The New Old Economy: Networks, Institutions, and the Organizational Transformation of American Manufacturing

THE NEW OLD ECONOMY

This page intentionally left blank

The New Old Economy Networks, Institutions, and the Organizational Transformation of American Manufacturing

JOSH WHITFORD

1

3

Great Clarendon Street, Oxford ox2 6dp Oxford University Press is a department of the University of Oxford. It furthers the University’s objective of excellence in research, scholarship, and education by publishing worldwide in Oxford New York Auckland Cape Town Dar es Salaam Hong Kong Karachi Kuala Lumpur Madrid Melbourne Mexico City Nairobi New Delhi Shanghai Taipei Toronto With oYces in Argentina Austria Brazil Chile Czech Republic France Greece Guatemala Hungary Italy Japan Poland Portugal Singapore South Korea Switzerland Thailand Turkey Ukraine Vietnam Oxford is a registered trade mark of Oxford University Press in the UK and in certain other countries Published in the United States by Oxford University Press Inc., New York ß Josh Whitford The moral rights of the author have been asserted Database right Oxford University Press (maker) First published 2005 All rights reserved. No part of this publication may be reproduced, stored in a retrieval system, or transmitted, in any form or by any means, without the prior permission in writing of Oxford University Press, or as expressly permitted by law, or under terms agreed with the appropriate reprographics rights organization. Enquiries concerning reproduction outside the scope of the above should be sent to the Rights Department, Oxford University Press, at the address above You must not circulate this book in any other binding or cover and you must impose the same condition on any acquirer British Library Cataloguing in Publication Data Data available Library of Congress Cataloging in Publication Data Data available Typeset by SPI Publisher Services, Pondicherry, India Printed in Great Britain on acid-free paper by Biddles Ltd., King’s Lynn, Norfork ISBN 0-19-928601-9 978-0-19-928601-0 1 3 5 7 9 10 8 6 4 2

Acknowledgments This all began in 2000 when I was brought on as a research assistant at the Center on Wisconsin Strategy (COWS). I was to take over a project that JeV Rickert had begun because JeV was moving on to the AFL-CIO. The project was an extension of COWS’ long-time interest in getting the Wisconsin economy onto a ‘highroad’ path by building institutions to push companies to be integral parts of strong communities and to provide workers with good jobs, training, and career prospects. The project focused on the implications of outsourcing—of decentralized production—and asked how and whether the small and medium-sized factories that employed so many more workers than they had in years past might Wt into that high-road vision. Wisconsin, after all, has an economy that depends on manufacturing to a degree paralleled by few other American states. I was interested—I had grown up in the Midwest and had seen the consequences of deindustrialization Wrst-hand. However, I told JeV, there was one big problem. I knew plenty about sociology, political economy, and geography; I had even had a year of graduate training in microeconomics. I was pretty sure I could cover that end of things. What I didn’t really know all that much about was manufacturing. He replied: ‘That’s a mixed blessing. You can ask them’ (meaning the people I was to interview). So that’s what I did. And he was right, It was a mixed blessing, though with more good than bad. I was very lost for a while. But I also began with about as few preconceptions as a social scientist can have about the empirical matter before him. JeV was around some to help me, and was always good for a phone call, but for the most part it was just me soaking up whatever I could from whomever I could get to talk to me. I hope those fresh eyes come through in this book, that I have managed to combine what I did know (social theory and political economy) and what I did not (the ‘old’ economy) in ways fresh to both. And I hope that this freshness does give some insight into how that high-road manufacturing economy might be achieved in a world of decentralized production. It is no longer the case that I don’t know all that much about manufacturing. I eventually learned enough to begin writing policy documents that were useful even to the people I was interviewing. That ledger, however, remains red; my debts to the many people in the Midwestern manufacturing economy who gave me that knowledge are greater than words can acknowledge. Most of these people cannot be named here. I guaranteed anonymity to all I interviewed so that they could speak freely even about sensitive subjects (the details of the interview process are in the Appendix). I have, however, also met some people along the way whose names I can and should mention as I know them in contexts outside the formal data-gathering process, with its human subjects approval and such whatnot. Foremost among these is Paul Ericksen, who has not only inspired me in this research but also in my life. He has become a

vi

Acknowledgments

treasured friend. Mike Klonsinski of WMEP, beyond Wnding ways to help support COWS Wnancially, has been free with ideas. Mike Schmitt and Trish Broskowski have done me favors for no good reason except that they are good people. The Wisconsin Regional Training Partnership and its employees and associates have been helpful. There are many others, but they must, unfortunately, remain nameless here. There are also many to thank inside the academy. This book began as the (now much better developed and presented) set of ideas that got me my Ph.D. from the University of Wisconsin and an award for best dissertation in sociology at that university in 2002–3. There, I had an excellent advisor with three heads, all of whom should be richly thanked. Jonathan Zeitlin, the closest thing to a library with legs I think I will ever meet, was (and is) always willing to read, comment, discuss, and tell me how if I would just read something or other else that he indicated, it would change my thinking (he was often right). Erik Olin Wright is a human analytic mill into which students can send any paper to discover every—and I do mean every—Xaw in an argument, all typed out in a relentlessly constructive and cheerful manner. He has won multiple teaching awards and as near as I can tell, he deserves them all. Joel Rogers painted the big picture to which I am still today committed, with his vision of a very diVerently functioning and more democratic economy. Gary Herrigel of the University of Chicago has been a good friend and an essential outside advisor. Without his counsel and discussion, I would never have understood what I was Wnding in the way that I did. Others at the University of Wisconsin to remember here include the other two members of my committee, Jamie Peck and Chas Camic (who also advised my master’s thesis), Gay Seidman (who should be counted as family), my coworkers at COWS, participants in the economic sociology seminar, and my fellow graduate students. In this last group, there are too many to name so I will not even start, save to say that two others, Matt Vidal and JeV Rothstein, who are today dear friends, are also bringing political economy back into economic sociology and doing good things in their own analyses of changes in American and global manufacturing. This book was not written at Wisconsin alone. I am very grateful to Wolfgang Streeck for oVering me a postdoctoral fellowship at the Max-Planck Institute for the Study of Societies from September 2003 to July 2004. A dissertation, even a good one, is very diVerent from a book, so I was fortunate that year to have that most precious of commodities: unobstructed time to write surrounded by very smart and interesting people. From that time, particular thanks go to Till Mu¨ller-Schoell, Helen Callaghan, Cornelia Woll, and Sabina Avdagic. Ulrich Glassmann at the University of Cologne was and is always good for Ko¨lsch and conversation on matters academic and not. Chuck Sabel has provided me with useful thoughts on my ideas and writings at various times over the years, and has always been very gracious in letting me use advance copy of his own work. Mark Granovetter read the manuscript at a fairly advanced stage and provided long

Acknowledgments

vii

and very insightful comments. The students in my comparative capitalisms seminar at Columbia had perceptive things to say about that advanced draft as well, and gave me conWdence that this book really is accessible and interesting to more than just specialists. I also have a real life, where there are again more people to thank than is possible here. I will begin, however, with my grandfather, Albert Whitford, an extremely accomplished astronomer who passed in 2002 but who ingrained a deep respect for scholarship and integrity in all his descendants, and then go on to his son, William Whitford, who is himself in the family (professoring) trade. I was never pushed to follow, but could still count on ceaseless encouragement and knowledgeable thoughts about both the research and the career (perhaps he was simply wily enough to recognize that the best way to get me to do something is to never push me to do it!). The support of the rest of the family, both nuclear and extended, has also been invaluable. In my nonfamily life, Issa KohlerHausmann deserves a mention for listening to many an addled thought back when I was beginning this project. Finally and especially, Sigrun Kahl deserves unquantiWable thanks for providing so much support in this last year even as she had to see vacations vanish into my screen and put up with lots of blathering about the fundamental role that anxiety must surely play in the relationship between manufacturing and social theory. Funding for the research on which this book is based was provided by COWS. Between 2000 and 2003, I was a research assistant at COWS, largely serving as part of the Advanced Manufacturing Project (AMP), a multistate consortium dedicated to investigating the determinants and possible policy supports of improved performance in the component-manufacturing sector. The principal investigators of AMP were Jonathan Zeitlin and Joel Rogers (University of Wisconsin at Madison), Gary Herrigel (University of Chicago), Dan Luria (Michigan Manufacturing Technology Center), and Susan Helper (Case Western Reserve University). AMP was funded primarily by the Alfred P. Sloan Foundation, with important additional funding provided by the Wisconsin Manufacturing Extension Partnership. My conversion of the ideas in that dissertation I wrote at Wisconsin into this book was funded Wrst by the Max Planck Institute for the Study of Societies, with a postdoctoral fellowship in the 2003–4 academic year, and since as a part of my job as an assistant professor at Columbia University. New York January 2005

J.D.W.

This page intentionally left blank

Contents Acknowledgments Introduction I The New Old Economy Introduction to Part I 1. A New Production Paradigm for a New Old Economy 2. Networks, Noise, and Institutional Change II

III

Networks and the Organizational Transformation of American Manufacturing Introduction to Part II 3. The Decentralization of American Manufacturing 4. Collaboration in Practice: The Cost Reduction (Incremental Innovation) Waltz 5. Uncertainty and Contradiction in the New Old Economy Institutions and the Relational Reconstruction of Regional Political Economy Introduction to Part III 6. It Couldn’t Happen Here? Public Policy, Regional Institutions, and InterWrm Collaboration in the United States 7. Toward the Relational Reconstruction of Regional Political Economy

Appendix Notes References Index

v 1 7 9 15 26

49 51 57 76 95

121 123 129 154 162 178 194 205

This page intentionally left blank

Introduction American manufacturing has seen a restructuring so fundamental in the last quarter century that its magnitude is almost hard to overstate. The biggest headlines, however, have been reserved for two developments that emphasize decline. We hear much about globalization and the outsourcing of production, which have permitted the near wholesale transfer of labor-intensive manufacturing to lower-wage areas in the developing world; and the American economy has been ‘deindustrializing’ at a rapid clip, losing roughly 3 mill ion manufacturing jobs in the Wrst four years of the twenty-Wrst century alone. All told, by the end of 2004, the sector had lost 44 percent of the 22 million manufacturing jobs it had at its 1979 postwar peak, falling from 23 percent of the labor force to just 13 percent.1 Does this mean that a book about changes in the organization of American manufacturing is, as the cliche´ goes, a mere chronicling of the rearrangement of chairs on the decks of the Titanic? It is, after all, not hard to Wnd assorted pundits, academics, and politicians to soothingly and correctly remind us that globalization and deindustrialization per se are nothing to worry about, and are on balance probably for the best for nearly all involved. The shifting of laborintensive production abroad lowers prices domestically and can drive development in areas of the world where it is sorely needed. Deindustrialization, as Rowthorn and Ramaswamy (1997: 2, 1) write for the International Monetary Fund (IMF), has occurred across all the world’s advanced economies and is largely independent of North–South trade: ‘in the 23 most advanced economies, employment in manufacturing declined from 28 percent of the workforce in 1970 to about 18 percent in 1994.’2 This widespread decline in manufacturing employment, they note further, ‘is not a negative phenomenon, but a natural consequence of further growth in advanced economies’. Demand for manufactured goods in the developed world has been stable over the last quarter century, which means that the rapid productivity growth in manufacturing relative to most service sectors comes accompanied, all else equal, by a necessary relative decrease in its employment share. Yet I have written this book about the implications of what I show to be a very substantial restructuring of American manufacturing, premised on the transfer of many productive responsibilities from large manufacturers to their smaller suppliers. And I do believe that what I say shows that what has happened, is happening, and will happen in American manufacturing is important for academics and policymakers alike. I readily accept that deindustrialization, globalization, and outsourcing are necessary correlates of growth in advanced economies, but I also show that there is nothing ‘natural’ about them. They are political and highly diVerentiated processes that play out in strikingly diVerent ways within the United States and across the developed world, with

2

Introduction

varied distributional consequences, in response to choices made by Wrms, states, and workers embedded into particular institutional and historical contexts. The research for, and writing of, this book took place with the American manufacturing sector as a whole in deep crisis: record numbers of manufacturing jobs were lost; stories about factory closures attributed to global competition, especially from China, had migrated from the back of the business section to the front page; and the American trade deWcit was soaring to new records almost monthly, imperiling the economic well-being of future generations and making clear that the travails of American manufacturing are a matter of more than parochial concern. Moreover, in the (large) parish of the American Upper Midwest, where manufacturing weighs heavily on state economies and where my empirical research has been located, job losses between 2000 and 2003 were especially punishing, occurring much faster than mere productivity growth or cyclical demand eVects would predict. It has been a time (hardly the Wrst) when the particular patterns of globalization, outsourcing, and deindustrialization have been quite obviously painful for workers and the communities in which they live—more so than necessary—and a time when there is ample reason to worry even for those who do not subscribe to the errant belief that jobs in manufacturing are somehow intrinsically (rather than contingently) better than those in services.3 Nonetheless, I certainly recognize, as Rogers and Luria (2003: 1) write, that ‘Absent a revolution in U.S. trade policy, high-wage/low-value-added American manufacturing will soon be dead, . . . driven to extinction by more eYcient or less labor-friendly domestic production, or by low-cost alternatives abroad’. But, as Rogers and Luria also note, there is much more to the story. Grounding my arguments and analyses in over 100 interviews at more than 50 metalmanufacturing Wrms in the American Upper Midwest, I show that there is great diversity in American durable manufacturing. There are many Wrms even in the central and historic ‘rust belt’ that are somehow groping their way towards the formation of globally competitive, highly Xexible, collaborative production networks, which allow them to jointly improve old products and processes and to develop new ones more rapidly than ever before; yet at the same time these eVorts have been contradictory at best and are only weakly supported by the existing American institutional infrastructure. This ‘more to the story’, I thus argue, is that there is enough new in the old economy to ensure that the ostensibly rusted metal manufacturers need not be written oV as an undiVerentiated mass invariably destined for a slow boat to the low-wage world, hopefully to be supplanted by an ever-expanding service sector. But the transition is a spotty one, with enough failures to give pause, to remind social scientists as well as policymakers, Wrms, and workers that eVorts to retain the remaining core of good-paying manufacturing jobs must be based on a real understanding—neither sugarcoated nor despairing—of what is possible in the high-wage world, of the feasible and the desirable in a world of global competition, short product cycles, and relatively unstable demand.

Introduction

3

I build this case around three main arguments. The Wrst follows directly from the observation that large American manufacturers—like large manufacturers across the developed world—have outsourced much of their productive capacity to smaller suppliers both in the United States and abroad. This devolution is not just quantitatively important. It represents a qualitative break with the recent past, the emergence of a new old economy in which most of what matters to manufacturing Wrms no longer happens under roofs they own or control. This has made the quality of relationships between Wrms much more important and their structure much more complex, even in the ostensibly mundane world of metal manufacturing. How (and where) these large Wrms choose and direct their armies of suppliers has tremendous consequences for the regional economies in which they are embedded, because Wrms selling in the more proWtable markets where competition depends as much on innovation and quality as it does on price are more likely to use skilled and better-paid workers. Importantly, these decisions are aVected by much more than trade policy, vagaries of currency markets, international regulatory arbitrage, and the like. They depend a great deal on the particularities of Wrms’ embedding in particular historical and institutional contexts. This recognition has led many of those who sought to understand the economy by looking at what happens inside large companies to also devote substantial attention to how activities are coordinated and governed between companies. My second core claim is that the existing social scientiWc literature has misdescribed the American transition to a new old economy. This is not to say, however, that it has gone unnoticed. A vibrant literature in sociology, political economy, and the business press argues that the demise of the centralized Fordist Wrm creates two fundamental and starkly bifurcated realities for Wrms and the regions in which they are embedded. The decentralization of production can entail the exploitation of market power, cost-shifting, union avoidance, and the chasing of lower wages; but it also creates the possibility of a normatively attractive new production paradigm, better for all stakeholders, premised on the creation of collaborative networks of Wrms that exploit multilevel relational networks to jointly compete in more proWtable qualityconscious niche markets. Which path is taken, the usual argument goes, is deeply aVected by path dependencies in regional, cultural, and institutional conWgurations—which does not bode well for the American Upper Midwest with its history of relatively atomized hierarchical and market governance. What I argue, by contrast, is that the reality is much more nuanced and that this nuance matters. Many large manufacturers in the region are opting to follow the prescriptive tenets of the collaborative new production paradigm even as their eVorts are deeply constrained by the need to hedge fundamental uncertainties caused by a history of poor relationships and a lack of institutional support. The result is a relational structure that is neither the collaborative production network that theorists such as Powell (2001) call the very building block of the twenty-Wrst century Wrm, nor is it an atomistic world of hostile arm’s-length

4

Introduction

contracting. Rather, it is a complex mix of the two, suggesting that the possibilities are considerably less bifurcated than the existing literature would have them be. My third claim goes to the institutional and policy implications of this misdescription, and bears particular topical relevance given the obvious diYculties faced by many American manufacturers. I make no pretense to know the silver bullet for all that ails American manufacturing, nor do I speak to all of the many policy arenas that aVect its fortunes. Rather, I focus on what follows from my second argument: Wrms’ microdecision processes can be signiWcantly and positively aVected by adjustments in economic development policy at the state and local levels (the responsible parties within the American federal structure). I acknowledge that the institutional legacy and the historic dominance of hierarchical and market governance in the American Upper Midwest do present genuine challenges to the region’s ability to stably sustain a high-collaboration decentralized manufacturing model. However, recognizing the empirical untenability of conventional and overly bifurcated understandings of the relational options employed by American manufacturers illuminates key barriers to collaborative network production and the tools for their resolution. Exploring complexities in relationships between manufacturers and in the strategies they employ, I argue, shows how the existing American economic development apparatus can be modiWed to support manufacturers’ very partial, problematic, but nonetheless promising eVorts to engage in regionally tied but globally competitive collaborative production models—that is, in the sorts of production models that might help American deindustrialization to in fact become the slow, steady, and relatively painless process it is sometimes (wrongly) advertised to be.

The Chapter-by-Chapter Structure of the Argument The book is divided into three parts. The introduction to Part I describes the changing patterns of American deindustrialization and argues that they are a consequence of radical changes in the organization of the American productive model. Chapter 1 then reviews the academic literature on what is in fact a worldwide transition from old to new old economy. It shows that there is consensus as to the general contours of a normatively desirable ‘new production paradigm’, but that there are disputes as to the degree to which it can be fully achieved in the American context. Chapter 2 examines some of the social theoretical implications of the increasing decentralization of production, and establishes a core theoretical claim of the book: prominent sociological theories of economic coordination too quickly dismiss systematic contradictions and hedging behavior by Wrms actively seeking to build collaborative network forms of organization. In so doing, these theories analytically obscure the need for, and possibilities of, policymaking to help build and sustain normatively desirable collaborative production models.

Introduction

5

Part II (Chapters 3–5) relies heavily on a case study of metal manufacturing in the American Upper Midwest—that is, a case study of the quintessential highwage manufacturing industry in the quintessential manufacturing region: just 15 percent of national employment is in the great lakes states, but one in four durable manufacturing jobs is located there. Extensively using direct quotes and observations from 100þ interviews conducted between 2000 and 2002 with large global manufacturing Wrms and their more territorially bound suppliers, I describe the enormous changes in the organization of American manufacturing. In the wake of these changes, Wrms are not simply making the strategic decision of whether to collaborate or not, but are instead continuously reconstituting and revising relationships as they cautiously feel their way towards (or away from) the joint deWnition and resolution of problems. These relationships are stably and systematically intermediate between arm’s-length and collaborative, and are characterized by ongoing contradictions that sit uneasily with the sociological literature on network production forms. Part III (Chapters 6–7) examines the policy implications of the Wndings in Part II and concludes the book. I argue that absent recourse to extra-Wrm institutional supports, there are clear limitations to the spread of the collaborative interWrm production that a ‘high-road’ American manufacturing economy would require. But at the same time, partial collaboration and active eVorts by large manufacturers and some of their suppliers to build long-term relationships is suggestive of the possibility of encouraging more. This contrasts prominent claims in comparative political economy that historically ‘liberal market’ economies—such as the United States—lack the business-coordinating capacity required to build such institutions. The argument relies heavily on the example of policy experiments in Wisconsin to show that it is in fact possible to mobilize latent business-coordinating capacity even in the ostensibly unfavorable context of Midwestern American manufacturing.

An Issue of Terminology: The ‘Original Equipment Manufacturer’ Throughout this book, I generally refer to manufacturing Wrms with one of two terms—‘original equipment manufacturer’ (OEM) and ‘supplier’—which I characterize as two fundamentally diVerent roles taken by organizations in today’s manufacturing (though in empirical fact, some companies do take on both roles). The latter term is straightforward: it refers to companies that sell what they make to other (usually larger) manufacturers, and that thus have their access to the Wnal consumer market mediated by those other companies. The former term—OEM—is standard in manufacturing, and refers to the (usually) large companies that sell products for the most part to retailers, though sometimes directly to consumers.4 Examples (none of the following were interviewed for this project) might include Ford, Caterpillar, Honda, Craftsman, General Electric, and so on. I use the term OEM throughout the book to refer to the large Wrms with market power that purchase components from suppliers for two

6

Introduction

reasons: Wrst, it is standard in manufacturing, and interviewees use it often; and second, using the more generic ‘customer’ risks confusion between consumers buying Wnished goods—who are not really discussed much in this book—and companies buying intermediate goods from other manufacturers, which are discussed.

Part I The New Old Economy

This page intentionally left blank

Introduction to part I Manufacturing employment weighs less heavily in all advanced economies than in years past, but how much less varies considerably. Between 1980 and 2002, as American manufacturing fell by 44 percent, from 22 percent to 13 percent of the labor force, the percentage of the German workforce employed in manufacturing fell by 31 percent, but remained at the relatively high Wgure of 24 percent. Italy and Japan saw smaller declines, 16 percent and 23 percent respectively, also to comparatively high percentages (23 and 20 percent). South Korea developed rapidly in the same years, and saw its manufacturing employment increase between 1979 and 1989 from 22 percent to 28 percent; but with continued development, deindustrialization occurred there as well, and the manufacturing sector employed just 19 percent of the labor force in 2002.5 In the abstract phrasing of the macroeconomist, these diVerences simply show that diVerent countries have diVerent historically determined relative trade specializations. Such terms are undoubtedly adequate to describe the larger global trend, with particular countries’ transitions to an inevitable and unproblematic postindustrialism reducible to variations on a single theme. Yet even a cursory analysis of patterns of American manufacturing job loss across time and space, and of the recent and dramatic increases in the American trade deWcit, makes clear that there is more that matters in the American story. Deindustrialization in the United States has been a very uneven process, which, when examined in closer detail, serves to remind that relative trade specializations depend on choices made daily by manufacturing Wrms within particular historical and institutional contexts. As can be seen in Figure 1, the decline in the number of manufacturing jobs in the American economy in the last quarter century has come almost entirely in three short steep declines punctuated by stability and even growth: between 1979 and 1982, 2.1 million jobs were lost; 1989–92 saw 1.2 million jobs disappear; and between 2000 and 2003, 2.7 million jobs vanished. The unevenness of American deindustrialization is most pronounced when one looks also at where jobs were lost across these three periods and at how the Wrst two periods of job loss diVer from the third. Between 1979 and 1992, the distribution of American manufacturing employment shifted radically. Beyond deindustrialization, the period was characterized even more strongly by deunionization, deurbanization, and decentralization (a relative shift of employment to smaller Wrms). Of the 3 million manufacturing jobs lost between 1979 and 1992, Wve in six were union, two in three were in just sixteen central cities in the

10

The New Old Economy

1979−82: 2.1 million jobs

18 1989−92: 1.2 million jobs lost

16

2000−3: 2.7 million jobs lost

14

1978 1979 1980 1981 1982 1983 1984 1985 1986 1987 1988 1989 1990 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004

Millions of American manufacturing jobs

20

Figure 1. Manufacturing employment in the United States Source : BLS

Northeast and Midwest, and three in four were in plants employing more than 500 workers.6 At the same time, there was growth in nonunion, ex-urban and small-Wrm manufacturing. Between 1980 and 1990, nonunion manufacturing employment increased by 1.3 million jobs, against a loss of 2.5 million union manufacturing jobs. And the reduction of 3 million jobs in large plants (> 500 employees) was substantially oVset by the creation in sub-500 employee plants of 2 million jobs (Luria 2000a). All told, the relative proportion of manufacturing employment at smaller establishments—those with less than 250 workers— increased by 18 percent between 1977 and 1997, from 43 to 51 percent. In key subsectors in durable manufacturing, the shift towards smaller production units was even more pronounced: for example, more than 88 percent of employment in metal fabrication was in establishments with less than 250 employees in 1997, up from 72 percent in 1977; in machinery production, the jump was from 41 to 55 percent; and even in transportation industries (which includes automotive assembly), employment at smaller production units soared by 60 percent (from 13 to 21 percent).7 Restructuring did not have to be accompanied by this painful and unprecedented assault on the American labor movement and the gutting of American central cities, but there is little doubt today that it was inevitable in some form or other. By the mid-1970s, fragmentation in once predictable markets for massproduced goods coupled with improvements in transportation and the opening of national markets had forced companies in many key end-user

Introduction to part I

11

industries—including but not limited to: automobiles and other transportation equipment; industrial, farm and construction machinery; and electrical appliances—to develop and produce a much wider palate of more customized products in ever shorter time frames. At the same time, they were shaken by the end of years of relative stability in core technologies (steel and mechanical engineering) and the need to incorporate technologies developed in other sectors, such as new materials and electronics. Some large American manufacturing Wrms failed to adjust to this rapidly altered competitive reality and simply shut down. Yet many others did not, managing somehow to actively engage the new environment. Very often, a big part of this new engagement involved retrenching to their so-called core competencies in design, marketing, and assembly by subcontracting (outsourcing) other activities to the many smaller suppliers that now do much of the ‘real’ manufacturing of components. By the mid-1980s, it was increasingly accepted across the developed world, by academics and managers alike, that the model of business organization at the center of the postwar growth model was in deep trouble. Business historian Langlois (2004: 355) notes the irony: just after Alfred Chandler published The Visible Hand, his crowning and ‘triumphalist account’ of the rise and inherent superiority of the large, vertically integrated manufacturing Wrm in all times and places, these same Wrms were rapidly collapsing and restructuring under ‘siege from a panoply of decentralized and market-like forms’. From this siege, American manufacturing came away with a deverticalized organizational structure, that is by no means simply a less urban, less unionized, and less concentrated version of what was there before. Rather, as I show in considerable detail in Part II, there was a radical reorganization of production and responsibilities both within and across manufacturing concerns. And this, as Sturgeon (2002) explains, has forced those seeking to understand the fate of industrial economies to shift attention ‘away from the logic and ramiWcations of the internal structures of the modern corporation to the external economies created by the ongoing interactions between Wrms’.8 The massive decline in American manufacturing employment that began in 2000 occurred, then, against a very diVerent and already highly decentralized national and global industrial context. It thus should not surprise that domestic patterns of employment shifts diVered from those in previous downturns. Union manufacturing employment again fell oV more sharply than did nonunion, but the diVerences are less stark than in past recessions. Rural areas have been hit at least as hard as cities and every state has lost manufacturing jobs—whereas between 1990 and 1992, for example, twelve states gained manufacturing jobs. And job losses seem to be split relatively evenly between large and small manufacturing Wrms (although the data necessary for a full accounting were not available at time of writing).9 However, if the movement of production out of large manufacturing Wrms (outsourcing) has been taken about as far as it can go, the movement of production out of the United States (oVshoring) has not.10 A long-standing

12

The New Old Economy

but heretofore relatively mild American vice, the failure to match imports with exports, spiraled well out of control at the end of the 1990s boom as the American trade deWcit grew rapidly in what is quite clearly a manufacturing— rather than a manufacturing-and-services—problem. The rapid development of new telecommunications and information technologies notwithstanding, international trade is still dominated by goods: The 2003 American goods trade deWcit of $547 billion was hardly dented by a service trade surplus of just $51 billion, a Wgure not only lower than it had been ten years previous, but that was less than the average annual growth in the goods trade deWcit between 1998 and 2004 (data from the Bureau of Economic Analysis, June 14, 2004 news release). Protests from the American administration and others fearful of protectionist impulses notwithstanding, the conclusion that some portion of the job losses in the period must be ascribed to the poor trade performance of American manufacturers is inescapable. As Bivens (2004:2-4) of the Economic Policy Institute notes, politics aside, the data speak for themselves: in real terms (which account for the price declines that follow from productivity improvements), manufacturing’s share of GDP was quite stable for the quarter century before 2000, at which point it fell by a full percentage point; more telling still, even as demand for manufacturing output remained roughly constant between 1998 and 2003, American ‘domestic production of manufactured goods . . . lagged this demand by a widening margin’ over the same period. Were this ‘wedge’ to have remained at its 1987–97 average of 90 percent, Bivens estimates, output in American manufacturing between 1998 and 2003 would have been some 16 percent higher, with employment likely not far behind. An intertwining of deindustrialization and international trade is not, by itself, anything new. As noted above, patterns of manufacturing job loss in the 1980s were driven in part by Wrms’ restructuring in response to global pressures. What is particular is that this twenty-Wrst-century manufacturing crisis has occurred within the context of an already highly decentralized industrial structure. OEMs that maintain only very minimal internal manufacturing operations face reduced barriers to chasing cheap parts made in countries with lower wages, or even to relocating the assembly process itself.11 Patterns of available trade data in intermediate goods sectors are suggestive: imports in the metal fabrication industry, for example, rose by 39 percent between 1997 and 2002 (exports increased by 5 percent in the same period); and in motor vehicle parts manufacturing, another obviously intermediate sector, imports grew by 52 percent between 1997 and 2002 (again, exports grew, but only by 8 percent).12 It would be wrong, however, to conclude from these trends that radical oVshoring—that is, more than the straightforward bleeding away of particularly labor-intensive production that has been happening for years—is in any way a necessary consequence of the sort of outsourcing that so characterized previous periods of restructuring. International trade in intermediate goods as a percentage of total imports has been growing across advanced manufacturing economies since at least the 1970s (Feenstra 1998), but the United States is almost unique in

Introduction to part I

13

the degree to which it runs an overall goods trade deWcit. Notably, this deWcit has grown even despite dramatic currency realignments in 2004,13 showing that the trade-induced portion of twenty-Wrst century American deindustrialization is not a matter of relative currency values alone (though these do certainly matter). Rather, some problems run deeper, with roots in the ways the American productive model evolved (or failed to) in the wake of the crises of the 1970s and 1980s. This book explores that evolution and its contemporary implications, beginning here in Part I with two chapters focused on two tasks. First, I show that there is a relative consensus in the academic literature as to the general contours of a new production paradigm that can sustain high-wage manufacturing in uncertain and globally competitive markets, but that there are also substantial disputes as to whether this paradigm has been or even can be realized in the American context. Second, I lay out the larger theoretical argument of the book and establish that there is reason to think that barriers to the spread of the new production paradigm are real but not insurmountable (the demonstration of this latter claim relies on empirical Wndings presented in Parts II and III).

This page intentionally left blank

1 A New Production Paradigm for a New Old Economy 1.1. C RI S I S A N D HY B R I D I ZATI O N I N P ROD U C T I V E M O DE L S In the 1980s, as it became clear that Fordist mass producers faced systemic rather than cyclical diYculties, a very excited literature emerged in the social sciences and even in the popular press to extol the virtues of more Xexible productive systems premised on cooperative relationships both within and between Wrms. The ‘Xexible specialization’ of the Italian industrial districts and German ‘diversiWed quality production’ were mined for concepts, in the hopes that some of their virtues could be reproduced elsewhere.14 But the big gorilla in the room was undoubtedly the ‘lean production’ of Japanese producers, portrayed most inXuentially, if imperfectly, in the popular press in The Machine that Changed the World. Written by Womack, Jones, and Roos (1991), this book made enormous waves by arguing that the Japanese model was both a uniWed and superior productive system and that it could be readily adopted outside Japan. Even as this literature has developed and matured, it continues to do more than merely document the reorganization and has retained a normative focus, seeking to identify a set of ‘best’ practices for Wrms buVeted by uncertain global markets. It—or at least the manufacturing practices it describes—is also highly inXuential: there is considerable empirical research to show that so-called ‘lean production’ and even ‘high-performance work organizations’ have signiWcantly diVused in the United States and that relationships between Wrms have changed enormously (DiMaggio 2001a; Helper and Sako 1998; Osterman 1999; Podolny and Page 1998; Smith 1997). Indeed, today, nobody credibly denies that there have been enormous changes in the organization of manufacturing in the United States (and in Europe) over the last twenty years, or that Wrms do not seek to borrow ideas—with varying success—from other productive systems.15 A prolonged Japanese recession notwithstanding, the so-called ‘Japanese production model’ retains its sway among manufacturers. When seen just as an ideal type abstracted from an exceedingly complex reality and never entirely realized across any economy (including the Japanese), it is quite clear that its constituent elements, albeit modiWed by local circumstance, have had a profound inXuence on onetime mass producers across the globe (Boyer et al. 1998; Durand et al. 1999; Kenney and Florida 1993; Liker et al. 1999; Zeitlin and Herrigel 2000). Such excitement over the emergence and diVusion of a new productive model following periods of crisis is nothing new. Highlighting the

16

The New Old Economy

attention given to Britain at the turn of the nineteenth century, to the United States ‘in the late nineteenth century through the 1960s and once again in the 1990s’ and to Japan in the 1970s and 1980s, Zeitlin (2000: 1) reminds that historical experience gives reason for caution when taking stock of the transitory success and failure of supposedly coherent national models of capitalism. Condemnations of the ‘rigidity’ of Japanese and German labor markets would have sounded odd not so many years ago when these relatively regulated systems were perceived to be outperforming their American counterpart (Osterman 1999). And, as I write, the implosion of the technology bubble, weak job growth and a post Enron awareness of crony capitalism, American-style, have left many quite skeptical of transferring the current American model in toto. The lesson here is not that the ups and downs of particular economies are random, or that every dog gets its day, but rather that pendulum-swings of faith in whichever system is performing measurably best at the moment follow from the often errant belief that productive models are fully coherent and importable wholes. Looking back at the attention lavished on the model-of-the-decade across historical periods, Zeitlin (2000: 1) writes that ‘foreign observers Xocked to the rising industrial power of the day to determine the secrets of its success, while business people and government oYcials sought through a variety of means to transplant the new methods into their own domestic soil. In each case, moreover, such experiments touched oV far-reaching debates about the essential features of the new production paradigm; its economic, social, cultural, institutional, and political preconditions; and its transferability across national borders.’ These ‘far-reaching debates’ have, however, generally failed to provoke the ‘wholesale imitation of foreign ‘‘best practice’’’. Rather, the typical historical experience is one of extensive but ‘piecemeal borrowing and selective adaptation to suit the divergent requirements of local economic and institutional circumstances’.16 In this chapter, I describe these ‘far-reaching debates’ and the sorts of ‘piecemeal borrowing’ and ‘selective adaptation’ of Xexible manufacturing practices they have provoked in American manufacturing. I argue that there is a general consensus on a set of practices that, if fully implemented, would amount to a sort of normatively desirable ‘new production paradigm’ (which is not to suggest that this paradigm has in fact fully diVused). I begin with the core element of the new model for purposes of this book—the network organization, premised on collaborative relationships between Wrms—and then treat questions of internal workplace reorganization. The chapter concludes by discussing the constraining/ enabling role of social and institutional context. The general approach I take Wts with what Carruthers and Uzzi (2000: 490, 487) describe as the dominant tradition in modern economic sociology. Unlike many in economics who strive to build macro-models on consistent microfoundations, I begin ‘somewhere in the middle ground (organizations and networks)’ and work my way outward, relying on ‘the idea of a two-way rather than a one-way relationship of information exchange between actors’. My

New Production Paradigm for New Old Economy

17

discussion of the potential transition to a new production paradigm is accordingly centered particularly on how and why the last quarter century’s decentralization and reorganization of production has made relationships between organizations both important and potentially problematic. However, I emphasize throughout that these relationships are not isolated from developments internal to Wrms and in the surrounding institutional, cultural, and market context. I thus quite deliberately refer to changes in the production—not supply chain—paradigm.17 1.2. T H E N ET WO RK ED F I RM AS TH E CRUX O F TH E N EW PRODUCTION PARADIGM Richard Florida drew in 1995 on a survey of Midwest manufacturers to describe a shift in Midwest manufacturing towards what he called ‘high-performance’ economic organization. By this, he meant that there was a trend away from a system ‘deWned by high levels of vertical integration and internationalization of capabilities’ in which ‘external suppliers tended to involve ancillary or nonessential elements [that] were generally purchased largely on price, and were stored in huge inventories in the plant’, towards one characterized ‘by a much higher degree of reliance on outside suppliers and the development of co-dependent complexes of end-users and suppliers’. An increased reliance on supplier Wrms is a result of the strategies that many OEMs have used to react to market volatility. Many of these once-verticallyintegrated manufacturers have deWned some limited set of ‘core competencies’ and then outsourced other operations (Prahalad and Hamel 1990), in a relative shift from capacity to specialized subcontracting. That is, whereas OEMs were once more likely to retain substantial internal capacity and to rely on subcontractors for demand peaks, OEMs today tend to rely more, at least in the short and medium term, on their suppliers’ specialized technology and/or labor skills. They are now more likely to use relationships with outside suppliers to mitigate the diYculties of volatile and fragmented markets that grant little cost leeway even as they demand increasingly diversiWed products with ever-shorter life cycles. Large Wrms that held substantial quantities of inventory and workin-progress rely instead on suppliers to deliver high-quality parts just-in-time (JIT), devolving day-to-day production and depending on substantial supplier assistance for process and design improvements to get products to market both more quickly and more cheaply. Needing to improve interWrm cooperation and information transfer, OEMs give more business to fewer suppliers, and forge closer relationships with a core strategic group that they hope to align with their own goals. Importantly, these key suppliers are not envisioned as mere satellites orbiting a dominant but benevolent patron, dependent and beholden. Rather, in a practice somewhat in tension with the desire to extract priority treatment when needed, OEMs push many of their suppliers to be more independent and to work closely with other customers and end-use industries. They share ideas,

18

The New Old Economy

technology and Wxed costs with these ‘partner’ suppliers in ways they recognize may beneWt competitors, but hope as well that these smaller Wrms will learn from those other customers, acquiring competencies that can transform the supply base into a vital source of new ideas and technology. This shift has formed an important part of the empirical basis for the claims of sociologists such as Powell (2001)—in DiMaggio’s edited volume The TwentyFirst-Century Firm—that the ‘networked Wrm’ has become the essence of a new production paradigm, as companies leverage partner Wrms and core employees to compete in the high-speed learning race he characterizes today’s economy to be. In these ‘twenty-Wrst-century Wrms’, hierarchies are Xattened, networks spread, and cross-fertilization between industries becomes essential. In his concluding chapter to the volume, DiMaggio (2001b: 222), emphasizes that despite the lack of good measures of organizational structure and strategy, ‘most knowledgeable observers believe that Wrms are engaging in more longterm collaborations—especially close information-sharing relationships with suppliers and collaborations around product development—now than during most of the twentieth century’. As companies externalize functions, he writes, ‘a far greater and more systematic functional burden is placed on relationships than was customary in the recent past’. This is pushed by structural incentives like ever shorter product cycles, regionally diVerentiated markets, and competition for sophisticated technologies that reward Wrms able to ‘adopt such techniques as Xexible production, JIT delivery, high-speed digital communication networks, and the use of subcontracting alliances to gain information’ (DiMaggio 2001b: 231). These happen together not just because they address common problems, but also because they are mutually supportive (i.e. leveling vertical control structures helps you get the rewards of information technology). Drawing on this ongoing evolution of the American productive model, Sabel (1996: 2–3) argued that ‘the organization of production in the US [had] become . . . substantially more collaborative or team-like: in a word, more ‘‘Japanese’’, and . . . [less] ‘‘American’’ ’, found most tellingly in the diVusion in the United States of Japanese customer–supplier relations, often taken to be the core of the Japanese production system as a whole. In these new supply relations a small number of top-tier contractors assume responsibility for codeveloping crucial modules or subsystems with the Wnal producer, coordinating the production of low-tier suppliers producing parts for subassemblies for their module, delivering the components just-in-time, and meeting targets for incremental improvement of production according to targets agreed with the customer.18

Importantly, such a sea change in the organization of the American productive economy does not require a full importation of the Japanese model, nor does it mean that every interWrm relationship must be collaborative.19 Rather, those who see the spread of a new production paradigm (or in Timothy Sturgeon’s terminology, the ‘production network paradigm’) in the changes that occurred in the 1980s and 1990s argue only that there was a strong relative shift from

New Production Paradigm for New Old Economy

19

arms-length to partnership relationships and that this shift reXects a diVerent logic of sourcing. Note also that none of this need imply that OEM–supplier collaboration is a wholly new phenomenon even in the United States. It again requires only that there has been a very substantial relative increase in the scale and scope of specialized subcontracting in American manufacturing since the 1980s.20 Indeed, authors in business history have shown parallels between contemporary and prewar patterns of collaboration between OEMs and suppliers.21 For example, discussing the American auto industry in the 1920s and 1930s, Schwartz (2000: 65) writes, ‘Relationships in old Detroit were characterized by cooperative product development, long-term contracts, and the ‘‘voice’’ system of resolving problems.’ Similarly, though referring to a slightly earlier period, Helper and Hochfelder (1997: 187) argue that ‘evidence indicates that many of the features of these so-called ‘‘Japanese-style’’ customer–supplier relationships were present in the American auto industry before 1920.’ It is thus worth a word on what is meant by a new paradigm, given both its aYnities to patterns of relationships between organizations in the earliest part of the century and the recognition that there certainly was some cooperative and specialized subcontracting throughout the Fordist period. My use of the word new does not mean that the practices themselves are either totalizing (since we are explicitly talking about a hybridized model) or had never seen the light of day. I mean simply that their spread is very substantial and it is recognized as ‘best practice’ in a way that it simply was not for most of the postwar period. Moreover, though some of the organizations that were looking to collaborate before World War II may have the same names as those looking to collaborate now, they are not in any real sense the same organizations. Much has changed since then, including all of the personnel, most of the organizational routines, and so on.

