicle industry Issuing of stocks and debentures of the German coal-tar dyes tuffs companies Wage rate table of the Hoechst Company in 1912 The working hours of the Bayer Company and BASF

x

36 86 143 145 148 149 178 178 187 208 212 213

CONTRIBUTORS

Paul L.Robertson, Associate Professor of Economics and Management, University of New South Wales, Australia. Lee J.Alston, Professor of Economics and Director, Center for International Business Education and Research, University of Illinois at UrbanaChampaign. S.R.H.Jones, Senior Lecturer in Economics, University of Dundee. Sachio Kaku, Professor of Economics, Kyushu University, Japan. Richard N.Langlois, Professor of Economics, University of Connecticut. Wayne A.Lewchuk, Professor of Economics, McMaster University, Canada. Deborah A.Savage, Assistant Professor of Economics and Finance, Southern Connecticut State University and Visiting Fellow, Institution for Social and Policy Studies, Yale University P.P.Saviotti, Professor of Economics, INRA-SERD, Université Pierre Mendes France, Grenoble. Dirk Schumann, Fakultät für Geschichtswissenschaft und Philosophie, Universität Bielefeld, Germany. Thomas Welskopp, Friedrich-Meinecke-Institut, Freie Universität Berlin, Germany.

xi

ACKNOWLEDGEMENTS

I wish to thank Lee J.Alston and Richard N.Langlois for their efforts throughout the activities that ultimately produced this volume, and most particularly for their work during the organisation of the session of the International Economic History Association Conference in Milan in 1994 at which most of the papers were originally presented. Thanks are also due to the University of Manchester for support to myself in the form of a Hallsworth Research Fellowship during the initial period of organisation. Finally, I wish to thank Sarah Robertson for her help in preparing the manuscript for publication. Three of the chapters are reprinted in slightly different versions with the permission of their copyright holders. Chapter 1 appeared originally in Maurice W.Kirby and Mary B.Rose (eds) Enterprise in Modern Britain from the Eighteenth to the Twentieth Century, London: Routledge, 1994, 31–60. Chapter 6 appeared in Business and Economic History 26, 2, 1997:662–75. Chapter 7 appeared in the Journal of Economic History 53, 4, 1993:824–56.

xii

INTRODUCTION Paul L.Robertson

With two exceptions, the chapters in this book were presented at a session of the International Economic History Association that was held in Milan in August 1994. The aim of the session, which was organised by Lee J.Alston, Richard N.Langlois and myself, was to present theoretical arguments and case studies on the relationship between organisational forms and the deployment of labour in modern capitalist economies. One of the main purposes of the session was to juxtapose views from different disciplines, including economics and sociology as well as economic history, in order to explore possible areas of agreement and dispute. Although such a broad terrain could hardly be adequately surveyed in only a few papers, the participants have provided thought-provoking discussions of several modern schools of organisational thought, along with a range of examples drawn from the history of Western Europe and the USA since the Industrial Revolution. Among the schools discussed are neo-classical, Marxist, evolutionary and transaction cost economics, the sociology of Anthony Giddens and Max Weber, and network analysis. Historical examples include factory-based labour in Britain during the Industrial Revolution, in the German chemical industry in the late nineteenth and early twentieth centuries, and in the American automobile industry in the inter-war period; agricultural labour in the American south after World War Two; and American medical specialists.

Authority and the management of labour The organisation of the workforce has been central to success in both civil and military endeavours since people first began to band together. Over the millennia, the coordination of labour has been undertaken in diverse activities ranging from the building of the Pyramids to modern industrial corporations, from subsistence family farms to latifundia, and in the transport and banking sectors and the armed forces. A variety of institutional forms has also been employed at various levels of aggregation, including markets of greater or 1

PAUL L.ROBERTSON

lesser degrees of competition, networks, trade unions and employer federations, and more elaborate and far-reaching forms such as feudalism. In most analyses, the effectiveness and efficiency of these authority relationships have depended primarily on two factors: the distribution of power, both within and between organisations; and the technologies employed. As both of these have changed, societies have also undergone changes in the ways in which authority has been distributed and implemented. As a rule, these changes in authority patterns have been worked out by trial-anderror methods, after existing power or technology relationships have been disturbed. Since change generates both winners and losers, it is hardly surprising that the transition to new authority patterns has often been contentious. Not only has there been inertia as the representatives of older orders resist change, but quarrels have arisen over the distribution of resources and rewards under the new orders. In some cases, these have centred on economic issues such as wages, but other, more broadly social or political issues may also have been involved as participants contended for limited time or space, for access to educational or religious resources, or, to put it baldly, for the right to control the destinies of themselves and others. But distributional issues are only one part of the story; matters of growth and of economic and technical efficiency have also been important in forming patterns of authority. Ways of implementing and using new technologies have also been worked out by trial and error. In some cases, the practices which govern authority patterns and the deployment of labour have been functions of particular technologies which have allowed very little leeway because of their very nature. In such cases, misused equipment could impair the safety of workers, people in the surrounding community and the equipment itself. Generally, however, more latitude has been allowed, and patterns of use might vary depending on factor costs, the skill levels of the workforce, the division of labour, the association between a given technology and complementary technologies, and numerous other factors. The use of new technologies and, as time passes, of existing technologies could be transformed, as learning-by-doing and learning-by-using progress and as environmental changes alter the relative prices of inputs and outputs. Rarely, if ever, have power relations and efficiency factors been independent of each other. Changes in each have usually led to changes in the other, resulting in modifications in overall patterns of authority. The chapters that follow analyse these interrelationships from a range of viewpoints. Technology and authority in the Industrial Revolution: Marglin and his critics In 1974, Stephen A.Marglin published an influential article entitled ‘What do bosses do? The origins and functions of hierarchy in capitalist production’, in which he challenged the view that the adoption of factories from the middle 2

INTRODUCTION

of the eighteenth century was largely determined by technological factors that increased the cost and size of equipment and required closer coordination among workers than the ‘putting-out’ system provided. In this and subsequent works, Marglin (1974, 1984, 1991) contended that the increased division of labour associated with the factory system, both during the Industrial Revolution and subsequently, was primarily a device used by the owners of capital to control the labour force and ensure that few workers could ever gain the broad knowledge needed to set up in competition with their employers. According to Marglin, the machines associated with the early stages of industrialisation could have been operated equally efficiently in dispersed workers’ cottages or small workshops, in the way that earlier technologies had been. By centralising production, however, ‘bosses’ were able to curtail leakages such as embezzlement, and to impose a fragmentation of work patterns that made it very difficult for hard-pressed workers to learn a broad range of tasks or even to appreciate where their individual efforts fitted into a larger production process. In short, the early development of factories was not the outcome of changing technology, but resulted from a conspiracy among capitalists to maintain their control over the factors of production and to exploit their workers, thereby enriching themselves. Subsequent technological changes, which were admittedly of a far greater magnitude and more dependent on factory-based centralisation and coordination, were not autonomous but were the result of patterns of organisation established during the Industrial Revolution (Marglin 1991). Like Piore and Sabel (1984) and Sabel and Zeitlin (1985), Marglin believes that a reliance on large-scale factories was not inevitable and that smaller units of production and different authority patterns could have led to greater economic justice without significant sacrifices in efficiency. Marglin’s New Left view that increased discipline was exerted under the factory system is not heretical in itself. Pollard (1965), for example, has shown how methods of controlling labour evolved in the course of the Industrial Revolution. Moreover, the exercise of authority by capitalists to ‘exploit’ (i.e. to profit to as large an extent as possible from the labour of) workers is entirely consistent with the profit-maximising assumptions used by classical and neo-classical economists. Marglin’s principal departure from earlier accounts generated by both orthodox and Marxist economists is his denial that technological change was an important force behind the centralisation of production. In this, he challenged the views of mainstream economic historians such as Mantoux (1961) and Landes (1969), and has in turn been severely criticised by Landes (1986), who has restated the importance of technological change to the growth of the factory system while conceding a number of Marglin’s other points. In Chapter 1 of the present volume, S.R.H.Jones reassesses the question of the role of technological change during the Industrial Revolution. After a careful outline of the nature of factories in eighteenth-century Britain, he 3

PAUL L.ROBERTSON

shows that the use of some of the most important new types of cotton textile machinery was not feasible in domestic workshops. Increases in productivity in this important industry may therefore be traced, at least in part, to technological change in a factory setting. The use of authority to provide coordination and to protect large fixed investments followed from the growth of large, relatively capital-intensive means of production. Jones also shows, however, that the factory system was adopted as a means of control in some cases in which the prevailing technology did not dictate centralised production. He concludes that ‘it is clear that whatever present day economists may think, to the nineteenth century world at large the factory system was synonymous with exploitation. What is less clear is that exploitation was its primum mobile.’ In Chapter 2, Richard N.Langlois explores the Marglin thesis from a different angle. He argues that the growth of the factory system stemmed largely from growth in the markets for manufactured goods—primarily textiles—in the early stages of industrialisation. This increase in the extent of the market validated the use of new technologies as embodied in various machines and the resulting changes in the division of labour and in authority relations. Improved efficiency was thus both a goal and an outcome of the adoption of the factory system. Following Clark (1994) and Lazonick (1990), Langlois also explores the question of exploitation in terms of the ‘effort norms’ of workers, i.e. the amount of work that they believed they should undertake. When there were backward-bending supply curves of labour, workers reduced their effort in response to higher wages because, at least on a conscious level, they were willing to settle for customarily-determined standards of living. But Langlois agrees with Clark that, in reality, workers desired the higher real incomes that factory employment brought but lacked the selfdiscipline to work as hard as the system required. Thus they submitted to the authority of bosses to provide the discipline that they would not impose on themselves. Langlois concludes that the factory system was a product of both technological and control factors, but that, in general, the workers were not exploited. Structuration, worker resistance, and the interaction of workers and employers The interaction between management, in this case mill owners, and their workers needs to be considered explicitly in order to achieve a convincing view of the relative contributions of technological factors and a desire to establish and maintain control over the workforce during the adoption of the factory system, at the time of the Industrial Revolution and subsequently. While the balance of power between individual factory workers and their employers was generally tilted heavily in favour of the latter, workers nevertheless frequently retained an ability to resist the authority of their wealthier employers. How large the scope for resistance was depended on 4

INTRODUCTION

many conditions, such as the skill level of the worker, the tightness of the labour market, local laws and customs, and the temperaments of workers and employers both. Even in the absence of trade unions, informal organisations established on the shop floor could lead to ‘soldiering’ or ‘ca’canny’ in which individuals restricted their effort to conform to norms which they and their colleagues had laid down explicitly or tacitly. When workers were formally unionised, their ability to resist conditions imposed by their employers could increase significantly, although it remained contingent on environmental conditions. As Robertson and Alston (1992) have shown, for example, the successful introduction of new technologies by employers depends on power relationships within a firm. Workers who are hard to replace, or who are supported by other workers who occupy strategic positions, may successfully block or control innovation, while those who less well-placed may be forced to bear the full brunt of adjustment on the job or, in more extreme cases, accept their own dismissal. Through his concept of ‘structuration’, Anthony Giddens has explored the relations between individuals and larger societal units.1 Giddens contends that individual and societal behaviour are not products of either individual activities or the structure of society. Individuals do not have unlimited scope to exercise free will, but neither are their activities totally determined by the structure of their social or economic environments. Instead, he posits, there is a ‘duality of structure’ in which The constitution of agents and structures are not two independently given sets of phenomena, a dualism, but represent a duality. According to the notion of the duality of structure, the structural properties of social systems are both medium and outcome of the practices they recursively organise. (Giddens 1984:25) In this schema, social systems enable as well as constrain human action because they provide the rules that people use when acting. As Held and Thompson (1989:4) put it, Giddens believes that we should cease to conceive of ‘structure’ as a kind of framework, like the girders of a building or the skeleton of a body, and…we should conceptualize it instead as the ‘rules and resources’ which are implemented in interaction. In interacting with one another, individuals draw on the rules and resources which comprise structure, in much the same way as an individual draws on the rules of grammar in uttering a well-formed speech act. Like the rules of grammar, structure is both ‘enabling’ and ‘constraining’: it enables us to act as well as delimiting the courses of possible action. 5

PAUL L.ROBERTSON

Giddens (1982) explicitly discusses power and authority relations in the works of two important writers who, he states, have denied that there is scope for independent action by workers in capitalist firms. The first, Max Weber, contended that modern bureaucratic practices constitute a ‘steel-hard cage’ that denies autonomy and spontaneity of behaviour to those who must work within it. Giddens argues, however, that in reality a ‘formal codification of procedures rarely conforms to actual practice’, since even a fully-specified set of rules is subject to interpretation (1982:37). In addition, he supports Crozier’s observation that the social and physical distance between subordinates and supervisors gives workers ‘space’ to elude control. The second writer, Harry Braverman (1974), wrote from a Marxist stance. Giddens (1982:40) accuses Braverman of having a ‘disinclination to treat workers as knowledgeable and capable agents’. According to Giddens, even in Taylorist situations of the sort discussed by Braverman, workers retain a significant degree of control of their work and are in a position to offer substantial resistance to their employers. To Giddens (1982:31), ‘the most seemingly “powerless” individuals are able to mobilise resources whereby they carve out “spaces of control” in respect of their day-to-day lives and in respect of the activities of the more powerful.’ In Giddens 1984, he writes: ‘… [A]ll forms of dependence offer some resources whereby those who are subordinate can influence the activities of their superiors’ (p. 16). Thomas Welskopp draws on the works of Giddens and others in Chapter 3 of the present volume to construct a theoretical framework for analysing the interaction of workers and firms. Welskopp contends that business history and labour history should be treated as linked fields rather than, as is frequently the case, two separate areas of study that focus on the respective activities of management and workers without giving adequate weight to the environmental circumstances and goals of the other. In his view, business history has been devoted too heavily to ‘economising’ concepts such as attempts at profitmaximisation by management, while ‘sociologising’ concepts involving classbased activities have exerted excessive influence on labour history. Like Giddens, Welskopp emphasises that workers do have power in large capitalist firms, power that they frequently use to achieve goals of a maximising nature in relation to wages and other types of returns for their labour. Management, on the other hand, is not free to pursue profit-maximising goals in an unimpeded way but, as indicated by Robertson and Alston (1992) and others, must consider the views and needs of their work-forces when administering their enterprises. According to Welskopp, his model provides ‘an avenue of research which directs attention to the interrelations of spheres. By “economizing” labor history and “sociologizing” business history, it acknowledges that both disciplines actually deal with the same sorts of problems.’ By extension, a similar convergence of analytical approaches could enrich contemporary studies of management and industrial relations. 6

INTRODUCTION

The transaction costs of authority and control Transaction and agency costs may affect the exercise of authority through several channels. One occurs when employers must devote resources to monitoring the activities of their labour force to ensure that work is performed as directed. ‘The metering problem’ (Alchian and Demsetz 1972) arises when it is difficult to measure particular inputs or outputs. In the case of workers, for instance, this might occur because one member of a team does not exert full effort, but chooses to be a ‘free rider’ who exploits the efforts of other team members. Because of the joint nature of the work, common and easyto-implement solutions such as payment by the piece will not eliminate the problem. If the employer is not in a position to measure inputs easily, such shirking could go undetected unless special, and possibly expensive, efforts are devoted to the monitoring and control of the performance of each team member. Another possible problem is the embezzlement of inputs by workers because their employer finds it difficult to measure outputs accurately. One example, already alluded to, was in the domestic handloom weaving sector where weavers could stretch the completed cloth, providing a thinner product than intended and keeping the yarn that was saved for themselves. If the finished product did not differ in an obvious way from what had been contracted for, the merchant could need to devote substantial resources to monitoring for possible fraud. The desirability of providing expensive supervision may also be intensified by the adoption of a capital-intensive production process in which the machinery is not owned by those who operate it. In such cases, which epitomise capitalism, the owner of the machinery may have an incentive to use the equipment intensively, especially when fuel or other inputs are going to be expended whenever the machines are turned on even if they are not being used actively, or when the rate of technological change is high, increasing the desirability of frequent changes of model. The use of piece rates may be one way of dealing with the problem, but only if the workers are willing to work as hard as the owner of the equipment wants them to. Ultimately, therefore, it may be necessary for the owners to bear the expense of closely supervising their workforce if they are to elicit enough effort to gain full value from the machinery. A third possibility is ‘hold-up’ arising from co-specificity between the workers and an organisation or its equipment (Williamson 1985; Putterman and Kroszner 1996). When there is only a limited number of people who have the skills to operate vital equipment or who understand the tacit procedures of the firm, they may be in a position, at least in the short run, to exact higher wages than they could if their skills were in greater supply. One common way of dealing with asset-specificity problems, namely vertical integration (Williamson 1985), is not practical since workers cannot be owned by their employers. Because firms must take care to ensure employment

7

PAUL L.ROBERTSON

relations are not severed in an inconvenient and expensive way, ‘[t]ransaction cost economics maintains that governance structures must be crafted more carefully as the degree of human asset specificity increases’ (Williamson 1985:243). Jones and Langlois examine the transaction costs arising from the growth of capital-intensive production processes in the Industrial Revolution. Jones also discusses in some detail the arguments of Williamson and Szostak concerning the role of a broad range of transaction costs in promoting the shift to the factory system. While he does find limited support for the thesis that transaction costs were an important consideration, Jones contends that Williamson’s argument is not persuasive because it does not ‘reflect what generally occurred in late eighteenth-and early nineteenth-century England’. Levels of transaction costs have been important in areas outside manufacturing. In Chapter 5, Lee J.Alston employs transaction cost theory to analyse the effects of mechanisation in cotton growing in the southern part of the United States (i) on the costs of labour contract bargaining between workers and cotton growers and (ii) on the boundary between farms and markets. He finds that the costs of monitoring labour were lowered in two ways by mechanisation. First, the opportunity costs of shirking by the remaining workers became higher as thousands of farm labourers were displaced by machinery, thus increasing the chances and costs of losing their own jobs in a looser labour market. Second, by standardising the value of the marginal productivity of labour, mechanisation made it easier for growers to assess hard-to-monitor labour inputs by looking at the more easilymeasurable output of labour. The evidence he presents is consistent with his hypothesis that increased mechanisation would generate a decrease in paternalism by employers and an increase in the relative use of wage contracts. Although Alston shows the value of transaction cost theory in analysing the role of authority and control within enterprises, the chapter by Jones suggests that substantial care must be taken to avoid substituting arguments based on the logic of transaction cost considerations for a careful examination of the empirical record. Networks, knowledge and ‘bossless’ enterprises The effects of change on coordination between enterprises may be as important as those on coordination within a single enterprise. Indeed, at times the boundaries of a particular organisation may be so blurred that it is hard to determine which employees are within the organisation and which belong to other organisations. One way of dealing with this blurredness is to look at networks both within and across organisations, rather than at organisations in isolation, but this does not necessarily clarify matters substantially. As Nohria (1992:3) has written,2 ‘Anyone reading through what purports to be network literature will readily perceive the analogy between it and a 8

INTRODUCTION

“terminological jungle in which any newcomer may plant a tree.”’ In view of this confusion, it is best for present purposes to adopt the definition proposed by one of the authors in this volume: ‘A network is an economic organization that accomplishes the exchange of capital, products, and/or knowledge without explicit equity investment or “ownership” ’ (Savage 1994:133). The role of authority and control within a network varies, but in some of the most interesting cases networks operate as ‘bossless’ organisations that are distinct from both markets and hierarchies. In his discussion of the use of teams in projects, Frame (1995) outlines the characteristics of ‘egoless teams’, which are in many ways similar to network forms such as professions. The members of an egoless team are for most, if not all, purposes equals in a hierarchical sense. Leadership is provided on an ad hoc basis and shared among the members. Although individual members possess different information and knowledge, both their existing stocks and new discoveries are shared without charging amounts much above the costs of communication. In some instances, the team may act collectively, but in others the members will act individually, in which case credit and blame are also apportioned individually. According to Savage (1994), professions are examples of networks with attributes similar to those of egoless teams. Professions, as she sees them, are not conspiracies designed to restrict the supply of services and raise costs to the public. Instead, they exist to foster the generation and exchange of knowledge and information as well as to certify competence. In contrast to typical organisations, however, there are no bosses in the sense of individuals who are able to hire and fire and to direct the work of others. Power and authority within networks are held and exerted collectively as networks operate through committees and other bodies chosen by the membership. Networks also accredit courses to train potential members and provide workshops, seminars and journals to keep existing members up to date. Work itself, though, is generally performed by individual members of the network who are not supervised by representatives of the network as a whole and are only required to account for their activities when something extraordinary occurs. Their major responsibility, in fact, is to their clients, and they trade primarily on their individual reputations and only secondarily on the reputation of the professional network. The occupational independence of professionals, however, is often hampered by their need to draw resources from other organisations in which authority is more heavily centralised. In what Frame (1995) calls a ‘surgical team structure’,3 in the course of using their own skills, professionals must operate within a more highly-structured environment that supplies support staff, physical facilities or other resources that individual professionals need but cannot economically provide themselves. This often causes problems because power and authority are divided between the professionals, who are paid to exercise their own judgement, and the 9

PAUL L.ROBERTSON

administrators, who are responsible for the coordination and maintenance of the support staff and facilities to ensure that, as far as is possible, ample provision is made for the needs of all of the professionals in the network who draw on those resources. In Chapter 6, Savage and Robertson look at the problems in maintaining a professional network structure that have arisen because of the need of surgeons and other medical specialists to coordinate their activities with those of hospitals in the United States. In recent decades, as a result of changes brought about by the increased leverage gained by funding bodies including the Federal Government and managed-care companies, the balance of power and authority between specialists and hospitals has shifted drastically. In order to attract and retain contracts with the government and managed-care companies, hospitals are engaging increasingly in economic credentialing under which they bar specialists who routinely use more resources than the funding bodies feel is necessary. This in turn severely undermines the professional authority of the specialists, whose medical judgement must now be tempered by considerations of resource usage as well as by expected outcomes for patients. Moreover, the professional networks within medicine may find it harder in the future to perform their training and monitoring roles because the new balance of power, between hospitals and insurers on the one hand, and specialists on the other, has eroded some of the premises on which the exchange of information within the networks has been based. Pier Paolo Saviotti investigates networks from a different angle in Chapter 4, in which he contends that the use of new technologies may give rise to coordination issues between organisations. He notes that rates of technological change have increased in recent years and are likely to remain high. Changing technologies are accompanied by radical changes in knowledge that may be difficult for hierarchical organisations to assimilate because they generate noise, require redundancy of information to compensate for that noise, and also reduce the intelligibility of the information that a firm acquires from its environment. Fast and radical changes in knowledge, therefore, degrade existing knowledge bases and organisational structures, and this in turn requires a redefinition of the division of labour within and between organisations, and generates coordination problems. Although hierarchical structures are often efficient at managing established knowledge bases, they may be unable to cope with high rates of change. New forms of organisations, with strong network components such as inter-institutional collaborative agreements (IICAs), are often better suited to deal with radical changes in knowledge. But, since IICAs may be less well-adapted to stable environments, it is possible that new technologies introduced by network organisations such as IICAs may be taken over by hierarchical organisations as they mature.

10

INTRODUCTION

Authority and control and the gender division of labour Employers have implemented various stratagems to deal with alienation among factory workers and to convince them to accept changes that might not be in the workers’ own interests. One approach taken by management has been to encourage the workforce to accept gender norms that could then be used to elicit behaviour that is beneficial to the firms. Although gender norms since the Industrial Revolution have not been unambiguous, they have frequently led to a pronounced stratification of the workforce. In nineteenth-century Britain, for example, women were generally paid less than men, but this did not lead to a wholesale displacement of male labour, because women were also denied by custom, and sometimes by law, the right to enter certain occupations that they were physically able to follow. In the latter set of occupations, many of which were termed ‘skilled’, higherpaid work was reserved for males who were presumed to be family ‘bread winners’ and therefore to need larger incomes. In a similar fashion, until after World War Two, female Federal public servants in Australia were required to resign their positions upon marriage on the assumption that they no longer needed to be employed. Wayne Lewchuk investigates the effects of gender norms on the employment practices of the Ford Motor Company in Chapter 7. As he puts it, ‘[f]raternalism replaced paternalism as a managerial strategy to convert labour time into effort, a strategy that limited employment opportunities for women.’ As he shows, however, male employees were also severely affect by Ford’s policy of portraying factory labour as ‘man’s work’. The logic of assembly line work, as well as the large fixed investments that it entailed, made Ford insistent on maintaining a continuous flow of effort from labourers who could be expected to be bored or antagonistic because of the repetitive nature of their jobs. Lewchuk contends that, in addition to providing good wages and close supervision, Ford sought to encourage worker cooperation by constructing a new image of the roles of men in the factory, in the family, and in the outside world. Men were told that, by working hard, they would enhance their status in the world at large. Women were deliberately excluded from factory work as part of the programme, since their presence would have undermined Ford’s attempt to use fraternalism as a way of eliciting increased effort from male labourers. Authority in Weberian and Chandlerian firms At the turn of the present century, Max Weber (1948, 1971) argued that the complexity of modern capitalist enterprises has compelled managers to ‘rationalise’ their operations by developing elaborate bureaucracies governed by systems of rules designed to eliminate ambiguity and guide workers when choices of action are necessary. More recently, Alfred D.Chandler

11

PAUL L.ROBERTSON

(1990) has again pointed to the value of bureaucracies in governing large industrial concerns. Chandler argues that At the core of this dynamic [of economic growth] were the organizational capabilities of the enterprise as a unified whole…. [O]nly if these facilities and skills [of manufacturing and research] were carefully coordinated and integrated could the enterprise achieve the economies of scale and scope that were needed to compete in national and international markets and continue to grow. (Chandler 1990:594) Although Giddens (1982) and others have criticised his view of bureaucracy as a ‘steel-hard cage’, it is clear that Weber was correct to emphasise the attempts of firms to systematise their operations through the generation of rules and procedures. In Chapter 8, Sachio Kaku illustrates the importance of this process in several large German chemical firms before World War One. He discusses how, as they grew in size, the labour forces of these firms came to be governed increasingly by rules covering hours of work, wages, eligibility for sickness benefits and housing, and other aspects of their employment relationships. Significantly, these rules also included explicit chains of authority that specified decision-making routes when questions arose. Chandler (1990) has also examined the respective roles of family members and salaried managers in German, US and British firms in the late nineteenth and early twentieth centuries.4 Following earlier writers such as Landes (1949, 1951), Chandler has questioned the ability of tightly-held family firms to grow successfully in capital-intensive industries with substantial economies of scale. He believes that such family firms are not only excessively riskaverse, but that it is improbable that successive generations of family members would possess the temperament and talent necessary to deal with growth and technological change. In particular, he argues that large British firms after 1870 suffered in comparison with firms in Germany and the United States because of the relative reluctance of the former to draw on the talents of salaried managers whose time horizons, technical knowledge and views on investment were better suited to promoting the success of enterprises in industries such as chemicals and electrical products that were among the dominant growth areas. Sachio Kaku in Chapter 8, and Dirk Schumann in Chapter 9, test Chandler’s views by investigating the respective roles of owners and salaried managers in Germany at the end of the nineteenth century. Schumann finds that family firms did not fade away but in many cases continued to prosper. Not only did new small-and medium-sized family firms continue to flow into the economy, but, as industrialisation proceeded, family members proved as adept as their salaried counterparts at building sophisticated managerial 12

INTRODUCTION

structures to govern large firms. For large German chemical firms, Kaku argues that salaried managers were common at the highest levels (the executive and supervisory boards) by 1914, but both he and Schumann note that the personal connections of family members were of importance to their firms even when managerial authority had been delegated to outsiders to a significant degree. Thus, while Weberian bureaucracies were established and salaried professionals played a greater role as time passed, bureaucracy did not provide a ‘steel-hard cage’ and families remained important influences in large segments of the German economy throughout the period before World War One.

Summary and conclusions The chapters collected here do not attempt to give a comprehensive view of managing and motivating labour in capitalist firms. Nevertheless, they do present some significant common themes. One is the importance of the horizontal and vertical divisions of labour. The pervasiveness of horizontal stratification is not surprising in a book devoted to authority and control. Scholars of virtually any school, including those who view the firm as a ‘nexus of contracts’, agree that someone is ‘in charge’ in most organisations. Even when workers are in a position to disregard commands, in the sense that they can offer their services elsewhere if they do not want to perform the tasks assigned to them, they are not free to do whatever they want without incurring penalties.5 To use Giddens’s (1982) term, the ‘space of control’ available to workers is limited. On the other hand, as Welskopp argues forcefully in Chapter 3, it does exist and it does place limitations on the scope of managerial action, a point illustrated in several of the chapters. The vertical division of labour also plays a substantial role in this collection, particularly in the chapters by Jones, Langlois and Lewchuk. In addition, Saviotti and Savage and Robertson illustrate how a vertical division of labour between organisations can affect and be affected by managerial procedures. Another common theme is the way in which technological change can influence the vertical division of labour. Since the middle of the eighteenth century, changes in the configuration of workers and machinery have resulted principally from alterations to the latter. While the extent of the market has helped to dictate the division of labour, both within and between organisations (Rosenberg 1963), changes in technology have first formed and then reformed the relations between workers and equipment as the scope of operation of individual types of machinery has become either more specialised or more integrated. These movements towards specialisation and integration seem to have been frequently driven by the economic logic of deploying particular technologies combined with the respective degrees of power of employers and workers.

13

PAUL L.ROBERTSON

There are also areas of disagreement between authors. Among economists, Lewchuk stakes out a position that is closer than that of other contributors to the approach of Marglin. Jones is sceptical about the use of transaction cost economics to explain historical events, while Alston demonstrates its value when adequate empirical back-up is provided. Langlois and, more strongly, Saviotti use evolutionary approaches, and Savage and Robertson venture into the new areas of resource-based and network economics. Both sociological and economic approaches are applied to authority and the division of labour. But, while all of these variations are shown to be of value, the basic assumptions that underlie them are often quite different. One major lesson to be drawn from this diversity is that compartmentalism of ideas can lead to an unnecessary narrowing of methodologies. As Welskopp has pointed out, in the fields of business and labour history both economic and sociological models are needed to provide realistic analyses. Much the same is true of strategic management and organisational behaviour, which seems to suggest that the attention of researchers could usefully be devoted to building models that combined assumptions from both areas rather than abstractly and rigidly eliminating one or the other from consideration.

Notes 1 Giddens has published an enormous amount on structuration since the 1970s. In The Constitution of Society (1984), he provides a rigorous treatment of many aspects of his work. A shorter account of facets of structuration that are of interest in the present context is given in Giddens (1982). Bryant and Jary (1991), Craib (1992), Clark et al. (1990) and Held and Thompson (1989) include a wide range of critical evaluations of Giddens’s work. 2 The literature on networks is huge and expanding rapidly. The collection edited by Nohria and Eccles (1992) is a good place to begin looking for concepts. 3 The surgical team approach is similar to what Mintzberg (1979) has termed the ‘professional bureaucracy’, in which a decentralised organisation responsible for training and the maintenance of standards is combined with a hierarchical support organisation. 4 The views of William Lazonick (1990, 1991) are similar to those of Chandler concerning the importance of hired managers in internalising and ‘rationalising’ activities that might otherwise be accomplished through market transactions. 5 Alchian and Demsetz (1972) claim that the relationship between a firm and its workers is the same as that between an individual and her grocer, in that in both cases the parties are continually engaged in the renegotiation of contracts on mutually-acceptable terms. This does not mean that employees can do as they please on the job, however, any more than one’s grocer is free to sell to customers whatever he wants. In both cases there is a primary decision-maker (the firm’s designated manager or the customer) who establishes the framework in which the transaction takes place.

14

INTRODUCTION

References Alchian, A. and Demsetz, H. (1972) ‘Production, Information costs, and Economic Organization’, American Economic Review 62:777–95. Braverman, H. (1974) Labor and Monopoly Capitalism: The Degradation of Work in the Twentieth Century, New York: Monthly Review Press. Bryant, C.G.A. and Jary, D. (eds) (1991) Giddens’ Theory of Structuration: A Critical Appreciation, London: Routledge. Chandler, A.D., Jr. (1990) Scale and Scope: The Dynamics of Industrial Capitalism, Cambridge, Ma: The Belknap Press of Harvard University Press. Clark, G. (1994) ‘Factory Discipline’, Journal of Economic History 54, 1:128–163. Clark, J., Mogdil, C. and Mogdil, S. (eds) (1990) Anthony Giddens: Consensus and Controversy, London: Falmer. Craib, I. (1992) Anthony Giddens, London: Routledge. Frame, J.D. (1995) Managing Projects in Organizations: How to Make the Best Use of Time, Techniques, and People, 2nd edn, San Francisco: Jossey-Bass. Giddens, A. (1982) ‘Power, the Dialectic of Control and Class Structuration’, in A. Giddens and G.MacKenzie (eds) Social Class and the Division of Labour: Essays in Honour of Ilya Neustadt, Cambridge: Cambridge University Press, 29–45. ——(1984) The Constitution of Society: Outline of the Theory of Structuration, Cambridge: Polity Press. Held, D. and Thompson, J.B. (eds) (1989) Social Theory of Modern Societies: Anthony Giddens and his Critics, Cambridge: Cambridge University Press. Landes, D.S. (1949) ‘French Entrepreneurship and Industrial Growth in the Nineteenth Century’, Journal of Economic History 9:45–61. ——(1951) ‘French Business and Businessmen in Social and Cultural Analysis’, in E.M.Earl (ed.) Modern France, Princeton: Princeton University Press, 334–53. ——(1969) The Unbound Prometheus: Technological Change and Industrial Development in Western Europe, 1750 to the Present, Cambridge: Cambridge University Press. (1986) ‘What Do Bosses Really Do?’, Journal of Economic History 46, 3: 585–623. Lazonick, W. (1990) Competitive Advantage on the Shop Floor, Cambridge, Ma.: Harvard University Press. ——(1991) Business Organization and the Myth of the Market Economy, New York: Cambridge University Press. Mantoux, P. (1961) The Industrial Revolution in the Eighteenth Century, London: Cape. Marglin, S.A. (1974) ‘What Do Bosses Do? The Origins and Functions of Hierarchy in Capitalist Production’, Review of Radical Political Economy 6 (Summer): 33–60. ——(1984) ‘Knowledge and Power’, in F.H.Stephen (ed.) Firms, Organization and Labour: Approaches to the Economics of Work Organization, London: Macmillan, 146–64. (1991) ‘Understanding Capitalism: Control versus Efficiency’, in B. Gustafsson (ed.) Power and Economic Institutions: Reinterpretations in Economic History, Aldershot: Edward Elgar, 225–52. Mintzberg, H. (1979) The Structuring of Organizations: A Synthesis of the Research, Englewood Cliffs, NJ: Prentice-Hall. Nohria, N. (1992) ‘Introduction: Is a Network Perspective a Useful Way of Studying Organizations?’, in N.Nohria and R.G.Eccles (eds) Networks and Organizations: Structure, Form, and Action, Boston: Harvard Business School Press, 1–22. Nohria, N. and Eccles, R.G. (eds) (1992) Networks and Organizations: Structure, Form, and Action, Boston: Harvard Business School Press. Piore, M.J. and Sabel, C.F. (1984) The Second Industrial Divide, New York: Basic 15

PAUL L.ROBERTSON

Books. Pollard, S. (1965) The Genesis of Modern Management, Cambridge, Ma.: Harvard University Press. Putterman, L. and Kroszner, R.S. (1996) The Economic Nature of the Firm: A Reader, 2nd edn, Cambridge: Cambridge University Press. Robertson, P.L. and Alston, L.J. (1992) ‘Technological Choice and the Organization of Work in Capitalist Firms’, Economic History Review 45, 2:330–49. Rosenberg, N. (1963) ‘Technological Change in the Machine Tool Industry, 1840– 1910’, Journal of Economic History 23, 2:414–43. Sabel, C.F. and Zeitlin, J. (1985) ‘Historical Alternatives to Mass Production: Politics, Markets, and Technology in Nineteenth-Century Industrialization’, Past and Present 108:133–76. Savage, D.A. (1994) ‘The Professions in Theory and History: The case of Pharmacy’, Business and Economic History 23, 2:129–60. Weber, M. (1948) ‘Bureaucracy’, in H.H.Gerth and C.W.Mills (eds) From Max Weber: Essays in Sociology, London: Routledge and Kegan Paul: 196–244. ——(1971) ‘Legitimate Authority and Bureaucracy’, in D.S.Pugh (ed.) Organization Theory: Selected Readings, London: Penguin: 15–27. Williamson, O.E. (1985) The Economic Institutions of Capitalism: Firms, Markets, Relational Contracting, New York: The Free Press.

16

1

THE RISE OF THE FACTORY SYSTEM IN BRITAIN Efficiency or exploitation? S.R.H.Jones

Introduction One of the most striking developments to have occurred during the Industrial Revolution in Great Britain was the rise of the factory system. Until the closing decades of the eighteenth century the majority of manufactured goods were produced by the domestic or putting out system. At the heart of the domestic system was the cottage workshop where the domestic artisan, often assisted by journeymen, apprentices and family members, used simple tools to perform one or more stages of the manufacturing process. The materials processed might belong to the artisan but they were frequently supplied by a merchant capitalist who paid for the work to be done on a piece rate basis. By 1871, however, the situation had radically changed, for the bulk of manufactured goods was now produced in the factory. This was certainly true for textiles, by far the largest single source of employment, whilst considerable quantities of pottery, hardware and other items of consumption also came from the factory (Usher 1921:362). The control of production had also changed, the role of the domestic artisan being taken over by the manufacturer who now supplied the plant and machinery as well as the materials and directed a factory labour force that might number hundreds of workers. Although factories came to dominate production in many trades during the course of the nineteenth century, substantial manufacturing activity continued to take place elsewhere. Even in the cotton industry, the archetypal factory industry, outworkers were weaving cloth on handlooms well into the second half of the nineteenth century, while for woollens the spinning jenny and handloom continued to be employed in homes and workshops even

17

S.R.H.JONES

longer. In some trades, such as garment manufacture, outwork was never entirely superseded. The flexibility conferred by a mixture of factory production and outwork was welcomed by most factory masters as the existence of a corps of outworkers/small subcontractors enabled them to cope with short runs or fluctuations in demand without incurring the fixed and quasi-fixed costs associated with factory production. Moreover, outwork and subcontract labour possessed the advantage that it was generally cheap, especially when it included large numbers of women and children (Jones 1987:87–8; and Hudson 1992:31) Yet it would be wrong to view the domestic and subcontractors’ workshops that existed in mid-Victorian England as little more than marginal units of production. It is quite clear that in a number of trades, particularly those catering for a growing body of middle-class consumers who typically placed a premium on quality, craftsmanship and fashion, the small workshop represented the optimal unit of production (Samuel 1977:56–70). This was the situation in the small-arms trade where the market for sporting guns remained the preserve of small makers who relied extensively on subcontract and outwork. Large military orders for cheap, standardised products, on the other hand, were filled by firms such as the Birmingham Small Arms Company which operated a large mechanised factory at Small Heath especially to meet this segment of the market (Timmins 1967:430–1; Allen 1966:116–19). The relationship between factory, workshop and outworker during the process of industrialisation was thus an exceedingly complex one. Where production might be located was the result of interaction between a variety of factors, including the nature of technology, the characteristics of the market, and relations between capital and labour which, on occasions, might block technical change (Jones 1987:85; Robertson and Alston 1992:330–2). Nevertheless, there was a general trend towards increasing factory production for most of the nineteenth century. What exactly constituted a factory is difficult to say. Indeed, the debate over the definitional characteristics of the factory is almost as old as the factory system itself. Most early commentators believed that the presence of power-driven machinery was an important distinguishing feature of the factory. Thus in 1835 Andrew Ure wrote The term Factory, in technology, designates the combined operations of many orders of work people, adult and young, in tending with assiduous skill a system of productive machines continuously impelled by a central power. (Ure 1967:13) Not all were persuaded that plants using power-driven machinery should be necessarily described as a factory. ‘I am engaged in scribbling, carding and fulling,’ complained Joshua Robinson, the operator of a West Riding woollen 18

THE RISE OF THE FACTORY SYSTEM IN BRITAIN

mill, ‘it is no factory—neither am I manufacturer but work the mill for country and domestic manufacturers.’ Early legislators, it seems, were not in the least concerned whether mill operators owned the materials they processed or not, deeming factories and mills to be synonymous for the purposes of inspection and regulation (Jenkins 1975:11–13). Nevertheless, they did agree that the use of powered machinery was an important feature of factory production, and factory inspectors were instructed to collect details of the nature and extent of power employed. The Factory Acts were gradually extended to cover a wide range of plants, many without machinery, and this necessarily gave rise to doubts about the usefulness of defining factories in strictly technological terms. Cooke-Taylor, writing in 1891, argued that the term ‘factory’ should be applied to all plants where several workmen are gathered together for the purpose of obtaining greater and cheaper conveniences for labour than they could procure individually in their own homes for producing results by combined efforts which they could not accomplish separately; and for saving the loss of time which the carrying of an article from place to place during the several processes necessary to complete its manufacture would occasion. (Cooke-Taylor 1891:1) Scale and efficiency, therefore, not technology, should be the defining characteristics of the factory. Given the variety of definitions, it is not surprising that there should be some disagreement concerning the origins of the factory system. Contemporaries, who marvelled at the complexity of machinery introduced by the likes of Thomas Lombe and Richard Arkwright, were inclined to regard the spread of factories as the necessary concomitant of advances in technology. They were well aware that not all factories contained machinery, and that sometimes masters adopted factory production merely to sweat labour. For the majority, however, the triumph of the factory system was inextricably linked to the introduction of new technology (House of Commons Committee on the Woollen Manufacture of England 1806:9–11; Reports from Assistant Commissioners, Handloom Weavers 1840:434–58; Ure 1967:12–13; Wing 1967: v-viii). Economic historians, too, have tended to point to the technological origins of the factory system. Mantoux, writing at the beginning of the twentieth century, was quite unequivocal. ‘The factory system’, he states, ‘was the necessary outcome of the rise of machinery’ (Mantoux 1961:246). Most have followed in the Mantoux tradition. Heaton, whose research was conducted a few years later, concludes:

19

S.R.H.JONES

the major part of the economic advantage of the factory springs from the use of machinery capable of performing work quickly, and the use of power which can make machinery go at high speed. Until these elements of speed became possible, the factory system did not possess any great advantage over the cottage industry. (Heaton 1965:352) And finally David Landes, in a recent restatement of his position: ‘No, what made the factory successful in Britain was not the wish but the muscle: the machines and the engines. We did not have factories until these were available, because nothing else would have overcome the advantages of dispersed manufacture’ (Landes 1986:606–7). The view that the rise of the factory system was due to the superior efficiency of the new factory-based technology has attracted increasing criticism since the mid-1970s, especially from those who are not economic historians. Stephen Marglin, for example, writing from the perspective of the New Left, focused on the exploitative aspects of the factory system. According to Marglin, factories were introduced not because they were technologically superior, but because they enabled capitalists to control the work process more closely and thereby extract greater output from workers for a given cost. Capitalists (who Marglin suggests were irrelevant to the productive process, be it the putting out or the factory system) were thus able to obtain an even ‘larger share of the pie’ (Marglin 1974:62). New institutional economists, while rejecting the exploitation thesis of the New Left, have also questioned the technological underpinnings of the factory system. Oliver Williamson, for example, argues that putting out was superseded by the hierarchically organised factory because the latter economised on transactions costs. Economies were achieved as factory production enabled transportation and inventory costs to be cut, embezzlement to be reduced and manufacturers to respond more readily to changes in fashion and demand. For Williamson, therefore, the main reason why the factory system was adopted was that it was organisationally more efficient than putting out and other earlier modes of production (Williamson 1980:28–9). These challenges to traditional explanations for the rise of the factory system have not remained unanswered. Landes, for example, has questioned the empirical basis of Marglin’s work, arguing that, while Marglin’s argument is logically elegant, it explains what might have happened rather than what actually did happen (Landes, 1986). Williamson has been criticised on a number of counts, including the use a model that uses dubious aggregative techniques, a lack of appreciation of the nature of British labour markets, and other empirical deficiencies (Jones 1982, 1983). Attempts by Rick Szostak to resuscitate Williamson’s work by arguing that a fall in transport costs led to a shift in comparative advantage from putting out to the hand factory in the late eighteenth century are subject 20

THE RISE OF THE FACTORY SYSTEM IN BRITAIN

to similar methodological and empirical criticism (Jones, 1992). Rather more satisfactory has been Gregory Clark’s recent work linking the introduction of new technology with a somewhat involuntary increase in work intensity, indicating that, in the modern factory system, efficiency and exploitation often went hand in hand (Clark, 1994). That the debate has continued unresolved for almost twenty years is not entirely surprising, for there is disagreement as to what is actually meant by ‘the factory system’, when it is supposed to have emerged, whether the empirical evidence has been properly addressed, and what was typical and what was not. Before we examine the merits of the various arguments, therefore, it might be helpful to establish the broad outlines of factory development in the centuries leading up to and including the Industrial Revolution.

Factories before the Industrial Revolution Factories that employed scores of workers using simple hand technology can be found in England as early as the sixteenth century. The most celebrated of these, perhaps, are the woollen textile factories of William Stumpe and Jack of Newbury, but there were similar establishments elsewhere (Patterson 1967:151–2). How cost effective these early factories were is difficult to say, although their ability to survive may well have been due to the unusual demand conditions of the time. Suffice it to say that they disappeared in the second half of the sixteenth century, the collapse in demand together with anti-factory legislation probably hastening their demise (Heaton 1965:90– 1). For the next two centuries the putting out system was increasingly adopted by those industries where processes were divisible, technology was simple and semifinished materials were not costly to transport. Rural labour surpluses encouraged this development, with a system of outriders and agents emerging that enabled capitalists to put out work to underemployed and unemployed labour throughout the countryside. Putting out work was not only confined to the countryside, for a number of urban trades came to rely on outwork too, but it was cheap rural labour that underpinned the growth of the putting out system. In some industries, technological non-separabilities and other characteristics meant that production in the domestic workshop was inappropriate. Manufacturers of iron, porter, glass, paper and sea-salt, for example, required plants of a certain scale if they were to realise production economies. For many goods, however, the small workshop with its simple technology constituted the least-cost mode of production. Certain stages of the manufacturing process, such as fulling cloth, might require more capitalintensive technology, but it was often possible to integrate this within the broad structure of domestic production.

21

S.R.H.JONES

Putting out offered significant advantages to the merchant capitalists who organised production. Labour was usually cheap, fixed costs were low, an extensive division of labour could be practised, and, in times of depression, production could be easily curtailed. Despite these advantages, however, the hand factory or manufactory, which utilised essentially the same technology as that employed by outworkers, began to appear in a variety of trades from the end of the seventeenth century onwards. The term ‘proto-factory’ has been coined by Freudenberger and Redlich to describe these technologically simple factories but, given its wider connotations within the protoindustrialisation debate, it will not be adopted here (Freudenberger and Redlich 1964:381). Some of the earliest factories were established as a means of setting the poor to work. The view that the poor might be usefully employed by the parish gained ground during the course of the seventeenth century, and by 1700 pauper factories run by the parish had come into being in London, Bristol and elsewhere (Furniss 1957:108–10; Thirsk and Cooper 1972:301–2). Philanthropists also set up factories in order to alleviate the twin evils of poverty and unemployment (Heaton 1965:354–5). The history of pauper factories, however, is scarcely a glowing testament to their efficiency. A study of forty-five workhouses-cum-factories undertaken in 1725 revealed that none earned sufficient to cover the costs of maintaining their inmates (Furniss 1957:108–9). Nevertheless, paupers continued to be set to work, often performing tasks in the workhouse on materials supplied by a manufacturer and under the supervision of an agent. This was particularly common in that archetypal factory industry, pin-making, where manufacturers continued to contract with overseers of the poor (and with gaols) to supply work well into the nineteenth century. An increasing number of manufacturers also employed paupers on their own premises (Rose 1989:5–29). One of the earliest factories in the hosiery industry, that of Samuel Fellowes of Nottingham, employed more than forty parish apprentices by the early 1720s. Few other Nottingham hosiers followed suit, not even in the 1760s when competition was at its fiercest, and domestic outwork continued to dominate the hosiery trade until the 1840s (Aspin and Chapman 1964:34; Nelson 1929–30:469). At the same time that experiments were being carried out with pauper factories, a growing number of hand factories came into being that relied on free labour. One of the earliest was that established by Ambrose Crowley at Winlaton, near Newcastle, which was set up in 1691 to manufacture nails. Expansion soon followed with a new, vertically integrated plant being erected in nearby Swalwell, and ultimately the firm possessed blast furnaces, slitting mills and a range of large workshops producing numerous items of hardware and ships’ chandlery. The Crowley enterprise was undoubtedly ‘a giant in the age of pygmies’. It also represented an attempt to break new ground by extending factory production to a hardware industry hitherto organised on 22

THE RISE OF THE FACTORY SYSTEM IN BRITAIN

traditional domestic lines. Yet in spite of the advantages of a reduction in embezzlement, an improvement in scheduling, better quality control and the evident profitability of the Crowley empire, few manufacturers sought to emulate Ambrose Crowley. Thus most items of hardware continued to be produced in small domestic workshops for a century or more. Indeed, the Crowley ironworks found it increasingly difficult to compete with domestic nailmakers of the West Midlands, the production of nails finally ceasing at Winlaton in 1816 (Flinn 1962:94, 184–93, 252–4). There were, nevertheless, examples of factory production in other hardware trades, especially in the production of fancy boxes and japanned ware. Among the first manufactories to appear was that of Charles Osborne, a tobacco box maker of Wolverhampton who set to work ‘a large shop of workmen’ during the 1720s. Others in the Birmingham district also gathered workers into the factory, including John Baskerville—who, by 1745, employed 300 hands in japanning—and the celebrated firm of Boulton and Fothergill. Yet in spite of contemporary interest in large factories, few manufacturers were prepared to adopt them. As Marie Rowlands points out: ‘[t]he large workshop with elaborate machinery and numerous employees aroused the admiration of visitors in the mid-eighteenth century precisely because it was exceptional. Only a minority of the workforce were employed in such places’ (Rowlands 1975:155–7). This continued to be the case well into the nineteenth century.

Technology and the factory At the same time that the hand factory was beginning to make an appearance in the Midlands, a very different type of factory was coming into being in the silk trade. This took the form of a multi-storeyed, water-powered silk throwing mill. The first English silk mill was erected in Derby by Thomas Lombe who, with his brother John, had acquired, improved and patented Italian technology. Defoe described Lombe’s silk mill in some detail, evidently fascinated by machinery that incorporated ‘26, 586 wheels and 97,746 movements, which work 73, 726 yards of silk thread every time the water wheel goes round’ (Chaloner 1963:13). Thomas Lombe’s patent for silk throwing expired in 1732. Thereafter, new entry occurred, and over the next four decades a number of mills were built incorporating the latest technology. By 1769 there were six silk mills at work in Stockport, principally supplying Spitalfields weavers, with a further twenty in operation in nearby Macclesfield. Additional mills were to be found at Derby, Congleton, Leek and Knutsford, further mills coming into operation in the south of England towards the end of the century. The success of these early silk mills did little, however, to persuade silk manufacturers to employ factories in the weaving section of their trade. A

23

S.R.H.JONES

few weaving factories were introduced at the end of the Napoleonic Wars, mainly in order to force down wage rates, but the majority only came into being with the introduction of new weaving technology later in the nineteenth century (Jones 1987:71–96). Despite the lead given by the silk industry in erecting large water-powered mills, the rise of the factory system has traditionally been associated with cotton manufacture. The first cotton spinning factories came into being during the 1740s, based on the patented machinery of Lewis Paul and John Wyatt, but they proved to be commercially unsuccessful. It was only after Arkwright patented his roller-spinning machine, or water-frame, in 1769 that cotton began to overtake silk as the premier factory industry in the country. The expiration of the roller-spinning patent in 1783, followed shortly afterwards by the revocation of Arkwright’s omnibus patents for preparatory processes, removed the principal barriers to entry to the factory-spinning section of the trade. By 1795 some Arkwright-type factories had come into existence (Chapman 1972:30). Substantial changes also took place in the woollen and worsted industries from around 1770 onwards. Power was applied to the preparatory processes of carding, slubbing and spinning, and to finishing with the mechanisation of milling and shearing. The result was a proliferation of mills to house the new technology, with old fulling mills being converted and extended and new water and steam mills being erected. There were some 243 mills in operation in the West Riding of Yorkshire by the beginning of the nineteenth century, as well as a number elsewhere (Jenkins 1975:17). Whether all should be termed factories is a matter for debate, for some employed less than a dozen hands and performed only a limited range of processes. Nevertheless, many operated on an integrated basis, using power for processes such as scribbling and carding (and sometimes worsted spinning) and hand powered machines to produce slubbings and weft, while some contained looms as well. The larger mills of this type often employed scores of workers and clearly merited the description ‘factory’, however that term might be defined. Yet not all advances in spinning technology required a large mill with a central power source. The spinning jenny, which was first introduced in the 1760s, was a hand-powered machine small enough for workers to operate in their cottages. Improvements in the method of construction soon permitted the jenny to be enlarged, however, and by the 1770s modest factories housing the larger hand-powered jennies were to be found in both Lancashire and Somerset (Wadsworth and Mann 1965:500–1). The larger jenny was slower to make its way in woollens than in cotton, mainly for technical reasons, although by the early nineteenth century there were a number of jenny workshops in both Yorkshire and Gloucestershire (Aspin and Chapman 1964:56–8). Crompton’s mule, which first appeared in the 1780s, was also a handpowered spinning machine suitable for domestic use. As with the jenny, 24

THE RISE OF THE FACTORY SYSTEM IN BRITAIN

enlarged versions were soon to be found in the factory, the application of power to the mule in the 1790s hastening its departure from the cottage (Kennedy 1819:129). By 1812, according to Samuel Crompton’s census, there were 573 mule workshops or factories, although how many of these were reliant on mechanical power is not known (Daniels 1930:109). The introduction of what might be regarded as ‘intermediate technology’ was also linked to the spread of hand factories in other trades. The stamp and press, for example, was introduced into a number of Birmingham trades, transforming the button industry, making inroads into the pin industry, and doing away with the hand punch for most types of needle-making. Like the early jenny, the stamp and press was small enough to be used in the cottage, although the not inconsiderable throughput of semi-finished goods, the fact that some of the machines were subject to patent, and the need for occasional adjustment by a fitter meant that there was a tendency to locate them in a large workshop or factory. In spite of these developments, outworkers and subcontractors using intermediate technology found themselves placed under increasing pressure as power-driven machinery was introduced, with those making pins, for example, being almost completely supplanted by automatic machinery by the middle of the nineteenth century. The introduction of capital-intensive equipment that significantly lowered the costs of production did not always rob artisans of their independence. In the West Riding woollen industry, domestic clothiers gained access to the new technology by combining to form joint-stock mills which performed a variety of processes for their shareholders, including carding, slubbing, fulling and sometimes spinning (Hudson 1983:135). In nearby Sheffield, labour chose to hire capital equipment rather than become petty capitalists themselves. Thus craftsmen in the cutlery trades were able to retain their independence by renting grinding-troughs at steam powered public wheels at which they worked for themselves, for factors and for larger manufacturers (Pollard 1959:54–7). Similar arrangements were arrived at by pointers and scourers in the Redditch needle industry (Jones 1980:187), while in the Birmingham sporting gun trade a combination of subcontracting and the extreme division of labour ensured the survival of artisans and small local and country gunmakers long after the application of steam power to barrel-boring and grinding (Timmins 1967:387–93; Greener 1910:407–10). Even in the cotton industry the practice of inside contracting meant that there was a role for the semiindependent artisan, with employers attempting to reduce the costs of hiring and supervision by employing labour through mule-spinners rather than directly themselves (Lazonick 1979:233; Kirby 1994:127). In these and other industries, therefore, the introduction of factory machinery with a high technical optimum did not result in all processes being carried out by large-scale hierarchically organised firms. Where technology was flexible and the characteristics of the market permitted, contracting out was seen by merchants, manufacturers and artisans as one way of sharing the 25

S.R.H.JONES

costs and risks of production. Nor was this structure necessarily lacking in vitality, for the existence of large numbers of small producers appeared to provide a competitive environment that encouraged invention and innovation, with new designs and new methods helping to ensure that the workshops of Birmingham and Sheffield were able to survive in an age of factories (Sabel and Zeitlin 1985:146–7). The symbiotic relationship that existed between factory, workshop and outworkers undoubtedly contributed to the growth in the numbers of both power-driven and hand factories between 1780 and 1850. Precise figures are not available, for the early returns from factory inspectors tended to exclude plants without power and provide only an imperfect record of the remainder. Even so, the returns for 1835 reveal 1,245 cotton mills at work throughout the United Kingdom, more than 1,300 woollen and worsted mills, 345 flax mills and 238 silk mills (Jenkins 1973:26–46). In addition, there were scores of hardware factories that produced pins, needles, nails, buttons, cutlery and other items, together with a considerable number of factories and workshops in other branches of industry. The majority of these establishments employed advanced technology, using steam and water power in conjunction with the latest machinery to ensure that, by the middle of the nineteenth century, Britain had become the workshop of the world.

Traditional explanations for the rise of the factory system The rise of the factory system was an exceedingly complex phenomenon. Factories appeared in certain trades long before others; there were differences in timing both between branches of a single trade and between geographical areas in which that trade was found; the shift to factory production was sometimes accompanied by the introduction of new technology and sometimes not, and on occasions artisans retained independence even though working within the factory. Generally speaking, though, the movement into the factory took place in most trades after 1780 and, from the turn of the century, was associated with steam power and the growth of factory towns. In spite of its complexity, the essential features of factory development have long been familiar to most economic historians. Certainly those who argue for the technological origins of the factory system are well aware of the wide variety of circumstances in which factories were adopted and that, in a number of instances, technological considerations were unimportant. Why, then, have so many economic historians, including Mantoux, Heaton, Usher, Ashton, Landes, Mathias and others, chosen to place so much emphasis on the role of technology? Before one can understand why these scholars consider technology so important, it is necessary to appreciate the nature of the dominant paradigm at the time they wrote. The dominant paradigm, it seems, was a general

26

THE RISE OF THE FACTORY SYSTEM IN BRITAIN

belief in the existence of an Industrial Revolution. Indeed, the notion of developments that resulted in fundamental—even revolutionary—changes in economy and society finds precise expression in the titles of the seminal works of both Mantoux and Ashton and, more allegorically, in Landes’s The Unbound Prometheus. At the same time, each of these scholars was aware that, in certain respects, change was more evolutionary than revolutionary. But, to quote Ashton, ‘the phrase “Industrial Revolution” has been used by a long line of historians and has become so firmly embedded in common speech that it would be pedantic to offer a substitute’ (Ashton 1962:2). Thus qualified, the belief in the existence of an Industrial Revolution has proved to be a powerful influence shaping the ideas of economic historians and others in the twentieth century (Coleman 1992:1–42). But what was it that so radically transformed the nature of economy, society and environment in late eighteenth-century England? A major force for change, in the eyes of many, was the modern factory system. This appeared in its embryonic form in the silk industry, reached its full bloom in cotton, was adopted by the rest of the textile industry and, in the nineteenth century, spread to other manufacturing sectors. The essential features of the modern factory system have been carefully spelt out by both Mantoux and Landes. First, there was the substitution of inanimate sources of power for the limited and irregular effort of human and animal muscles. And second, machines—as opposed to mere tools—performed tasks previously undertaken by hand. The difference between a tool and a machine is a matter for debate but, for the purposes of their arguments, both Mantoux and Landes envisage machinery that was of a scale that required sizeable premises and a central power source for efficient operation (Mantoux 1961:346; Landes 1969:2). In other words, both regard technological indivisibilities to have been a major feature of the modern factory system. For many historians, however, the development of the modern factory system entailed rather more than just the introduction of power-driven machinery. It also entailed the transformation of an industry that was capable of growing in size until it had major ramifications for the rest of the economy. Thus, even though there were more than thirty power-driven silk mills at work prior to Arkwright’s experiments, cotton is regarded as the key to the industrial revolution. This is because the cotton industry, unlike silk, enjoyed a favourable combination of highly price-elastic demand schedules and rapidly shifting cost curves (Landes 1969:81–3; Mathias 1983:116–17). The result was almost exponential growth, described by one contemporary as ‘absolutely unparalleled in the annals of trading nations’ (Aiken 1968:3). By the 1820s, few commentators doubted that both economy and society in Great Britain had undergone a radical change (Berg and Hudson: 1992). A modern factory system based on power-driven cotton mills is therefore seen as having been at the heart of the first Industrial Revolution. There is less unanimity concerning the processes that lay behind the appearance of 27

S.R.H.JONES

the factory system, with Ashton stressing, amongst other things, a reduction in the cost of capital, Mantoux paying particular attention to inventors and inventions, and Landes linking developments in technology with changes in the structure of demand. Indeed, the growth of demand, especially domestic demand, is seen by some as crucial to the development of the industry, with bottlenecks in supply leading to invention and innovation. Most are agreed, however, that the rapid increase in cotton factories placed substantial demands on capital goods and other industries. This induced further innovation, capital deepening and increases in labour productivity that were so important for sustained economic growth. Yet at the same time that power was being applied to spinning and other processes, there was also an increase in the number of hand factories that did not require power. The 1770s witnessed the spread of jenny factories, especially in Lancashire cottons, while towards the end of the century large loom shops began to appear in both the Yorkshire and Gloucestershire woollen industries. Given such activity, why have generations of economic historians tended to regard hand factories as unimportant in their contribution both to the development of the modern factory system and, more generally, to the process of industrialisation? The answer would seem to be that the spinning capacity of jenny factories was relatively small compared with that of power-driven factories. Loom shops were even less significant, being dwarfed by outwork until the 1830s. Furthermore, such establishments yielded neither the cost savings nor the linkage effects of the power-driven mills. Their introduction did not, therefore, represent a radical change in the organisation of production. Few would deny that there were advantages to be gained from gathering workers together in manufactories. The centralisation of production enabled the manufacturer to avoid the transportation and other costs of putting out and, through closer monitoring, reduce embezzlement, improve scheduling and exercise greater quality control. Yet there were also disadvantages, the need to provide both plant and equipment leading to an increase in fixed costs and a reduction in the manufacturers’ ability to respond to fluctuations in demand. Nor was it easy to persuade a labour force unaccustomed to time discipline to work regularly or diligently in the manufactory, especially in the absence of machinery that imposed its own work discipline (Pollard 1968:189; Harte 1977:41). But the principal disadvantage of the factory system, it would seem, is that masters were usually obliged to pay inworkers higher rates of pay than outworkers. This was the result not only of the premium that factory masters were obliged to pay labour to work in the much disliked factories, but because they were also unable to tap distant and rural labour markets in which employment opportunities were limited and wage rates were substantially lower. Putters out faced no such constraints, being able to employ on a casual basis persons of all ages, sexes and states of health—some of 28

THE RISE OF THE FACTORY SYSTEM IN BRITAIN

whom were otherwise unemployable—paying wages that might be below subsistence and which sometimes had to be made up by the parish (Select Committee on the Silk Trade 1831–2; 195; Bischoff 1968:185–6; Landes 1986:606–7). The judgement of a succession of historians, therefore, is that the gains from factory organisation were generally insufficient to offset the additional costs involved. When raw materials were particularly expensive, quality a consideration, prompt delivery important, secret processes in use, or other fixed costs involved, it might pay masters to have a proportion of their goods produced in the factory (Ashton 1962:109; Aspin and Chapman 1964:33; Landes 1986:603). The absence of substantial numbers of hand factories prior to the Industrial Revolution, however, is taken as prima facie evidence of the fact that such conditions were rarely met. The modern factory system, like the Industrial Revolution of which it was such an essential element, is thus seen to have been based on new technology that ultimately overwhelmed putting out and other earlier modes of production. This technology was not usually to be found in the hand factory or manufactory but in more technologically advanced establishments, of which the cotton factory, with its power-driven machinery, was the exem-plar.

Factories, discipline and the ‘new left’ The first major challenge to the view that the rise of the modern factory system was due to advances in technology was mounted by Stephen Marglin. In a seminal article published in 1974, Marglin argues that the adoption of factories had nothing to do with their technological superiority. Rather, it was so that the capitalist and not the worker might control the work process and quantity of output. This innovation in the organisation of work enabled the capitalist to get ‘a larger share of the pie’. The change in relative income shares was achieved through supervision and discipline, the closer monitoring and control of the factory system making it possible for masters to force the workers to work harder and produce more for little or no extra remuneration (Marglin 1974:62, 84). The advantages conferred by supervision and discipline were sufficient, according to Marglin, to persuade manufacturers to adopt the factory system even though it was no more technologically efficient than putting out. These advantages, he maintains, are widely recognised by a number of leading historians, including Mantoux, Ashton and Landes, though none are willing to admit that supervision and discipline were responsible for the emergence and success of the factory system. Marglin attributes their agnosticism to the fact that most employ a model in which greater output is achieved through technical innovation. The technologically simple factory, he says, which

29

S.R.H.JONES

only obtained greater outputs through dint of greater inputs, does not fit easily within their model (Marglin 1974:82–3). Simple manufactories, as Marglin is well aware, appeared long before the Industrial Revolution. Why then, if such institutions offered manufacturers the opportunity of appropriating ‘a larger share of the pie’, did they not appear in significant numbers before the end of the eighteenth century? Marglin offers two related explanations. First, he suggests that the factory system was slow to arrive because of the lengthy struggle between guilds and capitalism. Thus, in the sixteenth and seventeenth centuries, the small master and journeyman took advantage of divisions between more powerful classes ‘to forge temporary alliances that for a time at least were successful in stalling the advent of the factory’. Gradually, however, as the power and interests of the various classes changed, the statutory provisions that protected journeymen and small masters fell into disuse and were ultimately repealed. This opened the way for the introduction of factories (Marglin 1974:102). Second, Marglin argues that, even after power relationships changed, pressures to switch to the factory system did not become critical until the latter part of the eighteenth century, when outworkers sought to take advantage of buoyant product markets by demanding higher real wages. With greater remuneration, outworkers could afford to substitute leisure for work and take time off. If the wage rates offered still seemed inadequate, they might embezzle raw materials and engage in other types of fraud. Capitalists attempted to control such activity by recourse to the law, with statutes being passed to ensure work regularity and reduce embezzlement. ‘But’, says Marglin, ‘more direct action proved necessary. The capitalists’ salvation lay in taking immediate control of the proportions of work and leisure.’ Thus the very success of the putting out system in exploiting demand at home and abroad ‘contained within it the seeds of its own transformation’ (Marglin 1974:92–5). Although the need to control the work process is seen as the primary reason for the movement into the factory, Marglin is prepared to accept that technology also became important in later years. Technology, however, was not the independent cause of the factory system as Mantoux and Landes would suggest. On the contrary, the particular forms that technological change took were shaped and determined by factory organization. It is not accidental that technological change atrophied within the putting out system after Hargreaves’ jenny flourished within the factory. Why was technological change apparently biased towards the factory? Marglin suggests that this is because, on the demand side, factory owners provided a better market for patented inventions than domestic outworkers, while, on 30

THE RISE OF THE FACTORY SYSTEM IN BRITAIN

the supply side, it was easier to monitor patent usage within a factory rather than an outwork environment (Marglin 1974:89–90). Marglin, the factory system and the evidence The Marglin thesis is carefully structured and buttressed with numerous quotations drawn from both contemporary works and those of economic historians. What evidence is there, however, to support his principal contention, namely, that the success of the factory system ‘had little or nothing to do with the technological superiority of large scale machinery’? Mantoux and Landes, as we have seen, carefully spell out what they mean by ‘factory system’. Indeed, they employ a definition which excludes a ‘system’ consisting of hand factories. Marglin is less specific, for, after initially referring to the ‘factory system’, he then refers to the success of ‘the factory’. The confusion is compounded by his choice of examples that neither Mantoux nor Landes would consider as part of the ‘factory system’. Consequently it is never entirely clear whether he is discussing the rationale for hand factories or for a factory system based on power. Such a distinction is probably unimportant to Marglin, who seems to believe that hand factories represented the first stage of a process that paved the way for the introduction of powerdriven machinery. Yet while a considerable proportion of factory spinning was initially based on the hand-powered jenny, what evidence is there to support the view that manufacturers set up jenny factories to enable them to extract more output from the workforce through closer supervision? For Marglin, the fact that cottage and jenny spinners used the same ‘basic machine’ is conclusive proof that the success of the jenny factory ‘could only have been for organizational reasons’ (Marglin 1974:86–7). Yet what does Marglin mean by the term ‘basic machine’ and, more importantly, was there any difference between the small cottage machine and the larger factory jenny? Certainly those in the cotton trade believed that there was a difference, with riots in Lancashire in the late 1770s leading to jennies of over twentyfour spindles either being destroyed or having excess spindleage lopped off. Following the riots, a proposal was put forward to tax all jennies of over twenty-four spindles. The rate of taxation was to increase as the size of the jenny rose, presumably to offset economies of scale and enable those with the smaller ‘basic’ machine to compete (Wadsworth and Mann 1965:496–502). In the woollen industry, too, attempts were made to limit the number of spindles per jenny or workshop (Aspin and Chapman 1964:57). Given the economies of scale that the operators of large jennies clearly enjoyed, why were they not adopted by the cottage spinner? Sometimes they were, especially in Yorkshire where small clothiers might incorporate them into their operations. Space, however, was often a limiting factor, with many 31

S.R.H.JONES

persons finding it difficult to house machines that might have a three-foot draw and be more than 12 feet in length. Finding additional space, moreover, was not always easy, especially if William Radcliffe’s comments in the shortage of workshops are to be believed. Furthermore, the cost of renting, setting up and maintaining a small jenny (or mule) shop was beyond the resources of most workers (Podmore 1923:38–9; Aspin and Chapman 1964:44; Lazonick 1979:233; Landes 1986:606). The evidence to support Marglin’s contention that the jenny factory was generally adopted for purely organisational reasons is tenuous, to say the least. Large jennies were simply more efficient than small ones, their very size ensuring that they were usually located in workshops and factories. Here they were often to be found running in tandem with power-driven carding engines and slubbing billies which, because of their size, throughput and power requirements, were also unsuitable for use in the cottage (Aspin and Chapman 1964:50–1; Landes 1986:604). Organisational advantages are also seen by Marglin as being responsible for the spread of handloom factories: Long before the powerloom became practicable, handloom weavers were brought together in workshops to weave by the same techniques that were employed in cottage industry. Clearly handloom shops would not have persisted if it had not been profitable for the entrepreneur, and just as clearly the source of profits could not have been in superior technology. There is no evidence that the handloom in the capitalist’s factory was any different from the one in the weaver’s house. (Marglin 1974:87–8) It is by no means clear, however, that the spread of handloom factories was due to the advantages conferred by better supervision and discipline. Technical parameters in the industry changed significantly around this time, with improvements in the quality of yarn due to power-spinning reducing the skill levels required of weavers, thereby opening the way for the employment of cheap, unskilled labour working in factories under skilled supervision. Even so, Landes believes that the rationale for such factories usually lay elsewhere, pointing out that both loom shops and jenny factories were often part of larger integrated concerns. These might include mechanised preparatory process such as slubbing and carding, as well as heat-intensive and power-driven finishing processes such as dying and fulling. Landes concludes: ‘If clothiers were going to invest in factories at both ends of the production process, they wanted to be able to keep plant and equipment busy’ (Landes 1986:604). Capacity utilisation, he suggests, not the gains from supervision and discipline, was the raison d’être for most loom shops and jenny factories. 32

THE RISE OF THE FACTORY SYSTEM IN BRITAIN

There were, nevertheless, exceptions to this pattern of development that lend a measure of support to the Marglin thesis. In 1815, for example, a strike by Macclesfield silk weavers over a reduction in piece rates led manufacturers to establish hand factories and employ erstwhile apprentices and others at significantly lower rates of pay. Elsewhere in the trade, a system of ‘half-pay apprentices’ was employed in which nine or ten young persons were gathered together under one roof and set to work under the supervision of a skilled silk weaver. Silk weaving factories were not widely adopted, however, even though water-powered silk throwing factories had been in existence for almost a century by this time. Clearly the cost advantages that were supposed to flow from factory organisation were not glaringly obvious to those engaged in the silk industry! (Jones 1987:78–80). Handloom factories were also introduced by Gloucestershire woollen manufacturers as a means of forcing down wage rates. A few large loom shops appeared around 1800 but they were not adopted on any scale until a succession of strikes against a reduction of wages in the 1820s finally led, in 1828, ‘to the more general establishment of the shoploom system’. Whether manufacturers would have been so quick to transfer weaving to the factory had they not been so concerned about ‘the value of buildings and machinery rendered inactive by this strike’ is a matter for conjecture (Reports from Assistant Commissioners, Handloom Weavers 1840). Moreover, even though the Gloucestershire industry was burdened with significant fixed costs due to the investment in fulling mills, shearing engines and so forth, handloom factories were surprisingly slow to be adopted. Marglin is therefore correct to suggest that some hand factories were introduced so that capitalists could exploit their positions of power to extract more labour from workers for a given cost. What he fails to establish, however, is that the majority of jenny factories and large loom shops, upon which his argument is largely based, were adopted for those reasons. Far more compelling are arguments concerning the size and interrelatedness of much of the new equipment—and this brings us back to technology. Paradoxically, although Marglin is quick to play down the role of technology, his treatment of it is surprisingly cursory. Certainly he offers no empirical evidence to support his contention that, because of the patent system, invention was biased towards producing technology more suited to the factory than putting out. Indeed, had he examined the cotton industry in greater depth, he would have discovered that the unpatented mule, although initially located in the cottage, was transferred to the factory because larger machines of greater efficiency required more space and power. Moreover, as Berg observes, many advances went unpatented (Berg 1991:187). More generally, Marglin fails to analyse the reasons for the introduction of power-driven factories in silk, cotton or wool, or to assess their importance relative to hand factories. Nor does he comment on the timing of their introduction which, in the silk industry, occurred almost a century prior to 33

S.R.H.JONES

the adoption of hand factories and, in the case of cotton and wool, was contemporaneous with the emergence of jenny factories and loom shops. For many economic historians, it is the introduction of these power-driven factories that constitutes ‘the rise of the factory system’, and on this subject Marglin remains largely silent.

New technology and exploitation The failure of Marglin to explore the nature of exploitation within powerdriven factories is somewhat paradoxical, because it was the hours and conditions of work within such establishments that attracted the ire of factory reformers. Compared with hand factories, where piece rates were common and discipline often lax (Clark 1994:133–7), conditions in many of the new, increasingly steam-driven mills were sufficiently bad that repeated legislation and regular inspection were required to check the worst abuses. Of course, given a highly competitive environment and the need to keep expensive capital equipment fully occupied, it seems hardly surprising that discipline within the modern factory system should have been extreme. Whether the desire to keep workers at their posts and maintain throughput was the main factor lying behind enhanced discipline, nevertheless appears doubtful. As Gregory Clark has recently pointed out, it may well have been cheaper to hold larger inventories and adopt more flexible hours than bear the costs of ensuring regular attendance. Instead, he maintains that the primary purpose of enhanced discipline was to make people work harder because, with high fixed costs due to heavy investment in plant and machinery, the marginal productivity of labour rose sharply as effort levels increased. Higher marginal productivity not only enabled manufacturers to pay factory workers significantly more than outworkers, but as labour market imperfections resulted in factory workers being paid less than their marginal product, profit levels were enhanced. Workers did not relish a regime of unremitting toil and, Clark suggests, given the choice, would have willingly traded off a fraction of higher wages for less taxing working conditions. Nevertheless, a sufficient number were prepared to present themselves for work and submit to harsh discipline, allowing themselves to be ‘exploited’ by factory masters who, at the end of the day, were able to pay them higher wages due to the more efficient use of capital equipment (Clark 1994). Whether labour, under the competitive market conditions that generally prevailed during the Industrial Revolution, could be regarded as having been exploited is a moot point, raising questions about the nature of individual rationality, the costs of search, the transferability of skills, attitudes to risk, and the way in which cultural norms adapt in a rapidly changing environment.1 Yet it is clear that, whatever present-day economists may think, to the

34

THE RISE OF THE FACTORY SYSTEM IN BRITAIN

nineteenth-century world at large the factory system was synonymous with exploitation. What is less clear is that exploitation was its primum mobile.

Transaction costs and the factory system Marglin’s account of the origins of the factory system, although receiving a somewhat guarded response from traditional economic historians, has been welcomed enthusiastically elsewhere. Thus Oliver Williamson, while not agreeing with Marglin’s central thesis, welcomes the attention paid to the way in which work was organised. Economists, he argues, are too wedded to the ‘production function orientation of received micro theory’, and this approach can make us blind to the fact that increases in efficiency can be brought about by improvements in internal organisation as well as advances in technology. For Williamson, therefore, internal organisation matters, so much so that he seeks to explain many institutional developments in terms of organisation rather than technology. This includes the displacement of putting out by what Williamson refers to as the authority relation, another term for Marglin’s hierarchically organised factory (Williamson 1980:11, 29). For the purposes of analysis, Williamson is first obliged to journey inside the ‘black box’ of the neo-classical firm in order to unravel its internal structure. What we generally find, he tells us, is that a firm does not consist of a single workstation transforming raw materials into finished products but a series of adjacent workstations, each of which adds value to an intermediate product before passing it on. In principle, therefore, one can envisage independent (and possibly dispersed) workstations transferring product to each other via a series of market exchanges. Alternatively, production might be undertaken on a cooperative basis, or certain processes might be subcon-tracted out to others, either working in the same plant or elsewhere. Williamson maintains that the reason why the hierarchically organised firm is generally used to coordinate product flow rather than coordination taking place via the market or other modes is because it incurs lower transaction costs. Hierarchy, by economising on transaction costs, thus serves important efficiency purposes, a proposition that runs counter to Marglin’s thesis concerning the role of the boss and the rationale for the factory. Williamson concludes that the historical progression from putting out to the authority relation (factory) is best explained in terms of the superior efficiency properties of the latter (Williamson 1980:11– 12). To substantiate his argument, Williamson embarks on a ‘comparative institutional assessment’ in which he compares the efficiency properties of a variety of organisational modes, including putting out and the factory. Eleven efficiency criteria are chosen against which to compare performance. These fall under three broad headings: product flow attributes, which include

35

Table 1.1 Simple efficiency properties of the putting out and factory systems

THE RISE OF THE FACTORY SYSTEM IN BRITAIN

transportation expenses, inventory requirements and leakages of product from the system; assignment attributes, which reflect the efficiency with which work is assigned to those with relevant skills, the need for coordination and leadership, and the capacity to contract for specialist services; and incentive attributes, which cover matters such as work intensity, equipment utilisation, local shock responsiveness, the incentive to innovate locally, and the capacity to respond to system shock. Technology and factor inputs are assumed to be identical. Having established his criteria, Williamson then uses a combination of a priori reasoning and a little historical evidence to assign a score of zero or one to each mode against each of the efficiency criteria (Williamson 1980:22–4). His scoring reveals that the factory was organisationally more efficient than putting out (Table 1.1). Williamson’s conclusions have received a mixed response from economic historians. All accept that the adoption of the hand factory was, on some occasions, due to a desire on the part of manufacturers to economise on transaction costs, especially ‘interface leakages’. Thus, when one Huddersfield merchant was asked whether he derived an advantage from employing looms in the factory, he replied: ‘We have them in factories principally to prevent embezzlement as we now manufacture [expensive] Spanish wool’. The advantages were not entirely clear cut, however, for he continued: ‘but if we meet with men we can depend on for honesty, we prefer having them wove at their own houses’ (House of Commons Committee on the Woollen Manufacture of England 1806:220). Although such evidence lends a modicum of support to Williamson’s argument, there are those who believe that his paper suffers from such serious methodological and empirical weaknesses that his efficiency assessments are largely meaningless (Jones 1982, 1983). The principal weakness of Williamson’s paper, it is argued, is that he ignores the fact that the putting out and factory systems had quite different production costs. The major cost difference was attributable to the fact that the putting out system enjoyed significantly lower unit labour costs. As we have already seen, the ability of putters out to tap rural and distant labour markets effectively shifted their supply curve of labour to the right. At the same time, workers’ reluctance to enter the factory due to adverse conditions and connotations with the workhouse shifted the factory labour supply curve to the left. The result was that factory masters might pay at least a third more for their labour than putters out (Bischoff 1968:185–6). Given the labourintensive nature of production, the organisational economies of the factory would have had to have been substantial indeed for it to have offset the lower production costs of putting out. Williamson’s failure to disentangle production and transaction costs means that his comparative efficiency assessment cannot be used a basis for predicting the choice of organisational mode. Williamson, of course, believes that the factory was organisationally superior, 37

S.R.H.JONES

scoring it 9–5 against putting out in his evaluation of efficiency properties. His method of scoring, however, leaves much to be desired, a failure to ascribe weights to the various efficiency criteria being particularly serious. Thus, saving in transportation expenses, a daily occurrence, is given the same weight as the capacity to respond to shocks which, by definition, occurred irregularly. Whether the efficiency criteria are, in fact, additive must also be doubted, with some measures relating to short-run costs and others to long-run factors such as the ability to innovate. Williamson is not unaware of these problems but argues that they are unimportant as his results seem to be pointing in the right direction! (Jones 1982:124–5; Williamson 1983a: 59). Yet even if one is prepared to go along with Williamson’s method of scoring, a number of objections can be raised against the scores allocated. The most important set of efficiency criteria would seem to be product flow, in which the factory is judged to be superior to putting out under all three headings (see Table 1.1). Few would disagree that the factory system incurred fewer transportation expenses. However, the advantage that the factory held under the second heading, interface leakages, of which embezzlement was the principal component, is rather less clear-cut than one might suppose. For a start, embezzlement continued to occur inside the factory, even within the archetypal pin factory (Jones 1976:41–2). Moreover, there is ample evidence to show that manufacturers incorporated embezzlement in their costings and adjusted the wage rate accordingly. They, like the workers, also went behind the wage contract, indulging in long pays, truck, arbitrary deductions and other deceits so that they actually paid less than the agreed rate. Embezzlement and truck, etc., may therefore be seen as opposite sides of the same coin and, although an inefficient method of contracting, were part of the process by which real wage rates were established in the marketplace. Legislation was repeatedly introduced to abolish both, but there is little evidence to suggest that this was a reflection of increasing problems or the way to solve them (Jones 1982:129– 32; Styles 1983:173–204). Williamson’s treatment of inventories also raises questions, for he concentrates on buffer inventories between workstations as opposed to the overall level of inventories. Although he is correct in suggesting that factory masters may have been better able to economise on buffer inventories, it is quite clear that they frequently accumulated large inventories of finished goods simply in order to keep their plant moving and hands together (Jones 1982:129). Unlike putters out, who might easily discharge hands when depression struck, factory masters, faced with fixed costs and an investment in a trained labour force, were obliged to stay in production. This ran them heavily into stock and drained them of liquidity at a time when money was tight. These disabilities far outweighed any savings that they might make on buffer inventories. 38

THE RISE OF THE FACTORY SYSTEM IN BRITAIN

The other two broad efficiency categories are those of assignment and incentive attributes, and here also the hierarchically organised factory is seen by Williamson as being more efficient than putting out. His evaluation of assignment attributes seems broadly correct, but one must question whether the factory system was more responsive than putting out to changes in market circumstances. Most view the responsiveness of putting out as one of its major attributes, being far better able to cope with fluctuations in demand than the factory system, which was obliged to hoard labour and hold stocks to meet upturns in economic activity (Jones 1982:133; 1987:87–8; Hudson 1983:142). Far from being clearly superior in terms of its efficiency attributes, a reevaluation of the evidence suggests that there was little difference in organisational efficiency between the hierarchically organised factory and putting out (see Table 1.1). In any event, access to cheap labour enabled the outwork system to compete quite effectively with the hand factories analysed by Williamson, and to continue in operation long after more technologically advanced factories had come into being (Landes 1969:118–19). More generally Williamson, unlike Marglin and Landes, offers no explanation as to why the transition from putting out to the hand factory should have suddenly speeded up during the closing decades of the eighteenth century. Indeed, his comments on the bankruptcy of technological approaches to work organisation seem to indicate a certain myopia on his part, so much so that it has been suggested that he is merely attempting to replace ‘the determinism of technological production efficiencies with the determinism of transaction cost efficiencies’ (Englander 1988:339). Traditional economic historians, who for many years have grappled with the trade-offs that existed between cheap labour, transaction cost efficiencies and technology, would probably agree.

Transport improvements, transaction costs and technology A recent attempt has been made by Szostak to provide the Williamson thesis with the temporal dimension that it has hitherto lacked. He suggests that while putting out may have been a transaction-cost-minimising method of production until the middle of the eighteenth century, improvements in transport thereafter tended to be biased in favour of the hand factory, which then took over as the least-cost mode of production. Technology only accelerated the shift to the factory at a later date, by which time the factory system (based on hand factories) was fairly widespread (Szostak 1989:346–7). Szostak uses the eleven criteria proposed by Williamson as a basis for determining the way in which transportation improvements affected the efficiency of the two systems. He regards the debate about the absence of

39

S.R.H.JONES

weights and scores assigned as being irrelevant for his purposes—he is only concerned with relative changes in efficiency. Szostak concludes that, ‘In terms of ten of the eleven criteria, transport improvements served either to make factories more advantageous or less disadvantageous’ (Szostak 1989:345– 9). Transportation and not technology, therefore, explains why the factory system emerged in the late eighteenth century. But were transport improvements biased in favour of the factory? Like Williamson, Szostak bases his conclusions not on a careful appraisal of historical data, but on impressionistic judgements that are arrived at through a combination of a priori reasoning and limited empirical evidence. Indeed, Szostak admits that to ‘prove’ that the net effect of transport improvements during this period was in favour of factories would require extensive data on the costs of putting out and early factories, and such data does not exist (Szostak 1989:351). Szostak may be criticised, therefore, on the grounds that his argument relies more upon logic than fact and reflects what may have happened given certain assumptions as opposed to what actually happened in practice. Consequently whether one believes that his hypothetical constructs reflect reality, or not, is largely a matter of faith. Yet even if one accepts the contention that the factory system was the major beneficiary of transport improvements, it does not necessarily follow that costs fell far enough relative to putting out to offset the lower labour costs of the latter. To establish that this was the case, Szostak has first to provide some estimate of the initial cost advantages of putting out and, second, to demonstrate that changes in relative costs due to transport improvements were sufficient to overcome those advantages. This cannot be done without an explicit consideration of labour costs and the way they also changed during the period under consideration. Szostak, however, appears unconvinced by the wealth of scholarship pointing to the existence of a reserve army of cheap outwork labour and seems disinclined to consider the effects that declining labour costs might have upon the competitive position of the putting out system (Szostak 1992:396). Finally, it must be said that the weight of empirical evidence does run counter to Szostak’s thesis. If transport improvements did provide the hand factory with such a competitive advantage after the middle of the eighteenth century, why did the putting out system continue to expand so vigorously? Why in Yorkshire, where the number of broad and narrow cloths stamped annually increased by 125,000 pieces between 1791 and 1805, did factories account for only 8,000 of those pieces? (House of Commons Committee on the Woollen Manufactures of England 1806:11–13). Szostak, taking refuge in a priori reasoning, suggests that the continued coexistence of hand factories and putting out was probably due to the fact that diffusion commonly follows an S-shaped path and it simply took time to learn about the advantages of factory organisation (Szostak 1992:397). The real answer, it seems, is because in Yorkshire, as elsewhere, the putting out system was still able to deliver 40

THE RISE OF THE FACTORY SYSTEM IN BRITAIN

product at a competitive price, notwithstanding the improvements in transportation that had taken place.

Conclusion Monocausal explanations in history should always be viewed with suspicion, especially where there are ambiguities in both evidence and terminology. Nevertheless, the argument that the rise of the ‘modern factory system’ in the latter part of the eighteenth century was largely the result of the introduction of new and more efficient technology would appear to be substantially correct. This is not only a reflection of the fact that authors such as Mantoux and Landes define the ‘modern factory system’ in terms of new technology, but because the majority of factories that appeared from the 1780s onwards incorporated new technology which, because of both size and/or power requirements, could not be easily located in the cottage. At the same time, it must be said that there is a body of evidence that lends credence to the views of Marglin and Williamson. Some manufacturers clearly did turn to the factory to cut costs, both by extracting greater output from workers and by economising on transaction costs. In spite of such evidence, the arguments of Marglin and Williamson are not persuasive. The processes they outline, while occurring occasionally, fail to reflect what generally occurred in late eighteenth-and early nineteenth-century England. They focus on the atypical rather than the typical and build logically sound but historically inaccurate models to justify their central theses. Landes has dubbed this type of economic reasoning ‘economoneirics’ or dream economics—what might have happened but in actual fact did not (Landes 1986:591). This is not to imply that the organisation of production was completely technologically determined, and Landes would doubtless not wish for such an inference to be drawn from his work. Yet while the factory, both hand and powered, afforded scope for the exploitation of labour, one cannot understand the rise of the factory system without a proper appreciation of the vital role of technology.

Notes 1 Langlois examines some of these issues in Chapter 2 of this volume, in an extended discussion of Clark’s paper.

References Aiken, J. (1968) Description of the Country from Thirty to Forty Miles Round, Manchester, Newton Abbot: David and Charles.

41

S.R.H.JONES

Allen, G.C. (1966) The Industrial Development of Birmingham and the Black Country: 1860–1927, London: Frank Cass. Ashton, T.S. (1962) The Industrial Revolution 1760–1830, London: Oxford University Press. Aspin, C. and Chapman, S.D. (1964) James Hargreaves and the Spinning Jenny, Preston: Helmshore Local History Society. Berg, M. (1991) ‘On the Origin of Capitalist Hierarchy’, in B.Gustafsson (ed.) Power and Economic Institutions, Aldershot: Edward Elgar. Berg, M. and Hudson, P. (1992) ‘Rehabilitating the Industrial Revolution’, Economic History Review 45:24–50. Bischoff, J. (1968) A Comprehensive History of the Woollen and Worsted Manufactures, vol. 1, London: Frank Cass. Chaloner, W.H. (1963) People and Industries, London: Frank Cass. Chapman, S.D. (1972) The Cotton Industry in the Industrial Revolution, London: Macmillan. Clark, G. (1994) ‘Factory Discipline’, Journal of Economic History, 54:128–63. Clarkson, L.A. (1971) The Pre-industrial Economy of England, 1500–1750, London: Batsford. Coleman, D.C. (1992) Myth, History and the Industrial Revolution, London: Hambledon. Cooke-Taylor, R.W. (1891) The Modern Factory System, London: Paul, Trench, Tribner. Daniels, G.W. (1930) ‘Samuel Crompton’s Census of the Cotton Industry in 1811’, Economic History 2:107–10. Englander, E.J. (1988) ‘Technology and Oliver Williamson’s Transaction Cost Economics’, Journal of Economic Behaviour and Organization 10:339–53. Flinn, M.W. (1962) Men of Iron, Edinburgh: Edinburgh University Press. Freudenberger, H. and Redlich, F. (1964) ‘The Industrial Development of Europe: Reality, Symbols, Images’, Kyklos 17:372–403. Furniss, E.S. (1957) The Position of the Labourer in a System of Nationalism, New York: Kelley and Millman. Greener, W.W. (1910) The Gun and its Development, London: Cassell. Hammond, J.L. and Hammond, B. (1919) The Skilled Labourer, London: Longman. Harte, N.B. (1977) ‘The Growth and Decay of a Hosiery Firm in the Nineteenth Century’, Textile History 8, 7–55. Heaton, H. (1965) The Yorkshire Woollen and Worsted Industries, Oxford: Clarendon Press. House of Commons Committee on the Woollen Manufacture of England (1806), British Parliamentary Papers 1806, III. Hudson, P. (1983) ‘From Manor to Mill: the West Riding in Transition’, in M. Berg, P.Hudson and M.Sonenscher (eds) Manufacture in Town and Country before the Factory, Cambridge: Cambridge University Press: 124–44. ——(1992) The Industrial Revolution, London: Edward Arnold. Jenkins, D.T. (1973) ‘The Validity of the Factory Returns 1833–50’, Textile History 4:26–46. ——(1975) The West Riding Woollen Industry, 1780–1835: A Study in Fixed Capital Formation, Edington: Pasold. Jones, S.R.H. (1976) ‘Hall, English & Co., 1813–41: A Study of Entrepreneurial Response in the Gloucester Pin Industry’, Business History 18:35–65. ——(1980) ‘John English & Co., Feckenham: a Study of Entrepreneurship in the British Needle Industry in the Eighteenth and Nineteenth Centuries’, unpublished PhD thesis, London University. ——(1982) ‘The Organization of Work’, Journal of Economic Behaviour and Organization 3:117–37.

42

THE RISE OF THE FACTORY SYSTEM IN BRITAIN

——(1983) ‘Technology and the Organization of Work: a Reply’, Journal of Economic Behaviour and Organization 4:63–6. ——(1987) ‘Technology, Transaction Costs and the Transition to Factory Production in the British Silk Industry, 1700–1870’, Journal of Economic History 47: 71–96. ——(1992) ‘The Emergence of the Factory System in Eighteenth Century England’, Journal of Economic Behaviour and Organization 19:389–94. Kennedy, J. (1819) ‘Observations on the Rise and Progress of the Cotton Trade in Great Britain’, Literary and Philosophical Society of Manchester: Memoirs and Proceedings 3(25):115–37. Kirby, M.W. (1994) ‘Big business before 1900’, in M.W.Kirby and M.B.Rose (eds) Business Enterprise in Modern Britain from the Eighteenth to the Twentieth Century, London: Routledge. Landes, D.S. (1969) The Unbound Prometheus: Technological Change and Industrial Development in Western Europe, 1750 to the Present, Cambridge: Cambridge University Press. ——(1986) ‘What Do Bosses Really Do?’, Journal of Economic History 46:585–62. Lazonick, W. (1979) ‘Industrial Relations and Technical Change: the Case of the SelfActing Mule’, Cambridge Journal of Economics 3:231–62. Mantoux, P. (1961) The Industrial Revolution in the Eighteenth Century, London: Cape. Marglin, S. (1974) ‘What Do Bosses Do?’, Review of Radical Political Economy 6: 60– 112. Mathias, P. (1983), The First Industrial Nation, London: Methuen. Nelson, E.G. (1929–30) ‘The Putting-Out System in the English FrameworkKnitting Industry’, Journal of Economic and Business History 2:467–94. Patterson, R. (1967) ‘Spinning and Weaving’, in C.Singer, E.J.Holmyard, A.R. Hall and T.I.Williams (eds) A History of Technology, London: Oxford University Press. Podmore, F. (1923) Robert Owen: A Biography, London: Hutchinson. Pollard, S. (1959) A History of Labour in Sheffield, Liverpool: Liverpool University Press. ——(1968) The Genesis of Modern Management, Harmondsworth: Penguin. Reports from Assistant Commissioners, Handloom Weavers, British Parliamentary Papers, 1840, XXIV. Robertson, P.L. and Alston, L.J. (1992) ‘Technological Choice and the Organization of Work in Capitalist Firms’, Economic History Review 45:330–49. Rose, M.B. (1989) ‘Social Policy and Business; Parish Apprenticeship and the Early Factory System’, Business History 31, 5–29. Rowlands, M.B. (1975) Masters and Men in the West Midlands Metalwares Trades before the Industrial Revolution, Manchester: Manchester University Press. Sabel, C. and Zeitlin, J. (1985) ‘Historical Alternatives to Mass Production: Politics, Markets and Technology in Nineteenth Century Industrialization’, Past and Present 108, 133–76. Samuel, R. (1977) ‘Workshop of the World: Steam Power and Hand Technology in Mid-Victorian Britain’, History Workshop 3:6–72. Select Committee on the Silk Trade (1831–2) British Parliamentary Papers 1831–2, XIX. Styles, J. (1983) ‘Embezzlement, Industry and the Law in England, 1500–1800’, in M.Berg, P.Hudson and M.Sonenscher (eds) Manufacture in Town and Country Before the Factory, Cambridge: Cambridge University Press: 173–205. Szostak, R. (1989) ‘The Organization of Work: the Emergence of the Factory Revisited’, Journal of Economic Behaviour and Organization 11:342–58.

43

S.R.H.JONES

——(1992) ‘Transport Improvements and the Emergence of the factory’, Journal of Economic Behaviour and Organization 19:395–9. Thirsk, J. and Cooper, J.P. (1972) Seventeenth Century Economic Documents, Oxford: Clarendon Press. Timmins, S. (ed.) (1967) Birmingham and the Midland Hardware District, London: Frank Cass. Ure, A. (1967) The Philosophy of Manufactures, London: Frank Cass. Usher, A.P. (1921) An Introduction to the Industrial History of England, London: Harrap. Wadsworth, A.P. and Mann, J. de L. (1965) The Cotton Trade and Industrial Lancashire, Manchester: Manchester University Press. Williamson, O.E. (1980) ‘The Organization of Work’, Journal of Economic Behaviour and Organization 1:5–38. ——(1983a) ‘Technology and the Organization of Work: a Reply to Jones’, Journal of Economic Behaviour and Organization 4:57–62. ——(1983b) ‘Technology and the Organization of Work: a Rejoinder’, Journal of Economic Behaviour and Organization 4:67–8. Wing, C. (1967) Evils of the Factory System Demonstrated by Parliamentary Evidence, London: Frank Cass.

44

2

THE COEVOLUTION OF TECHNOLOGY AND ORGANISATION IN THE TRANSITION TO THE FACTORY SYSTEM Richard N.Langlois 1

Introduction Within the last two decades, the question of the origin and nature of the factory system has leapt from obscurity to fill thousands of pages. The seminal article was, of course, Marglin’s ‘radical’ interpretation of factory organisation, a paper now twenty years old. But Marglin’s broadside arguably aroused as much interest as it did because the questions it addressed were quite congenial to those in which the larger profession was becoming increasingly interested, namely, questions of institutions and organisation. As exemplified in the work of Douglass North in economic history and Oliver Williamson in the economics of organisation, this New Institutional Economics, as it was coming to be called, offered a fresh viewpoint on the nature of capitalist organisation during the Industrial Revolution. Economic historians like David Landes and S.R.H.Jones also took up the cudgels, adding historical insight and a perspective typically rather different from that of either the ‘radicals’ or the New Institutionalists. Despite the complexity and subtlety of the conversation, it might none the less be helpful to summarise the arguments in a simple schema. First of all, the questions, it seems to me, move along two different dimensions. The first dimension is what we may call that of origins: what caused the factory system to emerge? The second dimension is what we make call that of raisons d’être: what is the nature or essence of the factory system, and how do we characterise its cause? Figure 2.1 summarises the possibilities. 45

RICHARD N.LANGLOIS

Organisation

Technology

Efficiency

Williamson (1980)

Landes (1986)

Exploitation

Marglin (1974)

Marx (1867)

Figure 2.1 Explanatory alternatives

Along the horizontal dimension lies the issue of origins: did the factory system emerge because of its organisational form, or did it spring from new technology, notably centralised motive power? Along the vertical dimension is the issue of raison d’être: did the factory system emerge because it was more efficient than what went before, or did it emerge because capitalists found themselves able to use factory organisation as a mechanism for worker exploitation? The traditional Marxian view is that the essence of capitalism is, well, capital. What characterises the capitalist system is the mode of production— the technology—and it is technology that enables the capitalist to create and appropriate surplus value.2 What was remarkable about Marglin’s assault on capitalist work organisation was his rejection of the Marxian insistence on machinery as the engine of exploitation. For Marglin, it was the organisation of work, not the technology, that mattered. By subdividing tasks in the manner advocated by Smith in the Wealth of Nations, capitalists could deskill work, rendering each task so simple that an undifferentiated and untrained proletariat could replace skilled artisans. But the capitalists divided labour not because this process is more efficient than crafts production but because deskilling allows the capitalist to control workers more effectively—and therefore to reap a larger fraction of the joint surplus of production.3 Writers like Williamson (1980) and North (1981) also view the arrival of the factory system as a matter of organisation. But they see that system as emerging because of greater efficiency, which they understand largely in terms of the minimisation of transaction costs, especially the costs of material lost to embezzlement, the costs of coordinating a finely subdivided process, and the costs of monitoring product quality. Economic historians like Jones (1982, 1987, 1993) and Landes (1986) have criticised both Marglin and the transaction-cost theorists for a comparative lack of attention to history. And, despite all the arguments of a priori theory, history demonstrates, they assert, that it was the superior technology associated with centralised power sources that triggered the factory system.4 ‘No,’ writes Landes (1986:606), ‘what made the factory successful in Britain was the muscle: the machines and the engines.

46

THE COEVOLUTION OF TECHNOLOGY AND ORGANISATION

We do not have factories until these were available, because nothing less would have overcome the cost advantage of dispersed manufacture.’ This essay argues that, although proponents of the ‘efficient technology’ argument are certainly closest to the truth, none of these alternatives has it completely right. In the end, the explanation for the rise of the factory system does in fact lie in the realm of organisation, but not in the qualities of organisation envisaged by either the ‘radical’ view or the transaction-cost view. The factory system arose because growth in the extent of the market (for textiles principally, but eventually most other goods as well) opened up entrepreneurial possibilities for high-volume throughput. This meant not only an extended division of labour but also investment in new capabilities (including, but not limited to, capital equipment) that, by making production more routine, permitted lower unit costs. For reasons that we will see, these new capabilities implied high fixed costs, at least initially, and it was these fixed costs that called for the ‘factory’ mode of organisation. Was this efficiency or exploitation? Efficiency, without doubt. But the problem of explanation is a subtle one, and this essay closes with some musings on the logic of both efficiency and exploitation.

What is a factory? We need to begin by establishing the meanings of terms. First and foremost: what is a factory and what is the factory system? There are a number of characteristics, operating both singly and in conjunction, that one might offer as distinctive of the factory. Principal among these are: • • •

expensive or indivisible technology; the concentration of workers in a single location; and close monitoring or supervision of work.

As Fang (1978:16) suggests, the archetypal factory had all three. Does any of these by itself define a factory? The idea that large-scale central-power technology defines the factory is an idea that goes back at least to Ure (1861:13). From the point of view of this essay, however, defining the factory by the use of large-scale, expensive, or indivisible technology rather begs the question. Moreover, there are at least some examples—notably the famous cottage factories in the silk industry (Jones 1987:90)—suggesting that it is possible, if perhaps just barely possible, that indivisible central power could coexist with the putting out system. (There are plenty of examples, however, in which indivisible central power is fully compatible with inside contracting, a point to which I will return.) Conversely, as Axel Leijonhufvud (1986:205) has noted, if centralised power

47

RICHARD N.LANGLOIS

defines the factory system, are we not compelled to wonder why the factory system remained alive and well in the era of small electric motors? The agglomeration of workers in a single facility is also not a definition of the factory. Here, too, there are plenty of examples, going back at least to the Arsenal of Venice (Lane 1973), of clusters of workers that we would not want to classify as factories, at least not in the sense of the British factory system of the Industrial Revolution. Indeed, to the extent that the workers act as independent contractors, the resulting inside contracting system (Buttrick 1952) is in many ways closer to the putting out system than it is to the factory system. There is, of course, the issue of whether the contractor or the capitalist owns the tools of production. In the former case, one might want to say that labour (the contractor) hires capital (buys his or her own tools), whereas in the other case capital hires labour. Inside contracting when the capitalist owns the machinery, as in the case of mule spinning in Lancashire in the late eighteenth and early nineteenth centuries (Lazonick 1990:80–5), obviously comes closer to a fully-fledged factory than does inside contracting when the contractor supplies the tools. Indeed, one often hears the Marx-inspired criterion of capital hiring labour touted as the defining characteristic of the factory system (not to say of capitalism as such). And we might well want to describe as a factory a Lancashire mule-spinning establishment in which master spinners use the capitalist’s machines, power, and materials to produce yarn on a piece-rate basis. Yet, there is also arguably something more to the factory system. An equally strong tradition holds that what is essential about the firm, if not necessarily the factory, is that the contract between worker and capitalist within a firm is not a simple contract over output. For Coase (1937) and his followers, there is an essential difference between a spot-contract for product and an employment contract. In the former, it is relative prices that matter; in the latter, it is authority—or so many have interpreted it—that matters: ‘If a workman moves from department Y to department X, he does not go because of a change in relative prices, but because he is ordered to do so’ (Coase 1937:387). It is for this reason that Williamson (1975:71–2), following Simon (1957), characterises the employment contract as an ‘authority relation’—a loaded term5—and the capitalist firm as a ‘hierarchy’. In much of the literature on the emergence of the factory system, indeed, ‘capitalist hierarchy’ is assumed to be the explanandum (see, e.g., Berg 1991). There is both truth here and confusion. One wouldn’t want to dispute that the capitalist firm is a hierarchy, in one or more senses. Surely the boss ‘tells the worker what to do’, and this is crucial. But also crucial is the difference between entrepreneurship and supervision. 6 In Simon’s formulation of the ‘authority relation’, the capitalist pays a wage for the right to choose which action x ε Ω the worker will perform at any time, where Ω is the ‘job description’ or set of allowable actions to which the worker agrees. As in Coase, the accent here is on the flexible assignment 48

THE COEVOLUTION OF TECHNOLOGY AND ORGANISATION

of workers to tasks. Langlois and Robertson (1995) argue that economic change, which necessitates the flexible redeployment of economic capabilities in order to capture entrepreneurial opportunities, is a vital and neglected aspect of the theory of vertical integration and disintegration. And, as Peter Temin (1991) has emphasised, the neglect of this function of entrepreneurial coordination accounts in large measure for the inability of the ‘radical’ critics of capitalism to detect a non-exploitive function for those they indiscriminately call ‘bosses’. But this understanding of ‘capitalist hierarchy’ as flexible redeployment is also far from the experience of workers in the early factories of the Industrial Revolution. The key point—and here we come finally to the essence of the definition—is that the factory system consists in a change (relative to the putting out system or the inside contracting system) in the nature of the supervision exercised by the capitalist. Rather than monitoring output, as the putter out or merchant capitalist does with a contractor, the factory capitalist (or, more likely, his hired supervisor) monitors the work process itself. That is to say, the crucial difference between the merchant capitalist and the factory capitalist is that the latter exerts factory discipline (Pollard 1963, 1965). Now, one can argue that factory discipline also does not by itself define the factory system. There was plenty of ‘factory discipline’ under the putting out system. The discipline—the monitoring of the work process itself—was the province of the master of the cottage, whose charges were typically members of his own family as well as some casual labourers.7 A putting out cottage or artisan workshop was thus a factory by this definition. So the transition to the factory system represented not a shift away from supervision of the work process per se but a shift in the locus of that supervision from the subcontracting cottage master to the factory owner (Cohen 1981). Figure 2.2 summarises the possibilities.

Work force concentrated

Work force dispersed

Process supervision

Factory system

–

Product monitoring

Inside contracting

Putting out

Figure 2.2 Organisational alternatives

49

RICHARD N.LANGLOIS

Division of labour, routine and technology We can think of the putting out system and the factory system as alternative institutional trajectories,8 and the problem of explaining the rise of the factory system as a problem of explaining Britain’s transition from one trajectory to the other. As an institutional structure, the putting out system offered a number of advantages. Principal among these was low labour cost. Apart, in the early years, from evading urban guild regulations, the merchant capitalist or putter out could take advantage in the countryside of surplus labour time made available by the seasonal nature of agriculture. The rural location of work also meant that cottagers could keep to some agricultural pursuits, thus lowering their subsistence needs from outside sources and further reducing labour costs relative to urban areas. Moreover, as the cottager owned his or her own tools, and capital requirements were low in any case, putting out was a strategy that offered the advantage of flexibility: in times of low demand, the capitalist had little in the way of fixed costs to cover. The transaction-cost theorists, however, point to some of the shortcomings of this system. The very dispersion of work made monitoring difficult, encouraging, in particular, embezzlement of materials, which the domestic worker could then either resell or work up on his or her own account. The embezzlement was typically covered up by reducing the quality rather than the quantity (which could be more easily measured) of the finished product. As we saw, North and Williamson see the superiority of the factory in light of the easier monitoring of ‘inside’ production. But Jones (1982, 1993), among others, has disputed the importance of embezzlement, noting that the merchants compensated for expected embezzlement with lower prices and certain other tricks like the truck system, which required the workers to take their compensation in kind. Moreover, it is not clear that the benefits of avoiding embezzlement and shoddy work outweighed the advantages of putting out. When it was worth it—when the material, such as Spanish wool, was especially valuable, or when problems of quality control were especially serious— workshops did indeed spring up (Pollard 1965:33). That there were few examples of this in the heyday of the putting out system suggests that, for the most part, embezzlement costs did not outweigh the benefits of low labour costs and flexibility. There is a message here. Although transaction-cost theorists understand in principle that evaluating relative efficiency is a matter of counting up both transaction costs and production costs (Williamson 1985:103), in practice analysts often forget the production-cost part—and production costs frequently turn out to be decisive (Langlois and Robertson 1995: ch. 3). There is a more important point. This process of evaluating the relative efficiency of institutions—comparative-institutional analysis, as it is called in the Coasean tradition—is almost always conceived of as a static exercise.

50

THE COEVOLUTION OF TECHNOLOGY AND ORGANISATION

Seldom do the evaluators consider the rates of change of the relevant variables along with their magnitudes. (Put less neoclassically: they don’t seriously consider history.) What is significant about the transaction costs of the putting out system is not so much the costs of embezzlement but the rate of change in those costs. Before the second half of the eighteenth century, embezzlement and deteriorating quality were not serious problems. What made them problems—or, more correctly, symptoms of a far larger problem—was the increasing demand for the products of the outworkers, especially spun yarn,9 as final demand for fabric accelerated. Landes (1969:57ff.) argues, indeed, that by the fourth quarter of the eighteenth century, the putting out system was reaching its limits. Although flexible in down-turns, the system was difficult to crank up in the face of predictable, secularly increasing demand. The possibilities for geographical expansion had been exhausted, and pressure at the intrinsic margin—output per worker—was met, Landes tells us, with a backward-bending supply-ofeffort curve. Indeed, embezzlement was largely a reaction by the outworkers to the capitalists’ attempt to lower real piece rates through indirect means as diminishing returns set in (Landes 1969:59). Others would dispute the extent to which the putting out system had reached exhaustion in this period.10 Labour supply was growing, transportation costs were falling, and in many sectors the extent of putting out was growing both before and after the Napoleonic Wars. In the end, ‘exhaustion’ is a relative matter.11 And it is more than arguable that the growing extent of the market for manufactured goods had begun by the late eighteenth century to make profitable an alternative technological trajectory opened up by the invention of water-and steam-powered machinery. This was nowhere more significant than in cotton fabrics, the industry that became the avatar both of British manufacturing and of the factory system itself (Fang 1978). But the mechanism by which increases in demand led to or triggered the move to the factory system remain obscure—or at any rate subtle. With the geographic margin of the putting out system arguably reaching (if not having already reached) the point of diminishing returns, there remained two other margins on which to push: the workers’ level of effort and the organisation of production. (What about technology? I’ll come back to that.) And here Marglin enters the picture. One aspect of his argument is to draw our attention to the usefulness of factory organisation in pushing along the effort margin. Factory discipline can get more effort out of a given labour force, and in that way break the bottleneck of the putting out system to the owners’ (but not, of course, the workers’) advantage. We will look at this argument more closely in the next section. Notice here, however, that Marglin neglects the organisational margin. That is, he does not see the reorganisation of production in the factory (of which factory discipline may play a part) as another way of attacking the cost bottleneck of the putting out system. Organisation, for Marglin, is merely a stratagem that allows the capitalists to 51

RICHARD N.LANGLOIS

exert pressure on worker effort, and it conveys no efficiency benefits in its own right. Needless to say, there is reason to think that capitalists pushed on all margins simultaneously, and that the organisation margin yielded considerably. In what way did organisation change? By Adam Smith’s famous theorem (or its converse, at any rate), the increasing demand for textiles in the late eighteenth century should have called for increasing division of labour. And this, in turn, should have led to a more intricate sequencing of tasks. In the pinshop model, the time-sequencing of tasks becomes crucial, as one worker’s output is the input to the next worker. By monitoring the work process, the capitalist can make the workers work at the system’s pace rather than at their own, assuring that intermediate product flows smoothly between stages. Thus does Williamson (1980) argue the superiority of the factory system in part on the grounds that it economises on work-in-process inventories relative to an (inside or outside) contracting system. This is not implausible. Buttrick (1952), for example, lays the demise of inside contracting in the American small-arms industry of the nineteenth century largely to inefficient inventory systems.12 On the other hand, Clark (1994) has calculated that the cost of work-in-process inventories would in fact have been unimportant in the factories of the Industrial Revolution. Leijonhufvud (1986) suggests another reason why the division of labour may have led to the factory system. In the pinshop model, all the workers become complementary to one another, in contrast to crafts artisans, who are substitutes in production. This complementarity means that, if the workers owned their own tools, they could individually threaten to withdraw their capital from the production process in order to capture a larger share of the joint rents of production.13 This is the phenomenon of ‘hold up’ familiar in the transaction-cost literature (Klein et al. 1978). If, however, the physical capital were pooled under common ownership—and capital hired labour instead of the other way around—this problem would disappear (or be replaced, at any rate, with the problem of bargaining with a labour union). This does not explain, however, the existence of process monitoring in Fang’s archetypal factory, since, as we saw, the fact of capital hiring labour does not speak to the nature of the contract between capital and labour, and is perfectly consistent with inside contracting. Does this mean that organisational advantages do not explain the factory system? If we take organisational advantages to mean the transaction-cost problems of the division of labour, as those terms are usually understood, the answer is probably that they do not. If, however, we broaden our field to mean by organisational advantages an imperative of which the division of labour is itself only derivative, then organisation does indeed matter. To see what this means, let us consider the process of production more carefully. Under crafts production, labour is undivided in the sense that each artisan performs a wide range of tasks. This requires a relatively large 52

THE COEVOLUTION OF TECHNOLOGY AND ORGANISATION

investment in human capital, since, to be proficient, the artisan must be accomplished in a wide variety of skills or subskills. Crafts production also implies a certain kind of flexibility and a lack of standardisation, since the artisan controls the ‘interfaces’ between tasks and the connections between parts. If, with Nelson and Winter (1982), we think of production as a matter of exercising and choosing among certain ‘routines’,14 then crafts production requires the possession of and the ability to choose among a wide range of possible routines (Stinchcombe 1990: ch. 2). Crafts production thus obviously has advantages when production runs are small, for reasons of both demand and supply. On the demand side, as Smith reminds us, the division of labour is limited by the extent of the market, and, if ‘the number of potential buyers of a commodity were too small, it would not be possible to dispose of the increased output which differentiation permits, forcing a worker to perform several activities in order to earn enough to fend off star-vation’ (Robertson and Alston 1992:331). On the supply side, crafts production may be necessary or advantageous when the production process involves uncertainty, in the sense that the choice of routines must be fitted interactively to changing particular circumstances (Stinchcombe 1990:66–70). We can think of a spectrum of skill levels.15 At one end of the spectrum are deskilled—or, at any rate, unskilled—factory workers. These operatives have a small repertoire of routines, and they engage in a restricted range of active choice within that repertoire. In other words, unskilled workers perform routine activities (in the less-technical sense of the term). At the other extreme are professionals—physicians, architects, attorneys, academics—who must have large repertoires of routines and who must be able to choose deftly among routines to fit changing particular circumstances. In addition, professionals also engage in innovation, the introduction of new routines (Savage 1994). In between are the semiskilled occupations, like tradesmen— carpenters, plumbers, drywallers, electricians—or the crafts artisans of the eighteenth century. These workers must also choose among routines flexibly, but both the size of the repertoire and the range of application of the routines is more restricted. Semiskilled workers also are less likely than skilled workers to innovate routines. Obviously, artisans in crafts production are more difficult to monitor directly than are factory operatives. Indeed, as Minkler (1993) argues, workers— especially skilled ones—may possess knowledge that is qualitatively different from that of supervisors, making monitoring costly even in the absence of principal—agent problems of the standard neoclassical sort.16 It is not surprising, therefore, that, as skill level increases, workers are less likely to be employees (supervised in process) and are more likely to interact with the market through subcontracting relations (monitored in product by relative prices).17 I have argued that the key trigger—I will postpone using the word ‘cause’— of the transition to the factory system was the secular increase in demand for 53

RICHARD N.LANGLOIS

the products of manufacturing. What is significant here, however, is not only the extent of the market but also the predictability of the market. When the extent of the market for a product increases, especially if it does so without much fluctuation, the production process becomes less uncertain, in the sense that the selection of productive routines requires less interactive tailoring to particular circumstances.18 This reduction in uncertainty leads to two distinct effects, only one of which is captured in the traditional notion of the division of labour. I will call these the division-of-labour effect and the volume effect. The former is much discussed, if not always well understood. As we have seen, it is only when flexible interactive selection among routines is no longer necessary that labour can be divided in the manner Smith advocated. Each worker can concentrate narrowly and deeply on a smaller subset of the routines necessary for production precisely because the function of selection among the routines becomes effectively hard-wired into the system. Variability in the pace of individual workers can introduce a mild kind of uncertainty, but one, as we have also seen, that can be ‘buffered’ (in Stinchcombe’s terms) by work-in-process—or buffer—inventories, the cost of which may or may not be significant to the choice of monitoring system. In the Smithian story, labour starts out skilled (crafts production) but tools are specialised; with the division of labour, labour specialises (tools remaining specialised) and, through differentiation spurred by innovation, perhaps increases its level of specialisation. This does not exhaust the possibilities, however. It is also possible, through mechanical innovation, for tools to integrate previously separate tasks (Robertson and Alston 1992) and, in general, for machines to become more ‘skilled’, that is, to have a larger repertoire of routines (Ames and Rosenberg 1965). For example, to the extent that the advent of the self-acting mule after the 1830s ‘deskilled’ the spinner (that is, required less skill in our sense than the common mule), it did so not because it subdivided labour more finely but because the machine itself became more skilled.19 Notice that, like the subdivision of tasks, the introduction of more-skilled machinery requires both increased volume of output and predictability of output. Consider the simple jig. With a reduction in uncertainty—permitting an increase in standardisation—the sequencing of choice among routines can be hard-wired into a machine.20 In drilling the plate A without the jig the skilled mechanic must expend thought as well as skill in properly locating the holes. The unskilled operator need expend no thought regarding the location of the holes. That part of the mental labor has been done once for all by the tool maker. It appears, therefore, that a ‘transfer of thought’ or intelligence can also be made from a person to a machine. If the quantity of parts to be made is sufficiently large to justify the expenditure, it is possible to make machines to which all the required 54

THE COEVOLUTION OF TECHNOLOGY AND ORGANISATION

skill and thought have been transferred and the machine does not require even an attendant, except to make adjustments. Such machines are known as full automatic machines. (Kimball 1929:26, emphasis original) This transfer of skill to machines is another manifestation of the process that motivates the division of labour, namely, the increasing routine and standardisation of production. It is also an aspect of what I have called the volume effect. It is not, however, the only aspect. As the extract above suggests, the transfer to a machine of ‘intelligence’—that is, the ability to select among operational routines—often takes the form of a jig, pattern or die. And, as Alchian (1959) points out, the ‘method of production is a function of the volume of output, especially when output is produced from basic dies—and there are few, if any, methods of production that do not involve “dies” (Alchian 1959 [1977:282], emphasis added). Why? Because, with increased volume, it pays to invest in more durable dies. Consider the example of printing. If one is going to run off a few copies of a memo, a photocopy machine will do the trick. If one needs several hundred copies of documents on an ongoing basis, it might be worth investing in a small offset press. For even larger predictable production runs, it would pay to have a more serious printing press. As volume and predictability allow greater ‘durability of dies’, unit costs decline. This is an effect of growth in the extent of the market distinct from the division of labour narrowly understood. In the case of cotton textiles during the Industrial Revolution, it is arguable that the volume effect was more important than increases in the division of labour. For one thing, as Pollard (1965:34) and others have noted, the division of labour was one of the benefits that originally recommended the putting out system. And, although there was surely room for further division of labour in factories, most of the change in technology and organisation in cotton was in fact arguably of a sort that increased the skill of machines rather than more finely subdividing tasks. Moreover, the history of technological change in cotton textiles is one directed very much toward what we could call greater durability of ‘dies’. The spinning jenny, the waterframe, and later the mule, were ways of multiplying for many bobbins simultaneously the routines of the spinning wheel. Innovations in weaving, printing,21 and other departments could be described in a similar way.

‘Durability’, fixed costs and supervision Obviously, if increasing extent of the market led to what I have called the volume effect—more highly-skilled machines embodying more durable ‘dies’—

55

RICHARD N.LANGLOIS

then there is likely some connection between the extent of the market and the factory system. The precise nature of that connection, however, requires some elucidation. The first missing link in the argument is the relationship between the volume effect and fixed costs. It is far from implausible to postulate that, ceteris paribus, as the skill and durability-as-die of machines increases, so do fixed costs. For graphic simplicity, Figure 2.3 displays the relationship between throughput (which I will use as a shorthand for the volume effect) and fixed costs per unit as linear, but the second derivative of the relationship will likely depend in fact on the particular technology and industry under consideration. The upward-sloping relationship holds at any particular planning date t. Over time, however, the curve is likely to shift down. That is, with innovation and learning in the production process and the machinery industries that supply it, the costs of providing any particular level of durability will decline. Only in an atemporal sense, then, does increased throughput imply higher fixed costs.22 Recall that factory organisation means not only workers concentrated in a single location but also the direct supervision of work. By elaborating on a couple of recent models of worker effort and organisation (Lazonick 1990, Clark 1994), we can generate several different arguments for why increased fixed costs might lead to factory organisation. Consider Figure 2.4. MP0 is the marginal productivity of labour (equal to

Figure 2.3 Throughput and fixed costs 56

THE COEVOLUTION OF TECHNOLOGY AND ORGANISATION

the wage in competitive equilibrium) as a function of effort, assumed linear for convenience, for the representative firm with fixed costs F0.23 If the labour market is indeed competitive, and each worker’s marginal product is separately observable, then the firm will offer a piece-rate contract that rewards workers according to marginal product (and is thus identical to MP0); and the representative worker with utility function U(e, w) will supply effort e0 and receive payment W0. In this world, there is no need for direct monitoring of the work process, and fixed costs don’t change that. A firm with fixed costs F1 (and, plausibly, a steeper marginal-product-of-labour curve) can keep the worker on the same indifference curve by offering wage w1, which elicits effort e1. The worker works harder and receives a higher wage, but there is no need for discipline. Obviously, this could change if marginal product were costly to determine. In that case, a piece-rate contract might be infeasible, and the capitalist would have to contract for an hourly wage. The worker would agree to supply (to firm 0) effort level e0 in exchange for wage w0. But, to the extent that monitoring of output is costly, the worker could reach a higher indifference curve by shirking and supplying less than e0. Direct monitoring of work in such a case may be less costly than the productivity

Figure 2.4 Wage and effort when marginal products observable Source After Clark (1994) 57

RICHARD N.LANGLOIS

forgone. Indeed, for firm 1, the marginal cost of shirking (the marginal productivity forgone) is greater because of the steeper slope of MP1. This qualifies as an explanation for the transition to the factory system, since it explains why process supervision (which requires centralised location) would eventually become economical as the extent of the market (and with it throughput and fixed costs) grew.24 It is a transaction-cost explanation, but one rather different from those offered by Williamson, North or Leijonhufvud. It comes closest, in fact, to the story told by Alchian and Demsetz (1972), in which the inability separately to meter individual marginal products leads to process monitoring by a specialist monitor, who, as residual claimant, is in turn monitored by market prices. The problem with the shirking explanation is that it relies on a specific kind of monitoring difficulty, namely indivisibilities in team production. In the textile industry, however, individual marginal products were arguably quite distinguishable, and, indeed, the success of the putting out system in this and other important industries suggests that there was no general monitoring-cost problem in offering piece rates.25 An exploitation explanation is in many ways the flip-side of the shirking story. Instead of the worker reducing effort below what was contracted for, in the simplest version of an exploitation story, the capitalist squeezes more effort out of the worker than was contracted for. This is essentially Marx’s idea: the capitalist pays the going (subsistence) wage for abstract labour power, but then must apply discipline to get the concrete labour out of the worker, the labour value of which concrete labour is more than the wage. Clark (1994) offers a slightly different interpretation of what he calls a ‘coercion’ account of factory discipline (see Figure 2.5).26 Firm 0 (perhaps the putting out system) is in initial equilibrium at point A. By increasing fixed costs, the capitalist shifts the marginal productivity of labour to MP1. It now pays to increase worker effort, which the capitalist does by introducing discipline. But the workers must be compensated by a higher wage w1, and the difference between w1 and w0 is a ‘disgust premium’ for submitting to discipline. This is exploitation in one sense, since the worker is not paid at marginal product and the capitalist pockets the surplus. Since point B is not a competitive equilibrium, however, one has to introduce a mechanism to keep wages from being bid up (and effort bid down) to marginal product. ‘Radicals’ (e.g., Marglin 1991:243) find it easy to assert that the worker ‘has no choice’; neoclassicals find it less easy to do so. (I return to this issue later in the chapter.) Why is factory discipline necessary at point B? Obviously, if the capitalist announces a wage contract of (w1, e1), the worker will have an incentive to accept but then to supply effort less than e1. But, as we saw, unless we introduce transaction costs that prevent cheap monitoring of worker output, there exists a piece-rate contract that will elicit effort e1 for payment of w1. (It would be given by the slope of a line tangent to the indifference curve at 58

THE COEVOLUTION OF TECHNOLOGY AND ORGANISATION

Figure 2.5 An exploitation explanation

B.) Clark’s account, however, is more interesting. Before we turn to it, consider a broader class of explanations (to which Clark’s belongs) in which the worker’s preferences do not appear as fully formed and given over the entire relevant space. In a model that is similar in many ways to that of Clark, Lazonick (1990) formalises an explanation of factory discipline hinted at by both ‘radicals’ and economic historians: a backward-bending supply curve of effort. Put in terms of the story we have been telling, the representative worker may have a utility function such that the capitalist will not be able to elicit higher levels of effort with pecuniary incentives. For example, in Figure 2.6, there is no wage less than or equal to marginal product, and therefore no piecerate contract, that will elicit an effort level, such as e1, that makes the higher-throughput technology (MP1) economical. Lazonick does not think of these preferences as immutable standards of economic welfare, however, but as ‘customary effort norms’ (1990:348) that should not be allowed to impede the adoption of higher-throughput technology. He agrees with Marglin, he says, that ‘the success of the factory system depended not on 59

RICHARD N.LANGLOIS

Figure 2.6 Backward-bending effort supply

technology but on the creation of a social environment conducive to the imposition of work discipline’27 (1990:52). That is to say, the capitalists had to teach the workers a different set of effort norms.28 Factory discipline served this function. Clark places a somewhat different twist on this account. He notes that, although workers complained about the pace of work, and although they would have chosen both lower e and lower w if allowed to pick their own effort/wage trade-off, workers none the less voluntarily chose the high wages and hard work of the factories. This suggests a problem of marginal incentives: effectively, the workers’ indifference curves are either discontinuous (they exist only locally around specific points, as suggested in Figure 2.7) or exhibit local nonconvexities. In either case, workers cannot traverse the space from point A to point B along the same indifference curve when technology changes; given a choice on the margin, they will take leisure

60

THE COEVOLUTION OF TECHNOLOGY AND ORGANISATION

over higher wages. Yet, if they are ‘coerced’ into working at point B, they will accept the non-marginal trade-off between B and the lower-wage, lowereffort alternatives. ‘The workers dislike discipline, but they stay in the factory because at the end of the week their wage is 60 percent greater than that they can achieve without discipline’ (Clark 1994:160). Thus the workers are not ‘coerced’ into doing what the capitalist wants them to do—that is, they are not exploited; rather, the workers are ‘coerced’ into doing what they themselves would like to do but can’t bring themselves to do on the margin. Like Lazonick (and many others), Clark sees factory discipline—like discipline in other areas of life—as aimed at a problem of individual preference. But, like Alchian and Demsetz, he sees discipline as correcting a problem of externality rather than as changing preferences. As with the shirking workers in the Alchian—Demsetz story, monitoring here solves a problem of divergence between marginal incentives and the global optimum. In this case, however, it is a ‘shirking’ externality that occurs within each worker’s individual psyche. Like a dieter faced with a piece of cake, the worker sees a bit of leisure on the margin as far more enticing than

Figure 2.7 Discontinuous indifference curve 61

RICHARD N.LANGLOIS

higher wages, even though, like a more-svelte figure to the dieter, the longer-range goal of income is ultimately more desirable—by the individual’s own lights. Just as teams of bargemen in pre-Revolutionary China hired an overseer to whip them,29 workers in the Industrial Revolution acceded willingly, though not happily, to the ‘coercion’ of their capitalist masters. In a sense, Clark has bridged the gap between arguments from preference change and arguments from monitoring costs, which is to say that he has ‘neoclassicised’ the contention that the factory system required a new kind of industrial worker. Since Jevons, neoclassical economics has seen the labour process as a matter of preferences, which are assumed given. Clark s innovation is to suggest that one can ultimately explain even what may appear to be ‘coercion’ in terms of the traditional given-preference approach. In the end; indeed, Lazonick’s account is not far different. He too sees ‘customary effort norms’ as reflecting the preferences of workers, and couches the difficulties of moving to a Pareto-improving contract of higher wages for higher effort not in terms of preference change but in terms of the worker’s distrust of the capitalist. I am enough of a neoclassical to agree that preferences do in the end matter. But I also think that it is often quite difficult to disentangle preferences from skills.30 If we take the perspective on the work process suggested earlier, then work, even unskilled work, is not only a matter of supplying some homogeneous commodity called effort but also a matter of possessing and choosing among a repertoire of routines. Lazonick writes about the problems of encouraging workers to acquire skills, arguing that close supervision can be antithetical to a skilled work force to the extent that it fosters industrial conflict, which in turn encourages capitalists to deskill workers in an effort to control them. It may, however, be possible for the reverse to be true. Close supervision may sometimes be not a technique for maintaining effort per se but a way of conveying skills to the workers. What skills? The skills necessary to work effectively with technology and production processes to which in the beginning the workers would have been unused.31 Changing ‘customary effort norms’ may have been as much a matter of changing skills as of changing preferences. This is particularly relevant if we are trying to explain a transition from the putting out system (especially in textiles), where levels of effort among outworkers often rivalled those in the factories.32 The factories required new habits of work, and by no means all of these were the habit of working harder. Perhaps it was the need for new skills—skills complementary to a new technological trajectory—and not just the need for more effort that made factory discipline what Pollard (1965) describes as the central management problem of the industrial revolution.

62

THE COEVOLUTION OF TECHNOLOGY AND ORGANISATION

Evolution, explanation and the inevitable We now arguably have one part of the story. What caused the transition to the factory system? Increasing extent of the market led not only to greater division of labour but also to greater predictability in the production process. This volume effect permitted production processes to use more durable ‘dies’, which implied higher throughput and higher fixed costs, ceteris paribus. And these latter increased for various reasons the marginal benefit of direct process supervision, understood not merely as a way of keeping up effort but as a way of inculcating and reinforcing a repertoire of workers’ routines complementary to the production process. But was this efficiency or exploitation? Marglin (1991) reminds us, quite rightly, that our answer to that question necessarily depends on the ideological preconceptions we bring with us. This does not mean, however, that such preconceptions are beyond discussion, especially if we narrow the field by distilling from ‘ideology’ an underlying explanatory apparatus. For Marglin, the alternative ideologies are mainstream neoclassical economics and ‘radical’ economics. We can take these as convenient starting points, even if we will want to move beyond them. Efficiency explanations, of course, are the bailiwick of mainstream economics. ‘I think it fair to say,’ says Marglin, that mainstream economists, even if they do not see capitalism as the best of all logically possible systems, see the status quo, as did Mr [sic] Pangloss in Voltaire’s Candide, as the best of all realistically feasible systems…. Markets not only work, but when left free of meddlesome government intervention, work well; markets are efficient. Indeed, efficiency is the watchword of the mainstream economist. By contrast, ‘radical’ economists see the concentration of power in the hands of an élite of bankers, businessmen, and bureaucrats as an obstacle to the realization of the individual and the community. And they see the democratization of the economy—the extension to the factory and the office of the participatory principles on which Western political democracy is founded—as an essential part of the project of human liberation. (Marglin 1991:229, emphasis original) Cast in terms of explanatory frameworks, the distinction looks something like this. Neoclassical economics is the epitome of Panglossian explanation because it fuses (in Marglin’s view) the inevitable and the desirable. The factory system, to Marglin’s mainstream economist, could not but have emerged,

63

RICHARD N.LANGLOIS

as it was the product of efficient economic forces; and, precisely because it was the product of efficiency, it was for the best.33 By contrast, Marglin’s ‘radical’ explanatory framework is the epitome of a non-necessitarian explanation. The factory system did not have to emerge, and it is not to the good that it did emerge. This is not to say that institutional structures emerge in a completely arbitrary way: Marglin is sure power matters, though he does not endorse the complete Marxian theory of class struggle; and, at times, the profit motive even seems to matter. The point, however, is that things could— and, of course, should—have been other than they were. Society is not, or at least need not be, under the sway of impersonal forces or laws, but can and should be reshaped by human will.34 One might think these sketches of the explanatory alternatives to be caricatures or straw men. Sadly, they are not. But perhaps this should not be surprising, since these magnetic poles of explanation have a long intellectual heritage. F.A.Hayek (1967) has traced them back at least as far as the ancient Greeks, who thought all social structures to be either natural (completely independent of human will) or artificial (consciously created by human will). As David Hume, Adam Smith and the other philosophers of the Scottish Enlightenment understood, however, the interplay between necessity and will is far more complex than this distinction admits. For the Scots and their followers, social institutions—like the factory system—are the results of human action but not of human design.35 On the one hand, this means that institutions, and the process of their evolution, have a systematic structure susceptible to study. They are the results of innumerable individual wills; but they are not, as Marglin (1991:228) would have it, a ‘haphazard aggregation’ of human intention. The structures that emerge from the process of human action are not entirely arbitrary. On the other hand, however, those structures are not ineluctable. Even less are they ‘optimal’, except in a restricted sense. Institutional structures are the result, then, of an evolutionary dynamic. It would hardly seem worth pointing this out, except that so many participants in the debate over the factory system seem to forget it. Organisational structures and technologies emerge in a process of experimentation and are retained or rejected to the extent that they fit well with the environment (which need not be the market, in any of its senses, alone). At the same time, those structures alter the environment, which in turn affects what comes after. Moreover, technology does not determine organisation any more than the reverse; the two ‘coevolve’, which is to say no more than that they are really both parts of the same process, both ‘institutions’ at the fundamental level of systems of rules and repertoires of routines. Are these structures optimal? Only in the limited sense that, as Stephen Jay Gould puts it in the biological context, they must work well enough: they must satisfy an ‘engineer’s criterion of good design’ (Gould 1977:42). And, as Hayek (1967) points out, evolved social structures, as the product of often 64

THE COEVOLUTION OF TECHNOLOGY AND ORGANISATION

extended periods of trial-and-error learning, are repositories of knowledge more substantial than, and often different in character from, the knowledge of those who appraise those structures with an eye to redesigning them. But none of this implies optimality in the global or absolute sense to which neoclassical welfare economics often pretends. The effectiveness of a structure’s design is measured only relative to the environment, not against an absolute standard. Many different alternative structures might have solved the problem of the environment equally well, either because the selection mechanism was not particularly severe or simply because there are many different engineering solutions that are equally good. Moreover, the sequence of environments through which the structure has passed may be important (Hayek 1967:75), and historical accidents or crucial individuals may shunt evolution along a path that may or may not seem best in retrospect (David 1985). Lying as it does between the poles of necessity and arbitrariness, evolutionary explanation can be, and of course has been, tugged in one direction or another. Lately, indeed, the problem of path dependency has been wielded by many as a kind of all-purpose weapon of attack against various evolved institutional structures (Liebowitz and Margolis 1995). It is important to remember, however, that even a path-dependent institutional structure is still an evolved structure, one carrying a heavy burden of accumulated social learning. It is perhaps an open question whether the QWERTY keyboard layout, for example, is optimal from a human-engineering standpoint (Liebowitz and Margolis 1990); but it remains a formidable and functional institution structure comprising an enormous body of complementary technology and skills within society. In the end, the evolutionary mode of explanation does not so much endorse Pangloss as shift the burden of argument away from those who would defend the status quo and on to those who would attack or redesign it. Of the combatants in the debate over the emergence of the factory system, it is probably the economic historians who have best understood this mode of explanation, at least instinctively. By combining sequence with at least a modicum of theory, economic history forces one to confront both the processes of institutional evolution and the ‘engineering design’ arguments we might use to make sense of that evolution. Perhaps it is for that reason that the historians’ account of the transition makes the most sense—in theory as well as in history.