The Double-Edge of Productive Decentralization Visions of a collaborative interWrm new production paradigm turn in part on the claim that a shift of production from large to small Wrms need not necessarily be a straightforward power grab by large Wrms eager to push costs onto weaker partners, to avoid unions, and generally to decrease the wage bill. It is undoubtedly true that wage-cost driven outsourcing is a problem in the United States (especially in the automobile industry), due to its large union/nonunion wage diVerentials, plant-level bargaining, many southern ‘right-to-work’ states with lower union density, and dramatically reduced wage costs just across the Mexican border (though reduced transportation costs and improvements in communications mean that price competition is perhaps as likely to come from far-oV China as it is nearby Mexico). Nevertheless, though wage-costs are undoubtedly important in many cases, they are at best a partial explanation (Deavers 1997). In perhaps the most systematic quantitative study

20

The New Old Economy

of outsourcing decisions in manufacturing establishments, Harrison and Kelley (1993: 228), who can hardly be accused of turning a blind eye to the dark side of capitalism, found that ‘the search for numerical Xexibility through subcontracting cannot be understood as manifesting only (or even primarily) a strategy for reducing the company’s dependence on high-cost labour. And we Wnd no evidence that subcontracting is associated with union-avoidance per se.’ Notably, this study was undertaken in the period when much of the manufacturing outsourcing was occurring (see also Harrison 1994). Nevertheless, just as research documenting the diVusion of new forms of work organization inside manufacturing Wrms has highlighted the uneven diVusion of new work practices, analyses of decentralized production also unequivocally show that coordinating production across multiple Wrms neither implies nor requires the presence of cooperative relationships. It may reXect the opposite, especially insofar as the externalization of production may well make it easier to put particular components up for bid to manufacturers in the developing world. In Powell’s (1990: 319) oft-cited article distinguishing the collaborative network from markets and hierarchies, he writes that the shrinking of large corporate hierarchies forced a choice between the rediscovery of the market with its ‘hostile world of arms-length relationships’ and the reorganization of production ‘not so much through eliminating jobs, but by searching for new methods of collaboration among formerly antagonistic and/or competitive parties. . . . Both responses entail some form of vertical disaggregation.’ The upshot: manufacturers are ‘simultaneously deciding which suppliers are worth investing in a longterm relationship with and determining which components can be obtained on the basis of price rather than quality’ where collaboration has less payoV. Powell (1990: 322) writes, ‘These disparate options graphically illustrate how practices such as subcontracting have a double edge to them: they may represent a move toward relational contracting, with greater emphasis on security and quality; or they could be a return to earlier times, a part of a campaign to slash labor costs, reduce employment levels, and limit the power of unions even further.’ Harrison (1994: 18) picked up and amply developed this latter theme in his book, Lean and Mean, where he focused on the darker aspects of a decentralized production model that does ‘not imply the end of unequal economic power among Wrms’. Contrasting the optimism of some that locally embedded networks of Wrms could form the basis of a relatively egalitarian productive model in the United States, he argued that the growth in small Wrm employment was in many cases due to ‘managers . . . outsourcing work they used to perform inhouse’; thus, the ‘increasing number of small Wrms turns out to be in part a function of the core-ring, lean production strategies of the big companies’. Such Wrms, Harrison (1994: 13, 18) noted, very often oVer poorer work conditions and, despite their de jure independence, are ‘de facto dependent on the decisions made by managers in the big Wrms on which the smaller ones rely for markets, for Wnancial aid, and for access to political circles’, since ‘power, Wnance,

New Production Paradigm for New Old Economy

21

distribution, and control remain concentrated among the big Wrms’. The implication: a ‘concentration without centralization’ that is hardly indicative of a normatively attractive new production paradigm, but that reXects instead a desire to use market power and global reach to drive worker wages ever lower. 1.3. T H E N EW PRO DUCTI ON PAR ADI G M I N S I DE T H E F I R M : FLEXIBILITY IN WO RK ORGANIZATION Nishiguchi (1994) is particularly clear in arguing that there is a fundamental intertwining of interWrm relations and intraWrm organization, forcefully rejecting the many analyses of the Japanese model that have argued it to be a simple dualism, Xexible only because graced by legions of lower paid subcontractors. He attacks ‘the traditional perception that Japanese industry is sustained by unequal subcontracting relations’ and challenges dualist approaches for failing ‘to distinguish whether or not Xexible manufacturing beneWts are actually available. The important message here is that ultimately a customer can have a Xexible production system only if its subcontractors are also Xexible in their internal operations. In this respect, the symbiotic element in Japan’s contractual relations is not incidental, but, rather, is an essential part of the new system.’ Notably, Nishiguchi’s argument does not presume that workers in small Wrms are not paid less; they may also be more exploited than are workers in larger Wrms. His point is simply that the externalization of a process will limit productivity growth, Xexibility, and innovation if suppliers are unable to keep up—and lower wages may not be enough to keep them up. OEMs cannot depend upon Xexible and reactive suppliers to incrementally improve designs if those suppliers are not in fact somehow more Xexible and reactive and able to incrementally improve designs. At the shopXoor level, the new production paradigm is embodied in ‘lean’ or ‘high performance’ work systems in which Wrms are able to improve both quality and Xexibility by pushing workers to acquire new, often Wrm speciWc, skills and empowering them to participate in decisions that alter organizational routines (Appelbaum et al. 2000; MacDuYe 1995).22 Smith (1997: 318) reviews the literature to Wnd that the ‘speciWc organizational innovations most likely to enable workers to carry out these goals included quality circles, employee involvement programs, job enlargement and rotation, self-managing teams, continuous improvement processes, organizational decentralization, and the promulgation of a new ethos of participation.’ Such functional Xexibility aims to draw on multiskilled (‘polyvalent’) workers carefully nurtured by employers, often in internal labor markets, to improve the Wrm’s ability to adapt to uncertainty to and to rapidly adjust to quantitative and qualitative demand shifts without sacriWcing productivity (Vidal 2000). However, commentators on labor Xexibility in the new production paradigm emphasize that the functional Xexibility of the core workforce is too often coupled with numerical Xexibility and a casualization of the labor relationship,

22

The New Old Economy

with many workers pushed into a contingent dual labor force (Smith 1997; Vidal 2000). Smith (1997: 315–16) draws the contrast between the idealized visions of work reorganization and critiques of its implementation in practice. She writes that there certainly are descriptions of ‘newly skilled, continually learning, empowered and engaged workers, [who], aided by entrepreneurial managers, strive to relax and Xatten rigid bureaucracies, trim excessive use of organizational resources (including time, space, and people), and use their experiential knowledge to improve the way they produce goods or serve people.’ But she notes as well that there is a parallel literature that sees Xexible work arrangements as ‘little more than a new permutation of work that disadvantages workers but oVers employers signiWcant dividends. Flexibility, in this view, may call for new production techniques or novel work group formations but doesn’t entail a substantial break with traditional hierarchical modes of control and authority relations; rather, it embodies and even deepens them by obscuring power behind participatory language.’ Smith (1997: 332) concludes that there are ‘limitations to viewing the Xexible model in polarized terms, as being either progressive, enabling, or a high performance approach, or coercive, restrictive, or a lowperformance approach’. The new model in fact combines elements of both characterizations, and plays ‘diVerent groups of workers oV one another, eliciting participation from some by denying participative opportunities to others’, as ‘core workers beneWt from the new approach because they are buVered by a peripheral work force brought in to absorb Xuctuations in work cycles but speciWcally excluded from participating in these organizational innovations’. Recent comprehensive national studies of the diVusion, eVectiveness, and impact on workers of new work systems reXect these contradictions. Albeit spreading in a somewhat piecemeal fashion, elements of high performance work systems have become common: Osterman’s (1999) use of the 1992 and 1997 National Establishment Surveys show a substantial increase in the use of quality circles, job rotation, and total quality management, while ‘self-managed work teams’ remained steady at about 40 percent of establishments; furthermore most Wrms engaged in at least two of the aforementioned practices. Although measuring the eVectiveness of high performance work systems is notoriously hard to do, given the impossibility of adequately matching Wrms for direct comparison, the evidence suggests on balance that at least when fully implemented (including compensation and training) they have improved productivity (Osterman 1999).23 In perhaps the most sophisticated study to date, Applebaum et al. (2000) gathered substantial primary data in the steel, apparel and electronic medical instruments industries and were able to show that Wrms in each of these industries were able to improve performance through the implementation of High Performance Work Organizations. Their industry level survey shows that workers involved in the new systems tended to have higher earnings and job satisfaction. Osterman’s wider national surveys (1999: 114), however, suggest that while the diVusion of new Xexible work systems are generally a good thing—many workers prefer the higher levels of responsibility, and they can lead to training in

New Production Paradigm for New Old Economy

23

new skills—they also have their dark side: the ‘establishment of innovative work systems (in 1992) is associated with higher chances of layoVs in 1997 and bears no relationship to compensation gains. The only exceptions are the small number of establishments that are unionized.’ This, he argues, reXects a substantial shift in the power balance between workers and managers that seems to have allowed companies to implement new work practices without the implicit exchange of long-term employment for worker commitment once argued to necessarily underlie the Japanese model.

1.4. LOC AT I NG THE N EW PRO DUC TI ON PA R ADI GM: GEOGRAPHY, LEARNING AND LOCK-IN The Wnal important theme running through social scientiWc discussions of the changed organization of manufacturing is that both productive and exchange behavior are embedded in particular social and institutional contexts. These, in turn, are located geographically, often at the (subnational) regional level. This discovery, Storper (1995: 191) explains, came in the early 1980s when ‘something funny happened’: theorists that had long treated the region as a mere ‘outcome of deeper political-economic processes, not a fundamental unit of social life in contemporary capitalism’ were blindsided by new successful forms of production—diVerent from the canonical mass production systems of the postwar period—[that] were emerging in some regions and not others. . . . [S]ince [these] seemed to involve both localization and regional diVerences and speciWcities (institutional, technological), it seemed that there was something fundamental which linked late twentieth century capitalism, regionalism, and regionalization.

Walter Powell—again in the oft-cited 1990 piece—noted the importance of space in sustaining network organizational forms, writing that ‘notions of industrial districts and spatially concentrated production were [long] largely ignored, but their success [and rediscovery] has three key ramiWcations’: (a) the boundaries of the Wrm have been blurred, expanded to ‘encompass a larger community of actors and interests that would previously have either been fully separate entities or absorbed through merger’; (b) the importance of forces external to the Wrm—including local government cooperation, skilled labor pools, and ties to research institutes and associations—became apparent; (c) ‘The spread of technologically advanced, smaller units of enterprise—a growth that comes at the expense of larger companies and is not explained solely by the shift from manufacturing to services’—is now understood to depend in many cases on cooperative interorganizational relationships (Powell 1990: 313). The upshot of this wider recognition, Wolfe (2003: 4–5) explains, is that ‘the production paradigm of this ‘‘new economy’’ is highly dependent on localized, or regionally based, sources of knowledge and learning.’ A ‘paradoxical consequence of globalization’ has thus been the accentuation of the local context as

24

The New Old Economy

the concept of the learning region emerged to describe those places that oVer the right institutional environment to encourage both private and social learning at four diVerent scales: the individual worker, the individual Wrm, within groups of related Wrms, and within governmental bodies themselves. The literature on learning regions argues that even global Wrms need to draw their innovative sustenance from the social production systems surrounding them. (Wolfe and Gertler 1999: 8–9)

There remains, however, some ambiguity as to the details of just what it means in practice to be a learning region. For example, Florida (2000) increasingly speaks of networks of research institutions and areas’ relative abilities to attract ‘talent’; by contrast, especially (though not exclusively) European analysts tend to focus more on trust, social context, and how Wrms manage cooperation with each other, their workers, and governments and associations. But such disagreements aside—they are largely questions of emphasis—there is general accord that Wrms’ ability to implement high-performance work organizations internally and to beneWt from cooperative relationships externally is geographically located. That is, it is fundamentally aVected by the institutional and cultural resources available, by what Storper (1995) calls ‘untraded interdependencies’ (Cooke and Morgan 1998; Putnam 1993; Scott 1998; Trigilia 1999; Wolfe 2003)

The Double-Edge of the Learning Region Just as with the new production paradigm, there is also a dark side to the notion of the learning region. It is one thing to use, for example, the central Italian industrial districts to show that a virtuous interaction of Wrm strategy with regional culture and institutions can generate international competitiveness and good worker outcomes (Burroni and Trigilia 2001; Pyke et al. 1990). It is quite another to transfer the model. Still, given the wide-ranging political implications of the changing character of manufacturing competition and industrial organization, it should not surprise that many of the theorists writing on the topic—going back to The Second Industrial Divide (Piore and Sabel 1984), most particularly to The Machine That Changed the World (Womack et al. 1991), but including commentators too numerous to mention—do not limit themselves to documenting a partial shift. Rather, they argue further that Wrms and groups of Wrms must somehow manage to approximate the new production paradigm if they are to survive and thrive in a fragmented and changing market. Even Harrison (1994: 34), though pessimistic as to the likely long-term eVects of decentralization, agrees that ‘obviously, we do not want to destroy the remarkable innovative capacity of production networks’, questioning instead whether existing forms of regulation are adequate to new challenges created by ‘businesses whose organizational boundaries (and as a consequence, legal liabilities) are becoming increasingly fuzzy’. To be clear, the disputes lie not in whether we can Wnd isolated exemplars of Florida’s ‘high-performance economic organization’, but in whether or not such a collaborative model can be generalized across regional and/or national contexts.

New Production Paradigm for New Old Economy

25

It is obvious that open, collaborative, positive-sum relationships between Wrms are good (deWnitionally!), but there is doubt that they can stably occur in contexts—such as the American upper Midwest—lacking the necessary ex ante social and institutional conditions. The issue, as Florida (1996: 315, 317) explains, is that much ‘contemporary theorizing emphasizes the institutional rigidities and so-called lock-in eVects that constrain and limit the process of regional change’. These eVects are thought to be particularly vexing in historic manufacturing regions: the literature on deindustrialization claims that a ‘tendency for older regions to become locked into older forms of production organization is a key factor in the economic decay of those regions’.24 These concepts—lock-in and path dependency—are most famously applied to technology (see Arthur 1994), but Wolfe (2003: 23) notes, they have ‘been applied with increasing frequency to studies of the social and political environment that inXuences and facilitates innovation’. The key issue is ‘the ability, or inability, of the local or regional economy to develop the underlying conditions of trust and social capital that contribute to the presence of a learning economy’ and that can thus help regionally embedded manufacturing Wrms to sidestep low-wage competitors. Given strong evidence that the externalization of production in globalizing and rapidly changing markets privileges those Wrms and regions that are more able to sustain collaborative network relationships, the question raised for crisisprone American manufacturing is an obvious—and perhaps ominous—one: are institutional lock-in and path dependencies so strong that the long-term prospects for prosperity are inherently bleak in historic manufacturing regions long dominated by large hierarchies and relatively pure market relationships? Answering this question—insofar as it can be answered without a crystal ball—requires a mix of theoretical and empirical analysis. Hence, where this chapter has focused on what the new model is supposed to look like, the next discusses theoretical approaches that give some insight into how, if, and when collaborative production can be expected to stably occur.

2 Networks, Noise, and Institutional Change 2.1. I N TRO DUCT I ON There are continued disputes over exactly what changed in the organization of manufacturing in the last years of the twentieth century, but everyone agrees on a few things. In particular, it is indisputable that large vertically integrated manufacturers decentralized substantial production and design to smaller suppliers, which meant that theory had also to come to terms with a production model requiring signiWcantly more explicit coordination between Wrms. In particular, Coase and Williamson’s famous distinction between economic activity coordinated through exchange in markets and that coordinated by Wat in hierarchies came under sustained attack, accused of masking the spread of production networks that are ‘neither markets nor hierarchies’ (Powell 1990). The argument, made particularly but not exclusively by sociologists, is that standard theories rooted in transaction cost economics are trapped by a very particular view of opportunism and incentive conXicts, and that this obscures ways in which social and institutional context aVect Wrms’ strategic behavior. In an extensive review of the sociological literature on network forms of organization, Podolny and Page (1998: 71, 59) show that networks—the crux of the collaborative new production paradigm—are very common in modern capitalism. However, they caution as well that this same literature has an important gap: It has eVectively shown the existence and functionality of network production in highly uncertain environments, but this has been coupled with a relative neglect of the ‘constraints that underlie the formation of network forms of organization, problems that arise in their governance, and boundary conditions on their functionality. We may have a good understanding of why economic actors want to utilize network forms of organization, but we have less understanding of why they do not.’ Yet this is precisely the bit of theory that is needed to understand the very partial spread of collaborative manufacturing practices in the American context. In this chapter, I thus draw on three prominent theoretical approaches to explaining collaborative network production—social networks and embedded ties, comparative institutionalist ‘Varieties of Capitalism’, and neopragmatist ‘learning by monitoring’—to show that the tools to resolve this problem do exist in the sociological literature; they are just not all in one place. SpeciWcally, each of these approaches does provide some explanation for why production

Networks, Noise, and Institutional Change

27

might not be organized through collaborative networks, even when such a governance structure would be best for all involved; but each is also guilty of a theory-driven dismissal of intermediate cases of partial and contradictory collaboration. As such, useful though they are, they have too little to say about the need for, and possibilities of, constructing new institutions that could push American manufacturing further down the path towards a more attractive production paradigm. Fixing this gap, I argue, requires simultaneously drawing on some, and rejecting other, elements in each of these three literatures. In the remainder of this section, I describe them as brieXy as possible, giving only the details most essential to my argument. The remainder of the chapter outlines this larger argument, returning then to re-examine each approach more fully in light of my key claims. Though there is a (mild) redundancy to this structure, I do believe it to be the clearest way to present the material.

The ‘Social’ Explanation: Social Networks and Embedded Ties The best known sociological explanation of interWrm collaboration is rooted in a social networks literature built around Granovetter’s (1985) rejection of the ‘undersocialized’ Williamsonian actor. Reacting particularly to the new institutional economics, theorists of social embeddedness argue that trust and other social considerations that mitigate opportunism allow for the formation of network forms of economic organization: as Powell (1990: 303) writes, ‘the parties to a network agree to forego the right to pursue their own interests at the expense of others’. This approach does not predict that all of manufacturing will be organized into collaborative alliance capitalism, but it does explain how some of it could be. For example, Uzzi’s descriptions (1996; 1997) of production networks in the New York garment district rely on the claim that pockets of social relationships sustain networks of collaborating Wrms. Building theory around the concept of the socially embedded tie requires welldeWned in- and out-groups. Relations to the out-group look like the classic prisoners’ dilemma—any eVort to collaborate is truncated by low trust and defection—whereas the in-group is characterized by the parties’ reinforcement of a collective trust as they jointly worry about tomorrow’s business. Applied to a manufacturing context, this suggests that theorists should focus on identifying the presence (or absence of) social sustaining dynamics able to push particular OEM–supplier relationships towards one pole or the other.

The ‘Hard Institutional’ Explanation: Comparative Institutionalism and the ‘Varieties of Capitalism’ A structurally similar argument is found in the institutionalist approach in political economy known as the ‘Varieties of Capitalism’. This literature, associated especially with Hall and Soskice (2001; Soskice 1999), replaces the emphasis on a value-orientation with an argument that the activities of Wrms are deeply

28

The New Old Economy

inXuenced by national systems of hard-to-build but self-reinforcing coordinating institutions—business and labor associations, training systems, and the like—that have interlocking complementarities. These institutions coordinate strategic interactions, adjust incentives and discipline malfeasance so that in ‘coordinated market economies’ like Japan and Germany, Wrms can safely take a long-term view, and are thus much more likely to collaborate. This position questions whether or not a production model premised on substantial interWrm cooperation is even reasonably viable in the United States, especially in old-line manufacturing industries where competitive survival often depends on incremental innovation. The widespread diVusion of collaborative core-ring networks is seen as extremely unlikely given initial weaknesses in an American small Wrm manufacturing sector that has long been asked to do little more than ‘build-to-blueprint’, especially when coupled with historicalinstitutional factors that make these weaknesses hard to correct. American manufacturing suVers from low levels of interWrm trust, weak business associations and a system of corporate governance that is not supportive of the long-term commitment required to sustain interWrm collaboration.

The ‘Organizational’ Explanation: Neopragmatist ‘Learning by Monitoring’ Finally, an ambitious neopragmatist approach developed by Sabel and collaborators (Helper et al. 2000; Sabel 2004a) argues that contemporary economic and organizational theory has simply failed to grasp the full implications of decentralized production in a world of radical uncertainty. No longer in possession of knowledge required to fully design their own products and only dimly aware of what products can possibly be designed and built, some Wrms have been creating new ‘federated’ structures organized around a fundamentally diVerent set of principles from those that underlay most manufacturing Wrms in the postwar boom. Where the classic hierarchy relied on the decomposition of tasks into well-deWned but Wxed routines—a concession to uncertainty and the need to mitigate opportunism—the new organizations are built around practices that do not so much Wght uncertainty as use it to generate novel alignments of interest that render collaboration more feasible and more necessary. Sabel (2004a: 2) argues that many American manufacturers are mastering production techniques pioneered in Japan that allow federated pieces of the new organizations to ‘deWne methods for choosing provisional, initial designs and revising them in light of more detailed, partial proposals’. These techniques, which include benchmarking, simultaneous engineering, and various disciplined methods of error-detection, reward collaborators ‘for achieving broad goals according to standards deWned as part of the [joint] process by which the goals are set’. Driven by the recognition of all involved that they cannot know their own needs without the input of others, these Wrms have routinized information sharing in ways that were unthinkable in the era of vertical integration. This in

Networks, Noise, and Institutional Change

29

turn allows parties to discover novel solutions to problems in ways that were simply unknowable before interaction. The end result, Sabel (2004a: 2) argues, is that we are engulfed in a sort of ‘revolution’ in the underlying organizational paradigm, in which the ‘permanent uprising against routine’ becomes the ‘key to survival in otherwise unmanageably turbulent world’. Firms that experiment with the new pragmatist techniques are likely to join this revolution, making collaboration between OEMs and suppliers into a virtual prerequisite of competitiveness. As Helper, MacDuYe, and Sabel (2000: 445) write, ‘once the cooperative exploration of ambiguity begins, the returns to the partners from further joint discoveries are so great that it pays to keep cooperating’. 2.2. T HE N EW DETER MI N I S M VS . TH E C O N T R A D I C TO RY MIDDLE GROUND Across these three literatures—varied optimism regarding the likely relative prevalence of interWrm collaboration in diVerent contexts notwithstanding— there are three fundamental points of agreement: (a) the emergence of some form of network production consistent with the new production paradigm outlined in the previous chapter would be a good thing for Wrms and their workers in many industries; (b) there are mechanisms—be they patterns of social ties, formal and informal institutions, or learning routines—that allow Wrms to overcome diYculties of incentive alignment and knowledge sharing cited in the academic literature as barriers to successful productive cooperation; (c) as OEMs decentralize production to deal with volatile markets, the self-reinforcing character of the underlying coordination mechanisms means that we should see dichotomous and stable relational equilibria, with a world of hostile arm’slength contracting on one side, and deep collaboration involving multiple Wrms with shifting and ill-deWned boundaries on the other. These Wrst two points of agreement are valid and important contributions to the understanding of the social and institutional bases of economic coordination; the third obscures important aspects of what is in fact a very contradictory pattern of relationships in American manufacturing. The problem, I argue, is that none of these three positions leaves space for a stable middle ground, and each thus falls prey to variants of what Crouch and Farrell (2002; 2004) refer to as the ‘new determinism’.25 This, Crouch and Farrell (2004: 8) explain, is the ‘recent tendency in the social sciences, and in political economy in particular, to emphasize how institutional systems tend to crystallize around coherent logics of ordering’ creating path dependencies and ‘lock-in’. They write that path dependence, in its original formulation, implies the existence of alternative paths of development than the one taken, which may be ‘rediscovered’ when actors face a changed environment which makes new demands. Embeddedness theory, for its part, was formulated precisely in order to provide an alternative to both ‘oversocialized’ and

30

The New Old Economy

‘undersocialized’ conceptions of human action, and, properly speaking, examines how social ties may enable social action as well as constraining it. However, empirical work has tended to stress how increasing returns to scale and embeddedness lead to lock-in, rather than out of it. (Crouch and Farrell 2002: 5–6)

This leaves ‘authors . . . virtually bound to consider all modes of action which do not Wt their overall characterization of a given national or super-national system as untheorized empirical ‘‘noise’’ which needs to be disregarded in the interests of an elegant and sharply proWled account.’ This amounts, they argue, to a dismissal of the ‘incongruities, incoherence, and within-system diversities’ that form the basis not for ‘a series of ad hoc empirical objections—but a theory of crisis resolution and Schumpeterian change that does not require either exogeneity or prediction of inevitable failure’, and thus to an underestimation of the ‘possibilities of change and innovation’ (Crouch and Farrell 2002; 2004: 33). Sabel’s neopragmatist position, by contrast, aims precisely to capture change and certainly cannot generally be accused of simple path dependence—indeed, Sabel (1995: 2) is a well-known opponent of the belief that the ‘same mechanisms (historical contingency) that disable the invisible hand turn economic choice into a nightmare of unintended consequences, at best unforeseeable, and at worst incorrigible’. Ironically, however, Sabel’s recent analyses (2004a: 1) seem themselves almost to suggest a sort of inevitability in development, though hardly a nightmarish one, with the claim that the ‘permanent uprising against routine’ has become the ‘key to survival in otherwise unmanageably turbulent world’. By claiming that initial successes at those Wrms that do experiment with functionally superior Japanese-style customer–supplier relations will set all involved on the path towards ‘cooperative exploration of ambiguity’ (Helper et al. 2000: 445), Sabel again leaves the in-between case untheorized and dismissable as a temporary and transitional place, so long as markets are competitive enough to force a march toward this paradigmatically new industrial organization. Like Crouch and Farrell (2002; 2004), I reject any such ‘new determinism’, but not on a priori grounds. I do not suggest that there is no logical basis for even quite strong path dependence and/or lock-in dynamics; indeed, there is good evidence that such phenomena do occur. Rather, the objection is to those who overplay theoretically their empirical hands, underestimating possibilities or overestimating inevitabilities of systemic change by leaving incongruous elements or latent patterns out of the analysis in the service of an ‘elegant and sharply proWled account’. Such overplays wrongly posits false determinisms that obscure both the traps of history and keys actors might use to unlock them. The objection is certainly not to abstraction per se, but rather part and parcel of the view that the right level of abstraction depends on the problem at hand. My goal, in short, is to advance work begun by Crouch and Farrell; but where their article is largely analytic and exploratory—showing that under the assumption of Bayesian learning, even slight modiWcations in Polya-urn games26 open

Networks, Noise, and Institutional Change

31

space for strategizing actors to escape traps of path dependency—my arguments begin fundamentally with an empirical rejection not only of the overly coherent and polarized models found in the hard institutionalist and the social networks literature, but also of the hint of inevitability in the transition story spun by Sabel and collaborators. To do this, in Part II, I draw on interviews with manufacturing Wrms in the American Upper Midwest to show that the modal case of the OEM– supplier relationship is both contradictory and systematically intermediate between the arm’s-length and collaborative poles. That is, the empirical noise— incongruities, incoherence, and within-system diversities—creates theoretical noise, and that noise matters. Firms are making real eVorts to collaborate, suggesting that systemic change may be possible; but those eVorts are also rife with failures that have real costs even as they often do not wreck the parties’ ability to try again. Collaborative production is by no means self-actualizing despite functional advantages; rather, it is undermined by systematic—though often potentially correctable—blockages. Importantly, this is a constructive rejection, and not just a clariWcation of details. I do not believe the approaches engaged to be completely wrong-headed. To the contrary, I explicitly build on them, arguing only that each leaves the intermediate case out of its analysis. Shorn of this errant dismissal of the modal in-between case as noise, elements of each provide the tools to disentangle why eVorts to construct a more collaborative new production paradigm have been so uneven and why there is reason to think it possible to build new institutions to support the creation of such a paradigm. The reconstruction depends on three key arguments, which depend themselves on the empirical material presented in Parts II and III. I give them here in the form of assertions to structure the book’s larger argument and its relationship to the extant literature on economic coordination. 1. I accept the core of the neopragmatist position developed by Sabel and collaborators, which is that there are seeds of potential change in what others dismiss as noise. In the face of unstable markets and rapid technological change, many manufacturing Wrms are hybridizing aspects of Japanese production systems and experimenting with federated organizational structures premised on decentralizing production and building collaborative relationships with suppliers. That some Wrms succeed in this undertaking ‘even in a short-term oriented, individual interest-maximizing society like the USA’, simply does not square with the claims of those who would argue that full-blown ‘Japanese-style collective institutions’ and/or a general culture of trust are required to ‘generate the conditions necessary to maintain and nourish collaboration’ (Helper et al. 2000: 475). 2. Nevertheless, this must itself be squared with the empirical Wnding that the diVusion of collaborative relationships between Wrms is partial at best. Not only are many OEMs deeply cautious about genuinely relying on supplier Wrms, but even those that do make the decision to reform their structures and build

32

The New Old Economy

collaborative ‘network’ relationships with suppliers often Wnd those eVorts blocked. These blockages, documented in Part II, underscore the diYculty of reforming institutionalized behaviors through individual action and suggest that the relative ability of Wrms to sustain collaborative relationships does depend considerably on the historical context and the degree of institutional support for market and/or cooperative interactions between Wrms. 3. This means that issues of organizational change are also issues of institutional change, putting questions of institutional lock-in and path dependency at center stage. Here I follow Granovetter’s point (1992: 7) that ‘economic institutions do not emerge automatically in response to economic needs [but] are constructed by individuals whose action is both facilitated and constrained by the structure and resources available in social networks in which they are embedded’. Disentangling the factors that place relationships between Wrms systematically between the arm’s-length and collaborative poles allows for the identiWcation of ‘resources available’ in the networks in which Wrm and state actors are embedded. Some of these resources, I show in Part III, amount to keys that can unlock traps of path dependency. Regions that have existing and stable collaborative eVorts upon which to build, even partial or obscured, need not resign themselves to the polarizing stories’ inevitable snap-back of self-equilibrating systems, in which state actors unblessed with the ‘right’ historical conditions have little leeway to change their lot. 2.3. T H E ECO N OM I CS OF O RGA NI Z ATI O N: FRO M BLACK B OX TO CONTRACT AND COMPETENCE THEORIES All economic theory turns essentially, if implicitly, on the problem of coordinating actors. (Michael Storper and Robert Salais 1997: 27)

In the most vulgar and unfair understanding of economics, the only truly eYcient means by which actors can coordinate activities is in ‘the market’. But of course, the Weld is not really so vulgar. The economics of organization is devoted precisely to explaining when transactions are best coordinated by Wat within hierarchies and when spontaneously in the market—that is, to explaining the boundaries of the Wrm. Economic sociologists are also not so vulgar as to have missed this, and it is this more sophisticated understanding to which the approaches discussed in this chapter react (more so than to each other). Hence, in this section, I provide as background a discussion of contractual theories of the Wrm and critiques thereof in both mainstream and evolutionary economics. With this I hope to establish two things. First, the economics of organization cannot, by itself, suYciently account for contemporary patterns of networked production. Second, it has, however, shown that an adequate account must explain how groups of Wrms simultaneously manage two tasks: coordinating incentives; and acquiring tacit productive knowledge.

Networks, Noise, and Institutional Change

33

The Contractual Perspective: Peeking Inside the Black Box The orthodox neoclassical view of the Wrm is neatly summarized by Winter (1993: 180): Firms are characterized by the technological transformations of which they are capable—formally, by production sets or production functions. Like consumers, Wrms are unitary actors and are economically rational; more speciWcally, they maximize proWt or present value. They deal in markets for homogeneous commodities; in almost all cases these appear to be contemporaneous spot markets for inputs and outputs. Contractual arrangements and other institutional supports for the functioning of the business Wrm are, one infers, assumed to be suYciently close to being Xawless and costless so as to justify the virtual total absence of discussion of these topics.

As Coase (1937) famously noted, however, this very simple theory of the Wrm as a proWt-maximizing black box production function cannot even explain why Wrms exist at all. This observation went relatively ignored in economics for a long time, leaving the study of organizations almost exclusively to sociologists. But in the 1970s, Coase’s claim that there are (transactions) costs to using the market got much new attention. Most notably, Alchian and Demsetz (1972) and Williamson (1985) seeded what have since become two variants of the dominant contractual literature in the economics of organization27: the principal–agent ‘nexus of contracts’ view of the Wrm and Transaction Cost Economics (TCE). Both are concerned with explaining where transactions costs, and thus Wrms (since transactions costs explain why markets may be less eYcient than hierarchies), come from in a world of rational maximizers. And both are premised fundamentally on the belief that eYciencies of specialization are regularly undermined by incentive conXicts caused by information asymmetries and opportunistic hold-up. They do, however, diVer somewhat in their predictions and assumptions. Principal–agent models are closest to black box (production function) neoclassical theories of the Wrm, adding only that Wrms represent team production, that actors are opportunistic, and that information is asymmetrically distributed. The key concept is the ‘moral hazard’, and Wrm structures are assumed to minimize monitoring costs by aligning incentives. TCE, the dominant literature in the economics of organization, is itself dominated by the work of Williamson, whose actors are not just opportunistic but are boundedly rational as well. He argues that the Wrm per se should not be at the center of analysis; rather, economists should concern themselves with explaining the most eYcient governance of transactions as between markets— which oVer the gains of spontaneous coordination through the price system— and hierarchies—which add costs to production but reduce the scope for opportunistic hold-up. In TCE, the most important properties of transactions are their frequency, uncertainty, and relative asset speciWcity (the relative ability of parties to transfer their assets to uses other than for the speciWc transaction in question).28 If one of the parties makes up-front investments in

34

The New Old Economy

specialized assets, he risks being ‘held up’ by a partner refusing to commit the complementary assets. The organizational imperative then becomes to ‘align transactions (which diVer in their attributes) with governance structures (the costs and competencies of which diVer) in a discriminating (mainly, transaction cost economizing) way’ (Williamson 1993: 93). Each of these theories is undergirded by two core principles. Following a line of reasoning that dates at least to Adam Smith, an increase in scale permits greater specialization of the division of labor with ensuing eYciencies. Specialization makes contracting parties vulnerable, however, because it creates more complementarities in the subcomponents of products, reducing the ability of their owners to shift them to diVerent uses. The result, Helper et al. (2000: 461) note, is that ‘owners of highly specialized complementary resources cooperate, therefore, at great risk’, and thus retreat to vertical integration. These standard theories predict that complementary resources will generally be held by a single owner, the preferred solution to hold-up and other problems caused by the impossibility of writing complete contracts. Organizations are assumed to take the form that best aligns the incentives of all actors in the game. Long-term cooperative relations between Wrms, insofar as they are observed, are generally explained as the eVects of the shadow of the future in an inWnitely repeated prisoner’s dilemma (to which long term principal–agent situations can be likened), which standard game theory tools show to have cooperative solutions if both parties suYciently value future income.29 This does allow for some interWrm cooperation even when the ex ante incentives do not seem to align (i.e. lots of hold-up potential), but such cooperation is very fragile. The appearance of trust is only the self-interested decision by parties to give each other the beneWt of the doubt on the Wrst interaction, after which both parties continue to cooperate since neither party ever has an incentive to renounce the deal, though both would given changes in future discounting, payoVs, or a defection by the other party.30

The Inadequacy of a Pure Incentive Focus: Orthodox and Heterodox Critique Winter (1993: 93) has cracked: ‘In the past half-century, it has been clearly demonstrated that the economy is much better at changing itself than economists are at changing their minds.’ But diYculty changing minds or no, transactions cost theories have come under Wre even among economists, both heterodox and mainstream, for an evident inability to explain the tendency to productive decentralization described in Chapter 1. Already in 1988, Milgrom and Roberts reported on ‘Economic Theories of the Firm: Past, Present, and Future’ and argued that ‘as our analysis of the Wrm deepens the Wrm–market distinction blurs; for production itself involves exchange’. Gibbons (1999; 2001) has built on insights from organizational sociology to argue that standard economic theories of the Wrm overplay the degree to which vertical integration resolves

Networks, Noise, and Institutional Change

35

incentive conXicts, and that economists should incorporate relational contracts both within and between Wrms into their models. And Holmstro¨m and Roberts (1998: 91), in their excellent review article, ‘The Boundaries of the Firm Revisited’, draw on the experience of Japanese subcontracting and other evidence of longterm relational contracting to argue that ‘the theory of the Wrm, and especially work on what determines the boundaries of the Wrm, has become too narrowly focused on the hold-up problem and the role of asset speciWcity’. They write: We think that of the signiWcant organizational change that seems to be taking place, only a small part can be easily understood in terms of traditional transactions cost theory in which hold-up problems are resolved by integration. Many of the hybrid organizations that are emerging are characterized by high degrees of uncertainty, frequency, and asset speciWcity, yet they do not lead to integration. In fact, high degrees of frequency and mutual dependency seem to support, rather than hinder, ongoing cooperation across Wrm boundaries.

Since this is essentially opposite what Williamson (1991) predicts, it perhaps should not surprise that Holmstro¨m and Roberts advocate a broader view in which Wrms serve also to coordinate and motivate the activities of individuals. This conclusion, that economic incentives—narrowly construed—cannot explain important variation in the organization of production has long been one of the central claims of the heterodox evolutionary and competence theories of the Wrm stressing that production involves an immense coordination of specialized knowledge.31 Rejecting the hyper-rational actor of mainstream economics, Nelson and Winter (2002: 29–30) write that the only way in which organizations can master the ‘competence puzzle’—the ability to handle immensely complex tasks despite being ‘strikingly ‘‘bounded’’ in their rationality’—is by developing routines which in turn ‘provide a focal point for a learning based answer to the competence puzzle’. These routines persist, they add, because the diYculty of both storing and accessing knowledge means that ‘learning or relearning routines has costs’. Langlois and Foss (1999: 201) argue that recognizing the importance and diYculty of qualitative coordination changes our understanding of Wrm boundary decisions. For example, in product markets where change tends to be systemic and to require ‘simultaneous change in many parts of a complex system—internal organization may prove less costly ceteris paribus’. But when change ‘can take place in separate subsystems without greatly aVecting the way those subsystems are connected together—then markets, which can take advantage of specialized and decentralized knowledge, may be at a relative advantage’. Due to the diYculties of transferring tacit knowledge embodied in routines, Wrms keep some phases of production inside (‘make’) simply because they cannot explain to others what they need, and they may look to collaborate with external suppliers because they are unable to acquire the required knowledge in short order. For example, the decision to invest in vertical integration may have nothing to do with transactions cost minimization but reXect instead the

36

The New Old Economy

inability of the supply network to provide specialized goods of adequate quality at suYcient scale. The necessary productive knowledge, assumed by standard theory to be costlessly available to all, may in fact be all but impossible to convey to suppliers (Chandler 1992; Langlois and Foss 1999). Langlois and Foss (1999: 201) are explicit in characterizing the dispute, writing that the basic heuristic driving TCE ‘is to reduce literally all problems of economic organization to problems of incentive conXicts attendant on imperfect information’ and in so doing forget that knowledge may also be imperfect in the realm of production. That is, as Winter (1993: 185) explains, standard treatments of knowledge treat its ‘storage as costless—the analogue in this context of costless and perfect contracts’. Emphasizing only misaligned incentives, Langlois and Foss (1999: 201) continue, obscures ‘the fundamental role that institutions (including the Wrm) play in qualitative coordination, that is, in helping parties to align not their incentives but their knowledge and expectations’. Where TCE assumes production functions to be equally available to all players in the market, trying to explain variations in organizational form only with Wrms’ eVorts to mitigate opportunism, Langlois and Foss (1999) hold that it would make more sense to do the opposite: assume transactions cost equal across organizational forms and assign all variation to Wrms’ eVorts to coordinate knowledge and production.32 The incentive alignment questions raised by TCE and principal–agent theories may be relevant, but their resolution cannot ignore the uneven distribution of productive knowledge. 2.4. TH E SO C IO LOG Y OF E CO NO M I C ORGA N IZATI O N Sociological eVorts to theorize economic coordination almost unanimously jettison anything remotely resembling the black-box production-function Wrm in favor of more complex stories with room for internal variation and strategic action. The three approaches that are the focus of this chapter are no exception, with some variation in which aspects of the critical economic literature they accept and which they question.33 More importantly, though each emphasizes diVerent mechanisms—from social context to coordinating institutions to novel organizational forms—each provides an explanation as to how Wrms can sustainably collaborate in the face of the twin bugaboos of incentive and qualitative coordination. These bugaboos, according to the economics of organization, should all but force Wrms to integrate vertically to rein in opportunism and/or to permit the sharing of complex tacit knowledge.

The ‘Social’ Explanation: Social Embeddedness and the Network Organization Perhaps the core contribution of the sociological literature on network forms of organization has been to dismantle Williamson’s famous assertion (1985) that the distribution of organizational forms is ‘thick in the tails’ by showing it to be

Networks, Noise, and Institutional Change

37

both misleading and empirically wrong. Treating markets and hierarchies as a relatively pure dichotomy, the argument goes, had obscured that the embedding of actors in particular social contexts can sustain enough trust between Wrms to render networks a type sui generis with situational advantages over competing forms (Granovetter 1985; Podolny and Page 1998; Powell 2001). A network is most generally deWned as a collection of actors pursing repeated exchange relations with each other, but without a legitimate organizational authority to arbitrate and resolve disputes. However, such an inclusive deWnition covers virtually all real economic exchange between and even within Wrms. It is also generally not the usage in the sociological literature concerned with networks as an economic governance structure. These deWnitions generally impose at least some restrictions on the quality of the relationships themselves and require that—relative to market ties—network ties be characterized by greater information sharing and the use of voice before exit (Podolny and Page 1998).34 Also, consistent with the pioneering work of Cyert, March, and Simon on the Wrm as a political coalition, sociologists leave space in the analysis for independent relationships between subgroups within and across organizations. In his very well-known 1990 piece, ‘Neither Market Nor Hierarchy’—the standard citation for claims that networks are not a hybrid form—Powell lists examples of network governance. Along with industrial districts, strategic alliances, and networks in craft industries, he cites the vertical disaggregation of manufacturing (the empirical case analyzed in this book), and makes the essential point that decentralized production per se does not amount to a network. Rather, networks require economic arrangements where ‘one party is dependent on resources controlled by another’, where ‘there are gains to be had by the pooling of resources’, and where relationships are such that, It becomes more economically sensible to exercise voice rather than exit. BeneWts and burdens come to be shared. Expectations are not frozen, but change as circumstances dictate. A mutual orientation—knowledge which the parties assume each has about the other and upon which they draw in communication and problem solving—is established. In short, complementarity and accomodation are the cornerstones of successful production networks. (Powell 1990: 303–4)

As a strong proponent of a social dimension in network deWnitions, Powell (1990: 300–1) was aiming primarily to ‘identify a coherent set of factors that make it meaningful to talk about networks as a distinctive set of coordinating economic activity’. These factors, he surmised, were to be found by recognizing that though all economic exchange is ‘embedded in a particular social context’, it is ‘also the case that certain forms of exchange are more social—that is, more dependent on relationships, mutual interests, and reputation—as well as less guided by a normal structure of authority’. By distinguishing networks as relying on a particular value-orientation on the part of exchange partners, Powell (and those upon whose work he was building) shifted the emphasis from the structure of ties to the behavioral orientation of

38

The New Old Economy

the actors themselves, thus building a foundation for the claim that networks are governed diVerently than are markets or hierarchies. Uzzi (1997: 37) explains that embedded relationships are characterized by trust and personal ties in which ‘actors satisWce rather than maximize on price and shift their focus from the narrow economically rational goal of winning immediate gain and exploiting dependency to cultivating long-term cooperative ties’. In an extensive and excellent review of the networks literature, Podolny and Page (1998: 60–1) explain that for writers in the social embeddedness tradition—typiWed by such noted sociologists as Granovetter, Dore, and Uzzi—networks depend upon reciprocity such that buyers and sellers are ‘willing to make relationship-speciWc investments because each party expects that the other will not use the relationship-speciWc investments to its own advantage’. This relatively more trusting ethic, they write, ‘is one of the deWning elements of a network form of governance, and the network form of governance is therefore not reducible to a hybridization of market and hierarchical forms, which, in contrast, are premised on a more adversarial posture’.35 Uzzi (1997: 37) places the sociological position in stark contrast to the contractual literature: ‘The basic conjecture of this literature is that embeddedness creates economic opportunities that are diYcult to replicate via markets, contracts, or vertical integration.’ Emphasizing social context and interpersonal relationships is suggestive of the possibilities oVered by network forms of organization in a rapidly shifting market context, and embeddedness theorists argue that they mitigate both of the major problems raised by debates in the economics of organization. 1. Networks can lessen the opportunism that leads contractual theories to posit vertical integration and ownership as the best solution to situations in which hold-up seems likely. Williamson (1991) does concede a place for ‘hybrid’ governance forms but in his approach they remain fundamentally contractual. This is rejected by sociologists who argue that it is not so much contract as trust and ongoing social relationships that decrease the threat of opportunism. 2. Thicker interWrm relationships facilitate learning, making a more decentralized structure feasible (recall that the evolutionary perspective holds that vertical integration is driven partially by the diYculty of transmitting hardto-codify knowledge across organizational boundaries) (Podolny and Page 1998). These relationships do more than simply mitigate opportunism; they can also facilitate the exchange of tacit and proprietary know-how. Podolny and Page (1998: 62–3) write that ‘network forms of organization foster learning because they preserve greater diversity of search routines than hierarchies and they convey richer, more complex information than does the market’; they also promote the rapid transfer of information and encourage ‘novel syntheses of information that are qualitatively distinct from the information that previously resided within the distinct nodes’. I emphasize that this is not an unsophisticated literature that characterizes all networks as identical, or that ignores cross-national variation. For example,

Networks, Noise, and Institutional Change

39

DiMaggio (2001b) describes substantial variance in network types, with diVerences in the degree of explicit coordination and in the relative ability of some partners to dominate others. He compares Westney (2001) and Powell (2001) to argue that, relative to Japan—where collaboration is heavily based on long-term interpersonal relationships—networks in the United States tend to focus more on particular projects (though trust, reciprocity, and interpersonal relationships remain important). Likewise, most who take this approach are quite open to the view that institutions matter. For example, in the extensive Italian literature on industrial districts, even authors who give explanatory primacy to notions of a common social culture and interpersonal ties also assign a causal role to associations, local government cooperation, links to research institutes, and other such factors (see Whitford 2001). Although Dore’s (1983; 2000: 45) work on Japanese subcontracting speaks of a culture of long-term commitment and productivist ethic, he also writes of an ‘institutional interlock’ between lifetime employment relations, ‘long-term, obligated customer-market rather than mobile auction-market supplier relations’, and patient capital without risk of takeovers that underwrite a ‘motivational congruence’. Still, while the social networks approach is generally sympathetic to the idea, as Powell (1990: 326) described it, that ‘certain kinds of institutional contexts, that is, particular combinations of legal, political, and economic factors, are especially conducive to network arrangements as well as interorganizational collaborations’, the fundamental contribution of authors in the tradition has been to show how and why particular social contexts sustain network forms of organization and that these network forms have some advantages relative to markets and hierarchies (see Uzzi 1996; 1997). They have not, as Powell (1990: 326) also notes, systematically attacked questions about the ‘kinds of political and economic conditions [that] support network forms’.36 In this sense, the approach Wnds a useful complement in recent sociology-friendly work in comparative political economy that investigates precisely these issues.

The ‘Hard Institutional’ Explanation: Comparative Institutionalism and the ‘Varieties of Capitalism’ Amable (2000: 645) writes that ‘the importance of institutions in economics is more widely acknowledged nowadays than it used to be, say 15 or 20 years ago’ (noting, e.g. that North even got the Nobel Prize in 1993). However, he stresses that it is not nearly enough to simply recognize that ‘institutions matter’. Rather, ‘particularly in approaches concerned with long-run capitalist development as well as those which emphasize the existence of several possible institutional architectures for modern capitalism’, the development of models of capitalism requires ‘the study of how diVerent institutions are complementary to each other’ (Amable 2000: 680, 647). The basic hypothesis, Amable (2000: 656) explains, is that ‘several institutions—in the broad sense—taken together

40

The New Old Economy

reinforce each other . . . so that they form a coherent and stable but not everlasting structure.’ In my own review, critique, and borrowing from institutionalist explanations of the diversity of national models of capitalism I focus foursquare on one of the many approaches37: the actor-centered Varieties of Capitalism (VoC) approach developed by Hall, Soskice, and collaborators in the 2001 edited volume, Varieties of Capitalism: The Institutional Foundations of Comparative Advantage.38 The attention I give the VoC, to the relative exclusion of other approaches in comparative political economy, could be justiWed by its recent vintage, inXuence, and claims to both build on and supersede other accounts.39 But there is a more salient justiWcation: it speaks very explicitly to issues of interWrm collaboration, and its account of the interactions between Wrm behavior and institutional surround directly answers Powell’s questions about the ‘kinds of political and economic conditions [that] support network forms’. In the programmatic statement introducing the volume, Hall and Soskice (2001: 3–5) claim to improve upon and systematize the study of comparative capitalisms in two ways: (a) they ‘build a bridge between business studies and comparative political economy’ by placing Wrms, at ‘the center of the analysis of comparative capitalism and, without neglecting trade unions, highlight the role that business associations and other types of relationships among Wrms play in the political economy’; (b) they reconceptualize how institutions aVect behavior. They hold that conventional understandings of institutions—as socializing agencies, as mechanisms for conferring power on social actors, or as matrices of sanctions and incentives—permit at best only a partial understanding of the ‘institutions of the political economy’. Such views, they write, ‘do capture important ways in which the institutions of the political economy aVect economic behavior’; but they also ‘tend to miss or model too incompletely the strategic interactions central to the behavior of economic actors. The importance of strategic interaction is increasingly appreciated by economists but still neglected in studies of comparative capitalism.’ Answers to the big questions in comparative political economy, they believe, are best got by looking at questions of coordination and ‘the kinds of institutions that alter the outcomes of strategic interaction’ (Hall and Soskice 2001: 5).40 Hall and Soskice (2001: 6) write that the ability of a Wrm to exploit its core competencies depends on ‘the quality of the relationships the Wrm is able to establish’ with both internal and external actors. The implication is that ‘a Wrm encounters many coordination problems. Its success depends fundamentally on its ability to coordinate eVectively with a wide range of actors.’ They cite ‘Wve spheres’ in which ‘Wrms must develop relationships to resolve coordination problems central to their core competencies’: (a) industrial relations; (b) vocational training and education; (c) corporate governance; (d) employees; (e) interWrm relations (this last most important to my story, of course). They maintain that political economies should be ‘compared by reference to the way in which Wrms resolve the coordination problems in these Wve spheres’ and draw

Networks, Noise, and Institutional Change

41

a fundamental distinction between two ideal types: liberal market economies (LMEs), with the United States and the United Kingdom as archetypes; and coordinated market economies (CMEs), with Germany as archetype (Hall and Soskice 2001: 8).41 Though it is impossible in a short space to do justice to the VoC’s extensive framework of complementary coordinating institutions as described in the many essays in the 2001 book, it is worth brieXy explaining what is intended in the arena of interWrm relations. Using Germany as an example, Hall and Soskice (2001: 26) draw on Casper (2001) and Teubner (2001) to explain that not only do business associations work with public oYcials to help improve Wrm competencies, but ‘the presence of strong industry associations capable of promulgating standards and resolving disputes . . . is the precondition for a system of contract law that encourages relational contracting.’ In the United States, by contrast, relationships are based on ‘enforceable formal contracts’, a side eVect of antitrust regulations that make relational contracting more diYcult; furthermore, ‘extensive relationship building’ is diYcult given the lack of ‘dense business networks or associations that circulate relationships for reliability or sharp practice quickly and widely’ (Hall and Soskice 2001: 31). According to the VoC, economies can be expected to cluster with relatively little hybridity into one of the two types at a national level, relying on two interrelated claims about the nature of coordinating institutions: (a) Institutions are self-reinforcing in that they help to build the common expectations that allow ‘actors to coordinate eVectively with each other’, and are ‘inextricably bound up with history’ (Hall and Soskice 2001: 13). Institutions can in principle operate at the regional or sectoral level, but Hall and Soskice (2001: 16) emphasize variation at the national level because ‘so many of the institutional factors conditioning the behavior of Wrms remain nation-speciWc’; (b) ‘There are strong interlocking complementarities between diVerent parts of the institutional framework. Each system depends on the other systems to function eVectively’ (Soskice 1999: 109). These arguments, in combination, ‘suggest that nations with a particular type of coordination in one sphere should tend to develop complementary practices in other spheres as well’ (Hall and Soskice 2001: 18). Firms have strong incentives to ‘gravitate toward the mode of coordination for which there is institutional support’, which means that Wrm strategy and organization should vary systematically by country (Hall and Soskice 2001: 9). In LMEs, ‘Wrms coordinate their activities primarily via hierarchies and competitive market arrangements’ and ‘invest more extensively in switchable assets’ (Hall and Soskice 2001: 8, 17). LMEs provide institutional support largely just for market and hierarchical coordination. CMEs, however, also require institutions that: (a) allow for credible commitments and the exchange of information; (b) facilitate the monitoring of behavior and the sanctioning of defection; and (c) that ‘provide actors potentially able to cooperate with one another with a capacity for deliberation’ (Hall and Soskice 2001: 11). Examples include business

42

The New Old Economy

associations, strong trade unions, cross-shareholding networks, and legal/regulatory systems that facilitate information-sharing. Importantly, though Hall and Soskice (2001: 21) recognize that each of the two types of capitalisms ‘has its partisans’, they explicitly do not argue that one is globally better than the other. Rather, each seems ‘capable of providing satisfactory levels of long-run economic performance’, though they do expect substantial sectoral diVerences in performance across countries. Firms in LMEs—such as the United States, are relatively privileged in ‘radical innovation’ industries that entail substantial shifts in product lines (i.e. biotechnology, semiconductors, etc.). In CMEs by contrast, there is institutional support for interWrm collaboration to help Wrms in industries—which notably include durable manufacturing— dependent on ‘incremental innovation, marked by continuous but small-scale improvements to existing product lines and production processes’ (Hall and Soskice 2001: 39).42 These industries are favored because Wrms can ‘depend more heavily on nonmarket relationships to coordinate their endeavors with other actors’, using relational contracting and ‘collaborative, as opposed to competitive, relationships to build the competencies of the Wrm’. Thus, ‘Wrms and other actors in coordinated market economies should be more willing to invest in speciWc and co-speciWc assets’ (Hall and Soskice 2001: 8, 17). In short, the CME is a theorization of the institutional conditions that favor the emergence of the sorts of network organizational forms that form the crux of the new production paradigm outlined in Chapter 1. Notably, this is not intended by Hall and Soskice to be a mere exercise in classiWcation. Rather, it is a set of causal claims that predicts an almost inevitable long-term clustering of economies into one of the two types and that thus carry substantial implications for policymakers in the American Liberal Market Economy. In a CME framework, governments can work together with business associations to police defection and to credibly administer policies that favor implicit contracts and other forms of collaborative nonmarket coordination. Such policymaking can greatly aid Wrms in durable manufacturing industries. However, the theory predicts that it will fail in an LME because, Soskice (1999: 128) writes, ‘EVective business coordinating capacity cannot generally be built ‘‘spontaneously’’ to service an institutional framework.’ It requires that companies already ‘be engaged in long-term relational contracts’ that can sustain ‘common shared understandings’ and the ‘creation of expert communities across associations, research institutions, and companies’ that ‘can only take place over long periods’. In LMEs, the argument goes, policymakers should instead recognize and embrace their own comparative advantage in radical innovation industries favored by freer market coordination and thus amenable to such ‘blunt’ policy instruments as deregulation and market incentive policies that ‘do not put extensive demands on Wrms to form relational contracts with others’. These, Hall and Soskice (2001: 46) argue, are the only sorts of economic policies that are ‘incentive compatible, namely, complementary to the incentive structures and coordinating capacities embedded in the existing political economy’.43

Networks, Noise, and Institutional Change

43

The ‘Organizational’ Explanation: Neopragmatist ‘Learning by Monitoring’ In an ambitious synthesis of the theoretical implications of Wrms’ increasing collaboration with suppliers in the context of reduced vertical integration, Charles Sabel, along with Helper and MacDuYe (2000: 465; 2004) argues that: (a) the emergence in the United States of ‘nonstandard’ Wrms that are ‘federated, not centralized’ undermines both standard theories of the Wrm and too-static sociological stories that premise interWrm cooperation on ex ante social or institutional conditions; (b) the problem of creating cooperative relationships is thus resolvable through organizational design; and (c) the incipient use by these federated Wrms of pragmatist ‘learning by monitoring’ techniques even in the unfavorable American context amounts to an organizational ‘revolution’ that should push American manufacturing down the path towards the collaborative new production paradigm.

Federated production and the obviation of vertical integration Drawing especially on examples from the automotive industry (which is the source of much of their own empirical work), Helper, MacDuYe, and Sabel (2000: 465) claim that emergent forms of decentralized production represent the seeds of a new and more eYcient model of industrial organization, a ‘nonstandard’ Wrm in which ‘components or services crucial to the Wnal product of one Wrm can be provided by independent companies and the Wrm’s internal specialized producers can provide outsiders with crucial inputs’.44 This is contrasted to ‘standard’ models of the Wrm premised on the belief that eYciencies of specialization create vulnerabilities such that ‘owners of highly specialized, complementary resources’ must fear opportunistic hold-up and thus will cooperate only at great risk—leading to a ‘centralized, hierarchical and vertically integrated Wrm’ in which ‘goals set by headquarters were achieved by hierarchically ranked, specialized subunits, all part of a single organization’ (Helper, MacDuYe, and Sabel 2000: 461). In today’s uncertain markets, they argue, the vertically integrated behemoth is a dinosaur. This is not nearly so controversial a view as it was when Sabel Wrst began pushing it in the mid-1980s, but he has since gone further to maintain that the outlines of an ascendant and paradigmatically new model can be seen in Xexible federated Wrms built around ‘novel methods of iterated goal setting’ that create ‘an information symmetricizing machine in which actors must keep one another abreast of their intentions and capacities’ (Helper, MacDuYe, and Sabel 2000: 466, 472; Piore and Sabel 1984; see also Sabel 2004a; Sabel and Zeitlin 1985; Sabel and Zeitlin 2004). The core element of the nonstandard Wrm is the work group, free to change its own internal organization, to choose inputs from either in or outside the company, and to coordinate directly with other internal units and with external customers and suppliers. Production uses systems of error detection and correction that use ‘breakdowns in the new routines to trigger searches

44

The New Old Economy

for weaknesses of the design or production process that escaped earlier detection’, while ‘design follows a disciplined, decentralized process known as simultaneous engineering’ (Helper, MacDuYe, and Sabel 2000: 466). Based upon ‘information exchanges [that] lead the actors to convergent understandings of the world they are exploring’, federated Wrms have at their core a principle of collaboration called ‘learning by monitoring’ because it ‘ties mutual assessments of reliability to joint explorations of capability’ (Helper, MacDuYe, and Sabel 2000: 466; Sabel 1994). As work groups learn these ‘search routines, problem-solving disciplines and the re-conWguring of Xexible equipment . . . product-speciWc resources are ‘‘de-speciWed’’, coming increasingly to resemble general purpose assets, and thus no longer the instruments or object of hold-ups.’ As a useful side eVect, the information exchange at the core of the model protects Wrms from ‘incompetent or unreliable’ partners by ‘alert[ing] them to this danger before the consequences [are] ruinous’ (Helper, MacDuYe, and Sabel 2000: 471–2). The ‘upshot of all these mechanisms acting together’, the claim goes, ‘is that the construction of Japanese-style production systems does not presuppose the existence of long term relations, because the system in the course of its operation produces them’ (Helper, MacDuYe, and Sabel 2000: 474). Rooting emergent collaboration in the links between Wrms’ monitoring activities and the necessities of collaborative production does two things: (a) it rejects as too static evolutionary models of the Wrm that posit limits to organizations’ ability to question established practices and to diVuse new ones; (b) it challenges the social embeddedness and VoC view that substantial cooperation requires either a particular pre-existing social context that allows nervous parties to share essential information or an accommodating but hard-to-build CME-style institutional framework.

Learning by monitoring and the organization of cooperation Sabel (1994; 1996) has long held that the interWrm cooperation observed in Japan is rooted in an institutional structure that encourages intense interaction between the parties and that allows them to mutually reshape interests to solve emergent problems. He goes further to argue that observed elements of ‘nonstandard’ Wrms integrating aspects of the Japanese model in the United States undermine those who believe that full-blown ‘Japanese-style collective institutions’ and/or a general culture of trust are required to ‘generate the conditions necessary to maintain and nourish collaboration’ (Helper, MacDuYe, and Sabel 2000: 475). Sabel’s approach contrasts game theoretic and cultural explanations that assume that failures of cooperation are best understood as the defection by one or both of the parties when short-term beneWts suddenly outweigh the beneWts of cooperation. Rather, he writes, it is often better to see these as the failure of the parties to adequately pose the problem. Conventional explanations see agreements to ‘fail because of earthly self-regarding motives, not haplessness in the face of higher powers’, he writes, but this leaves them unable to explain

Networks, Noise, and Institutional Change

45

emergent cooperation, for they ‘assume that cooperation is the result of anterior conditions: the alignment of the actors’ self-interests in the one case and the normative characteristics of a group or habits of reciprocity in the other’ (Sabel 1994: 146). They do not ‘contemplate the speciWc possibility that the inner workings of cooperation might transform the actors’ understandings of one another in relation to the commonly deWned world in which their interests are rooted’ (Sabel 1994: 155). Firms reorganizing under learning by monitoring principles are not building ‘thicker’ relationships solely to better assess contractual compliance, nor because of social considerations; rather, they have structured production such that neither party can even know quite what to build or how to build it without regularly sharing information with partners.45 Performance and process data are now being shared in ways that would have once been unthinkable, but ‘in volatile markets, companies realize it is simply too risky to assume that one’s current processes, no matter how much they improve on past practice, are competitive, let alone superior’ (Helper, MacDuYe, and Sabel 2000: 467). As OEMs decentralize responsibilities to suppliers, they come to depend on the ‘continuous, disciplined exchange of information among all those collaborating in production, not the contractual relations among the autonomous agents, each presumed to know by itself what needs to be done to meet its obligations to the others’ (Sabel 1996: 3). By restructuring production in ways that force disparate groups to pool ‘the diverse capacities and experiences of its members to judge the alternatives produced by benchmarking, simultaneous engineering and problem solving searches’ (Helper, MacDuYe, and Sabel 2000: 469), manufacturers develop a model of the Wrm that is ‘a member of a new class of institutions deWned not by the Wxed routines to which they are oblivious, but rather by the routines they use for interrogating and altering their routines. Think of the new institutions as pragmatist: they systematically provoke doubt, in the pragmatist sense of an urgent suspicion that habitual beliefs are poor guides to current problems’ (Helper, MacDuYe, and Sabel 2000: 468–9).46 Sabel (2004a: 2) is careful to caution that ‘these new organizations are not utopia realized’, and is well aware that decentralization brings new risks both to OEMs giving up control of production and to suppliers that ‘cooperate more intimately, but [still] fear betrayal by customers’. Nevertheless, he is emphatic that the nonstandard organizations represent a fundamentally diVerent sort of organization that ‘outperforms hierarchies in volatile environments’. Most importantly, because it depends on a relatively formalized set of routines for questioning and rebuilding routines, the new organizational model ‘can be built in widely diVerent cultures’ and is thus ‘potentially as generally applicable as was the hierarchy’. Indeed, though recognizing that not all organizations are of this new type and that the transition to a new production paradigm is at best partial, Sabel argues that when markets are both competitive and uncertain, Wrms experimenting with learning by monitoring will quickly join this organizational ‘revolution’ in which a ‘permanent uprising against routine becomes

46

The New Old Economy

the key to survival in an otherwise unmanageably turbulent world’ (Helper, MacDuYe, and Sabel 2000; Sabel 2004a: 1). The argument, acknowledged to be both positive and normative, is that the information sharing and benchmarking at the heart of learning by monitoring will cause Wrms experimenting with it in less vulnerable areas to recognize their successes, so that ‘once the cooperative exploration of ambiguity begins, the returns to the partners from further joint discoveries are so great that it pays to keep cooperating’ (Helper, MacDuYe, and Sabel 2000: 445). 2.5. CO NC LUS I O N It is both theoretically and empirically problematic to expect market selection processes alone to be stringent enough—in general—that the mere functionality of a particular organizational form should explain its diVusion.47 Drawing on Sidney Winter’s seminal 1964 PhD thesis, Tolliday and Zeitlin (1991) explain that although Wrms must meet some minimal eYciency criteria, these are often quite loose in real economies in which uncertainty is endemic. Firms cannot know ex ante whether they are optimizing; and scale economies, product diVerentiation, variable access to Wnance, and a host of other reasons imply that the ‘wrong’ Wrms may be the ones to go out of business. The result, Tolliday and Zeitlin (1991: 15) write, is that ‘at any given moment, therefore, the population of Wrms in the economy will typically consist of a heterogeneous amalgam of maximizers and nonmaximizers with respect to current market conditions, and this pattern can be expected to persist indeWnitely in the absence of some longterm trend to static equilibrium.’ Podolny and Page (1998: 57, 67) underscore this concern: ‘as sociologists move away from critiquing what are now somewhat outdated economic views, they need to balance the exclusive focus on prevalence and functionality with attention to constraint and dysfunctionality.’ But they are also optimistic that this same literature already has the tools to better integrate evidence of network failure and nondiVusion ‘with the literature emphasizing the functionality of the network form of governance’, and point especially to work that discusses variations in social and historical context in ways attentive to the lock-in and path dependency eVects that can make changing organizational form diYcult.48 As regards the literatures discussed in this chapter, both the social embeddedness and VoC approaches do balance explanations of how social or institutional conditions permit interWrm collaboration with an ‘attention to constraint’, and explicitly hold that the necessary conditions for collaboration do not always obtain (for the VoC, especially in LMEs). But it is also true that each of these literatures has seen little need to investigate complexities and contradictions within collaborative and arm’s-length relationships; their conceptualizations of these types paints them as far more distinct and stable in theory than they turn out to be in reality.

Networks, Noise, and Institutional Change

47

Each of these positions has been developed in part by demonstrating the empirical salience of phenomena dismissed as unimportant by other theoretical positions—with social embeddedness showing the importance of mechanisms that permit economic coordination governed neither by market nor by hierarchy, and the VoC explaining why organized business sometimes opposes neoliberal restructuring in a globalizing world. I cannot emphasize enough that each has also generated important theoretical advances that I seek not to undermine but to improve with what I believe to be the next useful step. In the analysis that follows, I investigate blurring and within-type complexity to ask if, when and why it matters. This is not simply for the sake of making things complicated, but rather because these complexities can help to show how new and better coordinating mechanisms might be constructed. As American manufacturing adjusts to a changing world, it is essential to move beyond stories in which ex ante social and institutional context are only parametric, as overly static conceptions of social institutions can encourage caricatured debates such as those of the late 1980s in which people wondered if any of the Japanese organizational innovations could be brought into an unfriendly American context (i.e. Japan was able to utilize the Japanese production model because it is Japan; American Wrms could not because they are American, and in fact should not even have bothered to try). A partial solution is perhaps found in Sabel’s neopragmatist approach and its eVorts to distill organizational principles demonstrably replicated in highly varied institutional environments. This undoubtedly eases the explanation of how such organizations arise where they did not previously exist. But as currently formulated, Sabel’s approach falls prey to a problem endemic to the literature on network forms of organization: by focusing only on ‘nonstandard Wrms’ that have already successfully transformed themselves, it builds in a functionalist bias. Proving it possible for Wrms to operate by functionally superior pragmatic principles is valid and very useful, but Sabel’s use of neopragmatism overplays the empirical coherence of the nonstandard type, and is somewhat Xummoxed by the observation that the eVorts of many OEMs and suppliers to build collaborative relationships are both very partial and at times very contradictory. Insofar as theorists focus only on how network Wrms work when they work, they have too little to say about barriers to their construction and are poorly positioned to explain why Wrms actively seeking to build collaborative relationships might nevertheless Wnd those relationships riddled with contradiction and conXict even as they genuinely progress towards the new production paradigm. Rather than dividing Wrms into already coherent ‘standard’ and ‘nonstandard’ types, an analysis of the ongoing decentralization and reorganization of production must also look at Wrms that are merely trying to build thick relationships, focusing on their constraints and motivations and leaving open the possibility of hybrid forms (even when such forms fall far short of the functional ideal). It is not, of course, enough simply to object that existing theories leave little space for contradiction and failure in eVorts by Wrms to build collaborative

48

The New Old Economy

relationships. There are two fundamental challenges—one positive (empirical), one normative—that must be met for mine to be an important contribution to our understanding of the changing organization of production. The empirical challenge : In any analytic distinction, many cases will be hard to categorize, and it is by no means necessarily problematic to ignore these cases as noise; rather, the distribution of relationships as between arms-length and collaborative must be shown not to be ‘thick in the tails’ (the same challenge already met by the networks literature vis-a`-vis TCE). In Part II of the book I thus show that in American durable manufacturing today most relationships —the modal case—are multiplex, rife with contradictions and lie systematically between the collaborative and arm’s-length poles. The normative challenge: Even if I can show the systematic empirical importance of contradictory collaboration between OEMs and suppliers, for this to matter, its recognition must also imply that somebody should—or at least could—actually do something diVerent. Thus, I will build on the empirical claims established in Part II to argue: 1. The identiWcation of systematic blockages of eVorts by OEMs and suppliers to build collaborative relationships means that Helper, Sabel, and MacDuYe underestimate the constraints on the diVusion of network forms evidenced in the VoC and social embeddedness literatures. By implication, if interWrm collaboration is to be a replicable means to sustaining the base of manufacturing jobs, states should look to organize nonmarket coordinating institutions. 2. A combination of systematically contradictory OEM–supplier relationships and eVorts at the subnational regional level to build sustaining institutions for nonmarket coordination eVectively refute the new determinism. Hall and Soskice (2001: 54) explicitly state that theirs is not a static conception of the political economy—and I do not contest that in their framework there is ample space for both LMEs and CMEs to evolve in the face of a changing world—but they also maintain that LMEs and CMEs as types can be expected to retain their coherence because of national-level institutional complementarities. By ignoring within-system diversity and analytically circumscribing the feasible hybrid types, they are far too pessimistic about the tools available for the construction of institutions to facilitate collaborative nonmarket coordination even in the ostensibly unfeasible American context.49

Part II Networks and the Organizational Transformation of American Manufacturing

This page intentionally left blank

Introduction to Part II The three chapters that make up Part II of this book are built around a case study of durable metal manufacturing in the American Upper Midwest. Drawing on over 100 interviews with personnel at large manufacturers and their smaller suppliers, I explain just what the ballyhooed trend towards greater collaboration means in the day-to-day constitution of the interWrm relationship. I show that the decentralization of production in the United States has indeed been accompanied by substantial changes in organizational structure and strategy, but that these changes cannot be adequately grasped by looking only at whether OEMs and suppliers have made the strategic decision to collaborate or not. We need to be asking as well how Wrms are continuously reconstituting and revising relationships. We need to understand why it is that even relationships that OEMs, suppliers, and theorists would all classify as collaborative—in which Wrms openly share information and work towards common goals in some areas—are also characterized by hedging and by ruthless and sometimes deceptive hard zero-sum bargaining enforced by viable exit. It is this dichotomy within relationships that must be explained, which means that any passable conceptualization of changing OEM–supplier relationships must leave space for multiple dimensions of strategic decision-making and variation, as well as for inconsistencies in the behavior of Wrms that are not unitary actors. The simplest distinction to be made in analyzing variation in OEM–supplier relationships parallels the relatively standard core–periphery distinction, in which OEMs strategically segment their suppliers. It is summed up well by management scholar Dyer (2000: 19–20): The guiding principle with regard to arm’s-length relationships versus partnerships is to use [the former] when inputs are low value; commodities or standardized (open architecture) products; stand-alone, or modular with no or few interaction eVects with other inputs; and characterized by a low degree of supplier–buyer independence. In contrast, partnerships are preferable with outside suppliers that produce inputs that are high value; nonstandard inputs or (closed architecture) inputs with the potential to diVerentiate the Wnal product; and characterized by a high degree of supplier interdependence and which have multiple interaction eVects with other inputs.