Notes 1 The author would like to thank Lee Alston, Derek Johnson, Steve Jones, Trevor Knox and Paul Robert son for valuable comments. 2 For an excellent account of the Marxian system, see Roberts and Stephenson (1973). 65

RICHARD N.LANGLOIS

3 In addition to Marglin (1974), see Marglin (1984, 1991). 4 We might in fact call this the ‘traditional’ view among economic historians. See for example Mantoux (1961). 5 Many other followers of Coase would insist that, in the end, a contract is a contract, and ‘authority’ is not involved. ‘To speak of managing, directing, or assigning workers to various tasks is a deceptive way of noting that the employer continually is involved in renegotiation of contracts on terms that must be acceptable to both parties. Telling an employee to type this letter rather than to file that document is like my telling a grocer to sell me this brand of tuna rather than that brand of bread’ (Alchian and Demsetz 1972:778). The final station for this train of thought—a reductio, but by no means ad absurdum— is that the firm is nothing but a ‘nexus of contracts’ (Cheung 1983). 6 As Temin (1991) points out, there is a difference between entrepreneurs, who engage in non-routine command behaviour, and managers, who engage in routine or customary behaviour, including the exertion of factory discipline. ‘Managers, in short, were the workers’ bosses, but entrepreneurs were the managers’ bosses’ (Temin 1991:350). On the distinction between command and customary behaviour, see Temin (1980). 7 This was also true of inside contractors like the master spinners in Lancashire. These masters, who hired and disciplined their own ‘scavengers’ and ‘piecers’, were far more likely to use and abuse child labour than were capitalists directly employing labour, and they accounted for a significant fraction of the child labour in the industry (Ure 1861:290ff; Pollard 1965:43). 8 In the sense of Langlois and Robertson (1995: ch. 6). I return to this idea later in the chapter. 9 In the era before major mechanical innovations in cotton machinery, spinning was the bottleneck, as it took the output of upwards of five spinners to supply one hand loom (Landes 1969:57). 10 Notably S.R.H.Jones in private communication with the author. 11 And continued growth in the extent of putting out is not by itself inconsistent with the onset of diminishing returns. 12 On the other hand, the just-in-time inventory system, invented in the early American automobile industry as ‘hand-to-mouth buying’ (Flugge 1929:163), suggests that suppliers can also in principle regulate product flow carefully. Indeed, hand-tomouth buying is itself an instance of the division of labour, for it decouples the function of speculation in inventories from the manufacturing function (Stillman 1927:3). 13 This also depends, however, on the worker’s capital being firm-specific as well as process-specific. If there is a thick market for, say, weavers, a weaver who threatens to withdraw his looms from a firm might be easily replaced with less-recalcitrant alternates. 14 See also Ames and Rosenberg (1965), who talk about the activities performed in production as instances of rule-following behaviour. 15 Following Ames and Rosenberg (1965), I am here taking ‘skill level’ to be a measure of the size of the worker’s repertoire of routines. In fact, we can also think of being skilful as meaning skill-deepening, that is, a highly developed ability to perform one or a few routines. 16 Minkler (1992) uses this idea of specialised knowledge as an explanation for the franchising contract, a modern-day analogue of the putting out system. 17 Professionals, indeed, are autonomous not only in the sense that they are seldom employees but also in that ‘no one except another professional is able to challenge the day-to-day decisions of a professional’ (Savage 1994:139). And professionals

66

THE COEVOLUTION OF TECHNOLOGY AND ORGANISATION

18 19

20

21

22

23

24

25

26

are monitored not only by relative prices but by a complex set of institutions, including peer monitoring. In more technical terms, predictability reduces the behavioural entropy of the choice among routines. For an analysis of the effect of uncertainty and unpredictability on the selection of actions from a repertoire, see Langlois (1986a). The acquisition of skill by a machine does not, however, imply deskilling of labour. Consider the backhoe, which integrates a number of ditch-digging functions. It requires an operator more skilled than any manual ditch-digger (Robertson and Alston 1992). Machines, of course, can deal with some kinds of uncertainty. The prime example is the Jacquard loom, the ancestor of modern numerical-control techniques. But even in modern computer-aided manufacturing, the degree and type of uncertainty with which machines can deal is limited to what I call parametric uncertainty (Langlois 1984). Indeed, the case of calico printing is an example almost literally analogous to the printing example cited above: cylinder printing was invented in 1783 not for text but for the printing of calicoes, moving Baines (1966:265, cited in Mokyr 1990:99) to compare the advancement of this machine over block printing to the advancement of mechanical spinning over the spinning wheel. Whether the observed expansion path is upward-or downward-sloping (that is, whether we observe increased throughput and higher fixed costs in a particular industry) will depend on the relative strengths of the volume effect and the rate of innovation in machinery. That is, the firm will be willing (and able) to pay the worker a wage w=ve-F, where v is the value of a unit of effort to the firm, e is effort, and F is the rental cost of fixed capital (Clark 1994:138). Thus the appropriate wage for zero effort is not zero but -F, since the worker ties up machinery and other fixed inputs; and workers become more valuable as they provide more effort. Writing in the context of contractual choice in agriculture in the post-bellum American South, Alston and Higgs (1982:340–1) suggest a complementary reason why increased capital intensity might lead to closer supervision. If, perhaps because of the absolute amount of capital required, capital hires labour, then it is in the interest of the capitalist to monitor closely to ensure that the worker properly maintains the productive assets. As with the shirking explanation, this motive becomes more urgent the more capital-intensive the production process. Monitoring to avoid harm to capital assets is not, however, necessarily the kind of supervision that keeps up worker effort. Moreover, Lazonick (1990:350–1) maintains that supervision to keep up effort levels actually increases harm to capital assets by encouraging sabotage to slow the pace. None the less, it may well be that what appeared to be supervision to maintain effort level alone actually had other motives instead or in addition, a point to which I return later in the chapter. On the other hand, it is possible that tasks that were susceptible to piece-rate contracting under the putting out system might not be so susceptible when the workforce is concentrated. A number of writers have argued that the very concentration of the workforce lowers the transaction costs of (typically informal) collective action to manipulate the piece-rate system to the workers’ advantage (Csontos 1993; Lazonick 1990). When this is possible, direct process supervision may become less costly. As we will see, Clark’s story is not in fact obviously a ‘coercion’ account, since he argues in the end that the contract (w1, e1) plus discipline is both Pareto optimal and ultimately voluntary.

67

RICHARD N.LANGLOIS

27 As I have hinted, however, Lazonick’s high-fixed-cost model does in fact suggest that technology was indeed in part responsible for the factory system, in that it was fixed costs that made discipline desirable. In Lazonick’s defence, however, fixed costs do not always mean physical capital; they can also mean humancapital investments and fixed investments in organisational capabilities. How important these latter were compared with physical capital during the Industrial Revolution is an open question. Despite his inclination to heap praise on Marglin (see, e.g., Lazonick 1991:291–4), Lazonick’s account is on the whole quite at variance with that of Marglin. For Lazonick, changes in favour of highthroughput production (during the Industrial Revolution and at other times) do in fact reflect efficiency, and are moreover the principal engine of economic growth and competitive advantage. In addition, the imposition of these technologies is not typically exploitation, in that the most successful episodes of rapid economic growth have occurred when institutions permitted capitalists and workers to share the gains of new technology so that neither would have an incentive to impede those changes. Indeed, in Lazonick’s work the failure of economic growth and competitiveness often takes the form of labourers standing in the way of the efforts of capitalists to impose more-efficient, highthroughput methods. 28 Some historians (e.g. Voth 1998) would argue, however, that the workers did not need capitalists to inculcate in them norms of higher effort: it was the developing possibilities on the consumption side that led workers to supply more effort (largely by allowing the observance of ‘Saint Monday’ and various minor religious feasts to fall into desuetude) in order to generate the income necessary to acquire newly available consumer goods. This is, of course, an idea that goes back at least to David Hume’s ‘demonstration effect’. By contrast, Clark and van der Werf (1996) argue that medieval labourers worked just as hard as those of the late eighteenth and early nineteenth centuries and that there was in fact no such ‘industrious revolution’. 29 Or so Steven Cheung (1983) claims. 30 For an elaboration of this point, see Langlois and Cosgel (1998). 31 In the American South after the Civil War, it was common for agricultural workers to begin their careers as closely supervised wage labourers and then eventually to become share-cropping contractors, a phenomenon known as the ‘agricultural ladder’. Alston and Higgs (1982) argue that this phenomenon cannot be solely the result of the workers possessing greater physical capital with age, but must also involve increasing human capital, which is the neoclassical shorthand for a repertoire of relevant skills. This suggests that, once taught the necessary repertoire of behaviours through supervised wage labour, the workers could be monitored easily enough with pecuniary incentives. 32 As Pollard makes clear in his detailed account of the problems of factory labour, it was not level of effort per se that distinguished the factory from the cottage. The crucial difference was the regularity of the work, against which the otherwise hard-working operatives chafed. In addition, the factory workers required new skills in accuracy and standardisation and needed to take proper care of machinery that was not their own (Pollard 1965:181). There is more to ‘discipline’ than effort, and much of it is in the nature of skills. 33 In the case of the factory system of the late eighteenth century, it may be possible— with a little stretching—to associate the status quo (the factory system) with the ‘unfettered’ market. But it has always seemed to me absurd to claim in general that neoclassical economics is Panglossian on the grounds that it upholds the efficacy of free markets. Even in the most market-oriented countries, the ‘status

68

THE COEVOLUTION OF TECHNOLOGY AND ORGANISATION

quo’ is scarcely the free market: an absolutely enormous part of the modern economy is regulated by the state or otherwise under political control. In the modern world, a belief in the efficacy of markets makes one a ‘radical’, not a Panglossian, as recent events in the United States may be serving to demonstrate. To the extent that mainstream neoclassical economics is Panglossian, it is because it does not uphold the efficacy of markets with much conviction. ‘Market failure’ is as much the watchword of the normative neoclassical as is ‘efficiency’, and this malleable doctrine can be and has been used indiscriminately to assert the primacy of ‘Western political democracy’ over individual rights. 34 The exact nature of the human will involved is not at all clear, however. Only at its peril, Marglin argues, will a society ‘leave any important decisions to the haphazard aggregation of individual maximizing decisions, be these decisions expressed through a market, a polling booth, or what have you. Society may leave to individuals to determine whether they eat apples or nuts, but not what the rate of growth is, what the distribution of income is, or what the structure of relative prices is’ (Marglin 1991:228–9). As to who or what ‘society’ is, Marglin is understandably silent. 35 In the famous phrase of Adam Ferguson (1980 [1767]). For a more thorough discussion of these issues of institutional explanation, see Langlois (1986b) and Langlois and Everett (1994).

References Alchian, A. (1959) ‘Costs and Output’, in M.Abramovitz et al. (eds.) The Allocation of Economic Resources, Stanford: Stanford University Press; reprinted in Alchian (1977): 273–99. ——(1977) Economic Forces at Work, Indianapolis: Liberty Press. Alchian, A. and Demsetz, H. (1972) ‘Production, Information Costs, and Economic Organization’, American Economic Review, 62(5) (June): 772–95; reprinted in Alchian (1977):73–110. Alston, L.J. and Higgs, R. (1982) ‘Contractual Mix in Southern Agriculture since the Civil War: Facts, Hypotheses, Tests’, Journal of Economic History, 42(2) (June): 327–53. Ames, E. and Rosenberg, N. (1965) ‘The Progressive Division and Specialization of Industries’, The Journal of Development Studies, 1(4):363–83. Baines, E. (1966) History of the Cotton Manufacture in Great Britain, 2nd edn, New York: Augustus M.Kelly. Berg, M. (1991) ‘On the Origins of Capitalist Hierarchy’, in Bo Gustafsson (ed.) Power and Economic Institutions: Reinterpretations in Economic History, Aldershot: Edward Elgar: 173–94. Buttrick, J. (1952) ‘The Inside Contract System’, Journal of Economic History, 12(3): 205–21. Cheung, S.N.S. (1983) ‘The Contractual Nature of the Firm’, Journal of Law and Economics, 26 (April): 386–405. Clark, G. (1994) ‘Factory Discipline’, Journal of Economic History, 54(1) (March): 128– 63. Clark, G. and van der Werf, Y. (1996) ‘The Industrious Revolution or the Bourgeois Middle Ages?’ Working paper, Department of Economics, University of CaliforniaDavis. Coase, R.H. (1937) ‘The Nature of the Firm’, Economica (N.S.) 4 (November): 386– 405.

69

RICHARD N.LANGLOIS

Cohen, J.S. (1981) ‘Managers and Machinery: An Analysis of the Rise of Factory Production’, Australian Economic Papers (June): 24–41. Csontos, L. (1993) Inside the Black Box of the Firm: The Economics of Internal Organization, PhD Dissertation, University of Connecticut. David, P.A. (1985) ‘Clio and the Economics of QWERTY’, American Economic Review 75(2):332–7. Fang, H. (1978) The Triumph of the Factory System in England, Philadelphia: Porcupine Press. Ferguson, A. (1980) An Essay on the History of Civil Society, New Brunswick: Transaction Books. [First published in 1767.] Flugge, E. (1929) ‘Possibilities and Problems of Integration in the Automobile Industry’, Journal of Political Economy, 37:166. Gould, S.J. (1977) Ever Since Darwin, New York: Norton. Hayek, F.A. (1967) Studies in Philosophy, Politics and Economics, Chicago: University of Chicago Press. Jones, S.R.H. (1982) ‘The Organization of Work’, Journal of Economic Behavior and Organization 3(2):117–37. ——(1987) ‘Technology, Transaction Costs, and the Transition to Factory Production in the British Silk Industry, 1700–1870’, Journal of Economic History 47(1) (March): 71–96. ——(1993) ‘The Origins of the Factory System in Great Britain: Technology, Transaction Costs, or Exploitation?’, in M.Kirby and M.Rose (eds) Business and Industrialization, London: Routledge. (Reprinted as Chapter 1 in this volume.) Kimball, D.S. (1929) Industrial Economics, New York: McGraw-Hill. Klein, B., Crawford, R.G. and Alchian, A. (1978) ‘Vertical Integration, Appropriable Rents, and the Competitive Contracting Process’, Journal of Law and Economics 21(2):297–326. Landes, D.S. (1969) The Unbound Prometheus: Technological Change and Industrial Development in Western Europe from 1750 to the Present, Cambridge: Cambridge University Press. ——(1986) ‘What Do Bosses Really Do?’ Journal of Economic History, 46(3) (September): 585–623. Lane, F.C. (1973) Venice, A Maritime Republic, Baltimore: Johns Hopkins University Press. Langlois, R.N. (1984) ‘Internal Organization in a Dynamic Context: Some Theoretical Considerations’, in M.Jussawalla and H.Ebenfield (eds) Communication and Information Economics: New Perspectives, Amsterdam: North-Holland: 23–49. ——(1986a) ‘Coherence and Flexibility: Social Institutions in a World of Radical Uncertainty’, in I.Kirzner (ed.) Subjectivism, Intelligibility, and Economic Understanding: Essays in Honor of the Eightieth Birthday of Ludwig Lachmann, New York: New York University Press: 171–91. ——(1986b) ‘Rationality, Institutions, and Explanation’, in R.N.Langlois (ed.) Economics As A Process: Essays in the New Institutional Economics, New York: Cambridge University Press. Langlois, R.N. and Cosgel, M.M. (1998) ‘The Organization of Consumption’, in M.Bianchi (ed.) The Active Consumer, London: Routledge. Langlois, R.N. and Everett, M.J. (1994) ‘What Is Evolutionary Economics?’, in L. Magnusson (ed.) Evolutionary and Neo-Schumpeterian Approaches to Economics, Dordrecht: Kluwer Academic Publishing: 11–47. Langlois, R.N. and Robertson, P.L. (1995) Firms, Markets, and Economic Change: A Dynamic Theory of Business Institutions, London: Routledge.

70

THE COEVOLUTION OF TECHNOLOGY AND ORGANISATION

Lazonick, W. (1990) Competitive Advantage on the Shop Floor, Cambridge: Harvard University Press. ——(1991) ‘What Happened to the Theory of Economic Development?’, in P. Higonnet, D.Landes and H.Rosovsky (eds) Favorites of Fortune, Cambridge: Harvard University Press: 267–96. Leijonhufvud, A. (1986) ‘Capitalism and the Factory System’, in R.N.Langlois (ed.) Economics as a Process: Essays in the New Institutional Economics, New York: Cambridge University Press: 203–23. Liebowitz, S.J. and Margolis, S.E. (1990) ‘The Fable of the Keys’, Journal of Law and Economics 33(1) (April): 1–25. ——(1995) ‘Path-dependence, Lock-in, and History’, Journal of Law, Economics, and Organization 11:205–26. Mantoux, P. (1961) The Industrial Revolution in the Eighteenth Century, London: Jonathan Cape, revised edition. Marglin, S.A. (1974) ‘What Do Bosses Do?’, Review of Radical Political Economy, 6 (Summer): 33–60. ——(1984) ‘Knowledge and Power’, in F.H.Stephen (ed.) Firms, Organization and Labour, London: Macmillan: 146–64. (1991) ‘Understanding Capitalism: Control versus Efficiency’, in B. Gustafsson (ed.) Power and Economic Institutions: Reinterpretations in Economic History, Aldershot: Edward Elgar: 225–52. Marx, K. (1867) Capital, 1, trans. Ben Fowkes, New York: Vintage Books, 1977. Minkler, A.P. (1992) ‘Why Firms Franchise: A Search Cost Theory’, Journal of Institutional and Theoretical Economics, 148(2):240–59. ——(1993) ‘Knowledge and Internal Organization’, Journal of Economic Behavior and Organization, 21:17–30. Mokyr, J. (1990) The Lever of Riches, New York: Oxford University Press. Nelson, R.R. and Winter, S.G. (1982) An Evolutionary Theory of Economic Change, Cambridge, Mass.: Harvard University Press. North, D.C. (1981) Structure and Change in Economic History, New York: Norton. Pollard, S. (1963) ‘Factory Discipline in the Industrial Revolution’, Economic History Review 16 (December): 254–71. ——(1965) The Genesis of Modern Management, Cambridge: Harvard University Press. Roberts, P.C. and Stephenson, M. (1973) Marx’s Theory of Exchange, Alienation, and Crisis, Stanford: Hoover Institution. Robertson, P.L. and Alston, L.J. (1992) ‘Technological Choice and the Organization of Work in Capitalist Firms’, Economic History Review 45(2) (May): 330–49. Savage, D.A. (1994) ‘The Professions in Theory and History: the Case of Pharmacy’, Business and Economic History, 23(2) (Winter): 129–60. Simon, H.A. (1957) Models of Man, New York: John Wiley. Smith, A. (1976) The Wealth of Nations, Glasgow edn, Oxford: Clarendon Press. Stillman, K.W. (1927) ‘Hand-to-Mouth Buying: What Has It Done—and What Will It Do?’, Automotive Industries, (July) 2:1–4. Stinchcombe, A.L. (1990) Information and Organizations, Berkeley: University of California Press. Temin, P. (1980) ‘Modes of Behavior’, Journal of Economic Behavior and Organization, 1(2):175–95. ——(1991) ‘Entrepreneurs and Managers’, in P.Higonnet, D.Landes and H. Rosovsky (eds) Favorites of Fortune, Cambridge: Harvard University Press: 339–55. Ure, A. (1861) The Philosophy of Manufactures, 3rd edn, London: H.G.Bohn.

71

RICHARD N.LANGLOIS

Voth, H. (1998) ‘Work and the Sirens of Consumption in Eighteenth-century London’, in M.Bianchi (ed.) The Active Consumer, London: Routledge. Williamson, O.E. (1975) Markets and Hierarchies: Analysis and Antitrust Implications, New York: The Free Press. ——(1980) ‘The Organization of Work: A Comparative Institutional Assessment’, Journal of Economic Behavior and Organization, 1(1):5–38. ——(1985) The Economic Institutions of Capitalism, New York: The Free Press.

72

3

CLASS STRUCTURES AND THE FIRM The interplay of workplace and industrial relations in large capitalist enterprises Thomas Welskopp

The problem and the argument in outline Distinct positions in both economic theory and business history have come to dispute the relevance of the ‘firm’ or the ‘business enterprise’ as an analytical category for exploring production processes, business structures and management decisions in industrial capitalist societies. The present essay acknowledges that their criticism of well-established formulations of concepts delineating the role of ‘firms’ in capitalist economies may be convincing. Yet to conclude from this that we must discard altogether the notion of the ‘firm’ seems a step in the wrong direction. Revisionist theory so far has fallen short of introducing a conceptionally sound alternative to the ‘firm’-centred approach to economic theory and business history. This essay argues that any attempt to establish such an alternative will eventually be bound to reproduce the very theoretical limitations it actually aims to overcome. The concept of the ‘firm’ is indispensable for both theory and historical research, since it is uniquely suited to grasp the peculiar type of social institution linking markets to production processes which belongs to the constituent structural elements that define and compose ‘industrial capitalism’. As Max Weber has pointed out, the separation of the ‘enterprise’ from both the social life-worlds beyond the factory gates and the ‘civic’ organisational and political sphere represents the central structural principle which distinguishes ‘industrial capitalist societies’ from preceding or discordant societal formations and decisively marks ‘industrial capitalism’ as ‘enterprise capitalism’ (‘Betriebskapitalismus’) (Weber 1991). A theoretically sound re-conceptualisation of the ‘firm’ in modern society thus constitutes a necessary precondition for examining both the inner workings of industrial capitalism and the fabric of societies at large structured by it. 73

THOMAS WELSKOPP

Moreover, such an endeavour, which is the prime objective of this essay, can provide a basis for developing the synthetic perspective of a comparative history of modern society that aims at integrating economic history, business history, labour history and the history of technology, since it is in the institutional locale of the ‘firm’ that the research interests of these subfields intersect (Scranton 1989). The present essay will illustrate that, despite disciplinary fragmentation and insulation, there are surprising parallels in the limitations of approaches to the ‘firm’ as applied across the vast array of economic, business and labour history. The failure to bridge the ‘agency/structure’ divide will be identified as the common denominator underlying these deficiencies (Birke 1992). The essay will establish that, while recent studies by industrial sociologists, geographers and labour historians on the inner workings of capitalist enterprises have successfully tackled this issue and have broadened our understanding of the ‘firm’ by applying a decidedly relational and microtheoretically enriched perspective to the analysis of industrial ‘employment relations’, they, in turn, have fallen prey to the fallacy of conflating social theory and historical theory of society. This has fostered the tendency to dissolve intrafirm relations into a welter of contingent situative interactions, whereas firm—environment relations—a central concern of conventional economic theory—remain completely opaque (Welskopp 1994a; Welskopp 1994b; Welskopp 1996). In this type of revisionist theory, the peculiar traits that characterise the ‘enterprise’ and its ‘employment relation sets’ as an institutional centre of gravity in industrial capitalist society thus remain elusive. Therefore, the essay will delineate a concept of the ‘firm’ as a complex of systemically interrelated ‘employment’ and ‘industrial relations’ which allows for ‘grounding workplace relations both in the wider context of capitalist relations and in the concrete experience of production’ (Scranton 1989:14). The concept draws on a dual notion of the relation between ‘agency’ and ‘structure’ in order to incorporate the ‘firm’ as a specific ‘field of social interaction’ into a reformulated model of ‘class’ in industrial capitalist societies. The revitalisation of ‘grand theory’, it will be argued, is necessary for infusing industrial capitalist ‘substance’ into the patterns of intrafirm interactions. Yet the grounding of the concept in modern social theory in turn substantially modifies the category of ‘class’: stripped from their usual ‘structuralist’ connotations, ‘class relations’ and ‘class structures’ appear as ‘contextualist’ axes of structuration which maintain systemic integration at the macrolevel of society while accounting for structural diversity at the mesolevel of the ‘firm’ and for ‘bounded’ contingency at the microlevel of intrafirm social relations (Welskopp 1994b: 30ff.). Such a conceptualisation of the ‘business enterprise’ as the system-specific institutional focus at the mesolevel of analysis, moreover, helps avoid the ‘productionist fallacy’ of reducing social reality to workplace relations. Indeed, a precise reconstruction of the ‘firm’ as a ‘field of social interaction’ proves indispensable for conceiving the other two societal 74

CLASS STRUCTURES AND THE FIRM

spheres of structuration: the ‘private’ life-worlds of communities, neighbourhoods and social milieus, and the organisational-political sphere of society, in an analogous way. Only this procedure, in turn, allows for identifying the concrete interrelations between these distinct spheres. By virtue of this integrative perspective, business and labour history can not only regain a long-lost mutual compatibility; both disciplines can advance to integral and methodologically pioneering elements of a comparative history of industrial capitalist societies.

The limitations of conventional theory Although business and labour history barely communicate with each other, their respective concepts of the ‘firm’ share essential features. First, they regard ‘business enterprises’—albeit to varying degrees—as functional mechanisms shaped in response to changing environmental pressures. Second, this view of the ‘firm’ subordinates the individual enterprise to unilinear evolutionary models either of corporate development and of the trajectory of labour processes in advanced capitalism, or of ‘working class formation’. Third, all these approaches converge in a reductive treatment of ‘agency’ and ‘structure’ alike. This bears the paradoxical result that, instead of charting the ‘firm’ as a social system which is constituted by specific sets of social relations and reproduced by specific patterns of social interaction between all participating social groups, ‘agency’ is actually removed from the confines of the ‘enterprise’ proper and conveyed either to management, which then appears as a collective agent outside of the structure it produces, or to the workforce, whose capacity for action is then limited to spheres beyond the workplace. The relations among the social groups engaged in capitalist production thus seem obscured rather than illuminated by concepts which envision the ‘firm’ as an inanimate transmitter of external pressures rather than as a terrain of interaction where those relations are structured (Tolliday and Zeitlin 1991). In business history, the ‘organisational synthesis paradigm’ of the Chandler school and the ‘transaction-cost theory’ of Oliver Williamson and others have shaped influential conceptions of the ‘business enterprise’ in industrial capitalism (Chandler 1990; Williamson 1975). Although this broad array of approaches explicitly concentrates on the issue of transforming market constraints and impulses into organisational structures, the inner fabric of capitalist firms—aside from their inter-departmental and inter-divisional makeup—has remained a ‘black box’. ‘Employment relations’ as well as actual production processes almost completely escape the conceptual grasp of this approach. Thus, whereas the extent to which environmental market imperatives encroach on management’s discretionary behaviour has been the subject of controversial debate, the extent to which the latter may be

75

THOMAS WELSKOPP

‘bounded’ by the internal context of the firm has been ignored. The ‘firm’ appears as a functional ‘input-output machine’ and as a mere object of management decisions. Since this approach treats the social context of the ‘enterprise’ as virtually irrelevant for the organisational outcomes of market dynamics transmitted by management activity, the concept of ‘market’ becomes both highly abstract, i.e. remote from ‘production’, and deterministic. Abstraction from context also accounts for the fact that the individual producing unit—and not social interaction—constitutes the microlevel of analysis, while strategic management, synonymous for the ‘enterprise’ or the ‘corporation’, composes the mesolevel of inquiry (which from the viewpoint of social theory is appropriately considered as the complex network of departments, workshops and production facilities which make up the ‘enterprise’ at large). Since the microunits of the ‘firm’ are cut off from any active role in producing corporate structures, the ‘organisational synthesis paradigm’ views individual enterprises merely as cases indicating a unilinear long-term trend towards the decentralised multidivisional manager-corporation that Chandler and his followers depict as the ‘normal’ trajectory of advanced industrial capitalism. Institutional diversity at either level consequently hints at diversions not covered by the framework proper, which, therefore, can only be explained by recourse to exogenous variables. On a deeper theoretical level, ‘agency’ actually remains reserved for higher management, whereas ‘structure’ rules in the anonymous reaches of the ‘firm’s’ texture, which hence seems uninhabited by historical actors. Yet even at the level of management, a relational perspective on bow the transformation of specific market imperatives into concrete organisational structures actually occurs is conspicuously absent (Kleinschmidt and Welskopp 1993:93ff.). The framework’s monolithic conception of management rules out the analysis of this group’s performance as a social process, although, as critical geographers rightly contend, ‘management is itself a labour process and not one which can be reduced to [either] the control of the labour of others’ or to organisation building (Sayer 1985:10, 24). In a similar vein, concepts related to the ‘labor process debate’ sparked off by Harry Braverman regard the ‘enterprise’ and the workforce employed in it as passive objects of purposeful management activity informed by strategic knowledge superior to all other groups concerned (Braverman 1974; Edwards 1979). By means of implementing ever more sophisticated machinery and applying Taylorist and Fordist techniques to the labour process, capitalist management actively fosters the long-term trends towards the separation of task conception from execution, towards the dilution of complex craft skills by a progressively minute division of labour, and towards the expropriation of knowledge and control over work from an increasingly dequalified workforce. Although labour process theorists differ from business historians in so far as they have focused their attention on the production process itself, this shift of perspective has not obviated theoretical flaws which we have 76

CLASS STRUCTURES AND THE FIRM

encountered as inherent in the ‘organisational synthesis’ approach portrayed on p. 75. Despite its central position in the argument, they devote little effort to the production process, organised in ‘firms’, as a distinct social setting and as an arena where management-workforce relations are situated (Zeitlin 1983). Instead, the actual production processes appear as mere instances of the overarching logic of capitalist development. This, in turn, entails a view of phenomena at the microlevel of the ‘firm’ which considers them as direct manifestations of the larger trend. The ‘labour process’ is as abstract a notion as ‘organisation’ is in Chandlerian business history. Equally, the concept of ‘capitalist accumulation’, which serves as a substitute for the ‘market’ in much of neo-Marxist analysis, proves, just like its counterpart in business history, too unspecific to account for the concrete social changes allegedly initiated by the process. Finally, the contention that management is able progressively to wrestle knowledge and control from the workforce entails the assumption that managers acquire an equally increasing capability to social action as such in the process, whereas the scope of discretionary behaviour on the part of the workforce correspondingly declines. Again, ‘agency’ becomes the attribute of a privileged social group while all other groups are successively stripped of their very character as social agents. In US and German labour history, the ‘working-class formation paradigm’ has established a particularly influential interpretatory framework for exploring the links between the historical diffusion of industrial ‘wage labour’ and social organisation along class alignments in the process of industrialisation (Zwahr 1978; Kocka 1983, 1986; Katznelson 1986). But, in its model design—which distinguishes among economic, social, socio-cultural and political stages of class formation—the ‘firm’ or, more precisely, the ‘workplace’ as a social context does not figure prominently in explaining phenomena on the level of collective organisation and action, which by definition transcend the shopfloor context. The concept of ‘class formation’ regards the ‘workplace’ only as an objectively given structural backdrop of workers’ ‘experiences’ with ‘wage labour’. These ‘experiences’ coalesce with perceptions and mentalities developed in other spheres of society to form the foundations of class-based identities and dispositions which, in turn, motivate ‘class formation’ in terms of collective action, organisation and conflict. The problems of this framework are threefold. First, in this type of concept the institutional mesolevel of the ‘firm’ is virtually bypassed as a distinct focus of analysis, in so far as categories referring to the macrolevel of society are directly projected to the microlevel of individual interaction as expressed in the notion of ‘experience’. This forces the historian to search for social phenomena that most closely match the abstract substance of the categories ‘wage labour’ and ‘class’, which actually pertain to the structural principles of the society at large. This quest to discover some ‘pure specimen of class’ (Thompson 1963:9) in concrete interaction leads to the levelling of structural diversity in favour of charting a tendency 77

THOMAS WELSKOPP

towards an alleged ‘homogenisation’ of the internal structure and experiences of working classes’ which is seen as the necessary precondition for collective action and class-based organisation. Margaret R.Somers has rightly charged this position as representing a fallacy of misplaced concreteness and as embodying an ‘epistemology of absence’, since it places a set of social phenomena at the core of the model which, in its rigidity and degree of abstraction, is impossible to be discerned anywhere in historical reality (Somers 1989; Canning 1992). Second, from the perspective of modern social theory, the stages of ‘class formation’ also signify a movement from ‘structure’, via ‘consciousness’, towards ‘conduct’, which transforms into a substantive historical explanation of organisation building and class conflict the fundamental connection between elements of ‘structure’ and ‘agency’ that is constitutive of all social interaction. In consequence, capitalist production and the ‘workplace’ are, as far as the workers are concerned, completely determined by ‘structure’, whereas ‘agency’ is narrowed down to collective action. Collective action and organisation are implicitly treated as identical, and the explanation of both rests on a vague notion of shared experiences and collective learning processes. Third, although this concept— unlike the others—envisions the industrial worker as an historical subject, it contributes its share to the reproduction of the ‘agency/structure’ divide by attributing to management and labour distinct and independent spheres of ‘agency’ and ‘structural constraint’. Workers’ agency in the workplace is confined to ‘experience’, whereas management enjoys structurally unrestricted capabilities of conduct in the context of the ‘firm’. This view, however, virtually rules out a relational approach to the patterns of interaction between these social groups, which decide how ‘class relations’ manifest themselves in the shopfloor context (Breuilly 1989; Welskopp 1992). Our critical assessment of dominant approaches to the ‘firm’ in industrial capitalist societies has demonstrated that both the deficient treatment of the ‘agency/structure’ problem and the flawed conflation of social theory and historical theory of society must be held responsible for their obvious limitations. To put it precisely: all three strands of theory on the ‘enterprise’ converge in that they graft general social theory on to historically concrete theories of society. They do this by building system-specific uneven distributions of discretionary power among different social groups into the very notion of ‘structure’ and ‘agency’ itself. The consequence of this reductive procedure has been that none of the existing concepts of the ‘firm’ encourages a relational analysis of those peculiar patterns of interaction that constitute the ‘enterprise’ as a social context. In this model design, the ‘firm’ looms uneasily between the macrolevel of society and the microlevel of interaction as an inanimate ‘black box’ which actually dissects the linkage between macro-and microhistory, instead of figuring as the mesolevel link between these dimensions. Since the shortcomings of these ‘structuralist’ concepts of the ‘firm’ are deeply embedded in the 78

CLASS STRUCTURES AND THE FIRM

core elements of their frameworks, a profound reconceptualisation of the ‘firm’ as a ‘field of social interaction’ calls for a radical departure in the direction of both, grounding any such concept in modern social theory capable of straddling the ‘agency/structure’ divide and ‘contextualising’ theories of society in order to assemble ‘midrange concepts’ which fulfil the task of explaining concrete historical configurations and processes, as well as that of providing the basis of broader generalisation.

‘Micropolitics’ as macroreductionism In recent years, an avalanche of empirical studies has challenged the core assumptions of many of the concepts reviewed above. Yet most of those case histories have shied away from drawing systematic conclusions from their findings. Therefore, it has been in social geography and industrial sociology rather than in history that a theoretical debate about reconceptualising the ‘firm’ in capitalist societies has surfaced. It is not too sanguine to infer from recent impressions, however, that revisionist empirical research and revitalised theory are beginning to engage in a fertile dialogue. Concepts like that of the internal ‘micropolitics’ in organisations or that of the ‘enterprise’ as a ‘negotiated social order’ have increasingly made inroads into the provinces of business and labour history, for they seem to be aptly suited to grasp the social reality of intrafirm relations, whose complexity has been depicted in many empirical case studies but has fit uneasily, if at all, into the ‘structuralist’ rigidities of conventional theory (Burns 1961; Küpper and Ortmann 1992). Practitioners of the ‘micropolitics’ or ‘production polities’ approach start out from the observation that ‘structuralist’ accounts of the conduct of ‘firms’ have fallen short of grasping the social reality of the processes under investigation. By reducing the ‘firm’ either to an object of managerial activity or to an objective generator of workers’ ‘experiences’, they have actually removed the site of production and distribution from the focus of analysis. The inner workings of ‘business enterprises’ have been inferred from management decisions and workers’ reactions outside the workplace, rather than subjected to closer scrutiny in their own right. It is this disregard for the quality of the ‘firm’ as a social ‘reality sui generis’ which underlies the tendency to attribute to management a capability to act from a position of ‘systemic rationality’, while degrading workers to social ‘actors of secondary order’ (Kotthoff and Reindl 1990:14; Price 1983). Detailed studies of systems of craft production, of processes of rationalisation, of skill development, and of the conduct of shop stewards and works councils, have revealed, however, that the social context of production itself determines to a large degree how the patterns of interaction at this level ‘channel down’ concrete environmental impulses and technological choices into intrafirm organisation and shopfloor practice. The ‘enterprise’ is the peculiar institutional system in whose boundaries

79

THOMAS WELSKOPP

the transformation of labour power into workplace performance, the transformation of technology into production facilities, and the transformation of organisational concepts into organisation occur (Minssen 1992). The interrelatedness of these processes creates a scope of indetermination much broader than implied in the ‘input-output machine’ view of the ‘firm’ as advanced by conventional theory. Therefore, ‘enterprises’ and ‘workshops’ constitute ‘soft’ social systems that provide open arenas for bargaining and negotiation among all social groups involved in production. This conception entails, first, that there is no such thing as a ‘one best way’ of intrafirm organisation or as a technocratic logic of capitalist accumulation that excludes structural diversity at the level of the ‘enterprise’. Rather, the social reality of production emerges from an ongoing bargaining process which is inherently political. Second, the social context of the ‘firm’ is constituted by the social relations among the social groups involved and recursively reproduced in day-to-day shopfloor practice. From this strictly relational perspective, all members of the ‘enterprise’ appear as social agents of the same quality. The relatively stable internal order of the ‘firm’ is not secured by managers’ and workers’ access to different qualities of ‘agency’, but rather by constantly embattled balances between relations of power and consent. While it is beyond any doubt that these power relations are asymmetric, neither actor is completely bare of power resources (Crozier and Friedberg 1979; Burawoy 1979). This means, third, that managerial conduct, both in intrafirm relations and in firm-environment relations, is ‘bounded’ as much by the social context of the ‘firm’ as it is by external constraints. Like any other actor, management cannot gain access to a position of ‘systemic rationality’, since there is no way to act within a social system and simultaneously to be a neutral external observer of the same. In this sense, the term ‘bounded rationality’ aptly describes the substance and mode of strategic management conduct in the social context of the ‘firm’. On the one hand, management behaviour is constrained by the uncontrollable ‘zones of uncertainty’ and ineradicable ‘zones of autonomy’ inherent in the specific characteristics of the particular production processes. On the other hand, management itself is a social process including ‘micropolitical’ components. Management itself appears as a production process organised in what can be called a specific ‘workshop’ or ‘departmental’ setting designed to generate ‘focused’ intrafirm integration and to monopolise control over firm—environment relations. Both the social context of the wider ‘firm’ and the internal social context of management coalesce in determining how management interprets the cosmos of the ‘enterprise’ and the environment external to it. This monitoring eventually decides which options in management policy are considered feasible. Instead of reflecting the pursuit of preconceived goals by adhering to rigid abstract principles, managerial conduct rather bears a resemblance 80

CLASS STRUCTURES AND THE FIRM

to an incrementalistic ‘organic searching process’ along ‘corridors of decision’. This process constantly balances the potential prospects of flexibly adapting to current environmental pressures against the incalculable risks of forfeiting relatively stable and effective ‘production arrangements’ for returns which might prove advantageous only in the short-term perspective (Weltz 1988). The more a given industry depends on investment in non-deployable assets, consequently the more strongly management will rely on ‘satisficing’ rather than ‘maximising’. Finally, the ‘micropolitical’ approach to the ‘firm’ rightly insists on the comprehensive reconstruction of the shopfloor context in its entirety. This means that workers’ conduct can no longer be reduced to reactions to ‘experiences’ made with ‘wage labour’, but that the term encompasses the broad scope of work practices, routine procedures, patterns of cooperation, and communicative relations, both formal and informal, that make up the social process of production. The work process itself, including its technological and organisational features, has to be made an explanandum of analysis, since the spectrum of conceivable patterns of interaction within this context is not sufficiently determined by the capitalist relation of production underlying all of them. Therefore, industrial work in its routine aspects is as much in need of explanation as are patterns of conflict. The specific configuration of workplace and industrial relations shaped by the peculiar conditions of industries, business structures and spatial location may condition patterns of group formation, conflict-oriented action and organisational capabilities that extend into the spheres beyond the factory gates in a decisive and permeating way. Therefore, an approach allowing for structural diversity, while maintaining a notion of shared structural principles linking the patterns within this spectrum, proves superior to the concept of ‘homogenisation’ as a framework for explaining class structures in industrial capitalist societies. Furthermore, the relational approach to the workplace context proves capable of showing how closely intertwined relations of power and consent, control and autonomy, compliance and resistance, and cooperation and conflict appear, even in routine shopfloor practice (Welskopp 1994a). The ‘micropolitics’ approach to conceptualising the ‘enterprise’ as a ‘negotiated social order’ and a ‘terrain of contest’ provides a valuable starting point for recasting the analytical concept of the ‘firm’. However, it has to be stressed that, while this line of conceptualisation can inform theoretical debate in history, it is too fraught with limitations of its own to be adopted without qualifications. First, the necessary emphasis on the internal social relations of production has fostered a tendency to become fixated on this internal perspective. Hence, firm—environment relations and the feedback cycle between the ‘firm’ and the ‘non-firm’ sphere are unduly lost from sight. Second, while the approach appropriately establishes the relations between the ‘firm’ as a mesolevel concept and the microlevel of social interaction, it 81

THOMAS WELSKOPP

fails to address the relations between the institutional mesolevel and the macrolevel of society in a similarly convincing manner. This gulf between a relational concept of the ‘enterprise’ and a suitable theory of society, i.e. a modified concept of ‘class’ and of ‘industrial capitalism’, has to be bridged as a necessary precondition to explain how the structural principles of industrial capitalism condition the intrafirm patterns of social interaction in the peculiar way which marks off the ‘enterprise’ as a social institution from other types of organisation in modern society. Only the ‘linking back’ of ‘midrange concepts’ at the institutional mesolevel to macroconcepts of society equips ‘enterprise’ studies with conceptual tools needed to sort the diverse forms of interaction and relations into broader patterns, which can then be compared and contrasted typologically. Third, the concept of ‘micropolitics’ has taken great pains to establish that the ‘enterprise’ is a ‘field of social interaction’ governed by relations of power and consent. The concrete substance of those relations, however, has remained obscure. This points, on the one hand, to its lack of an appropriate macrotheoretical ‘backup’. On the other hand, it reveals that the concept has invited a reverse form of conflating social theory and historically specific theories of society: Here, theory of society appears as grafted on to social theory. Adopting the dual notion of ‘agency’ and ‘structure’ as propounded by British sociologist Anthony Giddens, the ‘micropolitics’ approach has convincingly established that the ‘firm’ is an arena of structured interaction. Yet it is a fallacy to contend that this already provides a sufficient explanation of how concrete structures pattern social relations in the shopfloor context. A theoretically sound reconceptualisation of the inner workings of capitalist enterprises, therefore, can build upon central elements as developed by the ‘micropolitics’ approach. Yet it has to transcend its limitations in a two-step procedure: 1 2

general social theory has to be revisited in order to ground the mesolevel concept of the ‘enterprise’ in a precise microtheoretical framework; this step will provide the basis for linking the mesolevel concept to a ‘contextualised’ and substantially modified concept of ‘class structures’ in industrial capitalist societies.

The concept of the ‘firm’ and critical social theory Anthony Giddens’s ‘theory of structuration’ as outlined in his seminal work Constitution of Society (Giddens 1984) proves particularly appropriate for historical research on three accounts: First, by replacing the dualistic notion of ‘agency’ and ‘structure’ with a conception of their respective ‘duality’, he offers a constructive solution to the key problem that has bedevilled historians

82

CLASS STRUCTURES AND THE FIRM

and sociologists alike for generations. Second, the contextualist nature and hierarchical structure of his social theoretical ‘guide’ to model construction enables the historian to distinguish between different levels of generalisation and hence to reconcile macrotheory and microtheory as well as macrohistory and microhistory. Third, Giddens’s concept distinguishes between general social theory, that lays out the features inherent to all social interaction, and substantive theories of society, which by definition relate to phenomena situated in concrete time—space. It therefore invites genuinely historical theory building (Giddens 1984, 1976). In Giddens’s view, it is flawed to depict ‘agency’ as a string of discrete intentional acts executed within the boundaries of rigid structural constraints, understood, for instance, like the girders of a building. Instead, ‘agency’ has to be conceived as a continuous process, a constant interweaving of internally stratified cycles of action related to one another. Therefore, ‘agency’ is inherently social and synonymous with interaction. This formulation incorporates a relational perspective into the very concept of ‘agency’. ‘Structure’ is not external to ‘agency’ and also not identical with ‘constraint’. Rather, ‘structure’ appears as the medium of interaction, thereby displaying ‘enabling’ as well as ‘constraining’ features. Interaction gravitates towards patterns of regularity due to the capacity of ‘structure’ to arrange the action cycle as ‘recurrent practice’. Instantiated in interaction, ‘structure’ denotes sets of ‘rules’ and ‘resources’ upon which agents draw in the course of social practice. ‘Rules’ refer to the constitutive (communicative) and regulatory aspects of interaction, to its modal characteristics. The term ‘resources’ relates to the allocative (‘power’) and authoritative (‘domination’, ‘legitimacy’) assets that agents are able to mobilise in order to achieve goals and to influence the outcome of situated practice. Agents are able to draw upon such ‘rules’ and ‘resources’ because they command a considerable practical knowledge of the conditions of their own actions—a knowledge which is mostly tacit but can at any time take on a discursive character. The flow of actions they produce contains a stratified array of behavioural forms ranging from more or less unconscious reactions to exogenous stimuli via the vast scope of routinised and habitualised actions, to complex ‘projects’ which combine and focus numerous elements of the former into purposive conduct. The cycle of action, consequently, involves the monitoring of the conditions of action, the recourse to practical knowledge which informs both this monitoring and the mobilisation of ‘rules’ and ‘resources’, the actual application of the latter, and the generation of practical consequences of action which appear as conditions of conduct at the onset of the next cycle (Giddens 1984). ‘Recurrent practice’ is the combined effect of unacknowledged conditions of action, of the quality and quantity of ‘resources’ at the agents’ disposal, of limited practical knowledge of ‘rules’ and ‘resources’, of ‘constraints’ of different qualities, and of unintended consequences of action. The extension and stability of patterns and regularities in interaction across time and space 83

THOMAS WELSKOPP

ranges from ephemeral ‘encounters’ to the lasting stamina of ‘institutions’, and varies with the ‘systemness’ of the respective social context. Stability increases with the degree to which ‘rules’ and ‘resources’ are systemically related and ‘transformable’ into one another. It is at this point where the dual nature of ‘structure’ enters the picture: ‘Structure’—as a set of ‘rules’ and ‘resources’—is a constitutive element of interaction and, by virtue of its relations to other such sets, part of larger social systems. ‘Structure’ mediates and generates interaction and is reproduced in the course of ‘recurrent practice’ (Kießling 1988:290). The concept of ‘system’ in ‘structuration theory’ is essentially a hierarchical one. At the macrolevel of society, ‘structural principles’ form a tight systemic context which lends cohesion to the organisation of societies as totalities. Descending down ‘axes of structuration’, which condition systemic coherence between mesolevel patterns of interaction and equip institutions with structural properties that determine their features and their high degree of integration, we encounter increasing diversity. This means that systemic arrangements of the same quality can manifest themselves in configurations of quite diverse forms. On the microlevel of social interaction, finally, the variety of forms may be immense, but the reciprocity of practices, conditioned by the systemic interrelatedness of sets of ‘rules’ and ‘resources’, patterns interactions in specific ways so as to secure the reproduction of the systems and institutions in which they occur. Along the ‘axes of structuration’, therefore, the ‘contextuality’ of structure progressively increases. While ‘structural principles’ are of ‘virtual existence’ in so far as they delineate ‘fields of potential interaction’ in which potential contexts emerge, ‘structural sets’ lend coherence to institutions as real social contexts within these fields and demarcate the spectrum of possible interaction. As ‘structural elements’, finally, ‘structure’ is embedded in real social interaction, constituting and reproducing the context. The grounding of historical concepts in the ‘theory of structuration’ so far has entailed a deliberate distinction between general social theory—as outlined on p. 88—and substantive concepts pertaining to theories of society. It is at this point that the links between the two have to be established. First, Giddens rightly emphasises the distinction between ‘structure’ and ‘structures’. To highlight the ‘structured’ nature of interactions does not anticipate their concrete outcome. The reference to ‘rules’ and ‘resources’ as determinants of the production and reproduction of ‘structure’ and ‘system’ does not already embrace statements about the kind and relative distribution of concrete authority and power potentials governing practice situated in concrete time and space. Therefore, substantive concepts have to be drawn upon to delineate the specific ‘rule-resource sets implicated in the institutional articulation of social systems’, which recursively pattern concrete interaction (Giddens 1984:377). This calls for a theory of society which-since it refers to social systems bound in time and space—is genuinely historical. The 84

CLASS STRUCTURES AND THE FIRM

conceptualisation of the ‘firm’, after being anchored in ‘structuration theory’, has still to fulfil the eminent task of grasping the peculiar characteristics of this type of institution by linking it to macroconcepts of ‘class’ and ‘industrial capitalism’. Second, as the argument above suggests, a sound conceptualisation of the ‘firm’ has to locate this social system at the mesolevel of ‘institutions’ and has to expand the concept in the direction of both microtheory and macrotheory. The macroconcepts of ‘class’ and ‘industrial capitalism’ are indispensable in order to avoid the ‘macroreductionism’ of the ‘micropolitics’ approach. Yet these concepts have to undergo substantial modification: They have to be recast into a hierarchical structure resembling the sequential frame of the ‘axes of structuration’, and they have to be transformed—on each stage of this hierarchy—into ‘contextualist’ models capable of both grasping structural diversity within the system as equivalent variations on shared structural properties, and of marking off this richness of form from phenomena exogenous to it. Table 3.1 charts the basic outlines of such a model’s internal structure.

‘Class relations’ and ‘class structures’ as axes of the structuration of ‘employment relations’ Combined with other structural principles, the ‘class relations’ inherent in capitalist production and the characteristics of ‘industrial capitalism’ segment modern societies into three distinct fields of social interaction which maintain close relationships among one another but whose respective peculiarities determine what forms these relationships assume. These fields are: 1 2

3

the enterprise as the unity of production and realisation, embedded in its environment of product, financial and labour markets; the (private) life-worlds exogenous to capitalist production, comprising the complex networks of family, neighbourhood, community, milieu and consumer relations; and the domain of formal organisation in the organisational—political sphere of society where the interplay of ‘civic society’ and the administrative apparatus of the state is situated.

‘Class structures’ and the constituent relations of ‘industrial capitalism’ are, of course, anchored in institutional backgrounds of the organisational-political sphere, most prominently as far as their legal and political aspects are concerned. Yet it is the network of enterprises and markets constituting the social cosmos of the capitalist economy which figures as the locale where these structural principles shape characteristic social relations (Welskopp 1994a).

85

THOMAS WELSKOPP

Table 3.1 Class structures and the ‘firm’ as a hierarchical and contextualist concept

‘Class structures’ are social configurations decisively patterned by ‘class relations’ between aggregates of social agents occupying antagonistic but mutually dependent ‘class positions’. The ownership of and/or the control over the means of production—or the lack of it on the side of ‘negatively privileged’ suppliers of labour power, skill and knowledge respectively— decide about the relative ‘chances [of the respective positions] to realize products and capabilities in the market’ (Marx 1979:182, 191; Weber 1980:177). 86

CLASS STRUCTURES AND THE FIRM

Yet this fundamental asymmetry of the ‘class relation’ does not override the fact that all ‘class positions’ are ‘bounded’ by their engagement in capitalist accumulation and consequently by the specific requirements of production processes and market relations. At this level, the labour market may be identified as the locale where latent ‘class relations’ are formed and reproduced, whereas the enterprise is the social sphere where these latent relations become manifest. ‘Class relations’ exist in the organisational—political sphere only in so far as organisations and state agencies themselves incorporate ‘employment relations’ and therefore display a hybrid character oscillating between ‘organisation’ and ‘enterprise’ with varying ratios of components. Although sometimes resembling ‘class structures’, as, for instance, is the case with landlord—tenant or bureaucracy—client relationships, and despite the often permeating influence ‘class structures’ exert in this sphere, social relations in the life-worlds outside of capitalist production are not ‘class relations’ proper. Similarly, it is right to contend in accordance with Anthony Giddens that preceding or discordant societal formations may elicit class-like segmentations, but that it is unique to industrial capitalist societies to be centred around ‘class relations’ as a structural principle that characterises them as ‘class societies’ (Giddens 1972). The peculiar character embodied in the ‘class relation’ consists in that only this type of relationship mediates the direct transformation of market power into formal authority relations within the enterprise, and vice versa. This character finds expression in the principal ambivalence of the ‘class relation’ which represents—simultaneously—a formally voluntary contractual relation, which in truth codifies a variable market power differential between agents who are in terms of the law equally entitled, and a relation of domination within a production setting whose contents and extension are only specified in broad outline and are substantiated by concrete power relations emerging from the specific conditions of production (Welskopp 1994a). The ‘class relation’ by definition bears a dichotomous character. Yet this does not mean that there are only two antagonistic ‘class positions’ in society. While economic independence bestows upon small capital holders, as well as upon members of certain professions who market arcane expertise shielded by limited entrance, an autonomy outside the context of large enterprises that places them on an intermediary position in the ‘class structure’ of modern societies, the command over scarce expert knowledge, special skills, and functional indispensability in the organisation of production processes translates into equivalent positions within larger-scale enterprises. These intermediary positions within the firm, on account of their strong power resources, block capital’s prerogatives to a large extent and utilise this superior autonomy in turn to infiltrate delegated and usurped management prerogatives in order to establish class-like power differentials versus subordinate groups of employees. This hybrid position indicates that an intrafirm ‘class structure’ combines ‘class relations’ of common principle but of different qualities 87

THOMAS WELSKOPP

within it. This mechanism concretises the long-disputed definition of ‘middleclass’ positions in class-based societies for the area of the ‘new’ intermediate social groups and describes these as firmly anchored in the ‘class structure’ of modern societies (Blumin 1985). This intermediary rank in a stable network of ‘class positions’ may be viewed as a source spurring social dynamism within its boundaries, since different social groups occupy this position in the course of their development, but the loss of specific power resources and ‘bargaining positions’ can cause their ousting by cadres of a different social and professional background. The highly skilled craft ‘team leaders’ in the iron and steel industry, for instance, were forced to surrender their organisational competence to ‘production foremen’ and their technical decision-making power to engineers with their knowledge and proficiency intact. In turn, just a decade later the engineers wrestled the organisational competence from the hands of the foremen and became departmental heads. In this function, they subjected the social organisation of work to their reductive logic and rhetoric of ‘technological feasibility’ which had been instrumental to their ascent at the expense of the ‘craft workers’ in the first place (Welskopp 1994b). This highly formal and abstract definition of the ‘class relation’ suffices to delineate the axis of structuration which demarcates the ‘enterprise context’ in industrial capitalist societies on two grounds. First, it is obvious that the ‘class relation’ is prone to generate conflict since it entails a structural contradiction, in that its elements depend upon and yet negate the others. In ‘class relations’, gains of the one position are inclined to be realised at the expense of its antagonist, yet both positions are forced to engage in factual cooperation in order to produce what is constantly embattled in precarious contests of power and bargaining processes. If both positions were maximisers without relying on conventions governing constructive cooperation and without access to intragroup relations superseding ‘prisoner’s dilemma’ constellations, the outcome would consist of low effort and low rewards (Mosner 1992:23ff.). Social systems of production, therefore, always represent fragile and transient balances of conflict and compliance. It is of utmost importance to consider in the definition of ‘class relations’ the obvious fact that capitalist enterprises are social systems which—to a large degree—are working. It is also significant to acknowledge this arrangement’s propensity to conflict which often immediately extends into the organisational—political sphere in the form of both organisational impulses and impulses forcing related issues on to the larger political agenda. It is misleading, however, to expand the very definition of ‘class relations’ in order to embrace the specific organisational and institutional outcomes of such processes as the practitioners of ‘working-class formation’ theory do. ‘Class’ is not a social entity which acts as a social group.

88

CLASS STRUCTURES AND THE FIRM

Class-based group formation and organisation building by such groups are to be analysed as complex social processes of organisation in their own right. The capability to form groups and consolidate them in organisations transcending the shopfloor setting depends too starkly on conditions inherent in particular production contexts. The evolving forms of organisation, of intraorganisational ‘micropolitics’ and of organisational conduct, are likewise too strongly co-determined by their institutional context to be covered appropriately by a definition at this level of abstraction. The exploration and explanation of both consequently calls for midrange concepts at the mesolevel of analysis. Second, the various social arrangements among and within enterprises in industrial capitalist societies and the diverse forms of group conduct in their context share the ‘class relation’ at their structural core. Hence even the most purely ‘economic’ accounts of the ‘firm’ are bound to relate to the social patterns conditioned by it. The concept outlined above has, however, pointed to the fact that the diversity of these forms cannot be covered by the abstract definition of either ‘class’ or ‘industrial capitalism’, since concrete arrangements are decisively co-determined by market conditions and the requirements of production processes which in their concreteness contribute to constitute specific contexts. It may be worthwhile to consider, as this concept suggests, that ‘industrial capitalism’ in this sense is at least as ‘industrialist’ as it is ‘capitalistic’. The structural principles of ‘class’ and ‘industrial capitalism’ manifest themselves in social interaction nowhere else except at the mesolevel of institutions, i.e. in the enterprise in its market environment. Therefore, we have to descend to this mesolevel of inquiry and subsequently to expand those macroconcepts to a theory of the spectrum of conceivable patterns of social relations within this context in order to apply an explanatory framework which indicates how ‘class’ and ‘industrial capitalism’ translate into the inner workings of capitalist enterprises and into institutional conduct. The explanatory burden of this type of analysis is carried by historical midrange concepts addressing relatively stable configurations of patterns within this spectrum. They invite comparison and provide a basis for constructing typologies which then, in turn, can inform broader generalisation.

‘Class relations’ in social systems of production: the ‘enterprise’ context The ‘enterprise’ in industrial capitalist societies is a relatively open social system clustered around determinate sets of ‘class relations’ and maintaining specific patterns of relationships to systemic environments consisting of product, financial, technology and labour markets, ties to institutions in the organisational—political sphere, and links to the social life-worlds of the social groups involved in production. This social system of production

89

THOMAS WELSKOPP

is of a nodal character, in that it represents a complex interplay of work procedures, cooperative relations, communicative relations, power relations and ‘firm’—environment relations. Since this system is bound by the requirements and intricacies of specific production processes, markets, and institutional links to the outside world, it is concretely situated in time— space and, therefore, bears a genuinely historical character. Social systems of production are hence to be grasped as relatively stable configurations whose stamina rests on the congruence of certain factors like market structures, the scale and scope of production, the product life cycle, the production technology applied, the social organisation of production, and business structures. The very same factors that figure as the elements lending stability to these systems can bring about their disintegration and subsequent replacement by other arrangements if and when inputs change in quantity and—often as a result of the former—in quality. A given industry, as can be demonstrated in the case of the iron and steel industry in Germany, the US and Great Britain, can pass through a sequence of such distinct configurations in the course of its long-term development (Welskopp 1994b; Nuwer 1988; Elbaum and Wilkinson 1979). Also, at any given time a society displays a determinate lateral spectrum of these social arrangements conditioned by its particular industrial structure and its institutional reverberations on business structures. These configurations are not simply reflections of environmental pressures, nor immediate responses to input impact. Neither do they mirror preconceived ‘master schemes’ implemented and unilaterally dominated by management. Rather, the social context of these systems determines how to interpret their environments and how to ‘channel down’ external inputs to the level of intrafirm organisation and shop floor practice. These multiple transformation processes occur as powerstructured bargaining contests and consensus-finding procedures between alliances among and between all social groups involved in production. They occur in the social practice of shopfloor performance. Although social systems of production involve ‘bounded’ strategic planning, particularly— if not exclusively—on the side of management, they ‘emerge’ rather than being brought forth by one-sided implementation (Birke 1992:7, 44ff.). In industrial capitalist societies, social systems of production are moderately stable constellations. These arrangements fall into three categories of distinct if interpenetrating social relations: workplace relations; hierarchical employment relations; and patterns of communicative relations. Workplace relations The social context of production contains more than the sum of the individual employees participating in the collective effort of work. It represents the ambivalent ‘unity of immediate work practice exerted to produce goods and of social practice within and around performance, which is specifically shaped 90

CLASS STRUCTURES AND THE FIRM

by the production process and simultaneously affects and modifies work practice’ (Lichte 1978:23f.). Hence the shopfloor setting is understood ‘not only as the unit that executes the material transformation (of raw materials and auxiliaries by means of production facilities into finished products by human labour) but also as the unit of social transformation, i.e. the interpretation and construction of both intra-and extrafirm reality in and by work practice’ (Pries 1991:121). The intricate web of production processes effected by the respective degree and form of task division constitutes a social cosmos, a network of ‘shopfloor community’ relations (‘Werkstattgemeinschaften’) which represents the social mode in which the collective effort of production is exerted. This social mode of ‘work group conduct’ (‘Belegschaftshandeln’), as it is termed by Rainer Lichte, depends in its character and consequences on the ‘material conditions of the specific production process, the form of task division, the cooperation among specialised [co-working] operatives, and on the technical standard of the production facilities’ (Lichte 1978:31). Task division normally does not simply mean task fragmentation, but rather signifies a specialisation and decentralisation of tasks that requires re-integration by internal coordination and versatile hand-in-hand cooperation among and between the work crews composing the intrinsic collectivity of the particular labour process (Robertson and Alston 1992). The production process proper elicits ‘shopfloor community’ relations irrespective of—if not uninfluenced by—its constituent members’ individual dispositions, social backgrounds, biographies and overall mentalities, because their function as a transformational mediator is a necessary precondition for its very working. The social context of production can, in reverse, produce social logics of ‘meaning’ that inform patterns of experience and dispositions to be translated into the life-worlds beyond the factory gates, if they correspond with perceptions made in those spheres. Workgroup autonomy in the iron and steel industry, for example, has informed a dualistic perception of ‘mill society’—pitting the ‘actually producing’ work crews against a distant management viewed as autocratic outsiders—which was easily translated into a polarised vision of ‘class society’ at large and was bolstered by regional and community identities displaying widespread distrust and antipathy towards ‘the company’ and ‘government’ (Welskopp 1994b). To varying degrees—according to types of industries and their respective stages of development—work procedures are infiltrated, appropriated and transformed into a ‘kind of social affair’ by those involved, not only in terms of filling out the niches and pores of ‘sub-optimally organised’ production processes, but rather in terms of accommodating themselves to their requirements, of mastering their intricacies, of coping with their strains, and of turning them as far as possible to suit their own purposes and interests. The web of functional task-execution, work routines, cooperation, coordination, and monitoring proper, constitutes the arena where formal and informal relations are formed and inextricably intertwined—not as a 91

THOMAS WELSKOPP

distortion of functional rigidities and an impediment to ‘rationality’ and ‘efficiency’, but rather as the mode of constructing the social reality of production. In varying combinations and ratios of components, this social appropriation of work takes place on all levels of intrafirm organisation and embraces commitment and compliance, as well as ‘distancing’ and ‘a multiplicity of miniature wars or of one-dimensional (partialized) intrafirm contests which reproduce everyday routine in perpetually changing qualities’ (Lichte 1978, 49). Basically, all work settings that coexist in a firm—be they manual, clerical or managerial—display these characteristics and hence lend themselves to the application of the same categories of analysis Workplace relations exist among employees engaged in an immediate production setting, and between the different types of such ‘workshop communities’ composing the production process at the departmental or plant level. The internal structures of the work groups, i.e. their size, internal differentiation (including their gendering), stability, their mechanisms regulating recruitment or cooptation and—depending on the characteristics of the respective production process—social patterns imported from the outside life-worlds combine to comprise one spectrum of conditions that pattern workplace relations. This cluster of conditions is linked to the configuration of ‘paratechnical relations’ (Anthony Giddens 1972) which denote a second set of conditions delineating the patterns of workplace interaction. These are dependent both on the requirements of the specific tasks to be executed and on the social arrangements emanating from the technical and organisational systems of production. They combine to determine skill levels and skill profiles (in terms of both formal qualification and the actual proficiency and knowledge necessary to master a particular job), degrees and forms of job autonomy, responsibility, and work-related discretion, the formal and informal organisation of work, and the social consequences of the production technology applied. Within this matrix of conditions, workplace relations appear as patterns of social interaction in the material process of work which emanate from the ‘cooperative collectivity’ among the employees involved. These embrace elements of internal power relations and competitive dispositions as well as those of mutual dependence, companionship, ‘peer group’ discipline, ‘group ethos’, solidarity and, not least, tenacious group rites and routines. The spatiality of work as embodied in workplace design and plant layout coalesces with working hours, forms of payment and other incentives, the scale and scope of production, and the physical strains involved in the work (e.g. heat, risk of injury, monotony, fatigue, and exposition to fumes, dust and grease) to compose a cluster of ‘intermediary’ conditions which nevertheless directly affect workplace relations and are—as are the other conditions—likewise the product and consequences of the patterns of interaction they help to establish. Workplace relations form systemically related sets shaped by and shaping the social systems of production in whose contexts they emerge. A term apt 92

CLASS STRUCTURES AND THE FIRM

to express the ‘systemness’ of these sets may be that of production milieus. To characterise particular production milieus and to subsequently create typologies thereof, it is important to elucidate the degree to which their elements can be turned by the employees into active resources utilised for group formation, the eliciting of solidaristic conduct, the appropriation of organisational capabilities, and the exertion of ‘bargaining power’ within the actual work processes. These resources can entail far-reaching consequences for the larger set of corporatist industrial relations, since they may instigate organisational processes that transcend the boundaries of the particular workplace context. Michael Burawoy has astutely observed that the execution of monotonous tasks often resembles the ‘games’ and ‘rituals’ by means of which operatives— both individually and collectively—accommodate to otherwise dull and numbing routines. These ‘games’ are pivotal to creating an informal ‘work rhythm’ which, on the one hand, ensures that a certain degree of effort is invested in even the most deadening work procedures, and, on the other hand, forms an invisible but obstinate barrier to the speeding-up of work processes, for example, by advanced machinery. Attempts by management to increase the pace of the process can be effectively blocked by ‘games’ which involve deliberately poor performance, or even the enforcement of slow-downs and standstills by wrecking parts of the machinery. Moreover, particularly if such attempts indeed are successful in superseding ‘blocking games’ of this vein, they can lead to the complete breakdown of the informal and tacit arrangements that previously had kept up the production process, so that the organisation of work only then poses a serious follow-up problem to management (Burawoy 1979). Since such ‘games’ tacitly ‘emerge’, rather than being overtly stipulated among otherwise highly fragmented workforces, they appear as a fairly weak and reactive resource not particularly suited to generate group formation and conscious solidaristic conduct. In this respect, those types of workplace relations differ from production milieus which comprise more skilled tasks and necessitate, as well as facilitate, internal coordination, negotiation, and hence ‘deep’ communication, be it formal or informal. Although representing a serial workplace structure broken down into mainly individual job positions, for example, shopfloor settings in the metal trades prove propitious to cohesive group formation, since precision work and task synchronisation demand sustained knowledgeable negotiation on the basis of shared and selfconsciously displayed skills. Since metal workers have to leave their workplaces in order to communicate with each other, their informal groups originating in such contexts bear a decidedly associational character. Shared job pride, a collective work ethos, and intragroup contests of position, combine to translate this associational group structure into the departmental ‘lunch break groups’ inside the factory as well as the ‘focused gatherings’ of local union meetings outside of the plant. Despite a considerable autonomy 93

THOMAS WELSKOPP

in the execution of individual tasks, the informal groups thus formed are usually not able to exert direct group power in the actual work processes, which would prove superior to the ‘games’ in lesser-skilled work settings. They constitute a mechanism, however, well suited to stipulate and sanction solidaristic behaviour, and they represent a formidable organisational resource both in intrafirm conflict and for the consolidation of group networks into broader-interest organisations like craft or industrial unions. Work settings centred around the constant regulation of intricate continuous flow processes involving the close cooperation among skilled operatives as a collectivity (which have, for instance, prevailed in the iron and steel industry since the 1920s) typify a context that generates a third type of group formation. Here, particularly tight work groups are formed in and by the work process; close hand-in-hand cooperation and high mutual dependence of the crew members necessitate solidarity as the very prerequisite of mastering their collective tasks. Functional group formation of this type creates ‘considerably stable forms of shopfloor community organisation…which [are] largely removed from encroachments by shopfloor line management because they are anchored in the contradictory conditions of the production process itself (Lichte 1978:505). As they occupy strategic positions in extremely vulnerable, high-throughput processes, functional work groups are capable of exerting direct group power in the actual production process by transforming their intrinsic collectivity and their inherent collective functional autonomy into a potential ‘blocking force’ instrumental to bring the entire process to a complete standstill. As an example (quoted in Welskopp 1994b: 574), Joe Gyurko, a steelworker at Inland Steel’s Indiana Harbor Works since the 1930s, recalls that in the 1936–42 pre-contract period, departmental strikes were common. When foremen or supervisors refused to deal with pressing issues affecting working conditions, the men thought nothing of stopping work and letting gondolas full of molten steel hang in mid-air. In these situations the rapidly approaching danger that production would be interrupted in order to clean out the gondolas and reheat the steel acted as a time clock, forcing the company to bargain with the workers. More often than not, the supervisors settled with departmental workers before much production time was lost. This force conveys upon the work crews a formidable if precarious intrafirm ‘bargaining power’. Intraworkgroup cohesion also directly translates into overall group solidarity fostered by collective job pride and a pronounced ‘esprit de corps’. Functional workgroups figure as potent pre-structured basic units of collective action, both at the department or plant level and beyond. Yet they are also strongly ‘bounded’ by the immediate context in which they originate, and, therefore, their organisational capabilities are limited by a fixation on departmental issues as well as by the spatial isolation of their domain. In this respect they differ from the associational groups encountered in the metal trades and elsewhere, which more easily translate 94