Dyer emphasizes that the potential gains to partnership relationships are situationally—not intrinsically—superior. They are more likely to be functional, however, in today’s economy because the swing from mass to Xexible production

52

Networks and Organizational Transformation

(or to mass customization) increases the potential gains from leveraging supplier knowledge and specialization. Though a useful starting point, a story of the evolution towards a more decentralized production regime that classiWes OEM–supplier relationships into two mutually exclusive sets—collaborative versus arm’s-length—ultimately obscures as much as it enlightens. There are two reasons why this ‘segmentation’ story makes it diYcult to see why the transition is potentially to a new production paradigm, rather than just a scaled move driven by changing product and market characteristics. First, it is misleading to assume that the optimal relationship will always be dictated by the good to be traded; the characteristics of the good itself are in many cases a primary object of negotiation and deWnition and thus themselves in part constituted by the relationship. And second, an analytic but overly hard relational dichotomy fails to capture important eVorts by OEMs and suppliers to hedge, cooperating in some areas while cautious in others. The Wrst limitation to any explanation for the changing patterns of interWrm relationships that stops at the segmentation story is that it assumes that the likely payoVs to collaboration are knowable from just the ex ante characteristics of the subcomponent to be purchased. Certainly, it is true that OEMs in many metal-using industries have been dramatically aVected by the incorporation of electronics and new materials into products and production processes, and it is relatively straightforward to understand why they would risk strong bilateral dependencies through full partnerships with companies already expert in these technologies rather than making enormous investments of their own. For example, the automotive industry does not have a historic expertise in electronics, and as anyone who has driven in the last 10 years knows, electronics now pervade virtually every subsystem in a car and often must be customized to the particular vehicle—requiring automakers to work quite closely with suppliers to ensure that designs of components work out successfully when integrated. Even apparently old-line manufacturers like farm equipment producers are looking at ways to integrate new electronic technologies, working with Silicon Valley researchers to develop Global Positioning Systems (GPS) attuned to the particular exigencies of farmers that allow for plowing at night or in heavy fog.50 However, the argument that developments in the world of manufacturing could lead to a new production paradigm requires much more than just this sort of high-level collaboration or the joint development of specialized hightechnology modules. It depends on the much stronger claim that there are currently unknown opportunities for specialized subcontracting and joint gains to cooperation to be discovered even in those mundane areas of the supply chain in which the OEMs themselves once were expert. That is, there are novel ways for OEMs to come up with gains from collaborating with their suppliers of metal and plastic components, the many so-called ‘process specialists’ such as metal stampers, machine shops, injection molders, and so on that have taken up the production that OEMs once did internally.

Introduction to Part II

53

In the relationships that these component suppliers keep with their customers, the optimal level of collaboration cannot be derived from the product, because in many cases, the product depends itself on the character of the relationship, including, but not limited to, the relative devolution of responsibilities to the supplier and the willingness of OEMs to adjust their own operations to accommodate changes suggested by suppliers. In this world of mundane but important manufacturing, it is only contingently the case that both parties will Wnd the payoVs to a collaborative relationship to outweigh its costs, where the contingencies follow from strategic and relational interactions between OEMs and suppliers that are conditioned by the larger institutional and market conditions. The second limitation to the simple segmentation story is that it reduces the options available to OEMs deverticalizing production to deal with volatile markets to a dichotomous forced choice: a world of hostile arm’s-length contracting on the one side is pitted against total collaboration involving multiple Wrms with shifting and ill-deWned boundaries on the other. But, as Herrigel (2004: 53–4) notes, reducing the options either to ‘pragmatic collaboration’, or to the ‘old style standard’ Wrm and its arm’s-length capacity subcontracting suggests that anything in-between is either uninteresting or a sort of messy ‘pathology’. Hard analytic dichotomies have their place in social science, but they can also blind to the possibility of a strategic and stable middle ground in which OEMs cultivate ‘a range of diVerent sorts of relations with suppliers, not simply cooperative pragmatic collaborative ones’.51 In empirical fact, Herrigel explains, OEMs often explicitly hedge their bets, juggling ‘their desires to extend their access to technological know-how, maintain quality standards in production and continually reduce costs’. Most, he writes, ‘do not seek to achieve all three goals in every relationship, every time. They attempt instead to meet aggregate goals for particular projects in particular time frames, and they do so through a variety of means. And, there are constantly trade-oVs.’ Moreover, competing strategies across departments and factions of OEMs (which are generally quite large organizations) mean that a ‘single coherent face’ to the supplier is more the exception than the rule, making any consistent pattern of collaboration or arm’s-length bargaining less likely still. In declaring inattention to strategic hedging and to the interconnection between relationship and product to be deeply problematic, I am not simply stating that there is more variation in OEM–supplier relationships than can be captured by a simple binary choice. That would be trivial. Rather, this injection of realism is necessary if we are to understand how American manufacturing could be made to more fully approximate the normatively attractive ‘new production paradigm’ outlined in Chapter 1. In manufacturing today, most relationships—the modal case—are multiplex, contradictory, and lie systematically between the collaborative/arm’s-length poles. Understanding that systematicity, and thus where the relationships ultimately fall, requires analyzing how OEMs and suppliers simultaneously deWne (and then redeWne) product and relationship in what I will show to be a dialogic back-and-forth, a sort of

54

Networks and Organizational Transformation

structured-yet-creative ‘waltz’ that intermingles competition, conXict and cooperation in joint eVorts to identify incremental innovations of product and process that can reduce (total) costs without sacriWcing quality.52

The Case Study and Interviews behind the Argument: Midwestern Metal Manufacturing The remainder of this book draws heavily on a case study of durable metal manufacturing in the American Upper Midwest. Restricting analysis of actual relationships between Wrms to this particular sector and to this particular region is by no means arbitrary. Beyond the need to choose something and somewhere in order to ensure reasonable comparability between the types and determinants of these relationships, the historic role that durable metal manufacturing has played in generating productivity growth and high-paying jobs makes it a logical enough candidate (it is also large, employing 36 percent of all manufacturing workers). The choice of region is virtually impelled by the choice of sector: Wisconsin, Illinois, Indiana, Michigan, and Ohio, the Wve states in the historic ‘rust belt’, contain 33 percent of metal manufacturing employment with but 16 percent of the American population.53 Within metal manufacturing, I focus particularly on the evolution of relationships between OEMs and their ‘component manufacturing’ suppliers. These usually small or mid-sized companies engage in the fabrication and/or assembly of molded, forged, formed, and machined goods made of metal and plastic (plastic components are increasingly utilized in the relevant end-use industries); that is, they perform many of the tasks that were internal to large vertically integrated Wrms in previous periods. Component manufacturers typically sell more than 90 percent of their products to other Wrms (usually large customers), and represent the core of intermediate good production. They absorb 38 percent of the output from the tooling and machine tool industry, and, along with tooling and machinery, provide 88 percent of the inputs to the automotive, construction equipment, and appliance sectors. Employment in the sector grew by 8.8 percent between 1980 and 2000 (up to 10.6 percent of manufacturing employment) and is even more regionally tied than is manufacturing employment generally, with 45 percent of employees concentrated in rust belt states. Component manufacturing Wrms are also increasingly suburban and much less likely to be unionized than are their larger customers, mirroring the trend towards a deurbanization, and deunionization of manufacturing described in the introduction to Part I.54 In short, the component manufacturing sector is an important and geographically concentrated post-outsourcing nexus of Xows that mirrors general patterns in manufacturing. It is thus perfectly placed to answer questions about interWrm vertical cooperation. Altogether, 102 semistructured interviews were conducted: 28 interviews at 12 OEMs in key durable manufacturing industries such as industrial, farm, and construction machinery, electrical appliances, machine tools automobiles and

Introduction to Part II

55

other transportation equipment; and 74 interviews at 44 supplier Wrms in the component manufacturing sector, primarily engaged in the fabrication and/or assembly of molded, forged, formed, and machined goods made of metal and plastic, principally for sale to other manufacturers. Because of a general overrepresentation of the automotive industry in other studies and an ensuing availability to me of secondary literature thereon, I have striven particularly to get good coverage of other areas of this very wide sector (there are of course some interviews with Wrms engaged in automotive supply). I also conducted background interviews with training providers, unions, academics, and others familiar with the relevant issues. The Wrm interviews quoted extensively in the text that follows were primarily (though not exclusively) conducted by me. Except in the relatively uncommon occurrence that the interviewee indicated that it would make him or her uncomfortable (interviewees were guaranteed anonymity), interviews were taped; direct quotations used in the text are in all cases taken from taped interviews. They were carried out under the auspices of the Advanced Manufacturing Project (AMP), a multistate research consortium dedicated to investigating the determinants and possible policy supports of improved performance in the component manufacturing sector. Further information and details regarding AMP, characteristics of the interviewees, and the interview process can be found in the Appendix. The interviews were conducted between August 2000 and January 2002 (that is, during the beginning of a downturn). They generally lasted between one and three hours, and were usually followed or preceded by tours of the manufacturing facilities. The format was ‘semistructured’, using a protocol of open-ended questions about internal organization, relationships to other Wrms, and whether and how these are aVected by the larger market and institutional context; followup questions were asked as appropriate. Interviewed Wrms were in the historic manufacturing rust-belt mostly in Wisconsin (73 interviews) but also in Illinois (23), Iowa (3), and Michigan (3). OEMs interviewed were in a wide range of end-user industries, including major players in truck and motorcycle markets, agricultural and construction machinery, machine tools, lawn and garden equipment and the automotive industry. Supplier Wrms ran the gamut of component manufacturing industries, from machine shops to plastic injection molders, to metal stampers, to producers of braking systems, and so on. At OEMs, interviews were largely with managers responsible for purchasing and managing other aspects of the supplier relationship; when suYcient access was granted, multiple interviews were done (see Appendix). At smaller supplier Wrms, the subject was often a plant manager or perhaps the company president or owner. To ensure that suppliers were comfortable speaking about their customers, and to increase the number of points of view on each of the interviewed OEMs practices, I tried to identify suppliers that worked for several of the OEMs interviewed. This allowed for multiple perspectives on the practices of particular OEMs and on particular OEM–supplier relationships (i.e. both nodes across particular ties). Interviews ‘down the chain’ were then supplemented with

56

Networks and Organizational Transformation

interviews at suppliers identiWed in a range of other ways, particularly through ties to associations. Because interviewees were guaranteed anonymity to allow them to speak freely even about sensitive issues, I never directly identify particular Wrms. I do, however, make my argument and illustrate points whenever possible with examples and real quotes (again, only from taped interviews) and provide contextualizing information as necessary; I am careful to say when an example is drawn from the somewhat particular auto industry and to explain when that does and does not matter. I generally describe bargaining in relatively abstract terms as consisting of two roles: the customer, generally referred to as the ‘OEM’ who contracts to purchase components and/or services; and the ‘supplier’ who accesses Wnal consumer markets by means of selling components to their (almost always) larger customers.

Structuring the Argument: A Brief Road Map for the Next Three Chapters Chapter 3 describes the changing structure of American durable manufacturing and the ensuing importance of the interWrm relationship. I argue that the decentralization of production has created space for innovation and gains to collaboration even at the middle level of small and medium-sized component suppliers, but that their creation requires that the parties cede some exit options. As OEMs and suppliers navigate uncertainties that follow from the devolution of innovation responsibilities, they hedge relationships at multiple levels, defying the predictions of the segmentation hypothesis. Chapters 4 and 5 introduce the metaphor of the ‘cost reduction waltz’ to explain how OEMs and suppliers develop incremental innovations through structured collaboration and information sharing. However, I emphasize that in empirical fact, OEMs and suppliers intermingle both collaborative and arm’s-length transactions as they navigate fundamental uncertainties owing to: (a) technological change and rapidly shifting Wnal market demand; (b) uneven supplier competencies and historically arm’s-length relationships; and (c) inconsistencies in the application of declared collaborative strategies. These Wrst two sources of uncertainty, I argue, need not undermine the vision of American manufacturing as engulfed in an uneven but essentially transitional passage to a more collaborative new production paradigm. However, there are many cases in which interactions between OEMs and suppliers drift beyond the bounds of accepted hard bargaining to include the contradictory and deceptive use of no-holds-barred tactics, injecting substantial uncertainties that in turn undermine even good-faith eVorts by both OEMs and suppliers to construct collaborative relationships. I show that despite eVorts at reform, conXictual and centralized intraWrm relationships plague interWrm relationships, owing especially to factional conXict both within and across departments and to the diYculty of aligning incentives within organizations to reward behaviors aimed at cementing long-term collaboration.

3 The Decentralization of American Manufacturing 3.1. I N TRODUCTI O N The automotive industry gets a great deal of attention in descriptions of the many changes that have occurred in American and global manufacturing in the last quarter century. And much of this attention is deserved. The sector is both mammoth and at the cutting edge in terms of manufacturing organization, a bellwether for trends soon to spread across other industries. It has also been the site of massive restructuring, with all North American and European assemblers adjusting their strategies in the face of relative sales growth in the developing nations and Xat but increasingly diverse demand for cars in their historic richcountry markets. They now produce a wider variety of models, each sold (on average) in smaller volumes than in the past. At the same time, the technological content of cars has increased, pushed in part by new regulations in environmental and safety standards and by the increased use of electronics in modern cars (Veloso 2000). As they reduce the asset intensity of their operations to improve shareholder returns, the assemblers purchase ever larger portions of production, and have focused their own operations largely on design, assembly, and marketing (including the client relationship and especially the Wnance end of things). This sort of retrenchment to just a few core competencies as a means to manage both Wnal market uncertainty and rapid technological change is not, however, limited to the automotive sector. It is in fact widespread across many durable manufacturing industries, in what is an enormous shift. Its magnitude in the last 25 years was summed up by a purchasing executive at a Fortune 500 producer of machinery: I used to tell people, when I was in college, the way it was taught then, a third was overhead, a third was direct labor, and a third was purchasing. That’s not true now. Today, this plant is one of the lower ones in the corporation, except for [strategic subcomponent]. We are probably 65 percent outside, and then your overhead would be around 25–30 percent and labor is 5 percent. We have some facilities that are as much as 75–80 percent purchased. We call them screwdriver factories.

In short, OEM strategies often Wt the claim of a Cap Gemini Ernst and Young (2002: 2) report on the changed investment environment: ‘physical assets have become a liability’.55 This view—dominant today—that purchasing large proportions of value has become a competitive necessity was forcefully underlined

58

Networks and Organizational Transformation

by an interviewed purchasing executive at a major machine tool producer explaining developments in his organization in the last few years: We essentially went from ground zero. What we had done in the past, when we were trying to do everything in our four walls, . . . we would manufacture anything we possibly could internally. We would design everything we possibly could internally. We would assemble everything we possibly could internally. So that we would buy as little as possible, other than raw material and deliver Wnished products. If we choose to follow that strategy we will be out of business very soon.

A supplier to the appliance industry was even more extreme in his analysis of the shift among his customers: ‘I would say the appliance industry manufactures as little as possible. And the trend is to manufacture nothing. The goal is to manufacture nothing.’ With so much production externalized, the need to manage technological change and product development has virtually forced OEMs to reorganize relationships and to reallocate important responsibilities to suppliers.

Automotive as a Lead Sector in the Reorganization of American Durable Manufacturing All major assemblers have ‘gone global’, seeking to produce cars in the markets where they are sold, and increasingly rely on suppliers for strategic aspects of production. As Veloso (2000: 9) notes, ‘Assemblers acknowledge that the critical issue in subcontracting is research and development cost. Manufacturing cost of modules and systems is often as high or higher in suppliers than in assemblers. Therefore, cost-wise, outsourcing becomes worth doing only if the supplier does all the engineering work.’56 It requires assemblers to replicate at least portions of their supply chains in each market, often asking their Wrst tier suppliers to follow them around the globe.57 It has also been coupled with both a rationalization and a reduction in the number of direct suppliers, many of whom are asked to provide full modules or at least subassemblies, to do signiWcantly more design and generally to be more explicitly integrated into the assemblers’ processes (i.e. JIT production).58 This has meant a dramatic increase in investments required to remain a Wrst tier supplier (research and design capabilities, the ability to produce globally and often also to set up new plants near many or all assembly facilities) and has caused a signiWcant restructuring of the automotive components industry (Caprile and Llorens 2000; Veloso 2000). Some of the Wrms eliminated from the assemblers’ direct supply base have either closed or left the industry, but many have simply become second or third tier suppliers, and now provide the reduced pool of Wrst tier suppliers with components for modules and subassemblies (Veloso 2000). Drawing on the numerous studies of the automotive industry by the International Motor Vehicle Program, Veloso (2000: 13) explains that other than raw material suppliers, the remaining direct suppliers ‘are becoming large global Wrms, which are either specialized in complex systems or integrators of several simpler subsystems’.

The Decentralization of American Manufacturing

59

These Wrms must have a global presence and be capable of coordinating a substantial supply chain. The remainder are generally component specialists, a category that encompasses the majority of Wrms in the automotive supply chain, and includes the many Wrms that previously sold directly to the assemblers but now contract instead with the Wrst tier module makers and systems integrators. These can be further divided into: component manufacturers, process specialists almost always at the second or third tier; and subassembly manufacturers with some assembly and integration capabilities, possibly (but increasingly rarely) supplying the assembler directly.59 Such Wrms are expected to aid the Wrst tier suppliers in new product design and development and to meet stringent cost and quality targets.

Not Just Auto: Similar Patterns Across Durable Manufacturing Unlike auto, with its enormous volumes and complex subcomponents, most other durable manufacturing industries operate in mid-to-low-volume batch production industries like trucks, agricultural and construction equipment, appliances, and so on. In such industries, production overhead costs such as tooling and transport must be spread over smaller volumes than in auto. This can protect domestic suppliers against more distant low-wage competition, but it also provides fewer parts over which to amortize investments in design capacities. Nevertheless, the trend of mass-customization leads to generally shortened product life-cycles and an OEM desire to build ‘to order’ rather than ‘to stock’ even here. It forces, for example, a tractor producer like Case New-Holland to be able to produce ‘millions of diVerent product conWgurations’ (Case supplier manual, 1999). This virtually requires these OEMs—who typically purchase between two-thirds and three-quarters (or more, it is common to hear 80 percent) of the cost of goods sold—to demand greater complexity, Xexibility, and eYciency from their suppliers, and to depend on these same suppliers for technologies in which the OEMs themselves are no longer (or perhaps never were) expert. As one mid-volume purchasing executive explained, technology has increasingly become the point of entry for new suppliers. He used plastics as an example: ‘If you look at the product that we shipped out of here 10 years ago, it was all steel outside. Now you will see 70 percent of the outside is plastic. And we’ve made that shift. . . . All of that plastic is purchased. The whole outside shell is plastic now. In the past that was all sheet-metal done inside.’ Later in the interview, using electrical parts as an example, he described the gradual process by which their engineering department was opening to outside expertise and bringing supplier engineers on-site: I call it supplier integration—of suppliers having engineers on-site at our engineering center. Electrical is a good example of that; they would have an engineer sitting right in our group, which 10 years ago at [our OEM], even supply management couldn’t go in those doors, let alone a supplier. So that is a pretty big shift when you go back 10

60

Networks and Organizational Transformation

years. . . . And we have a lot of room to improve. In the example I gave you, electrical, we are probably more using them as a sounding board—‘What do you think?’—as opposed to using them to do all the routings. . . . It is slow, but we are pushing it.

Another purchasing manager detailed his company’s increasingly complex demands of suppliers: We no longer want to buy piece parts; we want to buy things at an assembly level. We want to Wnd assembly houses who can acquire the pieces to build them. We would also want to see either an ability to subcontract engineering or have the internal engineering so that they can act as a competent subassembly supplier, again, and to become part of our design team at the right time. So instead of buying bit pieces we are going to an assembly house with the acquisition capability and potentially the engineering capability also.

The ongoing supply-chain revolution has been led by the automotive industry, where the discourse of low inventory, high quality, lean manufacturing has been around since the end of the 1980s, as has the practice of condensing the supply base both to build leverage for better pricing and to limit the number of interlocutors on which OEMs must focus attention. However, other industries are borrowing and implementing in full or modiWed form target pricing, simultaneous engineering, buying ‘higher in the billables’ (i.e. buying subassemblies), and generally increasing quality and service expectations of their suppliers. Just as in auto, OEMs are culling the supply base to work with fewer suppliers, devolving responsibilities and getting them involved in design earlier.60 Notably, however, component supply in lower volume markets tends to be less stratiWed than in automotive, in the sense that even Wrms selling directly to the OEMs can be relatively small, and may have primarily just raw-material suppliers upstream (Herrigel 2004; Herrigel and Wittke 2004; Whitford and Zeitlin 2004). 3.2. CH AR AC T ERI ZI N G TH E P OS T- O UT S O U RC I N G AM E R I C A N P RO D U C T I V E M O D E L A characterization of the American productive model focused primarily on the coordination of various actors internal to a particular corporation would have been a partial picture even in the Fordist era, but such a framing today, after so much production has been outsourced, is almost nonsensical. After all, most of the transformations of raw material and other additions of value are carried out under diVerent roofs and ownership. In many cases, decision-making about just what transformations to make and value to add requires extensive explicit coordination between Wrms. Ceding control of much of the ‘real’ manufacturing means that OEMs must now negotiate production of the lion’s share of the value in their products, requiring that they coordinate and share information with their suppliers. Improving coordination is not, however, unproblematic, as much of the information that must be shared is potentially quite valuable. Furthermore, as

The Decentralization of American Manufacturing

61

noted in Chapter 2, a prominent position in political economy—exempliWed in Hall and Soskice and the Varieties of Capitalism approach (2001)—argues that we should expect companies in many American manufacturing sectors to be institutionally constrained by the American ‘liberal market economy’ and to have trouble achieving the many innovations aided by the ‘close interWrm collaboration [that] encourages clients and suppliers to suggest incremental improvements to products or production processes’. There are (at least) two potential responses to the Hall and Soskice position. One, oVered by Timothy Sturgeon and collaborators (explained and challenged on empirical grounds in the next subsection) is essentially accomodationist, arguing that standardization and modularization along the lines of contract manufacturing in electronics can and will obviate many of the stickiest coordination problems in ways that render market coordination unproblematic. The other is the larger argument of this book. I accept that the devolution of production responsibilities does place questions of interWrm coordination and collaboration at center stage, and even that there are many ways in which American Wrms are institutionally challenged in building collaborative relationships. However, I dispute the path dependency of Hall and Soskice’s claims, and argue instead that systematic contradictions and hedging behavior of OEMs and suppliers in the construction of the collaborative relationship are indicative of potential paths of institutional construction that are obscured by a misleadingly hard analytic dichotomy.61

Modularization versus the Persistence of Integrated Design An important and inXuential view of the ‘new model of American industrial organization’, associated especially with Timothy Sturgeon (and collaborators such as Charles Fine and Richard Florida) draws on particular and incipient developments in the automotive industry to argue that component supply is likely to evolve similarly to contract manufacturing in electronics. Herrigel (2004: 47) explains that this ‘modularization position’ holds deverticalization to be a process driven by rapid product change and characterized by the deconstruction of product design into discrete subsystems or functional modules (example in an automobile: front end, cockpit, drive train, common chassis platforms, etc.) with standardized interfaces among them. By shifting the production of such modules onto suppliers, OEMs increasingly concentrate their energies on design and seek to structure product design in such a way as to reduce manufacture to a process of following standard guidelines and optimizing inventory and logistical processes. In this way, paradoxically, Xexibility is achieved through the increasing separation of conception from execution in the production chain.62

Modularization could permit the creation of relatively generic manufacturing capacity at the level of component supply, which would in turn ‘reduce transactions costs, build large economies of scale, and reduce risk to network actors.

62

Networks and Organizational Transformation

Standardization and codiWcation of component interfaces would dramatically reduce contracting uncertainty, as OEMs and suppliers could eVectively trade a (so-called) black box, permitting all problematic coordination (prisoners’ dilemma-like situations) to be isolated inside the separate Wrms (Sturgeon 2002). It is certainly true that several of the major assemblers are directing signiWcant energy, resources, and experimentation towards the creation of a viable modular strategy, and key steps necessary to make modular automobile production possible can be seen in the development of a relatively few mega-suppliers in the automotive components industry. Herrigel (2004: 49) notes that there are a ‘number of very well-known experiments with nearly complete modularity currently being undertaken by VW, Daimler Benz and Ford (with more being planned) as well as other more pervasive smaller modular subassembly in many of the automobiles produced by virtually all automobile companies currently in the global industry.’ Nevertheless, while a full modular strategy with a strong concentration of productive and design capacity at the Wrst tier of supply is perhaps theoretically possible, the entirely modular automobile does not yet exist. Sako and Murray (1999, cited in Herrigel 2004: 49) write: ‘A car is at once a group of physically contiguous subassemblies and a series of systems—climate control, safety, electronics and so on. System integration is essential to performance and yet systems may criss-cross physical subassemblies to a degree that renders their separate design almost impossible without sacriWcing performance.’ Herrigel (2004: 49) continues: ‘This obstacle to design modularity is intensiWed in the context of pressure for rapid change and/or special customer demands on the product. Designers and engineers in the heat of delivery pressures and market struggle scrap the tedious process of creating standards and design speciWc customer or product speciWc solutions instead.’ The principle of modular design requires a break between design and manufacture that has proved extremely diYcult in the automobile industry, showing that the sector does not share the same strong structural pressure for modularity as is found in consumer electronics. And if there is not this particular pressure in auto, with its enormous investments in new product development and relatively high volumes, there is even less in the many mid-volume industries that are also an important portion of component manufacturers’ end markets (which is not to say that there are not again eVorts in that direction; but the same arguments as to why it has not fully taken over in auto apply even more strongly to midvolume manufacturing). Herrigel (2004) explains that empirically, there is a persistence of integrated design despite the ongoing deverticalization of production. Using the automotive industry as an example, he gives three reasons why restructuring the component manufacturing industry along the lines of black-box contract manufacturing is a diYcult proposition. First, OEMs that wish to knowledgeably collaborate with their suppliers may seek to retain operations beyond their core competencies in order to maintain

The Decentralization of American Manufacturing

63

the knowledge base to collaborate with (and evaluate the performance of) suppliers.63 Second, the continued involvement of the assemblers in ‘design and manufacture of complex integrated product designs makes OEMs less interested in the scope economies that diversiWed mega suppliers can oVer and also less interested in ceding control over the selection of second and third tier component producers for all parts of the subassembly or module’ (Herrigel 2004: 50). And, most importantly, ‘The range of actual designs of components and subassemblies across automobile models and companies is so vast that no vertically integrated Wrst tier supplier can possibly master all the design, technology and manufacturing know how needed to be competitive’ (Herrigel 2004: 50). As the automotive mega-suppliers become ever more mega, they focus their own production on core competencies and subcontract operations to other suppliers—which means that for many component manufacturers, the real change has not been in what they make, but who they sell it to. Indeed, the ‘failure’ of many Wrms to remain at the Wrst tier in auto can be a strategic choice. Drawing on Automotive News data, Veloso (2000: 17) shows that if a Wrm’s ‘strongest capabilities and competences are associated with particular components, they may be able to do as well or better than systems manufacturers, even if that means working as a second tier Wrm’. 3.3. O U T O F T H E F I RE BUT I NTO A C AN OF WO RM S ? F ROM M ANAGING FINA L M AR K ET U N CERTA IN TY TO NE GOTIATIN G (INCREMENTAL) INNOVATION UNCERTAINTY It is accepted uncontroversially in both the ‘real’ world of business and in the social sciences that innovation of some sort or other is important to the longterm survival of Wrms and industries in market economies. A standard and quite common analytic distinction is made regarding types of innovation, between the radical and the incremental. As the typology is explained by Hall and Soskice (2001: 39) the former entails ‘substantial shifts in product lines, the development of entirely new goods, or major changes to the production process’, while the latter is ‘marked by continuous but small-scale improvements to existing product lines and production processes’. Incremental innovation, they note, is of particular importance in capital goods, such as machine tools and factory equipment, consumer durables, engines, and specialized transport equipment. Here, the problem is to maintain the high quality of an established product line, to devise incremental improvements to it that attract consumer loyalty, and to secure continuous improvements in the production process in order to improve quality control and hold down costs.

The need for incremental innovation poses an obvious risk to OEMs decentralizing production: shunting (downstream) Wnal-market risk onto suppliers generates an important new source of upstream uncertainty—innovation now

64

Networks and Organizational Transformation

depends upon suppliers.64 In a world of rapid technological change, OEMs specializing in just a few aspects of production quickly Wnd themselves unable to build or even to adequately design new products without assistance from other Wrms. But this raises a question: which suppliers fall into the group without which the OEMs cannot do? Certainly, nobody is claiming that every supplier provides a product so speciWc as to make switching suppliers unworkably costly. This recalls the concept of segmentation mentioned in the introduction to Part II. It is undoubtedly the case that OEMs segment their suppliers, maintaining very diVerent sorts of relationships with diVerent Wrms depending on the degree to which continued supply and/or technological and design assistance are viewed as important to the long-term health of the OEM. Counterposed to this group, OEMs also spend large sums on goods that—while involving highly varied levels of sophistication in engineering and technology (i.e. from personal computers to software to steel to plastic resin to rock salt for the parking lot, etc.)—are eVectively commodities with well enough deWned characteristics that purchasing Wrm can switch suppliers in a matter of days or even minutes. For this latter group, the image of purchasing as a sort of anonymous bazaar, perhaps in the form of an online auction, is ideal. At the poles, it is easy to distinguish between suppliers where the bilateral dependency is such that the payoV to—indeed the virtual inescapability of—some form of collaborative relationship is obvious, and those where it is hard to imagine why anyone would use anything other than a straight market transaction. In interviews with personnel at OEMs regarding the question of ‘with whom’ to have a collaborative relationship, the underlying idea of segmentation (that you need diVerent strokes for diVerent folks)—if not always the word itself—was ubiquitous. Everyone says that there are some suppliers from which exit would inevitably be unworkably costly for both sides. However, also ever-present was the idea that there is a middle ground of component supply in which there can be real gains from collaboration—but their creation raises exit costs because gains depend on the sharing of sensitive information and joint deWnition of product. To focus on the poles alone obscures what is the biggest potential positive-sum change that could emerge from a devolution that goes beyond production to include design and other process responsibilities: the possibility of leveraging expertise to simultaneously innovate at multiple levels of the supply chain with an ensuing raised proWle of mid-level component supply. This has not been lost on OEMs and their component manufacturing suppliers in American manufacturing, who are developing relationships that require much more intense coordination than in years past.

The Changing Role of Mid-level Component Supply in Auto The devolution of substantial portions of product development to major supplier Wrms creates bilateral dependencies that take exit almost oV the table, with

The Decentralization of American Manufacturing

65

automotive again as a lead sector. The reorganization of the industry has left key Wrst tier suppliers in charge of much of the design of new products. As it was described by an interviewee at a major supplier (using GM as the example, but in a statement applicable to the other assemblers as well), this means: ‘What happens of course is that the engineers who used to work for General Motors work for the suppliers now. And what you have left at General Motors is program managers.’ The assemblers do have what one interviewee called the ‘big economic hammer’. But, as Automotive News (13 May, 2002) explains, switching suppliers in the face of problems can still be diYcult; the assemblers have some suppliers that are ‘too big to fail’. The article, using GM as an example but stressing that Ford and Chrysler do likewise, explains: The unwritten policy, seldom discussed publicly by GM, acknowledges that the automaker is prepared to help some suppliers rather than risk parts shortages that could idle GM plants. To help these vendors, GM has hastened payments for parts, guaranteed future contracts, postponed price cuts, oVered consulting and even raised the prices paid for components. . . . The automakers do these things because they know it would be time consuming and costly to replace certain parts makers. In fact, the Big 3 calculate the time and cost of transferring parts production, say sources who have seen such lists. That calculation can lead automakers to oVer quick assistance to suppliers who are still solvent.

Although the direct automotive suppliers are ever fewer and ever bigger in a highly tiered structure, this does not mean that the second tier does not Wgure importantly in ‘with-whom-to-collaborate’ decisions—it just means that the responsibility for managing those relationships moves down a level— underscored again in Automotive News (16 December, 2002). An article drawing on an interview with Harald Bolster of Mercedes-Benz passenger car purchasing explained: ‘Subsupply management is a vital task’ for carmakers and their top suppliers, Bolster said. ‘Even small suppliers can bring so much risk to us. We need to ensure there are no cracks in the supply chain.’ Faced with a rising number of car recalls, US and European carmakers have been laying a large part of the blame on small suppliers—notably on makers of electronics components, such as chips. . . . The irony is that in the last decade carmakers have been giving more responsibility to suppliers. Therefore, in many instances, carmakers cannot directly inXuence quality control.

This shift has profoundly impacted the industry. A purchasing manager in a large Wrst tier automotive supplier explained that now, One of the things that the OEMs are looking for is tier-one suppliers who can manage the supply chains, which makes my job interesting and valuable. This is really where that leads to right? They’re asking those questions big time. They want to know how well we’re managing our supply chains. They’re asking for that earlier and earlier. Those of us who can do that, we will win business and we will also make proWts.

66

Networks and Organizational Transformation

This does not just mean coordinating deliveries and other such relatively pedestrian activities. The purchasing manager at another Wrst tier supplier who had worked previously at an OEM explained that now, his Wrm must do ‘what you call managing supplier relationships. That’s where my experience at the OEM level comes in. I know which suppliers can work well in certain areas. Other than just knowing people, too. . . .’ Things such as mutual cost-management strategies, requiring explicit coordination between Wrst tier suppliers and their own subsuppliers are ‘essential. Because we’re being told we have to reduce our costs every single year.’ In short, the highly tiered structure of automotive supply, with a relatively few direct suppliers, does not obviate the need for collaborative involvement of smaller component manufacturers—but it does in many instances move it down a level or two.65 An interviewee at a Wrst tier supplier (the same manager quoted above as having an ‘interesting and valuable job’) described his own company’s need for subsuppliers with particular technical abilities and for products they think will not become commodities for a long time: What is the characteristic of a strategic supplier? It tends to be a high dollar spend for us; a limited number out there; must have good performance; tends to be a highly engineered component; they’re willing to share their resources, to follow us, share their technology. With that kind of supplier, we’ll grow with them, work together with them on mutual cost-management strategies.

For Wrst tier automotive suppliers, this is complicated by their own collaboration in design and continuous improvement projects with their own customers, the assemblers, who at times mandate second and third tier suppliers (due to the persistence of integrated design). A critical supplier to the Wrst tier Wrm might simply be one over whom they cannot develop leverage without involving their own customer directly in the negotiations.

The Changing Role of Mid-level Component Supply Beyond Auto Again, many of the patterns that are exempliWed and best publicized in the gargantuan automotive sector are characteristic throughout durable manufacturing, though with variations that reXect the lower volumes and less deeply tiered structure of component supply in other industries. A high-level purchasing manager at a major mid-volume OEM explained that the massive increase in purchased content and short product development cycles mean that, ‘like it or not’, his company has been making the supply base more ‘strategic’. He deWned strategic in a very straightforward way: diYcult to replace, be it due to customer reliance on a supplier’s technology, risks due to the limited availability of alternative sources, or otherwise. His examples ranged from a plastic injection molder—which used dedicated tooling with a replacement cost of $300–500,000 to make co-developed products—to his wireform supplier, where he was not happy with the suppliers he had, but could not Wnd suYciently expert alternative sources.

The Decentralization of American Manufacturing

67

An interviewee at another mid-volume OEM, describing the structure of his own supply base, explained the pattern. He said that of his approximately 300 suppliers, he could ‘ballpark it and say, I would say probably about less than 20 of those suppliers would be—I believe it’s less than that, it’s maybe 15—would be real key, heavy interface type suppliers, you know, based on the uniqueness of their product. Those would be the engine guys, the transmission guys, the hydrostatic system guys, those types of things.’ And then, of course, there are standardized goods, both highly engineered and not (he gave tires as an example). ‘But then’, he added: there’s also a tremendous number of, you know, little widget type components, that Wt our own design: little injection molded pieces, little shafts, little rods, little gear case assemblies, that type of thing, made to our drawing. And one like that, you know, you can get that made just about anywhere. So it’s engineered, and we need some interface with the supplier, but there’s hundreds of suppliers out there that can do that work.

Even with these ‘widget’ suppliers, he emphasized that as they move the design up front . . . the development process is just a tremendous amount of work. And if you can get it right before you go live in production, both companies win. Unfortunately, it’s diYcult to develop that, that relationship and that information Xow, with too many suppliers. So we want the collaborative relationship but we don’t want the supplier to believe that now that he’s pegged to be our supplier of this particular group of widgets, that it’s a lifelong contract.