CLASS STRUCTURES AND THE FIRM

into forums of broader organisational activity beyond the confines of individual plants. Any consolidation of functional work groups into comprehensive and inclusive interest organisations thus requires an already established institutional infrastructure with a professional apparatus which, from the outside, supersedes group or plant isolation and effectively integrates and harnesses their strong but shopfloor-bound organisational impulses. The United Steelworkers of America, for instance, were organised by a professional apparatus recruited from other, established unions, and they consolidated as an organisation which remained a decentralised, syndicalist network of strong, plant-based locals integrated by increasingly bureaucratic professional cadres (Welskopp 1994b). The three production milieus sketched above only represent basic types in a spectrum that is considerably more comprehensive. Yet their features do exemplify the fact that production milieus are systemically related in multiple ways and varying intensity to the organisational-political sphere and the social life-worlds beyond the factory gates. These ‘interfaces’ to the other fields of social interaction derive their particular form and consequences from the specific characteristics of their respective shopfloor milieu. On the one hand, workplace relations generate social impulses which are launched into the external sphere of action: As has been shown, they can provide the basis for the formation of work-related ‘shopfloor communities’ which may either directly engage in combative ‘work group conduct’ at the plant level, or advance to peculiarly pre-structured units of collective action which, when channelled into comprehensive processes of mobilisation and organisation and harnessed by established institutions, contribute their specific sociological features to the very character of the movement at large. Conversely, the organisational ‘micropolitics’ inherent in these institutions and the broader institutional context in which they act may reverberate on the social basis at the shopfloor level. The degree to which organisational constraints penetrate to this sphere largely depends on the relative cohesion and autonomy of the group structures composing this basis. Likewise, production milieus can radiate into the social lifeworlds beyond the factory gates and influence the social milieus which blend gender, ethnic, family, neighbourhood and community relations into tenacious amalgamations. Yet the mutually reinforcing amalgamation of shopfloor and milieu relations into a tight network that extends from the workplace via the community and its adherent associations into the sphere of union and party politics—like practitioners of the ‘working class formation’ approach postulate as the single configuration matching a true ‘class’ constellation—might only represent a rather exceptional pattern of interrelations based on specific circumstances. Production milieus and social milieus may well coexist without bearing significant influence on one another or on their respective forms of organisational conduct. Likewise,

95

THOMAS WELSKOPP

patterns of mutual paralysis have to be reckoned with (Welskopp 1994b: 700–30). On the other hand, certain production milieus invite the intrusion of exogenous social relationships into the shopfloor context. The degree to which this may happen depends on the relative flexibility and porosity of the respective production setting. Such an ‘import’ of external social relations and networks—which usually rest upon gender, family, kin, and ethnicity— can occur irrespective of management’s objectives as a process of infiltrating the niches and pores carved out by the requirements of specific production processes. In other cases, lack of organisational competence and a high degree of organisational flexibility will prompt management to ‘colonise’ and harness groups already possessing a social structure by incorporating them into the internal fabric of the enterprise. Workplace relations, finally, may be influenced by reverberations from the organisational-political sphere which originate in political movements relatively independent of ‘shopfloor community’ conduct. On a more general level, strong production milieus will radiate into the other fields of social interaction and prove rather immune against outside influences. Inclusive production milieus have to be conceived as representing elements which combine with features of social milieus to form complementary unities that achieve cohesion only by mutual reinforcement. Open production milieus, by contrast, are sufficiently permeable to attract their infiltration by exogenous patterns of social interaction (Hareven 1982; Kleinberg 1989). Hierarchical employment relations Horizontal workplace relations as delineated in the previous section are inextricably and reciprocally interwoven with the vertical patterns of the hierarchial relations of production, or, in other words, the intrafirm industrial relations which characterise the enterprise’s unique ‘dual quality’ as a system of domination and—simultaneously—of production and realisation. It is in the variety of these ‘employment relation sets’ and the forms of social interaction governed by them that ‘class relations’ manifest themselves concretely, where vestiges of ‘class’ can be pinpointed empirically, and where ‘class relations’ broaden out into clusters of ‘class structures’ conditioning concrete social practice. ‘Employment relations’ represent complex configurations comprising elements of power and formal domination as well as those of material cooperation, bargaining, compliance and consent. The interplay of the following four factors demarcates the spectrum within which those configurations arise as components of the larger social systems of production they substantially help to structure: First, the respective constellation of workplace and intrafirm industrial relations decisively depends on the question of which groups of social agents involved in production acquire and sustain the capability and competence 96

CLASS STRUCTURES AND THE FIRM

to determine the technical organisation of the production processes. The way in which these capabilities influence both vertical and horizontal shopfloor relations further depends on the objectives pursued by those at the helm and on the practical technical knowledge at their command. The social agents in question act on the background of what they perceive to be the ‘enterprise’ as a social and technical system. Hence their conduct rests on interpretive monitoring, informed by the mapping of feasible options and objectives derived from their operative knowledge and the scanning of information flows as established both within the firm and in firm—environment relations. Yet the actual consequences for the entire complex of workplace and employment relations which technical decisions produce may substantially differ from what had been preconceived and intended (Minssen 1992:78ff.). Second, the same mechanism applies to the social organisation of labour processes. The interplay of both factors is paramount for the crystallisation of the firm’s ‘dual reality’ representing the tensions and incongruities between its formal organisational structure and its real internal power structure which can—and as a rule will—substantially deviate from and reverberate on the former. It is here that the usually universal claims to unrestricted entrepreneurial and managerial authority clash with barriers to their realisation, barriers anchored in the real power fabric of the enterprise and erected by (acknowledged and unacknowledged) labour market conditions. These claims are infused into intrafirm social relations by the establishment of control regimes aimed at obtaining necessary information, at channelling work practice into efficiency, at maintaining disciplinary standards (whether or not they improve actual performance), and at eliciting effort either by surveillance and sanctions (in low-trust organisations) or by incentives and the sanctioning of output performance (in high-trust organisations). These control regimes—which emerge ‘organically’ rather than representing sequences of implemented ‘pure’ and ‘neat systems’ as implied in neoMarxist theory—consist of joint planning, technical instruction and direction, various forms of direct performance control and surveillance, as well as incentive schemes and sanctions in diverse ratios of components. Although forming an integral element in the respective social system of production, these control regimes do not necessarily determine the social reality of the labour process in its full scope. Relative efficiency of production may, on the contrary, be the result of their very limitations. Under specific circumstances, close control can even hamper productivity if, for instance, extensive surveillance of behaviour proves irrelevant to production results but in turn stifles effort and initiative central to the effectiveness of the particular production arrangement. In short, there is an intrinsic ‘dialectic of control’ (Anthony Giddens 1972) which substantiates itself in the confrontations between the firm’s formal authority structure and actual workplace relations. The determination of the latter by the requirements of both the tasks executed and the cooperative nature of the work, as anchored 97

THOMAS WELSKOPP

in the peculiarities of the specific production process, carves out spheres of autonomy and work-related discretion which prove impenetrable to claims to full control and detailed regulation. The incorporation of the firm—and the particular industry as a whole—into specific product, financial, technology and labour markets produces a second set of constraints to the scale and scope of managerial control. On the one hand, the paramount task of the firm to secure its survival and achieve ‘satisficing’ returns under dynamic and uncertain, and indeed uncontrollable, market conditions, which are specific to the industry and its interfirm context, massively qualifies, focuses and constrains the scale and scope of intrafirm regulation and control. Contrary to Stephen A.Marglin’s (1974, 1984, 1991) contention that ‘control’ rather than efficiency is the essential objective of managerial conduct, it must be emphasised that the character of the enterprise as the unity of production and realisation entails ‘the fact that industrial relations decisions often constitute second order choices which are deeply influenced by first order business policy considerations’ (Plowman 1991:72). Management will as a rule employ those organisational means which it perceives to vouch for maximal managerial autonomy in an uncontrollable market and corporate environment. In order to cope with internal as well as external uncertainty, ‘satisficing’ control may be already attained if management considers itself to be in full command. The reduction of external uncertainty, by contrast, is often pursued by internalising and transforming contractual into organisational relations which may then produce new internal uncertainties. Maximal autonomy, therefore, means maximising the scope to manoeuvre between the contradictory objectives of sustaining maximal market flexibility and minimising the internal impact of market changes on an organisation once it is perceived as being effective and providing ‘satisficing’ control. On the other hand, firm survival persistently requires the commitment to invest in non-deployable assets, in technology whose influences on the social systems of production are neither fully predictable nor completely controllable. Since technology, once applied, lends itself only to limited remoulding to suit the purposes of a given control regime, management will follow a similar ‘satisficing’ course by balancing the perceived ‘certainty’ of a tried organisational system against the risks of adjusting it to the entire range of possible requirements inherent in the remodelled production process. The successful implementation of a new technology conceived as cost-effective and productive may need to be preceded by a reorganisation to cope with the potential social consequences that the technology may produce. Third, this matrix of conditions is further complicated by the fact that management itself on all levels of the hierarchy constitutes social contexts which obscure the linear vision of firm structure as mirrored in official ‘organisational charts’. Strategic management can only commit limited organisational resources to matters of production control because it is to a 98

CLASS STRUCTURES AND THE FIRM

large degree absorbed in the maintenance of firm-environment relations. Therefore, intermediary positions are occupied by social groups located between workforce and central management in the formal as well as the informal authority structure of the enterprise which do not act as mere transmitters and executors of managerial directives. The historical role of gang leaders, masters, foremen, engineers, superintendents and plant managers oscillates between that of technical experts, organisers of production and labour processes, personal instructors and controllers, and that of mid-level (plant) strategic management, according to the requirements of the respective social system of production. The conditions of the latter also determine the form and scope of autonomous group power, which is derived from both the delegation and the usurpation of authority and allocative competence originally resting with the entrepreneur and/or strategic management. Functional autonomy in organisation as well as surveillance, and the indispensability at this level of functions, combine to prompt the formation of intermediary power groupings which occupy—by virtue of arcane expertise and the control over information flows— autonomous informal power positions in the social structure of the firm which infiltrate and supersede its linear formal hierarchy. These groupings can be ousted from their prominent position if they cease to be capable of monopolising their knowledge and functions, but this degradation also affects their replacement by other social groupings so that the intermediary position proper remains untouched. Except in staff settings which take on a workshop character in their own right, these groupings exert their power not as collectivities but as aggregates of individual functionaries in the technical and organisational process of production itself. Their intermediary power position shapes the relations within these groupings as well as, more importantly, the vertical relations to the workforce they superintend, and the vertical relations to strategic management, vis-à-vis whom they claim and defend qualitatively extended spheres of autonomy and allocative competence. They are, therefore, capable of gaining access to prominent, ‘professional’, ‘bargaining positions’. Whereas they act as transformatory pivots of ‘class relations’ as far as the workforce is concerned, they largely block and usurp certain management prerogatives implicated in the ‘class relations’ concerning their own position (Welskopp 1994a). Fourth, the varying degree to which formalised zones of negotiation and bargaining are incorporated into an otherwise rigidly linear formal authority structure reflects the extent to which interest organisations in the institutional context of the organisational—political sphere prove capable of influencing the power structure of the firm via the ‘interface’ of corporate industrial relations.

99

THOMAS WELSKOPP

Patterns of communicative relations Patterns of communication are, of course, implicated in the web of workplace and intrafirm industrial relations in so far as these relations by definition involve communicative aspects. The specific modes of communication, however, whether formally established and overtly accepted or driven into the twilight zones of informality, do bring to bear a considerable leverage on their forms and consequences. The communicative modes of workplace relations determine the prospects of mutual stipulation, the conditions for group formation (both formal and informal), and the potential forms of solidaristic collective action. Communication, however deeply embedded in the work process proper, plays a pivotal role in the diffusion of collective practical knowledge and, therefore, links power resources inherent in a particular configuration of production to actual collective action, which may range from tacit complicity and the subversive transformation of work routines to the overt proceedings of ‘bargaining groups’ and combative ‘workgroup conduct’. It also creates power resources, predominantly when consolidating primary workgroups into movements transcending the perimeters of individual plants and firms. Primary group formation at the plant level, conversely, may subvert established modes of communication and create new ones, which then reinforce group cohesion and organisational capabilities. Patterns of communication which emerge as a functional necessity in the planning, coordination and synchronisation of work processes often simultaneously figure as informal channels serving to sustain intragroup coherence and solidarity. Group cohesion may even be more pronounced the deeper it is anchored in the communicative aspects of collective task execution and the tighter it is linked to the habitualised contacts of routine interaction. Yet the very strength exhibited in the immediate workplace context by primary workgroups clustered around such functional modes of communication may limit their ability to form alliances and coalitions capable of transcending the department or plant boundaries, since these modes are only to a limited degree convertible into organisational resources. The communicative modes established in vertical employment relations influence not only the scope of information flows which provide the data upon which managerial performance monitoring and decision-making rely, but also the degree to which conditions of production are negotiated in formal and informal bargaining or in overt and covert power contests. Open bargaining may constitute a formal element in the planning and coordinating aspects of employment relations. The intrafirm industrial relations inherent in craft production, for instance, resemble a loose but formally established, constantly renegotiated contractual cash nexus that harnesses the insurmountable competence and functional autonomy of skilled craft worker ‘teams’ into a system of ‘responsible autonomy’ which

100

CLASS STRUCTURES AND THE FIRM

elicits effort by remunerating and sanctioning production results. Work processes demanding a high degree of adaptability to idiosyncratic conditions and of swiftness in their adjustments to those circumstances may not display such a formalised, contractual mode of communication but still have to rely on constant open participation and compliance of the workforce involved. Indeed, in the light of these propositions, ‘Taylorism’ can be interpreted less as an attempt to deprive skilled workers of their knowledge and proficiency proper than as a managerial device to replace open task negotiation between foremen and skilled work teams by linear authority structures—with the perverse and unintended consequence that the foremen’s autonomy and power were actually enhanced, skill profiles remained largely intact, and modes of negotiation prevailed, since they were only driven into the realm of informal bargaining (Tolliday and Zeitlin 1991:7, 11). Historical evidence amply supports the proposition that this elimination of formal forums for negotiation backfired in the sense that ‘silencing’ bargaining and participation compelled otherwise intact solidaristic ‘shopfloor communities’ to take recourse to more tacit and intrinsically more combative forms of engaging in controversies over authority and work-related decisions. Rigidly linear management structures, which are the hallmark of most mass production industries, actually owe their design to the aims of focusing information flows, of minimising open vertical communication and negotiation, and of reducing complexity in order to secure maximal managerial autonomy. Yet the modern decentralised and flexibly re-integrated production processes, represented by the fully mechanised continuous-flow facilities dominating production in most largescale industries since the turn of this century, require a degree of internal synchronisation and skilful coordination in close proximity to the workplace not provided for by these linear management structures (Nuwer 1988). In consequence, informal spheres of functional workforce autonomy and participation arise, which perpetually clash with the persisting autocratic line management. Hence ‘human engineering’ essentially represents a rather ineffective managerial device to supplement the limited scope of mechanistic management structures without relinquishing this linear control regime. While systematic observation shall, in this case, replace bilateral communication and substitute for the acknowledgement of the participatory elements inherent in production, it actually fails to penetrate the functional autonomy of work groups in fully mechanised production. Consequently, the informal solidarity networks keeping up production appropriate informal participatory competences which persistently challenge the intransigence of management and become the basis for an ongoing struggle—both overt and covert—to establish formal work-related participation (or: codetermination) under adverse circumstances. This controversy in most core industries has eventually brought about the implementation of industrial unions, drawing their strength from syndicalistic networks of intrafirm rank101

THOMAS WELSKOPP

and-file groups incorporated into and consolidated by elaborate systems of plant-related employee representation of various forms. Yet it should be noted that this representation still tends to provide formal participation only at a subsidiary level, where work-related concerns have to be translated into social issues in order to be considered ‘negotiable’ by an otherwise conventionally mechanistic and autocratic management. This may be the reason why the ‘team work’ aspects of ‘lean production’ concepts—which represent a late adjustment of management to conditions of production already present for a couple of decades rather than a revolution in the organisation of work—are often considered as a potential threat to authority by shop stewards, grievance committee members and works councils (Birke 1992:65ff.). Linear formal management structures, moreover, rely on the assumption that the deliberate limitation of information is inclined to focus information flows on the essential data necessary for effective decision-making. Their combination of observation and surveillance as an alternative to bilateral communication, however, only drives negotiation and bargaining back into the informal niches and undersides of employment relations. In addition, the more rigid formal authority structures are, the more they are prone both to be blocked by zones of autonomy they cannot penetrate and to be infiltrated and usurped by intermediary groupings. In consequence, information flows are broken and diverted in manifold ways. Linear structures create the very autonomy they aim to eliminate. Thus employment relations prove to be a particularly malleable and ‘uncertain’ terrain of ongoing tacit power contests and ‘small-scale wars’ precisely when cast into the procrustean bed of rigid linear authority structures, since in that case arenas in which negotiation and controversy could be controlled by rules are lacking and the scope of issues to be tackled by bargaining remains undefined. Thus the reduction of internal uncertainty, which is a crucial prerequisite for managerial autonomy, often appears to be attained at the expense of incomplete and distorted information. However, since management decisions rely on this incomplete information, they can affect employment relations in totally unintended and often conflict-generating ways, instead of reasserting compliance and efficiency.

‘Class relations’ as system-specific power resources in the configuration of workplace and intrafirm industrial relations This complex web of workplace relations, intrafirm industrial relations and communicative relations among the workforce, low-and mid-level management, and top or strategic management, derives its systemic coherence from the interplay of four distinct if interrelated types of power resources which shape the intrinsic asymmetry of vertical employment relations centred

102

CLASS STRUCTURES AND THE FIRM

around them. Practically, the concrete constellation and distribution of these power resources and their range, which is delimited by structurally embedded ‘zones of uncertainty’, represent the ‘substance’ of ‘class’ in the enterprise context, the specific form in which the ‘class relation’ manifests itself at the level of the ‘firm’ and patterns concrete interaction. In reverse perspective, their configuration unites all the various sets of employment relations as variations on the common structural principle of the ‘class structure’ within the enterprise context. The power to make decisions Of these four types, the power to make decisions about investments, to allocate capital in non-redeployable assets, and to define the objectives of production and the criteria for measuring efficiency and success constitutes the most context-remote but also the most system-specific, ‘capitalist’, type of power resource, since it is based upon the (private) ownership and/or the command over the means of production and upon the ‘property rights’ allotted to it. This power resource, of course, fundamentally depends on structural conditions set by the society at large, especially on an appropriate legal system and a state which displays a ‘credible commitment’ to guarantee and defend those ‘property rights’ (Tilly et al. 1991:658ff.). The unequal distribution of this type of power resource inserts the most fundamental but also the most unspecific power differential into the relations among capital owners, top management, and the employees who are in the Weberian sense ‘negatively privileged’ or, to borrow Marx’s term, ‘free’ in the dual sense of being both independent to sell their labour power to capital as a commodity and at the same time forced to work for wages on account of their lack of capital. This fundamental asymmetry represents a social constant in industrial capitalist societies and proves largely immune to variations in the actual employment relations which result from specific production contexts once a decision to allocate capital in a certain line of production has been made. The implementation of this ‘allocative power’ depends on expectations of profitability and yields in product and financial markets, rather than on the internal conditions of particular systems of production. Conversely, precisely this monopolisation of discretionary power concerning investment and disinvestment on the side of management, which emanates from the capitalist private ownership of the means of production, can massively influence intrafirm power asymmetries irrespective of both the decision’s original intentions and the power differentials shaped by the conditions of specific production milieus. Although resembling a relation of an almost binary polarity, the differentials of ‘allocative power’ do not actually take on the form of zero-sum games. On the one hand, any concrete investment decision creates ‘zones of uncertainty’, since it commits scarce resources to allocation in non-redeployable assets and thereby to systems of production whose control poses problems which cannot 103

THOMAS WELSKOPP

be tackled exhaustively by recourse to follow-up large-scale investment decisions of this sort. Investment in particular industries can entail a lasting commitment of otherwise highly flexible capital, since prospective ‘separation costs’ might exceed the gains of a potential shift of resources. Therefore, such commitments lock in ‘allocative power resources’ to a certain degree, and in so far as this occurs, the relative significance of other power resources not completely determined by access to capital increases at the expense of this most onesided and asymmetrical power differential. The almost purely commercial cash nexus relations between capitalist merchants and dependent craft producers in early nineteenth-century putting out systems reflect a set of employment relations almost exclusively governed by the implementation of ‘allocative power resources’. Investment in capital-intensive mass production facilities (characteristic of heavy industry, for instance) represents the opposite pole in a wide spectrum of configurations signifying greatly reduced flexibility of assets and a high degree of commitment and ‘boundedness’ by the conditions and requirements of the particular industry. On the other hand, corporate growth, centralisation and consolidation expand intrafirm ‘organisational capabilities’ and hence increase ‘allocative power resources’, while decreasing the effect of ‘separation costs’for the enterprise at large. This accounts for the superior striking force that strong and diversified corporations, especially when organised on a multinational scale, develop in industrial conflicts by threatening their opponents with plant closures, the holding back of investments, or the spatial shift of entire lines of production. It also accounts for the structural supremacy lockouts display in relation to strikes (Hoerr 1988). Power to implement and control In a given production setting, market conditions, technological choice and the requirements of particular production processes significantly delimit the principally unrestricted power resources of owners and/or top management to implement and control the technical and social organisation of the firm. The allocation of capital in non-redeployable assets ‘contextualises’ the ‘allocative power resources’ and thereby decreases the one-sidedness of capitalist power relations on three grounds: First, even absolute ‘organisational power’, if at all feasible, does not guarantee that decisions produce precisely those social consequences which had been intended and anticipated. Management strategies, based on selective information and formulated to reduce uncertainty, are prone to ‘underdetermine’ their potential effectiveness, appropriateness and the scope of possible consequences. Therefore, while this type of power, especially if highly centralised, massively influences the variables of a given production process, it may do so beyond, and sometimes in outright contradiction to, pre-formulated objectives. 104

CLASS STRUCTURES AND THE FIRM

Second, the power to make technical and organisational decisions to significant degrees has to be delegated to technical, organisational and functional experts, since top management, however elaborately organised, is absorbed both in strategic decision-making and the maintenance of firm— environment relations. Its competence to perform tactical decision-making in close proximity to production and in the high frequency of routine procedures is limited by necessity. Yet this delegation of discretionary powers also invites the usurpation of management prerogatives by the experts and functionaries comprising middle and low-level management. Their infiltration of formal authority structures for group-egoistic purposes creates autonomous power positions based on (monopolised) arcane expert knowledge and functional indispensability. Thus the ‘uncertainty of consequences’ is complemented by an ‘uncertainty of power distribution’. This is exemplified in the discrepancy between the autocratic and strictly hierarchical formal structure of the firm and its actual texture, which rather resembles a ‘complex and labile web of compromises’ lending coherence to an elaborate network of autonomous power positions, shielded arcana, and tenacious centres of resistance that blend formal and informal structures into the multi-dimensional social reality of production. As a rule, the discretionary power of foremen, engineers and general superintendents, if not their actual competence, exceeds the scope of delegated formal authority conferred upon them by top management (Crozier and Friedberg 1979:50). Third, the power to make production-related decisions as distributed by the above-mentioned processes of diffusion across the higher and lower echelons of management does not eliminate the ‘uncertainty of task execution’ which emerges from the confrontation between authoritative decisions and actual social consequences. Relative efficiency is less a question of ‘organisational power’ than of the way in which modes of dealing with this impenetrable ‘zone of uncertainty’ are built into the social structure of the firm. Therefore, autocratic discretionary power of management and their functionaries can coexist with insurmountable power resources at the disposal of the workforce which are based in task autonomy, indispensable skills, knowledge and proficiency, and which effectively dodge, subvert or counteract the intended effects of management decisions. Specifically modern highquality mass production processes, moreover, require such a high frequency of both work-related decisions and decentralised regulation procedures, and entail such a great measure of internal coordination and integration that they by necessity generate a considerable functional workplace autonomy, which deprives any central control regime and linear chain of command of any comprehensive control of task execution. The increase of production-related knowledge and competence in both management and workforce in modern technology-intensive industries has produced the paradoxical result that, while management’s power to determine the parameters of the production process may have considerably profited 105

THOMAS WELSKOPP

from ‘scientific’ sophistication, the discretionary power of the workforce within those parameters has likewise increased substantially. This instead is reflected in the widespread recourse to indirect control regimes based on sanctioning work performance instead of task execution and organised as complex monitoring systems that necessarily involve the collaboration of those to be controlled. Similarly, the ongoing conflict between actual (informal) workforce participation in shopfloor decisions and management’s denial of formal participatory competences has spurred much of the controversy concerning employee representation and co-determination since the 1920s. In fact, no given management system, whether strictly linear and autocratic or employing formal procedures of negotiation and participation, decides about the extent of shopfloor autonomy and self-determination, but only about how the participatory requirements of a particular production process are built into the firm’s internal structure and to what extent they are acknowledged by management or driven into the realm of informality. Production, therefore, resembles an always situative balance of asymmetric, but by no means one-sided, power relations, a fragile compromise between control and autonomy, power and consent, incentives and sanctions, as well as effort and initiative (Minssen 1992:78ff.). The power to command hierarchical authority The formal distribution of the power to command hierarchical authority, i.e. to transform employment relations into a command-and-obey hierarchy, equals a zero-sum game. Formally, the settlement of the labour contract signals full acceptance of management’s claims to absolute and unrestricted authority. Yet the lack of the specificity of written company orders and regulations, as far as the relation between the endorsement of work discipline and the eliciting of work performance is concerned, already points to the fact that formal domination covers only a minor section of the social reality of production. Moreover, the actual internal power structure of the firm as a rule substantially deviates from its formal chain of command as exhibited in its ‘organisational chart’. Domination in the enterprise context is subordinate to efficiency—as are, at least by tendency, discipline and obedience to effort and performance. This distinguishes the firm from ‘total’ institutions like penitentiaries, asylums or the military. Domination in firms is not self-serving, but in its forms, range, and limits it remains ‘bound’ by the prime objective to support the production of marketable goods meeting the requirements of efficiency. Management is thoroughly bound by its dependence on largely incalculable and uncontrollable market processes, as well as by its reliance on an organisation of production which represents a precarious balance and blend of formal and informal structures. Management’s claim to absolute authority reflects the attempt to reduce both ‘zones of uncertainty’ by ‘ordaining away’ internal uncertainty in order 106

CLASS STRUCTURES AND THE FIRM

to increase managerial autonomy and flexibility in firm—environment relations. The achievement of this end, however, paradoxically requires a significant degree of—acknowledged or unacknowledged—self-restriction in the actual allocation of formal authority resources, since the quest to run complex production processes by detailed orders would divert scarce management resources from its prime tasks, and consequently would reduce external flexibility. Therefore, the firm’s formal authority structure, although governed by management’s claim to total command, represents something like a ‘subsidiary order’ which is reconciled with the complex social reality of workplace and employment relations in that it is largely not necessary to be drawn upon in routine production. This order reduces managerial uncertainty by endorsing the plausible ‘fiction’ of total control and by representing the latent ‘potentiality’ of authoritative intervention which management falls back on in case of acute crisis. As analyses of such crises reveal, however, it is often precisely the autocratic recourse to management’s ‘final say’ which actually aggravates conflict instead of settling it (Golden and Ruttenberg 1942:236ff.). Furthermore, it is a widely accepted fact that formal authority can enforce obedience but falls short of vouching for effort and initiative, which in most production settings are pivotal to generating efficiency. Hence, relatively efficient performance is instead the result of an informal powerstructured consensus-finding process, an implicit social arrangement among all social groups involved, which rests on compliance and a partial compromise between otherwise antagonistic interests rather than on obedience (Minssen 1992:78ff.). In many cases, production settings prove efficient despite their autocratic formal make-up, precisely because it is in their informal undersides where the necessary requirements of smooth cooperation and active coordination are met, and because the physical absence of management in routine production minimises intergroup contact and thereby the probability of clashes. Intrafirm employment relations, in sum, are shaped by complex if asymmetric balances of power and consent, of which formal authority relations form only a minor part, lending stability to the firm’s basic power differentials by virtue of their ‘latency’ and ‘virtuality’ rather than by their actual interference in routine production. These balances of power represent informal and latently embattled arrangements between management’s claims to full command and control and the spheres of autonomy, disposition and communication which are entrenched in the functional requirements of the production processes. First, managerial domination is filtered and effectively blocked by informal work groups whose functional autonomy and potential ‘blocking power’ resources render the enforcement of ‘iron’ discipline irreconcilable with smooth operation and efficiency. ‘Working to rule’, collective ‘soldiering’ or wildcat strikes by shopfloor crews that leave complex, sensitive production processes hanging in the balance and threaten to cause substantial production losses, 107

THOMAS WELSKOPP

are formidable weapons at the disposal of such primary workgroups that can be mobilised effectively to block managerial encroachments. This direct group power is ineradicable precisely because it emanates from the workers’ appropriation of the ambivalent and contradictory traits of the production process itself. The command over specific scarce skills or certain forms of workplace cooperation and communication, moreover, can instigate the transformation of direct group power and shopfloor solidarity into organisational power that enhances intrafirm bargaining positions and tends to shift conflicts into the organisational—political sphere. Second, the dependence of firms on expert knowledge, organisational competence and proficiency can substantially encumber the imposition of rigid authoritative measures to the extent that even central managerial prerogatives become challenged and subverted. This is, for instance, the hallmark of employment relations in industries dominated by highly skilled craft workers whose ‘responsible autonomy’ provides them with a formidable measure of veto power in the actual work process (Tolliday and Zeitlin 1991:7, 11). In addition, the formation of craft unions based on collective craft pride and group ethos reinforces and broadens this sphere of autonomy by creating organisational resources which reverberate on shopfloor relations, e.g. in the form of establishing ‘closed shop’ rules or standards for the restriction of output (Welskopp 1994b). Third, the same applies on an even more comprehensive basis to the intermediary ranks of company organisation: to the engineers and general superintendents as well as the middle echelons of clerical and R&D experts. Instead of acting as mere transmitters of top management’s orders, their expert knowledge and functional indispensability (which, though collective resources, are drawn upon by the individual functionaries in routine work as an aggregate of agents and not as a group) invite the usurpation of autonomous power positions which obscure the fixed formal chain of command by infiltrating authority relations to subordinate positions, while effectively blocking top management’s claims to full authority as far as their own stature is concerned. The uncertainty of top management in fixing standards of performance and behaviour for these experts results from the fact that their scientific training provides indeed the necessary prerequisite, but not a sufficient condition for an appropriate performance in production, where competence has to be brought to bear in securing output. On the one hand, this complex configuration is responsible for the ‘incomplete professionalisation’ of these specialists whose prominent ‘bargaining position’ may be shielded by cooptation and the segmentation of a labour market for academically trained experts, but still depends on uncontrollable cycles of this labour market due to changing business conditions and on the actual access to the decisive positions in the firm where expertise can be proven and utilised to defend privileges against encroachments by management. On the other, their 108

CLASS STRUCTURES AND THE FIRM

dependence on individual performance increases top management’s uncertainty vis-à-vis experienced and well-established experts, as can be concluded from the extremely wide and fragmented spectrum of material compensation granted to different members of the same profession within a given firm. By the same token, struggles for position, competition for competences, subversion of information, and combative ‘micropolitics’, accompany the obligation to ‘sell oneself favourably to top management which ensues from this precarious power position and demarcates the workplace relations of low-and mid-level management decisively from the intrinsically less competitive shopfloor relations among manual workers. The considerable width of the spectrum of hierarchical positions within these ranks, therefore, reflects an amalgam of individual and collective (professional) social distancing and power maintenance strategies on the side of these experts as well as top management’s strategies to curb these uncontrollable ‘micropolitics’ by inserting ever more fine-grained formalisations into the organisational fabric of the firm. The power to define and determine The power to define and determine working conditions (for example hours of work) and the forms and amount of remuneration does not immediately reflect the constellation of power differentials as portrayed in the previous sections. Only in the most extreme cases—if employees’ bargaining positions are unfavourable, collective bargaining does not exist or proves impotent, and the state abstains from providing a framework of minimum standards and obligatory procedural schemes—is management able unilaterally to impose conditions of work on a workforce that is quiescent (Gaventa 1980). Since anticipated productivity standards and industry-specific developments in product and financial markets decisively inform management strategies concerning conditions and wages, external ‘contexts’ interfere with decisionmaking even in this most one-sided variation of fixing the terms of employment. Business cycles have an immediate impact on ‘distributive power relations’, even in the largest corporations, as wages are usually paid from sales and the payroll is therefore more important than the share of wages in total costs as a determinant of the material conditions of employment. Furthermore, the structure of segmented labour markets—whose demand side may be dominated by industry and industry-specific requirements but whose supply side remains largely out of reach—composes another ‘context’ which deprives management of one-sided control. In the extreme case just mentioned, firms belonging to a particular industry may succeed in establishing particularly low standards of pay—especially if employing high percentages of unskilled labour and if taking advantage of gender or ethnic lines of division within the workforce. Yet they may do so only at the expense of having to invest 109

THOMAS WELSKOPP

disproportionate resources into the creation of a particularly elaborate control force whose members then tend to turn their functional indispensability into a lever to extract extraordinarily high compensation (Welskopp 1984b). Hence ‘distributive power’ in the firm—or in a number of firms comprising an industry or an industrial region—appears as the consequence of the translation of intrafirm power configurations into the sphere where they overlap with and are mediated by interfering market contexts in turn influenced by but not fully determined by those constellations. This relationship is further complicated by the degree to which collective bargaining and bipartite industrial relations are institutionalised, and to the degree to which interest organisations—both employers’ associations and trade unions or professional organisations—are capable of mobilising organisational power in negotiations over terms of employment. Since in ‘civic society’ large-scale industrial conflicts have a direct and far-reaching impact on the political sphere, the state tends to enter the stage of collective bargaining as an additional major force affecting ‘distributive power relations’. Depending on the type of society, the state intervenes to differing degrees by providing a specialised legal framework consisting of protective and regulatory rights; by establishing procedural rules for negotiation, sometimes arbitration, and generally the institutionalisation of industrial disputes; by defining and delimiting the scope of legitimate and legal industrial action; by setting and enforcing minimum standards of conditions of employment; and by decisively influencing terms of employment as a potent economic player in public sector labour relations, which reverberate on other industries (Keller 1991:78f.). The creation of social security systems which transform classbased social problems—both internal and external to industry—into insurance risks (that formally make no class distinctions) exercise a complementary, although mediated, influence on the balance of ‘distributive power’. The fact that the state is not a monolithic bloc but consists of sometimes competing and conflicting agencies and systems, each of which tends to develop a ‘micropolitical’ momentum of its own, underscores the fact that the analysis of ‘distributive power relations’ has to supplement the contextualisation of the approach to intrafirm industrial—relations with the contextualisation of the perspective on firm—environment relations.

Concluding remarks: the ‘firm’ as an incompletely organised ‘political coalition’; and consequences for historical research The model of the ‘firm’ in industrial capitalist societies developed in this chapter provides a framework well-suited to grasp the complex interplay of power-structured and consensual workplace and employment relations constituting the social context which shapes production and the conduct

110

CLASS STRUCTURES AND THE FIRM

of the enterprise as a precarious and fragile result of both ongoing formal or informal bargaining processes and open or hidden ‘power struggles’ among all social groups involved. As one of the most conspicuous institutions of industrial society, bearing, together with the ‘market’, the peculiar systemspecific traits of ‘industrial capitalism’ and its ‘class structure’, the ‘firm’ has to be viewed as a particularly ‘soft’ and ‘dynamic’ organisation, since organisational elements only constitute a part of its social reality and are embedded, surrounded and impinged upon by non-organisational social relations. The characteristic ‘openness’ of the ‘firm’ for impulses from its market environment and for the requirements of specific production processes is responsible for this mutual penetration of ‘Gemeinschafts-’ and ‘Gesellschaftsbeziehungen’, to borrow Max Weber’s terms, since all social agents engaged in intrafirm power relations draw on resources which, at least in part, are not provided by the organisation itself and, conversely, at least in part deprive any attempt at full control by organisational means. Whereas internal bargaining processes, ‘power games’ and ‘micropolitics’ characterise all organisations in general, thereby constituting social realities that substantially deviate from their formal order, it is precisely this dependence on system-specific ‘interfaces’ to markets and process technology which marks off the social relations in the ‘firm’ from those in institutions which almost exclusively draw on rules and resources provided by the organisation proper. In other words, the incorporation of the ‘class relation’ into the fabric of the ‘firm’, which distinguishes this institution from administrative bodies, associations, agencies and political parties, decisively checks and delimits the scope and impact of organisation on internal social relations (Weber 1991:16, 18). In this context, it is possible to make a reassessment of the role and development of ‘skill’ in industrial enterprises that allows for a productive reconstruction of ‘labour process’ theory, which has now virtually collapsed under the barrages of revisionist empirical scrutiny. ‘Skill’ has to be defined as the actual amount of knowledge, competence and proficiency necessary to perform a specific task, and it is subject to a detailed reconstruction of work processes in order to decide if a concrete work position requires ‘high’ or ‘low’ skill. As Paul Robertson and Lee Alston have shown convincingly, technological change in a majority of cases is motivated by general cost and productivity considerations rather than by purposeful skill dilution and task fragmentation. Moreover, technological choice—regardless of its underlying motivation—tends to produce, not fully predictable but, rather, diverse, ambivalent and sometimes directly upgrading effects on skill development, so that ‘[w]hether a technology enhances skills or deskills is frequently more a result of technological choice than a motivation for technological choice’ (Robertson and Alston 1992:347). Analyses of major large-scale industries indeed suggest that full mechanisation and the cooperative and coordinative requirements of flexible, 111

THOMAS WELSKOPP

decentralised high-quality production have spurred a discernible upgrading trend, both in skill structure and skill profile development, since technological change has replaced unskilled rather than skilled labour. Moreover, whereas manual proficiency may have vanished from many industrial work settings, production knowledge and process control competence have increased and been supplemented by coordinative and discretionary skills (Welskopp 1994b; Nuwer 1988). Therefore it might be worthwhile to analyse in depth the forms and consequences of the shift from craft skills to industrial skills rather than to stick to the indiscriminate presupposition of a comprehensive, linear trend of ‘dequalification’ as underlying ‘labour process’ theory. To derive statements on skill development from detailed reconstructions of production settings, furthermore, represents an appropriate way of coping with the fact that formal job classification and formalised training procedures frequently hint at power balances and differentials in the power to define skills rather than to the actual scope and quality of competence which may as well remain largely ‘hidden’ and ‘tacit’. It is of prime importance to regard skill as only one string among many in the intricate network of cooperative and power structures, of control, consent, and functional autonomy, which combine to constitute the diverse configurations workplace and employment relations form (Robertson and Alston 1992:346f.). The analytical distinction between skill, power and modes of cooperation, which is implied in such an approach, allows a firm grasp on the spectrum of potential social consequences produced by those configurations in their entirety. Skill, if paired with a high degree of functional autonomy, can prove a formidable resource of direct group power, especially among workers occupying strategic positions in vulnerable high throughput processes. Skill, embedded in tight cooperative networks, can also become the basis of solidaristic workgroup conduct and thereby spur direct collective action at the workplace or the consolidation of organisational power into broader interest movements reverberating on working conditions from the organizational—political sphere of society. By contrast, even high degrees of skill can fail to provide a basis of power or solidaristic behaviour if close personal control deprives skilled workers from autonomy and isolates them in ‘prisoner’s dilemma’ situations. In these cases, degradation does not have to entail or require ‘deskilling’. Rather, degradation here results from the lack of access of skilled workers to vital communicative resources originating in an almost complete shift of autonomy from the workplace into the sphere between top management and supervisory force (Montgomery 1987). In a general sense, the sets of workplace and employment relations typically ‘reveal complex political processes involving both conflict and cooperation between individuals, departments and interest groups rather than a clearly defined hierarchical structure of command corresponding to the formal organisational chart’ (Tolliday and Zeitlin 1991:12). Fragile ‘political coalitions’, shaped in open and tacit bargaining processes by power-structured consensus112

CLASS STRUCTURES AND THE FIRM

finding and continually embattled in open and covert ‘power struggles’, constitute the social reality of production. Efficiency, always relative, appears as the combined effect of balances between power and compliance, control and autonomy, incentive and effort. The working of an enterprise can rest on explicit or—more frequently—implicit ‘social arrangements’ between the social groups involved whose otherwise antagonistic interests converge in a tacit ‘productivity consensus’ as a ‘quasi-resolution of conflict’ as long as both sides profit from smooth and efficient production. Even if this is not the case, relative efficiency can emanate from the fact that production goals are achieved by patterns of behaviour whose motivations may differ among groups but prove complementary rather than mutually exclusive in their effects. Michael Burawoy’s ‘games’ fit into this pattern (Burawoy 1979), as does the compliance of skilled workers with procedures which, in turn, allow workgroups to maintain and reassert their autonomy and internal channels of communication. Compliance, far from reflecting opportunism and harmony, thus results from the fact that either production goals or the procedures and collective routines to achieve them prove acceptable to both sides, even if viewed from antagonistic angles. At any time, however, this ‘armed truce’ can erupt in open or hidden conflict, especially when the ‘subsidiary order’ of mechanistic managerial command intervenes in the day-to-day routine normally governed by habitualised custom and pragmatic conventions. Therefore, periods of tacit compliance and negotiated ‘quasiresolution of conflict’ alternate with those of explosive power struggles in different degrees of intensity and scope. It depends upon the way in which arenas and participatory procedures to compartmentalise and settle those conflicts by power-structured negotiation are built into the employment relations of a given ‘firm’ to determine if tensions can be isolated and tempered in controlled conflict settlement or if they culminate in wholesale mutual blockade and open war. An episode that occurred in an American steel mill in the 1930s may illustrate such a confrontation between management’s claim to total control and the workers’ factual participatory competence based on shared skills, collective job pride and functional autonomy in routine production. This partialised dispute arose in an acute crisis that prompted management to instantiate its otherwise ‘virtual’ claim to command. In the operation of an open-hearth furnace, a heat will sometimes develop a ‘boil’. Unless it can be worked out or guided to the rear of the furnace, it breaks through the front and the molten metal pours out on the floor. This is costly and dangerous. Every effort is made to prevent it. At the time an epidemic of ‘boils’ had been plaguing the department and the superintendent, a graduate of an outstanding university in metallurgy, had applied all his knowledge to solve the problem, but it was beyond him. For weeks Joe, a charging-machine operator, tried to tell him how to kill a ‘boil’. The superintendent would not listen. During a shift when the 113

THOMAS WELSKOPP

superintendent was out of the mill, a ‘boil’ developed and threatened to break out through the front of the furnace. Joe decided to put his idea to practical use. He loaded a four-thousand-pound box of ore on his charging machine and drew it up before the door of the furnace. Just then the superintendent arrived on the scene. He ordered Joe away, told him that he was a fool and would get ‘burnt to a crisp’. The first melter in charge of the furnace favoured trying Joe’s idea. The superintendent said ‘nothing doing’. Finally, as the ‘boil’ was almost at a head, the melter ordered the super off the floor. The super could fire him later, he said; but he was in charge at the moment and, come what may, Joe was going to get a chance to see what he could do, because ‘We’ve been losin’ too many heats and my pay envelope has been showin’ it’. The experiment proved successful. The ‘boil’ was broken; the heat continued and was tapped without a loss of a minute’s working time or an ounce of steel. For a month the super would not talk to Joe, who had so humiliated him. (Golden and Ruttenberg 1942:236f.) Yet the bulk of intrafirm ‘power struggles’ may be ‘played out’ in conflicts of lower intensity or may even be decided by the unilateral establishment of conventions and customary work practices on the side of the workforce which largely escape the grasp and the perception of management. These tacit ‘power games’ are equally based on the workforce’s contrivance to infiltrate and subvert production procedures for their own purposes without noticeably compromising productivity, and on management’s acknowledged or unacknowledged tendency ‘to look the other way’ as long as efficiency remains ‘satisficing’, which results from its urge to focus information flows and maintain maximal autonomy in firm-environment relations. Implicit ‘social arrangements’ of this kind largely rest on the functional autonomy of work groups and prove capable of eliciting effort as long as they do not clash with management’s forceful recourse to the ‘subsidiary order’ of command-andobey relations. Another episode may illustrate this. In the early 1950s, the plant management of a large steel rolling mill in the West German Ruhr district noticed that, for a considerable span of time, the night shift had turned out a significantly higher tonnage than the two day shifts. Neither plant management nor the superintendents were at first able to explain this phenomenon. Since each of the three rolling mill crews alternated in working the night shift and each achieved the higher output when doing so, skill differentials as between crews were ruled out as a possible cause. Sustained investigation by superintendents finally uncovered that the rollers had developed a procedural ‘trick’ by means of which the overheating of the rolls could be prevented and, consequently, stoppages in production could be reduced to a minimum. The workers had submitted this practice as a formal suggestion for improvement to management some time before. Yet management had not only turned down this suggestion but had subsequently overtly prohibited the application of this procedure. 114

CLASS STRUCTURES AND THE FIRM

Since the crews were not willing to forfeit the higher premium wages resulting from improved performance, they utilised their collective functional autonomy and solidaristic group structure to negotiate a clandestine arrangement collectively to disregard management’s orders and apply the trick during the relatively ‘control-free’ night hours, when plant management, engineers and superintendents were usually absent from the plant. Characteristically, the turn foremen, who closely interacted with the crews on a day-to-day basis, were included in the arrangement, so that the outer boundaries of functional autonomy demarcated the ‘in-group’ of the workforce from the ‘out-group’ of the mill bosses. Subsequently, management not only came to tolerate this night shift convention; it tacitly signalled encouragement to extend the practice to the day shifts as well. While the workers’ convention thus prevailed, management succeeded in upholding formal authority by refraining from authoritative intervention. (Bahrdt 1962:31) The theoretical framework of the inner workings of large capitalist enterprises developed here aims on the one hand at reconstituting a fertile dialogue between the insightful—if disparate—contemporary empirical research on business and labour history, with grand theory of society which has to incorporate microtheory and critical social theory in order to regain its waning synthesising capabilities. On the other hand, it aims at integrating industrial work, skill development, group cooperation, the formal and informal organisation of production processes, and the internal industrial relations between workers, supervisors, engineers and management into a comprehensive model which pays tribute to the widespread observation that all of these aspects contribute to the generation of workplace and employment relations precisely through the configurations they compose. Since the ‘firm’s’ internal make-up matters in all aspects conventionally covered by labour history, and since business history can no longer justifiably ignore that management is too ‘bounded’ by both market and production contexts as to foreshorten managerial conduct to ‘organisation building’, the model suggested here does provide an avenue of research which directs attention to the interrelations of spheres. By ‘economising’ labour history and ‘sociologising’ business history, it acknowledges that both disciplines actually deal with the same sort of problems; many of their respective research interests converge in the capitalist ‘firm’ and its inner workings, even if approaching the enterprise from diverging perspectives. It has to be stressed, however, that the model does not claim to provide a ‘cover-all’ explanatory framework that can readily be applied to historical data in its abstractness. Rather, it is designed as a systematic ‘guide’ to genuinely historical theoryconstruction which calls for typological and comparative research, enhances the mutual ‘compatibility’ of partial approaches, and reintegrates midrange concepts by grounding them firmly in social theory and a ‘contextualised’ theory of modern society. 115

THOMAS WELSKOPP

The microtheoretically enriched view on social interaction on the shopfloor, in the office, and in the realms of management, forces us to abandon constructions of secular, linear ‘ideal’ models of development that underlie most strands of conventional theory and reduce empirical richness of form to deviations from abstract norms. To acknowledge microdiversity without relinquishing the notion of structural identity instead allows us to classify and explain cases as variations on shared structural principles, and this means to emphasise typology and comparison in order to assemble midrange concepts of types of industries, of stages in their long-term development, or of sets of employment relations, which then fulfil their explanatory functions on the basis of the insight that concrete phenomena are not only at the same time products of structural regularities and situative idiosyncrasies, but that structural principles manifest themselves nowhere else and in no other way except in these configurations located in concrete time—space. To accept this does not mean to dissolve history into a welter of incoherent events and episodes; the conceptual axis linking the microlevel of social interaction to the mesolevel of institutions and the macrolevel of society calls for generalisation. Yet it forces us to define and reflect precisely the level of abstraction on which generalisations can be made. This prospective guide to historically conceptualising the ‘firm’ in industrial capitalist societies not only integrates business and labour history into a broadly conceived ‘industrial history’; it aims at inserting such an ‘industrial history’ into the comprehensive framework of a comparative history of modern societies.

References Bahrdt, H.P. (1962) ‘Die Industriearbeiter’, in M.Feuersenger (ed.) Gibt es noch ein Proletariat?, Frankfurt: Suhrkamp: 25–33. Birke, M. (1992) Betriebliche Technikgestaltung und Interessenvertretung als Mikropolitik. Fallstudien zum arbeitspolitischen Umbruch, Wiesbaden: Deutscher Universitätsverlag. Blumin, S. (1985) ‘The Hypothesis of Middle-Class Formation in Nineteenth-Century America: A Critique and Some Proposals’, American Historical Review 90: 299– 338. Braverman, H. (1974) Labor and Monopoly Capitalism: The Degradation of Work in the Twentieth Century, New York and London: Monthly Review Press. Breuilly, J. (1989) ‘The Making of the European Working Class’, in H.Konrad (ed.) Probleme der Herausbildung und politischen Formierung der Arbeiterklasse, Wien and Zürich: Europaverlag: 1–19. Burawoy, M. (1979) Manufacturing Consent. Changes in the Labor Process under Monopoly Capitalism, Chicago and London: University of Chicago Press. Burns, T. (1961) ‘Micropolitics: Mechanisms of Institutional Change’, Administrative Science Quarterly 6:257–81. Canning, K. (1992) ‘Gender and the Politics of Class Formation: Rethinking German Labor History’, American Historical Review 97:736–68. Chandler, A.D., Jr. (1990) Scale and Scope. The Dynamics of Industrial Capitalism, Cambridge, Ma. and London: The Belknap Press of Harvard University Press. 116

CLASS STRUCTURES AND THE FIRM

Crozier, M. and Friedberg, E. (1979) Macht und Organisation. Die Zwänge kollektiven Handelns, Königsstein: Anton Hain Meisenheim. Edwards, R.C. (1979) Contested Terrain: The Transformation of the Workplace in the Twentieth Century, New York: Basic Books. Elbaum, B.L. and Wilkinson, F. (1979) ‘Industrial Relations and Uneven Development: A Comparative Study of the American and British Steel Industries’, Cambridge Journal of Economics 3:275–303. Gaventa, J. (1980) Power and Powerlessness: Quiescence and Rebellion in an Appalachian Valley, Urbana, I11.: University of Illinois Press. Giddens, A. (1972) The Class Structure of Advanced Societies, London and New York: Harper and Row. ——(1976) New Rules of Sociological Method: A Positive Critique of Interpretative Sociologies, London and New York: Hutchinson. ——(1981) A Contemporary Critique of Historical Materialism, 1: Power, Property and the State, London and Berkeley: Polity Press. ——(1984) The Constitution of Society: Outline of the Theory of Structuration, Cambridge and Berkeley: Polity Press. Golden, C.S. and Ruttenberg, H.J. (1942) The Dynamics of Industrial Democracy, New York: Harper. Gordon, D.M., Edwards, R.C. and Reich, M. (1982) Segmented Work, Divided Workers: The Historical Transformation of Labour in the United States, Cambridge: Cambridge University Press. Hareven, T. (1982) Family Time and Industrial Time. The Relationship Between the Family and Work in a New England Industrial Community, Cambridge: Cambridge University Press. Hoerr, J.P. (1988) And the Wolf Finally Came. The Decline of the American Steel Industry, Pittsburgh: University of Pittsburgh Press. Katznelson, I. (1986) ‘Working-Class Formation: Constructing Cases and Comparisons’, in I.Katznelson and A.R.Zolberg (eds) Working-Class Formation: Nineteenth Century Patterns in Western Europe and the United States, Princeton: Princeton University Press: 3–41. Keller, B.K. (1991) ‘The Role of the State as Corporate Actor in Industrial Relations Systems’, in R.J.Adams (ed.) Comparative Industrial Relations. Contemporary Research and Theory, London: Harper Collins Academic: 76–93. Kießling, B. (1988) ‘Die “Theorie der Strukturierung”. Ein Interview mit Anthony Giddens’, Zeitschrift für Soziologie 17:286–95. Kleinberg, S. (1989) The Shadow of the Mills. Working-Class Families in Pittsburgh, 1870–1902, Pittsburgh: University of Pittsburgh Press. Kleinschmidt, C. and Welskopp, T. (1993) ‘Zu viel “Scale”—zu wenig “Scope”. Eine Auseinandersetzung mit Alfred D.Chandlers Analyse der deutschen Eisen-und Stahlindustrie in der Zwischenkriegszeit’, Jahrbuch für Wirtschaftsgeschichte (2/ 1993):251–97. Kocka, J. (1983) Lohnarbeit und Klassenbildung. Arbeiter und Arbeiterbewegung in Deutschland 1800–1875, Berlin and Bonn: J.H.W.Dietz Nachf. ——(1986) ‘Problems of Working-Class Formation in Germany: The Early Years, 1800– 1875’, in I.Katznelson and A.R.Zolberg (eds) Working-Class Formation: Nineteenth Century Patterns in Western Europe and the United States, Princeton: Princeton University Press: 279–351. Kotthoff, H. and Reindl, J. (1990) Die soziale Welt kleiner Betriebe. Wirtschaften, Arbeiten und Leben im mittelständischen Industriebetrieb, Göttingen: Otto Schwarz and Co.

117

THOMAS WELSKOPP

Küpper, W. and Ortmann, G. (eds) (1992) Mikropolitik. Rationalität, Macht und Spiele in Organisationen, Opladen: Westdeutscher Verlag, 21992. Lichte, R. (1978) Betriebsalltag von Industriearbeitern. Konflikthandeln einer Belegschaftsgruppe vor und während einer Betriebsstillegung, Frankfurt/M. and New York: Campus. Marglin, S.A. (1974) ‘What Do Bosses Do? The Origins and Functions of Hierarchy in Capitalist Production’, Review of Radical Political Economy 6 (Summer): 33–60. ——(1984) ‘Knowledge and Power’, in F.H.Stephen (ed.) Firms, Organization and Labour: Approaches to the Economics of Work Organization, London: Macmillan: 146–64. ——(1991) ‘Understanding Capitalism: Control versus Efficiency’, in B. Gustafsson (ed.) Power and Economic Institutions: Reinterpretations in Economic History, Aldershot: Edward Elgar: 225–52. Marx, K. (1979) Das Kapital, I Marx-Engels-Werke, 23, Berlin: Dietz. Minssen, H. (1992) Die Rationalität von Rationalisierung. Betrieblicher Wandel und die Industriesoziologie, Stuttgart: Ferdinand Enke Verlag. Montgomery, D. (1981) Workers’ Control in America: Studies in the History of Work, Technology, and Labor Struggles, Cambridge: Cambridge University Press, 21981. ——(1987) The Fall of the House of Labor. The Workplace, the State, and American Labor Activism, 1865–1925, Cambridge: Cambridge University Press. Mosner, S. (1992) ‘Effort Strategies, Labour Markets and Labour Relations: Managing British and German Skilled Metal Workers 1900–1935’, MPhil thesis, University of Manchester. Nuwer, M. (1988) ‘From Batch to Flow: Production Technology and Work-Force Skills in the Steel Industry, 1880–1920’, Technology and Culture 29:308–38. Plowman, D.H. (1991) ‘Management and industrial relations’, in R.J.Adams (ed) Comparative Industrial Relations. Contemporary Research and Theory, London: Harper Collins Academic: 56–75. Price, R. (1983), ‘The Labour Process and Labour History’, Social History 8:57–75. Pries, L. (1991) Betrieblicher Wandel in der Risikogesellschaft. Empirische Befunde und konzeptionelle Überlegungen, Opladen: Westdeutscher Verlag. Robertson, P.L. and Alston, L.J. (1992) ‘Technological Choice and the Organization of Work in Capitalist Firms’, Economic History Review 45:330–49. Sayer, R.A. (1985) ‘Industry and Space: A Sympathetic Critique of Radical Research’, Society and Space 3:3–29. Scranton, P. (1989) The Workplace, Technology, and Theory in American Labor History’, International Labor and Working-Class History, 35:3–22. Somers, M.R. (1989) ‘Workers of the World, Compare!’, Contemporary Sociology 18: 325–9. Thompson, E.P. (1963) The Making of the English Working Class, Harmondsworth: Penguin. Tilly, C., Tilly, L.A. and Tilly, R. (1991) ‘European Economic and Social History in the 1990s’, Journal of European Economic History 20:645–71. Tolliday, S. and Zeitlin, J. (1991) ‘Introduction: Employers and Industrial Relations Between Theory and History’, in S.Tolliday and J.Zeitlin (eds) The Power to Manage? Employers and Industrial Relations in Comparative-Historical Perspective, London and New York: Routledge: 1–31. Weber, M. (1980) Wirtschaft und Gesellschaft, Tübingen: J.C.B.Mohr, 51980. ——(1991) ‘Vorbemerkung’, in M.Weber Die protestantische Ethik, 1, ed. J.Winckelmann, Gütersloh: Gerd Mohn, 81991. Welskopp, T. (1992) ‘Arbeitsplatz, Staat und Arbeiteraktivismus. David Montgomerys

118

CLASS STRUCTURES AND THE FIRM

“Fall of the House of Labor” als neue Synthese amerikanischer Arbeitergeschichte’, Geschichte und Gesellschaft 18:94–106. ——(1993) ‘Von der verhinderten Heldengeschichte des Proletariats zur vergleichenden Sozialgeschichte der Arbeiterschaft: Perspektiven der Arbeitergeschichtsschreibung in den 1990er Jahren, 1999. Zeitschrift für Sozialgeschichte des 20. und 21. Jahrhunderts (3/1993):34–53. ——(1994a) ‘Ein Klassenkonzept für die vergleichende Geschichte industrialisierender und industrieller Gesellschaften. Kritische Skizzen und theoretische Überlegungen’, in K.Lauschke and T.Welskopp (eds) Mikropolitik im Unter nehmen. Arbeitsbeziehungen und Machtstrukturen in industriellen Großbetrieben des 20. Jahrhunderts, Essen: Klartext: 48–106. ——(1994b) Arbeit und Macht im Hüttenwerk. Arbeits-und industrielle Beziehungen in der deutschen und amerikanischen Eisen-und Stahlindustrie von den 1860er bis zu den 1930er Jahren, Bonn: J.H.W.Dietz Nachf. ——(1996) ‘Der Betrieb als soziales Handlungsfeld. Neuere Forschungsansätze in der Industrie-und Arbeitergeschichte’, Geschichte und Gesellschaft 22:118–42. Weltz, F. (1988) ‘Die doppelte Wirklichkeit der Unternehmen und ihre Konsequenz für die Industriesoziologie’, Soziale Welt (1/1988):97–103. Williamson, O.E. (1975) Markets and Hierarchies, New York: Free Press. ——(1985) The Economic Institutions of Capitalism, New York: Free Press. Zeitlin, J. (1983) ‘Social Theory and the History of Work’, Social History 8:365–74. Zwahr, H. (1978) Zur Konstituierung des Proletariats als Klasse: Strukturuntersuchung des Leipziger Proletariats während der industriellen Revolution, Berlin: Akademieverlag.

119

4

KNOWLEDGE, INFORMATION AND ORGANISATIONAL STRUCTURES P.P.Saviotti

Introduction Knowledge intensity in industrialised economies has been growing since the Second World War. R&D expenditures have been expanding at a very high pace and the R&D function has become internalised in most large firms (Freeman 1982). This phenomenon has caught the attention of economists and students of technological change (Freeman 1982; Coombs et al. 1987; Stoneman 1983), but the problem has often been analysed by comparing the different R&D intensities of, for example, different firms or countries to establish whether higher R&D intensity has led to greater innovativeness. Recently, however, increased attention has been paid to the processes of knowledge creation and utilisation (Dasgupta and David 1992; Foray 1991; David and Foray 1994). The main aim of this chapter is to assess the implications for organisational structures of recent developments in the economics of knowledge. These seem to be particularly important because new types of organisational structures, such as inter-institutional collaborative agreements (IICAs) (Chesnais 1988; Mytelka 1991; Hagedoorn and Shackenraad 1990, 1992), have become very common, and because we can expect that the trend towards increasing knowledge intensity will continue in the foreseeable future. In making its assessment, this paper begins by reviewing some recent developments in the field, paying particular attention to the distinction between information and knowledge. Later, knowledge itself is interpreted as a component of the external selection environment of firms and other organisations. I argue that problems of firm structure and boundaries cannot be discussed without paying due attention to the nature of the external environment. Finally, all types of organisational structures are analysed in terms of the two elementary 120

KNOWLEDGE, INFORMATION AND STRUCTURE

phenomena of the division of labour and coordination. The production of any type of composite output involves a series of coupled productive stages; that is, a particular combination of the division of labour and coordination. Thus existing and possible organisational structures can be interpreted as different combinations of the division of labour and coordination. Each such organisational form adapts better to, and thus is more stable and efficient in, a given type of external environment. Changes in knowledge as a component of the external environment are used in this chapter to analyse the relative stability of different organisational structures. The conclusion is that, while hierarchies are suited to stable environments, the additional flexibility offered by IICAs may be more appropriate in environments in which the rate of radical change is relatively high.

Knowledge: some basic considerations An approach which concentrates on R&D as an input may be sufficient when the distribution of R&D expenditures is very skewed; that is, when some countries or firms spend a lot more on R&D than others. However, the quality and usefulness of the knowledge thus produced cannot depend only on the money spent on it. Research and knowledge strategies become of fundamental importance when not only the general level of knowledge intensity has increased, but when many participants have achieved similar levels of financial commitment. In such cases, it becomes an important challenge to understand mechanisms of knowledge generation and use. This chapter both summarises a number recent developments and adds some personal contributions to the economics of knowledge. A more detailed analysis of the same problems is contained in Saviotti (1996). The considerations which follow are not intended as a generalised theory of knowledge, but only as a means to help us answer such questions as: how do firms use scientific and technological knowledge?; and how do they transform this knowledge into market success and competitive advantage? Two functions of knowledge which are particularly important for this purpose are emphasised here. First, scientific theories, and knowledge in general, correlate events and variables. For example, physical and chemical laws expressed as equations give us the relationship between two or more variables. From this point of view, theories can be considered to be correlational structures. Moreover, knowledge of certain theories allows individuals or institutions to interpret knowledge and information created externally by other individuals or institutions. For example, information created by the development of new disciplines and involving concepts qualitatively different from those of pre-existing disciplines can only be retrieved and interpreted by the agents possessing the new knowledge. Thus knowledge and theories can also be considered to be retrieval and interpretative structures.

121

P.P.SAVIOTTI

It must be noted that these two functions of knowledge are useful both for science and for technology. Understanding of the relationship/correlations of different variables is a fundamental part of what we mean by scientific theories. Furthermore, the transformation of inputs into outputs which takes place in production is facilitated by knowledge of the relationships of the properties/variables of the materials, of the prevailing energy forms, etc., which constitute the inputs. That these two functions of knowledge are useful in both cases does not imply that scientific and technological knowledge are the same. Traditionally they were considered to be so distinct that technology was denied a cognitive function (Layton 1974). ‘Science discovers, technology applies’ represents accurately this perception of the different knowledge roles of science and technology. The cognitive role of technology has recently been reassessed in order to take into account the non-linearities and feed-backs existing between science and technology in the innovation process (David and Foray 1994). In order to treat the flows of knowledge taking place between scientific institutions and firms, we have to use some variables and parameters which are common to the different types of knowledge that they produce and use. Such considerations are complementary with respect to those of Dasgupta and David (1992), who consider that science and technology do not differ in the knowledge that they use, but only in their reward mechanisms. The point of view taken here is that science and technology have some common underlying features which allows the existence of flows between them, but that we can define variables, such as their span, which allow us to establish differences of degree between them. A number of consequences follow if we accept that theories are correlational structures. First, any system of interest to us for scientific or technological reasons (physical, chemical, biological, etc.) may be described by a number of variables. Depending on their sophistication and maturity, theories will be able to establish correlations between some or all of the variables of the systems investigated. In other words, theories at different stages of development will be characterised by different correlation densities. The higher the correlation density, the greater will be the explanatory and predictive power of theories. Second, each theory or piece of knowledge will describe and correlate a number of observables. The collection of such observables will constitute the observation space of the theory. In turn such observables will be described by variables. We can imagine placing all the required variables corresponding to the observation space of a theory along a range, which represents the portion of the environment investigated by the theory. The larger the range over which the theory correlates variables, the greater its power and generality. Each theory will have a span, which is a measure of the width of the range of variables it correlates. A large span implies a general theory while a small span implies knowledge or a theory of local validity. The correlational function 122

KNOWLEDGE, INFORMATION AND STRUCTURE

of theories leads to the local character of knowledge (Nelson and Winter 1982). This local character implies that firms know how to do well things that they have done in the past, and other very similar things. The probability of success in search activities is higher if the search is on a range of events similar to those studied in the past. Scientific theories are likely to be more general/less local than technological knowledge, but even they do not cover an infinite range of variables. Therefore, the correlational structure of theories is common to scientific and technological knowledge, and leads to the local character of knowledge. From its local character, it follows that knowledge will accumulate preferentially in some ranges of variables or events, and that scientific and technological capabilities will grow preferentially in those ranges. That is, knowledge is cumulative. Another characteristic of knowledge that can be deduced from this approach is that knowledge development will be subject to path dependence. Once knowledge starts accumulating in a particular range it will continue to grow in the same range. Moreover, as firms’ technological capabilities increase in a given range, their ability to interpret and retrieve further similar knowledge increases. As Cohen and Levinthal (1989, 1990) have observed, R&D increases the absorptive capacity of firms (see also Foray 1991 on this issue). Of course, the returns to search activity and to other knowledge-creating mechanisms are crucial in leading to path dependence and cumulativeness. The presence of increasing returns in its early phases seems a factor that favours the emergence of a new paradigm. Once established, however, the paradigm could even be stable in the presence of constant returns to exploitation. Incremental innovation will then be the rule. During this period, the absorptive capacity of firms and other organisations that have started accumulating knowledge in the paradigm increases. When diminishing returns are encountered, there will be an inducement to switch to a different range of variables/events or to new theories. In this case the interpretative structures based on the old knowledge/theories will become redundant, corresponding to what Tushman and Anderson (1986) have called competence-destroying technological change. These discontinuities in knowledge temporarily eliminate path dependence and require a total change of the knowledge base of firms and organisations. The correlational function and the local character of theories/knowledge helps us to understand the nature of search activities. Search activities are a generalised analogue of R&D (Nelson and Winter 1982), and they constitute a form of learning by not doing. The distinction between basic research, applied research, and development has been widely used but it has never been completely satisfactory. They are very similar kinds of activities and there is a considerable continuity between them. It is possible to interpret search activities as explorations of the environment in given ranges of variables/ events. Search activities differ for the width (span) of the range studied and for the probability of obtaining the desired outcome. Thus basic R&D will 123

P.P.SAVIOTTI

have a wide range and a low probability of obtaining the desired outcome, while applied R, and even more D, will have a narrower range but a higher probability of success. We can expect search activities to evolve according to a life-cycle, starting with a wide range and a low probability (basic search), then identifying a more fruitful subset of the range, concentrating on it, and achieving a higher probability of obtaining the desired outcome. It is quite possible for the correlation density to increase within the narrow range chosen. The mature stage of a technology would correspond to the transformation of most search activities into D. Another important feature of knowledge that has been noted by a number of scholars is that it can be either tacit or codified (Polanyi 1962; Teece 1981; Nelson and Winter 1982). Tacit knowledge can be used to achieve particular outcomes, but cannot be easily communicated to others. On the other hand, knowledge expressed according to formal codes (that is, codified) can be much more easily communicated. In practice, there are many pieces of knowledge which are partly or imperfectly codified. It is probably better to conceive of both tacit and codified situations as extremes of a possible range, and to talk of the degree of codification of a particular piece of knowledge. A number of observations can be made about the tacit/codified nature of knowledge. First, when new knowledge is born there are always tacit elements in it. It is only during the growth and maturation of this knowledge that codification is gradually introduced. Some degree of codification is required because knowledge generation is a collective enterprise. All the participants must be able to know what has been done in the past in order to develop it further. Second, the degree of codification depends also on the structure of rewards (Dasgupta and David 1992; David and Foray 1994): researchers working for a firm may be discouraged from codifying knowledge, because that would make it more easily diffusable, in contrast to researchers working in public or academic institutions who are expected to disseminate their results. Third, while codification makes communication easier, there are costs involved in establishing the codes and in learning them once they have been established. Thus, while the knowledge produced in molecular biology is probably highly codified, it can only be easily communicated to people who know the code. In some developing countries there might be very few researchers able to retrieve such codified knowledge. Before considering the relationship between knowledge and information, we have to take into account that the knowledge used by a firm or other organisation has a collective character, and that it depends both on the individual human resources and on the mechanisms of interaction within the organisation. Such collective knowledge is called the knowledge base (KB) of the firm and it needs to be distinguished from its Revealed Technological Performance (RTP), that is the nature of its output. Such output can be an artefact or even a service. The KB and RTP of a firm are related, 124

KNOWLEDGE, INFORMATION AND STRUCTURE

in the sense that KB has to precede RTP in time. KB represents a potential that is later embodied in RTP. Each firm’s KB will differ in its location, that is the areas of knowledge where the firm has capabilities, and in the degree of accumulation of knowledge in these areas. The characterisations of knowledge previously described may be applied here. The KBs of different firms will differ in their location, span, specificity and accumulated knowledge. The more specific and tacit each KB is, the more appropriable it will be. The degree of accumulation can be estimated based on the total volume of search activities in the area. It is in principle possible to estimate a degree of similarity (or, equivalently, a distance) between the KB of an organisation and the external knowledge to be learned. These considerations can be reformulated as follows. The probability of learning a given type of external knowledge, and thus also the absorptive capacity of the firm, is inversely proportional to the distance (or directly proportional to the similarity) between the organisation’s knowledge base and the external knowledge under consideration. The difficulties that a firm experiences in learning increase, the greater the difference between the external knowledge to be learned and its pre-existing knowledge base. Thus an existing KB can be relatively easily extended to incorporate external knowledge in the case of an incremental innovation, but a firm experiences much more severe difficulties in incorporating the external knowledge on which a radical innovation is based. In fact, as Schumpeter (1942) foresaw, it is much more likely for a radical innovation to be introduced by new firms. Moreover, it may be difficult for an organisation to assimilate new external knowledge that is competence-destroying (Tushman and Anderson 1986). However, it must be remembered that the KB of a firm is constituted of many parts, usually located in different divisions of the firm. It is quite possible for an established firm to have a KB which is weak in a new technological area (e.g. genetic engineering) but to have strong competences in what Teece (1986) calls complementary assets, such as marketing. If the new product can be marketed using old channels, new firms would be in an advantageous position in the early stages of introduction, while the KB of pre-existing firms would have some decisive advantages in the later stages of large-scale production and commercialisation.

Knowledge and information Knowledge and information are two factors that have a considerable influence on the conduct and performance of firms. They are related but not identical, and it is important to distinguish between them. As we have seen in the previous section, knowledge provides us with correlations between variables, that is with generalisations, and it is also a retrieval/interpretative structure. Information, as used in information theory, however, gives us freedom of

125

P.P.SAVIOTTI

choice when we select a message. For example, one unit of information is required to choose between two messages a and b (Shannon and Weaver 1949; Saviotti 1988, 1991), and the number of units of information required increases with the number of messages. Information conceived in this sense has nothing to do with meaning. It is factual information. In a real system, for example a firm or an organisation, economically relevant types of information will be those required to describe the organisation itself, that is the types of human resources, capital equipment, etc., used (internal information), or those required to describe the external environment of the firm (external information). In each case we will need a certain number of messages (a certain quantity of information) in order to distinguish the human resources, the capital equipment, etc., existing from all other possible variations. As we have seen before, knowledge has a correlational function and it provides us with interpretative structures. Factual information can be understood and used only in the context of a particular theoretical framework. For example, the conditions to be used and the sequence of operations to be followed in order to prepare a particular composite material can only be understood by someone who knows some macromolecular chemistry and physics. Only individuals and organisations knowing a new theoretical framework such as molecular biology can correctly interpret the information that it creates. We can understand this problem from the viewpoint of information theory by remembering that information flows between the source and the receiver. Transmission of a message implies that it be encoded, sent through a channel, and finally decoded in order to make it intelligible to the receiver (Atlan 1972:32). Messages corresponding to radically new forms of knowledge cannot be interpreted by recipients holding only old forms of knowledge. Moreover, radical changes in knowledge can be interpreted as noise from the viewpoint of information theory. Noise reduces the intelligibility of messages for the receiver. The problem of noise can be overcome by sending several similar or identical messages instead of one, that is by redundancy (Atlan 1972). The consequence of noise is that the average quantity of information per symbol used in the messages transmitted falls. In other words, noise reduces the efficiency of information flows. In situations in which knowledge is changing rapidly, firms have difficulties interpreting the messages coming from their external environments. The greater the degree of novelty of external knowledge, the greater the number of similar messages firms will need to receive in order to acquire the same quantity of information; that is, the greater will be the redundancy required. Therefore, radical changes in knowledge reduce the efficiency of information flows. A firm not possessing the new knowledge will have higher search costs for knowledge and information. A number of implications of patterns of information use have been outlined in two previous papers by the present author. Changes in organisa-tional 126

KNOWLEDGE, INFORMATION AND STRUCTURE

structure can lower the quantity of information required to describe the organisation or, correspondingly, raise the quantity of information it can store (Saviotti 1988, 1991). By using informational entropy to measure this quantity of information, it is possible to prove that the introduction of hierarchical organisations and an increased degree of division of labour can lower the quantity of information required to describe such systems. Furthermore, it can be proved that the M form organisation is informationally more efficient than the U form in presence of a given output variety (Saviotti 1991:190–3). However, such demonstrations are valid only at constant levels of knowledge, when there are no problems arising from noise coding, and no redundancy is needed. Later, it will be argued that the organisational consequence of coding difficulties and of noise, and therefore of radical changes in knowledge, is an increase in coordination problems. At a constant level of knowledge, a highly structured organisation is informationally efficient because it can store a large quantity of information but only needs to use a small quantity. However, such an organisation cannot easily adapt to changes in its environment. The information it can store is of a given type that corresponds to a particular theoretical framework. But, as we have seen, new theoretical frameworks will generate information of new types, uninterpretable by individuals or organisations that possess only the old type of knowledge. Adaptation to new knowledge implies destroying the old organisational structure in order to create a new one. In this sense we can understand that a disordered organisational structure, while being able to store less information, can have a higher level of potential information. The disordered organisation can be more easily transformed into other organisational structures, each one efficient with respect to a given type of information. From this it follows that the distinction between potential and stored information is a particularly useful one when there are important qualitative changes in the environment of organisations, such as changes in knowledge or paradigmatic changes. More random, or organic, structures are likely to be less efficient but more adaptable than mechanistic structures (Burns and Stalker 1961). Another function of theories which follows from their correlational function, and which has important implications for this paper, is their capacity to save information. The quantity of information required to describe the system will increase with the variety of the elements composing it. However, if a constraint exists between the elements of the system, the quantity of information required may fall when there is constant variety. For example, if several elements of the system are tightly connected, we only need to know the position of one in order to know that of all the others. Likewise, if there is a theoretical relationship between different components, we only need information about one component in order to know the whole system. Theories can help us to save on the quantity of information required to describe a system. 127

P.P.SAVIOTTI

Further development of these considerations requires a more detailed discussion of the concept of the external environment of an organisation.

Knowledge and the external environment of firms The external environment of the firm is the subset of what is outside the boundaries of the firm but which influences the behaviour and performance of the firm itself. For the purposes of this paper, two problems are important. First, what are the most relevant aspects of the external environment?; and second, what relationship is there between the structure of the external environment and its changes on the one hand, and the nature of the organisations operating in it on the other hand? The second problem can be better understood if we think that the main goal of firms, and of other organisations, is to adapt to the external environment in which they operate. This process of adaptation includes both adaptation to a given environment and the modification of the environment in order to suit the firms goals. In other words, the process being described here is one of coevolution of organisations and of their environments. It seems logical that firms have to adopt strategic goals and organisational structures that improve their adaptation to the external environment. Moreover, the process of adaptation is not automatic, but requires learning. Furthermore, adaptation is limited by inertia (Hannan and Freeman 1977; Barley and Freeman 1991), that is particular structures developed to adapt to given environments may survive even when the conditions of the environment change. Organisation theorists have developed typologies of organisational environments. Such typologies reflect the complexity of these environments. Different typologies are broadly similar, but there is no agreed model. For example, there is a general agreement that environments may be viewed as having two components, climate and texture (see, for example, Aldrich 1979; McKelvey 1982). Climate is the general non-purposeful milieu surrounding the texture of all organisations in a particular environment. Texture is the set of all interactions amongst the (myopically) purposeful organisations operating in the environment. There is no agreed set of attributes of the climate, although the lists provided by different authors are broadly similar (see McKelvey 1982:122). A description of texture which is still widely used was provided by Emery and Trist (1965). They identified four types of texture: Type I, Placid Randomised; Type II, Placid-Clustered; Type III, DisturbedReactive; Type IV, Turbulent Field. Types I and IV differ in their rates of change and in the intensity of interaction of the participating organisations. For the purposes of the present work, a simplified version of these typologies will be useful. Amongst the attributes of the external environment there are both impersonal elements, such as location and the prices of raw materials, and organisations other than those on which the analysis focuses.

128

KNOWLEDGE, INFORMATION AND STRUCTURE

The most important kinds of these other organisations are: (a) competitors; (b) users/customers; (c) suppliers of inputs, equipment, knowledge and human capital; and (d) regulating institutions. Amongst the various attributes of the environment, those which seem to be common to most typologies are: (a) complexity/diversity; (b) interactivity; and (c) the rate of change (quantitative/qualitative). Environmental typologies are not useful per se, but because organisations are expected to develop strategic goals and organisational structures adapted to their external environments. Empirical evidence of this adaptation began to be gathered starting in the 1960s (see, for example, Burns and Stalker 1961; Woodward 1965; Lawrence and Lorsch 1967). These and other authors found evidence of a relationship between external environments, production technologies and organisational structures. Mintzberg (1979) derived the following generalisations from a number of empirical studies. 1 2 3 4 5

The more dynamic the environment, the more organic the structure. The greater the complexity of the environment, the more decentralised the structure. The more diversified the organisation’s markets, the greater the propensity to split into specialised divisions, given favourable economies of scale. Extreme hostility in its environment drives any organisation to centralise its structure temporarily. A high environmental heterogeneity encourages organisations to adopt different types of structure for different subsets of the environment.

A number of comments are relevant in order to focus the results of organisation theories on the goals of the present paper. First, the theories described concentrate predominantly on the effect of environmental characteristics and changes in those characteristics on the structure of organisations, rather than on their boundaries. The boundaries, except possibly for their location, are taken as given. Second, technological knowledge is an increasingly important component of the external environment of organisations. This is reflected in an increasing importance of the knowledge base of organisations in determining their performance. Third, the adaptation of organisations to their external environments is a slow process which requires learning. Furthermore, even after a particular organisational structure is developed in order to improve adaptation to the external environment, such a structure can acquire a stability which will confer on it a considerable degree of inertia in presence of environmental changes. In other words, there can be hysteresis in the process of adaptation of organisations to their external environments. Likewise, as was already mentioned, adaptation does not occur in a static and unchanging environment. In a number of cases firms and other organisations may try to modify their external environments to suit their goals. The knowledge created by a particular firm or organisation can 129

P.P.SAVIOTTI

become a necessary component of the knowledge bases of all comparable firms or other organisations. Thus the invention of the transistor by Bell Labs created a new knowledge area that other electronics firms could not neglect. Such a situation can best be described as the coevolution of organisations and of their environments. Knowledge is, therefore, one of the components of the external environment and one of the most relevant aspects of firm structure and behaviour. Firms scan the external environment in order to detect any possible pieces of external knowledge which are useful for their productive purposes. When they find such useful pieces of knowledge, they have to internalise them. As we have seen in the previous section, the capacity of a firm to learn and internalise knowledge depends on the firm’s previous knowledge. Such path dependency, however, is not infinite. As long as knowledge changes incrementally, it will also be cumulative, but at the moment of transition, when radically new knowledge appears, path dependency is likely to fall and the distribution of competitive advantage will change drastically. Recently, economists have themselves begun to develop a more complex characterisation of the external environment of firms. Nelson and Winter (1977, 1982) proposed a ‘selection environment’, which could be a generalised analogue of the market for the most varied types of organisations such as firms, schools, hospitals, prisons, and so on. Their selection environment can be specified by means of four elements: (a) the definition of worth or profit for the firms in the sector; (b) the manner in which consumers and regulatory institutions influence what is profitable; (c) investment processes; and (d) imitation processes. Metcalfe and Gibbons (1989) have developed a characterisation of the selection environment for the analysis of economic growth and competition, which emphasises the environment’s capacity to evaluate different product and process technologies, its rate of growth, and its degree of homogeneity.

The division of labour and coordination At least since Adam Smith (1982 [1776]), the division of labour has been considered to be one of the determinants of economic growth and development. Increasing the extent of the division of labour can increase productivity and, therefore, lead to a growth in total output. Any complete process can be divided into a number of production stages and provided with the level of coordination required to transfer the output of each stage to other stages. However, the division of labour changes continuously during the process of economic development, and each new type of division of labour requires a redefinition of the coordination stages. Furthermore, the coordination required determines the types of organisations existing in the economic system. In order to analyse the organisation of production, we can

130

KNOWLEDGE, INFORMATION AND STRUCTURE

start by imagining a simplified economic system characterised by the greatest division of labour possible at the time and constituted by individual producers. Each one of them produces the output of a single stage and sells it through the market. We can formulate the problem by asking ourselves why such an economic system has never existed, which is equivalent to asking why firms exist. A number of answers have been given, such as the theory of teams (Alchian and Demsetz 1972) and, more recently, transaction cost analysis (Williamson 1975, 1985). Transaction cost analysis provides a theoretical justification for the existence of hierarchical structures, whose historical development has been studied by Chandler (1962, 1977). It leads to the conclusion that the only stable organisational forms are markets and hierarchies. While these models were and continue to be extremely useful, they have recently been challenged by the emergence of inter-institutional collaborative agreements (IICAs) in the 1980s (Chesnais 1988; Mytelka 1991; Hagedoorn and Shackenraad 1990, 1992). The approach adopted in this paper is that all types of organisational structures reflect combinations of the extent of the division of labour and of the extent of coordination that prevail, and that the stability of different combinations depends on the nature of their external environments. We can start by trying to understand what factors influence the division of labour and coordination. Can the division of labour be increased indefinitely, and would indefinite increases in the division of labour produce constant or falling marginal increases in productivity? An answer given to these questions in the past has been that the division of labour is limited by the extent of the market (Stigler 1951). According to Becker and Murphy (1992) the division of labour is not only limited by the extent of the market, but also by knowledge and by coordination costs. Increased knowledge allows a greater degree of specialisation to take place by reducing coordination costs. However, these authors by no means make clear how greater knowledge can reduce coordination costs. This chapter adopts a similar approach by taking into account several factors, including knowledge, that can influence the division of labour and coordination. The problem of the division of labour is placed here in the context of knowledge-intensive production processes, taking into account the structure of the external environment in which firms operate, and the role played by knowledge generation in environmental change. As has been pointed out, we can in principle conceive of dividing the production processes taking place in an economy into the maximum possible number of stages. Each stage, consisting of the transformation of an input into an output, is performed by an individual. In this sense the economy would then be composed of a collection of individuals performing elementary stages and trading their output through market transactions. Quite clearly no economy has ever been organised in this way. Instead of a collection of individual workers producing and trading output in elementary stages, we have collections of groups of workers, each group separated from other 131

P.P.SAVIOTTI

groups by means of an institutional interface. The location of such institutional interfaces was taken as given before transaction cost analysis. In order to analyse all types of organisational structures, including interfirm collaboration, we can use the following properties of the institutional interface: 1 2

3

Location: What factors determine the location of the institutional interface between a firm and other firms, and between a firm and markets? Discreteness or diffuseness/fuzziness with respect to flows of information or human or physical capital: Is the institutional interface clearly defined, in the sense that we can always tell of which organisation each individual is a member? Or is the interface diffuse, as when individuals are members of two or several organisations? Intertemporal stability: Is the institutional interface stable or does it change frequently?

At least three types of changes in an economic system can influence the division of labour: (a) market expansion; (b) a change in market structure, leading, for example, to (bl) completely new types of output, or (b2) the specialisation of existing types of output; and (c) process innovation, with an unchanged output structure. Case (b) is almost invariably due to new knowledge, with a greater degree of novelty required for (bl) than for (b2). In particular, if a growth in variety is a necessary condition for economic development (Saviotti 1994), we can expect that the division of labour will increase systematically. While variety growth in this case would lead to income growth and, therefore, to market expansion, this mechanism is different from case (a), because (a) in principle could take place without any change in output structure. Case (c) can in principle take place with a constant market size, if knowledge changes allow a redesign of the productive processes used, thereby improving their efficiency. Case (c) would in this case not require previous expansion in the market, but it might induce a subsequent market expansion if lower output costs were accompanied by a high price elasticity of demand. In short, the extent of the market is but one of the factors influencing the division of labour. Changes in knowledge, either directly or indirectly are an equally important factor. First, at least some production processes follow a life-cycle pattern of development (Abernathy and Utterback 1975, 1978). Such processes begin with low-volume production, a flexible organisation, using general purpose machine tools, a multiplicity of product designs, with the primary emphasis on product performance and product innovation. As the product matures and as production volumes increase, more rigid organisational structures emerge, based on specialised equipment and a dominant design. Product innovation is replaced by process innovation. Increased production volumes are likely to lead to an increased division of labour. Second, as the technology 132

KNOWLEDGE, INFORMATION AND STRUCTURE

matures and a new paradigm becomes established (Dosi 1982), changes in knowledge and the greater availability of new vintages of human capital will also allow a greater division of labour. In production processes characterised by such patterns of development, we can expect the division of labour to increase systematically as the market expands and the technology matures. It is important to note that such process evolution is only likely in the presence of a stable external environment, in which a large-scale, rigid and dedicated organisational structure emphasises efficiency rather than adaptability. If rapid and radical changes were to take place in the external environment, small and flexible organisations adopting an organic style, such as IICAs (Burns and Stalker 1961), would have a greater survival ability than large, rigid, mechanistic ones. A constant external environment allows a high degree of process standardisation and is likely not only to increase the divisibility of labour, but also to reduce both time and coordination problems. To summarise this section, we can say that changes in knowledge and in process organisation, which are components of environmental changes, have a profound influence on both the division of labour and coordination.

Organisations, organisational boundaries and environmental change The problem of coordination costs underlies transaction cost analysis (Williamson 1975, 1985). Production processes can be divided into a number of stages connected in an input—output fashion. The passage of the output of stage k to stage (k+1), of which it becomes an input, is a transaction. It is precisely in transactions that problems of coordination occur. The more difficult the coordination, the higher the transaction cost. Transaction costs are essentially coordination costs. The definition of coordination here has to be used in a sufficiently broad way to include, for example, transport costs. Then, following Williamson, the particular governance structure used, and therefore the boundaries of the firm, depend on the balance between production costs (the costs of the production stages) and transaction costs. When transaction costs are very low, markets are the best governance structure, but when transaction costs are very high relative to productions costs, hierarchical organisations are the best governance structure. That is, the balance between production and transaction costs determines the location of institutional boundaries and interfaces. Two problems arise at this point. First, since markets and hierarchical organisations can be conceived as the extremes of a range of possible situations, with intermediate cases falling in between, a relevant question which emerges is whether interfirm collaboration is an intermediate case between the two extremes (De Bresson and Amesse 1991). Second, transaction cost analysis

133

P.P.SAVIOTTI

tends to address the problem of the location of interfirm boundaries, but does not consider the nature and the permanence of such boundaries. The problem is in part linked to the absence in transaction cost analysis of the external environment of the firm. The boundary of an organisation separates its internal and external environments. Such a boundary can be considered as constituted by all the members of the organisation who communicate with other organisations. This boundary can be very diffuse, as when many members of the organisation communicate externally, or very sharp, as when only one or a few members of the organisation communicate. Organisational boundaries are themselves part of the process of adaptation, and they change together with organisational structures following changes in the external environment. Changes in knowledge in the external environment are likely to require a new division of labour and to give rise to greater coordination problems. One of the advantages of the division of labour is that it reduces the quantity of information required by each worker (Saviotti 1988, 1991). In a very static environment, and for very well-known processes, the worker in stage (k+1) only needs to know what to do with the output of stage k, and not how such an output has been obtained through the sequence of the previous k stages. In these conditions, coordination of the subsequent productive stages is easy. Coordination costs are very low with respect to production costs, and the division of labour is limited only by the extent of the market (Smith 1982 [1776]; Stigler 1951; Becker and Murphy 1992). However, if we imagine a production process which changes continuously as it is carried out, then coordination problems will become very significant. The output of each productive stage will have to be accompanied by very detailed instructions specifying what to do with it. Alternatively, each worker must know what the other workers are doing in order to be able to perform their operations. The knowledge of each worker must overlap with that of other workers. In this sense the growth in the division of labour that takes place within the evolution of a given process can be compared to standardisation. A change in knowledge also has consequences for the division of labour between organisations. For static and well-known processes, each organisation can purchase inputs (materials, equipment, knowledge) in an embodied form, where knowledge would be embodied in texts. Knowledge has to be retrieved and decoded by the appropriate vintages and types of human capital in the organisation. Radical changes in technological knowledge, however, will require a redefinition of the division of labour within the organisation, and will make the old vintages of human capital in the organisation redundant. At the beginning of a new technological paradigm, new vintages of human capital will be in scarce supply, given the delays of teaching institutions. No firm will be able to purchase all the new human capital required to retrieve and decode the new knowledge. Interfirm collaboration will then be needed to reach the critical mass of human capital 134

KNOWLEDGE, INFORMATION AND STRUCTURE

required to use the new knowledge. In these conditions, the knowledge base of each firm—the collective knowledge that the firm uses for its productive purposes—will overlap different institutional boundaries, including individuals who are members of different organisations. Overlap of the knowledge bases of different organisations implies that each of them has to hold more knowledge and more information than would otherwise be required. If we imagine placing the productive stages of two different firms A and B on a line, we can see that the number of stages about which each firm has to have knowledge and to acquire information is greater the greater the degree of overlap (Figures 4.1 (a) and 4.1(b)). A period of paradigmatic change is characterised by a high degree of uncertainty. The expectation of further radical changes as the new paradigm moves towards normalisation or standardisation is likely to limit the commitment of each firm to new productive processes and new outputs. Interfirm collaboration would allow firms to enter new fields with a minimum commitment and to increase the reversibility of entry by minimising exit costs. Thus both risk limitation and the scarcity of the new vintages of human capital would tend to favour interfirm collaboration with respect to more stable arrangements. Thus one could say that radical changes in knowledge will lead (a) to a redefinition of the intra-organisational division of labour, with a greater degree of overlap of the knowledge of individual workers, and (b) to an increased fuzziness of institutional boundaries as a result of the overlapping of the knowledge bases. As a new technology matures and becomes more

Figures 4.1 (a) and 4_1 (b) Overlap of knowledge base. The number of stages about which knowledge has to be held and gathered increases with degree of overlap of the knowledge bases of the two firms A and B. There is no overlap in Figure 4.1(a), while there is some overlap in Figure 4.1(b)

135

P.P.SAVIOTTI

standardised, however, the shortage of human capital will decrease and coordination problems will be reduced. We can then expect that the overlap of the knowledge of individual workers and of organisations will be reduced, as will the need for interfirm collaboration.

Summary and conclusions In this chapter the relationship between information and knowledge and its implications for organisational structures have been analysed, with emphasis placed on three functional aspects of knowledge—correlation, interpretation, and the saving of information. Information has been considered to be of a factual nature and defined in a way similar to that employed in information theory. However, the factual nature of information is relative to a particular theoretical framework or knowledge type. Radical changes in knowledge introduce noise, require redundancy to compensate for it, and reduce the intelligibility of the information transmitted from the environment to the firm. In order to be able to compare traditional organisational structures such as markets and hierarchies, and newer ones such as inter-institutional collaborative agreements (IICAs), it has been assumed that all types of organisations utilise particular combinations of the division of labour and coordination, with the stability of each combination being determined by the external environment in which the organisation operates. Fast and radical changes in knowledge invalidate previous knowledge bases and organisational structures, which in turn requires a redefinition of the division of labour and exacerbates coordination problems. If the environment remains stable after a qualitative transition, gradual adaptation may lead to stable, hierarchical structures. On the other hand, if the rate of radical environmental change remains higher than the rate of adaptation, more flexible and reversible structures such as IICAs may be favoured. If we assume that particular knowledge fields, corresponding to sets of applications, follow a life-cycle leading to a progressive reduction in the rate of qualitative change and to a certain degree of standardisation in procedures, we can then expect flexible structures such as IICAs to be particularly well suited to the early stages of development of a new knowledge field. On the other hand, hierarchical structures could be more appropriate to more stable, even if complex, environments. From the environmental typology presented here, it follows that different organisational structures differ along the three dimensions of rate of change, degree of complexity and degree of interactivity. We could say that interfirm collaboration is an intermediate case between the extremes of market relations and hierarchical organisations if the three governance structures could be placed along a single dimension. But the three governance structures

136

KNOWLEDGE, INFORMATION AND STRUCTURE

cannot be placed along a single dimension because they differ along at least three. Furthermore, interfirm collaborations arise for rates of change and degrees of complexity greater than those at which either markets or hierarchies are found, and are therefore beyond their range. Finally, interfirm collaboration and networks require the existence of a certain minimum amount of trust (De Bresson and Amesse 1991), while the existence of hierarchical forms of organisation is justified in part by the presence of opportunism in transactions between individuals. Thus it is not completely accurate to represent interfirm collaboration as an intermediate case between markets and hierarchies. Whether IICAs are a temporary or a permanent organisational form depends on the balance between the rate of radical environmental change and that of adaptation. The problem then centres on the degree of permanence of the environmental conditions which give rise to interfirm collaboration. These environmental conditions are a greater rate of change and greater uncertainty and complexity than those leading to hierarchical organisations. Such conditions are particularly common in the transition to a new technological paradigm and in its early stages. Rate of change, complexity, and uncertainty fall during the normalisation phase of a paradigm and the maturation of a technology. It would then be possible for new technologies to be introduced by networks of collaborating firms and for hierarchical organisations to take over these same technologies as they mature. The concentration of interfirm collaboration in the economic system, as opposed to that of market relations and hierarchies, is a function of the percentage of new technologies in use and of their pervasiveness. To the extent that the approach taken here is correct, we can expect the economic weight of interfirm collaboration to be proportional to the extent and rate of radical, knowledge-intensive change in the economic system. Such change was almost certainly greater and more knowledge-intensive in the 1970s and 1980s than ever before in recent history, and we can reason-ably expect knowledge intensity to increase in future. In this case, interfirm collaboration will have an assured place amongst the prevailing forms of industrial organisation.

References Abernathy, W.J. and Utterback, J. M (1975) ‘A Dynamic Model of Process and Product Innovation,’ Omega, 3(6):639–66. ——(1978) ‘Patterns of Industrial Innovation’, Technology Review, (June/July): 41–7. Alchian, A. and Demsetz, H. (1972) ‘Production, Information Costs and Economic Organisation’, American Economic Review, 62:777–95. Aldrich, H.E. (1979) Organisations and Environments, Englewood Cliff: N.J.: Prentice Hall. Atlan, H. (1972) L’organisation biologique et la théorie de l’information, Paris: Hermann. Barley, S.R. and Freeman, J. (1991) ‘Niches as Networks: The Evolution of Organisational Fields in the Biotechnology Industry’, mimeo, Cornell University. 137

P.P.SAVIOTTI

Becker, G.S. and Murphy, K.M. (1992) ‘The Division of Labour, Coordination Costs and Knowledge’, Quarterly Journal of Economics, 107:1137–60. Burns, T and Stalker, G.M (1961) The Management of Innovation, London: Tavistock. Chandler, A.D. (1962) Strategy and Structure, Cambridge Ma: MIT Press. ——(1977) The Visible Hand, Cambridge Ma: Harvard University Press. Chesnais, F. (1988) ‘Technical Cooperation Agreement Between Independent Firms, Novel Issues for Economic Analysis and the Formulation of National Technological Policies’, STI Review, 4:51–120. Cohen, M. and Levinthal, D. (1989) ‘Innovating and Learning: The Two Faces of R&D’, Economic Journal, 99:569–96. ——(1990) ‘Absorptive Capacity: A New Perspective on Learning and Innovation’, Administrative Science Quarterly, 35:128–52. Coombs, R., Saviotti, P.P. and Walsh, V. (1987) Economics and Technological Change, London: Macmillan. Dasgupta, P. and David, P. (1992) ‘Toward a New Economics of Science’, Center for Economic Policy Research: Stanford University. David, P. and Foray, D. (1994) ‘Accessing and Expanding the Science and Technology Knowledge Base’, Paris OECD, DSTI/STP/TIP(94) 4. Davies, S. (1979) The Diffusion of Process Innovation, Cambridge: Cambridge University Press. De Bresson, C. and Amesse, F. (1991) ‘Networks of Innovators: A Review and Introduction to the Issue’, Research Policy, 20:363–79. Dosi, G. (1982) ‘Technological Paradigms and Technological Trajectories: A Suggested Interpretation of the Determinants and Directions of Technical Change’, Research Policy, 11:147–62. Emery, F.E. and Trist, E.L. (1965) ‘The Causal Texture of Organisational Environments’, Human Relations, 18:21–32. Foray, D. (1991) ‘Towards an Economic Analysis of the Organisations of Research and Development’, Revue d’Economie Politique, 101:779–808. Freeman, C. (1982) The Economics of Industrial Innovation, London: Pinter. Hagedoorn, J. and Shackenraad, J. (1990) ‘Interfirm Partnerships and Cooperative Strategies in Core Technologies’, in Freeman and Soete (eds) New Explorations in the Economics of Technological Change, London: Pinter. ——(1992) ‘Leading Companies and Networks of Strategic Alliances in Information Technologies’, Research Policy, 21:163–90. Hannan, M.T. and Freeman, J. (1977) ‘The Population Ecology of Organisations’, American Journal of Sociology, 82:929–64. Lawrence, P. and Lorsch, J.L. (1967) Organisations and Environment: Managing Differentiation and Integration, Cambridge, Ma.: Harvard University Press. Layton, E.T. (1974) ‘Technology as Knowledge’, Technology and Culture, 15:31–41. McKelvey, B. (1982) Organisational Systematics, Berkeley: University of California Press. Metcalfe, J.S. and Gibbons, M. (1989) ‘Technology, Variety and Organisation: A Systematic Perspective on the Competitive Process’, Research on Technological Innovation, Management and Policy, 4:153–93. Mintzberg, H. (1979) The Structure of Organisations, Englewood Cliffs: N.J.Prentice Hall. Mytelka, L.K. (ed.) (1991) Strategic Partnership and the World Economy, London: Pinter. Nelson, R. and Winter, S. (1977) ‘In Search of Useful Theory of Innovation’, Research Policy, 6:36–76.

138

KNOWLEDGE, INFORMATION AND STRUCTURE

——(1982) An Evolutionary Theory of Economic Change, Cambridge, Ma.: Harvard University Press. Polanyi, M. (1962) Personal Knowledge: Towards a Post-Critical Philosophy, New York: Harper Torchbooks. Saviotti, P.P. (1988) ‘Information, Entropy and Variety in Technoeconomic Development’, Research Policy, 17:89–103. ——(1991) ‘The Role of Variety in Economic and Technological Development’, in Saviotti and Metcalfe (eds) Evolutionary Theories of Economic and Technological Change: Present State and Future Prospects, London: Harwood. ——(1994) ‘Variety, Economic and Technological Development’, in Shionoya and Perlman (eds) Technology, Industries and Institutions: Studies in Schumpeterian Perspectives, Ann Arbor: University of Michigan Press. ——Z(1996) Technological Evolution, Variety and the Economy, Aldershot: Edward Elgar. Schumpeter, J.A. (1942) Capitalism, Socialism and Democracy, New York: Harper. Shannon, C.E and Weaver, W. (1949) The Mathematical Theory of Communication, Urbana: University of Illinois Press. Smith, A. (1982) The Wealth of Nations, Harmondsworth: Penguin Books, [original edition 1776]. Stigler, G.J. (1951) ‘The Division of Labor is Limited by the Extent of the Market’, Journal of Political Economy, 49:185–93. Stoneman, P. (1983) The Economic Analysis of Technological Change, Oxford: Oxford University Press. Teece, D. (1981) ‘The Market for Know-How and the Efficient International Transfer of Technology’, Annals of the American Academy of Political and Social Science, 458:81–96. ——(1986) ‘Profiting from Technological Innovation: Implications for Integration, Collaboration, Licensing, and Public Policy’, Research Policy 15:285–305. Tushman, M.L. and Anderson, P. (1986) ‘Technological Discontinuities and Organisational Environments’, Administrative Science Quarterly, 31:439–65. Williamson, O.E. (1975) Markets and Hierarchies: Analysis and Anti-trust Implications, New York: Free Press. ——(1985) The Economic Institutions of Capitalism, New York: Free Press. Woodward, J. (1965) Industrial Organisation: Theory and Practice, Oxford: Oxford University Press.

139

5

TECHNOLOGICAL CHANGE, TRANSACTION COSTS, AND THE INDUSTRIAL ORGANISATION OF COTTON PRODUCTION IN THE US SOUTH, 1950–1970 Lee J.Alston1

The cotton economy of the US South underwent a dramatic transformation during the period 1950 to 1970. Cotton went from a highly labour intensive to a capital intensive crop. The number of sharecroppers and tenants in the ten leading cotton-producing states in the South fell from 807,013 in 1950 to 113,351 in 1970 (US Department of Commerce 1952, 1971). The complete mechanisation of cotton production, i.e., ploughing, cultivation and harvesting, was responsible for the decline in labour intensity (Aiken 1978; Alston 1981; Fite 1984). The mechanisation of cotton also brought about other changes in the industrial organisation of cotton production, and these are the focus of this paper: (a) mechanisation increased the costs of using sharecrop and tenant contracts compared to wage contracts; and (b) the adoption of mechanical pickers necessitated a technological compatibility between cotton pickers and cotton gins that influenced the farm/market boundary and the survival of cotton production in some regions.2

The organisation of cotton production before the cotton picker, circa 1950 Before the widespread adoption of the mechanical cotton picker, beginning in the 1950s and 1960s, the production of cotton was highly labour intensive. Both small owner-operated farms and large plantations, employing wage labourers, sharecroppers and tenants, co-existed and produced for the market. 140

INDUSTRIAL ORGANISATION OF COTTON PRODUCTION

The advantage of the small farms was low costs of monitoring labour while the advantage of the large plantation was more traditional economies of scale. Before the mechanisation of cotton production, monitoring costs were high on plantations because workers were spread over a considerable physical space and it was difficult to judge labour effort by measuring the output of cotton. In pre-mechanised and pre-scientific agriculture, the variation in output is high and the causes of the variation are often, but not always, the result of insufficient work effort. To limit the costs of monitoring labour effort directly, plantations used a mix of contracts, with the mix depending on the endowments of owners and workers (Alston and Higgs 1982). The contracts varied in two important respects: who bore the risk; and how great were the incentives for expending labour effort. Under a wage contract the landlord bears all of the risk associated with poor yields or prices, whereas under a sharecrop arrangement the risk is shared, and under a fixed-rent arrangement the tenant bears the majority of the risk. Under a wage contract, because the income of workers is independent of their labour effort, workers have an incentive to expend as little effort as possible unless monitored. Labour effort can be elicited with less supervision by tying effort to reward. Sharecrop contracts give more incentive for labour to work without supervision than do wage contracts, and fixed-rent contracts give the greatest incentive. But giving labour an incentive for effort comes at a cost. Under competitive conditions, labour earns more on average as the contract moves from a wage to a sharecrop to a tenant contract, because competition among landlords ensures that labour captures some of the gains from the lower risk and monitoring costs that landlords face. To produce cotton requires land, labour effort, labour know-how (human capital) and physical capital (e.g. mules and tractors). The landlord contributes land to the production process and workers contribute labour effort. The remaining inputs of know-how and physical capital may be owned by either party at the time of contracting. Holding risk constant, contract choice will be driven by an incentive to economise on the transaction cost of monitoring. The monitoring costs will vary across and within contract form, depending on who brings what inputs to the production process. For example, for a sixteen-year-old worker with little experience in agricultural production and no physical capital, the owner will have to supply all of the inputs except labour effort. Given that the landlord has to be on the farm to give advice (know-how) and monitor the care of his physical capital, the marginal cost of monitoring labour effort is low, so a wage contract should be transacted. Alternatively, if a worker owns his own work-stock and knows how to farm, the marginal cost of monitoring effort alone would be high, so tenant contracts should be used to economise on transaction costs. Landlords used paternalism as well as contractual form to give workers an incentive for greater work effort without supervision.3 Paternalism served as 141

LEE J.ALSTON

an efficiency wage of sorts.4 Landlords increased the cost of job loss for workers caught shirking by providing them with a variety of in-kind services such as a house, credit, intercession in markets and, most importantly for black workers, protection from a discriminatory state. In return for paternalistic goods, workers gave dependable labour services. Landlords used both tractors and mules as power sources for ploughing, though the tractor became more widely adopted during World War Two. The reluctance of many farmers to adopt the tractor resulted from the seasonality of labour requirements in cotton (Aiken 1978; Fite 1984; Whatley 1985). The demand for labour over the season can be decomposed into a demand for ploughing/planting, cultivation (weeding) and harvesting. Before the mechanical harvester, the largest demand occurred during harvest and the least demand during ploughing/planting. The tractor mainly reduced the demand for labour for ploughing/planting. If a farmer dispenses with labour during the ploughing/planting season, labour may migrate out of the region and not be available when demand increases for cultivation and harvest. In short, the lack of a cotton picker meant only a limited adoption of the tractor in the South compared to other agricultural regions of the country. For example, in 1950, 66 per cent of the farms in the Census North reported having tractors, while the comparable figure for the Census South was only 24 per cent (US Department of Commerce 1956:30, 206) In the latter half of the nineteenth century, ginning technology dramatically increased the capacity of individual gins, which in turn resulted in large economies of scale.5 The result was that, by the turn of the century and certainly by 1945, ginning was no longer an activity undertaken by an individual plantation owner but was instead a service purchased in the market. From the turn of the century to 1945 the number of cotton bales per gin increased fourfold (Aiken 1973:208). The ‘new’ gin plants were frequently part of a large enterprise that leased land, supplied credit, sold fertiliser, and, later, crushed cotton seed and purchased and stored cotton. Over the first half of the twentieth century the ginning of cotton was disintegrated from the production of cotton and the ginning activity became integrated with other input and output activities of farmers. By World War Two the farm/market boundary was one where farmers: (a) purchased their seeds and fertilisers in the market; (b) employed household labour or purchased labour in the market, which used the capital of the landlord (tractor or mule); or rented their land to tenants who owned their own capital; and (c) landlords or tenants sold their cotton to firms that owned gins, sometimes the same firm from which they had purchased seeds and fertilisers.

142

INDUSTRIAL ORGANISATION OF COTTON PRODUCTION

The adoption of the mechanical cotton picker, and changes in the industrial organisation of cotton production The percentage of cotton picked by machine across the Southern states is shown in Table 5.1. There was considerable variation in the use of cottonpicking equipment across the Southern states, with adoption of machines tending to be slower in the South-eastern states.6 For example, in I960 the percentage of cotton picked by machine in North Carolina, South Carolina and Georgia was about 11 per cent, while the average for Texas and Oklahoma exceeded 60 per cent. For the South overall it took about twenty years for universal adoption. For widespread adoption to be profitable, several advances were required: (a) genetic changes in seeds such that the bolls ripened uniformly in a shorter span of time; (b) better defoliants so that the lint was less trashy; and (c) adoption of gins that had cleaners and dryers. The new seeds and defoliants became available over the course of the 1950s. The technology for the new gins was readily available by the end of

Table 5.1 Percentage of upland cotton mechanically harvested, 1949_1972

143

LEE J.ALSTON

World War Two, but not all gin owners perceived it as profitable to purchase cleaners and dryers. The initial lack of improved seed varieties and defoliants after World War Two partially explains the overall rate of adoption over time, while the differential rate of adoption of cleaners and dryers for gins partially explains the variation in percentage of cotton picked by machine across the South. My purpose, though, is not to explain the adoption per se, but rather to examine how adoption affected contracting and the firm/market boundary. The mechanical picker and the costs of contracting7 It is well known that mechanical adoption in agriculture displaced labour, but it is less well known how it influenced the decision to contract with the remaining labour. Because the census collects information on farms based on operators rather than ownership, we do not have precise figures on the mix of contracts used in agriculture. Nevertheless, we can get an idea by looking at the total number of tenant and sharecrop contracts and at the amount of dollars expended on wage labour across the ten leading cotton-producing states in the South, as presented in Table 5.2. I use dollars of wage expenditures rather than number of wage labourers, because the census did not collect data on full-time wage labourers. As you can see, the number of sharecroppers and tenants fell each census year for each state. For the South as a whole the decline was 86 per cent. The expenditures on wages also fell, but not nearly as dramatically: the total decline in real dollars was 47 per cent.8 As argued below, cotton was labour intensive, with high total costs of monitoring labour. Sharecrop and, even more so, tenant contracts give labour an incentive to supply greater work effort without direct supervision. As a result, before mechanisation, employers used a mix of contracts to economise on monitoring costs. The optimal mix of contracts would change with mechanisation because machines standardise the production process. In economist’s jargon, the variation in the marginal productivity of labour falls with mechanisation. In addition, because mechanisation displaced so much labour, the value of retaining a job increased. The result would be that labour effort without supervision should increase, because workers would fear losing their jobs if caught shirking.9 Finally, the tractors and cotton pickers that replaced mules and labour tended to be owned by employers. This was in contrast to the ownership of mules which were within the financial reach of many tenants. Because the employers owned the physical capital, they had an incentive to monitor its use in order to limit depreciation of their asset. As a result, the marginal cost of monitoring labour work effort fell because landlords were monitoring the use of physical capital. The combined effect is an increase in the relative use of wage contracts as mechanisation proceeded, as illustrated in Figure 5.1. To test for the relationship across states between mechanisation and wage 144

Table 5.2 Tenants, sharecroppers and real wage expenditure

LEE J.ALSTON

Figure 5.1 Increase in the relative use of wage contracts with mechanisation

contracting, I correlated for each state and census year the percentage of cotton mechanically picked with the ratio of wages to tenants plus croppers. The correlation coefficients (for each census year) are: 0.91 (1950); 0.62 (1954); 0.59 (I960); 0.60 (1964); and -0.27 (1970). Except for 1970, the associations between mechanisation and wage labour are high, indicating that mechanisation increased the incentive to contract for wages rather than sharecrop contracts or rentals.10 Mechanisation, along with the accompanying advances in science-based technology, also reduced the incentive for landlords to supply paternalism. There is little direct evidence that paternalism declined, but other pieces of circumstantial evidence suggest that mechanisation led to the demise of paternalism. Because landlords used tenant and sharecrop contracts and paternalism as means to reduce supervision costs, the evidence demonstrating the negative relationship between mechanisation and tenancy and sharecrop contracts suggests that paternalism would have fallen as well. In addition, paternalism rested on the Southern institution of social control, defined as the variety of racially discriminatory laws and practices. Planters acted in the political arena to maintain social control because paternalism was valuable. If paternalism was becoming less valuable with mechanisation, we should observe an erosion of social control where mechanisation proceeded. In an earlier study (Alston 1986) I tested for the relationship 146

INDUSTRIAL ORGANISATION OF COTTON PRODUCTION

between tenancy in the early 1960s and the perception by blacks of race relations in their area. In a regression analysis controlling for other causal factors of perceptions of race relations, I found that tenancy was the only variable strongly associated with perceptions of poor race relations. And finally, before the mechanisation of cotton, Southern Congressmen consistently acted to thwart the expansion of the welfare state, which would have reduced the value of paternalism to workers by giving them a substitute for some of the services that the landlord provided in part to lower monitoring costs (Alston and Ferrie 1985b, 1989, 1993a). As mechanisation reached high levels in the early 1960s, the welfare state expanded despite no apparent decline in the seniority of Southern Congressmen on committees in Congress, the political institution through which Congressmen exercised agenda control (Alston and Ferrie 1993a). This result would be paradoxical if paternalism was still highly valued. Cotton pickers and cotton gins: technological compatibility As we have already seen, part of the delay in adopting mechanical cotton pickers is attributable to the need to have technological compatibility between cotton pickers and cotton gins.11 This is especially true for spindle pickers, which need to be lubricated the result of which is picked cotton that has a higher moisture content than hand-picked cotton. In addition, machines do not have the same discretion as humans, and consequently stems, leaves and other debris get picked as well as the cotton boll. This is known as ‘trashy cotton’. Although some farmers tried to clean their cotton before taking it to the gin, it became obvious by the early 1940s that the least-cost method of cleaning and drying was at the gin. But installing cleaners and dryers was not cheap: Aiken (1973:217) reports that ‘In 1940 a new cotton gin incorporating the latest machinery could be built for approximately $35,000. In 1948 a plant equipped to process machine-harvested cotton, but with the same ginning capacity as in 1940, cost about $100,000.’ Naturally, many gin owners would be reluctant to purchase cleaners and dryers whose use was specific to mechanical pickers, unless they could be assured that growers would adopt mechanical pickers. Growers in turn would be reluctant to purchase a mechanical picker unless they knew that a nearby gin had installed cleaners and dryers; in the early 1950s a two-row cotton harvester cost about $15,000 (Prunty and Aiken 1972:292). Because of the transportation costs of moving pre-ginned cotton, a bilateral monopoly issue arose: mechanical pickers required gins that had cleaners and dryers, and gins could not recover the capital costs of cleaners and dryers unless growers brought them machinepicked cotton.12 The percentage of gins that had cleaners and dryers in 1945 was considerably higher than was the percentage of cotton that was mechanically picked (see Table 5.3). This suggests that in some areas ginners acted before they had 147

LEE J.ALSTON

the assurance that growers would adopt pickers, though it is interesting to note that in the more marginal cotton-growing South Atlantic states the adoption of new technology by ginners was less. Despite adoption rates by ginners being higher than adoption rates for growers, there is a strong correlation across states in the percentage of cotton mechanically harvested and the percentage of gins with airline cleaners or overhead cleaners: 0.66 for airline cleaners and 0.53 for overhead cleaners. Picking cotton by machine also created a bottleneck problem at gins. When cotton was hand-picked, the harvest season extended from early September through to early December, allowing gin owners to process cotton as it was picked. Mechanical pickers had greater capacity and compressed the majority of the cotton harvest into several weeks in October. The answer to this dilemma was either to store seed-cotton or increase the capacity of gins. Because machine-picked cotton is wetter and trashier than hand-picked cotton, storage was not practical. The result was that gin capacity had to be increased, leading to further capital costs and to the further need for gin owners to be assured that growers would adopt a mechanical picker before investing in cleaners, dryers and greater capacity. In Table 5.4, I present data on the percentage of gins with high capacity (greater than 36 bales per day). The data suggest the same story as that told by Table 5.3: ginners added capacity ahead of growers adopting pickers. Nevertheless, there is still a high correlation across states between highcapacity gins and the adoption of mechanical pickers: in this instance the figure is 0.71. As with cleaners, the Eastern cotton states lagged behind the Western states. With the advent of mechanical cotton pickers the technological interface

Table 5.3 Percentage of gins with cleaners and dryers, 1945

148

INDUSTRIAL ORGANISATION OF COTTON PRODUCTION

Table 5.4 Percentage of gins with daily capacity greater than 36 bales, 1945

between harvesting and ginning became blurred. There are four possible scenarios that we might expect: (a) independently of one another, the adoption by gin owners of cleaners and dryers, and by cotton growers of mechanical pickers; (b) long-term contracting between ginners and growers; (c) integration of harvesting and ginning; (d) exit from growing cotton and ginning cotton. Before assessing the responses of growers and ginners to technological change, it is necessary to describe the overall macro conditions in cotton production. Growing cotton was a declining activity in the post-World War Two period, largely due to the reinstatement of federal crop allotments in 1950, but also because other activities became relatively more profitable. For the South as a whole, the acreage devoted to cotton declined, though output nevertheless increased because yields increased dramatically. The declines were in no way uniform, and in some areas acreage expanded. The decline in acreage was greatest in the Piedmont, a region stretching from North Carolina to Georgia, and secondarily in the Black Waxy Prairies of Eastern Texas. Acreage and output expanded in the High Plains area of Texas and Oklahoma, the Inner Coastal Plains stretching in an arc from North Carolina to Alabama, parts of the Mississippi Valley, and the South Texas Coastal Plains. Unfortunately, data on the degree of vertical integration and long-term contracting do not exist. We do know from Table 5.3 that, in those states that adopted pickers, gin owners were more likely to have installed cleaners and dryers. The question was whether they acted independently or through long-term contracts or vertical integration. I have no evidence that planters acted formally in concert with gin owners. My intuition is that in those areas where acreage expanded because of dominant soil conditions, i.e. soil that made certain regions the least-cost producers, the obvious choice for growers 149

LEE J.ALSTON

with sufficient acreage was to purchase pickers and the obvious choice for gin owners was to install cleaners and dryers. The assurance needed for both sides to adopt the new technology was the soil itself. For planters whose scale was insufficient to warrant the purchase of pickers, other alternatives to exit were available: Usually these people would ‘swap work’ with neighbours who possessed the machine, or pay for the services of neighbouring machines and operators on a custom basis. Custom services available from the nearby cotton ginner are used extensively in most cotton regions, and have been particularly important in the Piedmont (Prunty and Aiken 1972:292). The situation was different in regions where soil conditions were marginal, such as in the Piedmont region or the plains of Eastern Texas. In these regions, unless some kind of assurance was given through long-term contracts, we should expect to see vertical integration or exit. We know that a considerable amount of exit occurred. More interesting is to observe what prompted the continuation of cotton production within the regions of overall declining acreage and output. When Merle C.Prunty and Charles S. Aiken undertook such an investigation for the Piedmont region (Prunty and Aiken 1972), they found islands of cotton production within a larger region of declining acreage and output. To assess what accounted for the islands, they interviewed the owners of ninety-eight cotton gins. They found considerable vertical integration: seventy-six of the ninety-eight gin owners also grew cotton. We do not know in what percentage of the cases the gins integrated backward or aggressive growers integrated forward, but it appears that the overwhelming majority of integration was backward from the gin to growing (Prunty and Aiken 1972:303–4). Other gin owners provided cotton pickers for planters on a custom basis. Ginners also contracted with planters: ‘The ginner is central to the whole cotton effort in the one-variety cotton community. In order to obtain the cotton and seed which they promote, ginners tie farmers to their plants by supply and ginning contracts arranged each spring (Prunty and Aiken 1972:303, emphasis added). Exit, though, appears to have been the dominant choice, in part brought about by the uncertainty arising from the need to have technological compatibility. The decline in cotton acreage for the period 1944 to 1964 in the Piedmont area was greater than 63 per cent, twice as great as the average for the Southeast (Prunty and Aiken 1972:285–7). Prunty and Aiken argue that the critical period for the survival of the Piedmont region as a major cotton-producing area was the 1950s. In that decade Piedmont growers lagged in adopting mechanical pickers and Piedmont ginners lagged in installing cleaners and dryers: Perhaps the Piedmont’s small growers would have been foolish to move to the modern production technology (mechanical harvesters) at a time when the ginners in the region could not adequately 150

INDUSTRIAL ORGANISATION OF COTTON PRODUCTION

process the machine-harvested product. The ginners, too, exhibited surprising technologic lag. Their failure to recognize the new production technology as essential to survival of cotton growing produces a fascinating speculation: had they equipped their gins to handle machine-produced cotton and had they engaged in their 1967 level of promotional and grower-supporting activities, most of the regional cotton complex might have survived. (Prunty and Aiken 1972:304–5)

Concluding remarks The mechanisation of the cotton economy of the US South prompted political, social and economic changes in the South. Recently, scholars have produced much good work on the overall impact of cotton mechanisation (Fite 1984; Kirby 1987; Schulman 1991; and Wright 1986). Less attention has been paid to the micro effects of mechanisation on the industrial organisation of the farms remaining in cultivation of cotton. In this chapter, I have used the theory of transaction costs to develop hypotheses about how mechanisation affects the costs of labour contracting between cotton growers and workers, and the farm/market boundary. Mechanisation reduced the supervision cost of monitoring labour effort in two ways: (a) mechanisation displaced thousands of farm labourers, which in turn raised the opportunity costs of stinting on effort across all types of contracts; and (b) mechanisation standardised the value of the marginal productivity of labour, leading to an increased ability to judge labour inputs by looking at labour output. The resulting decrease in supervision was expected to lead to an increase in the relative use of wage contracts and a decline in the prevalence of paternalism. The evidence presented is consistent with these hypotheses. The mechanisation of cotton required changes in the extant gins. Cotton picked by machines is wetter and ‘trashier’ than hand-picked cotton and, as such, gins had to have installed cleaners and dryers in order for growers to adopt mechanical pickers. The cleaners and dryers were expensive additions to gins, and ginners would be reluctant to purchase the equipment unless they could be assured that growers would purchase mechanical pickers. The result was a bilateral asset specificity problem which should lead either to integration of ginning and harvesting, or to exit from the industry, unless some other type of bonding could be found. The evidence presented indicated that in the prime cotton growing regions the suitability of the soil was enough assurance for both sides to adopt the new technology. In other areas, most notably the Piedmont region, considerable exit occurred, along with many ginners integrating backward into growing cotton or purchasing mechanical pickers and either leasing them or harvesting on a custom basis.

151

LEE J.ALSTON

Notes 1 Paper prepared for the 1994 International Economic History Meetings, Milan, Italy. I thank Ann Carlos, Joe Ferrie, S.R.H.Jones and Bernardo Mueller for comments, and Bernardo Mueller for research assistance. 2 To my knowledge, the only authors to analyse the issue of technological compatibility between mechanical cotton harvesters and cotton gins are Charles Aiken and Merle Prunty. See Aiken (1973, 1978) and Prunty and Aiken (1972). For examples of how technology affects the transaction costs of firms, see Robertson and Alston (1992). 3 For this issue, see Alston and Ferrie (1985a, 1989 and 1993b). 4 For a discussion of the theoretical and empirical functioning of efficiency wages and compensating payments, see Alston and Fairris (1994). 5 For my description of cotton ginning I rely most heavily on Aiken (1973). 6 The adoption of the mechanical cotton picker is a story that has been well-told elsewhere (see, e.g., Fite (1984), Maier (1969), Street (1957) or Whatley (1985). 7 This section draws heavily from Alston (1981 and 1985). 8 Part of the relative increase in wage expenditures may be attributable to rising real wages, but this alone could account only for a small fraction of the increase. 9 In labour economics, this is known as an efficiency wage. For the use of efficiency wages in the Southern agricultural labour market, see Alston and Ferrie (1993a). 10 The result for 1970 is attributable to the relatively low variation across states in 1970 in percentages of mechanically-picked cotton, and to the impact of Texas and Oklahoma where tenants tended to be more independent and not part of larger plantations. To test for this possibility, I excluded Texas and Oklahoma from the data set for 1970. The resulting correlation for the remaining states changed to 0.21. 11 For this discussion I rely heavily on Aiken (1973). 12 The issue of asset specificity leading to bilateral monopoly is most clearly put forward in Williamson (1985).

References Aiken, C.S. (1973) ‘The Evolution of Cotton Ginning in the Southeastern United States’, Geographical Review, (April): 196–224. ——(1978) ‘The Decline of Sharecropping in the Lower Mississippi River Valley’, Geoscience and Man, 19 (June): 151–65. Alston, L.J. (1981) ‘Tenure Choice in Southern Agriculture, 1930–1960’, Explorations in Economic History, 18 (July): 211–32. ——(1985) Costs of Contracting and the Decline of Tenancy in the South, 1930– 1960, New York: Garland. (1986) ‘Race Etiquette in the South: The Role of Tenancy’, in P.Uselding (ed.) Research in Economic History, 10, Greenwich, Ct: JAI Press: 193–205. Alston, L.J. and Fairris, D. (1994) ‘Wages and the Intensity of Labor Effort: Efficiency Wages versus Compensating Payments’, Southern Economic Journal (July): 149– 60. Alston, L.J. and Ferrie, J.P. (1985a) ‘Labor Costs, Paternalism, and Loyalty in Southern Agriculture: A Constraint on the Growth of the Welfare State’, Journal of Economic History, 45 (March): 95–117. 152

INDUSTRIAL ORGANISATION OF COTTON PRODUCTION

——(1985b) ‘Resisting the Welfare State: Southern Opposition to the Farm Security Administration’, in R.Higgs (ed.) The Emergence of Modern Political Economy, Greenwich, Ct: JAI Press: 83–120. ——(1989) ‘Social Control and Labor Relations in the American South Before the Mechanization of the Cotton Harvest in the 1950s’, Journal of Institutional and Theoretical Economics, 145 (March): 133–57. ——(1993a) ‘Paternalism in Agricultural Labor Contracts in the U.S. South: Implications for the Growth of the Welfare State’, American Economic Review, 83 (September): 852–76. ——(1993b) ‘The Bracero Program and Farm Labor Legislation in World War II’, in G.T.Mills and H.Rockoff (eds) The Sinews of War, Ames: Iowa State University Press: 129–49. Alston, L.J. and Higgs, R. (1982) ‘Contractual Mix in Southern Agriculture Since the Civil War: Facts, Hypotheses, and Tests’, Journal of Economic History, 42 (June): 391–424. Fite, G.C. (1984) Cotton Fields No More: Southern Agriculture, 1865–1980, Lexington: University Press of Kentucky. Kirby, J.T. (1987) Rural Worlds Lost: The American South, 1920–1960, Baton Rouge: Louisiana State University Press. Maier, F. (1969) ‘An Economic Analysis of the Adoption of the Mechanical Cotton Picker’, PhD dissertation, University of Chicago. Prunty, M.C. and Aiken, C.S. (1972) ‘The Demise of the Piedmont Cotton Region’, Annals of the Association of American Geographers, 62 (June): 283–306. Robertson, P.L. and Alston, L.J. (1992) ‘Technological Choice and the Organization of Work in Capitalist Firms’, Economic History Review, 45 (May): 330–49. Schulman, B.J. (1991) From Cotton Belt to Sunbelt: Federal Policy, Economic Development, and the Transformation of the South, 1938–1980, New York: Oxford University Press. Street, James H. (1957) The New Revolution in the Cotton Economy, Chapel Hill: University of North Carolina Press. U.S. Department of Agriculture (1974) ‘Statistics on Cotton and Related Data, 1920–73’, Statistical Bulletin 535, Washington DC: Government Printing Office. U.S. Department of Commerce, Bureau of the Census (1946) Cotton Ginning Machinery and Equipment in the United States: 1945, Washington DC: Government Printing Office. ——(1952) United States Census of Agriculture: 1950, I, Washington DC: Government Printing Office. ——(1956) United States Census of Agriculture: 1954, I, Washington DC: Government Printing Office. ——(1961) United States Census of Agriculture: 1959, I, Washington DC: Government Printing Office. ——(1966) United States Census of Agriculture: 1964, I, Washington DC: Government Printing Office. ——(1971) United States Census of Agriculture: 1969, I, Washington DC: Government Printing Office. ——(1972) Historical Statistics of the United States, Colonial Times to 1970, Washington DC: Government Printing Office. Whatley, W. (1985) ‘A History of Mechanization in the Cotton South: The Institutional Hypothesis’, Quarterly Journal of Economics 4 (November): 1191–215.

153

LEE J.ALSTON

Williamson, O. (1985) The Economic Institutions of Capitalism, New York: Free Press. Wright, G. (1986) Old South, New South: Revolutions in the Southern Economy Since the Civil War, New York: Basic Books.