With such Wrms, he said: The old, take three quotes and take your best one, and to think . . . you’re doing the best for cost is shortsighted. The best way to do it is to continually develop a relationship with a given supplier, and if you don’t quote every single little widget to three diVerent people, that’s okay, you’re probably not risking much. In the short term, you’re not risking that much; in the long term you’re risking even less

because, in a pinch, it is not too onerous to redesign the piece and move it to a new supplier. From such suppliers, moving is costly, but not prohibitive.66

Component Manufacturer Strategy in a Decentralized Production Regime In the face of rapidly shifting demands from their customers, suppliers are unsurprisingly adjusting their own strategies to respond to the new commercial environment created by their customers’ reorganization. The many interviews conducted for this book with component manufacturers in the American Upper Midwest suggest that there are two underlying logics of supplier response: specialization and diversiWcation, often crossed with each other in a variety of combinations.67 These represent a toolbox of strategies used to exploit niches created by the changing practices of their OEM customers. The most prominent strategy seems to be specialization on a particular set of processes or products, with the smaller Wrms mirroring the OEMs’ focus on a

68

Networks and Organizational Transformation

limited set of core competencies. There are, however, risks to a pure logic of specialization: given OEMs’ focus on cost-containment, it requires an expanding customer base, lest cost reductions simply translate into declining turnover.68 Thus, it is unsurprising that many suppliers are coupling specialization with a logic of diversiWcation that can run: (a) vertically, adding more value in a single supply chain and capitalizing on the OEMs’ desire to reduce the size of the supplier base; (b) horizontally, spreading risk across a wider range of customers and industries; or both. These diVerent logics—each of which bears risks—are not mutually exclusive, but combine to form a strategy space. Specializing suppliers, rather than seeing JIT production as a form of inventory shifting, reduce cycle times in an eVort to drive stocks, work-in-progress, and thus costs out of the entire supply chain. These companies become the real experts in production, using this position to reduce costs from or to add value to the product by focusing on process or design improvements in ways not possible, when they simply built to speciWcations provided by OEM engineering departments so unconcerned with interWrm collaboration that they did not even tell suppliers the end use of components. The rationale behind an intensive focus on a single business was clariWed by a wire harness manufacturer who had previously also done welding, but had eliminated that aspect of the business and shifted the composition of its end-user industries (and hence customers) by investing in capital equipment to focus exclusively on harnesses. This Wrm had recently acquired a contract from a major OEM known for keeping work inside because ‘[the OEM] realized their internal costing was [bad]. They don’t use automated equipment, we use automated equipment’ and can thus make the parts more cheaply. The OEMs do not necessarily have more advanced capital goods and better productivity; suppliers can specialize. Indeed, a machining supplier described a major customer’s attempt to sell him their capital equipment when they outsourced a part family to his company, saying that although the customer thought the equipment was good and was using it, ‘it was not good. I asked them to do a study on their own parts’ using an outside source. Of 148 part numbers, only 12 were up to the OEM’s speciWed tolerances, which is ‘why they were farming it out, they did not want to invest in the capital, as it was outside of their core competence’. Additionally, many—though certainly not all—component manufacturers are providing increased design and services and generally developing relationships with their customers’ engineering departments. For example, a small supplier that has steadily become increasingly engineering-driven (from two engineers to six in just Wve years) commented, ‘The OEMs don’t have the [design] horsepower anymore, they don’t want to mess with it. They recognize that we are the experts at handling metal bending. . . . They ask us to help with design and we do that’, adding, ‘In my short time (12 years) with this company, I have watched it go from OEMs going, ‘‘No, no, no, don’t touch our drawings’’ to OEMs saying, ‘‘Hey what’s the matter with my drawings?’’ or, ‘‘Give me suggestions.’’ They have

The Decentralization of American Manufacturing

69

gotten to that point, now all they have to work on is making the changes’. For suppliers doing more intensive or specialized design, customers often do not come to them with blueprints ready, because, another supplier explained, much of the information required to complete those prints resides with suppliers’ engineers. The plant manager at a relatively large machine shop explained that for his Wrm, The logic to developing engineering, it started [because] the owner always did all the quoting, so he wanted someone with an engineering background to take over that job. So we added our Wrst process engineering. From then on, our customers truly pushed [us] into adding that as a cost center. Because as they outsourced machining, they had expectations about us being able to tell them about machinability, design for manufacture, and it forced us into that issue. It has pushed [us] into a place where we really need to have it.

The goal, of course, is to build a space in which suppliers can—as one said—get a ‘leg up and the power to say I’ve got some control over my destiny’. He said that in today’s market, With the advent of QS9000 [quality certiWcation standards] and everybody being QS, . . . you’re not going to establish a higher price because your quality is better. The expectation is that quality and service is a level playing Weld, same with engineering know-how. From a positioning standpoint, what we don’t want to be is a commodity. That’s a four-letter word. Once we become a commodity, then on an annual basis when they’re asking for price reductions, the threat can ruin and really be true. We want to be able to have the engineering and process know-how not only in just the complexity of the part, but also over the next guy. . . . You always want to make things simple for manufacturing. But if it becomes so simple that it becomes a commodity, then anybody can do it. It’ll be a self-fulWlling prophecy that you’ll be designed out or everything will be reduced to labor costs.

And there is space for component manufacturers to decommodify themselves. Another supplier explained that ‘the more we can get them feeling dependent in that relationship, the more value that we have’ although, he recognized, ‘the bottom line is that it is a cost-reduction to them and it is a price increase to us, . . . but I think it is a risk I think we should be taking.’ Specialized suppliers are able to take advantage of the substantial learning curve that exists in many component supply markets to resist pressure tactics. A mid-volume metal stamper stressed, ‘It’s really diYcult to go online [to an auction] with something that you can’t really spell out what it really takes to make it. You really gotta know all of the tooling that’s involved—the size of the presses—there are just so many operations that you’d have to know about it. If somebody is going to take the time to know all that, it’d take a long time.’ In another example, a supplier of automotive components (a Japanese subsidiary) recounted how they were able to leverage the multiplexity of their relationship to a customer and their technological expertise to resist strong-arm tactics in an online auction. He explained,

70

Networks and Organizational Transformation

We knew, for example, on that one project, we were bidding and it started out at 35 million. And 20 minutes later this was at 20 million dollars. It ended up at 19 million dollars, this project. That is pretty tough pricing. Especially, when you looked at the 35 million mark and you looked at it and said, ‘That’s when we were looking at our competitor’s technology.’ Now of course there is this low cost technology over here [which is] the guys we thought we were going to compete against. . . . [And] we thought it was going to be tough for them to make their product for that volume for 35 million dollars. Then when we saw it plummet all the way down, our bid ended up very low and there were other companies that were below us. We think it was [a particular] technology. . . . So, we were having a hard time understanding how they did that. Now, we still had good margins. But the reason that we did not continue dropping our prices was because of our strategy. We had gone through and had made a concerted eVort to go through and talk with the key players on these particular launch teams, . . . where you have got the design engineer, software engineer, the program manager, and we made them very aware of our product’s capabilities. Because we knew eventually, that when purchasing got done, it was going to go back to a group decision, because a purchasing agent will not buy a technology that won’t work.

Beyond adding design capabilities and becoming experts in their area, many suppliers are further diversifying vertically by adding new process capabilities under a single roof, with the goal of becoming a one-stop-shop for OEM customers. As OEMs lurch towards buying more complete subsystems, suppliers with suYcient capacity to provide multiple services hope to capitalize on OEMs’ culling of the supply base. In cost-competitive markets with thin margins, these Wrms also hope that they will be able to capture a larger portion of the value chain by encompassing multiple processes, and believe that having many operations in-house complements the OEMs’ need for ever shorter lead times. Among suppliers interviewed, elements of this verticalization strategy were common, from a product specialist buying a small gearbox manufacturer so that they could provide a more complete system, to a process specialist who had gotten into proprietary products in the early 1990s to supplement irregular orders from OEMs but that was now getting out. The interviewee claimed that although his customers are better, perhaps, at designing and marketing products, his Wrm is good at production. Upon this realization, his Wrm sold the proprietary product line, restructured to build exclusively for OEMs, and is now looking for ways to increase value-added by performing adjacent operations. They have a contract to make parts and then do Wnal assembly work for some large items designed and marketed by an OEM, and will quote jobs in multiple ways (when customers give them suYcient information about a part’s eventual use), with and without supplementary steps to see if they are competitive on these additional steps. Note that such strategies—adding design and/or new process capabilities— aim primarily at revenue enhancement by capturing more of the same value chain, either by improving quality to move upmarket and adding design capabilities, or by adding complementary upstream and/or downstream operations

The Decentralization of American Manufacturing

71

without encroaching too much on the core competencies of either their customers or suppliers. Followed to its end point—though there is no reason suppliers cannot stop somewhere on the continuum—the logic of vertical diversiWcation leads to a convergence of both product and process specialization strategies. The supplier becomes a full Wrst tier systems supplier, co-designing the product with the OEM, building those portions that Wt their own core competencies and sourcing the rest to other process specialists (upon whom they perhaps also depend for some design help). There are, however, two fundamental risks to a strategy of vertical diversiWcation. First, it is not easy to manage, and potential diseconomies of scope abound. Despite eVorts to maintain coherence in the competencies acquired, diversifying suppliers may have diYculty managing multiple processes or products without adding costly overhead, leaving their market share vulnerable to leaner process specialists. Second, as suppliers tool up to become module-makers and position themselves to acquire Wrst-mover quasi-rents, some will inevitably run ahead of the market, face OEMs slow to devolve full responsibility (and hence, share of value-added) to the supply base, and Wnd themselves burdened with capacities they cannot sell. A horizontal diversiWcation strategy resembles in certain respects that of the capacity subcontractor in that the supplier takes existing process capabilities and seeks to diversify the customer base, ideally expanding across multiple sectors. Among interviewed Wrms, a horizontal diversiWcation is exempliWed by a metal fabricator producing parts and subassemblies for OEMs in agriculture and construction markets that also does contract manufacturing for the computer industry, making the metal parts for servers and doing other subassemblies; likewise, by a tube-bender that uses its capabilities to make both motorcycle frames and muZers for oV-road vehicles. This strategy may complement aspects of vertical diversiWcation (especially in terms of adding new process capabilities), but diVers in its underlying logic by focusing primarily on spreading risk across supply chains, rather than enhancing value-added within them. Notably, it diVers from a verticalization logic in that the supplier will utilize its varied capacities separately, rather than integrating them into a single product, thereby underscoring the point that we are dealing with a continuum of strategies. The advantages of risk diversiWcation are obvious, and horizontalization is particularly desirable for suppliers operating in highly seasonal markets, which would otherwise be left with excess capacity in oV-peak seasons. Horizontalizing Wrms may also beneWt from industrial cross-fertilization, acquiring ideas from one sector and creatively applying them to others. But this strategy has its own problems. The loss of focus inherent in horizontal diversiWcation can become problematic in the context of constant demands for creative cost reduction suggestions. Furthermore, Wrms diversifying vertically into new processes can build up Wxed capital debt, creating a structural tendency towards horizontalization and an incentive to ‘buy business’ by cutting margins to keep the machines

72

Networks and Organizational Transformation

busy, even though the Wrm may never become suYciently expert in the new Weld to make the investment pay oV.

The Upshot: A Raised ProWle of the InterWrm Relationship even at the Mid-level The implications of an increased OEM reliance on supplier technological and design competency even with process-specialists was well explained by a die caster, most of whose customers, like many OEMs today, are no longer ‘manufacturing concerned. They are really assembly companies.’ While they know what they want their parts to do, ‘more and more, our customers are including us in the design of their products because we understand more about die casting than they will ever understand.’ He went on to say that it is not that the customers necessarily had internal die casting, but they had their own big engineering departments. And they had a guy who understood die casting, stamping, injection molding. Now those people are gone. They’ve downsized their engineering departments, and they’ve gone back to their suppliers and said we want to partner with you and when we need a die casting, we’re going to bring you in to work with our people, and when we need a stamping, we’ll bring in Joe Stamper.

Later in the interview, he described how they now will sit in design meetings and often times, we’ll sit there with people who aren’t direct competitors—they have competing processes, and the designers will say something like, ‘We kind of want to make something like this, and we’re not sure, but we think it’s going to be about this big and that long, and here’s sort of a sketch’ and they’ll throw something up on the screen, and, ‘It’s got to have this property and that, and what do you think?’ And I’ll shrug my shoulders and say, ‘You can’t make that out of die casting.’ And I’ll look at the guy next to me and he says, ‘Well, it sure as hell isn’t stamping.’ They don’t know how to make a [appliance]. And that doesn’t bother them. They Wgure it out—sooner or later. And not always correctly. We will sometimes get drawings from people we don’t do business with. They’ll say, ‘We’d like to get a quote on this and we think we’re going to buy three million a year.’ And we’ll look at it and say, ‘You know what? This is not a castable part.’

But the reliance on supplier Wrms to aid in the search for ‘continuous but smallscale improvements to existing product lines and production processes’ goes well beyond suppliers’ ability to answer questions about technological capabilities in the development of new products. There is also an expectation that suppliers actively be on the lookout for ways in which they can improve OEM designs or processes without sacriWcing quality. This is best seen in OEM requirements that their suppliers help them meet targets for ‘cost reduction’—where cost is explicitly distinguished from price. These reductions, at least in theory, do not lower suppliers’ margins, but are supposed to reXect performance improvements, redesigns that ease manufacture, the elimination of steps in the process, and other such changes that improve the ratio of price to value.

The Decentralization of American Manufacturing

73

OEMs so regularly expect suppliers to reduce their prices every year that it is virtually impossible to Wnd suppliers who do not echo in some form or other the comment of an interviewed process specialist who said that even the customers you feel that you’ve got the best relationship with, the reality is, price is important. It gets to where quality and delivery is expected, and price is what diVerentiates. And the reality is that no matter what your relationship is with your customers, in this day and age, if you’re not constantly trying to drive the prices down, . . . somewhere along the line it’s going to hurt you.

The big issue, another explained is that ‘our customers have been unyielding in their demand for lower costs’ and ‘there are only two ways to lower costs: you either reduce your proWt margins or you increase your productivity. We try to increase our productivity.’ In a decentralized production model, many of the opportunities for productivity improvement are found in changes that go beyond the boundaries of the single Wrm. Coupled with the ‘persistence of integrated design’, a major implication of the devolution of production and design responsibility is that much of what should be classiWed as incremental innovation is heavily inXuenced by negotiations between OEMs and their suppliers around the issue of cost, price, and ensuing discussions of what is in fact to be traded.69 Negotiations over cost in fact often focus as well on Wnding improvements of some sort (innovations) in product or process, which forces substantial interWrm coordination, not only because suppliers must have an understanding of how their part Wts into the whole in order to suggest changes but also because changes often force concomitant adjustments by the customer (process in this sense should be broadly construed to include interWrm aspects, such as delivery practices). ModiWcations in one subsystem, or even one subcomponent of a system, may require numerous other changes at multiple points in a single production network, and thus necessitate new testing and redesign of other parts or systems. And though these steps might have been necessary also under vertical integration, then the points of change would be internal to the OEM and would not require such complex coordination across Wrms. 3.4. B E TWE EN S EGM EN TAT I O N A N D H E D G I N G Sturgeon’s descriptions (2002) of modular production networks in consumer electronics make clear that under decentralized production, frequent adjustment to process and product at multiple levels of the supply chain do not inherently require complex coordination. There will certainly always be some standardized interfaces, and there will also always be both radical and incremental changes to product and process that can be had internal to the black box of a particular Wrm. However, as long as much of manufacturing is characterized by relatively integrated product and process design features (as appears likely into the indeWnite future), there will be countless areas ripe for incremental innovations

74

Networks and Organizational Transformation

that do require substantial mutual adjustment. These adjustments are, at a minimum, between a customer and a particular supplier, but may also require coordinating activities with third parties to ensure that the entire product and process Wt together. This means that interpreting the change as sharper segmentation—the idea that there are ‘diVerent strokes for diVerent folks’ that vary with the perceived criticality of particular suppliers—only begins the analysis. In actual practice, the easy division into collaborative and arm’s-length relationships is too simplistic to capture the evolving dynamics of coordination in the complex production systems that characterize contemporary metal manufacturing industries; these systems are, empirically, simply too rife with new uncertainties created by the devolution of innovation responsibilities. The products over which the parties are contracting are certainly important; but when suppliers are actively involved in decisions over what will eventually be made and are expected to be sources of innovation, deciding what is and is not critical is a highly contingent process. Certainly, the good being purchased now matters, but both the product and the circumstances of its purchase—that which renders a supplier critical— can (and does) change for reasons currently unforeseen. However, as game theorists (and others) across the social sciences have amply shown, the mere recognition that cooperative behavior would leave all parties better oV hardly means they will in fact cooperate to become better oV. Using analogies to the prisoners’ dilemma, much work has been done to show that cooperative equilibria require either external interventions to punish defection or some means of guaranteeing continued interaction between the parties. However, this work does not necessarily apply in a straightforward manner to these multidimensional and very negotiated relationships between OEMs and suppliers. Understandings of interests change too quickly and are too poorly deWned before interaction. In a system in which OEMs require technical and process innovation from the supply base but know only approximately where relational returns will be highest, where suppliers investing in competencies to meet new demands need assurances of markets for those competencies, where exit (market) power is unequally distributed even in clear cases of bilateral dependency, there is too much room for slippage and hedging in OEM–supplier relationships. Companies do Wnd ways to collaborate and do structure production in ways that virtually force reliance on others. But even as they are exposing themselves to deep risks of hold-up with great regularity in some areas, they exhibit hard-nosed bargaining in other areas, even with the same bargaining partners. Though vertical integration as a solution to hold-up risk has largely gone by the wayside, OEMs remain wary of getting caught over a barrel in negotiations. They are careful to retain some ability to leverage suppliers against each other and against producers in other areas of the world, sometimes for access to the lower wage costs, sometimes for fears systemic rigidities due to lock-in to particular technologies and/or suppliers, and sometimes for both. And everyone worries about excessive supplier dependence on a single customer.

The Decentralization of American Manufacturing

75

So, while the underlying principle behind segmentation does speak to something real, there must also be space for shifting segmental boundaries and hedging within segments. This claim is developed and defended in the next two chapters, where I analyze the relational implications of the devolution of production and design responsibility to show how and why interviewed OEMs and suppliers actively negotiate new space for positive-sum collaboration even as they carefully hedge that collaboration in response both to each other and to the vagaries of Wnal markets. The interaction of these strategies, I argue, results in multidimensional relationships that are in a sense deWning of product (not the other way around) and that range systematically between the arm’s-length and collaborative poles, both within and across divisions and product lines.70

4 Collaboration in Practice: The Cost Reduction (Incremental Innovation) Waltz 4.1. I N TRO DUCT I ON It has become standard practice for OEMs to expect annual cost reductions from the suppliers to whom they have increasingly devolved both production and design responsibilities, often by selectively borrowing target costing techniques originally developed in Japan.71 These techniques are designed to ensure the protection of supplier margins and need not translate into wage reductions or even into immediate changes in the organization of production, given the numerous and imaginative ways in which suppliers and OEMs can jointly develop incremental innovations of product and process to reduce total costs. In this chapter, I use this idea of the ‘cost reduction’ as a point of entry to explain the relational intertwining of incremental innovation and interWrm collaboration. I justify this narrowing of focus on three grounds. First, as Veloso (2000: 6) notes of the auto industry (in a statement that could be applied to much of durable manufacturing), an increased diversity of models and technological advances notwithstanding, ‘the industry focus on lowering costs has never been as acute’. Second, it gives greater coherence to the narrative and sacriWces little generality; the persistence of integrated design forces Wrms to simultaneously explore changes in cost, quality, and technology as they seek out incremental innovations. Third, my use of ‘cost’ as an organizing frame to discuss negotiations between OEMs and suppliers over the good to be traded is not a rhetoric that I imposed on interviewees, but is one that I learned from them. There are many ways to raise productivity using incremental innovations to reduce costs through a sort of structured yet creative dialogic waltz of coordination and collaboration in which OEMs and their suppliers search for ways to jointly adjust product or process without destroying supplier margins.72 However, both the decision to ‘waltz or not to waltz’ and how well companies waltz vary considerably, with substantial consequences for both OEMs and suppliers (and for the regional economies in which they are embedded).

What Is Target Costing? Target costing has two main dimensions: reducing the initial cost of newly designed products; and ensuring that part costs are further reduced in ongoing

Collaboration in Practice

77

production. In new product development, as the practice is described by Nishiguchi (1994: 126), OEMs use a ‘market-price-minus’ principle, rather than ‘cost-plus’. They work together with a selected set of suppliers who are expected to help evaluate design possibilities, ‘keeping in view what the consumer needs and desires’, but ‘the combined costs of the parts are reduced step by step, toward the target cost while keeping constant the required speciWcations’. For ongoing production, suppliers are expected to meet target productivity improvements. As developed in the Japanese auto industry, Smitka (1991: 142) explains, targets took into account the customer’s own experience of productivity improvement, and were ‘chosen to be achievable, but were kept uniform across suppliers rather than being set higher for Wrms with a good track record or lower for Wrms with a poor one. Fixed targets (and hungry rivals) thus provided an incentive for Wrms to engage in internal process improvements.’ The cost breakdowns used in the bidding process were also used to focus engineering eVorts on areas ‘most amenable to improvement on the basis of interWrm experience. The reduction was thus intended to be fair across suppliers, and achievable in fact—even if it took the aid of the auto Wrm itself.’ The principle behind such contracts is extremely straightforward. OEMs recognize that devolution to suppliers puts this latter group in a much better position both to stay abreast of changes in manufacturing technologies and to see potential product design and process improvements. Suppliers are asked to take part in ‘value engineering’ and ‘design for manufacture’ as well as to take on services—such as inventory management—that are intricately linked to the manufacturing process. OEMs often also require suppliers to meet a costreduction suggestion target, using programs similar to Chrysler’s SCORE program, a formalized mechanism by which suppliers submit ideas for ways to modify product or process to reduce costs without aVecting functionality. The suggestions must be potentially ‘implementable’, though the OEM may of course elect not to explore them. Most importantly, the search for positive-sum cost reductions is widely understood to be most eVective when it is an open and collaborative practice: it requires not only considerable information transfer and joint exploration of production cost drivers, but also concomitant changes at various places in the supply chain. This search can be conceived of as a sort of ‘waltz’ of continuous mutual adjustment, in which suppliers and OEMs intermingle both competitive and collaborative interactions. 4.2. WALTZ I NG AT I TS BES T: C OO PER ATI VE DECOMMODIFICATION IN A DECENTRALIZED PRODUCTION REGIME Spreading the production process across multiple Wrms creates an almost limitless number of areas in which customers and suppliers can react to each other in ways that provide eYciencies. The leaner suppliers get, the more creative these

78

Networks and Organizational Transformation

become, and the more likely that they will require explicit coordination between Wrms. Both OEM and supplier interviewees are emphatic that it is possible to reduce costs without, in the words of an OEM interviewee, ‘absorbing someone else’s margins’. They do not just expect suppliers to improve internal processes, but, he said, When we go on our yearly crusade, we say, ‘Here’s our expectation, you have a right to tell us how to get there, and we’ll review those with you.’ The important point is getting there. And normally we can get a list [of suggestions] out of our partnering suppliers. We normally get a pretty good list about ‘here’s some opportunities that we can see going forward to help you’. Whether you can do it or not, I don’t know.

The reason that cost reductions often require collaboration was well explained by an interviewee at a Wrst tier automotive supplier who contrasted his experience there with other less ‘integrated’ experiences. Although they will look for opportunities, ‘In other industries that I’ve worked [in], it’s always easy to come up with a value-engineering change for our products—change what we oVer and let the consumer buy them. And here, because we’re making for the [OEM] customer, any change that we make must be approved by them before we put it into practice.’ He added that when they have value-added engineering opportunities that we’re able to share with our customer, . . . each customer has its own speciWc nuances, but there is a parts submission warrants of some type which says we have done this change, we are asking for your approval, and depending on the nature of the change, . . . how much technical or testing data the customer would [require] to review the change. One recent change we had was very simple. We changed the part to another supplier, and at the same time, we took out a small piece of the plastic to make the tool more eVective. To anybody else, this is not a functional change. There is no diVerence in form, Wt, or function. But because it was a physical change, we had to go through full approval with our customers to make that happen. The perpetual pressure to reduce costs and the continued interconnections between manufacturing and design combine to encourage dialogue through which incremental innovations can be developed. Explaining why Wrms can no longer count on steady pricing, a supplier explained that uncertain markets mean that their own customers, some of which are higher-tier automotive suppliers, ‘realize that maybe the pricing agreements and the things we agree on today are okay for today’s business, but three years from now they are going to continue to have pressure from their customers to reduce cost, they want as part of the agreement a two and half percent cost containment, cost reduction.’ He went on to say, There is only one way to do that: sit down together. Some of those ideas come from the OEM in optional design changes because . . . that is where manufacturing comes in. If I could make this diVerently, do you really need that close of a tolerance in the application? The engineer who is buying the product is specifying and saying, ‘Well,

Collaboration in Practice

79

no I don’t really need that service machine, I just liked it because it was nice and clean, but I really don’t have to have it.’ That takes cost out, if you don’t have to have it. Do you have to have that hole with that tolerance, because the only way to really get it is to wire the hole, versus simply drilling it? If I can drill it, I can take cost out. Do you really need it plated? That adds cost. What is your real requirement? So through that collaborative eVort we can take cost out of it. . . . But a lot of times, those things are just overlooked. The engineers say half the time, ‘Well I guess that is what is required because that is what’s in the catalog.’ But it doesn’t have to be produced that way. It is that way because we never know who is going to buy it. If you don’t need those machine services, we’ll remove them. We will give you a custom solution even though it looks like a standard application, we’ll customize it.

This waltz of mutual adjustment is comprised of almost endless ways in which OEMs can rely on supplier expertise and direct knowledge of process details to invent ways to improve the manufacturability, cost, and performance of their products through a cumulation of small changes. Examples abound. A sheet metal process specialist explained that they are regularly brought in by the customer to develop ideas on reducing the cost of major products by their customers. In the two-day process, they were just brainstorming with the engineers or the manufacturing guys and asking, ‘What if we did this?’ We’ll keep a list and sometimes there’ll be 200 diVerent items on that list, and then you try to attach a value to that list, and then you focus obviously on risk factors—will the customer accept it, the safety concerns, that type of thing, before you take the next step and put the true costs on that. Ultimately, you end up with some signiWcant redesign on those machines.

A producer of specialized exhaust components emphasized the sorts of creative rethinking of the entire product that can come out of cost-reduction discussions, given that changes to parts in one part of a complex machine often have implications for other subsystems. He gave an example of a discussion in which they were oVering suggestions on how to reduce costs on a part, but there was a safety issue, a heat issue. And so, we were oVering suggestions as to how we could reduce costs . . . all the way around, getting on the end, so and so forth. And it kind of prompted them, some of their engineers, to say, ‘Well, maybe it would be cheaper for us to put a heat shield on the body rather than having the muZer insulated and shrouded.’ Maybe, and they were going to explore that. It might give them an overall reduction. Although they’re going to have to maybe add a little bit more cost into another area. But, you know, it’s worth it because you’re reducing [overall] cost by more than the cost you’re adding. It forces people to look at all these sorts of things.

Altogether, interviews with OEMs and suppliers were peppered with descriptions of successful joint cost-reduction activities far too numerous to list (though I try to give a Xavor here). They run the gamut, from changes in product design, in manufacturing process, in logistics, in sourcing strategies,

80

Networks and Organizational Transformation

and beyond. As Herrigel (2001) explains in an internal research Wndings memo to the Advanced Manufacturing Project describing the cost reduction waltz: Reductions can also be achieved, e.g. through the cooperation of customer and supplier to gain leverage on raw materials purchases OR by redesigning the interface between supplied part and its environment OR by changing the details of delivery conditions OR by giving the supplier business in a new area in exchange for a price reduction in an older area. The possibilities here are very nearly limitless.

Further examples include suggestions that require approval of the OEMs’ engineering departments, such as requests to relax tolerances to allow the supplier to use faster machines, design modiWcations that ease manufacturability or that allow the use of cheaper materials (it is common for products to be ‘overengineered’ or for designs not to fully take into account improvements in available raw materials). Suppliers may be invited to ‘tear-downs’ of OEM products to see how the parts they make Wt into the product in the hopes of stimulating ideas on ways to redesign the interface between the supplied part and its environment: an interviewee explained, ‘We saw a lot of parts that could be made, two could be made out of one, parts that were heavier than needed to be, welds that could’ve been eliminated, parts that could’ve been eliminated, etc. . . . Then they take this information [and ask,] ‘‘Can we live with this, will this be strong enough?’’ That’s a way to cut their costs, and at the same time, we don’t lose any value-added.’ Suppliers with cheaper processes might be given the opportunity to ‘take’ parts from other processes: in areas of fertile but incremental technological improvement such as powdered metal and plastics, suppliers may be given the opportunity to bid on parts currently machined or stamped (more expensive processes) to try to show that they are now able to make the parts (technology improvements mean that parts made from powdered metal are considerably stronger than in the past; in plastics, they are able to injection mold esthetically pleasing parts in ways previously impossible). For example, an interviewed powdered metal supplier explained the creative ways that they try to use the cost reduction process to redirect attention away from ‘mature’ parts to other areas in which they can reduce the OEM’s overall costs: Let’s say you’ve got an existing customer and you’re making some parts for them, and you’ve been making them for them for a long period of time. . . . They want to drive costs out of that. And we’re saying, ‘Hey, we’ve done that, and there ain’t nothing, any more to squeeze out of those things, in our opinion.’ So what we say is, ‘Look, can you give us some more opportunities to cost reduce some existing products that may be made out of something else and get your cost reduction that way?’ You know, can the customer work out a situation where I’m going to get an overall cost reduction, it’s not coming on the parts that I identiWed, but I got some other parts here, that if I go to [my company] and we convert those to powdered metal, I can get the cost reductions there. And that balances out what you can’t give them on the old, established parts.

Collaboration in Practice

81

OEMs are even asking their suppliers to provide full breakdowns of all their process costs—showing margins and thus weakening supplier bargaining power—in order to jointly identify areas in which supplier manufacturing costs are out of line. The goal is to Wnd ways for the OEMs to help reduce costs by, for example, using their leverage with raw material suppliers, or by providing access to expertise (i.e. manufacturing engineers or other specialists). At its limit, a supplier explained, ‘Where you can truly partner is where you are opening up your cost structure . . . [to] say, ‘‘I’m going to be so trustful with you that I am going to show you where my costs are and trust that you are not going to take that and work that against me.’’ We are doing it right now with [Fortune 500 mid-volume OEM].’ 4.3. MU LTI D I MEN S I ONA L I T Y A N D T H E C OM P O N E N T S OF A S UCC ES SF UL WA LTZ These stories of successful waltzing have a clear common denominator: given the persistence of integrated design, the actualization of joint innovations of product and process very often requires explicit cross-Wrm coordination in multiplex relationships. This is why it is the character, quality, and content of the interWrm relationship that pushes particular products and subcomponents back and forth between commodity status—unchanging black boxes bought on bid from the lowest price provider—and the object of creative joint inquiry—in which initial product and process speciWcations serve only as a point of departure. The cost reduction waltz depends upon the involvement of multiple divisions at both suppliers and OEMs, with all parties looking proactively at their own manufacturing processes and product designs, but then looking at other dancers to learn both what is needed and what is possible. The issue, an OEM interviewee explained, is that internally, ‘Many times—and I believe this is the problem that many companies have—purchasing needs engineering and engineering needs purchasing to really eVect these cost reductions.’ While the company in the past ‘would do the design and then go out and try to get the part to our design, [now], I think that we need to move to involve our suppliers more.’ But he added, suppliers must proactively come to them with ideas, because ‘one of the biggest struggles we have is getting the right people to talk at the right times’. He went on to say that ‘probably the biggest thing I’m trying to get out to our suppliers is: Don’t wait for us to spoon-feed you. You need to be part of the solution. You need to come in here and help us Wgure out how to hook up the right people at the right time.’ Echoing a theme common across interviews, he emphasized that open communication is especially important in the early stages of product development, where too often feedback comes only ‘after the engineering print is Wnalized. And what we’re Wnding now is [that] up front in the brainstorming phase is a much better point to get supplier input.’ Underlining this, a supplier explained

82

Networks and Organizational Transformation

that close relationships with customers help them to direct the design process. He said that: Along with the customer, you have to be in that [early development] stage just so you can deWne what [parts are used]. If we want them looking at that tolerance, we’ve got to know exactly what the tolerance of that shaft is, and the stack-up tolerance of the whole application. So if you’re in the beginning stages, you can have an eVect on deciding what parts are used throughout the application to help us Wnd the most costeVective route.

Early discussion is useful even for parts that are largely designed by the OEM. A supplier of machined parts explained: We do design for manufacture, but they still develop the prints in house, and then we get it, try to go back and clean up as we can. Being earlier on would help eliminate it. On some of those projects, we have literally driven out 10–20 percent of the cost before even making a chip. But year two, you get into production, now you are trying to cut additional cost. . . . It is good to be involved up front, you know your [own] suppliers, you can lay it out on the table . . . but down the road, another 3–5 percent, it is extremely diYcult to get.

Suppliers are well aware that cost reduction need not stop with initial design and can be ongoing, relationally multiplex, and incremental. One interviewee explained that they will tell customers, ‘If you do X, Y and Z instead of insisting that you have it like it is, you know, we can take some cost out’; but in those conversations, he emphasized, ‘You’re normally talking to supply people or engineering/manufacturing people, and it’s surprising how big a say all of the marketing people have got in those sort of decisions. You know, ‘‘It distinguishes our product’’, etcetera.’ Underlining the beneWts of proactively engaging the OEMs and of the incremental character of the waltz, another commented: We might say, if we change the resin from Brand X to Brand Z, we might be able to give you Wve cents oV. But then I have to go, I have to requalify everything, it takes an engineering change. . . . It could be, ‘Forget it’, or it could be, ‘Hey that’s a good idea, let’s run the gauntlet.’ It’s not like you can say this is cost reduction month. It’s a continuous thing because come October and it’s time to talk, you have to have already been working on these issues and either already have brought it up to their engineering department, or working out, or sampling things, or coming up with other designs or whatever. So . . . it’s a continuum.

The motivation for open, collaborative, positive-sum relationships between OEMs and suppliers is very straightforward. OEMs that have ceded control of not just production but also product development and overhead services must, as one OEM purchasing manager put it, ‘see that strategically we’re aligning with people. . . . We not only look at what they have today, [but whether] we think we can develop a product in the future together.’ To the extent that suppliers and OEMs are successful in this quest, products that could have been fully deWned by

Collaboration in Practice

83

the OEM and then sent as commodities to the lowest bidder are treated instead as potential objects of joint inquiry and innovative supplier input. That is, they are eVectively ‘decommodiWed’ in decisions the beneWts of which cannot be known from the ex ante speciWcations of the products themselves; these beneWts depend as well on the relative ability of the suppliers in question to actually do something with that freedom to innovate. The upshot, an OEM interviewee explained, is that increasingly ‘we want to establish a partner relationship with suppliers. And if we really want a partnership relationship, we shouldn’t be arbitrarily saying, ‘‘this is what we need from you’’ with no regard to what your business is. We have to look at ways to eliminate fat and waste without taking it out of your margins.’ But even though his Wrm worked with many fewer suppliers each of whom is more strategic than they had been in the past, he said that suppliers rarely try to leverage that power: ‘They know that they could be changed. But when things get to that point, we should have a strategic alliance, a partnership. And our top suppliers know this’; there is substantial information exchange, because ‘we need to understand their cost, just basically, we need to know their business, just as they need to know ours’. For these ‘partner’ suppliers, it is common (though a practice that has somewhat run its course at the automobile assemblers) for OEMs to send teams of manufacturing engineers, often called supplier development, on site to supplier Wrms having trouble developing Xexible manufacturing systems or otherwise unable to meet the emerging needs of a production system that has changed dramatically in what is, after all, a relatively short period of time.73 However, OEMs also recognize their dependence on suppliers’ ability to keep up with technological change, so they do hedge to some degree. The OEM interviewee quoted in the previous paragraph also said, ‘Certainly we are doing more of the single sourcing now than we did in the past’, but, giving an example, he cautioned that in key areas ‘we still may have two suppliers of [a key component], but not both building for a single product line or a single application’. Rather, suppliers are divided across product market segments. That way, ‘at least we have a tune in to general technology, so that if one supplier is deWnitely falling substantially behind, you have a parity check about what is going on. But they can still credibly say: ‘‘Look, you’re going to be the supplier’’ so that we can ask the supplier to open up with their technology on that particular application, without the concern that once we know that, we’ll go to three other suppliers on that particular product application, and try to take their ideas.’ His approach was seconded by a colleague explaining why it is so important to have the ‘right’ supplier. Inventing a nonexistent process to make his point, he said, If you are a supply base manager, the only thing that will let you sleep tonight is knowing that you have your components sourced at suppliers that have competitive advantages. . . . What about technology? Am I the best inertia welder in the world (if that was a process . . . )? What is the competitive advantage? Or is it just margin? Because if it is just margin, it’ll be gone tomorrow.