154

6

THE MAINTENANCE OF PROFESSIONAL AUTHORITY The case of physicians and hospitals in the United States1 Deborah A.Savage and Paul L.Robertson

Introduction Most of the models in this book deal with hierarchical organisations in which authority is clearly defined and, for the most part, centralised. Typically, hierarchical organisations undertake routine operations in which the bulk of the workforce is expected to perform the same activities or groups of activities repeatedly. One of the main goals of these organisations is therefore to restrain people from exercising their individual judgement, since deviations from routine patterns may disrupt the entire flow of work. Hierarchy and centralisation are inappropriate, however, when there is a substantial amount of uncertainty present in the production process. If producers must frequently make complex choices, even instruction books may be inadequate and the use of a rule-based Weberian bureaucracy (Weber 1946, 1947) becomes impractical. Professions provide an example of an organisation whose work is highly uncertain and contingent, requiring professional practitioners to rely heavily on their individual skill and judgement within the norms of accepted practice for their particular professions. But professionals must often draw on resources that are concentrated in institutions that they do not control. While these organisations may be intended specifically to assist professionals in the independent use of their skill and judgement, they nevertheless pose problems because they are most often not owned by these professionals, and their administration is not entirely under professional control. This creates the potential for conflicts between independent practitioners, who seek to preserve their authority and autonomy, and the administrators of complementary institutions, who have responsibilities of their own. 155

DEBORAH A.SAVAGE AND PAUL L.ROBERTSON

In this chapter we trace the nature of these conflicts and examine how they may be resolved, using as an example the interaction of hospitals and medical staffs in the United States of America. Hospitals pose a particularly interesting problem because they are not owned by physicians, but, whether non-or for-profit, instead represent a community’s collective interests in health care. Ownership notwithstanding, physicians have a major effect on a hospital’s resource acquisition and use. As financial resources for health care become tighter, debate about how to reorganise the relationship between hospitals and physicians in the US becomes more heated. This problem of how to organise hospital-based health care has been studied from a variety of academic perspectives. As we will show, all of these approaches come down to designing a production process that minimises the need for ex post monitoring of quality. We will explain how the medical profession is organised as a network, how it has evolved mechanisms over time to monitor quality and resource use while maintaining professional authority, and how hospitals have evolved to accommodate professional self-regulation as their major monitor of quality. However, current privately-sponsored health-care reforms such as managed care, as well as various government proposals, have shifted the emphasis in monitoring by giving an increasing priority to resource use, even though quality of care remains an important consideration. These reforms are implicitly based on the assumption that physicians are perfectly substitutable and independent labourers who, through the employment relationship or contracts with strong financial incentives, will be able to change the way that they practise. The hoped-for result is that health care will be produced differently, but will nevertheless be produced with the same or better quality by dint of external monitoring, ‘clinical pathways’, outcome studies, and other techniques. If this is indeed the goal of health-care reform, then we must be careful to evaluate any hospital governance system as either a substitute for or a complement to the existing system of professional production. In this chapter we examine the growing tension between monitoring usage of hospital-based resources and the maintenance of profession authority and autonomy in the medical profession in the US since 1918. We show that, from the end of World War One, the professional behaviour of physicians was monitored on the Joint Commission Model in which boards composed of local medical practitioners sought to maintain professional standards and allocate local resources but did not inquire into cost factors at the system level. More recently, however, the use of physical resources by physicians has also been monitored in ways that may impinge on their authority and autonomy. We begin by describing the connection between individual and collective autonomy and authority within professions as network organisations, and then trace the development of monitoring procedures and discuss their impact on the ability of physicians to use their individual judgement in treating patients. 156

THE MAINTENANCE OF PROFESSIONAL AUTHORITY

The professional mode of production One of the ways to differentiate among economic institutions is to identify the kinds of knowledge problems that each solves well, and to study its strengths and weaknesses in structuring production and exchange. As Jensen and Meckling (1992:251) point out, economic organisation must solve two different kinds of problems: ‘the rights assignment problem (determining who should exercise a decision right), and the control or agency problem (how to ensure that self-interested decision agents exercise their rights in a way that contributes to the organizational objective)’. These two problems arise because of the need for decentralisation implied by the specialisation of knowledge in a complex production process. As suggested in the modern literature on the capabilities of organisations (Teece and Pisano 1994; Langlois and Robertson 1995), productive knowledge is not merely idiosyncratic but often sticky, since it consists principally in repertoires of tacitly understood routines and procedures. These routines and procedures are the capabilities of the organisation (Richardson 1972; Nelson and Winter 1982). Efficiency demands that the appropriate knowledge find its way into the hands of those making decisions. There are basically two ways to ensure such a ‘collocation’ of knowledge and decision-making: ‘One is by moving the knowledge to those with the decision rights; the other is by moving the decision rights to those with the knowledge’ (Jensen and Meckling 1992:253). Markets (in the widest sense of the term) take the latter approach. The Coase theorem suggests that, so long as rights are well defined and alienable, decision rights will tend to end up in the possession of those whose specialised knowledge can make the most of them. This also solves the agency problem, since the alienability of a right means that market prices can measure its value, which in turn creates an incentive for the owner to maximise value by using the right appropriately. But there are also potential costs to such extreme decentralisation, costs that arise in the interactions among the decentralised holders of rights. These might include the familiar sorts of transaction costs arising from moral hazard and asset specificity (Alchian and Woodward 1988). More interestingly, however, they may arise from the need to bring otherwise decentralised knowledge together and to coordinate it (Milgrom and Roberts 1992: ch. 4; Kogut and Zander 1992), especially in circumstances involving learning and the generation of new productive knowledge (Langlois and Robertson 1995). One alternative is to organise production under common ownership in order to gain the benefits of synergies and the integration of knowledge, albeit at the cost of imperfect collocation of knowledge and decision-making. Firms are strong at executing known routines in which complex management teams oversee large-scale production and distribution processes of a repeated, consistent and planned character. Such routines tend to be measurable at various stages of production, and so are suited especially well to formal

157

DEBORAH A.SAVAGE AND PAUL L.ROBERTSON

monitoring schemes, including documentation, accounting trails, and the supervision of employees (Barzel 1982, 1987). As a result, hierarchical governance lends itself well to formal, rule-based monitoring schemes of the sort long ago discussed by Max Weber (1946, 1947). Such organisations reflect what Mintzberg (1979) calls Machine Bureaucracy. But when production involves uncertainty and requires highly flexible adjustment of routines to tasks, then the benefits of knowledge synergy in a hierarchy come at a cost that is large in terms both of agency and of the poor collocation of knowledge. Professional production is very much a sphere of activity in which uncertainty and task variability are important, and in which rigidly pre-programmed routines work poorly. As Arthur Stinchcombe (1990: ch. 2) so nicely puts it, professionals are information processing systems that must wield and apply a wide repertoire of routines to fit widely varying concrete circumstances. Fortunately, as researchers are coming increasingly to notice, markets and hierarchies do not exhaust the types of organisational form available. In particular, networks (Powell 1990; Nohria and Eccles 1992) are coming to be recognised as a distinct form with unique knowledge-flow and transactioncost properties. It is our claim that professions are instances of the network form. Although economists have written extensively about individual professions, they have not produced a theory of professions as distinct economic institutions. As a working definition of professional networks, we can think of them as communities of independent practitioners who share a core competence, and who form strategic alliances across ownership boundaries.2 Professional networks identify core competences, build capabilities, share them across the membership, and internalise knowledge flows without integrating ownership. Each professional’s decisions and abilities are constrained by the capabilities of the network as a whole, as well as by other institutions, and their decisions must be implemented within the system. For example, physicians have developed core competences in diagnosis and the design of treatment regimes, but their ability to gain maximum benefit from these capabilities has been constrained at various times by the availability of supporting technologies (surgery without anaesthesia being an unpleasant experience) and by the absence of effective health-financing schemes like insurance. The current health care ‘crisis’ in the US is in part another such episode, in which changes in liability, financing and hospital ownership have outstripped the profession’s ability to respond strategically. While individual practitioners remain independent, they make a longterm commitment of their substantial human capital to a ‘hubless’, indeed a bossless, network. A network’s coordinating structure is horizontal, an organisation comprising equals. Without the exchange of cash payments, members willingly exchange information and technology and collaborate in 158

THE MAINTENANCE OF PROFESSIONAL AUTHORITY

production—that is, share routines—without authoritarian supervision, and without integrating external management functions into their day-to-day operations. In fact, network members remain competitors across many dimensions, attempting to take advantage of their capabilities more quickly and ably than others. So professions operate a complicated production strategy, furthering the interests of individual members as well as the interests of the network as a whole. Networks thus provide a vital part of the framework in which professionals operate, but they by no means define the entire environment. As we saw, for example, a professional’s use of his or her repertoire of routines is highly judgemental. To be sure, all members of a particular profession operate under a shared regime. Lawyers are constrained by the cumulative precedents of previous cases, most of which were decided long before the current generation entered the profession. This is true as well for physicians, whose day-to-day decisions are affected by existing treatments administered to patients by other physicians. Nevertheless, professional knowledge is fungible to a variety of tasks. The exact routine employed by a professional is unique to each case: non-routine routines, if you will. In this sense, it is fruitless to try to define narrowly what a professional does, since professionals can apply their routines across a wide array of job descriptions and play several different roles simultaneously. A further characteristic of professionals is that they have multiple responsibilities. They are responsible to their larger professions for the maintenance of set standards, and they often also are bound by law to behave in certain ways. Their primary responsibility, however, is generally to their clients. To take an obvious example, lawyers, priests, physicians and many other types of professional may legally claim the right to withhold information communicated to them by clients, parishioners or patients. And when clients are displeased with the service they have received, they may in most cases lodge complaints only against individual practitioners and not against a profession as a whole. In fact, it is often professional bodies that are called upon to investigate the conduct of an individual practitioner when a complaint is registered. Figure 6.1 summarises the internal and external relationships of professional practitioners. The profession as a whole is represented by the circle, in which all members are connected to each other through ties of training, knowledge, the setting of standards, and membership in bodies established by the profession to express its collective views as well as to enforce behaviour when there have been deviations. It is as individuals, however, that professionals are connected to their clients. Each professional trades on the basis of his or her own judgement, skill and reputation when dealing with clients, and clients usually have no right of redress against other members of a profession when one practitioner fails to meet accepted standards.3

159

DEBORAH A.SAVAGE AND PAUL L.ROBERTSON

Figure 6.1 The internal and external relationships of a profession

Standards are one mechanism used to coordinate independently executed routines while preserving professional judgement. When describing ‘the Professional Bureaucracy’, Mintzberg (1979:348) lists ‘[s]tandardization of skills’ as the ‘Prime Coordinating Mechanism’. Members of a given profession possess common skills that allow them to function quickly and accurately4 in a variety of contexts. Properly-qualified lawyers should be able to perform well in any court that employs the system of law in which they have been trained, and similar flexibility should be characteristic of pharmacists, physicians, accountants and other professionals.5 Mintzberg (1979) discusses the relationship between professionals and their environments at some length. He contends that ‘the structure of [a Professional Bureaucracy] is essentially bureaucratic, its coordination—like that of the Machine Bureaucracy—achieved by design, by standards that predetermine what is to be done’. But, he continues, [w]hereas the Machine Bureaucracy generates its own standardsits technostructure designing the work standards for its operators and its line managers enforcing them—the standards of the Professional Bureaucracy originate largely outside its own structure, in the selfgoverning associations its operators join with their colleagues from other Professional Bureaucracies. (Mintzberg 1979:351) Networks thus provide a vital part of the framework in which professionals operate, but they by no means define the entire environment. Professionals 160

THE MAINTENANCE OF PROFESSIONAL AUTHORITY

often work in institutions such as courts of law or hospitals that professionals do not own or manage, either collectively or individually—and that do not, in turn, own or manage them. In this setting, an important issue for professionals is to maintain professional authority and autonomy. Autonomy means that no one except another professional, the network’s unspecified representative, can challenge the day-to-day decisions of a professional. It legitimises judgement without managerial oversight. In medicine, as we will see, the Joint Commission Model ensures that medical practice outside of the hospital setting is almost entirely at the individual’s unfettered discretion: patients are free to choose physicians, and physicians are free to choose practice settings. Physicians are responsible for a patient’s care, and cannot point fingers at others for errors of omission or commission. Inside the hospital, however, physician performance is continually monitored and assessed, as is usage of common resources. Networks prepare individuals to use their judgement independently, as well as to share routines in a complementary setting. Autonomy represents the formal recognition of their individual responsibility to do so. The mechanisms that make autonomy operational are designed to select potential entrants, and to convey to them the theory and practice of shared routines and competences. Included in these routines are incentives for individual and collective self-restraint, which help both members and non-members to perceive professional authority as legitimate. The multi-locational nature of practice increases the scope and incentive for individuals to develop innovative uses for existing routines, invent new ones, and share information about the external environment and the strengths and weaknesses of existing and emerging capabilities and strategies. Authority emphasises that professionals possess command capabilities not available to economic agents outside the professions. In this paper we are obviously most interested in command over resources that the profession does not own, manage or even make payments for the use of. Here, we can think of an attorney’s use of court time, or—the case at hand—physician control within a hospital. This is further complicated by the problem that hospital employees are explicitly not supervised by physicians, but the decisions of physicians effectively allocate these as well as non-human assets. Until the advent of aggressive managed care, physicians decided which patients were eligible to enter hospitals for treatment. Such authority enables production, but does not mean that professionals have the ability to force individuals, such as clients or patients, into specific actions based on their opinion and advice. Quite the opposite, physician authority effectively confers autonomy on patients. Coercive powers commonly belong only to the state, and actually complicate rather than enhance professional productive ability. In a hierarchical structure, authority is delegated from the top down, from owners through managers to employees. In contrast, networks delegate 161

DEBORAH A.SAVAGE AND PAUL L.ROBERTSON

authority to individuals and institutions, and have no central authority structure. Eliminating the management of individuals in medicine after training avoids expending resources on monitoring the opportunistic behaviour of managers. Defining ownership in terms of the right to participate in shared routines returns us to the concepts of authority and autonomy. In a sense, the autonomy and authority that the network grants to professionals are the analogue to ownership in a market: they are a kind of quasi-ownership. The professional, who possesses the relevant localised knowledge, does get to exercise decisionmaking; but the right to do so is not alienable, and the exercise of decisionmaking is always circumscribed by the constraints of network participation.6 Networks are complex mechanisms that put discretionary differences among practitioners to their best use. Each professional interacts with a large but distinct set of the network’s capabilities, which enables the individual practitioner to see parts of the big picture as well as the small. Moreover, not all practitioners are equally skilled at using routines. Some accountants are better than others at dealing with the taxation authorities; some architects design buildings that are stronger or more pleasing to the eye; some surgeons have higher success rates in performing particular operations; and some lawyers are better at securing acquittals for their clients. Indeed, judgement and skill are what differentiates one member of a profession from another. This may benefit individual professionals through two channels. The first is internal to a given profession, when other members recognise a person’s intelligence or skill by directing difficult cases in his or her direction. This is especially valuable when, as in the case of many medical specialists, clients are normally referred to a practitioner by other professionals. When there is a realistic choice, responsible physicians do not direct their clients to surgeons with poor reputations. For specialists such as radiologists, pathologists or anaesthesiologists, who are practically anonymous as far as the general public is concerned, reputation within the larger medical profession may be by far their most important asset. Professionals view reputation as a signal of the availability of assets that other professionals can access in the execution of their own routines. Reputations are flexible and informal review mechanisms that are particularly useful in the arena of shared non-owned assets. Evaluation occurs during direct contact, as during consultations, but also indirectly when professionals see the results of one another’s work or simply talk about each other. Given institutional and geographical limitations, it is impossible to know how every practitioner is running his or her own office, but frequent contact gives clear insight into the use of shared resources. This is particularly useful in professions like medicine and academics, where post-entry specialisation implies that only a few other subspecialists will be able to judge the merits of one’s contributions. In place of hierarchical supervision, networks rely on reputation as the basis for peer monitoring. One advantage of peer monitoring is that, unlike 162

THE MAINTENANCE OF PROFESSIONAL AUTHORITY

contracts, it relies on self-interest rather than the availability of third-party enforcement. The need to maintain one’s reputation among one’s peers creates disincentives and sanctions for bad behaviours. Thus, peer monitoring works to help individuals internalise norms that would be difficult to enforce or even define externally. Peers are good monitors, not only because of what they know, but because they are the ones to whom good monitoring matters most. In an organisation without hierarchy or ownership, residual risk is borne by all members. The members want to make sure that shared resources are not mismanaged, whether individual members are keeping up with their fields, and whether their practice style reflects well on the network. ‘When agents interact to produce output, they acquire low-cost information about colleagues… Mutual monitoring systems derive their energy from the interests of agents to use internal agent markets or organizations to enhance the value of human capital’ (Fama and Jensen 1983:310). It is possible to think of many ways to monitor physicians in both their office and the hospital setting, and managed care companies are becoming quite innovative in doing so. They attempt to collect data on billable hours, number of patients seen, patient satisfaction, number and types of diagnostic tests ordered, and a host of other factors. But in a network setting, no-one cares about monitoring these measures, because the individual practitioner puts as much or as little effort into private practice as he or she wishes on behalf of patients. This is as close as professions get to the usual story of ‘self-employment’ as a way to reduce monitoring costs. Moreover, it is in the network’s interest to discourage this kind of monitoring because it changes the focus from the network to the individual practitioner. The potential punitive uses of external monitoring would cause individual practitioners to be less likely to share information about their practices, and therefore diminish the value of shared competences, with destructive repercussions for the network, and therefore for patient choice. It is characteristic of professional networks to provide practitioners with up-to-date information about the latest developments in their profession, as occurs at professional meetings. But external monitoring presupposes some proprietary use of information, again lessening the incentive to share.

Professionals and complementary institutions Although the independent practitioners we discuss are ‘bossless’, in the sense that their professional actions are presumed to be correct and they are generally subject to the authority of other professionals or outsiders only when they are accused of misconduct, this does not imply that they are not associated with more formal, hierarchical organisations. For one thing, even bossless professionals may require assistants who must be supervised. Some

163

DEBORAH A.SAVAGE AND PAUL L.ROBERTSON

of these assistants perform clerical jobs, others (such as paralegals) act as professional auxiliaries, and still others are themselves professionals who are still in training or whose activities involve providing professional assistance. For example, trials require judges as well as lawyers, and in the United Kingdom a senior barrister often appears in court with a junior. Similarly, fully qualified specialists provide supervision and guidance for medical residents who assist them in order to develop professional skills. Second, professional practice may require the use of buildings or items of capital equipment that are too expensive to be owned by individual practitioners. Even when the legal system relied primarily on private prosecutions, inns and courts of law were still owned either collectively or publicly. As surgeons typically devote only a few hours of each week to performing operations, it would be wasteful for all surgeons to have their own operating and recovery rooms, full-time support staff and so on. Therefore, these professionals, who are bossless in the use of their judgement and skills, must nevertheless frequently conduct their activities in conjunction with complementary institutions that are owned and controlled by others. And by their nature, these complementary institutions need a degree of central control to function smoothly. Within US hospitals, much of the medical staffs clout stems from the hospital’s dependence on physician referrals as the most important source of patients. From the perspective of hospitals and third-party payers, which patients physicians choose to admit and which hospital resources are committed to an individual patient’s care largely determines the financial viability of the institution. Hospitals hope that physicians will bring them lots of patients, but not those who are very sick. Increasingly, hospitals contract directly with several managed-care companies for in-patient and ambulatory care. Managedcare companies decline to associate themselves with hospitals whose medical staff is too ‘expensive’, as measured by intensity of resource use. This means that hospitals may lose access to entire blocks of patients, since subscribers are forbidden or at least penalised for going to facilities outside their plans. Thus, both competitive pressures and the need to manage resources more generally mean that non-physician managers—Boards of Trustees and their paid representatives—require a balancing influence over the use of the resources they also do not own, but have the obligation to direct and protect. On the other side, physicians are dependent on hospitals for access to technologies, facilities, specialities and ongoing training, which are important determinants of their ability to diagnose and treat patients, and therefore to attract and keep a patient base. Because the use and quality of these inputs is so important to them, and because they are in a good position to judge the immediate quality and usefulness of these technologies and other inputs, physicians require some influence over the use of resources that they do not own. They, too, must meet the requirements of managed-care companies or lose patients who join these plans. This puts them in a position of competition 164

THE MAINTENANCE OF PROFESSIONAL AUTHORITY

within the medical staff for resources and patients. But for all of the reasons described above, professions have neither the capabilities nor the interest in coordinating day-to-day supporting services, and are quite happy not to expend resources playing at the administration game. In the next section of the paper, we provide some historical context for the current organisation of hospitals and medical staffs in the USA. We show how the Joint Commission Model has evolved as an interface between autonomous, competing physicians and the administration of hospitals, in the process changing from an institution designed to monitor the quality of medical practice from within the profession to one that, even if under duress, is now called upon by external authorities to pass judgement on the way in which physicians use the resources of hospitals and other complementary institutions. As a consequence, the focus has changed from affirming the responsibility of physicians to individual patients to monitoring the use of resources paid for by patients as a body, as represented by insurance groups and the government.

The Joint Commission Model In 1918, the American College of Surgeons began a process called the Hospital Standardization Program (Roberts et al. 1987). Its founders recognised that surgical medicine was developing quickly, but unevenly. Its modest goal was to take the small step of improving surgical practice through standardisation—not, as one might think, of private or even hospital practice, but through the creation of a system of uniform medical records. With these open records, they hoped to disseminate information about, and to evaluate the procedures and methods of, their fellow surgeons. They instituted a policy of regular meetings at which clinical experiences were reviewed and interpreted. The process enabled them to educate themselves about the safety and efficacy of competing procedures, and to decide which to include in surgical training and continued education. Until this time, medical staff self-governance had consisted largely in the process of deciding who should be allowed to practise at the hospital. Breaking with this tradition, the College of Surgeons—the speciality most dependent on emerging hospital capabilities—recommended that hospitals adopt an open, but defined, medical staff model. They recognised that their future lay not in keeping fellow surgeons out, but in monitoring the quality and ensuring the cooperation of those who shared hospital resources. Closed hospitals were a form of monitoring that clearly would not serve medicine well in an era of increased physician dependence on hospitals and on emerging thirdparty payment schemes that provided patients with more choice of physicians. An open model came to make more sense. At this time, they also recommended

165

DEBORAH A.SAVAGE AND PAUL L.ROBERTSON

abolishing the fee-splitting system that was an integral part of the closed-staff model. So, while there is a body of literature in health policy across many disciplines that focuses on the ‘monopoly’ aspects of medicine, it is clear that under the original conception of medical staff self-governance, competition -real rivalry— was an important goal. Physicians did not design this system as a way to control hospitals. Instead, the system was a recognition of the fact that, by World War One, any physician who hoped to have a respected practice needed to establish an ongoing relationship with at least one hospital, and, for competitive reasons, more if possible (Rosenberg 1995). And, of course, physicians with choice over institution represent patients with choice as well. By World War Two, technology and financing were concentrated in hospitals, so doctors were increasingly dependent upon them. This is particularly true to the extent that certificate-of-need regulations made it increasingly more difficult to purchase equipment or set up competing private laboratories apart from hospitals. That is, when faced with the decision of whether to allow physicians to open private facilities at their own financial risk in competition with hospitals, regulators said no. Over time, other specialities adopted similar formats, and in 1951 they merged and incorporated to form the Joint Commission of Medical Accreditation. Although the Joint Commission seemed to think of itself as an educational association, the effect of its programme was to build monitoring into the production process. At least in the abstract, this was not lost on the designers of the Medicare Act (1966). Because Medicare would finance health care, its designers wanted to have quality-assurance mechanisms in place. Since external ex post monitoring seemed impossible, they imposed upon the Joint Commission self-governance structures the additional task of monitoring hospital-based care on their behalf. They required hospitals to have Joint Commission structures in place in order to receive Medicare reimbursement. Other third-party payers followed suit, and the Joint Commission Model became a formal link between external and internal production processes. What had been self-governance was now externally imposed, institutionalised self-regulation. The Health Care Financing Administration (HCFA), which administers Medicare, still struggles with the problem of direct regulation of quality, precisely because there is no one governance structure that leads to optimal quality, however defined. The agency was correct in identifying the Joint Commission as a particularly effective example of such a structure, but incorrect in assuming that redefining its role would not change its processes. One important change was an alteration in the focus of self-governance, away from the quality of local medical staff in the context of managing the joint assets of the professional network, and towards satisfying the national Joint 166

THE MAINTENANCE OF PROFESSIONAL AUTHORITY

Commission regulators. The Joint Commission became an outside agency for accrediting and regulating hospitals (Jost 1983). Not surprisingly, it is held in some disregard by local practitioners, who resent the ‘surveyors’ who show up every three years to assess their procedures, take up their time, range through their records, and make suggestions that, while marginally helpful, seldom lead to significant improvements in the quality of care. The Joint Commission for Accreditation of Health Care Organizations (JCAHO) is currently a private, non-profit structure, governed at the national level by twenty-eight commissioners representing hospitals, nurses, physicians, surgeons, dentists and the public. Through JC procedures, physicians are still presumed to practise self-governance. JC rules specify how each individual hospital is to set up processes by which the medical staff elects its own officers and establishes procedures to carry out credentialing, privileging, reappointment and peer review. The processes by which credentialing, privileging, reappointment and peer review are carried out are important because they help to support and maintain physician authority and autonomy in the context of hospital production. The staff agrees to take it on because monitoring matters most to them. Every physician on the medical staff is eligible to participate, and gets just one vote. Physicians weigh the interests of the hospital and medical staff against the imperatives of their own practice (and patients) when casting it. Cartel-like behaviour rarely emerges, since various specialities compete against each other for resources. Credentialing is the process by which physicians are granted privileges to admit and treat patients in a specific hospital. Joint Commission regulations require that the medical staff be defined, but it must remain open to potential entrants. Individual competence is demonstrated by presentation and review of credentials, including graduation from an accredited medical school, possession or eligibility for a state medical licence, and, recently, the requirement that applicants reveal prior actions taken against them. Credentialing does not require that other physicians observe the practitioner in performance of daily activities. Though it is largely a formality, there is a phasing-in process which contains a probationary period before full privileges are technically awarded. And, emulating the university-based division of medicine into departments, appointment is generally to a department (like internal medicine), although in practice privileges often extend across such boundaries. This is evidence that, contrary to much that is written about medicine, self-regulation is not about erecting barriers to entry at the local level. The credentialing committee cannot refuse to grant privileges because they believe that the local market for, say, obstetrics, is saturated and fear that adding another physician will result in lower incomes for others in the speciality. Private practice is an entrepreneurial activity, and the successes and failures of individual practitioners provides important feedback to the professional 167

DEBORAH A.SAVAGE AND PAUL L.ROBERTSON

network. Currently, however, there is also pressure on credentialing committees to provide economic credentialing, which examines the economic efficiency of a physician’s practice and denies hospital privileges to individual physicians who practise medicine in a way deemed too expensive by managed-care companies (Blum 1991). Joint Commission rules explicitly forbid such evaluations, but hospitals desperate to win managed-care contracts have implemented them anyway. Several aspects of this are significant departures from the original intent of self-governance. First, economic credentialing is not based on the medical staffs incentive to surround themselves with practitioners whose reputations and skills are complementary to their own. Particularly, but not only, when economic credentialing is done across hospital systems, the incentive is to reward uniformity in practice at the low end of the resource-use spectrum, since ‘savings’ will be distributed to physicians who work in low-cost hospitals. This is different from the previous process of physicians conserving shared resources by internal rivalry but not concerning themselves with the level of income of other doctors. Physicians denied privileges on utilisation grounds have no real recourse, and without the ability to admit patients to that hospital effectively lose the right to earn a living from their private practices. The effect on specialists is particularly chilling. In response, some states have enacted ‘any willing provider’ laws, which would require companies to impanel any physician willing to participate under a plan’s guidelines, but this does not address the ratcheting downward in general, or the effect on physician independence in particular. Ironically, it is managed care that erects barriers to entry, and the professional network that opposes it. These companies entice specialists to accept reduced reimbursement rates in exchange for a promise not to sign on other specialists within the geographic area. Second, the current system is the antithesis of the open system that surgeons established in response to the problems caused by closed panels at the turn of the century. When a hospital has to meet certain aggregate cost requirements in order to be named as a provider by a particular company, even doctors who have no wish to participate in that plan are forced to abide by the changing utilisation guidelines and staff cutbacks in the hospital. It means very little to have privileges in a system with exclusive managed-care contracts. Under the original Joint Commission goals, physicians remained competitors because the barriers to getting privileges were low and uniform. Physicians didn’t own the hospital or pool their own capital, so they had no financial incentive to exclude competent providers. They certainly felt responsible for participating in governance mechanisms aimed at rational allocations of scarce resources, but the bottom line, profits or losses, did not come out of their pockets. High consumers of resources were dealt with in formal and informal peer review and in daily competition for resources; but the goal was not to 168

THE MAINTENANCE OF PROFESSIONAL AUTHORITY

save money in order to line one’s own pockets. So the Joint Commission goals of an open panel and no kickbacks has been unravelled in less than thirty years. Most importantly, the JC medical-staff model was not supposed to be the professional version of a union. Current innovations in health-care financing and delivery have put the model in the position of bargaining with hospitals and insurers, but medical staff have neither the capabilities not homogeneous interests to do this. Whereas the amount of remuneration that a physician received had depended on his or her own individual entrepreneurial abilities, boards are now pressured by insurers to design incentives structured to reward or punish individual physicians based on the whole staff’s use of resources. Moreover, the JC medical-staff model was not meant to coordinate hospital finances or non-physician production. Increasingly, however, a hospital’s ability to be named as a plan’s preferred provider depends on physician behaviour, and the distinction between self-governance and responsibility to patients, and the staffs role in hospital governance, is further muddled. The Joint Commission Model does not solve the problem of the relationship between physicians and hospital employees either. The most obvious problem is that of nurses, whose contractual obligation is to follow written hospital policy and obey supervisors who are responsible for schedules, hiring, firing and training, but whose day-to-day responsibilities are to implement the treatment plans of physicians as ordered. Employed physicians in pathology, radiology, emergency medicine, and anaesthesiology are also in increasingly uncomfortable positions. Finally, interns and residents have always been counted as employees under supervision, and have not been included in the JC self-governance structure.

Conclusion Traditionally, the medical profession has used a network organisation to provide for the training and monitoring of practitioners within a decentralised context. The advantage of employing a network rather than a more hierarchical form of organisation is that it has allowed physicians to establish and maintain standards and to disseminate new knowledge with low transaction costs. Equally importantly, networks also permit members of the profession to maintain their individual autonomy and authority and to compete against each other, while preserving the ability to act collectively when necessary. The Joint Commission Model evolved in this context as a way of monitoring individual performance without imposing excessively uniform rules on a profession in which the ability to deal with contingencies is vital. The Model also opened up hospitals to a wider range of physicians, thereby increasing

169

DEBORAH A.SAVAGE AND PAUL L.ROBERTSON

the options available to their patients as well. It was therefore supportive of the network, because it improved the ability of the medical profession to maintain standards while augmenting the scope of opportunities potentially available to practitioners. Most significantly, this was accomplished without impinging on the ability of the network to function as a means of sharing vital information among members. In recent decades, moves sponsored by the Health Care Financing Administration and health insurance groups have altered the operations of the Joint Commission Model. Under the new arrangements, the primary focus has shifted away from the maintenance of health care standards by individual physicians to economic credentialing in order to conserve resources. This has had a number of serious effects on the operations of the network. In particular, rather than function as complementary and open institutions as they have in the past, hospitals have increasing been obliged by insurers and the government to become more exclusive and directive in administering requests by physicians on behalf of their patients, in the process compromising both the autonomy and the authority of the profession. Furthermore, the new balance of power has undermined the functioning of the professional network by destroying some of the premises on which the exchange of information within the network has been based. As a consequence, the networks may no longer be as able as in the past to perform their training and monitoring roles.

Notes 1 Deborah Savage gratefully acknowledges the support of the Robert Wood Johnson Foundation. Opinions expressed are those of the authors and not necessarily those of the Foundation. 2 Much of the argument on the next few pages is explained in more detail in Savage (1993). 3 However, when the practitioner is a salaried employee of, or works within, a professional institution—as is frequently true of teachers (schools), accountants (accounting firms) or nurses (hospitals)—clients may sue the institution if they feel that they have not been treated properly by an individual associated practitioner. 4 Mintzberg’s definition of a professional bureaucracy combines the characteristics of both professions and the various ancillary institutions which assist professionals in their work—for example, surgeons and hospitals or lawyers and law courts. Although we insist strongly that professions and the ancillary bodies are distinct institutions, his comments on learning, training and the standardisation of skills apply strongly to professions per se 5 Notice that this is not what is meant by standardisation when the term is used by hierarchical supervisors of managed-care organisations. As we suggest later in the chapter, in managed-care organisations standardisation means removing professional judgement and replacing it with short-run least-cost clinical pathways. 6 Of course, ownership always involves a ‘bundle’ of rights, some of which may be alienable. Thus doctors and lawyers in private practice may have conventional ownership rights over some assets like computers, office buildings, smaller pieces 170

THE MAINTENANCE OF PROFESSIONAL AUTHORITY

of equipment, and even, as Grossman and Hart (1986) suggest in a relevant context, a list of clients. Moreover, professionals own their own human capital. But the value of the owned human and physical capital is highly sensitive to participation in the network; and the right to such participation is not alienable and reflects an ongoing reciprocal relationship. This is in part what distinguishes membership in a profession from participation in a system of, say, fast-food franchises or cooperative hardware stores. The incomes of individual hardware stores depends on their participation in the cooperative, which provides wholesale services, advertising, etc. (Hansmann 1988). But the right to participate is fully alienable.

References Alchian, A.A. and Woodward, S. (1988) ‘The Firm Is Dead; Long Live the Firm: A Review of Oliver E.Williamson’s The Economic Institutions of: Capitalism’, Journal of Economic Literature 26(1):65–79. Barzel, Y. (1982) ‘Measurement Costs and the Organization of Markets’, Journal of Law and Economics 25:27–48. ——(1987) ‘The Entrepreneur’s Reward for Self-Policing’, Economic Inquiry 25: 103– 16. Blum, J.D. (1991) ‘Economic Credentialing: A New Twist in Hospital Appraisal Processes’, Journal of Legal Medicine 12:427–76. Burns, L.R. and Wholey, D.R. (1993) ‘Adoption and Abandonment of Matrix Management Programs: Effects of Organizational Characteristics and Interorganizational Networks’, Academy of Management Journal 36(1):106–38. Fama, E.F. and Jensen, M.C. (1983) ‘The Separation of Ownership and Control’, Journal of Law and Economics 26(2):301–27. Grossman, S. and Hart, O.D. (1986) ‘The Costs and Benefits of Ownership: A Theory of Vertical and Lateral Integration’, Journal of Political Economy 94: 691–719. Halpern, S.A. (1992) ‘Dynamics of Professional Control: Internal Coalitions and Crossprofessional Boundaries’, American Journal of Sociology 97(4):994–1021. Hansmann, H. (1988) ‘Ownership of the Firm’, Journal of Law, Economics and Organization 4(2):267–304. Jensen, M.C. and Meckling, W.H. (1992) ‘Specific and General Knowledge, and Organizational Structure’, in L.Werin and H.Wijkander (eds) Contract Economics, Oxford: Basil Blackwell: 251–74. Jost, T.S. (1983) ‘The Joint Commission on the Accreditation of Hospitals: Private Regulation of Health Care and the Public Interest’, Boston College Law Review 24: 835–923. Kogut, B. and Zander, U. (1992) ‘Knowledge of the Firm, Combinative Capabilities, and the Replication of Technology’, Organization Science 3:383–97. Langlois, R.N and Robertson, P.L. (1995) Firms, Markets, and Economic Change: A Dynamic Theory of Business Institutions, London: Routledge. Lathrop, J.P. (1991) ‘The Patient-Focused Hospital’, Healthcare Forum Journal: 17–20. Milgrom, P.J. and Roberts, J.D. (1992) Economics, Organization, and Management, New York: Prentice-Hall. Mintzberg, H. (1979) The Structuring of Organizations, Englewood Cliffs: PrenticeHall. Nelson, R.R. and Winter, S.G. (1982) An Evolutionary Theory of Economic Change, Cambridge, Ma.: the Belknap Press of Harvard University Press.

171

DEBORAH A.SAVAGE AND PAUL L.ROBERTSON

Nohria, N. and Eccles, R.G. (1992) Networks and Organizations: Structure, Form, and Action, Boston: Harvard Business School Press. Powell, W.W. (1990) ‘Neither Market Nor Hierarchy: Network Forms of Organization’, Research in Organizational Behavior 12, 295–336. Richardson, G.B. (1972) ‘The Organization of Industry’, Economic Journal 82: 883–96. Roberts, J.S., Coale, J.G. and Redman, R.R. (1987) ‘A History of the Joint Commission for Accreditation of Hospitals’, Journal of the American Medical Association 258:936– 40. Rosenberg, C.E. (1995) The Care of Strangers: The Rise of America’s Hospital System, Baltimore: Johns Hopkins University Press. Savage, D.A. (1993) Change and Response: An Economic Theory of Professions with an Application to Pharmacy, unpublished PhD dissertation, University of Connecticut. Stinchcombe, A.L. (1990) Information and Organizations, Berkeley: University of California Press. Teece, D.J. and Pisano, G. (1994) ‘The Dynamic Capabilities of Firms: an Introduction’, Industrial and Corporate Change 3(3):537–56. Weber, M. (1946) From Max Weber: Essays in Sociology, ed. and trans. H.H.Gerth and C.Wright Mills, New York: Oxford University Press. ——(1947) The Theory of Social and Economic Organization, trans. A.M.Henderson and T.Parsons, ed. T.Parsons, New York: Oxford University Press.

172

7

MEN AND MONOTONY Fraternalism as a managerial strategy at the Ford Motor Company Wayne A.Lewchuk1

Mr. Ford’s business is the making of men, and he manufactures automobiles on the side to defray the expenses of his main business. (Rev. S.S.Marquis, Director, Ford Sociology Department, 1915–21)2 One of the most priceless possessions still retained by modern man is what is called manhood…. Would you be a MAN—free, proud, independent, POWERFUL? Then get together with your fellow worker, ORGANIZE YOURSELF, and you will be in a position to proudly look into the eyes of foremen, straw bosses, and all the world and say, I AM A MAN. (Auto Worker News Oct. 1927:4)

Until very recently, the shopfloors of America’s leading automobile manufacturers, and the unions that bargained for those who worked there, were the domain of men. For example, throughout the interwar period, the level of female employment in production departments hovered around 1 per cent at the Ford Motor Company. At its massive River Rouge complex outside Detroit, Ford employed over 70,000 men but not one woman in a production department in the early 1940s.3 This article examines why this was the case. I will argue that, as the production process became more capital intensive and integrated, the importance of converting labour time into a steady stream of effort increased. But converting time into effort became more difficult as work became increasingly unskilled, repetitive and monotonous, and as workers, especially male workers, became alienated from their tasks. For men, this alienation was partially the result of the growing gap between the nature of work under mass production and the gender norms of skilled men who, building on their and their fathers’ experiences in craft shops, associated

173

WAYNE A.LEWCHUK

independence and decision-making power at work with masculinity. At Ford, the attempted solution to this problem went far beyond simply raising wages to compensate for the deterioration in working conditions. The increase in wages was part of a broader strategy to reconstruct the concepts of masculinity inherited from the nineteenth century. The new focus was on hard work and making useful products in the company of other men. Fraternalism replaced paternalism as a managerial strategy to convert labour time into effort, a strategy that limited employment opportunities for women. Research during the 1970s and 1980s directed at understanding the experiences of women at work is now having a profound impact on our understanding of men at work.4 Two critical themes have emerged. First, a deeper understanding of the role of women within the family, kin networks, and the ‘informal economy’, can illuminate the relationship of women to work and work-based organisations. This is described as a ‘gender model’, as distinct from a ‘job model’ which explains behaviour on and off the job by looking only at work-based experiences. Second, these studies suggest that gender norms change over time and are socially constructed as men and women interact in their daily lives. The literature on gender and work encourages us to think of men and women as bringing both economic and gender interests to the workplace, and to acknowledge that patterns of conflict and cooperation between employers and employees need to be studied in terms of both sets of interests. For example, in the case of men, their role within the production process has implications for their status and authority in the larger society and within the family. Furthermore, although male managers and male workers may have opposing economic interests in some contexts and complementary ones in others, they remain men throughout. The possibility of male workers participating in the restructuring of work contrary to their economic interests in order to protect their gender interests must be considered. From this perspective, managerial strategies are gendered strategies, in the sense that they build upon existing gender norms in pursuit of economic objectives. I will argue later that, in some cases, employers try to change these norms in order to achieve their economic goals. It is beyond the scope of this chapter to put forward a general model of the relationship between gender, technology and effort norms. Through their impact on labour relations and effort levels, gender norms may bias the process of technical change by focusing interest on certain types of innovation or by influencing which innovations succeed. On the other hand, innovations, stimulated by economic forces, may undermine existing gender norms and cause changes in the social construction of masculinity and femininity. It is likely that both processes are at work in most periods. The objective of this chapter is limited to showing that gender norms were important in the transition to mass production and that studies of managerial strategies and the process of technical change need to be sensitive to the gender context. 174

MEN AND MONOTONY

In what follows, the focus is the American automobile industry and the workshops of Henry Ford between 1903 and 1930.5 During this period a technology based on a predominantly skilled male workforce was displaced by one employing large numbers of less-skilled but still predominantly male workers, who were machine-paced and subjected to an extreme division of labour. Ava Baron has argued that periods of major technical change can create a crisis in masculinity, ‘a crisis for men workers both as men and as workers’ (Baron 1987:61). She has shown how the late nineteenth-century introduction of the linotype machine in the printing trades reshaped social norms regarding how men should work and how hard they should work. Building on these insights, I suggest that a similar crisis occurred during the transition to mass production in the automobile industry. How the crisis was resolved may shed light on how Ford revolutionised labour productivity standards and why other automobile producers, especially those in Europe, had difficulty matching Ford’s success.

The automobile industry as a male domain One of the critical points made in the literature on women at work is that the social construction of gender norms gives substance to concepts of masculinity and femininity. Ava Baron, Sonya Rose and Elizabeth Faue have shown how lived experiences, interpreted through the lenses of language and discourse, shape and give meaning to gender identities and how men and women come to understand their roles in society (Baron 1991b; Faue 1991:69–99; Rose 1992:1–21). In this section I examine how gender norms interact with more conventional economic forces to shape the gender division of labour and why women and men rarely work together on the same tasks in a workplace.6 Large sections of the American automobile industry, despite initial trends towards a mixed-gender workforce under mass production, became male domains in the interwar period. Why and how Henry Ford encouraged this arrangement will be examined in the final two sections. Recent research on nineteenth-century Britain has shed new light on how gender norms gave men an advantage in securing skilled work outside the home. Notwithstanding the bias in labelling men’s work as skilled regardless of the tasks actually performed, there is evidence that many nineteenthcentury occupations required workers to make decisions based on specialised knowledge and that these occupations were dominated by men (Rose 1992:22– 31; Valverde 1988:621–5; Phillips and Taylor 1980). Men monopolised such work in part thanks to social norms that identified control, independence and the ability to make decisions as inherent masculine traits. There was also social pressure on employers to allocate work outside the home to men, leaving women, especially married women, the responsibility for maintaining the domestic sphere. Even when fortunate enough to find work outside the

175

WAYNE A.LEWCHUK

home, women’s opportunities were limited because many positions were deemed unladylike, as distinct from too strenuous or requiring specific knowledge. For instance, one British brass founder, pressing to employ women in his shops, was rebuked by a male member of a parliamentary committee; the parliamentarian did not question the ability of women to perform the tasks, but argued instead that women should not be employed on the grounds that wearing a dress was impractical in parts of the shop (Rose 1992:31). The gender division of labour in nineteenth-century Britain was never determined by gender norms alone. Employers were under pressure to minimise costs, and the employment of low-wage women and children was tempting, whatever tensions may have been created because of prevailing views of appropriate gender roles. The supply of labour, although arguably itself a product of gender norms, exerted its own effects, interacting with gender factors in complex ways. For instance, British men were more likely to pursue the training needed to master the mysteries of nineteenth-century trades. This imbalance reflected both differences in the expected return to training for men and women—itself conditioned by gender asymmetry in the allocation of work—and the influence of norms of appropriate male and female behaviour that ensured that only men would pursue specialised training (McClelland 1989:169). Technological changes that took place during the nineteenth and twentieth centuries exacerbated the tensions between adhering to prevailing gender norms and minimising costs, as knowledge and strength requirements were reduced and the ease of substituting low-wage women and children for male workers was increased. Claudia Goldin has argued that, as a general rule, ‘technological advances are accompanied by an increase in the female intensity of an industry or a sector of the economy’ (Goldin 1990:94). The historical record in Britain and elsewhere suggests a diversity of responses, depending on existing gender norms, the availability of suitable female workers, and the ability of men to organise to defend their economic interests, a defence that itself was often organised around gender norms. During the first stages of the Industrial Revolution in both Britain and the United States, women readily found employment in industries such as textiles (Kessler-Harris 1982:20–2; Dublin 1975; Rose 1987, 1991). In Massachusetts in 1840, 40 per cent of the entire industrial workforce was female (Goldin 1990:50). Women found their way into the British hosiery, carpet and silk trades and worked many of the new steam-powered looms (Rose 1992:31, 102–25, 154–84). However, in other trades, employers either made no effort to use women or were prevented from doing so by organised men. In the lace trades in Britain, men worked the new machines for making lace, whereas in the United States attempts by employers to use women on the new linotype machine were successfully resisted by men (Rose 1992:25; Baron 1991b). The spread of mass production created new opportunities for employing women, and in industries such as the manufacture of electrical goods, women 176

MEN AND MONOTONY

were employed in large numbers. But even here variations can be found, with firms like Philco in the United States relying more heavily on men than their competitors (Cooper 1991). The automobile industry, despite major changes in technology between 1903 and 1930, employed very few women in the United States.7 At first, a number of Detroit firms did turn to low-wage female workers during the shift to mass production. Hudson began employing women in 1910 when it built special facilities for them in their new plant, and may have employed as many at 5,000 women by the mid-1920s. In 1927, women trimmers were employed at Murray, who supplied bodies to Ford. At Studebaker, women were employed in large numbers on drill presses, lathes and internal grinders, and they could be found on the assembly line at the Piquette plant in the 1920s.8 Hòwever, by the late 1920s, before the onset of the Depression, many automobile employers lost their early enthusiasm for employing women. A study by the Women’s Bureau in 1928 charged that employers showed little interest in exploring how the labour of women could be more effectively used in automobile plants. Their study concluded: Throughout the interviews…there was little to indicate that anything scientific in the way of personnel work had been undertaken to find which jobs were suitable for women…. Managers who were progressive and alert to changes in machinery, to new ideas in sales and advertising policies, when they turned to personnel problems, especially those affecting the employment of women, were likely to revert to rote reasoning.9 By 1941, as shown in Table 7.1, the purge of women from the plants of Detroit’s major assemblers was almost total. In 1942, when women were brought into the plants in large numbers on war work, they tended to work in isolation from men in predominantly female departments. Once the war ended, women were quickly ‘invited’ to leave the automobile plants and to return to the home, thereby maintaining the masculinity of the industry (Milkman 1987:99–127; Kossoudji and Dresser 1992). At Ford there was never any great interest in employing women. As shown in Table 7.2, the spread of mass production was associated with a gradual decline in the proportion of women production workers. In 1914 women were found in the magneto, top-making and upholstery departments, whereas in 1925 women worked in magnetos and coils.10 At the new Ford River Rouge plant, which became the main production facility in Detroit by the late 1920s, no women were employed. It was men who worked the sewing machines in the upholstery department. What explains the low level of female employment in this industry, the

177

WAYNE A.LEWCHUK

Table 7.1 Employment of women in Detroit automobile plants, 1941 and 1942

Source: Employment of Women and Negroes, April 1943, UAW Research Department, Reuther Library, Ace. 350, Box 11, File 11.

Table 7.2 Female employment at Ford Motor Company, 1912_1941 (percentage of workforce)

Sources: 1912: Free Press Dec. 1912, in Edward Levinson Papers, Reuther Library, Acc. 85, Box 1, File Automobile History 1910_15; 1914: Arnold and Faurote, Ford Methods: 58; 1917 and 1921: Labor Employment Totals, Ford Archives, Ace. 940, Box 16; 1925_1929: Factory Counts, Ford Archives, Ace. 732; 1941: Employment of Women and Negroes, April 1943, UAW Research Department, Reuther Library, Acc. 350, Box 11, File 11. Notes: The Ford data are for the home plant and prior to 1929 do not include the River Rouge plant.

178

MEN AND MONOTONY

trend away from employing women in the late 1920s and early 1930s, and the absence of women on the shopfloor at Ford? The question is even more puzzling if one accepts the contemporary opinion that trends in technology after 1910 reduced the strength and knowledge requirements for working in automobile plants, opening possibilities for employing more women.11 A mid-1920s review claimed: Quickly—overnight as it were—the machine, gigantic, complex and intricate, has removed the need of muscle and brawn…. Instead we have a greater demand for nervous and mental activities such as watchfulness, quick judgments, dexterity, guidance, ability and lastly a nervous endurance to carry through dull, monotonous, fatiguing rhythmic operations. (Reitell 1924:43) The author went on to argue that such work was perfectly suited for women. An internal survey of Ford jobs in 1913 or 1914 concluded: At the time of the inquiry…there were then 7,882 different jobs in the factory. Of these, 949 were classified as heavy work requiring strong, able-bodied, and practically physically perfect men; 3,338 required men of ordinary physical development and strength. The remaining 3,595 jobs were disclosed as requiring no physical exertion and could be performed by the slightest, weakest sort of men. In fact, most of them could be satisfactorily filled by women or older children. (Ford 1923:108; Emphasis added) The cost-effectiveness of employing women on mass-production tasks in the automobile industry is supported by observations that, when employed, they were as productive as men, but paid much less. According to the Auto Worker in 1927, ‘They [women] can do many of the operations as well as men, but they get much less money for doing the same work’ (Auto Worker Dec. 1927; Gartman 1987:256–7). Nancy Gabin has shown that women doing identical jobs to men earned from one-half to two-thirds of the male wage rate (Gabin 1990:14–15). Nor is it plausible to suggest that Ford needed to maintain maximum flexibility in allocating labour and hence resisted hiring women who may have been ill-suited for one-half of all positions. By 1917, nearly one-fifth of the workforce at Ford was classified as crippled or physically substandard, and, one suspects, as limited in the number of jobs they could do as able-bodied women might have been (New York Times 9 May 1928). A number of recent studies have pointed to the role of labour laws and factory legislation in restricting industrial wage work for women. 179

WAYNE A.LEWCHUK

However, a Department of Labor study of the automobile industry in the United States in 1928 dismissed this explanation. Conditions in the industry already exceeded most of the standards imposed by law, and there were no restrictions in Michigan on night work by women. Instead, it was argued, ‘In no other industry studied was there found such violent prejudice [by employers] against women’s employment because of the mere fact that they were women. It is this prejudice that is a stronger force than any legislative requirements in closing opportunity to women in such occupations.’12 Ruth Milkman has argued that Ford’s decision to follow a high-wage and high-effort strategy, abandoning at least temporarily the more traditional strategy of employing the least-costly labour time available, precluded the employment of women. She suggests that, even during periods such as the 1930s, when firms were more concerned about minimising costs, the dominant ideology that in times of high unemployment men should be offered jobs first shaped hiring policies (Milkman 1987:12–26). There is little doubt that the general acceptance of male priority in employment had negative consequences for women seeking work, especially in times of high unemployment. It is less obvious why the decision to pursue a highwage policy in 1914 should close the door to women seeking jobs. Ford could have increased the wages of women relative to their alternative opportunities and extracted more effort from them as well. They were almost certainly available, even outside of the war years, given their employment in the thousands by other Detroit assemblers mentioned above. Had Ford employed women at high wages, there is no reason to believe that the effort-to-earnings ratio of women would have been less attractive than that of men on many jobs. But there is no evidence anywhere in the Ford archives of an attempt to hire women during this period, nor of their unsuccessful use in production departments. It was an option that simply was not considered. In the remainder of the chapter, I show that women were excluded from production, not because men were being paid a high wage, but rather because it was unclear if time could be converted into effort as efficiently in a mixed-gender workforce. The exclusion of women was part of a broader strategy by Ford to reshape masculinity along lines more consistent with conditions in a mass-production factory. Ford consciously excluded women from the workplace and created a fraternal system, a men’s club, to help male workers adjust to a world of monotonous repetitive work. In the process, Ford and his managers shifted both gender norms at work and standards of labour productivity; they also helped to remodel the family and the role of working-class men in society. The Ford strategy gave real meaning to Marquis’s claim that ‘Mr. Ford’s business is the making of men.’

180

MEN AND MONOTONY

Masculinity in the period before mass production In order to understand why Ford might believe that the conversion of male time into effort would be more difficult with a mixed-gender workforce, despite evidence that on individual tasks women were a cost-effective alternative, it is useful to look more closely at the norms of masculinity in effect at the end of the nineteenth century and at how work in the new massproduction plants was incompatible with these gender norms. Our current understanding of working-class masculinity is shaped by studies of nineteenthcentury British, and to a lesser extent North American, skilled male workers. David Montgomery argued that, for skilled men, the concept of masculinity was ‘embedded in a mutualistic ethical code’. In both Britain and North America, the ‘manly’ worker ‘refused to cower before the foreman’s glare— in fact, often would not work at all when a boss was watching.’ He accepted group-imposed quotas or stints, even when management offered attractive incentives to speed up production. The skilled male worker was expected to behave in a ‘manly’ fashion towards his fellow male workers, which implied refusing management deals that might undermine a fellow worker’s job.13 Nineteenth-century Toronto printers linked the cultivation of one’s intellect and the application of acquired knowledge at work and at home with manliness (Burr 1993a: 11–19; 1993b). Keith McClelland’s study of the mid-nineteenth-century British ‘representative artisan’ suggested that such a worker was a man in control of his life and worthy of respect (McClelland 1989). In the words of a British miner in 1873, a man wanted ‘…the independence of the workshop, and he wanted to be able to pursue his work in such a manner and under such a condition that it should not be a degradation to him in his eyes. He wished to be independent in following his ordinary daily occupation’ (McClelland 1989:172). Many of these values were taught during an apprenticeship, during which boys learned the mysteries of a trade, appropriate attitudes towards employers, male codes of sexual conduct, and male social responsibility. In the process, a boy went through a period of ‘unfreedom’, from which he emerged a free and independent man, that is, ‘one of the lads’ (McClelland 1989:170). Many crafts in North America continued the tradition of apprenticeships. Armed with new technical skills, boys proved their masculinity by defending their right to a decent standard of living and autonomy in applying their skills (Baron 1991b: 50–61). We are only beginning to understand what happened to this artisan sense of masculinity during the early decades of the twentieth century. British employers were clearly concerned that the systems of mass production being pioneered by the Americans were incompatible with elements of existing gender norms, especially the level of autonomy and independence of workers on the job. For instance, Charles Bayley argued before the British Institute of Automobile engineers that:

181

WAYNE A.LEWCHUK

In America, I understand, the labor available is much more amenable to systematized working. In England there is difficulty in getting a man to do exactly what he is told, because he is apt to think a great deal more for himself than do his fellows in America. Therefore, a system in this country has to be more elastic and less precise than many American systems are said to be. (Reeves and Kimber 1916/17:396) Austin’s Perry Keene made the following observation on American labour management methods: ‘In America you have to employ methods which a crowd can carry out, but the British individual will not have that…the Britisher will not have ‘herd’ methods. He has the individualistic tendency, and it is a British tendency that you have to allow for’ (Keene 1916/17:31). Even after British employers had shifted to what was in essence mass production along American lines in the years following 1945, it has been argued that British men constructed a mythology that retained for them the aura of nineteenthcentury craft control, despite the reality of deskilling and loss of authority (Thompson 1988). In North America, this period was also one of conflict between existing gender norms and the changing nature of work, conflicts exacerbated by the richer ethnic and racial composition of the work-force.14 For instance, early twentieth-century Northern cigar makers prided themselves on their ability to challenge employer prerogatives and assert control over their workplaces. Disputes over poor stock, victimisation of members, dictatorial foremen and violations of apprenticeship rules, often occasioned walkouts. But in areas such as Tampa, where immigration had changed the composition of the labour force, men and women worked together in the cigar trades. Northern union leaders reacted by questioning the manhood of Tampa men and called on them to join the brotherhood and return women to their rightful places in the domestic sphere (Hewitt 1991; Cooper 1987, chs 1, 3). In New England, immigrant Lancashire textile workers defended their vision of masculinity built on autonomy in the shops, control of one’s trade, support for trade unions, and the recognition of the right to bargain collectively. In Britain, mutual recognition between workers and employers was viewed as a manly arrangement; male workers showed deference to male employers and employers agreed to negotiate with the unions. But American employers in Fall River wanted no part of such an arrangement. They blacklisted union leaders, forced men to accept women in the shops, and undermined craft control in an attempt to lower labour costs and purge their workers of ‘chronic insubordination’. An elderly Lancashire spinner, upon being forced to renounce his union membership to obtain work in New England, declared ‘I’m humiliated—I’m less of a man than I was’ (Blewett 1991:95). The introduction of linotype technology in the printing trades led to conflict between employers wishing to employ less expensive women and 182

MEN AND MONOTONY

male printers trying to defend the masculinity of their trade. The men argued that printing was men’s work because it combined intellectual and manual labour, and that the traditions of tramping, social drinking while waiting for work and an aggressive shopfloor culture made printing unsuitable for women. In the end, the men maintained control of the trade, but only by redefining their masculinity on the basis of how hard they could work. Gradually the quantity of work a man could do became as important as alleged advantages in intellectual capacity in defining masculinity among printers (Baron 1987:62– 5). Race also forced some American men to redefine the basis of their masculinity. Dolores Janiewski explored the contradictions between a Southern male view of society—where it was claimed that white men loved to control, whereas blacks loved to be controlled—and the reality of the developing Southern economy in which more and more white male workers were directly controlled by white employers. She argues that the physical segregation of black and white workers was in part a response to these tensions. What emerged was a cross-class brotherhood of white men that allowed white wage-earning males the illusion of superiority over blacks in Southern society, despite their loss of autonomy (Janiewski 1991:72–7, 88). The process of modernisation was especially troublesome for middleclass American men. The rapid pace of urbanisation and the transition to white-collar employment required new ways of defining masculinity. Mark Carnes’s study of secret rituals and fraternal societies in late nineteenthcentury America indicates that the collapse of the Puritan family and the growing control of women over the household and children—as men left the home to work in offices and factories—created identity problems for men. They responded by creating secretive fraternal organisations where women could be excluded and men could reaffirm their masculinity (Carnes 1989:107–27). Anthony Rotundo has argued that, among middle-class workers, these same factors resulted in the development of what has been called the ‘masculine primitive’ norms that valued men’s alleged natural instincts. The physical strength and energy of men became a critical characteristic in defining their masculinity and their power over women in society (Rotundo 1987; Grossberg 1990; Kimmel 1987). There is thus no single set of characteristics that defined masculinity in North America or Britain during the process of modernisation, and clearly gender norms were reconstructed in various ways as technology changed the nature of work. However, at least for early twentieth-century North American men, the themes of control, skill, autonomy and independence were important in how they defined their masculinity. The centrality of these themes to the gender identities of working men created serious problems for employers in the American automobile industry, as they tried to make the transition to mass production and the institutionalisation of work that was regimented, boring and monotonous. This transition created a crisis for working 183

WAYNE A.LEWCHUK

men, not only in respect to their patterns of work, but also in respect to the ways they constructed their identities as men. It remains to show in detail how Ford responded to this crisis not only by hiring new workers from regions less affected by the norms of male craftworkers and paying his workers a premium to accept the new conditions, but also by setting out to change male notions of appropriate manly work in line with the realities of mass production and monotonous repetitive tasks. Masculinity and mass production at the Ford Motor

Masculinity and mass production at the Ford Motor Company In the period prior to the advent of mass production, Ford relied heavily on skilled male workers and had little choice but to leave them a degree of control over how vehicles were assembled. These workers played a key role in suggesting improvements in the production process, and had some control over getting tools and fixtures designed for their tasks. The social organisation of work was paternalist, with Ford visiting the shops frequently and knowing most workers by their first names.15 The spread of mass production and the transition to an unskilled workforce created both new problems and new possibilities for converting time into effort. In 1910, about 60 per cent of Ford’s workforce were classified as skilled mechanics. By 1913, skilled workers comprised only 28 per cent of the workforce, and by 1917, they represented only 8.6 per cent of all employees (Meyer 1981:48– 51). Skilled workers had been replaced by unskilled assemblers and machine minders, many of whom were machine-paced after the introduction of the moving assembly line in 1913.16 For men, the new conditions of work undermined any sense of control or independence on the job, hitherto the key characteristics of manly work among skilled workers. For male managers, the rising capital investment and the organisation of production along flow lines increased the importance of extracting as much effort as possible from labour and, as argued by Daniel Raff, at as regular a pace as possible (Raff 1992). In contrast to our earlier account of British employers, Ford and his managers pointed to the removal from workers of virtually all decisionmaking power as a great social advance. R.L.Cruden claimed that, in the 1920s, Ford tour guides boasted about the simple and repetitive nature of work at Ford. They claimed that Ford workers were well suited to such an arrangement and that ‘most of the workers have had little or no schooling…. They have never been taught to think; and they do not care to think…. All of which means that they get to like their monotonous jobs.’17 Ford argued that ‘the vast majority of men want to stay put. They want to be led. They want to have everything done for them and to have no responsibility’ (Ford 1923:99). Upton Sinclair captured this relationship in his character

184

MEN AND MONOTONY

Abner: ‘He [Ford] was going to do the thinking, not merely for himself, but for Abner—and this was something which suited Abner perfectly. His powers of thinking were limited, and those he possessed had never been trained’ (Sinclair 1937:15). Although Ford may have believed that his male workers were indifferent to the loss of skill and autonomy, one of the few accounts by a contemporary autoworker paints a different picture. In 1914, Frank Marquart, then a young factory hand but eventually a key player in the rise of the UAW, recounted how the boys and men employed in the industry would pass their time in the local saloon, talking shop, each trying to impress the others of how important his job was and how much skill it required (Marquart 1975:12). In fact, the introduction of mass production at Ford coincided with growing labour relations problems. As early as 1908, as employment levels grew, the company began experiencing problems converting labour time into effort and experimented with time studies and profit-sharing plans.18 By 1913, daily absenteeism reached 10 per cent of the workforce and the annual staff turnover rate was 370 per cent (Meyer 1981:83). Furthermore Arthur Renner, who began working at Highland Park in 1911, recalled that during this period relations between male workers were fractious. It was not unusual to see five or six fist fights in progress while walking from one department to another.19 Detroit workers were also looking to unions to defend their interests. Raff argues that at least part of Ford’s motivation in doubling wages in 1914 was to ensure that the Wobblies (the radical Industrial Workers of the World) did not gain a foothold in his shops (Raff 1988). The high rates of turnover, the turmoil on the shop floor, and the growing interest in unions, all suggest that existing male workers were less than enthusiastic about the new conditions of work under mass production. In 1914, Ford moved to stabilise his workforce by effectively doubling real wages to five dollars a day. Turnover rates fell to a fraction of their pre-1914 levels. Raff has argued that the payment of five dollars a day was more than simple compensation for the new working conditions. The long queues of men seeking employment at Ford both before and after wages were raised suggests that the company was paying them above their opportunity cost in 1914 in return for stability and more effort, in other words an efficiency wage (Raff and Summers 1987). However, according to Raff, it is difficult to explain the doubling of wages in terms of the factors stressed in the efficiency wage literature. Neither turnover costs, the need to control shirking, nor the ability to select only the most able workers warranted the doubling of wages. Raff accepts that the wage increase improved the morale of workers and led to higher productivity, but argues that the timing and the magnitude of the pay increase were determined by the company’s desire to keep newly active unions such as the Wobblies out of the Ford plants (Raff 1988).

185

WAYNE A.LEWCHUK

Hard work, useful products, and masculinity There remain, however, troubling questions about the Five-Dollar Day, the high effort norms at Ford, and the absence of women that, I argue, a gendered analysis can help resolve. One of the problems is that many of the factors that Raff suggests motivated the payment of a higher wage in 1914 remained in place through the 1930s, by which time the premium offered by Ford had eroded in real and relative terms, as can be seen in Table 7.3. Why Ford would raise wages 100 per cent to combat unions who rarely succeeded in raising wages by more than 20 or 30 per cent is also puzzling. In addition, the payment of a premium to men does not help to explain the absence of women. All of the reasons given by Raff for paying men more money apply equally to women. If keeping the union out was the main reason for paying the higher wage, one might have expected Ford to employ women, which would have compounded the organising tasks of the union by forcing it to face a mixed-gender workforce. Any argument that seeks to identify the reasons behind the Five-Dollar Day needs to explain why only men could profitably be employed at a wage above their opportunity cost and why a male-only workforce was the most effective way of preventing unionisation. A reinterpretation of changes taking place at Ford during this period suggests that the Five-Dollar Day was one component of a broader strategy to revise norms of masculinity in keeping with the new conditions of work. The first signs of a tendency to extol the virtues of work in Ford shops in order to sustain labour productivity appeared in 1908, when the Ford Times, a company magazine, encouraged workers to take a new approach to their jobs, offering them a New Year’s Eve resolution that included an exaltation of the ‘gospel of work’ and affirmed the need ‘to keep the head, heart, and hand so busy’ that a worker would not have time to think about his troubles. It suggested that ‘idleness is a disgrace’ (Ford Times 15 Dec. 1908). Obviously such messages from management to workers were not particularly novel, and could be applied to women as well as men, although in fact Ford was employing only men at this period. It is the focus on how much work, rather than the intrinsic characteristics of work and the skill needed to execute it, that marks the beginning of a campaign that would eventually redefine masculinity on the job. The campaign to promote monotonous and repetitive work as masculine intensified with the introduction of the moving assembly line. Again, the focus was not on the skill requirements and inherent qualities of work at Ford, but rather on the virtues of working hard and the usefulness of the products being produced. An appropriately titled but short-lived company publication, Ford Man, suggested in 1917 that:

186

MEN AND MONOTONY

Table 7.3 Hourly earnings in the American motor vehicle industry (cents per hour)

Sources: Industry Average. 1908/09 and 1914: Census, US Motor Industry, Ford Archives, Acc. 96, Box 10; 1919: US Department of Labor, Industrial Survey: 37; 1922 and 1925: US Department of Labor, Handbook; 1934: Tolles and La Fever, ‘Wages, Hours.’ Ford Average. Plant Analysis, Ford Archives, Acc. 96, Box 8. Ford Minimum Plus Bonus. Hours Worked and Wage Rates, Ford Archives, Acc. 572, Box 32. Ford minimum wage was deflated by the consumer price index from US Bureau of the Census, Historical Statistics: 211. Note: a average wage is less than the minimum because some Ford workers did not qualify for the bonus share when it was first introduced in 1914.