84

Networks and Organizational Transformation 4.4. T HE LI M I TS O F THE WALTZ

At its best, the collaborative cost reduction waltz leads important relational incremental innovations that can add up to very real money. But it would be foolish to expect the mere goodness of collaboration and decommodiWcation to cause their inevitable occurrence. The goal—about which both OEMs and suppliers are quite explicit—is not to create surplus per se, but to bring it home with you. It is thus perhaps unsurprising that the waltz is also characterized by hard-nosed bargaining and the strategic and cautious withholding of information as parties hedge by intermingling competitive and collaborative interactions. A somewhat extended example can perhaps make clearer how and why incremental innovation works through mundane collaboration and information sharing in relationships that are simultaneously close and prickly. The plant manager at a 140-person small-town machine shop in Illinois told of his eVorts to reduce costs on a long steel rod for a piece of farm equipment that required about three inches of cosmetic zinc-plating at the end. His company relied for its plating on a subsupplier that was 80 miles away, so the zincing was adding $80,000 a year to the cost of making the part. He suspected this was probably not worth it to the OEM, but did not know how the part was used so he could not be sure. He went to a local farm equipment dealer and dug around in the relevant machine (he had grown up on a farm), and discovered that about a half an inch of the rod was visible, making him think he was right. Still, were he to unilaterally make the change his parts would be rejected at the receiving dock. The supplier did not know why the part was still plated, though the possibilities are many. It was likely a legacy design from the days of vertical integration, and was probably more visible on previous designs. Also, when it was made internally by the OEM there may well have been an internal plating department doing batch production; with less expediting, the plating would have added less cost. In any case, stories like this about how OEMs need suppliers to tell them about unnecessary costs and apparently trivial improvements that nobody thought of yet are very common. They also add up to very real money. The OEM did have formal channels for cost-reduction suggestions, so the plant manager submitted his idea. However, he got no response. Some months later, he was at the OEM for another reason, so he grabbed an engineer and brought him over to the relevant machine and showed him the visible half inch of zinc. He explained that it was costing $80,000 per year and that he had already submitted it as a suggestion but that nothing had happened. The engineer replied, ‘Do it again, and put my name on top.’ The next time, it went through, saving the OEM $80,000 per year. Despite such clear shows of commitment, the supplier was nevertheless extremely wary of the same OEM in other areas. This particular customer requires all suppliers to provide step by step production costs when they quote

Collaboration in Practice

85

jobs, ostensibly to allow the OEM to benchmark supplier processes, which should beneWt the suppliers themselves. But when I asked the plant manager if he gave up that information, he said, ‘Well, sort of ’, and laughed. He then explained that he was always careful to muddy the numbers by doing his labor costing in an archaic way. He simply did not trust the OEM to know his margins, and when he was called and told ‘certain numbers are too high or low’, I say, ‘Tell me what you want that number to be.’ He would then recalculate, but without doing anything ‘real’ such as changing processes or the Wnal price. Quite simply, he openly showed the OEM that there were limits as to how far he would go and to what he would let them know, because he simply did not know what tomorrow would bring.74 The back and forth between arm’s-length and collaborative interactions should not be ascribed only to the segmentation of suppliers along product lines—across relationships—as is often assumed. In fact, it often reXects as well strategic hedging within relationships. That it is generally a mix of interactions was emphasized by a large automotive supplier. Discussing the downstream relationship to the assemblers but expressing a sentiment common in interviews, he stressed that at the end of the day, by hook or by crook, the OEMs ‘need their cost to go down. And sometimes they will work with you to make it a winwin situation where you both take out cost. On other occasions, they will leverage you, either by the reverse auction process, electronically or even not electronically, you’re still going to get leverage.’ There is also variation in just how collaborative interactions will be, he said, where ‘[The assemblers] will open up information to allow you to proceed with the project. They are not going to sit there and give you all their design secrets, but they will give you enough information so you can work together.’ This less utopian reality of cost reduction negotiations and pressures was put most starkly by a purchasing manager at a large Wrst tier automotive supplier, speaking of the Chrysler SCORE model premised on an open exchange of ideas to ferret out joint cost reductions. He said: The whole model is that suppliers submit cost reduction ideas. Chrysler, at the time, acted on them. You reduce cost by reducing real cost and not impacting margins. I still think that’s a fantastic way to do business. On the other hand, you can see from [my] model, that I’m also a realist. You don’t always need that. That was one of the weaknesses at Chrysler. They went too far; they acted as if [commodity buying] didn’t exist, and it was just too limited of a way to look at the world. You’re right— any supplier would love that [collaborative] model. But even when that model exists, and it’s working well, we’ll still keep [some commodity buying] because we’ve got shareholders. Economic proWt is driving it.

The point he was making is that although it is not always apparent what action will best return proWts in the long term, there are cases in which the best way for OEMs to make money is to not waltz, and to instead treat the subcomponent in question as a commodity. This eVectively says to a supplier ‘it’s your

86

Networks and Organizational Transformation

problem, deal with these constraints. If you can’t or won’t, we’ll Wnd someone who will.’ There are two fundamental reasons—detailed in the next two subsections—why OEMs may have an interest in refusing to waltz: (a) treating parts as commodities enhances exit options, which favors the party with greater market power; (b) depending on suppliers to deliver incremental innovations only makes sense if suppliers in fact have the competencies to deliver those innovations.

Positional Bargaining, Big Economic Hammers, and Five Percent Letters It is common for powerful OEMs to send what is sometimes referred to as the ‘Wve percent letter’, meaning a demand that the price go down by some Wxed annual percentage by hook or by crook. Such demands are backed by the threat of quick exit, perhaps to lower-wage areas. The tactic was famously used by Daimler-Chrysler shortly after the Daimler takeover, but is in widespread use across manufacturing. It can be extremely high-pressure and result in the breach of current deals. As a supplier to mid-volume OEMs explained, ‘The tone of these letters is diVerent than the three percent automotive, you know: ‘‘We request that you Wnd a three percent decrease for us every year.’’ These are, ‘‘As of January 1, we are going to start paying existing and future POs [purchase orders] at this price, at Wve percent less.’’ I mean, there’s no question about it.’ The use of high-pressure tactics was described by another supplier to midvolume OEMs: One of our customers wants 90 days payment, pay 90 days after he gets his product! . . . Some of them are saying, ‘We’ll take the product in house, and we’ll pay you when we use it.’ I’m serious! That’s not even consignment. It’s ‘when we’ll use it!’ So it just goes on and on. One of our major customers wanted major relief immediately— like three percent every year. You just have to sit down and say, ‘That’s just not realistic.’ But they do that. And then you run into threats. . . . They go to the free markets online [an online auction company] which is a real disaster for many, many people. I really think that sooner or later. . . .

The eVectiveness of positional bargaining is premised on two conditions, both of which often obtain in OEM–supplier relationships: the diYculty of enforcing contractual commitments; and the ability of the OEMs to Wnd alternative sources willing to ‘buy the business’. Asking the interviewer if he had ever seen the contracts typical of the automotive industry, an interviewee said that they can be cancelled at any time: ‘Have you ever read their contracts? They are pretty ugly. . . . Well, essentially, they are one sided.’ He described sending one ‘to our attorney and he said, you know, ‘‘All these things are wrong, but shit, if you want to make parts for them, you’ve got to agree to it.’’ ’ What happens, explained another supplier—who noted that the customers ‘have all the leverage and believe me, they understand that’—is that even if there is a long-term contract,

Collaboration in Practice

87

Most customers have come back in and violated those kind of agreements. Most customers come back and say, ‘We know we negotiated this deal, however, business conditions have changed and we need your help, partner, to help us out of this situation. You know, our competition has changed, our needs have changed, market conditions have changed and we need more from you now than what we had already agreed to.’ So, long-term contracts, they sound nice and are nice things to talk about, but . . . we have found that there are problems in our customers adhering to those contracts.

This reality is echoed by mid-volume suppliers: one, for example, observed, ‘You get long term agreements, but I can’t outspend them in court’, saying that ultimately, the ‘contract is just a formalized handshake that says that your intention is to put business in here and my intention is to at least go to the bank with something to help me Wnance the equipment.’ The ability of the OEMs to get their way with contracts was particularly interestingly, if tacitly, provided by a strategic buyer at a major mid-volume OEM in a conversation about mechanisms of dispute resolution, when the buyer suggested that the use of lawyers was essentially only to cheat suppliers. The conversation was as follows: [Josh Whitford] ‘When there are problems at a supplier, where there is something in the process where it is unclear who is at fault, what is done?’; [Buyer]: ‘We work with that supplier. If there is a manufacturing fault it is theirs, if engineering is at fault, it is ours’; [JW] ‘Are there disputes about whether or not it is manufacturing or engineering?’; [B]: (laughing) ‘Oh there’s always disputes’; [JW] ‘But you don’t call in the lawyers?’; [B] ‘No sir. We do not try to cheat our suppliers. We try to be fair and honest.’ OEMs can also use their global reach to move business to suppliers in lowerwage regions. With widespread recognition among American manufacturers that quality has dramatically improved at suppliers in both Mexico and East Asia, suppliers know this threat can be credible, even when business does not actually move.75 As one interviewee put it: ‘Do I feel global competition? Absolutely. In reality, are my customers buying parts from oVshore suppliers? No. But they’re taking those prices and they’re forcing me to reduce my prices’ (I do not mean to suggest that work never moves abroad—because it certainly does—only that the credible threat of re-sourcing work can obviate the reason to move it). Likewise, OEMs are able to take advantage of suppliers both in the United States and abroad who are willing to buy business by lowering margins to win contracts with the hope of getting a foot in the door. Speaking of this pattern of activity— probably more extreme in, but not exclusive to, the high volumes of the automotive industry—an automotive supplier described what he called the ‘big economic hammer’ of the assemblers: They spend a lot of money and everybody gets attracted to it. And sometimes people think that the automotive companies will be loyal to them, which they are not. And so, what they want to do when they Wrst want to gain entry into a particular automotive arena . . . sometimes, they buy the business. . . . Automotive companies love that because essentially, what happens is that they end up with . . . product, probably below

88

Networks and Organizational Transformation

cost. In which case, if you did that all the time, your company would go out of business. But these companies that do that think that, ‘Okay, I’m now in the automotive arena. I will establish a good relationship with the automotive company and maybe the next project, I will make money on.’ Never happens because there is always someone else lined up and ready to come into the automotive arena who is trying to buy business.

Arm’s-length tactics are used with suppliers with which customers were collaborating around cost reduction, with which they may collaborate again if the relationship’s maintained, and may occur along with other collaborative steps— but are clearly not the same thing. Especially in down markets, pressure tactics are common. Table 1, drawn from the reports of small and medium sized supplier Wrms in the proprietary database of the Performance Benchmarking Service, shows the impact of the substantial price pressures on suppliers, with a median annualized giveback of 3.7 percent (Luria 2002). Suppliers are generally aware that there are limits to the protections oVered by co-development or other such decommodifying practices. For example, a supplier said that while it was useful to get customers to use their patented materials, there were limits to the protections that aVorded them in price negotiations because ‘a supplier who is trying to put forward his two to three percent [increase] every year is in big trouble. You will get shopped around. If you are not holding prices, you are a bad supplier’ and you will be replaced, it is just a matter of time. Or as a supplier of electrical parts remarked of his customer regarding a co-developed part: ‘I know they are honest. I trust them. And I believe that if they are going to stay in that product line, they have got to Wnd a diVerent way to do it.’ The part in question is a ‘fairly sophisticated unit’, where the customer is Wnally saying ‘I know this partnership has been fantastic and we’ve all worked together and we really appreciate it, but we cannot aVord it. We have to go with the lower price.’ So there is a perfect example of the fact that it does not necessarily work when . . . something gets to be of such competitive nature. The co-development, [they were] . . . respectful of it, we’ve had a lot of good business out of it, it has helped us keep ourselves where we are, but in the long term it has not been able to keep the staying power.

Table 1. Price reductions granted, 12 months ended March 31, 2002 Able to raise prices No givebacks A little, but < 2.5% Some: 2.5–4.99% SigniWcantly: 5–9.99% A lot: 10% or more

2% 14% 25% 42% 16% 2%

Collaboration in Practice

89

OEMs’ ability to impose their will is not unbounded, however, as was well explained by a metal stamper who described a time that they ‘lost $1.2 million worth of business Wve years ago because they took it to Mexico’. He said that when ‘the purchasing agent came in and said, ‘‘you gotta reduce your price by 30 percent or I’m going to move it to Mexico’’, I said, ‘‘God bless you.’’ And ‘within one year we had it back’ because it had ‘damn near put [the customer] out of business when he found out he wasn’t going to get his part.’ Another supplier described a time that an OEM customer had come in and said, ‘You know, we really need better pricing’ on a part that he had bid on. The supplier replied, ‘Well, why is it right now you’re looking for a new supplier’, to which the purchasing agent replied, ‘Well, our other one went out of business.’ The supplier said he then asked, ‘And how much of your business did they do?’ Oh, it was about 80 percent of their business. And as he said it, he started turning red, because he knew exactly where we were going. And we said, ‘So, it’s our turn next to go out of business, huh? That’s what you want?’ ‘Well, no, I don’t want anybody to go out of business’, he says, ‘but, yeah, that probably is the reason they went out of business, because they didn’t know how to price.’ And he backed way oV.

In short, leverage is a relative thing, and suppliers are often far from powerless. The collaborative creation of surplus is not independent of its relative disbursement, rendering potentially Pyrrhic Wghts the OEMs could undoubtedly win. Choices already made, such as the devolution of design and overhead services to suppliers often virtually mandate some collaborative interface. OEMs recognize (in the words of an interviewee) that especially ‘when you get into product development, picking the supplier right up front is so key. I’m not going to be able to change it much over time. It’s there.’ OEMs may later threaten to move business as a tool to gain leverage, but they recognize that doing so is diYcult. An OEM manager summed up the issue: A lot of times we are working with suppliers who have historical poor performance, or unacceptable performance. So why do we stay with them? Well, a lot of times we are so engineered into that supplier. . . . You can’t take this design, which was probably jointly developed, and go take it to their competition and get a better valve. We have to redraw, re-test, re-everything. . . . The cost of re-sourcing is huge. We recognized a long time ago that we have to stick with some of our conditional suppliers because of the cost of re-sourcing it. We give them a lot of rope.

Supplier Competencies in Historical Context The OEMs relative market power advantage is perhaps a necessary condition for the proWtability of the exit option, but the use of global sourcing initiatives as well as other exit-based bargaining tactics in the purchaser’s toolbox reXect also an avoidance of lock-in to particular relationships: they are a hedge not only against technological and market uncertainties, but also against deWciencies in the supply base. Oligopoly hardly precludes brutal Wghts over market share (e.g.

90

Networks and Organizational Transformation

the automotive industry), so OEMs that rely on suppliers to develop incremental innovations to reduce costs, to help them pool capacities to amortize market swings, and to stay on the cutting edge of technologies need their suppliers in fact to be competent in those tasks. However, in the historic American context dominated by arm’s-length relationships and capacity subcontracting, in which many small Wrms functioned only as low-overhead shops batch producing OEM-designed parts, suppliers often have substantial diYculties adjusting to the new realities. As OEMs devolve greater responsibilities to these suppliers, a frequent problem, an OEM interviewee explained, is that ‘the relationship has traditionally been build-to-spec, don’t ask any questions, if it doesn’t work, that’s our problem, kind of relationship. I think that there’s probably a history of that that works against this, . . . and a lot of those manufacturers haven’t developed that expertise.’ An OEM manager in charge of supplier development complained that ‘A lot of our suppliers are owned by people that have got old thinking in terms of manufacturing: ‘‘Let’s just get the people working harder’’’ without assessing their own management style. ‘We see a lot of suppliers that have got such old thinking that . . . corrective action and root cause analysis quite often is ‘‘Well, we have a new operator.’’ They don’t [see] that training and dealing with turnover is a part of manufacturing.’ Such suppliers, he said, may even avoid employees who are ‘too good’, because if you have ‘too good’ people, you are likely to lose them. Unsurprisingly, the many suppliers that are increasingly responsible for production are on average less productive than their OEM customers, producing less value-added per direct worker—and the productivity gap has grown signiWcantly since the 1980s. They tend to employ a lower ratio of managers to production workers, and therefore have fewer professional staV to develop strategic plans to reorganize work in eYcient ways that reduce cycle times and cut costs while improving output. These enterprises spend less annually on capital upgrades and new equipment, instead adjusting labor costs to regulate production. They are less likely to be unionized, and invest less in their workforce. These features, together with low capital intensity, mean that less is spent annually on training, jobs tend to require fewer skills, and wages on the whole are lower (Helper and Sako 1995; Helper and Sako 1998; Luria 1996a; Luria 1996b; Luria 2000a). These averages matter and make clear the challenge, but it is also important to recognize that the small Wrm sector is not monolithic and that Wrms can and do change. Drawing on the database of the Performance Benchmarking Service and using value-added-per-full-time-employee as a metric, Luria (2002) has shown that there is high variance in the productivity and wages of sub-500 employee manufacturing Wrms, and that the distribution is very right-skewed—so much so that the top 10 percent of small Wrms in the database are twice as productive and pay twice as much as the median shop (see Figure 2).76 Perhaps more importantly, his data shows that the issue does not seem to be plant size per se but rather a question of management, technology, and labor quality: small plants

Collaboration in Practice

200,000

180,000

160,000

140,000

120,000

100,000

Upper Third

80,000

Middle Third

60,000

40,000

20,000

0

Lower Third

91

Value-added per full-time employee

Figure 2. Manufacturing SME productivity is highly right-skewed

owned by large companies are no less productive, even on average, than are large plants. 4.5 H E D G I N G IN T H E WA LTZ The diYculties that many suppliers have in meeting the needs of the collaborative model and the ability of powerful OEMs to pressure suppliers to jump, scramble and sacriWce margins in order to meet cost targets provide powerful incentives for OEMs to use arm’s-length positional bargaining, so it is perhaps tempting to see shifting patterns of collaboration between OEMs and their smaller suppliers as a simple function of the exit options available to each party (a reasonable measure of market power). However, while power diVerentials form a constant and partially endogenous background condition, to focus on power alone renders static a process better characterized as dynamic decision-making under conditions of uncertainty. Interactions between Wrms are not so much a series of discrete transactions as they are part of the ongoing reconstitution of relationships in which what occurs today conditions notions of what might be possible tomorrow.77 The process, as Herrigel and Wittke (2004: 11) explain (also drawing on interviews with OEMs and suppliers in metal manufacturing), is one in which ‘strategic interest in the present and concern for future business make customer– supplier relations into a conXictual partnership’. They write that ‘OEMs maximize the know-how gains and cost reduction contributions they receive from

92

Networks and Organizational Transformation

suppliers at an aggregate level, rather than at the level of each individual supplier relation, because it gives them more Xexibility.’ This leaves suppliers perpetually unsure ‘what role they will play, or even are playing, at any moment in time— will they be courted for their know-how, integrated into a collaborative process of combined development and cost reduction, or will they simply be pressed for cost concessions on components that were once understood to be part of the previous two categories?’ Herrigel and Wittke’s image of the ‘conXictual partnership’ in which relations are characterized by a combination of collaboration and wariness, of intermingling both spot market and collaborative interactions Wts easily with the descriptions of OEM–supplier relationships oVered by the many people interviewed for this book.78 OEMs and suppliers are both extremely wary of following any pure strategy, and are constantly renegotiating the degree to which they will collaborate in the face of fundamental uncertainties that leave them unsure of their needs, time horizons, the likely payoVs to collaboration, and even of their own and their partners’ ability to follow a coherent strategy. By decentralizing production, OEMs have done more than just transfer some Wnal market and technological uncertainties onto supplier Wrms. Some of the costs caused by these uncertainties have actually been eliminated. Not only do OEMs now have access to a much more varied knowledge base than they could maintain internally, but pooling specialized assets and investments (in supplier Wrms) allows productive capacity to shift more quickly within and across markets. A simple example of this would be the greater ease with which a particular machine shop can not only shift capacity between two farmequipment producers if one is selling while the other is not, but can also work for, say, the heavy truck industry. This is good for everybody, and was logistically more diYcult under vertical integration. Nevertheless, OEMs still Wnd themselves buVeted by pressures beyond their control that leave them uncertain as to their needs from the supply base and very wary of any sort of lock-in.79 At times they are able to invest for the long-term, other times failure to somehow meet sales targets this quarter will mean mass Wrings, and so on. Time horizons are constantly changing in unpredictable ways, so everybody keeps their options open, making OEMs understandably cautious about ceding the ability to use market power to garner short-term gains—even when they hope not to need that power. Describing even those of his suppliers that pay quite low wages, an interviewee at a large market-leading OEM emphasized the pressure on his company to shop around: We have suppliers who are paying eight or nine bucks an hour, but if they are competing with Mexico at three dollars. . . . They’ve got to be the best darn proposition for [our OEM], we can’t buy at a premium. And that is where our suppliers don’t really get it. They don’t really get that our competition wants our market share just like their competition wants theirs. Our message needs to be that. Our message shouldn’t be: ‘My ass is getting kicked by my boss for Wve percent.’

Collaboration in Practice

93

Nonetheless, OEMs decentralizing crucial elements of production often Wnd a price-only strategy to be unsustainable in practice. OEMs certainly do not need supplier innovation in all products at all times, but they absolutely do need it in some products, somehow, at all times. They can treat components as commodities, asking suppliers to, as they say, ‘build-to-blueprint’ by doing incremental innovation themselves around parts treated as unchanging black boxes, but this not without costs. By expecting and allowing suppliers to somehow improve these very same parts (or portions of the process) using relational practices like the cost reduction waltz that depend on the open but structured exchange of information and ideas, OEMs are more likely to stumble onto new ideas that can help them to garner market share. In short, parts can be relationally decommodiWed; this goes back to the counter-intuitive but quite important point that relationships deWne products, as the parties negotiate the making of an integral product across multiple roofs and ownerships. Or perhaps more precisely, there is a constant mutual adjustment between relationship and product driven by the substantial uncertainties as to just what the gains to collaboration will be (because you are really talking about tomorrow’s money). This point is perhaps subtle, but it is also important. The segmentation hypothesis holds that OEM decisions over whether or not to collaborate will be made based on the ex ante deWnition of the subcomponent to be purchased. But though this is certainly a part of the story, so is the inverse. The ultimate shape and characteristics of the Wnal product, when all the subsystems are deWned and integrated, is fundamentally conditioned by the quality of the relationships a particular OEM has with its suppliers; and OEMs and suppliers know this. I thus use the term ‘deWne’ to capture that a Wnal product begins as a more or less inchoate and only partially deWned idea, a sketch, whose eventual design and manufacture must be negotiated across multiple roofs and ownerships. Which subcomponents are treated as unchanging ‘black-box’ commodities is not given, but depends on whether or not OEMs decide it worth the trouble to give suppliers the right to change them; this depends greatly on the guesses its managers and engineers make about the relative ability of particular suppliers to deliver incremental innovations; and this depends on relationships. Once the decision to treat a part as the object of joint inquiry—rather than as a commodity—successes and failures in the sharing of performance and process information will fundamentally impact the shape and characteristics of what is ultimately made, as well as whether that part might again be deemed a commodity in the future. Recognizing this, many in OEMs now accept that supplier aid in improving today’s and designing tomorrow’s products is best got through the credible though partial cession of exit power, so that suppliers feel they can safely invest—lest there be no surplus to divide. However, the unevenness of the American supply base and spotty historic relationships once again force hedging by both sides. OEMs’ interest in moving in the collaborative direction

94

Networks and Organizational Transformation

is conditioned by uncertainties that go far beyond just those intrinsic to the innovation process. The American supply base has traditionally been asked to do little more than ‘build-to-blueprint’, with signiWcantly less intense and open exchange than is necessary in virtually all OEM–supplier relationships today. And though suppliers do generally believe that their best and perhaps only protection against the margin pressures endemic to the so-called commodity product markets is to invest in higher-level capacities such as design and development capacities, they are very wary when told by customers that investments in new competencies will be rewarded with orders, especially given that there are not strong associations or state actors to discipline Wrms defecting from agreements or that can keep suppliers from buying each other’s business by bidding below cost to get in the door. It is not enough, however, merely to recognize that OEMs and suppliers in fact hedge and that this places their relationships somewhere between the collaborative and arm’s-length poles. Because, as detailed in Part I, there are normative reasons to hope that more of these relationships might move toward the former of these poles, Chapter 5 investigates the sources of hedging in greater detail with a particular eye to isolating those which might be correctable through policydriven institutional intervention (the subject of Part III).

5 Uncertainty and Contradiction in the New Old Economy 5.1. I N TRODUCTI O N Chapters 3 and 4 described the tremendous changes that have occurred in the organization of American manufacturing in the last quarter century. Once vertically integrated OEMs have reacted to market and technological uncertainties by radically decentralizing production, and there is ample evidence that many firms are responding to the pressures of deeply uncertain markets by joining forces to develop joint innovations through structured yet Xexible relational practices like the cost reduction waltz. Even the partial diVusion of these and similar practices demonstrate the possibility of a new production paradigm even in the unfavorable historical-institutional context of the American Upper Midwest, built around eVorts to use collaboration as a means to navigate the uncertainties of Wckle global markets and ever more rapid technological change. Yet, as evidence of hedging in the waltz suggest, this possibility is far from an inevitability. There seem empirically to be limits to the spread of that paradigm. In this chapter, I draw once again on descriptions of the cost reduction waltz to show that the partiality of collaboration, and thus the spread of the new production paradigm, is likely to remain problematic unless manufacturing Wrms are provided with greater external institutional support for collaborative initiatives. In spite of very real eVorts by OEMs to reformulate organizational structures and to build collaborative relationships with suppliers, these relationships are nevertheless frequently characterized by ‘bad waltzing’ that diVers fundamentally from the simple use of hard bargaining tactics backed up with the threat of exit power. Simple hard bargaining is widely understood by suppliers to be well within the norms of everyday business and predictable enough that it need not undermine collaboration. But interviews with OEMs and suppliers in the American Upper Midwest show that relationships are also systematically plagued by ambiguous signaling and rife with no-holds-barred tactics used by OEMs exploiting vulnerabilities opened up by the new relationships for short term gain. These deviations, which unequivocally deviate from oYcial Wrm strategies and which both sides agree should not occur, happen often and unpredictably enough that they virtually force suppliers to hedge their own collaboration,

96

Networks and Organizational Transformation

undermining the eVorts even of those in OEMs and suppliers who are genuinely working to create systemic change. 5.2. U NC ERTA I NT Y, TRUS T, A N D T HE S TA BI LI T Y OF TH E INTERMEDIATE CASE: A RETURN TO THEORY In Chapter 4, I argued that hedging in the OEM–supplier relationship is fundamentally a response to uncertainty. And in Chapter 2, I argued that my description of OEM–supplier relationships as stably intermediate between the arm’s-length and collaborative poles Wt poorly with existing theoretical approaches to interpreting these relationships. It is thus worth beginning this chapter with a return to these theories in light of the discussion thus far, not least because it is hardly news to the sociological literature on economic coordination that manufacturers negotiate the market and technological uncertainties that drove them to decentralize in the Wrst place. Nor would many Wnd it odd that contemporary manufacturers are troubled by what could be called competence uncertainty—the problem of getting innovations from suppliers who were once asked only to execute. Indeed, the predictions regarding patterns of cooperation made by each of the theoretical positions outlined in Chapter 2—social networks, institutionalist ‘Varieties of Capitalism’, and neopragmatist ‘learning by monitoring’—can be understood as turning at least implicitly on diVering stories of how Wrms negotiate these uncertainties. Nevertheless, though each of these theories does provide tools for understanding patterns of relationships in American manufacturing, as currently formulated they do not adequately explain how and why the modal case of the interWrm relationship could be stably intermediate between the poles. This can be easily seen when the positions are redescribed with a relatively straightforward modeling of the decision by OEMs and suppliers to engage in collaborative interactions using the familiar notion of ‘trust’ as a means to negotiate uncertainty, though making a distinction between two types— competence trust and trust as reliability—that allow for a clearer discussion of interorganizational (as opposed to interpersonal) relationships.80 Competence trust refers to beliefs about a contracting partner’s ability, or competence, to follow through on stated objectives. Of course, greater OEM trust in supplier competencies should be expected to lead to greater collaboration, since (deWnitionally) one expects competent suppliers to more successfully generate suggestions for product and process improvements, and to generally better meet customers’ needs. Levels of competence trust are obviously not independent of actual competencies of supplier Wrms (which depend in turn on historic choices etc.), and thus can be improved over time with investment in physical and human assets. These investments, however, are unlikely to occur without trust as reliability—the conWdence parties have that their contracting partners will reliably do what they say they will do. Though this conWdence need not be—and often is not—secured by formal contractual means, it is essential in allowing the parties to know exactly what game is being played. Leverage is a

Uncertainty and Contradiction in the New Old Economy

97

relative thing, and suppliers do have some means to resist the hard bargaining of arm’s-length transacting—but if they are to resist, they have to know to try. Suppliers can invest in the ability to come up with collaborative incremental innovations, but they have to believe those investments will be rewarded with sales. In short, while collaboration does not require that there be no hard bargaining or arm’s-length transacting, the parties do need to be able to reliably interpret when the other party is in fact collaborating in good faith. Framed in these terms, the question for everyone becomes one of understanding how and if such reliability can be guaranteed. For the social networks theorists, social embeddedness and the multiplexity of value-orientations serve this function in, for example, Japanese manufacturing (e.g. Dore 1986; 2000), or in the central Italian regions trumpeted by Putnam (1993). For the hard institutionalists in the varieties of capitalism framework, any signiWcant generalization of interWrm cooperation would require an institutional framework that allows for the sanctioning of unreliable behavior (which, as noted in Chapter 2, is thought diYcult to build in the American context). These complementary approaches vary on the exact mechanism ensuring reliability, but both focus on the fundamental facilitating role of particular antecedent conditions, be they initial historical relationships and/or institutional conditions. As such, they oVer a coherent story for why such problems might always get resolved among certain parties—a value-orientation or forced continuous interaction and the ability to sanction defection—and why they might never get resolved among other parties—the lack of same. Following the mechanisms cited, one would predict some cooperative production networks in the United States with Hall and Soskice’s descriptions of dynamics in Liberal Market Economies leaving ample space for the sorts of fast-moving networks with rapid partner switching that Saxenian (1994) identiWes as important in the development of Silicon Valley. There is also reason to expect pockets of sustaining social ties, as Uzzi (1996; 1997) has shown to be important to understanding patterns of cooperation in New York Garment manufacturing. But in general, the factors cited in the previous section as the limits of the cost reduction waltz—traditionally arm’s-length relationships, unequal bargaining power, and historically weak competencies—should be expected to substantially constrain the spread of collaborative network production in American manufacturing, leaving these approaches poorly positioned to explain why so many Wrms are to some degree making initial forays into building collaborative relationships without the (ostensibly) right antecedent conditions. Drawing on these successes—this ‘other’ half—in the cost reduction waltz, Charles Sabel’s neopragmatist position argues that each of these views misses the full implications of decentralized production in a world of radical uncertainty. This, he argues, has made it possible for Wrms to use a new logic of organizing premised on Learning by Monitoring techniques that allow Wrms to relatively quickly build quite collaborative relationships even absent a history of pre-existing trust or formal external coordinating institutions (Helper et al. 2000; Sabel 2004a).

98

Networks and Organizational Transformation

To brieXy reiterate the Learning by Monitoring model: as OEMs decentralize production to navigate the uncertainties of deeply fragmented markets, they are forced to begin learning both from and about their suppliers (who in turn learn from the customers). That is, the parties begin to learn about (monitor) each other’s relative competencies as they explore what to do next. This monitoring is not about ferreting out opportunism, but rather about deciding just what to build and how to build it—the federation of production has left each side unable to make that determination on its own—and has the welcome side eVect that the parties glean numerous clues about each other’s reliability. The dynamic is akin to the structured yet creative search for, and construction of, coordination games in what wrongly appears to outside academic observers (who assert the knowability—or at least estimability—of interests before interaction) to be a thicket of potential prisoners’ dilemmas (where suboptimality is the inescapable outcome, barring external intervention). The result: an eVective facsimile of trust can be built through the interaction of interdependent OEMs and suppliers, as the parties’ initial experimentation with positive-sum collaboration paves the way for new experimentation. This leaves each party more conWdent in the reliability of the other, stimulating the investment in new competencies bolstered by the conWdence of each that hold-up is relatively unlikely. Sabel’s explanation for emergent collaboration contrasts the more generalized, hard-to-build, often cultural, and static understanding of interWrm trust found in much sociological discussion—such as Powell’s description (1990: 303) in which ‘the parties to a network agree to forego the right to pursue their own interests at the expense of others’—with a more formalistic conception that draws on a conceptualization of Japanese production as rooted not in deepseated cultural mores, but in a set of imitable organizational routines that force the parties to jointly question and perhaps realign their very understandings of their own interests.81 This sort of organizationally rooted trust-as-reliability is behind, for example, Dyer’s claim (2000: 88) that the basis of collaboration in Toyota’s supplier relationships is the ‘perceived fairness and predictability of Toyota’s routines for managing external relationships’ as these allow the automaker and its suppliers to depend on ‘rules, norms and procedures (just as Wrms do) for coordinating and sharing the knowledge required to produce complex and customized products’. Sabel (2004a: 1) builds on this idea, arguing that although these new organizational forms were ‘pioneered in Japan’, they ‘can be built in widely diVerent cultures’ and have since been ‘mastered and sometimes improved’ by, for example, ‘the Americans, Danes, and Irish’. This very rational and clean story is an attractive one. It allows for power inequalities and weak competencies to present initial problems but can still explain the otherwise puzzling Wnding that many Wrms are making very real eVorts to collaborate even in the ostensibly unfavorable American institutional context. Using the organizational mechanisms of learning by monitoring, OEMs can hedge against Wnal market uncertainty even as they make informed judgments about the likely beneWts of collaboration based on an assessment of the

Uncertainty and Contradiction in the New Old Economy

99

relative competencies of the supply base. Suppliers, for their part, return the hedge by maintaining arm’s-length relationships when necessary, but also seize the opportunities oVered by their trading partners’ new openness, and thus invest in new competencies, work to improve OEM designs and processes, and generally divulge even risky private information when there is a mutual decision to collaborate. The claim, in short, is that the forced openness of joint design and learning by monitoring creates the conditions for a virtuous circle. They permit good assessments of OEMs’ reliability, which leads suppliers to cautiously but steadily release more information and to invest in new competencies. This encourages OEMs to further decommodify components and to rely on suppliers for help in innovation (making investments in new competencies pay oV ). It creates the conditions for a stepwise move toward a paradigmatically diVerent and more collaborative production model, in which the players move strategically back and forth between arm’s-length and collaborative interactions, but inching ever toward the latter in a steady, virtuous and self-reinforcing pattern in which one instance of reliability encourages another, and so on. This need not end up in total and utopian collaboration everywhere, but it does suggest that those OEMs and suppliers which do experiment with structured joint problem solving should move asymptotically towards collaboration. Today’s intermediate state, one can surmise, thus represents but a transitional stage, the occasional failure but an epiphenomenon, in the ongoing organizational revolution driven by Wrms’ recognition that the ‘permanent uprising against routine’ has become the ‘key to survival in otherwise unmanageably turbulent world’ (Sabel 2004a: 1–2). But attractive or no, this virtuous circle of steadily improving collaboration— in which cooperative practices are marred only by hedging against uncertain Wnal market demands and guesses about the likely ability of suppliers to meet new needs—captures only a part of the relational patterns in contemporary American manufacturing and does not fully acknowledge the diYculties of organizational transformation. Rather, as I will show in the remainder of this chapter, even when OEMs and suppliers in the American Upper Midwest have taken substantial steps to restructure their organizations in ways that force reliance on collaborative relationships, many of those relationships are yet plagued by disingenuous and failed collaboration that does not simply reXect strategic defection in a prisoner’s dilemma dynamic. Rather, it results from systematic failures of Wrms to follow their own collaborative strategies, owing often to intraWrm factors such as factional conXicts within and across OEM departments and the diYculty of realigning incentives to reward behaviors aimed at cementing long-term collaboration. These failures generate substantial organizational uncertainty that encourages both suppliers and OEMs to hedge against the unreliability of contracting partners, blocking the virtuous circle in which reliability begets competency begets reliability. The result is that reform eVorts put forth by some in both OEMs and suppliers are themselves blocked, ensuring that the transition to a more collaborative new production paradigm in

100

Networks and Organizational Transformation

American manufacturing is unlikely to advance much further absent improved external support for such eVorts. My argument that organizational uncertainty is important to understanding the intermediacy of relationships between OEMs and suppliers is in three parts. First, in Section 5.3, I simply show that what I am calling organizational uncertainty exists, matters and can be seen in the failure of many OEMs and suppliers to waltz well even when they are ostensibly committed to try. Section 5.4 then argues that it is a mistake to interpret these failures as random dysfunction or as simple short-term strategic defection by OEMs; rather, they result from a complex interaction between suppliers and groups internal to OEMs. Even Wrms that are quite explicitly restructuring and developing strategies that rely explicitly on eVective collaboration with suppliers nevertheless Wnd those eVorts systematically undermined by problematic intraWrm relationships, a Wnding that squarely contrasts Sabel’s argument that the emergent new logic of organizing built around ‘nonstandard’ Wrms that are ‘federated, not centralized’ and ‘open, not vertically integrated’ has begun an inescapable organizational revolution. Something is undeniably going on, but whatever it is remains altogether partial (and stably so). Section 5.5 concludes this second part of the book by arguing that organizational uncertainty has serious systemic consequences. Wary of OEM unreliability, even suppliers anxious for new responsibilities—believing it to be their best protection against Wnding themselves in total price competition—are virtually forced to hedge, both in the sharing of strategic information for purposes of joint problem solving and in investing in the sorts of engineering and development resources that could enable them to take bigger roles in the development of new products. However, this makes it more attractive for OEMs to treat some subcomponents as commodities and to use arm’s-length transacting (thus undermining the aforementioned virtuous circle). 5.3. WA LTZ I NG BA DLY: CO N TRA DI CTO RY CO LLA B O RATI O N Arm’s-length bargaining tactics such as the Wve percent letter, while certainly not loved, are, as an automotive supplier said, just ‘the way that the market is right now. That is the business. It is not so much that it is unfair because it is only unfair if you’re the only one getting treated that way, but we are all getting treated that way. That is just the way the business is with respect to some of these commercial things that have been happening.’ He went on to explain that ‘most of the time, it is not confrontational or harsh, [customers] just have very high wants. You just have to understand where your business is, understand what your competition is so you know how to respond to it. But by and large, we try to keep the emotion out of it. This is just business.’ The point, an interviewee at another the automotive Wrst-tier explained in talking about his own suppliers, is that ‘everyone has to understand what is fair and unfair’ and ‘that business is business’.

Uncertainty and Contradiction in the New Old Economy

101

Interviewed suppliers clearly interpreted even quite diYcult interactions with customers in this light: hard bargaining, an unloved reality that is well within the rules of the game and, when done above board, predictable enough that it need not undermine trust as reliability. However, they also spoke of another sort of no-holds-barred bargaining that is not interpreted as ‘just business’, that does undermine trust as reliability, and that thus gives suppliers reason to be quite cautious with the open exchanges of information that are the central step in the cost reduction waltz. The name most often associated with disingenuous no-holds-barred bargaining tactics is undoubtedly that of Jose Ignacio (‘Inaki’) Lopez, the executive in charge of purchasing at GM between 1991 and 1993. The process as it happened then was explained by an interviewee at an automotive supplier: before Lopez, he said, ‘If we got a bid, we did our best to come up with the best concept and the best price. If we won it, we won it. If we didn’t, then you continue to work and try to improve until you get business.’ But under Lopez, ‘all of the orders would come in, they would take a look at it, and then come back and say, ‘‘Okay, by the way, we’re not going to accept any of these and here is your target price now.’’’ After looking, they would come back out and drop it about 20 percent and say, ‘Let’s go out and bid again’, creating a new target price lower ‘than the lowest price that came out’. This, he said, was called ‘being Lopezed’, and had the longterm consequences one might expect: many suppliers simply stopped playing the game. This manager said that at his company, ‘If General Motors came out with a bid to us, we would decline to quote it, and I know that it was more than our company alone that did that. . . . Our company was so disgruntled with that, we refused to do business. In fact, quite a few people in our association refused to do business with General Motors.’ Lopez’s name and the strategy associated with it draw rapid and emphatic negative reactions today—as one OEM purchasing manager described the practice: ‘My term for it is unethical. I really like doing business so that it’s . . . no problem being tough and working under pressure, but there’s a lot of reward in being able to look each other in the eye and shaking hands.’ And certainly, among all of the OEMs interviewed for this book, Lopezing had been rejected and the use of a no-holds-barred approach was generally eschewed as a matter of explicit OEM strategy. Nevertheless, interviews with suppliers were almost inevitably peppered with stories of often quite creative deviations from ostensibly collaborative strategies that both sides say should not occur. That is, bargaining is characterized by numerous and corrosive contradictions, in the very literal sense that one thing is said but another is done, with practices that both OEMs and suppliers generally recognize to breach norms of collaborative interaction. Such contradictory collaboration happens often enough and unpredictably enough that it virtually forces suppliers to react in ways they know to undermine future collaboration. Deviant bad waltzing includes, but is hardly limited to: irrationally inXexible price targets; misleading forecasting and other means of inventory shifting; the

102

Networks and Organizational Transformation

sharing of inside information; hold up and other abuses of trust; and the deceptive use of online auctions.

Nonnegotiable Cost Reduction Targets: When the Waltz Is All Lead and No Follow When the OEMs’ legitimate (and eVective) eVorts to push suppliers to ferret out cost reductions evolve instead into the inXexible application of a hard and nonnegotiable target, systemic irrationalities such as gaming and even disinvestment can arise. It is not always feasible to reduce costs by any Wxed annual percentage, especially when raw materials represent the major component of costs. It is not uncommon to hear the view articulated by a supplier who said that the ‘whole program of regular cost-reductions is sort of being seen as what it was, which was sort of a bogus attempt on the part of very big companies to squeeze their suppliers further and further’. Speaking of the same thing—what he called ‘this game they play called supplier suggestion’, an automotive supplier said: If I was successful in driving the cost down in the product, I would try and not let them know about it because I think for the most part, they are going to come back to you anyway and say, ‘This next year, you have to lower the price by this amount.’ And if I do it, Wne, I will do that but if I can continue to drive the cost down on my own, I am not going to pass it on to them; because next year, when I have to bid on that contract again, or when the contract is up, they won’t care if I saved them money.

A little later in the interview, he underlined the point: If I was to do everything like this and tell them, ‘I have made the improvement and by the way, here’s your money from all of the work that I have done to improve it’, I am not silly enough to believe that two years later when I am bidding on a new project, they are going to look at the scorecard and say, ‘[interviewee name], you helped me reduce the price on that previous program. I am going to give you that reward even though your cost is a little bit higher than someone else’s.’ No, that’s not going to happen. . . . What have you done for me today? I don’t care what you did for me yesterday.

Exacerbated by buyer turnover—as one supplier said, ‘We might have made some guy a hero, and next year, you have someone who [could not care less] about what you did for that guy’—the failure to give adequate credit for past success coupled with a nonnegotiable enforcement of annual targets (including suggestion targets) pushes suppliers to play the cost reduction programs like a game in which they temporarily withhold ideas to make sure they correctly fall into Wscal years (that is, they are careful not to suggest ways to save 7 percent one year, lest they fail to meet 5 percent the next). Such gaming of systems can be a particularly attractive hedge when suppliers are asked to help in the initial design of a products, where success early in eliminating the so-called ‘low-hanging fruit’ makes removing costs later more diYcult. As a supplier of parts for nonautomotive vehicles explained,

Uncertainty and Contradiction in the New Old Economy

103

The biggest problem we had with [a large Japanese OEM] was that we got beat up because we did not cost down their product enough, and our argument was that we took all the cost out before we built it. Their engineer would say, ‘Here is the blueprint’, and we would take their $50 [component] and make it into a $28 [component]. . . . And then they come and say, ‘We paid $28 last year and we are still paying $28.’ Yeah, but you would have paid $50. I kind of thought, ‘Okay, we’ll play that game, . . . we will keep those ideas in our factory and we will deal them out to you.’

The reason to game, obviously, is that cost reduction targets might otherwise only be met out of margins, advancing too little for new investment; and without such investment, it is diYcult to come up with the sorts of productivity improvements that can make annual cost reductions sustainable. A machine shop manager heavily involved in design for manufacture explained that care is required ‘because you can only run the part for maybe a year, and then they put it out for quote, and somebody who has maybe less overhead, less costs [underbids you]. . . . You lose the work and all that work you did is gone; happens every day.’ For the OEMs, causing such disinvestment—eating the seed corn—makes long-term sense only if they believe that the supplier can be replaced with minimal cost because of generalized overcapacity in the industry and minimal asset-speciWc investment. The short-term incentives are obvious. THE M ISUSE OF I NSIDE INFORMATION Stories of OEM misuse of private information often revolve around the relaying of supplier cost information or of supplier-provided cost-reduction ideas— submitted to meet targets set by customers—to the supplier’s competitors. One supplier reported, for example, that they had submitted a detailed proposal for cutting out a step in the production of a particular component at the request of the customer, who then passed on the suggestion to another Wrm, which subsequently outbid them for that part. Another told of receiving a nine-page document of cost reduction ideas prepared by a competitor on projects bid on by both suppliers. This did not engender trust, of course; the recipient had also prepared a similar document and given it to the customer, and now assumes that their competitors have all those ideas. A powdered-metal process specialist, speaking of the same sort of thing, said, ‘Sometimes you think you’re getting something good when you get some inside information about some other supplier, but then you stop and think: ‘‘Well, if he’s giving you this information from this supplier, well, then he’s giving information from you to him.’’. . . It’s not like he’s just telling me everything he knows about this. I’m sure he’s telling other people too.’ Suppliers who fear that their customers will use process information to ratchet down their margins are careful to ‘muddy the waters’ as they pass on such information, and will at times even refuse OEM oVers of help because they fear its impact on price negotiations. At its most basic, they are careful not to give the sorts of cost information that might allow OEMs to benchmark supplier

104

Networks and Organizational Transformation

processes. As a supplier explained, they ‘supply [the customer] the info . . . [but] you can work the numbers. We like to show a 10–15 percent proWt. If we show more, we feel that would be suspect. We do need some extra margin to reduce in future years. We do parts where we almost lose on; overall, we do end up around 15 percent. You can adjust, you throw your burden up, you arrive at that 15 percent.’ Another said, ‘I can shuZe the numbers around and get whatever I want. Sometimes they call and say certain numbers are too high or low. I say, ‘‘Tell me what you want that number to be’’; numbers are measured in diVerent ways.’ Resistance can run much deeper than just the muddying of numbers and suppliers may even reject oVers of concrete assistance from their customers for fear of compromising information. For example, a supplier explained that he had refused to participate with a customer in creating a mutually valuable process map of his own plant, as the exercise would have involved his giving up proprietary information that could, in theory, be used to drive down his margins in bargaining. A high-level OEM supplier development manager, charged with helping suppliers to improve processes, described the diYculties he had with suppliers who had felt whipsawed by departments in his own company, and who were thus wary of sharing the information required to generate joint process improvements. He described the reaction his staV heard from some suppliers: ‘Hey, Mr. supplier development, your strategic sourcing was here last week and they broke my arm and got every red cent that I’ve got.’—Now that wasn’t their mission, but that is how it was interpreted—‘So now you guys are coming in here and, what do you think, that they left some on the table?’ Or: ‘We know what all of our processes are and we want to be in control, and we want to decide how to Wx them and what we will do with the beneWts of Wxing them. And by the way I still have your buddies from strategic sourcing busting my chops, so if this gives me the kitty without you folks being here, I’ll be able to decide for myself what strategic supply gets for annual cost reduction.’ They don’t want supplier development coming in and closing the loop on that, saying ‘You saved X percent; you’d better give it all to me.’ We hear that from suppliers a lot.

Suppliers are also wary of providing ideas to their customers, lest they simply be re-bid. For example, a supplier described an innovation they had developed (the ability to use a cheaper material on some of his product) and shared only with a customer with which they had a sole-source relationship. They were careful not to mention it to another that maintains multiple sources for all product families. Were he to share it, he believed, the idea would immediately be passed to his competitor and would provide him no advantage whatsoever.

Abuse of Trust and ‘Classic’ Hold-up Beyond the short-term misuse of private information, there are also many abuses of trust of a standard character that should not be ignored.

Uncertainty and Contradiction in the New Old Economy

105

For instance, one interviewee reported that a customer OEM had encouraged his Wrm to invest in additional production facilities to handle growing orders for a particular process, only to see new managers pull the most proWtable back jobs in-house, leaving the supplier with only the ‘nuisance work’. A tube bender was convinced by a buyer to buy raw steel wholesale rather than from distributors, and to pass on the cost savings; when the market fell oV, this process specialist was left with high inventory carrying costs even as the lower price was already locked in. The supplier ended up with a lower sale price and higher costs—and the buyer who had cut the deal had moved on to another position in the OEM, so the supplier did not expect future good faith for their action to be forthcoming. An interviewee at a foundry explained that taking on greater responsibilities had helped their survival: ‘We do the machining, we do the development work with them’—but, giving an example of ‘a product line that was [co-]developed Wve years ago’, he said: In spite of all that, . . . now the new buyer comes in to the company and looks at it and gets quotes from outside sources and says, ‘Hey, I can buy it 20 percent cheaper’, even though we helped co-develop it. That was with a guy Wve years ago and they have gone through three buyers in Wve years. There is not that sense of commitment that I see. Even though we have done all that, it is not buying us longevity.

Another problem is that modiWcations of product or process bargained ostensibly to be cost reductions are not always such; bad-faith or incompetent collaboration can lead them to become price reductions (lowering of margins). Many interviewed suppliers emphasize that cost reductions achieved may not match price reductions demanded. By claiming to have discovered ways to eliminate costs in the production process, OEMs try to convince suppliers to deliver price cuts before the savings are secured. Even suppliers skeptical of such claims can nonetheless feel compelled to deliver price cuts rather than challenge the customer’s interpretation. If the projected cost reductions then do not pan out as the part goes into production, they are ultimately Wnanced from the suppliers’ margins. Examples from suppliers illustrate what can happen. Speaking of a redesign they had come up with and oVered to their customer, the sales manager at a supplier specialized in small exhaust systems explained that they had come up with a ‘print change [that] was initiated to try to become cost-eVective again, after we gave them a price reduction in June’. But in November, they were being asked to give the OEM those cost savings although ‘they haven’t even gotten the approval on the print change. . . . So we are giving some money away on something that was designed to give us some money back after we gave them a price concession. And they thought it would be a real good idea, no reason [not to] go for that.’ (He added sarcastically, ‘I think if we just gave stuV away for free, it would still be too expensive.’) Such unilateral restructuring of terms is not exclusive to supplier-provided ideas, of course. Another supplier described a situation in which the customer

106

Networks and Organizational Transformation

informed his Wrm that the product would be redesigned to engineer out cost, and requested a 6.5 percent cost reduction in anticipation of the projected savings. The customer told him that failure to meet target price would lead the part to be bid to competitors. However, the supplier had not been included in the redesign process, and had not seen blueprints before the price reduction request. Upset, he told the buyer on the engineering team that he was uncomfortable agreeing to make a part he had never seen—since he was unsure that he could produce it in such a way as to achieve the projected savings. The buyer claimed that no other supplier had objected to this procedure, but accepted an alternative agreement whereby the supplier agreed to meet the target subject to review of the plans. The supplier met the requested price on the Wrst product redesign, but when the Wnal design arrived, it failed to deliver a part with the requisite cost savings. The supplier reported that the buyer told him that engineering was still ‘throwing prints over the wall’ to them, meaning that purchasing was being left in the dark as to the Wnal product design. The diVerent functions within the customer Wrm were not communicating in a way that would allow the supplier to provide information about how they could produce the part nor did it allow purchasing to mediate eVectively between engineering needs and supplier capabilities.

Shifting Unremunerated Risk and Cost to Suppliers An important feature of the emergent manufacturing model is the outsourcing not only of production functions and parts, but also of costs that had previously been assumed by the OEM. Of course, asking suppliers to provide new services is unproblematic—it is not just within the rules of the game, such devolution of responsibility is the game—and eliminating redundancies is a frequent and fruitful step in the cost reduction waltz. However, there are again many instances of disingenuous cost-shifting that are later denied in price bargaining. Although some supplier complaints regarding perceived cost-shifting can be understood simply as a strategic decision by the OEM to forgo collaboration, there are important examples, such as the interrelated issues of inventoryshifting and market forecasting, in which particular (short-term oriented) OEM practices transform a potentially positive-sum game into a negative-sum aVair that dumps risk and/or unremunerated costs on suppliers. A central principle of the new production paradigm is that the OEMs—onetime holders of large stocks of inventory with ensuing high carrying costs, risk of obsolescence and slow discovery of defects—now expect their suppliers to deliver exactly the right amount of goods on very short notice. OEM eVorts to reduce their own inventories are not supposed to push carrying costs down a level, but should remove stocks and work-in-progress from the system as ‘lean’ and specialized suppliers reduce cycle times and lot sizes. To help suppliers plan the use of their facilities, buy raw materials, and so on, OEMs typically also provide order forecasts that then become ‘hard’ orders at some Wxed time prior to delivery, after which the OEM is required to purchase supplier production

Uncertainty and Contradiction in the New Old Economy

107

(though not necessarily on the date forecast). This is in principle a positive-sum game: it is in everybody’s interest to eliminate inventory, but OEMs also need suppliers to have suYcient capacity when orders do materialize. In practice, these systems do not always work so well. First, the systems by which the OEMs transfer information between functional departments and then down to suppliers are relatively new, and have numerous kinks that cause suppliers to receive inaccurate information, forcing them to hold buVer inventory (since late delivery still reXects badly on the supplier). For instance, several suppliers blamed the new computer system at one interviewed OEM for this problem. OEMs also may inXexibly presuppose shorter cycle times for some suppliers than are feasible in the short and medium-term. Constraints inhibiting suppliers from suYciently reducing cycle times can include the need for operations at second tier suppliers, diYculties in obtaining specialized raw material, and the like. Second, as many OEMs move in the direction of ‘build-to-demand’ strategies and an almost constant revision of production schedules—with potentially large changes at quite short notice as new market information comes in—predicting order volumes can be diYcult.82 In and of itself, this is simply objective uncertainty, a systemic cost to be bargained over; the optimal solution is for the OEM to provide suppliers with the best information available. However, OEMs—or at least elements within them—have a short-term incentive to err on the high side (‘be optimistic’) in their forecasts, to ensure that suppliers have suYcient capacity to meet upswings, particularly for highly seasonal products where sales can be won and lost quickly by meeting the whims of the market more quickly than competitors. Suppliers see the prediction of orders in the future, do not realize that these are hoped-for ‘phantom’ orders, and either build inventory or maintain suYcient unused capacity. After consecutive years of signiWcant OEM underperformance relative to forecasts, one supplier angrily stated the problem succinctly and directly to his customer (at a supplier conference I attended): ‘Your reservation of our capacity bankrupts us.’ Another supplier to the same OEM complained that he had lost money in what should have been boom years because he had overhired in anticipation of an upswing that never came. Such misleading forecasting and other means of inventory shifting can create systemic problems. Many suppliers see the OEMs’ shift to JIT production as providing them with a competitive advantage vis-a`-vis foreign competition. They are right, so long as they are in fact able to remove inventory from the entire production system by reducing cycle times, and do not simply become a warehouse for OEM parts (‘just-in-case’ production). If they take the latter approach—which, as noted above, is sometimes necessitated by OEM practices—nobody gains. Thus, the obvious response to misleading OEM forecasts—common among suppliers interviewed—is to simply discount (i.e. if they are 25 percent high on average, simply remove that amount), which suppliers certainly do. A supplier to the OEM cited above for a particularly

108

Networks and Organizational Transformation

egregious under-build in recent years said, ‘I guess the anticipated build that never came was a little better [than it could have been]. Two years when they Wrst did it, it was terrible. We hired up people, we added equipment, and when the bubble didn’t come . . . . This time we just didn’t buy it.’ An OEM interviewee recognized that such forecasting undermines suppliers’ sense of his company’s reliability. He said that suppliers are saying, ‘Jesus Christ, you forced us to have the capacity to handle this, I bought capacity, I hired capacity, I manage capacity, and where are your damn orders? You are not there for me, and you are costing me money up the wazoo. And if you saved inventory anywhere, it was you guys and your Wnished inventory, because come down to my factory and we have inventory up to the ceiling!’ And then the suppliers pick up our annual report and they say, ‘Hey you guys made a lot of money, your Wnished goods inventory went way down. You guys saved money, great, but you still have my arm behind my back for Wve percent. What’s that partner thing? A share of proWts?’

The systemic risk generated by such conXicts is that suppliers are likely to discount the OEMs’ forecasts at varying rates, which is extremely problematic. The entire supply chain depends on sales by the OEM; given JIT production, if a single key supplier is signiWcantly late on deliveries because of insuYcient capacity in an upswing, the OEM and its entire supply base loses business too—not just the errant-forecasting supplier.

The Deceptive Use of Online Auctions As more of the purchasing process ‘goes paperless’, many OEMs are experimenting with online auctions. A request for quotes is made on a group of parts (or speciWcations for subcomponents requiring design, etc.). On its face, this is simply the migration of an existing process into the electronic arena. However, the real-time eVect of the online auction is sometimes used for its greater threat potential by bringing underqualiWed suppliers into the bidding pool. As a supplier primarily in automotive markets explained, OEMs will sometimes compare apples to oranges without telling bidding suppliers that is what is being done in an eVort to dupe incumbent suppliers into reducing their own bids. The OEMs, he said, allow suppliers using a cheaper technology into the auction, because ‘when you’re in that room and you have 30 minutes and every time someone puts a bid up there is a ding sound and psychologically it is very stressful and your Wrst reaction is that we must lower the [price] . . . and these guys know it. They are trying to drive the price down and they’re using this competing technology which is inferior to drive your pricing down.’ A metal stamper working primarily for mid-volume OEMs recounted the story of his discovery that some of the bidders in a particular online auction were, with the OEM’s knowledge, unlikely to be able to make the parts being quoted (this story was not an isolated case). The interviewee explained that they got to the point where

Uncertainty and Contradiction in the New Old Economy

109

we were only quoting against one other company—and I said ‘That’s crazy, we’re making it for practically nothing. We’re not going to go lower than that.’ And we dropped out. So then, we get a phone call, and he says, ‘Why’d you drop out?’ I said, ‘Well, we’re done.’ He said, ‘Well, you gotta quote on the tooling.’ And I said, ‘Well, why would I want to quote on the tooling when I just told you I didn’t even want to make the part?’ [The buyer replied:] ‘Well, you may make the part because we have to come out and qualify you, and maybe the other person won’t qualify.’ So we quoted on the tool just for the hell of it. Well, to make a long story short, they came out, because whoever was quoting on it wasn’t using the right material [as a basis for their bid].

The Big Question: Where Does Contradictory Collaboration Come From? Like Wve percent letters and other hard-bargaining tactics, online auctions are disliked by suppliers who generally think they are exclusively about driving down margins, ‘a real disaster for many, many people’ in the words of one interviewee. But, they can be useful and are by no means necessarily a deceitful tactic. ‘The bottom line of that whole story’, an OEM interviewee explained, ‘is the old adage of attitude: how do you look at it? The problem comes in when it’s not done right: a lot of incompetence, a lot of abuse. We have a set of rules, a guide for our buyers, etcetera. That way, it works. Before having this [new] guy on board, we made some missteps and screw-ups. None of it ended up being malicious or really destructive, it just wasn’t handled well.’ In a similar vein—that of the ‘screw-up’, but referring to private information—a supplier spoke of a period when a particular customer was ‘really pushing us to give the cost breakdowns, one of the [OEM] engineers—I am sure it was inadvertently—they gave us a complete cost breakdown from one of our competitors. We didn’t ask for it, but here it is, . . . emailed it to us. Well, if it is that easy to get this stuV, who is to say that our breakdown won’t show up in a competitor’s hands?’ With interviews so consistently peppered with stories of contradictory collaboration that force suppliers to hedge their own collaboration, an obvious question arises. How and why do these ‘mis-steps and screw-ups’ occur? The most straightforward of answers are perhaps tempting: contradictions might simply be written oV as the inevitable and interesting but ultimately random and systemically unimportant random consequence of the occasional incompetent rogue acting alone; or perhaps they are just an unmasking of OEMs’ ‘true’ shortterm interest, revealing that beneath the rhetoric the OEM–supplier relationship is but the most straightforward of prisoners’ dilemmas. Such write-oVs, however, are far too facile. They forgo real explanation and ignore that many of the stories in the previous section—such as those about departments ‘throwing prints over walls’ or of suppliers ascribing the loss of codeveloped work to turnover in the buyer position—are fundamentally stories of organizational screw-ups rooted in interactions between internal OEM departments and suppliers.

110

Networks and Organizational Transformation 5.4. CO N TR A DI C TORY C OL LA B OR ATI ON A ND THE N EW OLD O RGANIZATION

Recognizing that eVorts to build collaborative relationships are bedeviled by problems internal to OEMs recalls in the Wrst instance the post-Weberian synthesis in organizational sociology. As Robert Gibbons notes, countless empirical studies have built on ideas Wrst systematized in Cyert and March’s (1992) Behavioral Theory of the Firm and in Simon’s Administrative Behavior to show that organizations are rife with misaligned incentives, factionalism and misaligned goals that leave them altogether quite ‘messy’ and thus hardly ‘oblivious to the conditions that wreck markets’.83 However, it is not entirely clear just what the observation that Wrms are often quite messy means for today’s networked production systems. The postWeberian synthesis was built empirically and theoretically through the study of vertically integrated organizations with clear boundaries at a time when many more parts were made internally and Wrms generally purchased only those items for which product deWnition was easy enough that spot market relationships could be the norm. For the ‘twenty-Wrst century’ manufacturer, by contrast, so very much of what matters does not happen under roofs they own, or even control, but also cannot happen or even be understood independent of their ongoing coordination of numerous and dispersed actors. Both reacting to and causing this new reality, large manufacturing Wrms have been radically federating their structures and proliferating work groups that include personnel from multiple departments and even from other companies. This systematic blurring of organizational boundaries by one-time vertically integrated behemoths is in open deWance of the predictions of classical organizational theory, and provides the empirical ammunition for Sabel’s arguments that we are today engulfed in a (nonutopian) organizational revolution driven by the apparent inadequacies of market and hierarchy. This revolution (recall from Chapter 2), is premised on the posited proliferation of ‘nonstandard’ Wrms using new principles of industrial organization that break down distinctions between ‘mutually ignorant specialists, each tempted to exploit the ignorance of the other’. This encourages actors with diverse competencies to ‘teach each other important elements of their respective specialties and [to] reveal the logic of their intentions in order to make themselves comprehensible’ (Helper et al. 2000: 465). These practices culminate in a sort of forced openness that allows actors to jointly discover previously unknowable ways in which goals can be aligned, thus rendering moot the sorts of opportunism that classical organizational theory (especially the Williamsonian synthesis) posited as intrinsic to relationships both between and within organizations. The question we need to ask now is not whether Wrms are oblivious to the conditions that wreck markets; we know they are not. Rather, we need to understand why the solution—networks with their ‘new logic of organizing’ (Powell 2001: 35)—nonetheless so often falls prey to the conditions that

Uncertainty and Contradiction in the New Old Economy

111

wreck Wrms, as the evidence of bad waltzing presented in Section 5.2 suggests it does. I do not purport in this book to be able to fully answer this very large and as yet unanswered question in organizational theory. It is simply too big. I focus instead on those aspects of the answer that bear most directly on the prospects for reforming the American production model along collaborative lines. I begin by noting an interesting parallel. Stark (1996: 995) has observed that even in the very conscious building of capitalism in Eastern Europe, the emergent organizational forms are simultaneously new and continuous with what they have replaced, because ‘actors in the postsocialist context are rebuilding organizations and institutions not on the ruins but with the ruins of communism’. The ongoing transformation of American manufacturing is following a similar logic: the new economy is not being built on the old economy; rather, what we have is very much a new old economy, in the sense that the new organizational forms in manufacturing are being built with the pieces, practices and in the institutional context of the organizations that came before, not on them. Sabel’s Learning by Monitoring is both important and useful in its systematic theorization of a new logic of organizing to explain how vertically integrated behemoths of yesteryear can, and sometimes do, reconstruct themselves as network organizations that, in Powell’s famous formulation (1990; 2001: 35), are ‘neither market nor hierarchy’. But again, the functionality of a logic of organizing simply does not drive its inevitable diVusion. There are beneWts but there are also costs to interpreting today’s network organizations in manufacturing as a discontinuous shift in type. It is not exactly wrong (if one has reason to draw epochal distinction between periods) to characterize collaboration between Wrms and/or internal departments some years ago as more or less ‘organized informality’, and to contrast it to today when collaboration is often built around Xexible but formalized ‘rules, norms and procedures’ (Dyer 2000: 15) that give suppliers the predictability and reliability they need to make investments in new competencies. But drawing that hard epochal distinction suggests that insofar as one does observe the contemporaneous use of older organizational logics in Wrms that have restructured themselves using Learning by Monitoring practices, these older logics are essentially transitional noise— interesting perhaps, but in the big picture an almost irrelevant blip in an organizational revolution driven by the virtuous circle described at the end of Chapter 4 (reliability begets competency begets reliability). Yet as Kristensen and Zeitlin (2004) show in their book on the reform of the multinational corporation, it is far from obvious that learning by monitoring can so easily sweep away ingrained practices in large multidivisional corporations that remain rife with ‘opportunistic behavior on the part of managers at various levels seeking to advance their careers according to the preexisting ‘‘institutional logic’’ of the corporation’. Sabel’s approach as currently formulated, they explain, does ‘not appear to recognize that ‘‘mutually ignorant specialists’’ may already be engaged in a pre-existing game, which continues

112

Networks and Organizational Transformation

within and between Wrms despite deliberate eVorts to move towards pragmatic collaboration and learning by monitoring’. My own interviews with OEMs and suppliers actively seeking to establish collaborative relationships support Kristensen and Zeitlin’s Wndings. The twenty-Wrst-century manufacturing Wrm is a changed beast, but it is a beast made from, not in place of, the twentieth century Wrm. Notwithstanding very real changes in Wrm structure and a substantially increased reliance of OEMs on suppliers for incremental innovations in product and process, eVorts to establish collaborative relationships are nevertheless bedeviled by substantial internal organizational obstacles including especially the continued salience of departmental barriers and the failure to fully restructure the incentives of OEM actors to avoid short-term decision-making to the detriment of long-term organizational interests. These would by themselves lead to ambiguous signaling, but could perhaps be thought mere transitional bumps if they were not fed also by factional conXicts both within and between OEM departments over the degree to which it makes sense to invest in collaborative relationships with suppliers. The result is a systematic and understandable weakening of suppliers’ conWdence that OEMs can consistently and reliably follow through on their own collaborative strategies. 5.5. TH E O RG A N I ZATI O NAL RO OT S O F CO N T R A D IC TO RY COLLAB ORATION Helper, MacDuYe, and Sabel (2000: 443) persuasively argue that collaboration is most eVective when customers and suppliers consistently follow procedures that force both sides to share process and costing data in ways that allow them to ‘continuously improve their joint processes without the need for a clear division of property rights’. Nevertheless, even as OEMs are almost uniformly developing cross-functional teams to force a more open sharing of the information, interviewees at both OEMs and suppliers said that inconsistent incentives, departmental inWghting and factional disagreements over Wrm strategy mean that these procedures are not followed consistently enough to circumvent Wghts over ‘property rights’. Even suppliers with which they have generally collaborative relationships remain quite cautious about investing in new competencies and about opening their operations and books to customer scrutiny. To understand how and why this happens, it is useful to begin by looking at how relationships between suppliers and the ‘buyer’ role in OEMs have evolved with productive decentralization.84 This particular job bears primary responsibility for maintaining the supplier relationship, for coordinating engineering, and so on. It has grown in importance and changed considerably with the devolution of both production and innovation responsibilities to supplier Wrms and many OEMs have made determined eVorts to formalize relationships between their buyers and suppliers. They rank supplier performance using rating systems based on a mix of subjective and objective metrics, which are then made

Uncertainty and Contradiction in the New Old Economy

113

known to suppliers and personnel at the OEM. OEMs often rotate people through the buyer position relatively quickly because, a supplier explained, ‘They don’t want you to get familiar with them’, thus forcing suppliers to rely more on measurable competence trustworthiness than on informal ties as a means to secure new business. As the manager of a machine shop explained, they know that buyers ‘are going to turn. Typically we see the good ones get promoted, the poor ones turn over and leave the company. So again, it’s: we have to try to build our relationship based on what we can do, so that when a new buyer comes in, they can look at that history.’ Likewise, a heat treating and plating process specialist, when asked if his had strong personal relationships with buyers, said: ‘We have very few. We do have long-term relationships, but we’re working to push those up line to not be so vulnerable, that all of a sudden, Josh [using me as an ‘‘example’’ buyer] is at [OEM X] and now Josh decides to go to [OEM Y]. And . . . you know, Josh’s brother comes to [OEM X] and he says, ‘‘Screw it, I’m moving everything to Milwaukee’’; and boom, we’re out of business.’ The upshot, a component specialist explained, is that there is today much less relationship selling, [where] you’d show up and you’d wine and dine and you’d take them out to dinner. If you were a buddy and you had a good product and price was, eh, okay, you were relationship selling; they were buying because of you and your company. Now we’re getting down to where we’re earning every piece of business and there’s less relationship selling, less wining and dining. There’s more sharing of information, [it’s] more professional.

This formalization and its encouragement of competence trustworthiness, however, is by no means the end of the story. Increased responsibilities for the management of incremental innovation have not relieved buyers of their traditional role of Wnding ways to lower costs subject to minimum defect and delivery targets. This should in theory privilege the use of collaborative processes like the cost reduction waltz. But, as an OEM interviewee acknowledged, the incentives are often always so well aligned: ‘[Buyers] are rated on cost reduction, at the same time they are on all these new product design teams. That is where there is some disconnect: ‘‘What am I supposed to be working on? I am trying to integrate the supplier, but you are measuring me on current product that is going to be gone in a few years.’’ ’ He said, We put a lot of pressure on our strategic sourcing people and cost management people to come up with the Wve percent [cost reduction target] and . . . it is not our goal to go out there and take it out of margins, but at the end of the day, we don’t really care where it comes from. If you are a strategic sourcing person—I don’t mean to make that sound like a shot—you are under tremendous pressure. The goal is out there, get that goal.

Or as a supplier who said, ‘We are learning to understand it’ frankly described the dynamic from his end: what is ‘driving it [is] the buyers’ motivators: incent the behavior you want. You don’t reduce prices, you don’t get a raise, a bonus, nothing.’

114

Networks and Organizational Transformation

A combination of rapid turnover and the basing of promotions on the meeting of cost reduction targets creates strong temptations for buyers to jump from underbidding supplier to underbidding supplier, because problems that are discovered long enough down the road and hard to trace to a particular decision (i.e. low-margin driven underinvestment by a supplier) are not the problems of the people making that decision. For example, a machined parts supplier described with disgust a time when they ‘were actually the low cost provider on a very large package, several million dollars. We didn’t get the work, where the incumbent supplier said he had to have the work or he was out of business. And they actually chose to allow the guy to drop his price to keep it.’ This, he said, was a ‘very short term decision’ in the sense that it encouraged the competing supplier to reduce margins and thus to undercut his ability to invest in genuine productivity improvements. The basic problem, a die caster explained, is quite straightforward: buyers [are under] tremendous pressure to source globally and to allow unknown suppliers to take a shot at it. And, that’s problematic. At the end of the day, they might Wnd out that they don’t get parts on time, or that they don’t get good parts. Or that the other company didn’t really understand the design and it wasn’t really what they thought it was. But that doesn’t help me. That’s still business that I don’t have.

OEMs are aware that this is not good for anybody, and are thus almost universally developing teams that include members of multiple departments and perhaps even multiple organizations.85 However, these eVorts have been only partially successful, with suppliers nevertheless whipsawed by very mixed messages from often quite fragmentary OEMs. For example, a supplier of relatively complex components described a situation where they were at risk of losing a contract with a particular OEM. He said that ‘if you go to the plant, which we did, the plant people want to have nothing to do with our competitors, because they live with the short deliveries, the quality problems, everything else. [The manufacturing] people at the plant, they don’t give a damn about the price. . . .’ Yet, he said, ‘The purchasing people are the ones that are driving using this other supplier. So that’s something that we have to deal with and that’s very hard to understand, why customers that you feel you’ve been very loyal to will sell you out for a few pennies’ against their own rhetoric. Such conXicts can create problems not only for suppliers seeking to develop ‘win–win’ collaborative relationships as a protection against cost competition, but also for the OEMs themselves. For example, a supplier recounted a story of his buyer telling him his product was a ‘commodity. I can buy them from anywhere of ten places.’ He said that ‘all of the engineers, their faces went white, you know, when she said that, because they know how hard it is to Wnd a good-quality wire-harness supplier.’ Still, for a short period this customer went away . . . at the request of the strategic buyer . . . about a year ago, based on a very low, low bid by a competitor. And, I mean, I’m talking cost; we know our costs well enough that they were not making any money on it. They basically did not know what

Uncertainty and Contradiction in the New Old Economy

115

they were doing, or they were doing it as a loss leader to buy the rest of the business. Came back with price increases, got told no, didn’t deliver on time, held them to things that they were inXexible on. They mailed us for instance in January, wanted to push out a PO [purchase order] that they had cut, to the next month or the month after, or even later, because they had anticipated a boom year like last year, and it didn’t materialize in January. Orders got pushed out. They wanted to push orders out. That supplier refused to push out the orders [and said,] ‘No, we’ve cut ’em, we’re already halfway through building them. You’re taking them and you’re going to pay for them.’ Gave them inventory through June on something they did not want. They were so pissed they told them, ‘That’s it. We’re done with you, and we’re back with [my company].’ And they have since given us everything.

Supplier confusion caused by strife and miscommunication between departments is common, and was perhaps most amusingly described by a major component specialist who likened the relationship to their main customer (to which they are the single largest and most important supplier) to that of the child of dysfunctional parents, a pass-through of information that for various reasons OEM departments failed to share with each other. Supplier personnel regularly found themselves in the position of brokering agreement between the OEM’s purchasing, marketing and engineering departments on a design that could be built at a reasonable cost. Another supplier, a process specialist commented: We suVer that conXict every day. And big, big companies like to talk about team building and partnering, and those things. But the fact of the matter is that big companies tend to be very self-destructive. Engineering is charged with innovation, quality. Procurement is charged with buying the cheapest thing they can buy in the world. And manufacturing is charged with keeping the lines running without having any rejects. Well, those things don’t always work together. And we’re often caught in the crossWre. We’ll spend two years working with engineering on the development of the product which has a lot of features that are not readily apparent. And procurement will put it out on the Internet to bid it out all over the world, and some company from Bangladesh will say we can do it for half that price if you just make this one change, and procurement will say, ‘Great—I just saved a million dollars.’ Then it gets to the plant Xoor and the guy says, ‘Well how am I supposed to attach this screw to it?’ Well, guess what? You can’t. So there, we start all over.

DiVerences between departments are by no means the only source of mixed messages to suppliers, and in fact often reXect deep factional struggles and diVerences of opinion that only partly coincide with structural interests related to people’s departmental connections. For example, the component specialist described above as the ‘child of dysfunctional parents’ recounted a time that the President and CEO of their major customer came to their factory and said, ‘You know, some of the things that . . . we are building here may far exceed the customer’s expectations, so . . . all of these visual standards are fair game.’ Yet, he said, when you talk to the ‘marketing and styling group in that company’, they say that esthetic standards are in fact ‘what sells that product’. He

116

Networks and Organizational Transformation

said that ultimately, they are designing and developing ‘products with one group’, but forced to do so warily because there ‘seems to be a disconnect inside’ the OEM. Problems with factional conXict, a major automotive supplier emphasized, are an issue with all the customers. There is not a real consistency there. I’d like to think, because I [personally] work with [major assembler], and I think that we have a very good relationship with [them], that [they are] more prone to try to look at the [entire] value stream. But at the same time, they have some very short-term needs that have to be fulWlled. Otherwise it gets to be a fairly threatening situation about what might happen to your future. But at the same time, they keep saying that they are envisioning that they are going to work with you as a partner and treat you like a partner and these types of things and we—the optimist part of us—keep wondering when that will play out. But there may be another part of their organization over here because they have inconsistencies in their organization where they’re just . . . , you know. They don’t care about all of this other stuV, you know, just lower your price. So, that’s a challenge that they have and I think they recognize that and they are trying to bring in other people to convey a consistent message to the suppliers, but it’s tough.