The nobility of labor rests in the practical merit of the product it makes…. A common complaint against modern labor is that it breeds dissatisfaction. This complaint cannot be lodged against Ford workers…. Even when the extreme specialization of the work obliges the man to make the same motion over and over again, the monotony of the action has not smothered the consciousness or the importance of his task…. He is building something that is a benefit to humanity and why should not his work be a pleasure. (Ford Man, 20 Sept. 1917:2–3) In 1919, the paper again stressed the importance of making useful articles, claiming that ‘anyone working for an organization that has good working conditions, is paying good wages, and is making a thing that people need can be sure that he is in a good calling’ (Ford Man, 3 Sept. 1919:2). This was a theme Ford would stress in his autobiographies, where he claimed: ‘There is one thing that can be said about menial jobs that cannot be said about a great many so-called more responsible jobs, and that is, they are useful and they are respectable and they are honest’ (Ford 1923:278). In focusing on the usefulness of the Model T automobile, Ford was both tapping an undercurrent of American populist culture that stressed the usefulness of work and reflecting pressing material demands in a developing economy (Rodgers 1974:9–13). Again, the strategy calling for more work and pride in producing a utilitarian product could have been applied equally to men or women. However, other

187

WAYNE A.LEWCHUK

evidence suggests that these characteristics were being promoted as components of a new masculine respectability. Effort and speed came to be identified as laudable features of working at Ford, affirming the positive selfimage of workers as men. Those who could not work at the expected pace were ridiculed and characterised as suitable only for women’s work, such as cutting ribbons. An early Ford manager wrote: ‘…many of them—often who have been measuring ribbon, perhaps and doing such things in life—find they can not, in the language of the shop, “stand the gaff” and they immediately resign…. They would rather go back to the ribbon counter for $12 or $15 a week than to stand what is necessary to become efficient in the Ford Motor Co.’20 In 1928, the journal Labor Age noted that, on the Ford assembly line, ‘their sweat is the sign of their labor; the impression of the whole factory is sweat; the whole district reeks with sweat. For sweat is a sign of speed, and speed is Henry Ford’s god.’21 Even Upton Sinclair’s fictional character Abner glorified hard work: [Abner] did not think of the Ford plant as an immense and glorified sweatshop; he thought of it as a place of both duty and opportunity, where he did what he was told and got his living in return…. It passed Abner’s comprehension how any man or women could fail to be grateful for such divine compassion on the part of Mr. Ford. (Sinclair 1937:22, 30) The production of family men A careful reading of the implementation of the Five-Dollar Day, and the adoption of the eight-hour day, reveals that these were more than an attempt to motivate through the cash nexus. Not only was Ford raising wages, he was also promoting a particular vision of men and women at work, in the home, and in society—a vision that he hoped would enable and motivate his male workers to work harder (May 1982). When the new wage standard was introduced in early 1914, it applied only to men.22 Its objective was, in the words of a contemporary Ford manager, ‘to help men to be better men and to make good American citizens and to bring about a larger degree of comfort, habits and a higher plane of living among our employees by sane, sound and wholesome means.’23 Rev. Marquis, who had taken over the company’s Sociology Department in 1915, argued that Ford’s intention was ‘to uplift the community; make for better manhood and character of his employees…and fix in their minds such ideas of right living as go to make better American citizens.’24 He suggested that: We have in mind a man who is right in his relations toward his employer, his family and towards the community in which he lives. This is the kind of man we have in view. This is the human 188

MEN AND MONOTONY

product we seek to turn out, and as we adapt the machinery in the shop to turning out the kind of automobile we have in mind, so we have constructed our educational system with a view to producing the human product we have in mind.25 With the higher wages, men were expected to provide for their families and participate more fully in family life. In return, their status within the household would be enhanced. Ford argued that ‘if a man feels that his day’s work is not only supplying his basic need, but is also giving him a margin of comfort and enabling him to give his boys and girls their opportunity and his wife some pleasure in life, then his job looks good to him and he is free to give it his best’ (Ford 1923:120). Men were encouraged to buy homes and life insurance and were discouraged from drinking outside the home or taking in boarders. Ford’s promotion of men as responsible heads of households is evident in its policy of allowing a divorced wife to request the Sociology Department to deduct a set amount from her ex-husband’s pay. If he objected, he was fired.26 This new focus on home life, made possible by both the FiveDollar Day and the eight-hour day implemented in 1914, represented a distinct departure from the culture of nineteenth-century skilled men.27 As the focus of male activity outside the workplace, the family was to replace the working-men’s societies, the Mechanics Institutes, and perhaps even trade unions, all of which played an important role in the lives of respectable nineteenth-century skilled working men and gave meaning to their identities as men. For Ford, social respectability was a product of earning a high wage, as distinct from the nineteenth-century concern with performing skilled work and retaining a degree of independence on the shopfloor. According to George Brown, who started working at Ford in 1907, every man wore his Ford badge in public after the granting of the Five-Dollar Day because he was proud of it.28 Allan Nevins also reports that Ford workers wore their numbered company badges to social functions (Nevins 1954:549). To J.R.Lee, Ford’s first personnel manager, the Five-Dollar Day was a comprehensive strategy that linked together monetary inducements to effort, the home as a nurturing centre enabling high effort, and the redefinition of masculine work as labour that required abnormal effort. In 1914, Lee wrote: ‘Mr. Ford believes, and so do I, that if we keep pounding away at the root and the heart of the family in the home, that we are going to make better men for future generations, than if we simply pounded away at the fellows at their work here in the factory’ (Meyer 1981:124). Charles Reitell, a contemporary journalist, argued that ‘to attain a normal day’s production the worker is timed so as to keep up an energetic gait for eight hours a day—this can only be done when a well-regulated living is carried on by the worker in his home life’ (Reitell 1924:38). This point is elaborated by Martha May in her analysis of the Five-Dollar Day: 189

WAYNE A.LEWCHUK

Ford’s family wage implicitly recognized the contribution of women’s domestic labor to a stable and secure family life. In all likelihood, Ford believed that women’s contribution was greatest in their emotional, nurturing, and motherly roles. This emphasis on psychological rather than material comfort parallels the arguments of many Progressive reformers, who saw the female emotional, affective role as a necessary aspect of family life which should be supported by adequate wages. (May 1982:416) Some Ford managers made a direct link between stability at home and effort at work. Marquis asserted that: …almost immediately after the profit share plan was inaugurated, there was a voluntary increase in efficiency of about 20 per cent…. It stands to reason that a man who comes out of a home well fed, whose children are properly clothed, whose wife is happy, and where there is a freedom from debt and all its worries, cannot help but be a great deal more efficient.29 In another paper Marquis announced: ‘We have made a discovery. You have all heard that the family is the foundation of church and state. We have found that the family is the foundation of right industrial conditions as well. Nothing seems to lower a man’s efficiency more than wrong family relations.’30 A community of men Accompanying the increase in wages was a campaign to promote the Ford Highland Park plant as a fraternal community, a male club. The company publication Ford Man was a key contributor to this new campaign; its stated objective was ‘to cultivate and establish the broadest fellowship among Ford workers through understanding each other’ (Ford Man 20 Sept. 1917:2). It stressed that ‘the Ford shop is a community in which all kinds of men’ could be found (Ford Man 20 Sept. 1917:3). Marquis, in a piece on the ‘Eight Hour Day’, asserted that ‘The attitude of the Ford Motor Company toward its employees is not paternal, but fraternal. The “Help the Other Fellow” spirit runs through all we do. The men catch this spirit and are practicing it in their relation toward one another.’31 A pamphlet distributed to Ford workers in 1915 claimed that the objective of the Five-Dollar Day was ‘to better the financial and moral standing of each employee and those of his household…to implant in the heart of every individual the wholesome desire to Help the Other Fellow, whenever he comes across your path.’32 George Brown, an early Ford worker, recalled that on one occasion when ‘the boys’ got together to help out a fellow worker, a Pole who had 190

MEN AND MONOTONY

tuberculosis and whose wife was having their eighth child, ‘it was a nice procedure. It made them feel that it was a family affair, not to just go there and labor. It was what Mr. Ford always figured. Let us be a family. Let us look after one another.’33 The ‘Help the Other Fellow’ philosophy was a regular component of the preachings of Ford Man. In 1918, when the newly introduced Ford suggestion plan was not living up to expectations, the editors admonished the men and foremen to pull together and to ‘speedily learn the solid truth that in “helping the other fellow” we can’t miss helping ourselves’ (Ford Man 17 May 1918:1). It was the central idea of what was described as the Spirit of Ford. ‘The Ford spirit is toward Achievement!…“Help the Other Fellow!” that’s the Ford spirit, a splendid spirit of co-operation…. It is a spirit that brings out the best there is in every man… Heed then, the spirit of Ford!—be dependable, be thrifty, be efficient, be all good things to all men.’34 The fraternal and community aspect of the Ford strategy was enhanced by other initiatives. Ford established a company store to combat the persistent high cost of staples in Highland Park. This was accompanied by new efforts to help Ford workers find housing through the pages of Ford Man.35 Ford workers were also invited to invest in the company through Ford Investment Certificates.36 Once the Ford Hospital was opened, the Sociology Department arranged to have workers and their families treated on account, with a portion of the bill being deducted from each week’s pay.37 Employment policies also reflected the climate of community and fraternalism being developed at Ford. The changes implemented went much further than the rationalising of hiring procedures in 1913 and the removal of the foreman’s autocratic power. In 1914, when demand for vehicles fell, married men were given preference over single men for employment.38 Ford’s stated policy was ‘to give employment to those who are in the greatest need. As a result, a slight preference is given to married men, or single men who have others totally dependent upon them for support, although single men with no dependants are not barred.’39 According to William Baxter, who worked in the Sociology Department—renamed the Employment Department in the 1920s—if a worker selected for laying off by a foreman was ‘in dire financial straits which could be substantiated, we would make arrangements to return the man to his department and ask the foreman to replace him with some other employee in more fortunate circumstances.’40 Over time, however, this practice created ill feeling and opportunities for abuse; it was among the reasons that the unionisation campaign was eventually successful.41 Ford also had a policy of hiring men who traditionally found it difficult to find employment in manufacturing shops, including convicts, disabled men and petty thieves. It was suggested that Ford was ‘the first corporation to undertake on a large scale the work of securing efficiency from defectives by 191

WAYNE A.LEWCHUK

the proper adjustment of the man to his work. In this instance the inception of the idea came through the motto, “Help the Other Fellow,” originated by Ford’ (Mead 1918). By November of 1917, Ford employed 6,095 crippled or physically substandard men, representing 18 per cent of the workforce. By 1928, there were 2,600 male convicts working in the plants (New York Times 9 May 1928). In 1926, Ford declared that he would help solve Detroit’s crime problem by hiring 5,000 boys to keep them out of mischief.42 Even the limited employment of women after 1919 contributed to the restructuring of Ford as a fraternal community where men supported each other for the common good. ‘Like the substandard men, the women are employed not because they are women. Most of the Ford women are wives or daughters of Ford men who have been in some way temporarily or permanently disabled. A women whose husband is an active worker in the factory cannot obtain employment in the Ford plants.’43 Together, the changes I have described replaced the paternalist strategy employed by Ford in the pre-mass-production era with a fraternalist strategy, one that was more in tune with the new conditions of work. Collectively, the campaign to glorify hard work and to generate pride in making useful articles, the promotion of a fraternal community through selective hiring policies, and the granting of a family wage that raised the status of men within their households and the community, created a package intended to make monotonous and repetitive work manly and hence respectable, if not enjoyable, for Detroit men. The adoption of this strategy was conditioned by numerous external factors such as the availability of workers from areas less affected by the norms of skilled workers, the emerging populist culture in the United States, the changing role of work in defining masculinity in the United States, as well as Ford’s own religious beliefs—which leaned heavily towards social gospel teachings and shaped his vision of the proper roles of men and women and of the relationships between work and family in everyday life.44 None the less, the weight of evidence supports the conclusion that, during the transition to mass production, Ford managers consciously promoted new norms of masculinity to maintain high effort levels in an environment where workers had little autonomy on the job. Such a fraternal community might also have been resistant to organising by outsiders and hence have helped secure the labour peace essential to an integrated production process. In any case, it was a strategy that left little room on the shopfloor for women. Although not the main focus of this article, it is worth discussing briefly the role of male workers in supporting the new norms of masculinity. Male workers and the unions that tried to bargain for them were generally supportive of the gender hierarchy implicit in the Ford fraternalist strategy. In the early 1920s, union papers such as the Auto Worker carried articles discussing alleged differences between men and women. One article suggested that women were less suitable for factory work than men because 192

MEN AND MONOTONY

women valued employer approval more than the size of their pay check, which led them to ‘unconsciously expanding energy which is unpaid for’ (Auto Worker Dec. 1920:13). This same paper, recognising the existence of female workers in the industry, served their interests by running a page dedicated to women’s concerns. On the ‘Women’s Page’, women and girls could read about the value of flirting, how to build children’s minds, and what to do about cranky children, as well as find household hints—including recipes for sour cream pie (Auto Worker Apr. 1921:9–10). Nancy Gabin, in her study of women and the UAW, argued that men viewed women as less than equal partners in the drive to organise the auto industry and that men were supportive of a gender hierarchy that insured their access to high-paying jobs. Union organisers called on male workers to vote for their union to raise wages and to restore their manhood. When employers refused to raise wages, the union dressed a number of men in women’s clothes, symbolising their understanding of how the level of wages differentiated men from women. The union’s approach to job classifications and seniority and their reluctance to organise women workers or to demand pay equity, or even equal pay for strictly equal work, confirmed the existing male-dominated gender division of labour (Gabin 1990:28–31). The extent to which the unions’ position on gender hierarchy was itself a product of strategies implemented by employers such as Henry Ford remains to be explored. What is clear is that, in many areas, Ford, his male workers and the unions trying to organise these workers were in agreement that men should be given preferential access to employment in the industry and that women should remain in the domestic arena.45 Ford’s strategy was not without its contradictions. We have already suggested that allocating lay-offs based on financial need created serious strains in the community during the 1930s. The decision to hire blacks was also problematic. One might have expected that, if Ford was trying to run a men’s club, it would have been exclusively a white men’s club. This was not the case, for in 1919, following Packard’s lead, Ford began hiring large numbers of blacks and quickly became the largest employer of blacks in the area. Although many found jobs in dirty and dangerous parts of the plant, Ford opened all areas of employment to blacks, including the apprenticeship schools. Employment at Ford became a mark of superiority, and it was not unusual to see black workers wearing their Ford badges on their lapels on Sundays. To Ford, blacks were quiet, compliant and loyal, both needing guidance and willing to be guided.46 However, racial conflict was a serious problem; it has even been claimed that in the late 1920s the possibility of laying off all black workers was seriously considered, given the amount of unrest in the shops.47

193

WAYNE A.LEWCHUK

Was the fraternalist strategy successful? Was Ford successful in creating a fraternal community and making monotonous repetitive work desirable to men? The telling of this story has depended heavily on statements by Henry Ford and Ford managers. It is difficult to get evidence regarding the actual impact on workers. It is likely that Ford workers were neither as joyful about monotonous work, nor as enamoured with the Ford community spirit—which was rarely extended to workers beyond the age of 40—nor as thankful for a wage packet that was eroded by inflation and employment instability as Ford, Lee and Marquis would have us believe. Reports suggest that in 1918 absenteeism and lateness had again become problems (Ford Worker 18 Mar. 1918). The pages of the union papers such as Auto Workers News and The Ford Worker are filled with complaints regarding labour speed-up and intolerable working conditions in the Ford plant, especially during the 1930s.48 One contributor to Ford Worker complained: Such is the wonderful thing called the ‘Ford Spirit.’ It makes sure that we workers feel like slaves, we are kicked from pillar to post. The star men, the foremen, the straw bosses, all of them have but one reason to keep their jobs and that is more slave driving, more speed-up. We can only stop this when we put up a united resistance. I for one say let’s organize. (Ford Worker Dec. 1929:3) When given a chance, Detroit workers, some of them from Ford, did organise and join fledgling unions in 1919 and again in the mid-1920s (Meyer 1981:171, 197). Racial and ethnic tensions within the male workforce also speak to the limits of the fraternal model as proposed by Ford. For some of the company’s workers, the Ford system was simply a massive intrusion into their lives. In the words of a young married polisher in the 1920s, ‘You’ve got to stand for a lot of things…. You just have to see that the boss gets nothing on you, and if he does just let him bawl you out’ (Cruden 1929:698). The strategy of fraternalism itself yielded to change. The shift to the Rouge plant in the late 1920s, the decline of the Model T, and the ascendancy of Charles Sorensen mark the rise of a much harsher regime—one more dependent upon fear than pride—to get the work out.49 None the less, the gender die had been cast, and women remained outside the Ford shops. There is evidence, none the less, that Ford was at least partially successful in redefining masculinity, and in doing so, labour productivity. The company easily attracted male workers to its shops, and many stayed for a long time. In an industry with a history of high turnover rates due to its seasonal nature, Ford had 25,000 workers with twenty or more years of seniority on the

194

MEN AND MONOTONY

payroll in 1945.50 There is also ample evidence that, from the point of view of supervisors, a community spirit did exist among Ford workers, at least at Highland Park, until the late 1920s. P.E.Haglund, who began working for Ford in 1915 and became a supervisor in the open hearth in the 1920s, recalled that ‘the employees worked willingly. They would tackle jobs and they tried to get results without any particularly excessive pressure. There was an elective drive in the men to do a good job at Highland Park.’51 K.C.Klann, another early Ford worker, who became the supervisor of motor assembly in 1912, referred to the ‘help the other fellow spirit’ at Highland.52 In a similar vein, Alex Lumsden, who began working in the tool room in 1913 and eventually became head of the department in 1923, recalled that ‘Our infatuation for the Ford Motor Company in those days was something that I could write a story about…. It created a tremendous loyalty on the part of a large, large group of men in the shop…. One of the common remarks was, Now listen, that’s not the Ford spirit.’53 Absolute measures of effort norms are difficult to find, but, by most anec-dotal accounts, Ford was more successful than most employers in enforcing high levels of effort prior to the 1930s. The wife of one Ford worker declared ‘The chain system you have is a slave driver!… That $5.00 day is a blessing—but oh they earn it’ (Russel 1978:45). Another worker returned to a position at Dodge after a period at Ford, claiming he was ‘too fatigued after leaving the Ford factory to do any serious reading or attend a play or concert’ (Raff and Summers 1987:74). A Yale student on a work assignment at Ford in the 1920s reported: ‘You’ve got to work like hell in Ford’s. From the time you become a number in the morning until the bell rings for quitting time you have to keep at it. You can’t let up. You’ve got to get out the production…and if you can’t get it out, you get out’ (Meyer 1981:41). Stories are also told of Ford Rouge workers riding the Baker streetcar so exhausted after their shifts that they fell asleep, whether sitting or standing (Marquart 1975:30). During the 1920s, Ford was viewed by some of his sternest critics with begrudging admiration as somehow different from other employers. One critic wrote: Really, up until that summer (1927), the workers of Detroit had set up Henry as a little tin god, to whom the poor forlorn workers could always run in times of stress and unemployment, and always get a job at five dollars a day. Dodge might close; Hudson might shut-down; even General Motors might lay off but Henry Ford; never!… Had not Ford himself said that he was interested in the lives of his workers?54 This tallies well with an earlier assessment that ‘The conviction that character and manhood are requisites for employment has spread throughout the 195

WAYNE A.LEWCHUK

entire shop; in fact throughout the whole city of Detroit’ (Rumely 1914:668). To some at least, Ford was successful in ‘making men’.

Conclusions The implementation of mass-production technology at Ford brought with it a heightened managerial focus on effort levels. A capital-intensive integrated system could be profitably employed only by sustaining high and stable levels of effort on the part of workers. However, enforcing such requirements was problematic. The new technology generated tensions for some men as it removed many of the attributes of work prized by skilled male workers in the nineteenth century. The evidence presented in this chapter indicates that Ford went beyond raising wages and tightening up supervision to ensure an adequate supply of effort. There was a conscious attempt by management to revise notions of masculinity. Through company papers, public speeches, and employment policies, a new ideal of what a man was supposed to be—at work, within the family, and in society—was promoted. Working hard—in the company of other men, on a useful product, and being paid well for it-would make Ford workers manly, even though the work itself was repetitive, boring and devoid of many of the elements of autonomy and control that were characteristic of nineteenth-century skilled labour. The Ford strategy balanced economic and gender interests as it reached beyond the workplace to shape relations of men in the household and in society at large. The loss of control, independence and status at work was compensated for by gains in the household and in the rest of society. The exclusion of women from the Ford shops was a component of this strategy.

Notes 1 This chapter is reprinted, with revision, from the Journal of Economic History. Comments and encouragement by the editor of the Journal of Economic History, two anonymous referees, Ava Baron, Les Robb, Robert Storey and Pam Sugiman, have significantly improved the paper. Dale Brown was instrumental in ensuring clarity and focus. The research was partially funded by the Social Sciences and Humanities Research Council. 2 Rev. S.S.Marquis, Director, Ford Sociology Department, ‘The Ford Idea in Education’, (1916) 12, Ford Archives, Acc. 293. 3 Employment of Women and Negroes (April 1943), United Automobile Workers (UAW) Research Department, Reuther Library, Acc. 350, Box 11, File 11. On the role of women in early automobile unions, see Gabin (1990: ch. 1). 4 See the collection by Baron (199la), especially Baron (1991b) and Blewett (1991). See also Parr (1990), Blewett (1988), Valverde (1988), Cockburn (1983), Pollert (1981: introduction) and Milkman (1987:1–26). 5 On transitions in masculinity at the Reo plant in Lansing, see Fine (1991). On

196

MEN AND MONOTONY

6

7

8 9 10 11 12 13 14 15 16 17 18 19 20 21 22

23 24 25 26 27 28 29 30

the gendering of work in the American automobile and electrical goods industries, see Milkman (1987) and May (1982). In 1904, of the 483 industrial occupations listed by the US Commissioner of Labor, only 44 employed both men and women. See Goldin (1990:77). On the historical roots of occupational segregation, see Hartmann (1976) and Leibowitz (1986). Women probably comprised from 10 to 15 per cent of the workforce in British firms affiliated with the Engineering Employers Federation. These would have included the major British producers, but neither Ford nor Vauxhall (GM). See data from the Engineering Employers Federation, file labelled ‘Number Employed in 1910’. News (27 Oct. 1910) in the Edward Levinson papers, Reuther Library, Acc. 85; Misc. Clippings, R.W.Dunn Collection, Reuther Library, Acc. 96; and Auto Workers News (June 1927) in R.W.Dunn Collection, Reuther Library, Acc. 96. US Department of Labor (1928). Meyer (1981). See also Factory Counts, Ford Archives, Acc. 732. On women in the coil unit department, see Ford Man (3 Apr. 1918:1). On the success of women in metal shops during World War One, see US Department of Labor (1920). US Department of Labor (1928):220. See Montgomery (1979:13–14); See also Kimmel (1987) and Clawson (1989:145– 9). For an excellent collection of essays on this topic, see Baron (1991a). Arthur Renner, Reminiscences, Ford Archives: 9; Ford Old Timers, Notes on George A.Brown, Ford Archives, Acc. 616. For details on these changes, see Hounshell (1984) and Lewchuk (1987). R.L.Cruden, ‘Ford’s Flimflammery’, Labor Age (June 1928), in R.W.Dunn Collection, Reuther Library, Acc. 96, Box 2, File 2–21, 15. Letter from Couzen to employees announcing profit sharing, 29 Dec. 1909, Ford Archives, Acc. 693, Box 1. Arthur Renner, Reminiscences, Ford Archives, 14. G.Bundy Work of the Employment Department (Bureau of Labor Statistics, Bull. 196, May 1916), in Frank Hill Papers, Ford Archives, Acc. 940, Box 6, File LaborRacial & Group Discrimination, 68. R.L.Cruden ‘The Worker Looks at Ford’, Labor Age (June 1928), in R.W. Dunn Collection, Reuther Library, Box 2, File 2–9, 3. Shortly after the introduction of the Five-Dollar Day, public pressure forced Ford to allow women heads of households to qualify for the higher wage. It was not until late 1916 that Ford allowed all women the right to participate in profit sharing. See New York World (25 Oct. 1916), in Frank Hill Papers, Ford Archives, Acc. 940, Box 16. Letter dated 26 Jan. 1914 to C.L.Gould, Omaha branch, Five Dollar Day, Reuther Library, Acc. 683, Box 1. Untitled manuscript, Marquis Papers, Ford Archives, Acc. 293, 1. S.Marquis ‘The Ford Idea In Education’, Marquis Papers, Ford Archives, Acc. 293, 12. James O’Connor, Reminiscences, Ford Archives, 31. On the tension between long hours of work in the nineteenth century and the role of men within their families, see Cross (1989). George Brown, Reminiscences, Ford Archives, 83. United manuscript, Marquis Papers, Ford Archives, Acc. 293, 16. S.Marquis ‘The Eight Hour Day’, Marquis Papers, Ford Archives, Acc. 293, 15.

197

WAYNE A.LEWCHUK

31 S.Marquis ‘The Eight Hour Day’, Marquis Papers, Ford Archives, Acc. 293, 20. 32 Helpful Hints and Advice to Employees to Help Them Grasp the Opportunities Which are Presented by the Ford Profit-Sharing Plan (1915), Vertical Files, Reuther Library, Box 14, folder titled Ford Motor Company 1920s’, 3. 33 George Brown, Reminiscences, Ford Archives, 97–8. 34 Ford Man, (3 May 1918:2). See also Ford Man (17 Jan. 1919:2). 35 Ford Man, (3 Dec. 1919:1). See also Ford Man (17 Dec. 1919:1). 36 Ford Man (31 Dec. 1919:1; and 3 June 1920:1). By June 1920, over six million dollars worth of certificates had been applied for. See Fordson Worker (1 Oct. 1920:3). 37 H.S.Ablewhite, Reminiscences, Ford Archives, 15. 38 Journal (16 May 1914), Edward Levinson Papers, Reuther Library, Acc. 85, Box 1, File Auto Industry History, 1910–15. 39 Memo entitled ‘Applicants’, Marquis Papers, Ford Archives, Acc. 293. See also, Rumely (1914:669). 40 William P.Baxter, Reminiscences, Ford Archives, 29. See also Willis Ward, Reminiscences, Ford Archives, 26–7. 41 Arthur Renner, Reminiscences, Ford Archives, 35. According to Willis Ward, who worked in the Employment Department during the period when Marshall was a prominent figure, Marshall would secure employment for prospective purchasers of Ford cars referred to him by local dealers. In return, Marshall earned a handsome commission on these sales. See Willis Ward, Reminiscences, Ford Archives, 15–16. 42 Dunn (1929:71). At the time, there was little growth in the Ford workforce, so the net result of this was to bring in 5,000 fresh, young, low-paid male workers in place of 5,000 older, higher-paid men. 43 Li ‘A Summer in the Ford Motor Company’, in R.W.Dunn Collection, Reuther Library, Acc. 96, Box 1. 44 On Ford’s sympathy for the social gospel within the Protestant Episcopal church, see H.S.Ablewhite, Reminiscences, Ford Archives, 7–10. S.Marquis, the first head of the Sociology Department, had been Dean of a Detroit Protestant church. Ablewhite was also a Protestant Episcopal minister before becoming a supervisor in what remained of the Sociology Department in the 1940s. 45 In her research on industrial workers in Minneapolis, Elizabeth Faue argues that, as productive work became isolated from the family household economy, the basis of working-class protest shifted from the community to the workplace. Increasingly the workplace was defined as a male domain and trade unionism as a masculine activity. The imagery of the strong, aggressive, organised male worker dwarfing capitalism, war and hunger provided men in the industrial sector with a sense of security in the face of deskilling and loss of control at the workplace. See Faue (1991:4–15, 71–5, 98–9). 46 On blacks in Detroit, see Meier and Rudwick (1979). 47 Willis Ward, Reminiscences, Ford Archives, 11–12. 48 See Auto Workers News (5 May 1934) and Ford Worker (Oct. 1926, Dec. 1929, 20 Apr. 1932). 49 See the account of working at the Rouge by Walter E.Ulrich, published by the League for Industrial Democracy (1929:8), Vertical Files, Reuther Library, Box 14. 50 Length of Service Report (As of May 1945), Ford Archives, Acc. 616. 51 P.E.Haglund, Reminiscences, Ford Archives, 62. 52 K.C.Klann, Reminiscences, Ford Archives, 88. 53 Alex Lumsden, Reminiscences, Ford Archives, 10–12.

198

MEN AND MONOTONY

54 R.L.Cruden, ‘Ford’s Flimflammery’, Labor Age (June 1928), in R.W.Dunn Collection, Wayne State University (WSU), Acc. 96, Box 2, File 2–21, 15.

References Arnold, H.L. and Faurote, F.L. (1916) Ford Methods and the Ford Shops, New York: Engineering Magazine. Auto Worker (Chicago, various dates). Auto Workers News (Detroit, various dates). Baron, A. (1987) ‘Contested Terrain Revisited: Technology and Gender Definitions of Work in the Printing Industry, 1850–1920’, in B.Wright (ed.) Women, Work and Technology, Ann Arbor: University of Michigan Press: 58–83. ——(1991a) Work Engendered: Towards a New History of American Labor, Ithaca, NY: Cornell University Press. (1991b) ‘An “Other” Side of Gender Antagonism at Work: Men, Boys, and the Remasculinization of Printers’ Work, 1830–1920’, in A.Baron (ed.) Work Engendered: 47–69. Blewett, M.H. (1988) Men, Women, and Work: Class, Gender and Protest in New England Shoe Industry, 1780–1910, Chicago: University of Illinois Press. ——(1991) ‘Manhood and the Market: The Politics of Gender and Class among the Textile Workers of Fall River, Massachusetts, 1870–1880’, in A.Baron (ed.) Work Engendered: 92–113. Burr, C. (1993a) ‘“The Rights of Labor Are the Rights of Man”: Masculinity and Labor Reform in Late Nineteenth-Century Toronto’, photocopy, University of Toronto. ——(1993b) ‘Defending “The Art Preservative”: Class and Gender Relations in the Printing Trades, 1850–1914’, Labor/Le Travail, 31:47–74. Carnes, M.C. (1989) Secret Ritual and Manhood in Victorian America, New Haven: Yale University Press. Clawson, M.A. (1989) Constructing Brotherhood: Class, Gender, and Fraternalism, Princeton: Princeton University Press. Cockburn, C. (1983) Brothers: Male Dominance and Technological Change, London: Pluto Press. Cooper, P. (1987) Once a Cigar Maker: Men, Women, and Work Culture in American Cigar Factories, 1900–1919, Urbana: University of Illinois Press. ——(1991) ‘The Faces of Gender: Sex Segregation and Work Relations at Philco, 1928–1938’, in A.Baron (ed.) Work Engendered: 320–50. Cross, G. (1989) A Quest for Time: The Reduction of Work in Britain and France, 1840–1940, Berkeley, CA: University of California Press. Cruden, R.L. (1929) ‘No Loitering: Get out Production’, The Nation (12 June): 698. Dublin, T. (1975) ‘Women, Work, and the Family: Female Operatives in the Lowell Mills, 1830–1860’, Feminist Studies, 3 (Fall): 30–9. Dunn, R.W. (1929) Labor and Automobiles, New York: International Publishers. Faue, E. (1991) Community of Suffering and Struggle: Women, Men and the Labor Movement in Minneapolis, 1915–1945, Chapel Hill: University of North Carolina Press. Fine, L. (1991) ‘“Our Big Family”, Masculinity and Paternalism at the Reo Motor Car Company of Lansing’, photocopy, East Lansing: Michigan State University. Ford, H. (1923) My Life and Work, New York: Garden City Publishing Co. Ford Man (Ford Motor Company, various dates). Ford Times (Ford Motor Company, various dates).

199

WAYNE A.LEWCHUK

Ford Worker (The Ford Shop Nuclei of the Worker [Communist] Party of America, Detroit, various dates). Fordson Worker (Ford Motor Company, various dates). Gabin, N. (1990) Feminism in the Labor Movement: Women and the United Auto Workers, 1935–1975, Ithaca, NY: Cornell University Press. Gartman, D. (1987) Auto Slavery: The Labor Process In the American Automobile Industry, 1897–1950, London: Rutgers University Press. Goldin, C. (1990) Understanding the Gender Gap, New York: Oxford University Press. Grossberg, M. (1990) ‘Institutionalizing Masculinity: The Law as a Masculine Profession’, in M.C.Carnes and C.Griffen (eds) Meanings for Manhood: Construction of Masculinity in Victorian America, Chicago: University of Chicago Press: 133–51. Hartmann, H. (1976) ‘Capitalism, Patriarchy, and Job Segregation by Sex’, in M. Blaxall and B.Reagan (eds) Women and the Workplace: The Implications of Occupational Segregation, Chicago: University of Chicago Press: 137–69. Hewitt, N.A. (1991) “‘The Voice of Virile Labor”: Labor Militancy, Community Solidarity, and Gender Identity amongst Tampa’s Latin Workers, 1880–1921’, in A.Baron (ed.) Work Engendered: 142–67. Hounshell, D. (1984) From the American System to Mass Production, Baltimore: Johns Hopkins University Press. Janiewski, D. (1991) ‘Southern Honor, Southern Dishonor: Managerial Ideology and the Construction of Gender, Race and Class Relations in Southern Industry’, in A.Baron (ed.) Work Engendered: 70–91. Keene, A. (1916/17) ‘Production—A Dream Come True’, Journal of the Institute of Production Engineers, 11:385–99. Kessler-Harris, A. (1982) Out to Work, A History of Wage-Earning Women in the United States, New York: Oxford University Press. Kimmel, M.S. (1987a) ‘The Contemporary Crisis of Masculinity in Historical Perspective’, in H.Brod (ed.) The Making of Masculinities: The New Men’s Studies, Boston: Allen and Unwin: 121–53. ——(1987b) ‘The Cult of Masculinity: American Social Character and the Legacy of the Cowboy’, in M.Kaufman (ed.) Beyond Patriarchy: Essays by Men on Pleasure, Power and Change, Toronto: Oxford University Press: 235–49. Kossoudji, S.A. and Dresser, L.J. (1992) The End of a Riveting Experience: Occupational Shifts at Ford After World War II’, American Economic Review, Proceedings, 82:519–25. Labor Age (various dates). Land, H. (1980) ‘The Family Wage’, Feminist Review, 6:55–77. Leibowitz, L. (1986) ‘In the Beginning…: The Origins of the Sexual Division of Labor and the Development of the First Human Society’, in S.Coontz and P. Henderson (eds) Women’s Work, Men’s Property: The Origin of Gender and Class, London: Verso: 43–75. Lewchuk, W. (1987) American Technology and the British Vehicle Industry, Cambridge: Cambridge University Press. McClelland, K. (1989) ‘Some Thoughts on Masculinity and the “Representative Artisan” in Britain, 1850_1880’, Gender and History, 1:164–77. Marquart, F. (1975) An Auto Worker’s Journal: The UAW from Crusade to One-Party Union, University Park: Pennsylvania State University Press. May, M. (1982) ‘The Historical Problem of the Family Wage: The Ford Motor Company and the Five Dollar Day’, Feminist Studies, 8 (Summer): 399–424.

200

MEN AND MONOTONY

Mead, J.E. (1918) ‘Rehabilitating Cripples at Ford Plant’, Iron Age, 112 (26 September): 739–43. Meier, A. and Rudwick, E. (1979) Black Detroit and the Rise of the UAW, Oxford: Oxford University Press. Meyer, S., III (1981) The Five Dollar Day: Labor Management and Social Control in the Ford Motor Company, 1908–1921, Albany: State University of New York Press. Milkman, R. (1987) Gender at Work: The Dynamics of Job Segregation by Sex during World War II, Urbana: University of Illinois Press. Montgomery, D. (1979) Workers’ Control in America, Cambridge: Cambridge University Press. Nation (12 June 1929). Nevins, A. (1954) Ford: The Times, the Man, and the Company, New York: Charles Scribner’s Sons. New York Times (9 May 1928). Parr, J. (1990) Gender of Bread Winners: Women, Men, and Change in Two Industrial. Towns: 1880–1950, Toronto: University of Toronto Press. Phillips, A. and Taylor, B. (1980) ‘Sex and Skill: Towards a Feminist Economies’, Feminist Review, 6:79–88. Pollert, A. (1981) Girls, Wives, Factory Lives, London: Macmillan. Raff, D.M.G. (1988) ‘Wage Determination Theory and the Five-Dollar Day at Ford’, Journal of Economic History, 48 (June): 387–99. ——(1992) ‘The Puzzling Profusion of Compensation Systems in the Interwar Motor Vehicle Industry’, photocopy, Cambridge, Ma: Harvard University; 1–28. Raff, D.M.G. and Summers, L.H. (1987) ‘Did Henry Ford Pay Efficiency Wages?’, Journal of Labor Economics, 5 (October): S57—S86. Reeves, A.W. and Kimber, C. (1916/17) ‘Comments on a Paper Titled “Works Organization” ’, Proceedings, Institute of Automobile Engineers, 11:385–99. Reitell, C. (1924) ‘Machinery and Its Effects Upon the Workers in the Automobile Industry’, Annals of the American Academy of Political and Social Science, 116: 37–43. Rodgers, D.T. (1974) The Work Ethic in Industrial America, 1850–1920, Chicago: University of Chicago. Rose, S.O. (1987) ‘Gender Segregation in the Transition to the Factory: The English Hosiery Industry, 1850–1920’, Feminist Studies, 13 (Spring): 163–84. ——(1991) ‘From Behind the Women’s Petticoats: The English Factory Act of 1874 as a Cultural Production’, Journal of Historical Sociology, 4 (March): 32–51. ——(1992) Limited Livelihoods: Gender and Class in Nineteenth-Century England, Berkeley, CA: University of California Press. Rotundo, A.E. (1987) ‘Learning About Manhood: Gender Ideals and the MiddleClass Family in Nineteenth Century America’, in J.A.Mangan and J.Walvin (eds) Manliness and Morality: Middle Class Masculinity in Britain and America, 1800– 1940, Manchester: Manchester University Press: 35–51. Rumely, E.A. (1914) ‘Mr. Ford’s Plan to Share Profits’, World’s Work, 27:664–9. Russel, J. (1978) ‘The Coming of the Line’, Radical America, 12:28–46. Segal, L. (1990) Slow Motion: Changing Masculinities, Changing Men, London: Virago. Sinclair, U. (1937) The Flivver King: A Story of Ford—America, Detroit: United Automobile Workers of America. Thompson, P. (1988) ‘Playing at Being Skilled Men: Factory Culture and Pride in Work Skills Among Coventry Car Workers’, Social History, 13 (January): 45–69.

201

WAYNE A.LEWCHUK

Tolles, N.A. and LaFever, M.W. (1936) ‘Wages, Hours, Employment and Earnings in the Motor Vehicle Industry, 1934’, Monthly Labor Review (March): 521–53. US Bureau of the Census (1975) Historical Statistics of the United States, Washington, DC. US Department of Labor (1920) The New Position of Women in Industry, Bulletin of the Women’s Bureau, no. 12, Washington, DC. ——(1928) The Effects of Labor Legislation on the Employment of Women, Bulletin of the Women’s Bureau, no 65, Washington, DC. US Department of Labor, Bureau of Labor Statistics (1920) Industrial Survey in Selected Industries in the United States 1919, Wages and Hours of Labor series, no. 265, Washington, DC. ——(1934) Handbook of Labor Statistics, 1924–26, Miscellaneous series, no. 439, Washington, DC. Valverde, M. (1988) ‘Giving the Female a Domestic Turn: The Social, Legal and Moral Regulation of Women’s Work in British Cotton Mills, 1820–1850’, Journal of Social History, 21 (Summer): 619–34.

Archives and collections Engineering Employers Federation, Modern Record Centre, University of Warwick, Coventry. Ford Archives, Greenfield Village and Henry Ford Museum, Dearborn, Michigan. Reuther Library, Walter P., Wayne State University, Detroit, Michigan.

202

8

MANAGEMENT AND LABOUR IN GERMAN CHEMICAL COMPANIES BEFORE WORLD WAR ONE Sachio Kaku

Introduction The large-scale industrial enterprises that appeared in Europe and America in the last quarter of the nineteenth century have played an important role in economic development. Accordingly, there is a large and growing literature on this subject. It would be appropriate to say, however, that the most influential research contribution in this area is the work of the American historian Alfred D.Chandler. Chandler defines the modern business enterprise as having two basic characteristics: (a) it contains a number of distinct business units, operating in different locations, carrying on different types of economic activities and handling different lines of goods and services; and (b) in order to realise the advantages of such internalisation, a managerial hierarchy has been created. Combining this model of the modern business enterprise and detailed empirical analysis, Chandler has described the origins and growth of the modern industrial enterprise in the United States of America, England and Germany (Chandler 1977, 1990). His research is all the more outstanding because this field abounds with studies of single companies, single entrepreneurs and/or single aspects, and synthetic analysis is rare. We can usefully call this ‘the Chandler paradigm’ (Schmitz 1993:10–11). Chandler’s work has found a response among German historians. Jürgen Kocka takes a general view of the development and characteristics of modern German industrial enterprise on the basis of information on the hundred largest manufacturing and mining firms in 1887, 1907 and 1927, with the intention of placing the German case in comparative perspective 203

SACHIO KAKU

and of examining its relationship to Chandler’s findings and hypotheses. He points out the striking similarities in the development of modern corporations between Germany and the United States, as well as a sharp contrast between both of them and Britain. He also argues that the earlier German bureaucratic tradition helped firms in implementing principles of scientific organisation within the factory, and he stresses the relationship between the relative backwardness of an economy and the relative modernity of some of its large firms.1 In his discussion of the growth of German industrial enterprise since 1840, Richard Tilly divides the era into four periods: 1840–73; 1873–1914; 1914–45; and post-1945. Tilly argues that individual enterprises succeeded in attempts to gain control over market forces, but that this development robbed industrial capitalism of one of its more important stabilising mechanisms and increased the level of government intervention; that although the introduction of the corporation (Kapitalgesellschaft) into heavy industry in the 1850s established intra-enterprise division of labour between entrepreneurial operation and ownership, most of the large-scale enterprises have remained, until very recently, under the control of a single entrepreneur, or small groups of entrepreneurs; and that financial factors were a key to understanding the investment policies and growth of large-scale industrial enterprise.2 On the development of the structure and management of German large industrial enterprises at the turn of the century, Hans Pohl confirms the following tendencies: (a) through growing horizontal and vertical concentration, a series of large, complex enterprises were formed, which required a different organisational structure from that of personal enterprises in the period of the industrialisation; and (b) the internal and external growth of enterprises increased their need for capital. Thus many large enterprises changed from personal enterprises to corporations in order to satisfy their demand for capital with the help of banks and the public financial market. According to Pohl, however, the predominance of joint-stock companies among leading large enterprises does not directly mean the dominance of non-shareholding managers. Rather, he notes the coexistence of family and managerial enterprises. As the examples of AEG and Siemens show, it became crucial to adapt to changes in production and market structures (Pohl 1981). Comparing the research of these German historians with that of Chandler, some points attract our attention. First, they suspend judgement on Chandler’s hypotheses on the change from personal enterprises to managerial enterprises. Second, they also discuss the problems which Chandler excluded or dealt with only secondarily. One typical example is the financial dimension of industrial enterprises, especially the relationship between banks and industrial enterprises which Tilly discusses in detail. Third, German historians stress the social and economic background of enterprises, as Kocka’s analysis strikingly shows. Nevertheless, the description of the German historians of 204

MANAGEMENT AND LABOUR IN CHEMICAL COMPANIES

the structure of enterprise and managerial hierarchies is still too general or fragmentary. That is the point on which I would like to concentrate. Little research has been undertaken so far with regard to the organisational structure and other procedures through which top-level managers administered large integrated and diversified enterprises in Germany, although historians have a long and honourable tradition of accumulating archival materials and writing company histories and entrepreneurial biographies.3 Moreover there remain some areas the importance of which they acknowledge but have scarcely discussed. It seems to me that consideration of questions of industrial relations and of problems in the workplace are indispensable to understanding the history of modern industrial enterprises. To do so, we need to carry out more detailed empirical studies, examine the hypotheses on which they are undertaken, and reform them if required. Through such a ‘round trip’ between empirical studies and modelling hypotheses, we can further advance the historical study of the modern industrial enterprise. The aim of this paper is, as a first step, to make clear the structure of management and labour of large German chemical companies before the World War One. Although the German chemical industry is one of the fields in which large-scale enterprises emerged, it has rarely been researched.

Development of the German coal-tar dyestuffs industry The coal-tar dyestuffs industry, as a new branch of the chemical industry, was born in England, with the development of aniline purple or mauve which William Perkin (1838–1907) of the Royal College of Chemistry discovered in 1856.4 In Germany, as well, many coal-tar dyestuffs enterprises were founded in the 1860s. The scale of these enterprises was generally very small. For example, in addition to the three founders, the Hoechst Company (Meister Lucius & Co., or Farbwerke vorm. Meister, Lucius & Brüning) consisted of five factory workers, one chemist and an office girl, when it was established in 1863 in Hoechst, near Frankfurt am Main, for the purpose of producing aniline dyestuffs. The work within the enterprise was personally divided at the beginning. Two of the founders, Eugen Lucius and Adolf Brüning, worked every day in the laboratory for eight-ten hours, while the other founder, Wilhelm Meister, worked with the commercial staff in the office. The working day was twelve hours long, shorter than the standard in those days. One worker, Johann Barthel, was called the allmaker (Allesmacher) by his colleagues because he was not only a coachman, a porter and a nightman, but also substituted as a bath attendant on occasions. A similar situation prevailed at the Bayer Company (Friedr. Bayer & Co.) which was founded in Barmen in 1863 (Bäumler 1988:48–49, 57; Bayer Magazin 1988:9–11). The coal-tar dyestuffs industry grew very rapidly by developing new

205

SACHIO KAKU

products one after another and deriving natural and older synthetic dyestuffs from them. Such a development was accompanied by a fall in price. Thus the ability to develop new products was the key to the growth of coal-tar dyestuffs firms. Success in producing synthetic alizarin on an industrial scale introduced a new era in the history of the coal-tar dyestuffs industry. Following BASF (Badische Anilin-und Soda-Fabrik), Perkin and the Hoechst Company, which first produced alizarin, many other enterprises followed. Output increased rapidly and exceeded that of natural madder by the middle of the 1870s. This increase was accompanied by a sharp fall in price. In Germany, the price per kilogram of 100-per-cent alizarin fell from 270 M. in 1869 to 200 M. in 1870, 120 M. in 1872–3, and 23 M. in 1878. This extreme price decline made the cultivation of madder practically extinct and caused severe competition among the enterprises which produced alizarin. Only the enterprises that could continue to raise productivity survived (Kockershied 1905:68–9). In this way, the coal-tar dyestuffs industry took on a new look in the 1870s. First, the German coal-tar dyestuffs industry already held a dominant position in the world market in this period. Germany accounted for more than 30 million M. of the total world output of coal-tar dyestuffs (about 53 million M.) in the middle of the 1870s, while output in England was 9 million M. and in France and in Switzerland 7 million M. each (Pinnow 1938:33). Second, dyestuffs enterprises became larger and larger in scale and more and more technologically intensive, which made the barriers to entry ever higher. Several enterprises tried to produce coal-tar dyestuffs, but were unsuccessful in their business. Here, only one case will be discussed. The Verein Chemischer Fabriken in Mannheim, stimulated by the excellent results of the dyestuffs industry, began to produce coal-tar dyestuffs in 1879, but the quality of its products was not satisfactory. Efforts at improvement did not bear fruit and the enterprise had to stop producing coal-tar dyestuffs in 1884 (Caro 1904:1355; Borschied 1976:188). Thus the coal-tar dyestuffs industry was now comprised mainly of large enterprises. Above all, the position of the three largest enterprises—BASF, the Hoechst Company and the Bayer Company—was prominent, and they were already among the hundred largest industrial enterprises of Germany in 1887. Each company employed approximately nine—ten thousand people around 1910, and the shareholders’ capital amounted to 36 million M. each in 1913 (Kocka 1979:74_5; Redlich 1914:35, 60). Third, however, such large enterprises themselves had to solve various problems: how to attract capital, how to develop a workable research and development system, how to structure the sales organisation, how to recruit workers, how to form a managerial organisation, and how to create a

206

MANAGEMENT AND LABOUR IN CHEMICAL COMPANIES

disciplined work environment. In this chapter, I deal only with the last two of these questions.

The managers of the German coal-tar dyestuffs companies The joint-stock company (Aktiengesellschaft), which was legally established in the second third of the nineteenth century, became the principal legal form of large firm in Germany. By 1887, seventy-nine of the hundred largest industrial enterprises had adopted the joint-stock company form (Kocka and Siegrist 1979:80). The predominance of this form is explained by their demand for capital. In the coal-tar dyestuffs industry, as well, large enterprises had become joint-stock companies by the beginning of the 1880s. BASF started as a joint-stock company in 1865, the Hoechst Company was reorganised as a joint-stock company with capital of 10 million M. in 1880, as was the Bayer Company with capital of 5.4 million M. in 1881.5 Table 8.1 shows that the peak period for the formation of joint-stock companies in the German coal-tar dyestuffs industry was the 1870s in regard to both the number of the companies and the total amount of capital employed. The three largest companies increasingly issued shares and debentures, and their position was overwhelming in this branch. But the banks did not exert direct influence on the coal-tar dyestuffs companies, in contrast, for example, to the iron and steel industry. This is because it was an obligee industry (Gläubigerindustrie) (Riesser 1910:547); the coaltar dyestuffs companies were not in debt to banks but rather held large deposits with them. As a reflection of this, there were few bank directors on the supervisory boards of the three largest German coal-tar dyestuffs companies before World War One. German corporation law after 1870 prescribed a dual board structure, that is an executive board (Vorstand) and a supervisory board (Aufsichtstrat) (Loeb 1902). In the three largest German coal-tar dyestuffs companies, the founders predominated on these boards at first, and their positions were often subsequently filled by members of their families. At the time of its reorganisation as a joint-stock company, three founders formed the supervisory board of the Hoechst Company. When Eugen Lucius, who had no son, resigned as the chairman of the supervisory board in 1902, he was succeeded by Walter vom Rath, W.Meister’s son-in-law. After Adolf von Brüning died in 1884, his oldest son, Gustav, who had studied chemistry, was elected to a seat on the supervisory board in 1889 in place of his father. He became a member of the executive board in 1899, and was its president from 1908 until his death in 1913. He was succeeded by Herbert von Meister, Wilhelm’s second son, who had had a seat on the executive board since 1898, after studying chemistry. Adolf and Walter von Brüning, the second and third sons of Adolf, and Graf Louis Schimmelpenninck-

207

SACHIO KAKU

Table 8.1 Issuing of stocks and debentures of the German coal-tar dyestuffs companies (millions of Marks)

Source: S.Kaku, ‘Die finanzielle Entwicklung der deutschen Teerfarbenunternehmen in den Jahren von 1880 bis 1913’, Scripta Mercaturae, XXIII (1989), 145. Note: ()=number of joint-stock companies.

Nyenhuis, husband of the only daughter of Eugen Lucius, were also members of the supervisory board in 1913, but it is not clear if they concerned themselves with the management of the company.6 When the Bayer Company was reorganised as a joint-stock company in 1881, its executive board consisted of Friedrich Bayer, Jr and Friedrich Weskott, Jr, who were both the sons of the founders and entered the company in 1877 as joint owners. Besides them, Carl Rumpff, Friedrich Bayer, Sr’s son-in-law, August Sillar, the son-in-law of Friedrich Weskott, Sr, and Eduard Tust, the only non-family manager, comprised the supervisory board. Henry Theodor Böttinger, who was also the son-in-law of Friedrich Bayer, Sr, was a member of the executive board from 1883 until 1906, and joined the supervisory board in 1907 (Plumpe and Scultheis 1988:63–4). However, the descendants of the founders lost power on the supervisory and executive boards by and by, and non-family specialists such as chemists and lawyers increasingly joined the boards. The executive board of the Hoechst Company was, at its reorganisation as a joint-stock company, composed of A.de Ridder, who had rendered great services to the development of the company as the sales manager since its foundation, and Carl König, who entered the company as a chemist in 1869. Gustav Adolf Dichl, who was in charge of accounting for the company, was promoted from the manager legally empowered to act as a company (Prokurist) to a directorship and then became a member of the executive board in 1907. August Laubenheimer, a university professor of chemistry, entered the Hoechst Company in 1883 and established its pharmaceutical department. He was nominated as a member of the executive board in 1887, after the death of Carl König, and of the supervisory board in 1904 (Bäumler 1988:73–5, 95, 126, 145, 175). It is well known that Carl Duisberg, who entered the Bayer Company as a chemist in 1883, contributed significantly to the growth of the company; he was accepted

208

MANAGEMENT AND LABOUR IN CHEMICAL COMPANIES

on to the executive board in 1900 (Duisberg 1933:26; Flechtner 1959:64, 172). The shift of weight from the founders towards the non-family managers occurred much earlier in BASF than in the other two companies. After Friedrich Enghorn and Carl Clemm, who were both founders of the company, resigned their positions in 1883 over a confrontation with other managers concerning the manufacturing of synthetic indigo, Heinrich Brunck and Carl Glaser, who were chemists and entered BASF in 1869, controlled the firm’s management.7 But it must also be kept in mind that matrimonial alliances were formed not only between the families of the founders but also between those of non-family managers, as well as between the former and the latter. Three of Carl König’s four daughters married members of the Hoechst Company: Frieda and Amalie, chemists; and Louisa, a lawyer, Adolf Haeuser, who entered the Hoechst Company in 1889 and received a seat on the executive board in 1907 (Bäumler 1988:75). Carl Duisberg married the niece of Carl Rumpff (Duisberg 1933:40–1; Flechtner 1959:92–5)

Terms of labour With the increasing size of the companies, the design of a managerial organisation became a major problem. The central administration building of the Hoechst Company, which was erected in 1893, housed top management and the officers of the accounting, sales, research and development and legal departments. In another building was the office of personnel affairs. This enables us to guess that the Hoechst Company was at that time functionally organised (Bäumler 1988:174–8). The Bayer Company not only came to produce many kinds of dyestuffs but also added departments for handi-craftsmen (1874), small dyeworks (1880), the drawing office (1886), work-in-progress (1886), the central laboratory (1887), etc. A functionally organised structure was formed, conforming to the ‘Memorandum on the Construction and Organisation of the Dye Works at Leverkusen’ (‘Denkschrift über den Aufbau und Organisation der Farbenfabriken zu Leverkusen’) written by Carl Duisberg in 1895. This structure remained little changed until the formation of IG-Farben in 1925 (Plumpe and Schultheis 1988:45–89, 110–15; Chandler 1990:475–7). It was also crucial to the development of these companies that they be able to deal with personnel matters, because management was becoming increasingly more distant from the workers and personal contact between the extremes of the managerial hierarchy was lost. We can outline the managerial hierarchy within the factory as in Figure 8.1, using as clues the few materials which remain, such as the description left by W.Grandhomme and others (Grandhomme 1880:8). In the Hoechst Company, the hiring and firing of workers was decided by

209

SACHIO KAKU

Figure 8.1 Managerial hierarchy within the coal-tar dyestuffs factory

the factory office (Fabrik-Comptoir) from the beginning of the 1880s, although a part of the investigation at the time of hiring was in the charge of the overseers (Aufseher) (Grandhomme 1880:46). In the provisions for the workers of the Bayer Company in 1902 (Vorschriften für Arbeiter der Farbenfabriken vorm. Friedr. Bayer & Co. vom 15 März 1902), we find the following description: The factory managers (Betriebsführer), not the foremen (Meister) or the overseers, are empowered to hire the workers in all the workshops including those of handicrafts…. We direct that only the factory office may accept applications for jobs hereafter, in order to introduce a better register system of the workers and handi-craftsmen. (Jacobi 1963:444–5) The Bayer Company set up committees on labour management (Ausschuss für Arbeiterangelegenheiten) at the Leverkusen and Elberfeld factories in 1904, after settling a strike. Their task was to hire and fire workers as well as to deal with the complaints of the workers (Bayer 1911:43–6). By the 1880s, the German large dyestuffs companies had laid down their factory or working rules. The first set of factory rules of the Hoechst Company, enacted in 1880, had been revised seven times by the outbreak of World War One.8 At the Bayer Company, the first set of factory rules was established in 1888 and revised in 1899, 1905 and 1911.9 Through these revisions, the rules were rearranged, expanded and increasingly systematised. Roughly speaking, the factory or working rules of both companies consisted of provisions on the payment of wages, working hours and discipline within the factory.

210

MANAGEMENT AND LABOUR IN CHEMICAL COMPANIES

The working rules of the Hoechst Company contained a provision dealing with the payment of wages for the first time in 1891. Article 10 states: When piece rates have not been introduced, wages are calculated on working days. The wages are paid on a fixed day of the week for each department by the settlement payment office. Article 9 of the working rules of 1902 added the provision that ‘the rate of wages is agreed when a worker enters the company, and the piece rate is determined before the commencement of work.’ Subsequently, the Hoechst Company endeavoured to establish a systematic set of wage rates. The fruit of this was the wage table (Lohnskala) which was drawn up in 1906 and revised thereafter. In these tables, the wages were determined according to the occupation, age and period of service of the workers, as is shown in Table 8.2. In the factory rules of 1888 the Bayer Company also provided that wages were based on shift and hourly rates or on piece-work, and that contributions to health insurance, fines and deposit money were to be deducted from the wages. In 1905, a provision was added that the rate of wages by the hour or by the shift was agreed at the time of engagement or later, while, in the case of the piece-work, a customary rate for the work applied, or the rate of wages was agreed before the commencement of work (article 7 of the factory rules for 1905). According to the working rules of 1911, wage rates were similarly determined at BASF.10 The provision on working hours in the factory rules of 1880 of the Hoechst Company was very simple and stated only that ‘bells announce to the workers the beginning and the end of the working hours in the morning and in the evening’. But article 6 of the factory rules of 1891 prescribed the working hours clearly: The working day begins at 6 a.m. and ends at 5 p.m. In the continuously-operating workshop, [the working day] is from 6 a.m. until 6 p.m., or from 6 p.m. until 6 a.m.… The rest periods are 8– 8.30 or 8.30–9 in the morning and in the evening (in the night workshop), and 12–1 or 1–2 in the afternoon and at night. Article 4 of the working rules of 1910 went into more detail and provided shorter working hours for female workers than for male workers. The working day for the female workers began at 6.45 a.m. and ended at 4.45 p.m. The provisions on working hours in the factory rules of the Bayer Company and those of BASF are shown in Table 8.3. As we have seen, the fundamental issues of working conditions—such as payment of wages and working hours—were expressly stipulated in the text of the working rules by the beginning of the twentieth century at the latest 211

SACHIO KAKU

Table 8.2 Wage rate table of the Hoechst Company in 1912 (Pfg/hour)

212

MANAGEMENT AND LABOUR IN CHEMICAL COMPANIES

Table 8.3 The working hours of the Bayer Company and BASF

in the German large coal-tar dyestuffs companies, and the responsibility for these matters lay with the central offices. The factory rules of the Hoechst and the Bayer Companies also covered punctuality, manners in the working place, safety devices, measures in case of disorder, abstinence from alcohol and tobacco, fines, etc., and indicated

213

SACHIO KAKU

which supervisory personnel were responsible for keeping discipline as specified in the rules. According to the factory rules of 1880 of the Hoechst Company, the workers were directly responsible to an overseer whose orders they had to obey willingly. Thus the workers informed the overseers if they were late or absent, and no worker might leave the workplace without the permission of the overseers. Furthermore, ‘if the machines or implements break down, the worker immediately has to make a report to his overseer, who is solely able to indicate requirements as to the repair.’ These provisions remained virtually intact up until World War One. In the Bayer Company, the factory rules of 1888 provided that the workers were subordinate to the senior staff of the factory (vorgesetzte Betriebsbeamte) and stated that ‘If any kind of machine or implement malfunctions, or is broken, the worker has to make a report immediately to his overseer or his factory manager about the occurrence. Only the overseer or the factory manager can bring about the measures as to the required repair.’ These provisions were continued with modifications in the factory rules of 1899, 1905 and 1911. Thus the competency of the overseers or the foremen was more limited in the Bayer Company than in the Hoechst Company, and the role of the factory manager at Bayer was much larger. BASF’s working rules of 1911 had the following provisions on discipline within the factory: the workers were hired and fired by board of management (Direktion) or by the men it delegated; the workers had to obey all the orders of their chief (Vorgesetzte) conscientiously, and decisions on fines or firing were made by the manager of the workshop if the workers violated the rules. At all times, the overseers or the factory manager occupied the key point of day-to-day contact between management and labour in the German coaltar dyestuffs companies (Duisberg 1918:641–2).

Company welfare facilities While the provisions of the factory or working rules outlined the framework of the relationship between management and labour, they did not reveal its underlying realities. In the Hoechst Company, the daily wages ranged very widely from less than 1 M. to 5 M. in 1893, but 87 per cent of the workers belonged to the stratum from 2.1 to 4.0 M., with a concentration of 43 per cent between 2.1 to 3.0 M. (Grandhomme 1893:36). At BASF, about 70 per cent of the workers earned wages of less than 45 Pfennig per hour in 1906/7 (Drösser 1908:216– 17). There are some indications that these wages were not enough for many workers in German coal-tar dyestuffs companies to make a living, even if various kinds of bonuses were added.11 As a result they had to depend not

214

MANAGEMENT AND LABOUR IN CHEMICAL COMPANIES

only upon their wages but also upon small farming, the wages of their wives and grown-up children, and their own jobs on the side. This also increased the attractiveness of overtime work, and the welfare facilities of the companies came to be important for the life of the workers. In fact, overtime was very frequent in the German coal-tar dyestuffs companies. Officials of the Hoechst Company reported that they found difficulty in making out a uniform wage rate table, because the greater part of the workers did overtime.12 At BASF, the lengthening of the working hours by 25 per cent was so common that the phrase ‘a quarter-creation’ (Viertel-Schaffen) was used in Ludwigshafen (Breunig 1976:122–3). The welfare facilities which the German coal-tar dyestuffs companies established were broadly divided into three groups: dwellings, health insurance and pensions, and other facilities. By 1909, the Hoechst Company had built 44 dwellings for the staff, 662 for the overseers and the workers, and a dormitory for single workers. The rent of the dwellings for the overseers and the workers was 2.3–4.2 M. weekly. This was equivalent to about 15 per cent of the earnings of the workers. Furthermore, W.Meister and E.Lucius and the supervisory board of the company donated money to build houses which were let out without rent for the overseers and the workers who had served in the company for more than twenty years. In 1913, 13 per cent of the workers lived in the company houses or the houses which were built by the fund the company provided (Hoechst 1910:27–34). In the Leverkusen factory of the Bayer Company, about 1,200 dwellings built before 1913 were occupied by more than one fifth of the workers (Pollay 1952:212–16; Jacobi 1963:437). In 1871, even earlier than its two main competitors, BASF began to build houses close to the factory and had 663 dwellings at the beginning of this century.13 The German coal-tar dyestuffs companies established various funds and gave relief to sick and invalid workers, widows, and orphans from the 1870s. Here we will mainly look at the case of the Hoechst Company,14 where the sickness benevolent fund was set up in 1876 and was managed by the company. If a worker who served for more than six weeks in the company fell sick and could not work, 50 per cent of his wages was paid for four days (from the third day through the sixth) and 60 per cent for six weeks thereafter. Three-quarters of his wage was paid to a worker who was injured during work. The fee for medical treatment was waived in both cases. Ten years before the passage of the old and invalid pension law, three founders of the Hoechst Company donated 150,000 M. to establish the KaiserWilhelm-Augusta fund in 1878, in memory of the Emperor’s golden wedding anniversary. Its aim was to loan money to overseers and workers in case of temporary hardships or for the purchase of property, and to give relief to invalids, widows and orphans. It was available to the overseers, to workers who had served more than five years, and to their families. These funds of the Hoechst Company were reorganised and expanded, 215

SACHIO KAKU

partly with the enactment of the social insurance laws in the 1880s. In this regard, it should not be overlooked that the administrative authorities supervised the management of these funds, and that the benefits they provided were more than the amount dictated by the provisions of the law. Furthermore, the Hoechst Company and its founders donated 30,000 M. and 50,000 M. respectively, in order to establish a superannuation fund for the staff of the company in 1886. The staff, who contributed 3 per cent of their salaries to the fund, received a lifetime pension if they retired after service of more than 30 years and were more than 60 years old, or if they became invalids. Widows and orphans also received pensions. BASF and the Bayer Company established similar relief funds for their employees from the 1870s, and expanded them thereafter in the same way as the Hoechst Company.15 Moreover, the three companies provided the following welfare facilities for employees and their families: a messroom, cooperative store and savings bank (Sparkasse) to improve life; a bath house, dispensary, sanatorium and a maternity home to promote health; educational and training facilities such as a school of domestic science, an evening school and an apprentice school; and facilities for recreation.16 The welfare facilities that were built by the German coal-tar dyestuffs companies were of deep significance for the lives of their workers. Their value was increased because they played an important role in anchoring the workers and/or in raising their sentiments of loyalty, because, in contrast to wages, they were characterised by an attitude of caring, even if it was paternalistic.

Conclusion The following points standout from the evidence I have presented: 1 In all three companies under review, non-family managers had become common at the levels of the executive and supervisory boards by World War One. 2 Functionally organised structures were developed in parallel with the growth in the size and complexity of the companies. 3 Aspects of industrial relations, such as the hiring and firing of workers and decisions regarding working conditions, became increasingly systematically regulated by the central offices as the enactment and subsequent development of factory and working rules demonstrates. Thus the development of the organisational structures of the chemical companies was well advanced by the beginning of World War One. 4 On the other hand, personal connections and interrelationships continued to play an important role, as matrimonial alliances among entrepreneurs and the development of welfare facilities show. It should also be noted that the on-the-job relationship between overseers and foremen, on the one hand, and the workers under them, on the other, remained more traditional than might be imagined, even

216

MANAGEMENT AND LABOUR IN CHEMICAL COMPANIES

after overseers and foremen had been stripped of many of their powers and functions.

To fully understand the importance of these findings, however, further work is necessary to place them in their proper historical and comparative perspective by analysing other industries and other periods and by further comparisons with conditions in other countries.

Notes 1 See Kocka (1980). See also, Kocka (1969, 1975, 1981a, 1981b, 1990). 2 Tilly (1974, 1993). See also, Tilly (1978, 1982). 3 See Jaeger (1974) as well as Tilly (1978:153); Pohl (1981:143); Feldenkirchen (1987); Dornseifer (1991, 1993). 4 For a good overview of the history of the chemical industry in the nineteenth century, see Haber (1958). For the history of the German coal-tar dyestuffs industry, see Beer (1959) and Kaku (1980). 5 Zur Feier des 25 jahrigen Bestehens der Farbwerke vormals Meister, Lucius, & Brüning zu Hoechst am Main 2. Juni 1888:7–11; Farbwerke vorm. Meister, Lucius & Brüning, 1863–1913:19; Pinnow (1938:46); Plumpe and Schultheis (1988:64– 6). 6 Fischer (1958, 1961), Bäumler (1988:157–8, 160). 7 Leber (1959). See also Schuster (1973, 1976:67) Borscheid (1976:174–6). 8 All the factory or working rules of the Höchst Company from 1880 until 1972 are given in Hromadka (1979). 9 The factory rules of the Bayer Company which are used in this paper are stored in its company archive. According to Pohl et al. (1983:15), in the Bayer Company, the factory rules for the individual factories were already established in the 1870s. 10 The working rules (Arbeits-Ordnung) of BASF were put into practice on 1 January 1911. 11 Ehrhart (1892). See as well Grandhomme (1880:41, 1893:33); Breunig (1976:119); Redlich (1914:62–6); P.Borsheid (1976:237). 12 An in-house memo dated 10 May 1906 (materials in the archive of the Hoechst Company; no reference number.) 13 Badische (1923:113–19); Im Reich (1965:77); Breunig (1976:68–71). 14 Grandhomme (1880:63–4, 1883:72–8, 86–93, 1896:55, 63–6); Hoechst (1910:13– 20, 75–89); Richter (1993). 15 Badische (1923:99–100, 130–1); Im Reich (1965:69); Breunig (1976:131); Schwellwien (1913); Bayer (1913:9–10). 16 Grandhomme (1896:41–2, 49–50); Hoechst (1914:5–10, 23–4, 109–13, 135–9); Badische (1923:94, 100–13); Breunig (1976:130); Schwellwien (1913:682–3); Bayer (1913:2–6).

References Badische (1923) Die Badische Anilin-&-Soda-Fabrik, Ludwigshafen a. R.: BASF. Bäumler, E. (1988) Die Rotfabriker. Familiengeschichte eines Weltunternehmens, München and Zurich: Piper.

217

SACHIO KAKU

Bayer (1911) Handbuch für die Chumiker u. Ingenieure der Farbenfabriken vorm. Friedr. Bayer & Co., Elberfeld: Friedr. Bayer & Co. Bayer (1913) Die Wohlfahrtseinrichtungen der Farbenfabriken vorm. Frieder. Bayer & Co. Leverkusen bei Cöln, Halle a. S.: Friedr. Bayer & Co. Bayer Magazin, Jubiläumsausgabe (1988). Beer, J.J. (1959) The Emergence of the German Dye Industry, Urbana: University of Illinois Press. Borscheid, P. (1976) Naturwissenschaft, Staat und Industrie in Baden (1848–1914), Industrielle Welt, 17, Stuttgart: Ernst Klett. Breunig, W. (1976) Soziale Verhältnisse der Arbeiterschaft und sozialistische Arbeiterbewegung in Ludwigshafen am Rhein 1869–1919, Ludwigshafen a. R.: Archives of the city. Caro, H. (1904) ‘Uber die Entwicklung der chemischen Industrie von MannheimLudwigshafen a. Rh.’, Zeitschrift für angewandte Chemie XVII. Chandler, A.D. Jr (1977) The Visible Hand: The Managerial Revolution in American Business, Cambridge, Ma.: The Belknap Press of Harvard University Press. ——(1990) Scale and Scope: The Dynamics of Industrial Capitalism, Cambridge, Ma.: The Belknap Press of Harvard University Press. Dornseifer, B. (1993) ‘Zur Bürokratisierung deutscher Unternehmen im späten 19. und friihen 20. Jahrhundert’, Jahrbuck für Wirtschaftsgeschichte, 1:69–91. Drösser, E. (1908) Die technische Entwicklung der Schwefelsäurefabrikation und ihre wirtschaftliche Bedeutung, Leipzig: W.Klinkhard. Duisberg, C. (1918) ‘Selbsterlebtes und Schlussbetrachtungen’, Geschichte und Entwicklung der Farbenfabriken vorm. Friedr. Bayer & Co., Elberfeld in den 50 Jahren, unpublished. ——(1933) Meine Lebenserinnerungen, Leipzig: Philipp Recalm Junior. Ehrhart, E.J. (1892) Die Zustände in der Badischen Anilin-und Soda-Fabrik, Mannheim: Mannheim Actiendruckerei. Feldenkirchen, W. (1987) ‘Big Business in Interwar Germany. Organizational Innovation at Vereinigte Stahlwerke, IG.Farben, and Siemens’, Business History Review, LXI: 417–51. Fischer, E. (1958) ‘Meister, Lucius und Brüning, die Gründer der Farbwerke Hoechst AG’, Tradition, III: 65–78. ——(1961) ‘Die Gründer der Farbwerke Hoeschst AG. Eugen Lucius, Wilhelm Meister, Adolf Brüning’, Nassauishe Lebensbilder, 6, 248–62. Flechtner, H.-J. (1959) Carl Duisberg: Vom Chemiker zum Wirtschaftsführer, Düsseldorf: Econ. Grandhomme, W. (1880) Die Theerfarben-Fabriken der Herren Meister, Lucius & Brüning zu Hoechst a. M. in sanitärer und socialer Beziehung, 1st edn, Berling: L. Schumacher. ——(1883) Die Theerfarben-Fabriken der Actiengesellschaft Farbwerke vorm. Meister Lucius & Brüning zu Hoechst a. M. in sanitärer und socialer Beziehung, 2nd edn, Heidelberg: Gustav Köster. ——(1893) Die Fabriken der Aktien-Gesellschaft Farbwerke vorm. Meister, Lucius & Brüning zu Hoechst a. M. in sanitären und socialer Beziehung, 3rd edn, Frankfurt a. M.: Mahlau and Waldschmidt. ——(1896) Die Fabriken der Actien-Gesellschaft Farbwerke vorm. Meister, Lucius & Brüning zu Hoechst a. M. in sanitärer und socialer Beziehung, 4th edn, Frankfurt a. M.: Mahlau and Waldschmidt. Haber, L.F. (1958) The Chemical Industry during the Nineteenth Century. A Study of the Economic Aspect of Applied Chemistry in Europe and America, Oxford: Clarendon Press.

218

MANAGEMENT AND LABOUR IN CHEMICAL COMPANIES

Hoechst. (1910) Die Wohlfahrtseinrichtungen der Farbwerke vorm. Meister Lucius & Brüning Hoechst a. M., Hoechst a. M.: Hoechst Company. Hoechst. (1914) ‘Farbwerke vorm. Meister Lucius & Brüning Hoechst am Main’, Bericht des Vorstandes und Aufsichtsrates an die Generalversammlung vom 9. Mai 1914. Hromadka, W. (1979) Die Arbeitsordnung im Wandel der Zeit am Beispiel der Hoechst AG, Köln, Berlin, Bonn and München: Heymann. Im Reich der Chemie (1965), Düsseldorf and Vienna: Econ. Jacobi, F. (1963) ‘Vorschriften’, Beiträge zur hundertjährigen Firmengeschichte 1863– 1963, Leverkusen: Friedr. Bayer and Co. Jaeger, H. (1974) ‘Business History in Germany: A Survey of Recent Developments’, Business History Review XLVIII: 28–48. Kaku, S. (1980) ‘The Development and Structure of the German Coal-Tar Dyestuffs Firms’, in Okochi, A. and Uchida, H. (eds) Development and Diffusion of Technology, The International Conference on Business History 6, Tokyo: University of Tokyo Press: 77–94. Kocka, J. (1969) ‘Industrielles Management: Konzeptionen und Modelle in Deutschland vor 1914’, Vierteljahrschrift für Sozial-und Wirtschaftsgeschichte LVI: 332–72. ——(1975) ‘Expansion-Integration-Diversifikation. Wachstumsstrategien industrieller Großunternehmen in Deutschland vor 1914’, in H.Winkel (ed.) Vom Kleingewerbe zur Großindustrie, Schriften des Vereins für Sozialpolitik, new ser., LXXXIII, Berlin: Duncker and Humblot: 203–26. ——(1980) ‘The Rise of the Modern Industrial Enterprise in Germany’, in A.D. Chandler Jr and H.Daems (eds) Managerial Hierarchies. Comparative Perspectives on the Rise of the Modern Industrial Enterprise, Cambridge, Ma.: Harvard University Press: 77–116. ——(1981a) ‘Großunternehmen und der Aufstieg des Manager-Kapitalismus im späten 19. und frühen 20. Jahrhundert. Deutschland im internationalen Vergleich’, Historische Zeitschrift CCXXXII: 39–60. ——(1981b) ‘Capitalism and Bureaucracy in German Industrialization before 1914’, Economic History Review, 2nd ser., XXXIV: 453–68. ——(1990) ‘Germany: Cooperation and Competition’, Business History Review LXIV:711– 16. Kocka, J. and Siegrist, H. (1979) ‘Die Hundert großten deutschen Industrieunternehmen im späten 19 und frühen 20. Jahrhundert’, in N.Horn and J.Kocka (eds) Law and the Formation of the Big Enterprises in the 19th and 20th Centuries, Kritische Studien zur Geschichtswissenschaft 40, Göttingen: Vandenhoeck and Ruprecht: 55–122. Kockershied, J.W. (1905) Uber die Preisbewegung Chemischer Produckte, Jena: Gustav Fischer. Leber, A. (1959) ‘Engelhorn, Friedrich’, Neue Deutsche Biographie 4, Berlin: Duncker and Humblot: 514–15. Loeb, E. (1902) ‘Das Institut des Aufsichsrats, seine Stellung und Bedeutung im deutschen Aktienrecht und der deutschen Volkswirtschaft, die Notwendigkeit und Möglichkeit seiner Reform’, Jahrbücher für Nationalökonomie und Statistik, 3rd ser., XXIII: 1– 28. Pinnow, H. (1938) Werksgeschichte. Der Gefolgschaft der Werke Leverkusen, Elberfeld und Dormagen zur Erinnerung an die 75. Wiederkchr des Grundungstages der Farbenfabriken vorm. Friedr. Bayer & Co., München: Brunckmann. Plumpe, G. and Schultheis, H. (1988) Meilensteine. 125 Jahre Bayer 1863–1988, Leverkusen: Friedr. Bayer and Co. Pohl, H. (1981) ‘Zur Geschichte von Organisation und Leitung deutscher

219

SACHIO KAKU

Großunternehmen seit dem 19. Jahrhundert’, Zeitschrift für Unternehmensgeschichte XXVI: 143–78. Pohl, H., Schaumann, R. and Sconert-Rohlk, F. (1983) Die chemische Industrie in den Rheinlanden während der industriellen Revolution. 1, Die Farbenindustrie, Zeitschrift für Unternehmensgeschichte, Supplement 18, Wiesbaden: Franz Steiner. Pollay, K. (1952) Die wirtschaftliche Entwicklung der Stadt Leverkusen, dissertation, Universität Köln. Redlich, F. (1914) Die volkswirtschaftliche Bedeutung der deutschen Teerfarbenindustire, Staats-und sozialwissenschaftliche Forschungen hrsg. v. G.Schmoller/M.Sering, No. 180, München and Leipzig: Duncker and Humblott. Richter, P. (1993) ‘Die Kaiser Wilhelm-und August-Stiftung’, Dokumente aus HoechstArchiven, 53, Frankfurt a. M.: Hoechst Company. Riesser, J. (1910) Die deutsche Großbanken und ihre Konzentration im Zusammenhang mit der Entwicklung der Gesamtwirtschaft in Deutschland, 3rd edn, Jena: Gustav Fischer. Schmitz, C.J. (1993) The Growth of Big Business in the United States and Western Europe 1850–1939, Basingstoke and London: Macmillan. Schuster, C. (1973) Vom Farbenhandel zur Farbenindustrie, Schriftenreihe des Unternehmenarchivs der BASF Aktiengesellschaft, 11, Ludwigshafen: BASF. ——(1976) Wissenschaft und Technik, Schriftenreihe des Unternehmensarchivs der BASF Aktiengesellschaft, 14, Ludwigshafen: BASF. Schwellwien, J. (1913) ‘Die sozialen Einrichtungen der Farbenfabriken vorm. Friedr. Bayer & Co., Jahrbücher für Nationalökonomie und Statistik, 3rd ser., IXVI: 678– 87. Tilly, R. (1974) ‘The Growth of Large-Scale Enterprise in Germany since the Middle of the Nineteenth Century’, in H.Daems and H.van der Wee (eds) The Rise of Managerial Capitalism, Louvain and The Hague: Martinus Nijhoff: 145–69. ——(1978) ‘Das Wachstum industrieller Großunternehmen in Deutschland 1880–1911’, in H.Kellenbenz (ed.) Wirtschaftswachstum, Energie und Verkehr vom Mittelalter bis ins 19. Jahrhundert, Forshungen zur Sozial-und Wirtschaftsgeschichte, 22, Stuttgart and New York: Gustav Fischer: 153–82. ——(1982) ‘Mergers, External Growth, and Finance in the Development of LargeScale Enterprise in Germany, 1880–1913’, Journal of Economic History XLII: 629– 58. ——(1993) ‘Großunternehmen: Schlüssel zur Wirtschafts-und Sozialgeschichte der Industrieländer?’, Geschichte und Gesellschaft XIX: 145–69.

220

9

BUDDENBROOKS REVISITED The firm and the entrepreneurial family in Germany during the nineteenth and early twentieth centuries Dirk Schumann

The family firm has received mixed reviews in economic history. For a long time it has been portrayed as an obsolete form of organisation which retarded economic growth in the advanced stages of industrialisation. Recently, the picture has changed somewhat. Now, while its dynamic influences on economic development are stressed, the perseverance and flexibility of the family firm are pointed out. The purpose of this paper is to place the recent findings on the German family firm in the context of this debate. Particular emphasis will be put on the results of studies dealing with entrepreneurial history in the framework of the history of the whole German Eürgertum (the middle classes). My argument is that the family firm and the entrepreneurial family continued to play an important role in the German economy after the first stage of industrialisation. They proved to be adaptable to the requirements of the economies of scale and scope and found niches to survive. Moreover, their existence rested on strong cultural foundations which did not weaken during the second half of the nineteenth century.

The debate about the family firm For a long time, the discussion about the family firm stressed its negative impact on the development of a national economy. France in particular, but also England, became the points of reference in this debate: Their rates of economic development appeared to be retarded in the nineteenth century as compared to those of the USA and Germany, and it was assumed that, for various reasons, family firms and the behaviour of entrepreneurial families

221

DIRK SCHUMANN

were mainly responsible for this bad performance.1 The family firm was described as too small and inflexible to generate substantial growth in the advanced stages of industrialisation. Families alone could not raise the capital required for further expansion; in addition, heirs who did not feel committed to the firm could withdraw their capital, thus weakening its financial foundations. Furthermore, the way family firms were directed did not seem sophisticated enough in the long run. Patriarchal owners who stayed in the top position too long failed to take the necessary strategic decisions and blocked reforms of the administrative structures of their firms. Inheritance, rather than qualifications, provided access to the top positions, allowing incompetent heirs to ruin the business. Some authors have driven this point so far as to posit a ‘law of the third generation’, according to which the founder of a family firm would be followed by a dedicated son who would make for its continuous and successful expansion, while in the third generation a growing number of heirs together with a lack of competence or commitment of the grandsons would bring about the decline and fall of the business.2 In addition to these general problems of family firms—which remain important today—there were other detrimental aspects of the relationship between the entrepreneurial family and its enterprise that were specific to the nineteenth century. Considerable parts of the profits of family firms were not reinvested but used to buy real estate. In many cases, it has been argued, this was not meant to serve as a safety measure but as a first step towards obtaining a position as a civil servant or to gaining access to the aristocracy before becoming ennobled or marrying into it. As David S.Landes has put it with regard to France, ‘the affaire was never an end in itself, but the means to an end’ (Landes 1949:53). Thus, entrepreneurial families would gradually transform themselves. They would adopt an aristocratic lifestyle, lose their drive for economic activity, give up their businesses and end up leading the lives of landed aristocrats. This process, which has been termed ‘gentrification’ for England and ‘feudalisation’ for Germany, was described as having had ideological and political ramifications as well. In Germany, especially in the western parts of Prussia, entrepreneurs in the first half of the nineteenth century were marked by a frugal lifestyle, a liberal outlook on politics, and a clear distancing of themselves from the Prussian bureaucracy and aristocracy. But after the unsuccessful revolution of 1848, and definitely after the foundation of the German empire in 1870/71, a new generation of entrepreneurs came to the fore who, according to this thesis, made peace with their former enemies and formed an alliance with traditional conservatism.3 Other studies have stressed the cultural atmosphere of the late nineteenth century as a cause of the erosion of the entrepreneurial family and the family firm. While the financial means of entrepreneurial families increased, their education and taste became more refined, their sons and heirs began to develop specific intellectual interests or critical views of their social class and their own future lives. As a result, it is claimed, they often were unwilling to 222

BUDDENBROOKS REVISITED

follow in their fathers’ footsteps but turned to careers and activities in science and the humanities or in the fine arts.4 Thomas Mann’s famous novel Buddenbrooks, about a merchant family in Lübeck, provides a classic portrayal of this change in the interests and abilities of four generations of an entrepreneurial family. This basically negative long-term view of the family firm in the nineteenth and early twentieth centuries has been debated anew in recent years, encountering both criticism and support. It has been modified but not fully revised. The discussion has again centred on England, France and the USA as countries with a long industrial tradition, and has also addressed developments in Japan and other Asian countries as newcomers on the industrial scene. Research on the German experience, however, has been relatively scanty. The theses of Alfred D.Chandler have been major points of reference in this ongoing debate. According to Chandler, the family firm was an endangered species from the last quarter of the nineteenth century as the leading economies developed the modern strategies emphasising scale and scope. The driving forces behind this change were the expansion of markets, which was fostered by better communication and transportation facilities, changes in the technology of production, and the impact of the depression after 1873, which encouraged rationalisation and concentration in the major industrial branches. The new large industrial enterprises, which came to be marked by functional integration and product diversification, now also needed sophisticated managers and other well-trained personnel. The entrepreneurial family in general did not seem capable of fulfilling this new role (Chandler 1990). Chandler describes three types of enterprises that succeeded each other as the relationship between owners and managers evolved. The personal enterprise was typical of an early stage of the industrialisation when one or a few owners took care of the strategic operations as well as the day-to-day affairs of a firm. When firms grew in size they became entrepreneurial enterprises; strategic decisions were still taken by the owner(s) but routine operations were left to salaried managers. When size and capital requirements grew further, firms were transformed into managerial enterprises. Here, all decisions were put into the hands of salaried managers while the multitude of owners/shareholders exerted only limited influence through supervisory boards (Chandler 1974:1–34, esp. 5f). For Chandler, the dividing line clearly runs between the personal and entrepreneurial enterprise on the one hand, and the managerial enterprise on the other. The former retarded further growth, while the latter ranked among its necessary preconditions. Great Britain, with its long-lasting personal enterprises, and Germany, with its cooperative managerial capitalism, are presented as cases in point. Chandler’s evidence is taken from the 200 largest firms in each country. He makes it clear that, although transitional forms can be found in advanced countries, those that he discusses were representative of the prevailing forms of enterprise in the countries compared. 223

DIRK SCHUMANN

The recent contributions to the debate on the family firm have not challenged the proposition that very large firms cannot be organised like early personal enterprises.5 They have instead concentrated on the dynamic potential of small and medium-sized firms. Those firms, which in general were organised as family firms, exerted an outstanding creative influence on the whole economy because they could react more quickly than big ones to new challenges in the marketplace generated by new products or new services. This was particularly true in industries and situations in which relatively little initial capital was required. Most of these enterprises, however, did not survive into the third or even the second generation. This was mainly due to the general problem of succession which has already been sketched above. It was not due, however, as recent studies agree, to any form of feudalisation or ‘Buddenbrooks-effect’. Recent scholars regard a high number of newly founded family firms as evidence of the dynamic nature of an economic sector, whereas they see a low rate of establishment of family firms (but not their perseverance) as evidence of sluggish development. Family firms which survived for several generations seem to have been no less adaptable to new challenges than big public companies and were able to grow further. Several studies have made a point of presenting examples of such enterprises and have described their strategies in detail. Two tentative results emerge from the recent debate. First, it seems that the family firm as an organisational form has been quite flexible and left room for very diverse business strategies. These strategies were primarily determined by environmental factors, in particular by cultural influences, which could vary even from one region to another. No answer, however, has yet been found to the question of how such factors related to quantifiable market forces such as economic booms and crises or to the specific influences affecting particular sectors and industries.6 Second, size apparently has had a major influence on the structures and strategies of family firms. At a certain point in the growth of a family firm, problems may arise, in particular with regard to the recruitment of enough qualified managers. However, where exactly this point is to be placed has not yet been worked out.7 All in all, while a theory of the success and survival of the family firm in history still cannot be proposed, it is evident by now that the family firm has been a much more dynamic and adaptable form of enterprise than was for a long time believed.

‘Family firm’: a definition A discussion of the family firm and the entrepreneurial family requires a definition of the term ‘Family firm’. Unfortunately, there is no generally accepted meaning. Legal form cannot be regarded as the decisive element: while most family firms in Germany were founded as sole proprietorships or

224

BUDDENBROOKS REVISITED

general partnerships (Offene Handelsgesellschaft, or OHG), some of them were later transformed into joint-stock companies that remained under family control while others underwent the same transformation but became genuine managerial enterprises. After 1892, firms could also adopt the form of a limited liability company (Gesellschaft mit beschränkter Haftung, or GmbH), which had been created to enable small and medium-sized firms to attract more capital without exposing themselves too much to the general public. Thus GmbH status served the needs of many family firms quite well. Apart from legal form, ownership is another elusive criterion. If a family owns a firm but leaves its operation entirely to salaried managers, such an enterprise could not be called a family firm in a meaningful sense. On the other hand, there are sufficient examples of families still exerting decisive influence on firms which had been transformed into joint-stock companies even though they held less than five per cent of the shares (Barker and LévyLeboyer 1982:19f). Therefore it seems appropriate to use a rather loose definition of the term ‘family firm’: if an entrepreneurial family owns at least a small share of a firm and displays a deep and long-lasting commitment to it by somehow being involved in its strategic decisions, such an enterprise could be called a family firm (Sachse 1991:10f). This definition certainly is not very precise in quantitative terms, but any other would probably fail to grasp central aspects of the subject. It may only be added that the term ‘family’, too, is hard to define and should, in general, encompass distant relatives as well as brothers-in-law and sisters-in-law and their descendants if they are involved in the activities of the firm.

Bavaria as a point of reference The following parts of the paper present evidence of the long-lasting influence of entrepreneurial families and family firms on the German economy. It should be noted that quite a few examples will be taken from the history of entrepreneurs in Bavaria (in the south-east of Germany). Bavaria is an interesting case in point, particularly with respect to international comparisons. The pace of its overall industrialisation lagged far behind that of Germany throughout the nineteenth and early twentieth centuries. To judge by its percapita income on the eve of World War One, Bavaria was a backward German region. However, from the very beginning of German industrialisation, Bavarian entrepreneurs took part in it and played leading roles in several consumerand producer-goods industries, with the important exception of mining. Several large textile mills were the first major firms, followed by the machine builders M.A.N. and Maffei, and after 1870/71, by BASF, one of the leading German chemical firms, and Schuckert, an electrical company which eventually merged on almost equal terms with Siemens. Bavarian entrepreneurs were not backward, but they were not numerous enough. As far as size of enterprises

225

DIRK SCHUMANN

is concerned, however, Bavarian firms were somewhat smaller than their German counterparts in other regions. Among the 200 largest German companies which Chandler has analysed, only a few were located in Bavaria. Thus, while Bavarian entrepreneurs in general fit the general profile of German entrepreneurs, in terms of the sizes of their firms they can be better compared with entrepreneurs in England and France than with the average German businessmen.8

The family firm in the early stage of industrialisation It is widely accepted that the family played a very important role in the industrialisation of Germany up to around 1870.9 Except for a few branches like the railroads, where huge amounts of capital were necessary from the beginning, the dominant type of enterprise was the family firm organised as a sole proprietorship or a general partnership. Jürgen Kocka has shown in detail why the close connection between family and firm was essential for German economic development during early industrialisation: The education of the future entrepreneurs mainly took place, not at schools, but within the family. Since most entrepreneurs were sons of small businessmen, artisans or entrepreneurs themselves, they not only learned the standard bourgeois values like thrift, order and performance of duties, they were also motivated to work as independent businessmen and to strive for economic success. Very often they were also provided with the necessary qualifications for their future jobs by their fathers and/or other members of the family, as special schools and colleges for that purpose did not yet exist. Providing the capital for an enterprise was another important function of the early entrepreneurial family rather than just the nuclear family. Since a network of banks and other credit institutions was not yet available, the family had to furnish the capital for the foundation of a firm and be ready to contribute further financial resources to it in periods of expansion or crisis.10 Marriage connections did not necessarily match the middle-class ideal of romantic love, but served as a way to broaden the financial base of a firm. Parity of the financial status of groom and bride was usually required, and thus the dowry would be a substantial addition to the capital of the firm. Moreover, while a labour market for managers had not yet come into being, the family also provided the necessary staff if the firm created new management positions or established branches in other places. The way Werner Siemens placed his brothers is a well-known example of this recruitment strategy (Kocka 1969:58–62). Finally, through frequent intermarriages with families of other businessmen, entrepreneurial families established networks which could act as safety nets, help to establish contact with other businessmen, and, if necessary, support approaches to the government. Thus, the family network functioned as a

226

BUDDENBROOKS REVISITED

substitute for the trade associations and other interest organisations that would eventually arise (Kocka 1979; Sachse 1991). Kocka’s examples are mainly taken from prominent families like the Krupps and the Siemens, and from Rhineland—Westphalia, the core region of Germany’s economic development in the nineteenth century. Subsequent research has shown that the patterns and structures described by Kocka were also found elsewhere. In Bavaria, important firms were established during the 1830s and 1840s by one or several families in the main industrial branches of cotton spinning and weaving and machine-building, e.g. the factories of (Cramer-) Klett in Nuremberg and of Sander in Augsburg which were later to become the M.A.N. (Schumann 1992:32ff., 59ff; Büchner 1940). Among the consumer-goods industries (which Chandler has described as on the whole backward in Germany; Chandler 1990:398ff), brewing (Chandler’s exception) was a particularly advanced trade in the first half of the nineteenth century. One of the leading firms was the Lederer-brewery in Nuremberg. (The first freight transported by a German railway was two barrels of Lederer beer.) Christian Lederer, who had completed an apprenticeship as a brewer, founded the brewery as a modern industrial enterprise after buying a traditional brewery in 1814. His own savings were not sufficient to pay the price of 18,000 fl., but some years before he had married the daughter of a brewer from a nearby town and he probably used her dowry. The couple had five children, two daughters and three sons. Both daughters married merchants from Nuremberg. Georg Lederer, the eldest son, learned the trade in his father’s brewery and was then sent abroad to study modern brewing technology in England. He married a brewer’s daughter from nearby Erlangen, was placed in charge of the brewery while his father still owned it, and became the sole owner on his father’s death. Johann Lederer, the second son, received financial support from his father to establish his own brewery, while the youngest son apparently suffered from ill health and was not able to join his brothers. Christian Lederer’s last will, which he prepared together with his wife, stipulated that no payment exceeding 3,000 fl. could be made to the two daughters and the youngest son, in order to avoid jeopardising the finances of the brewery, which made up the bulk of the family’s fortune. Georg Lederer, who died just three years after his father, had only two children, both sons. They succeeded him jointly and modernised and expanded the brewery, which was to remain one of the leading Bavarian breweries until the turn of the century.11 The Lederer case illustrates how early industrial entrepreneurs used their families to accumulate capital, ease competition, build up a network of commercial relations, and provide their successors, while also taking measures to prevent the family from destroying the enterprise. The case also shows the long-lasting commitment of the family to the enterprise, which does not seem to have come to an end in the third generation. In addition to that and to Kocka’s findings, the example demonstrates that a reduced number of 227

DIRK SCHUMANN

children was quite consistent with the development of the enterprise. While the firm’s structure became more complicated, with technical and commercial tasks being the first for which special managerial positions were created, there were still enough heirs to take over other important posts. Equally importantly, there were fewer surplus heirs to demand money from the firm without contributing. Very similar stories could be told about brewers in Munich, such as the Pschorr family. In these cases the ‘founding father’ left each of his sons their own breweries and they later repeated the same process of expansion, modernisation and separation of functions as in Nuremberg.12 While the Lederers were Protestants, the Pschorrs were Catholics. There is no obvious reason to assume that their respective stories were specifically determined by their confessional denominations. Another thing that Protestant and Catholic entrepreneurs in Bavaria had in common was the limited geographical ranges of their intermarriage relationships. They apparantly did not establish marriage and subsequent family ties which spanned a whole region, as did their counterparts in Rhineland—Westphalia and other German regions (Schumann 1992:95ff). Whether this impeded the growth of their firms, however, remains an open question.

The family firm in the advanced stage of industrialisation After 1870/71 the close connection between family and firm seems to have lost its main functions. A number of important banks were founded which eventually covered the German regions with a network of branches. Financial aid for new industrial enterprises no longer had to be furnished by the entrepreneurial family. The older technical colleges (Polytechnika) were transformed into technical universities and expanded greatly, while the science departments in the traditional universities also grew. Therefore, as industrial technology became more complicated, technical experts no longer had to be trained on the job but were now recruited from academic institutions. After 1900, commercial training also acquired an academic dimension. Finally, in the new German Kaiserreich, in addition to the reformed chambers of commerce, interest organisations were founded on national, regional and branch levels, particularly from the 1890s onwards, eroding the third main function of the extended family. In addition, as competition increased and large-scale enterprises took the lead, the general problems of succession and finances of family firms, which have been described in the previous section, had a greater impact (see Schumann 1992: If.; Kocka 1979:132–5). However, the family firm did not disappear and the ties between entrepreneurial families and their firms remained strong. Quantitative and qualitative evidence from the study of leading industrial enterprises and from the collective and individual biographies of entrepreneurs point in the

228

BUDDENBROOKS REVISITED

same direction. Recent research in German economic history has concentrated on the structures and strategies of the largest firms. Kocka and Siegrist, comparing the 100 largest enterprises in 1887 and 1907 with regard to their legal forms, found that seventy-three joint-stock companies were among the top 100 as early as 1887, a number that rose to seventy-five in 1907. The number of sole proprietorships and general partnerships fell from fifteen to seven over the same period (Kocka and Siegrist 1979:55–122). Figures that concentrate on legal form, however, convey a misleading impression. One of the enterprises owned by a single proprietor in 1887, for example, was Krupp, then by far the largest German enterprise in terms of capital and the number of workers. Another was Siemens, the emerging giant of the electrical industry. Although both firms had been transformed into joint-stock companies by 1907, the former owners and their families still exerted decisive influence on the enterprises. The same seems to have been true for the majority of the other big companies. While exact figures on the degree of family influence are probably impossible to come by, several authors have estimated that, before World War One, ‘entrepreneurial enterprises’ were still dominant in German industry. Hans Pohl even has assumed that this situation prevailed up to World War Two (Siegrist 1980:88; Pohl 1981; Brockstedt 1984). This apparently was true not only in general, but for most individual industries as well. According to Brockstedt, entrepreneurial (or personal) enterprises predominated in the traditional capital-intensive branches of mining, in the modern capital-intensive branches of the chemical and electrical industries, and in the backward consumer-goods industries, while they also had a strong position in iron and steel production and machine-building (Brockstedt 1984:249). Apart from Krupp and Siemens and several Silesian magnates, well-known families such as Thyssen, Haniel, Hoesch or Stinnes still controlled their large enterprises. Up to World War One, Maffei, one of Bavaria’s leading machine-building firms, was owned by the nephew of the founder, who still took part in its strategic decisions (Möhl 1937; Martin 1914:123–7). Legal form and de facto ownership and control do not tell much about the long-term success of a firm. It seems safe to say, however, that those entrepreneurial enterprises were no less adaptable and flexible than their managerial counterparts. Krupp, for example, became a modern enterprise in terms of internal organisation quite early, when Alfred Krupp institutionalised a top position for administration apart from technological and commercial matters in his ‘Generalregulativ’ of 1872. Later on, his son expanded the range of decisions which the board of directors were allowed to take, and finally appointed a director-in-chief without a specific portfolio, albeit without surrendering his own supreme rights. Krupp, moreover, created a ‘Rechnungsrevisionsbureau’ which as early as 1875 comprised the functions of controlling, internal revision, and internal control. Parallel organisational reforms were implemented at the Gutehoffnungshütte (Bongartz 1984:33– 55, 73–113). Siemens successfully reclaimed its role as a pioneer of the 229

DIRK SCHUMANN

electrical industry after it had been challenged in the field of high-voltage products by AEG, which was a managerial enterprise from the start. While a sophisticated management structure was built up, the Siemens family, who still owned the majority of the shares, reinforced their control by creating the position of a ‘delegate of the supervisory board’ and filling it with a son of the founder. The ‘delegate’ was both a member of the supervisory board and the coordinator of the work of the board of directors, which did not have a chairman.13 All in all, it was not the legal form and the ownership structure which mattered, but the ability of a firm to adapt its organisational structures rapidly to changing production and market conditions (Pohl 1981:177–8). In the field of small-and medium-scale enterprises it was clearly the family firm which predominated. In 1907, according to official German statistics, 99.6 per cent of factories with more than ten employees had fewer than 1,000 employees, who together comprised 85.8 per cent of the total German industrial workforce.14 There are no figures about the legal forms of all these enterprises, but we can probably assume that the overwhelming majority of them were sole proprietorships or general partnerships. Thus, there obviously was ample room for family firms to exist and to develop successfully. For Bavaria alone there are many examples. Of the owner-entrepreneurs in my sample who were active between 1871 and 1914, 48.8 per cent were founders of their firms (Schumann 1992:155). Family firms were established in those branches of consumer-goods production which were rather backward and had not yet introduced mechanisation and mass production. Thus, for example, in furniture, cloth making and food production many former artisans’ shops were transformed into factories and further expanded by the descendants of the founders. In 1884, for instance, after receiving commercial training, Heinrich Metzger, the grandson of a Nuremberg gingerbread baker, took over the family firm together with his two brothers. Under their direction the firm, which already had been mechanised in 1872, gained access to the world market and employed almost 200 workers by the turn of the century. In 1920, Heinrich Metzger was largely responsible for arranging the fusion of his firm with another Nuremberg family firm. He then became deputy head of the supervisory board of the new joint-stock company.15 There were also still chances to found an enterprise in certain branches of consumer-goods production. Ignatz Bing, a Nuremberg merchant who had started his business by trading in hops and yarn, turned to making oil lamps, a cheap mass-produced product, in the 1870s. He then established branch factories in Saxony and other regions, built up his own marketing network, and gained access to the world market. When he transformed his enterprise into a family joint-stock company in 1898—at least one son and one son-inlaw were among its directors—he employed more than 1,000 workers.16 Other family firms were established in the building trades where the rapid growth of cities opened up opportunities for new enterprises, but rationalisation and the introduction of capital-intensive machinery were possible only to a 230

BUDDENBROOKS REVISITED

limited degree. Here, too, former artisans’ shops were sometimes transformed into industrial enterprises. Moreover, family firms could still be established in the most modern and capital intensive branches like the machine-building and chemical industries. Here, founders had to discover niches, e.g. the production of special machinery for dairies or tanneries that the big enterprises like the M.A.N. were not (or not yet) ready to make (Schumann 1992:147f). They also could be successful when they invented new products. The chemist Albert Knoll, after earning a PhD at Göttingen, spent some time in an English firm where he invented a procedure to obtain the pain-and cough-killing drug codeine from morphine. In 1886 he founded a factory at Ludwigshafen/ Rhine with his brother and his brother-in-law. More inventions and innovations followed, branch factories were built abroad, and in 1925 the enterprise became a joint-stock company, apparently still under the control of the family, with Albert Knoll serving as chairman of the supervisory board until his death in 1938 (Deutsche 1980:207f). In general, quite a few founders or heirs of family firms after 1870 proved to be successful because they were inventors or inno-vators. As, in this respect, they were dynamic entrepreneurs in Schumpeter’s sense, the family firm had (and still has) a chance to hold its own.

Heirs In addition to these findings about the firms, the results of other studies on the entrepreneurial family point in the same direction. An analysis of twentyfour such families from all over Germany who started their businesses in the eighteenth and nineteenth centuries demonstrates that more than half of their descendants in the third generation were independent entrepreneurs or held top positions as managers (Stahl 1973:278ff). At least two-thirds of the grandsons of Bavarian entrepreneurs who were active in the nineteenth century became entrepreneurs themselves, most of them, it seems, in the firms of their grandfathers (Schumann 1992:213–16). The brewing family Pschorr in Munich is again a case in point. On the eve of World War One, three of the four great-grandsons of the founder jointly managed the firm, while their younger brother held a post as a university professor of chemistry and functioned only as a sleeping partner. They had expanded and modernised the brewery after the death of their father in 1894, sold their beer on the world market, employed 400 workers and accumulated fortunes of 6 million Marks each (whereas their father had bequeathed them a total 7 million).17 From all the information available it is very likely that the development of those family firms that did not suffer from a lack of heirs took a steady path from generation to generation up to World War One.18 When, after the war, a number of firms changed their legal forms and their proprietors, this probably not only occurred because of the economic turmoil in the course of military

231

DIRK SCHUMANN

defeat and inflation, but was also due to the fact that during the war quite a few heirs and potential successors had been killed or severely wounded and therefore were not able to assume the top positions in their firms. Under ‘normal’ conditions, some family firms would probably have been able to persevere. As far as the financial resources of entrepreneurial families are concerned, one group of them has been particularly scrutinised in recent years: the very rich. Here, it was assumed that the evidence for the ‘feudalisation thesis’ would be particularly strong. Now, however, we know that basic assumptions of the thesis are wrong. A study of more than 400 very wealthy businessmen (multimillionaires) in Wilhelmine Germany has shown that more than twothirds of their sons also became entrepreneurs, whereas only six per cent chose the life of a landowner and sixteen per cent embarked on careers as higher civil servants, army officers and in higher education. Figures for the Bavarian wealth élite are very similar.19 Even the very few who were granted a title of nobility did not give up business. Lothar Faber, the leading German pencil producer from the vicinity of Nuremberg, was made a baron (Freiherr) in 1881 but continued to direct his enterprise as the personal proprietor until his death in 1896. His successor was Count Castell, a member of the German high aristocracy by birth who had married Faber’s granddaughter, the sole heiress of the firm. Castell, who changed his name to Faber-Castell, does not seem to have had any trouble with his new role and even used his coat-ofarms for advertising purposes.20 The Oppenheim family, a well-known case from the west of Germany, was ennobled in the late 1860s. The family acquired a landed estate and took on an aristocratic lifestyle but did not terminate its banking activities. Family members were still present in the top management of the private bank in the 1920s and again after World War Two, when the Nazi regime which had tried to oust them had fallen from power (Stürmer et al. 1989:194ff).

The entrepreneurial family and the middle classes Presenting figures and facts that show the continuity of the close connection between family and firm is easier than explaining its causes. More research will be needed that follows entrepreneurial families through several generations and combines the social and cultural history of the families with the economic history of their firms. As far as the intentions and motives of entrepreneurs as family heads at the turn of the century are concerned, last wills are a telling source. Their basic contents do not seem to have been different from those of wills made in the early stages of industrialisation: the firm was to be handed over to the male heirs, while its capital would be protected by clauses to prevent other heirs from being paid off so quickly that they undermined the financial base of the firm. Often the testator explicitly urged

232

BUDDENBROOKS REVISITED

that the firm be kept by the family and continued undivided (which, in an age of very low inheritance taxes, would not create substantial strains). The firm was conceived not as an instrument to make money, but as a ‘mission’ which had to be served and held in high esteem. The religious spirit of quite a few of the older wills usually now was absent, but it had been replaced by a kind of secular awe towards the creation of the entrepreneur. Apparently, there were not many protests against such stipula-tions. This compliance of the family members could serve as a proof of the continuing existence of a strong family identity. The modern process of individualisation and the conflict between the generations may not have been as intense at the fin de siècle as has sometimes been claimed.21 Recent findings on the history of the whole German Bürgertum shed further light on the reasons for the continuing commitment of entrepreneurial families to their firms. After 1870 the Bürgertum certainly could be regarded as more homogeneous than before. Social ties between entrepreneurs, civil servants, army officers and professionals, in particular through marriage, were strengthened, while all groups of the middle classes increasingly shut themselves off from the working class (Kocka 1988). Social contacts were culturally reinforced. In his portrait of the Bassermanns, an important middleclass family in the south-west of Germany, Lothar Gall has described how the construction of a family identity was taken up after 1850. Now families of the Bürgertum started to collect material about their ancestors, write their own histories, invent coats-of-arms, and use seals. Gall suggests that the families who started such activities belonged to an older generation of the Bürgertum that placed particular emphasis on cultural interests but less on economic success as such, and who were often no longer successful in business (Gall 1989:393ff). It may well be that those entrepreneurial families who had only recently acquired their wealth were less sophisticated in their efforts to build up and consolidate a family identity. There is reason to believe, however, that, while steadily reducing the number of children,22 they made efforts to inculcate in their sons and heirs a sense of responsibility for the enterprise and for the family as a whole. The emerging family ideology might even have been functional in an economic sense. By underlining the bonds and obligations between the family members and by ascribing an elevated status to the common enterprise, it helped to place some constraints on the spending habits of the individual family members and enforced the practice of reinvesting profits. Other developments in the German society after 1870–1 yielded similar results. Now entrepreneurs, especially those with a training as artisans or technicians, no longer had to suffer from an inferiority complex with regard to civil servants. The title of a commercial councillor (Kommerzienrat), up to then limited to Prussia, was introduced in the other states as well and was granted in increasing numbers. The new councillors were left with the feeling of having achieved the same status as the ‘normal’ councillors in the civil 233

DIRK SCHUMANN

service. Thus there was no longer any need to abandon their business activities in order to gain status. If one of the sons of an entrepreneur chose a different career, however, it could even be helpful for the firm, as he would not claim a management position and thus avoided possible conflicts with brothers and cousins. As long as there was at least one male heir or an appropriate son-in-law to take command, the family firm was not threatened. Therefore, if scions from entrepreneurial families did not follow in their ancestors’ footsteps, it should not be assumed that this indicated the decline of the family firm unless the familial context of such behaviour has been explored.23 The same could be true for sons becoming lawyers and politicians. Kocka is certainly right in arguing that the emergence of interest organisations after 1870 relieved the entrepreneurial family of one of its functions and thus contributed to weaken it. However, this process could work both ways. Gall describes the case of Ernst Bassermann, one of the leading politicians of the liberal—conservative ‘establishment’ in Wilhelmine Germany. Bassermann not only held important political posts, he also sat on a number of supervisory boards (Gall 1989:427). This obviously had no direct influence on the remaining firms being run by other members of the family, but it helped to create contacts for business transactions and thus exerted an indirect influence. While the old family networks had been functional for the family firm, the new organisational ones could serve the same purpose. Future research, therefore, should conceive of family firms even in the late nineteenth century as elements of a greater family network, and the focus should shift from individual family firms to the extended families behind them. This would make it easier to differentiate between the influences of market forces, on the one hand, and the collective efforts of the family, on the other, on the performance of such firms (Jones and Rose 1993:9).

Summary The family firm was one of the essential features of the German economy from the beginning of industrialisation far into the twentieth century. This judgement is based on a broad definition of the term ‘family firm’ which includes all forms of major influence by an entrepreneurial family on a firm. Large family firms proved to be no less adaptable than their managerial competitors when sophisticated management structures had to be built up. Small and medium-size family firm continued to be established in a wide range of industries as industrialisation progressed. Entrepreneurial families became strongly interrelated with other families of the Bürgertum. While their commitment to the firm continued to carry a high ethical value, the cohesion of the family as an expanding network of relatives became rooted in new cultural foundations. When the essential economic functions which

234

BUDDENBROOKS REVISITED

had upheld the entrepreneurial family at the beginning of industrialisation faded away, cultural and ideological forces arose as compensation.

Notes 1 See Landes (1949:45–61, 1951:334–53, 1966:274–601, esp. 553ff); Cochran (1965/ 66:25–38). As a systematic overview over the general shortcomings of family firms, see Mittelsten Scheid (1985), and Albach and Freund (1989:264ff). 2 A German example is Paulsen (1941:271–80). 3 Zunkel (1962:99ff). 4 The most recent example of this position for Germany is Bauer (1991). 5 See the articles in Hannah (1982); Okochi and Yasuoka (1984); Jones and Rose (1993), in particular Church (1993) who places particular emphasis on the Japanese zaibatsu; and Hannah (1983). For recent case studies, see French (1993); Carter et al. (1976), in particular the article by Lévy-Leboyer (1976); and Berghoff (1991). A stimulating combination of social and economic history is Davidoff and Hall (1987). 6 See Church (1993:26ff) on the different performances of family firms in England and Japan; Barker and Lévy-Leboyer (1982:10–25, esp. 20–2) on the regional and religious differences in France; Hannah (1982) on the problems of sale or transformation of an enterprise. 7 This problem has been pointed out by Payne (1984). 8 Schumann (1992:23ff). The sample includes 647 entrepreneurs and their families from a wide variety of branches in industry and commerce. 9 Wehler (1987:185ff, 1995:112ff); Nipperdey (1983:178ff). 10 Pierenkemper has recently demonstrated that the critical foundation period of industrial enterprises in the nineteenth century (and today) took several years during which a steady influx of capital was needed in order to prevent firms from going bankrupt and before they were in a position to generate profits. The history of Krupp under the young Alfred Krupp is a particular case in point. See Pierenkemper (1990). 11 Lederer (1968); Jegel (1952:286ff); Staatsarchiv Nürnberg, AG Nürnberg NA 1849/ 509; 1852/630. 12 Roth (1921), Schumann (1992:115). 13 Wilhelm von Siemens, the son and successor of the founding father Werner, acted as first ‘delegate’. After his death, his brother Carl Friedrich followed him in this position. See Kocka (1969:383ff., 453f); Goetzler and Schoen (1986); Homburg (1991:343ff., 381ff). 14 Statistik des Deutschen Reiches N.F. vol. 213:42. In 1882 the figures were 99.7 per cent and 91.1 per cent (Statistik des Deutschen Reiches N.F. vol. 119:43). 15 Verwaltungsbericht der Stadt Nürnberg 1930:432. 16 Hilpert (1950:202–7); Staatsarchiv Nürnberg, AG Nürnberg 1918/533. 17 Roth (1921); Martin (1914); Staatsarchiv München, AG München IA NR Not. XV, 637. 18 It should not be overlooked that even among salaried managers entrepreneurial dynasties emerged. Walther Rathenau, who succeeded his father Emil as head of AEG, is only the most prominent example. Brockstedt (1984:262); Schumann (1992:202ff). 19 Augustine-Perez (1988:307), Schumann (1992:209) and Berghoff and Möller (1993:373ff) come to the same conclusion.

235

DIRK SCHUMANN

20 Deutsche (1959:722f); Bitterauf (1919:90–4); Bayerisches Hauptstaatsarchiv, Adelsmatrikel Grafen F 79. 21 Schumann (1992:222–9); Stürmer (1989:212ff., 257ff). Hellige (1979:476–518) has emphasised that severe conflicts between fathers and sons arose particularly in Jewish entrepreneurial families. Mosse (1989:115f) has criticised this view and argued that most such conflicts eventually were settled and ended with the sons accepting their place as successors in the family enterprise. By contrast, Kocka (1979:125ff) assumed that individualisation threatened the existence of family firms at the end of the nineteenth century. 22 Entrepreneurs in Bavaria around 1900 had only three children on the average, whereas before 1870 it had been five. Schumann (1992:201); von Nell (1973). 23 Stürmer (1989:264ff) gives an example of how a member of the family who became a professional archaeologist was well supported by the family and enhanced its prestige without having any negative impact on the family business.

References Albach, H. and Freund, W. (1989) Generationswechsel und Unternehmenskontinuität— Chancen, Risiken, Maßnahmen, Gütersloh: Verlag Bertelsmann-Stiftung. Augustine-Perez, D.L. (1988) ‘Very Wealthy Businessmen in Imperial Germany’, Journal of Social History 22:299–321. Barker, T.C. and Lévy-Leboyer, M. (1982) ‘An Inquiry into the Buddenbrooks Effect in Europe’, in L.Hannah (ed.) From Family Firm to Professional Management: Structure and Performance of Business Enterprise, Budapest: Akadémiai Kiadó. Bauer, F.J. (1991) Bürgerwege und Bürgerwelten. Familienbiographische Untersuchungen zum deutschen Bürgertum im 19. Jahrhundert, Göttingen: Vandenhoeck and Ruprecht. Berghoff, H. (1991) Englische Unternehmer, 1870–1914. Eine Kollektivbiographie führender Wirtschaftsbürger in Birmingham, Bristol und Manchester, Göttingen: Vandenhoeck and Ruprecht. Berghoff, H. and Möller, R. (1993) ‘Unternehmer in Deutschland und England 1870– 1914. Aspekte eines kollektivbiographischen Vergleichs’, Historische Zeitschrift 256:353–86. Bitterauf, T. (1919) ‘Lothar Freiherr von Faber, Grossindustrieller, 1817–1896’, in A.Chroust (ed.), Lebensläufe aus Franken, vol.1, Munich: Schöningh: 90–4. Bongartz, W. (1984) ‘Unternehmensleitung und Kostenkontrolle in der rheinischen Montanindustrie vor 1914. Dargestellt am Beispiel der Firmen Krupp und Gutehoffnungshütte’, Zeitschrift für Unternehmensgeschichte 29:33–55 and 73– 113. Brockstedt, J. (1984) ‘Family Enterprise and Large Scale Enterprise in Germany (1871–1914)—Ownership and Management’, in A.Okochi and S.Yasuoka (eds) Family Business: 237–65, Tokyo: University of Tokyo Press.. Büchner, F. (1940) 700 Jahre Geschichte der Maschinenfabrik Augsburg-Nürnberg, Frankfurt: Societäts-Druckerei. Carter, E.C. II, Forster, R. and Moody, J.N. (eds) (1976) Enterprise and Entrepreneurs in Nineteenth-and Twentieth-Century France, Baltimore.: Johns Hopkins University Press. Chandler, A.D. Jr (1974) ‘The Rise of Managerial Capitalism and its Impact on Investment Strategy in the Western World and Japan’, in H.Daems and H.van de Wee (eds) The Rise of Managerial Capitalism, Louvain and The Hague: Leuven University Press: 1–34, esp. 5f. 236

BUDDENBROOKS REVISITED

——(1990) Scale and Scope: The Dynamics of Industrial Capitalism, Cambridge, MA.: Belknap Press. Church, R. (1993) ‘The Family Firm in Industrial Capitalism: International Perspectives on Hypotheses and History’, Special Issue of Business History 4: 17–43. Cochran, T.C. (1965/6) ‘The Entrepreneur in Economic Change’, Explorations in Entrepreneurial History, 2nd.ser., 3:25–38. Davidoff, L. and Hall, C. (1987) Family Fortunes: Men and Women of the English Middle Classes, 1780–1850, London: Hutchinson. Deutsche (1959) Neue Deutsche Biographie, 4, Berlin: Duncker and Humblot. Deutsche (1980) Neue Deutsche Biographie, 12, Berlin: Duncker and Humblot. French, M. (1993) ‘Structure, Personality, and Business Strategy in the U.S. Tire Industry: The Seiberling Rubber Company, 1922–1964’, Business History Review: 246–78. Gall, L. (1989) Bürgertum in Deutschland, Berlin: Siedler: 393ff. Goetzler, H. and Schoen, L. (1986) Wilhelm und Carl Friedrich von Siemens. Die zweite Unternehmergeneration, Stuttgart: Steiner-Verlag-Wiesbaden. Hannah, L. (ed.) (1982) From Family Firm to Professional Management: Structure and Performance of Business Enterprise, Budapest: Akadémiai Kiadó. Hellige, H.-D. (1979) ‘Generationskonflikt, Selbsthaß und die Entstehung antikapitalistischer Positionen im Judentum. Der Einfluß des Antisemitismus auf das Sozialverhalten jüdischer Kaufmanns-und Unternehmersöhne im Deutschen Kaiserreich und in der K.u.K.Monarchie’, Geschichte undGesellschaft 5:476–518. Hilpert, H. (1950) ‘Ignaz Bing’, in Stadtrat von Nürnberg (ed.) Nürnberger Gestalten aus neun Jahrhunderten. Ein Heimatbuch zur 900 Jahrfeier der ersten urkundlichen Erwähnung N ürnbergs, Nürnberg: Ulrich and Co.: 202–7. Homburg, H. (1991) Rational is ierung und Industriearbeit. Arbeitsmarkt— Management —Arbeiterschaft im Siemens-Konzern Berlin 1900–1939, Berlin: Haude and Spener: 343ff., 381ff. Jegel, A. (1952) Die wirtschaftliche Entwicklung von Nürnberg-Fürth, Stein und des Nürnberger Raumes seit 1806. Mit Berücksichtigung des allgemeinen Geschehens, Nürnberg: Spindler. Jones, G. and Rose, M.B. (eds) (1993) ‘Family Capitalism’, Special Issue of Business History 35(4):1–251. Kocka, J. (1969) Unternehmensverwaltung und Angestelltenschaft am Beispiel Siemens 1847–1914. Zum Verhältnis von Kapitalismus und Bürokratie in der deutschen Industrialisierung, Stuttgart: Klett. ——(1979) ‘Familie, Unternehmer und Kapitalismus. An Beispielen aus der frühen deutschen Industrialisierung’, Zeitschrift für Unternehmensgeschichte 24: 99–135. ——(1988) ‘Bürgertum und bürgerliche Gesellschaft im 19. Jahrhundert. Europäische Entwicklungen und deutsche Eigenarten’, in J.Kocka (ed.) Bürgertum im 19. Jahrhundert. Deutschland im europaischen Vergleich, Munich: dtv; 11–76. Kocka, J. and Siegrist, H. (1979) ‘Die hundert größten deutschen Industrieunternehmen im späten 19. und frühen 20. Jahrhundert. Expansion, Diversifikation und Integration im internationalen Vergleich’, in N.Horn and J.Kocka (eds) Recht und die Entwicklung der Großunternehmen im späten 19. und frühen 20. Jahrhundert, Göttingen: Vandenhoeck and Ruprecht; 55–122. Landes, D.S. (1949) ‘French Entrepreneur ship and Industrial Growth in the Nineteenth Century’, Journal of Economic History 9:45–61. ——(1951) ‘French Business and Businessmen in Social and Cultural Analysis’, in E.M.Earl (ed.) Modern France, Princeton: Princeton University Press; 334–53. ——(1966) ‘Technological Change and Development in Western Europe, 1750–

237

DIRK SCHUMANN

1914’, Cambridge Economic History of Europe, VI, Cambridge: Cambridge University Press: 274–601. Lederer (1968) Fünf Jahrhunderte Geschichte der Lederer-Brauerei, Frankfurt: Societats-Druckerei. Lévy-Leboyer, M. (1976) ‘Innovation and Business Strategies in Nineteenth-and Twentieth-Century France’, in E.C.Carter et al. (eds) Enterprise and Entrepreneurs in Nineteenth and Twentieth Century France: 87–135. Martin, R. (1914) Jahrbuch des Vermögens und Einkommens der Millionäre in Bayern, Berlin: Siegismund and Volkening. Mittelsten Scheid, J. (1989) Gedanken zum Familienunternehmen, Stuttgart: Poeschel. Möhl, F. (1937) 1OO Jahre Krauss-Maffei 1837–1937, Munich: Krauss-Maffei. Mosse, W. (1989) The German-Jewish Economic Elite 1820–1935. A Socio-Cultural Profile, Oxford: Oxford University Press. Nell, A.von (1973) Die Entwicklung der generation Strukturen bürgerlicher und bäuerlicher Familien von 1750 bis zur Gegenwart, PhD dissertation, University of Bochum. Nipperdey, T. (1983) Deutsche Geschichte 1800–1866. Bürgerwelt und starker Staat, Munich: Beck: 178ff. Okochi, A. and Yasuoka, S. (eds) (1984) Family Business in the Era of Industrial Growth: Its Ownership and Management, Tokyo: University of Tokyo Press. Paulsen, A. (1941) ‘Das “Gesetz der dritten Generation”. Erhaltung und Untergang von Familienunternehmungen’, Der praktische Betriebswirt 21:271–80. Payne, P.L. (1984) ‘Family Business in Britain: A Historical and Analytical Survey’, in A.Okochi and S.Yasuoka (eds) Family Business in the Era of Industrial Growth: Its Ownership and Management, Tokyo: University of Tokyo Press, 171–206. Pierenkemper, T. (1990) ‘Zur Finanzierung von industriellen Unternehmensgründungen im 19Jahrhundert—mit einigen Bemerkungen über die Bedeutung der Familie’, in D.Petzina (ed.) Zur Geschichte der Unternehmensfinanzierung, Berlin: Duncker and Humblot: 69–97. Pohl, H. (1981) ‘Zur Geschichte von Organisation und Leitung deutscher Gr oßunter nehmen seit dem 19. Jahrhundert’, in Zeitschrift für Unternehmensgeschichte 26: 143–78. Roth, H. (1921) Ein Jahrhundert Pschorrbräu 1820–1920, München: J.Lindauersche Universi tätsbuchhandlung. Sachse, W. (1991) ‘Familienunternehmen in Wirtschaft und Gesellschaft bis zur Mitte des 20. Jahrhunderts. Ein historischer Überblick’, Zeitschrift für Unternehmensgeschichte 36:9–25. Schumann, D. (1992) Bayerns Unternehmer in Gesellschaft und Staat, 1834–1914. Fallstudien zu Herkunft und Familie, politischer Partizipation und staatlichen Aus-zeichnungen, Göttingen: Vandenhoeck and Ruprecht Siegrist, H. (1980) ‘Deutsche Großunternehmen vom späten 19. Jahrhundert bis zur Weimarer Republik. Integration, Diversifikation und Organisation bei den hundert größten deutschen Industrieunternehmen (1887–1927) in international vergleichender Perspektive’, Geschichte und Gesellschaft 6:60–102. Stahl, W. (1973) Der Elitekreislauf in der Unternehmerschaft. Eine empirische Untersuchung für den deutschsprachigen Raum, Frankfurt: Deutsch: 278ff. Stürmer, M., Teichmann, G. and Treue, W. (1989) Wägen und Wagen. Sal. Oppenheim jr. & Cie. Geschichte einer Bank und einer Familie, Munich and Zurich: Piper: 194ff. Verwaltungsbericht der Stadt Nürnberg (1930), Nürnberg. Wehler, H.-U. (1987) Deutsche Gesellschaftsgeschichte, 2. Von der Reformära bis

238

BUDDENBROOKS REVISITED

zur industriellen und politischen ‘Deutschen Doppelrevolution’, 1815–1845/49, Munich: Beck. ——(1995) Deutsche Gesellschaftsgeschichte, 3, Von der ‘Deutschen Doppelrevolution’ bis zum Beginn des Ersten Weltkrieges, 1849–1914, Munich: Beck. Zunkel, F. (1962) Der Rheinisch-westfälischer Unternehmer 1834–1879. Ein Beitrag zur Geschichte des deutschen Bürgertums im 19. Jahrhundert, Cologne: Westdeutscher Verlag.

239

INDEX

Abernathy, W.J. and Utterback, J.M. 132 agency/structure 4_6, 5, 75_6, 78_9, 82;and general social theory 84_5; interaction 83;and recurrent practice 83_4;and resources 83;and rules 83_4 Aiken, C.S. 140, 142, 147, 153, 154 Aiken, J. 27 Albach, H. and Freund, W. 238 Alchian, A.A. 54;and Demsetz, H. 7, 16, 58, 61, 69, 131;and Woodward, S. 157 Aldrich, H.E. 128 Allen, G.C. 18 Alston, L.J. 140, 146;and Ferrie, J.P. 147, 153;and Higgs, R. 70, 71, 141 Ames, E. and Rosenberg, N. 54, 69 Ashton,T.S. 26, 27, 29 Aspin, C. and Chapman, S.D. 22, 24, 29, 31_2 Atlan, H. 126 Augustine-Perez, D.L. 239 authority 1_2, 6, 14, 48, 101, 155_6, 161_2, 169_70; effectiveness/ efficiency of 2;and gender division of labour 11;in Industrial Revolution 2_4;and Joint Commission Model 165_9;and professional mode of production 157_63;professionals and complementary institutions 163_5;and provision of coordination 4;transaction costs of 7_8;in Weberian/Chandlerian firms 11_13

Badische Anilin-und Soda-Fabrik 206, 209, 211, 214, 220 Bahrdt, H.P. 115 Barker, T.C. and Lévy-Leboyer, M. 225, 238 Barley, S.R. and Freeman, J. 128 Baron, A. 175, 176, 181, 183, 200 Barzel, Y. 158 Bäumler, E. 205, 208, 209, 220 Bayer Company 205_16, 220 Bayley, C. 181 Becker, G.S. and Murphy, K.M. 131, 134 Beer, J.J. 220 Berg, M. 33, 48;and Hudson, P. 18, 27 Berghoff, H. 238; and Möller, R. 239 Birke, M. 74, 90, 102 Bischoff, J. 29, 37 Bitterauf, T. 239 Blewett, M.H. 182, 200 Blum, J.D. 168 Blumin, S. 88 Bongartz, W. 229 Borschied, P. 206, 220 Braverman, H. 6, 76 Breuilly, J. 78 Breunig, W. 220 Brockstedt, J. 229, 239 Bryant, C.G.A. and Jary, D. 16 Büchner, F. 227 Burawoy, M. 80, 93, 113 Burns, T. 79;and Stalker, G.M. 127, 129, 133 Burr, C. 181 business enterprises see firm/s

240

INDEX

Buttrick, J. 48, 52 Canning, K. 78 capitalism 29_30, 46, 48_9, 52, 60, 82, 89 Carnes, M.C. 183 Caro, H. 206 Carter, E.C. et al. 238 Chandler, A.D., Jr 11_12, 16, 75, 131, 203_4, 209, 223, 227 Chandler paradigm 203 Chapman, S.D. 24 Chesnais, F. 120, 131 Cheung, S.N.S. 69, 71 Church, R. 238 Clark, G. 4, 21, 34, 52, 55, 58_62, 70; and van der Werf, Y. 71 Clark, J. et al. 16 class: defined 86_7;formation 77_8; structures 85_9 class relations 85_9, 111;at structural core 89;and conflict 88_9, 94; dichotomous character of 87_8;in social systems of production 89_102; as system-specific power resources in workplace/intrafirm industrial relations 102_10 Clawson, M.A. 201 coal-tar dysteuffs industry 216_17;and company welfare facilities 214_16; development of 205_7;managers of 207_9;and terms of labour 209_11, 212, 213_14 Coase, R.H. 48, 69 Cochran, T.C. 238 Cockburn, C. 200 Cohen, J.S. 49 Cohen, M. and Levinthal, D. 123 Coleman, D.C. 27 Cooke-Taylor, R.W. 19 Coombs, R. et al. 120 Cooper, P. 177, 182 cotton picker, adoption of 143_4;and costs of contracting 144, 146_7; and cotton gins 147_51; organisation before 140_2 Craib, I. 16 Crozier, M. and Friedberg, E. 80, 105 Cruden, R.L. 194, 202 Csontos, L. 70 Daniels, G.W. 25

Dasgupta, P. and David, P. 120, 122, 124 David, P.A. 65;and Foray, D. 120, 122, 124 Davidoff, L. and Hall, C. 238 De Bresson, C. and Amesse, F. 134, 137 division of labour 109_10, 204;and change in market structure 132;and coordination 130_3;and crafts production 52_3;effect of 54_5;and factory system 52;and gender 11; and hand-to-mouth buying 69;and market expansion 132;and process innovation with unchanged output structure 132;vertical/horizontal 13 Dornseifer, B. 220 Dosi, G. 133 Drösser, E. 214 Dublin, T. 176 Duisberg, C. 209, 214 Dunn, R.W. 202 economic approach 6, 14, 115 Edwards, R.C. 76 efficiency criteria 2, 36;assignment attributes 37;evaluation of 50_1; incentive attributes 37; objections to 38_9;product flow attributes 35, 37; and transportation improvements 39_40;weakness of 37_8 Ehrhart, E.J. 220 Elbaum, B.L. and Wilkinson, F. 90 Emery, F.E. and Trist, E.L. 128 employment relations: and class relations/structures 85_9; hierarchical 96_9;vertical 100 Englander, E.J. 39 factory system 41, 71_2;advantages of 28, 29_30, 32_3, 51_2, 56;before Industrial Revolution 21_3; challenges to explanations of 20_1; defined 31, 47_9;disadvantages of 28_9;and durability 55_62;and economies of scale 31_2, 52; efficiency/exploitation in 34_5, 63_5;essential features of 27; evolution, explanation and the inevitable 63_5;and fixed costs 55_62;growth of 4;internal organisation of 35, 37_9;and the ‘new left’ 29_34;origins of 45_7; and outwork/subcontract labour

241

INDEX

17_18;power relationships in 30_1; raison d’être of 45_6;rise of 17_21, 26_9;and routine 50_5;spread of 32_4;success of 59_60; supervision/ discipline in 29_34, 55_62;and technology 19_20, 23_6, 27_8;and transaction costs 8, 35, 37_9;and transport improvements 39_41 factory/ies: craft production 79;defined 18_19, 47_9;discipline/supervision in 49, 50;hand 22_3, 33, 34, 41; inside contracting 49, 50;pauper 22; and putting out system 21_2, 37, 40_1, 49, 50_1 Fama, E.F. and Jensen, M.C. 163 family firm 12_13, 204;in advanced stage of industrialisation 228_31; in Bavaria 225_6;debate about 221_4; defined 224_5;in early stage of industrialisation 226_8;heirs in 231_2;and the middle classes 232_4;portrayal of 221 Fang, H. 47, 51 Faue, E. 175 Feldenkirchen, W. 220 Fergusson, A. 72 firm/s: as ‘bossless’ 8_10;boundaries of 115, 133_6;and class relations/ structures 85_9;and common ownership of production 157;conceptualisation of 73_5; contextualist approach to 82_5; control/agency problem 157;and critical social theory 82_5;and decentralisation 157; development/ characteristics of 203_5;external environment of 128_30;hierarchical 10, 136;as incompletely organised ‘political coalition’ 110_16;as inputoutput machine 76, 80;inter-firm collaboration 136_7;internal context of 75_6;limitations of conventional theory 75_9;micropolitics of 81_2, 95, 109, 111;relational approach to 76, 78, 80;rights assignment problem 157;and routines 157_8; social approach to 76_9, 80_1;as soft/ dynamic 111; synthesis approach to 75_6, 77; vertical integration of 8, 149 Fischer, E. 220 Fite, G.C. 140, 142, 151

Flechtner, H.-J. 209 Flinn, M.W. 23 Flugge, E. 69 Foray, D. 120 Ford, H. 178, 184, 187, 189 Frame, J.D. 9 Freeman, C. 120 Freudenberger, H. and Redlich, F. 22 Furniss, E.S. 22 Gabin, N. 178, 200 Gall, L. 233, 234 Gartman, D. 178 Gaventa, J. 109 gender 173_5;automobile industry as male domain 175_7, 179_80; community of men 190_3;and division of labour 11, 109;and employment of women 174, 177, 178, 179_80, 193;fraternalist strategy 174, 192, 193_6;hard work, useful products, masculinity 186_8;masculinity and mass production 184_96;masculinity in period before mass-production 181_4;production of family man 188_90 Giddens, A. 5_6, 12, 13, 16, 82_5, 87, 97 Goetzler, H. and Schoen, L. 238 Golden, C.S. and Ruttenberg, H.J. 107, 114 Goldin, C. 176 Gould, S.J. 64 Grandhomme, W. 209_10, 214, 220 Greener, W.W. 25 Grossberg, M. 183 Grossman, S. and Hart, O.D. 172 Haber, L.F. 220 Hagedoorn, J. and Shackenraad, J. 120, 131 Hannah, L. 238 Hannan, M.T. and Freeman, J. 128 Hansmann, H. 172 Hareven, T. 96 Harte, N.B. 28 Hartmann, H. 200 Hayek, F.A. 64_5 Health Care Financing Administration (HCFA) 166, 170 Heaton, H. 19_20, 21, 22 Held, D. and Thompson, J.B. 5, 16

242

INDEX

Hellige, H.-D. 239 Hewitt, N.A. 182 Hilpert, H. 238 Hoechst Company 205_16, 220 Hoerr, J.P. 104 Homburg, H. 238 Hromadka, W. 220 Hudson, P. 25, 39 human engineering 101 IICAs see inter-institutional collaborative agreements Industrial Revolution 2_4, 21_3, 27, 34, 48 information 125_6, 136;factual 126; and noise 126;patterns of 126_7; potential/stored 127 inter-institutional collaborative agreements (IICAs) 10, 120, 121, 136, 137 intrafirm/workplace industrial relations 96;and balances of power 107_8; and class relations 102_10;and control regimes 97_8;in craft production 100_1;decision-making 96_7;formalized zones 99; intermediary positions 98_9;and networks 101_2;patterns of communicative relations 100_2;and power resources 100 Jacobi, F. 210, 215 Jaeger, H. 220 Janiewski, D. 183 Jenkins, D.T. 19, 26 Jensen, M.C. and Meckling, W.H. 157 Joint Commission Model 165_9 Jones, G. and Rose, M. 234, 238 Jones, S.R.H. 18, 20_1, 24, 25, 33, 37_9, 46, 47, 50, 69 Jost, T.S. 167 Kaku, S. 220 Katznelson, I. 77 Keene, A.P. 182 Keller, B.K. 110 Kennedy, J. 25 Kessler-Harris, A. 176 Kießling, B. 84 Kimball, D.S. 54 Kimmel, M.S. 183, 201 Kirby, J.T. 151

Kirby, M.W. 25 Klein, B. et at. 52 Kleinberg, S. 96 Kleinschmidt, C. and Welskopp, T. 76 knowledge 8_10, 120_1;basic considerations 121_5;changes in 134_6;collocation of 157; correlational function/local character of 122_3;and division of labour/ coordination 130_3;and external environment of firms 128_30;and information 125_8;and organisational boundaries/ environmental change 133_6;as retrieval/interpretative structure 121_2;and search activities 123_4; as subject to path dependence 123;as tacit or codified 124 knowledge base (KB) 124_5 Kocka, J. 77, 204, 206, 220, 226_7, 228_9, 233, 238;and Siegrist, H. 207, 229, 234 Kockershied, J.W. 206 Kogut, B. and Zander, U. 157 Kossoudji, S.A. and Dresser, L.J. 177 Kotthoff, H. and Reindl, J. 79 Küpper, W. and Ortmann, G. 79 labour 205;bargaining power of 93, 94; coercion of 58, 59_62, 70;and company welfare facilities 214_16; discipline of 58_62;exploitation of 58;flexible redeployment of 48_9; games played by 93, 94, 113;hiring of 48;and male alienation 173; management of 1_4;marginal productivity of 56_8;and phenomenon of hold-up 52;process 76_7;resistance 4_6; shirking/ embezzlement by 7, 50, 58, 61, 70;skills of 52_3, 54_5, 62, 69, 71, 79, 93, 111_12, 192;supervision of 56, 62, 70, 71;terms of 209_11, 212, 213_14;welfare facilities 214_16;workplace relations 90_6; see also management/worker interaction Landes, D.S. 3, 12, 20, 26, 27, 28, 29, 30, 31, 32, 39, 41, 46, 51, 222, 238 Langlois, R.N. 70, 72;and Cosgel, M.M. 71;and Everett, M.J. 72;and Robertson, P.L. 49, 50, 69, 157 Lawrence, P. and Lorsch, J.L. 129

243

INDEX

Layton, E.T. 122 Lazonick, W. 16, 25, 32, 48, 55, 59, 70, 71 lean production 102 Leber, A. 220 Lederer 238 Leibowitz, L. 200 Leijonhufvud, A. 47, 52, 58 Lévy-Leboyer, M. 238 Lichte, R. 91, 92, 94 Liebowitz, S.J. and Margolis, S.E. 65 Loeb, E. 207 McClelland, K. 176, 181 McKelvey, B. 128 Maier, F. 151 management 76;autonomy 101;and blocking of power 107_8;in coal-tar dyes tuffs companies 207_9; constraints on 80;hierarchies 203, 204_5;as production process 80; strategic 98_9;structures 101_2; uncertainty of 108 management/worker interaction 4_6, 16;and hold-up 8;and metering problem 7;power struggles 113_15; power/authority relations 6;and structuration 4_5;and supervision 7; see also labour Mantoux, P. 3, 19, 26, 27, 28, 30, 31, 68 Marglin, S.A. 2_3, 20, 29_31, 32_4, 58, 63_4, 68, 71, 72, 98;views/ critics of 2_4 market/s 1, 13, 85, 133;and division of labour/volume effect 54_5, 56; extent/predictability of 54_6;and the status quo 71_2 Marquart, F. 185, 195 Marquis, S.S. 188_9, 190, 201, 202 Martin, R. 229 Marx, K. 86 masculinity see gender Mathias, P. 26, 27 May, M. 188, 190, 200 Mead, J.E. 192 mechanisation of cotton 140;adoption of cotton picker 143_4, 146_51; organisation before cotton picker 140_2 Metcalfe, J.S. and Gibbons, M. 130 Meyer, S. 184, 185, 189, 194, 195 Milgrom, P.J. and Roberts, J.D. 157

Milkman, R. 177, 180, 200 Minkler, A.P. 53, 69 Minssen, H. 80, 97, 106, 107 Mintzberg, H.A. 16, 129, 158, 160, 172 Möhl, F. 229 Mokyr, J. 70 Montgomery, D. 112, 181, 201 Mosse, W. 239 Mytelka;L.K. 120, 131 Nell, A.von 239 Nelson, E.G. 22 Nelson, R.R. and Winter, S.G. 53, 123, 124, 130, 157 networks 1, 8_10, 85;and new technologies 10;and professions 9_10, 158_61, 162_3 Nevins, A. 189 New Institutional Economics 45 New Left 3, 20, 29_34 Nipperdey, T. 238 Nohria, N. 8;and Eccles, R.G. 16, 158 North, D.C. 46, 58 Nuwer, M. 90, 101, 112 Okochi, A. and Yasuoka, S. 238 organisations see firm/s Parr, J. 200 Patterson, R. 21 Paulsen, A. 238 Payne, P.L. 238 Phillips, A. and Taylor, B. 175 Pierenkemper, T. 238 Pinnow, H. 206, 220 Piore, M.J. and Sabel, C.F. 3 Plowman, D.H. 98 Plumpe, G. and Scultheis, H. 208, 209, 220 Podmore, S. 32 Pohl, H. 204, 220, 229, 230 Polanyi, M. 124 Pollard, S. 3, 25, 28, 49, 50, 54, 62, 69, 71 Pollay, K. 215 Pollert, A. 200 Powell, W.W. 158 power resources 2, 6, 102_3;allocative 103_4;balances of 107;blocking of 107_8;coercive 161;discretionary 105;distributive 110;in factory system 30_1;formal distribution of 244

INDEX

106;organisational 104;to command hierarchical authority 106_9;to define and determine 109_10;to implement and control 104_6;to make decisions 103_4; and zones of uncertainty 103, 105, 106_7 Price, R. 79 Pries, L. 91 production milieus see social systems of production production process 77, 79, 173;life-cycle pattern 132_3;professional mode of 157_63 professions/professionals: and authority 161_2;and autonomy 161;as bossless 9;characteristics of 159;and coercive powers 161;and complementary institutions 163_5; internal/external relationships 159_60;and Joint Commission Model 165_9;mode of production 157_63;networks 158_61, 162_3; occupational independence of 9_10; and reputations 162 Prunty, M.C. and Aiken, C.S. 147, 150_1, 153 Putterman, L. and Kroszner, R.S. 7 Raff, D.M.G. 184, 185;and Summers, L.H. 185, 195 Redlich, F. 206, 220 Reeves, A.W. and Kimber, C. 182 Reitell, C. 178, 189 Renner, A. 185 Revealed Technological Performance (RTP) 124_5 Richardson, G.B. 157 Richter, P. 220 Riesser, J. 207 Roberts, J.S. et al. 165 Roberts, P.C. and Stephenson, M. 68 Robertson, P.L. and Alston, L.J. 5, 6, 18, 53, 54, 70, 91, 111, 112, 151 Rodgers, D.T. 187 Rose, M.B. 22 Rose, S. 175_6 Rosenberg, C.E. 166 Rosenberg, N. 13 Roth, H. 238 Rotundo, A. 183 Rumely, E.A. 196 Russel, J. 195

Sabel, C.F. and Zeitlin, J. 3, 26 Sachse, W. 225, 226 Samuel, R. 18 Savage, D.A. 9, 53, 69, 171 Saviotti, P.P. 121, 126_7, 132, 134 Sayer, R.A. 76 Scheid, M. 238 Schmitz, C.J. 203 Schulman, B.J. 151 Schumann, D. 227, 228, 230, 231, 238, 239 Schumpeter, J.A. 125 Schuster, C. 220 Schwellien, J. 220 scientific theories 121_3 Scranton, P. 74 Shannon, C.E. and Weaver, W. 126 Siegrist, H. 229 Simon, H.A. 48 Sinclair, U. 184_5, 188 Smith, A. 130, 134 social systems of production 89_90; associational group structures 93_5; functional group formations 94_5; hierarchical employment relations 96_9;patterns of communicative relations 100_2; shopfloor communities 93, 95, 96, 101; workplace relations 90_6 sociological approach 6, 14, 76_9, 80_1, 82_5, 115 Somers, M.R. 78 Stahl, W. 231 Stigler, G.J. 131, 134 Stillman, K.W. 69 Stinchcombe, A.L. 53, 158 Stoneman, P. 120 Street, J.H. 151 structure see agency/structure Stürmer, M. et al. 232, 239 Styles, J. 38 Szostak, R. 39_40 Taylorism 101 technological change 4, 13, 13_14, 30_1, 55, 134_6, 149 technology 39_41, 41, 50_5, 70_1; efficient 46_7;and exploitation 34_5;and the factory system 23_6; implementing/using new 2, 5, 148_9, 182_3;in Industrial Revolution 2_4;role of 33_4

245

INDEX

Teece, D.J. 124, 125;and Pisano, G. 157 Temin, P. 49, 69 Thompson, E.P. 77 Thompson, P. 182 Tilly, C. et al. 103 Tilly, R. 204, 220 Timmins, S. 18, 25 Tolliday, S. and Zeitlin, J. 75, 101, 108, 112 transaction costs 14, 39_41, 158;of authority and control 7_8;and coordination costs 133, 134; and the factory system 35, 37_9;and growth of capital-intensive production processes 8;levels of 8;and location of interfirm boundaries 134;and putting out system 51 transport 39_41, 147 Tushman, M.L. and Anderson, P. 123, 125

Valverde, M. 175 Voth, H. 71

Ulrich, W.E. 202 Ure, A. 18, 19, 47, 69 Usher, A.P. 17

Zeitlin, J. 77 Zunkel, F. 238 Zwahr, H. 77

Wadsworth, A.P. and Mann, J. de L. 24, 31 Weber, M. 73, 86_7, 111, 155, 158 Wehler, H.-U. 238 Welskopp, T. 74, 78, 81, 85, 87, 88, 90, 91, 94_6, 99, 108, 110, 112 Weltz, F. 81 Whatley, W. 142, 153 Williamson, O.E. 7_8, 20, 35, 37_41, 46, 48, 50, 52, 58, 75, 133, 152 Wing, C. 19 Woodward, J. 129 workers see labour working class 77_8, 95 workplace relations see intrafirm/ workplace relations Wright, G. 151

246