Importantly, the ups and downs in these factional struggles can be quite opaque to suppliers. A machined parts supplier speaking of his time at a major midvolume OEM’s supplier advisory council said that what they ‘learned the most up there was about the struggles within the [OEM]; people change so fast. . . . The same suppliers were there, but change in buyers, in leadership, the focus was changed a couple, three times.’ Describing swings of this sort, another said that ‘[Our major customer] started swinging back the other way. [They were] really price sensitive, then they swung over here with the partnerships, and they were at least publicly saying [about us that we have] excellent quality, delivery, if the price is a little high, we will give them the resources to get that down.’ This OEM had even sent people to the supplier’s plants to help ‘jointly take some cost out of this product’, but now we are starting to see it going the other way. It is like two diVerent camps within [the OEM]. You have the people saying suppliers are precious; nothing is cheap if you don’t get it or it is all junk. Quality is worth paying a dime for. . . . And you get the people who would kill their moms for a nickel. I don’t know who is going to win out, but it is real frustrating for some of the supplier development people because they are out there preaching relationships and you have the other group with the club saying: ‘Lower your price or else.’ And you are saying: ‘Is this a partnership or not?’

The core issue was perhaps best summed up by the president of a small supplier Wrm. He explained that the ‘reality of partnering is that there’s politics in both organizations, there’s oxes to be gored and issues to be dealt with. Do we really want to sit down and solve the problems, or do we want to just talk and put window dressing on them and make our area look good?’

Uncertainty and Contradiction in the New Old Economy

117

5.6. CO N CLU D I NG PART II : TH E S YSTEM I C I MPL I CATIONS OF ORGANIZATIONAL UNCERTAINTY The obvious implication of endemic bad waltzing and contradictory collaboration is that suppliers hedge too—not just against market vagaries, but also against organizational uncertainties. This does not mean that there have not been signiWcant changes in the relational structure of American manufacturing. It is essential that this chapter’s stories of bad waltzing be read against the background of Chapters 3 and 4 with their many stories of OEMs and suppliers successfully waltzing as they jointly exploit promising new niches opened up by the decentralization of production in a world of radical uncertainty. Suppliers recognize that the game has fundamentally changed; for example, a die caster said that although there is this ‘disjointed process and the big companies are using it to their advantage, if the design department of my customer calls me and says, ‘‘We’re working on a new product and we’d like you to come in and get involved’’, I’d be foolish to say, ‘‘No, thank you.’’ ’ Or as a supplier of high-end industrial bearings explained, after saying that his Wrm’s design improvement suggestions to customers get shopped to competitors ‘all the time’: Your Wrst reaction is to want to be more cautious doing that, once you get quoteunquote ‘burned’ by one of those, you want to say ‘I’m not going to do that again.’ The reality is that if you don’t, you’re not in the market. It is part of business. You are building a partnership. There is a personal relationship there. There is still a certain amount of moral business ethics that says, ‘We came into this in all good faith to help you and we will provide you a solution, and if you choose to take that and work it against us, we have nothing to bind you that way, nothing more than a handshake. We are together on this thing we’ll help you, we’ll try to help you be successful at business.’ And you hope it works out. If it doesn’t, it doesn’t.

The claim that the ongoing transformation of American manufacturing is characterized by endemic bad waltzing and contradictory collaboration even in relatively collaborative relationships does, however, suggest that it is a very contested transformation. Even OEMs that have radically decentralized production and design and which unquestionably depend fundamentally on suppliers for incremental innovation still hedge their collaboration with those suppliers, sometimes strategically, and sometimes contradictorily due to internal diVerences over how best to proceed. Suppliers are cautious in turn, basing their own bets on how to waltz—the degree to which they themselves follow the new logic of organizing—on their guesses about how their customers are going to dance. And as this chapter’s many stories of contradictory collaboration make clear, this can be quite hard for suppliers to know with enough certainty to permit investments in new competencies and the open sharing of information. As one supplier asked (rhetorically): ‘That is the frustrating part, [OEM], if we are not competitive, tell us that. Don’t play games. . . . If we can’t be competitive, then I guess you need to go someplace else, but . . . .’ Similarly, a tiny Metal Spinning

118

Networks and Organizational Transformation

supplier in Milwaukee told of a time that he ‘had another metal spinner tell me that some people were quoting out some of my parts. You know, and so that was good to know when I was meeting with the OEM so I could negotiate properly.’ When asked why he was bothered by the OEM testing the market for better pricing on the parts he currently made, he Wrst responded, ‘Oh, because I’m a big baby!’, but he quickly went on to explain that he had ways of resisting—not sharing design ideas for example—but that he had to know if and when to use them. The fundamental dilemma was highlighted by the vice president of manufacturing at a relatively large supplier with the ability to produce modules and integrated subsystems. He said that the engineering departments of his customers remained somewhat reluctant to trust the competence of supplier engineering (in general, not just his company’s). This placed him in a diYcult position: his customers were telling him that they want to cull the supplier base by purchasing subassemblies and modules from existing suppliers, but when his Wrm pressed for such larger jobs to quote, nothing was forthcoming and they were not given the opportunity to ‘walk the line’ at the OEM to learn of and to suggest ways to integrate the components they currently build into subassemblies. As a result, they were left with expensive capabilities (large engineering staV and assembly) that matched customers’ purchasing strategies but not what he was ultimately asked to quote. This left them in a bad spot because ‘in our type of business, it is mostly 25-to-50-person shops with no overhead and pretty lean. . . . You know, for us, it is kind of a double-edged sword, having the full breadth of capability—it is good because it is one-stop, but on the other hand, we don’t have a focus either. If a shop is focusing strictly on machining or stamping, . . . we are competing against those people too.’ An automotive supplier speaking to the same issue underlined the point: ‘these small companies don’t have a very large engineering staV or an overhead. All they consist of is maybe a small assembly area with a machinist and skilled trades like that and they take the concept and build equipment. It’s diYcult for you to compete with someone like that if you have a large engineering staV to come up with the concepts and ideas.’86 The implication, a mid-volume metal fabricator explained, is that being able to meet the rapidly changing demands of customers and to help in product design is, paradoxically, ‘a double-edged sword. The more our customer pushes on us, it increases our overhead and sometimes, it comes to a point that you get challenged for your pricing and they start to look at other suppliers. [Our key OEM customer] is doing that now for benchmarking purposes. We add a lot of value to the supply chain by adding new engineers. I have eight engineers who work on my team alone’, but they are concerned, because ‘when you get squeezed, it is like, where are you going to get the money to reinvest?’ These pincers have consequences for relationship building as well: the plant manager at a mid-sized machine shop that had been forced to reduce white-collar staV (though still retaining some engineering capabilities) explained, ‘Guys [learn]

Uncertainty and Contradiction in the New Old Economy

119

to do without it, you cope, but things don’t happen when that happens. When a customer wants us to go spend three days at a supplier meeting, who do I send?’ In short, even suppliers anxious for collaboration as a protection against excessive price competition are wary of OEM declarations of ‘partnership’. For example, a wireform supplier remarked that he withheld costing information from a customer to get a ‘little leverage to keep the market here. And, is that being a partner? I don’t know.’ Faced with bad waltzing and organizational uncertainty, real changes at OEMs not withstanding, suppliers have good reason for caution in the sharing of process information and of investing in new product development on the promise of business down the line. Yet this gives OEMs greater reason to treat particular subcomponents as commodities and emboldens marketeer factions within OEMs who believe that the best way to lower costs is jump from underbidding supplier to underbidding supplier to take advantage of excess capacity in global markets. Linking these pressures back to the already weaker average productivity at small suppliers, the long-term implications are distressing. As low-overhead competitors—whether in the United States or abroad—undermine those looking to upgrade, suppliers asked by customers to reduce prices or to take on responsibility for additional services may simply reduce their already weak margins rather than Wnancing them through parallel cost reductions or process improvements. This in turn means less capital available for strategic plans, R&D, technology development, capital investments and worker training, placing many Wrms at risk of being mired in an ultimately self-destructive vicious cycle.87 And though there are limits to this vicious cycle, both because the decentralization of production does create enough market for supplier innovation to allow some investment, and because collaborationists within OEMs do sometimes win, the picture of that emerges American manufacturing going forward is not so pretty as one might hope. It is one of continued and strategic hedging by Wrms buVeted by the vagaries of technological and market dynamics, the uncertain returns of joint innovation, and the failure of production networks to adequately sidestep the conditions that wreck Wrms. What we see in American manufacturing is an altogether partial construction of the normatively attractive new production paradigm outlined in the Wrst part of this book. However, this partiality is also a story of ongoing and systematic contradictions that sits uneasily with sociological literatures on network organizational forms in manufacturing. It is certainly not a clean-but-uneven pattern of collaborative alliance capitalism, with insider and outsider positions explained by the oft-times inhospitable cultural and institutional conditions in the United States. It also does not fully square with the notion of an ongoing organizational revolution driven by the federation of production and experimentation with functionally superior Japanese-style customer–supplier relations in global markets competitive enough to deny oligopoly rents.88 Even Wrms’ good-faith eVorts to cooperate are often subverted by a complex interaction between pre-existing weaknesses in the supplier base and intraorganizational conXicts. Dismissing

120

Networks and Organizational Transformation

these systematic contradictions as interesting but ultimately unimportant leads to a fundamental misunderstanding of the reorganization of American manufacturing and the prospects for its reform (which are not as bleak as this chapter might suggest; there are ways in which these same contradictions can be leveraged to support existing partial collaboration, as I show in Part III).

Part III Institutions and the Relational Reconstruction of Regional Political Economy

This page intentionally left blank

Introduction to Part III Recall from Chapter 2 that Hall and Soskice’s conclusions rely heavily on the observation that incremental innovation industries—like metal manufacturing—are substantially favored by ‘institutions that reduce the uncertainty actors have about the behavior of others and allow them to make credible commitments to each other’ and that ‘provide actors potentially able to cooperate with one another with a capacity for deliberation’. Such institutions matter because ‘deliberative proceedings in which the participants engage in extensive sharing of information about their interests and beliefs can improve the conWdence of each in the strategies likely to be taken by the others’ (Hall and Soskice 2001: 10–11). The problem—as far as durable manufacturing is concerned—is that the United States is not blessed with such an institutional infrastructure. Firm strategies that rely on nonmarket modes of coordination and ‘extensive relationship building’ are unlikely to be very successful in an American context bereft of the sorts of ‘dense business networks or associations that circulate reputations for reliability or sharp practice quickly and widely’ and that are generally found in North European and Japanese coordinated market economies (Hall and Soskice 2001: 38–9). Importantly, in making this argument, Hall and Soskice (2001: 21) are in no way arguing that institutional constraints on incremental innovation caused by liberal institutional frameworks represent limitations for the economy at large. Liberal Market Economies (LMEs) have their own ‘comparative institutional advantage’ in radical innovation, as evidenced by the growth of Silicon Valley and similar high-technology industries in the United States. Firms in these industries, they argue, beneWt from their greater freedom—relative to Coordinated Market Economies (CMEs)—to coordinate economic activity in the market. Hall and Soskice explicitly note that, despite their sectoral implications, CMEs and LMEs have both shown themselves to be ‘capable of providing satisfactory levels of long-run economic performance’. Nonetheless, Hall and Soskice do not divide national economies into these two varieties of capitalism merely for purposes of classiWcation, to say, ‘both are Wne’, but hold that the polar typology bears clear lessons for policymakers. They write that the goal of sound economic development is to induce ‘economic actors to cooperate more eVectively with each other’ because the ‘economy is an arena in which multiple actors develop competencies by devising better ways of coordinating their endeavors with one another. . . . Accordingly, one of the principle ways in which policymakers can improve national economic

124

Institutions and Relational Reconstruction

performance is to secure better forms of coordination among private-sector actors.’ In principle, this can be in either market or nonmarket arenas, but in practice, ‘economic policies will be eVective only if they are incentive compatible, namely, complementary to the incentive structures and coordinating capacities embedded in the existing political economy’ (Hall and Soskice 2001: 46). Hall and Soskice (2001: 47) distinguish between ‘ ‘‘market incentive’’ policies’ that use ‘incentives to induce actors to perform more eVectively’ and ‘ ‘‘coordination-oriented’’ policies’ that ‘attempt to improve the competencies of Wrms, such as their skill levels or technological capabilities, by addressing Wrm needs with relative precision’. The latter type of policies require access to ‘high levels of information about the activities of the Wrm’ that Wrms ‘are reluctant to share . . . with governments whose position as powerful actors under a range of unpredictable inXuences raises the risks that they will defect from any agreement and use the information they have acquired against the Wrm’. Two issues of incentive compatibility follow: Wrst, governments must have conWdence that established producer groups are in a position to eVectively administer coordinationoriented policies; and second, ‘business associations and their members’ will enter into ‘implicit contracts’ that require information sharing only if ‘the government’s commitment to abide by [the implicit contracts] is credible’ (Hall and Soskice 2001: 48). It follows that governments in CMEs should utilize a very diVerent palate of policy instruments than do governments in LMEs. In CMEs, governments can and do work together with independent business associations and strong trade unions in the administration of policies that favor asset speciWc investments by companies and other forms of collaborative nonmarket coordination. These policies are viable because CME governments can work with established producer groups that have good information on member needs and that can sanction those of their members who defect. Producer groups also have an incentive to cooperate with the governments insofar as their established role in the policy structure puts them in a better position to ‘mobilize a serious constituency’ to sanction governments that themselves deviate from agreements. As a result, Hall and Soskice (2001: 48) write, ‘coordination-oriented policies should be more feasible in nations with both a coordinated market economy and a political system in which producer groups enjoy substantial structural inXuence.’ These same policies, by contrast, should be relatively ineVective in LMEs because, Soskice (1999: 128) explains, ‘eVective business coordinating capacity cannot generally be built ‘‘spontaneously’’ to service an institutional framework’ in the CME style: it would require that companies already ‘be engaged in longterm relational contracts’ that can sustain ‘common shared understandings’ and the ‘creation of expert communities across associations, research institutions, and companies’ that ‘can only take place over long periods’. The eVectiveness of coordination-oriented policies is further damaged by the relative inability of disorganized producer groups in LMEs to exercise structural inXuence vis-a`-vis the state.89 By way of example, Hall and Soskice (2001: 48) draw on work by

Introduction to Part III

125

Culpepper (1998) and Levy (1999) to argue that French policymakers have ‘had diYculty implementing schemes for regional or technological development that require coordination among private sector actors, partly because it concentrates power in Paris and cannot Wnd encompassing producer groups to operate them.’ The upshot is that LMEs—including the United States—cannot be expected to eVectively hybridize CME-style institutions. Instead: In general, liberal market economies should Wnd it more feasible to implement market-incentive policies that do not put extensive demands on Wrms to form relational contracts with others but rely on markets to coordinate their activities. These include regional development schemes based on tax incentives, vocational programs focused on formal instruction in marketable skills, and government subsidies for basic research. Because of the bluntness of the instruments available to states and the importance of markets to these economies, deregulation of various kinds is often the most eVective way to improve coordination in LMEs. (Hall and Soskice 2001: 49)

Certainly, the evidence presented in Part II—showing that there are systematic failures of cooperation between OEMs and suppliers in the American Upper Midwest—is consistent with some of Hall and Soskice’s claims regarding the eVect that historical–institutional factors can have on Wrms’ strategic interactions and their relative propensity to seek cooperative relationships with each other. But at the same time, the documentation of partial collaboration and active eVorts by some American OEMs and suppliers to build long-term relationships is suggestive of the possibility of encouraging more. It gives reason to doubt that the equilibrating forces of hard complementarities really mean that state eVorts to create nonmarket institutions in the service of interWrm collaboration are not only unfeasible but perhaps even counterproductive. And if such policies can be implemented without damaging those aspects of the institutional structure that favor the country’s already advantaged radical innovation industries, there is no reason that they should not be. Their implementation ward be all beneWt and no cost. Here, much follows from the empirical claims established in Part II, where I show that the strategies American manufacturing Wrms use to navigate a combination of market and technological competence and organizational uncertainties cause the modal case of the vertical OEM–supplier relationship in American manufacturing to range systematically between the collaborative and arm’slength poles. There are real eVorts to build a more collaborative production model in the United States, but because they are so partial and so often blocked, they do not in fact amount to the organizational revolution posited by Sabel and collaborators. Rather, it seems clear that any substantial generalization of the new production paradigm outlined in Chapter 1 requires the building of external institutions to help Wrms sustain collaborative relationships. Sabel (2004a: 85) has recognized that cooperative ‘customer–supplier relations are not self-stabilizing and therefore in need of second-order stabilization through governance’, arguing that such institutions should themselves follow

126

Institutions and Relational Reconstruction

Learning by Monitoring principles, but he is silent on how such institutions must themselves be constructed, leaving ‘for another time discussion of the morphology of and methods for constructing the networks presumed by the organizational revolution’ (Sabel 2004a: 21). However, because eVorts to establish collaborative network relationships are so systematically blocked in ways that seem far more stable than transitional, questions of morphology—both institutional and organizational—cannot be left for another time. There is a need to understand and to disaggregate the multiple factors that determine the relative collaborativeness of the mix of relationships, so that we might identify hooks and points of leverage that can aid the construction of institutions to push those relationships in socially optimal directions. The argument I want to develop here in Part III, brieXy stated, is that there are elements in American manufacturing that represent the sorts of ‘incongruities, incoherence, and within-system diversities’ that Crouch and Farrell (2004: 33) show analytically to give actors tools they can use to escape the traps of path dependency. Two ‘incongruities’ are particularly relevant. First, as shown in Part II, there is an existing—if limited and sometimes latent—relational structure in American manufacturing; there are important collaborationist factions that could be mobilized to ensure the eVective functioning of ‘coordination-oriented’ policies that would require the building of new institutions. Second, the extreme fragmentation of the American federal structure in the area of economic development policy creates enormous space for local and regional experimentation.90 Notably, this second ‘incongruity’ again contrasts a key tenet of the VoC position—that the national-speciWc character of ‘so many of the institutional factors conditioning the behavior of Wrms’ means that variation at the regional level is largely insigniWcant (Hall and Soskice 2001: 16). Treating regional institutions as peripheral too-quickly dismisses viable options hidden in interstices and inconsistencies in an American national institutional framework that is not nearly so coherent as theory would have it be. In the context of American federalism, the state level oVers the best chance for ‘collaborationist’ producer groups desirous of institutional change to exercise ‘structural inXuence’ to ensure the continuity of eVorts to hybridize CME-style institutions. Although these two incongruities are suggestive, they are but tools. Recall Granovetter’s historically rooted point (1992) Wrst cited in Chapter 2: ‘Economic institutions do not emerge automatically in response to economic needs’ but must be actively constructed. There is reason to believe it possible—and good— to create nonmarket sustaining institutions for interWrm collaboration in LMEs; but nothing in my arguments suggests that their construction is inevitable (nor that it is easy). Nevertheless, if the space I claim for the construction of nonmarket coordinating institutions goes unused in fact, a major implication of the empirical work presented in Part II manages the dubious feat of being theoretically important yet practically trivial. Chapter 6 thus turns more concretely to questions of regional level policymaking and institution-building to show that it is feasible

Introduction to Part III

127

to build coordinating institutions at the state-level in the United States, showing that Hall and Soskice are indeed too pessimistic in their assessment of the possibilities of incremental institutional change absent the ‘right’ historic conditions. The core limitations on ‘coordination-oriented’ policymaking in the United States are far more political than they are historical–institutional.

This page intentionally left blank

6 It Couldn’t Happen Here? Public Policy, Regional Institutions, and InterWrm Collaboration in the United States 6.1. I N TRODUCTI O N In this chapter, I argue that the Wisconsin Manufacturers’ Development Consortium (WMDC)—a consortium of seven OEMs that formed in 1998 to work jointly with the state’s manufacturing modernization service to provide training to suppliers—is suggestive of the sorts of public–private institution building that can both enhance supplier performance in the United States and proactively encourage greater collaboration between OEMs and their suppliers. Without overplaying the importance of this policy experiment—which to date remains a limited and ever tenous solution—I argue that its structure and evolution show that it is both possible and useful to leverage and strengthen existing partial collaboration between OEMs and suppliers through the construction of CMEstyle institutions premised on substantial business coordinating capacity even in the prototypical American LME. My focus on the possibilities of institution-building at the subnational level does not mean that decisions taken at a national level regarding trade, tax, and monetary policies do not tremendously aVect the fates of American manufacturers; the relative value of currencies, the cost of capital, and tariV levels are strongly aVected by decisions and policies made at the national level and are of obvious importance to manufacturing Wrms that must compete in global markets. There are, however, still important policy arenas where decisions taken at the subnational level substantially aVect strategic interactions between Wrms, most particularly in what is usually called ‘industrial policy’. There, as Eberts and Erickcek (2002: 21) note (paraphrasing a remark by the former Secretary of Labor John Dunlop), the United States has no real national policy but ‘pursues 5,000 industrial policies, as each state and local municipality pursues policies to promote their own economic development’. Or as local economic development expert Timothy Bartik (2003a: 2) argues, even federal eVorts to ‘promote the economic development of advanced US manufacturing’ should be concentrated at the subnational level: not only has ‘the aggressive promotion of economic development in manufacturing . . . primarily been a state and local role during the post-World War II era’, it should remain there because so ‘many of the key

130

Institutions and Relational Reconstruction

inputs for advanced manufacturing development are provided locally’ (e.g. education of workers through universities or community colleges). I thus focus particularly on what can be done at the state level (without ignoring national-level constraints) to encourage the diVusion of the sorts of collaborative organizational practices that can allow Wrms to successfully compete in demanding and quality-conscious product markets that require almost continuous incremental innovation (that is, in the markets most likely to support the relatively higher wages paid in the developed world). In looking to the possibility of systemic changes that begin with incremental state-level policy experimentation, my arguments Wnd an important parallel in the related ‘sphere’ of American welfare and labor market policy. There, Zeitlin (2003: 6) notes, ‘the most innovative initiatives toward building new systems of welfare provision and labor-market governance have come from state and locallevel actors in response to devolution, privatization, and regulatory gaps in federal policies and programs.’ The next section describes these parallel developments, drawing on work by Dresser and Rogers (2003) and Osterman (2003) to show that there is space in the interstices of the American LME for the construction of nonmarket coordinating institutions in that particular institutional ‘sphere’.91 The remainder of the chapter then shows that the regional concentration of American manufacturing and the highly decentralized American industrial policy environment make it both feasible and desirable to (incrementally) build new institutions to help sustain existing but partial collaboration in Hall and Soskice’s sphere of interWrm relations. 6.2. PA RA LL EL DEVE LO PM EN TS : WO RK FO RCE INTERMEDIARIES AND COORDINATION I N T HE AMERICAN LAB OR MARKET Osterman (2003: 240) writes that ‘in the last two decades as the American labor market has radically changed, public policy has been very slow to catch up’. Dresser and Rogers (2003: 266) concur, explaining that most new policies have created only lacunae, a negative Xexibility that brings ‘little general beneWt’ with social costs that are ‘exceedingly high’. However, Dresser and Rogers (2003: 284–5) add, ‘of greater and more positive interest may be some of the institutional experimentation that has followed from such negative labor-market restructuring’. They refer speciWcally to the wide ‘variety of new ‘‘workforce intermediaries’’—serving both workers and employers, and sometimes performing functions previously assigned the state—that have become more evident in recent years’. These new institutions ‘are helping to forge local solutions to labormarket problems. By targeting multiple employers and building workforce development solutions for them, these programs have begun to show real promise as a means of improving outcomes for workers and the unemployed.’ The examples cited are myriad, and include: the Industrial Areas Foundations around San Antonio, the Paraprofessional Healthcare Institute and Cooperative

It Couldn’t Happen Here?

131

Home Care Associates in the South Bronx; the San Francisco Hotels Partnership; the Culinary and Hospitality Academy of Las Vegas; the Hospital and Health Care Workers Union Training and Upgrading Fund in Philadelphia; the Wisconsin Regional Training Partnership; and the Jobs with a Future Initiative in Dane County, WI. These many examples, Dresser and Rogers (2003: 267) write, suggest that ‘in the limiting case’, new workforce intermediaries ‘appear to oVer, under the radically changed labor-market conditions of the present, some promise of reinserting a worker voice throughout the micro-institutional choices involved in restructuring—perhaps even to the point of making it again serve the general welfare.’ This Wnding notably contrasts a key element of the VoC position: Hall and Soskice’s argument that institutional complementarities are such that policy in each of the ‘Wve spheres’ in which Wrms must resolve coordination problems—including vocational training and education—should be ‘incentive compatible’ with the national LME institutional framework. Without disputing that it may be diYcult, Dresser and Rogers and Osterman show that it is neither unfeasible nor undesirable to build mechanisms for formal nonmarket coordination in the United States. However, Dresser and Rogers (2003: 284–5, italics added) caution that what they describe is very much a ‘resistible rise of workforce intermediaries’ that have been consciously built, often with considerable struggle, against an ‘infrastructure supporting, informing, and extending these local eVorts [that] is fairly weak’, and that has grown in many instances only because ‘the organizations themselves have helped to build it’. Likewise Osterman’s discussion (2003: 251) of the Industrial Areas Foundation describes not only the changing structure of the labor market as creating space for institution building, but recognizes that Wlling that space was not easy and owed much to particular actors—especially Saul Alinsky—who created this ‘organization of organizations’. Despite these constraints, both Dresser and Rogers and Osterman are emphatic that workforce intermediaries can form the basis for systemic change—a ‘new sort of ‘‘American model’’ in training’—by building on the incipient provision at the regional level [of] what is not provided nationally—a genuine infrastructure of industry and union collaboration that both drives industries toward more demanding skill demands and provides the Xow of information, and assurances against freeriding, needed to meet them. Given pressures for devolution, moreover, there is no reason why such eVorts could not be more eVectively integrated into public labor-market administration.

The workforce intermediary most relevant to this book is the Wisconsin Regional Training Partnership (WRTP). The WRTP is a union–management cooperation initiative founded in 1992, with a 2003 membership of greater than 100 unionized Wrms collectively employing around 60,000 workers. Along with the signature issues of incumbent worker training and future workforce development, the WRTP, Dresser and Rogers (2003: 281) explain,

132

Institutions and Relational Reconstruction

‘focuses considerable attention on modernization in member Wrms’ because often ‘new technology and new work organization go hand-in-hand’. This is done by ‘collaborating closely with the state MEP [manufacturing extension partnership; discussed below] to ensure that member Wrms have access to its resources and that those resources will help to serve member Wrms’ because so many of the organization’s smaller Wrms ‘do not have suYcient resources to commit to modernizing their own Wrms’. To date, however, coordination between WRTP member Wrms has been largely horizontal. A series of meetings among OEM union representatives to discuss responses to outsourcing and ways of helping to direct work towards union suppliers has certainly touched on the vertical OEM–supplier relationship, but the WRTP has not yet undertaken substantial activity to stimulate the vertical coordination of activities between OEMs and their suppliers. Nevertheless, the ‘resistible rise of workforce intermediaries’ such as the WRTP in the spheres of industrial relations and of vocational training and education show it possible to build institutions for nonmarket coordination even in the prototypical American LME. It thus helps to ground a counterclaim to the notion that institutional complementarities necessarily render hybridization unfeasible or unpalatable. If there is space for deviation from type in the institutional sphere of vocational training and education, why not also in the sphere of interWrm relations, especially if government need not stimulate interWrm collaboration ex novo, but can instead direct and improve an existing— if limited—collaborative relational structure (in part by organizing supportive factions within Wrms)? 6.3. S TAT E AN D LO CA L I NDUS TRI A L PO LI CY I N A WOR LD O F D E C E N T R A L I Z E D P RO D U C T I O N The extreme decentralization of American economic development policy creates space for local level experimentation, but it can—and does—also put states and localities into ‘zero-sum’ competition using Wnancial and tax incentives to try to inXuence Wrm location decisions. As Erickcek (2001: 1) observes, ‘practitioners argue that tax and Wnance incentives are crucial as states battle for that next plant opening’, despite the insistence of ‘most researchers . . . that smart economic development would avoid the use of tax incentives and state grants to lure business across state lines’ (see also Wolman and Spitzley 1996). This strategy, however, has become ever more expensive as more states and localities play the same game: in 1980, jobs in Tennessee’s new Nissan plant went for a mere $11,000 each; by 1993, Alabama’s Mercedes plant built in the early 1990s cost the state $168,000 per job (Bartik 1995).92 In addition to having become extremely expensive, zero-sum economic development strategies make ever less sense in a world of decentralized production, since a majority of the jobs associated with a particular manufacturing plant are not located in that plant. Productive decentralization means that most of the jobs

It Couldn’t Happen Here?

133

‘created’ by a particular OEM are found in the external supplier Wrms that carry out key production processes, so these jobs can be at some considerable distance from the OEM in question. For example, an OEM interviewee in Wisconsin explained that his company had been enticed in part by incentives to build a new plant in the southeastern United States. The plant they built, however, did only assembly and thus some 80 percent of the ‘cost-of-goods-sold’ was purchased material (that is, only 20 percent of the value of the product was added at the plant). Furthermore, though the OEM could Wnd some competitively priced local suppliers, they continued to purchase substantial portions of their inputs from suppliers near their historic Wisconsin plant with which they continued to maintain collaborative relationships and which gave them, he said, ‘more than just parts’. Indeed, fourteen years after its opening, the southern plant continued to purchase 40 percent of its materials from suppliers based in Wisconsin. From a policy standpoint, the Wnancial consequences should not be underestimated: even though the spin-oV factory was located in the southeastern United States rather than Wisconsin, the factory spends about $30 million more per year in Wisconsin than it does in its new home (and thus almost certainly provides more jobs in Wisconsin than it does in the southeast). This is also not an isolated case. Another interviewed OEM had also opened a plant (with incentives) in a southern state and seconded the point that ties to historic suppliers made that plant less valuable to its new home than to its old one (which is not to say that Wisconsin would not have appreciated having both the OEMs and the suppliers; but it is the tax incentives that are the deadweight loss). That OEM opened its new southern plant in 1997 and reported that in 2001 purchased content was 75 percent of the cost-of-goods sold, or about $44 million annually. Of that 44 million, $16.7 million (28.5 percent) was purchased from Wisconsin suppliers, greater than the estimated $12.6 million (25 percent) ‘created’ in the southern state due to the location of the OEM’s plant. The decentralization of production thus means that landing a branch plant does not have as large an impact on a state’s economy as it would have in years past, whereas eVorts to maintain the competencies of existing component manufacturers can pay oV more than they used to. Many policymakers do of course recognize this, and—although simple tax and Wnancial incentives are still extremely prevalent in economic development—public authorities at various territorial levels are also experimenting with policies intended to ease the transition to a more decentralized production regime in ‘positive-sum’ ways that increase the overall economic activity of the region in question rather than simply moving production from place to place. Often using the rhetoric of ‘cluster’ development popularized by Porter (see especially Porter 1998; 2000), these strategies vary widely. However, they have in common a belief that economic development requires linking the many entities that create regional competitiveness, crucially including suppliers of specialized inputs and infrastructure, governmental and other institutions that provide needed training and technical support, as well as trade associations and unions.93

134

Institutions and Relational Reconstruction

Beyond just an emphasis on services—such as training or infrastructure—to serve sectors that already have ‘critical mass’, there is wide variation in what this actually means on the ground. Besides the simple provision of an educational infrastructure, such as the technical or community college system, there are a variety of modes through which government resources can be and are used to upgrade the base of small and medium sized suppliers. For example, the ‘Industrial Training Program’ in the state of Illinois disburses approximately $12 million directly to manufacturing Wrms for training.94 Approximately 15 percent of this money goes to a competitive grant program in which individual Wrms apply for money, while the remainder goes to ‘multicompany’ programs that supply training to many diVerent manufacturing Wrms at 50 percent subsidized rates. The multicompany training programs can be administered either ‘horizontally’—grants are given to associations that subsidize training for their members, often at local community colleges or from private training providers—or ‘vertically’—with money provided to single OEMs who then use it to subsidize relatively unrestricted training programs for their own suppliers. Note, however, that both of these modes Wt quite easily with Hall and Soskice’s notions of ‘incentive compatibility’ in LME policy: they are aimed entirely at the improvement of supplier performance per se—a useful but certainly not suYcient condition for successful collaboration. The federal government also contributes to the upgrading of the American supply base, most notably through its sponsorship of the Manufacturing Extension Partnership (MEP) housed in the National Institute of Standards and Technology (NIST). Although a few industrial extension programs began as early as the 1950s, the Clinton administration Wrst made them into an important part of the American industrial policy by providing substantial—though partial—federal funding for a series of locally based not-for-proWt manufacturing modernization centers to provide technical assistance to small and mediumsized manufacturers. This led to the formation of many new MEPs in the 1990s, and currently all Wfty states are served with centers that have consistently been shown in studies to signiWcantly improve the productivity of the Wrms they serve (Bartik 2003a; NIST 2003; Shapira 1998). The MEP, though certainly a nationally coordinated program, mixes resources from the public, private, state, and federal levels, and is ultimately quite consistent with the general decentralized tenor of American industrial policy for two reasons.95 First, the centers are very explicitly partnerships, have substantial local autonomy, and must raise a majority of their funds from other sources, usually a combination of state government subsidy and the sale of services to Wrms (which are then partially matched by NIST). Furthermore, even this partial federal commitment has been somewhat tenuous. The program suVered its Wrst near death experience after the 1994 elections, averted after congressional visits by MEP assisted Wrms, and suVered a two-thirds funding cut (from $106 million to $39.6 million) in the 2004 budget, leading to substantial cuts at many of the

It Couldn’t Happen Here?

135

centers. That budget cut did, however, lead to a substantial public outcry. The program is relatively popular on Capitol Hill and has been the rhetorical beneWciary of many of the calls for the federal government to do something for manufacturing competitiveness in the wake of the job losses between 2000 and 2003. The network ultimately had its budget restored to a reasonable $109 million for the 2005 Wscal year. Second, particularly relevant for the argument of this book, the MEP per se Wts quite easily with Hall and Soskice’s descriptions of the sorts of ‘blunt’ policies that are feasible in LMEs bereft of business coordinating capacity. As Turner’s dissertation (1999: 10) exploring three diVerent strategies of manufacturing extension explains, there is variation in MEP models—he compares the business consultancy, associative, and entrepreneurial state models—but ‘the overwhelming majority of state centers’ simply provide a discounted business consultancy by selling oV-the-shelf technologies at subsidized rates on a Wrstcome-Wrst-served basis, with a ‘persistent theme across all three programs’ he studied being a ‘diYculty in promoting increased cooperation among businesses and public sector programs’.96 However, Turner (1999: 218–19, 222) stresses as well that the extremely decentralized nature of the MEP also oVers novel opportunities: ‘states and localities’ are given ‘the responsibility for designing centers to address regional issues’ resulting in what is—at its best—an extremely Xexible program that reXects the ‘trappings of a new division of labor between the national government and states regarding economic development’ by providing ample space for variant strategies and local experimentation. The WMDC—the subject of the remainder of this chapter—is a result of just such experimentation. It relies heavily on the Wisconsin MEP (WMEP), which, in combination with the state’s technical college system, represents the ‘public’ in this public–private eVort to upgrade the Wisconsin supply base. But it adds a novel wrinkle that straightly breaks with traditional liberal market industrial policy: it relies for its governance on substantial business coordinating capacity Wnalized towards the nonmarket coordination of strategic interactions and improved collaboration between OEMs and their suppliers. 6.4. T HE WI S CO N S I N M A N U FAC TU R E R S ’ D EV E LO P M E N T C O N S O RT I U M Originally called the Wisconsin Supplier Training Consortium, the WMDC began as a joint eVort between the WMEP and John Deere. Its foundation owes much to the vision of Paul Ericksen, a committed ‘collaborationist’ at the John Deere Horicon Works who also serves on the WMEP board of directors. In this dual role, Ericksen saw not only the growing importance of suppliers to his company’s own manufacturing activities but also the signiWcance of OEM purchasing strategy to Wisconsin’s economy. Deere already had a stand-alone supplier training program in Illinois subsidized by the Illinois Industrial

136

Institutions and Relational Reconstruction

Training Program. However, in constructing the Wisconsin program, Ericksen felt that a more expansive vision that included the development resources of the MEP and the coordinating capacity of multiple OEMs would provide tangible beneWts to all involved, and would also be more politically viable—if he could mobilize the support of other OEMs. He thus joined with the executive director of WMEP, and together they managed to convince and recruit representatives from Wve other OEMs that they considered to be ‘good citizens’ and that they knew had made eVorts to cultivate good relationships with many of their suppliers.97 The consortium was founded primarily to provide suppliers with a problemcentered training program to improve Wrm performance in lead and cycle time reduction, delivery, product quality, and cost. However, it also aims more generally to aid supplier viability by enhancing supplier–OEM business relationships, increasing understanding of OEM needs and introducing suppliers to other potential customers. The vision is very much one of both leveraging and enhancing collaborative relationships between OEMs and suppliers to align the organizational models of involved Wrms (that is, to spur investment in cospeciWc assets). It was inaugurated in summer 1998, aided by a $500,000 allocation from the state budget. This grant paid for a 50 percent discount on high-quality training (often in collaboration with the state technical college system, with which WMEP already had a close relationship) to small and medium-sized enterprise (SME) supplier Wrms in Wisconsin nominated by one of the governing OEMs.98 The OEMs are required to select Wrms they consider ‘strategic’, and with which they have at least a 24-month relationship that they intend to continue. The resulting organization, while of governable size, is still a signiWcant chunk of the state’s manufacturing economy. As of May 2003, the governing OEMs reported spending some $713 million annually with 212 nominated suppliers that collectively employ about 50,000 people in Wisconsin (since that announcement, the consortium added a seventh large OEM and some of its suppliers).99 After a Wrst year in which the consortium drew on Deere’s existing supplier training infrastructure for administrative services and a large portion of the training, the full administrative management passed to WMEP in July 1999. This shift in administrative governance went smoothly, and both allowed and obligated the OEM partners to give greater input into the direction of activities. It also put the WMEP into the position of ‘honest broker’, to ensure that the costs and beneWts are shared out fairly among the participants, and to discourage opportunism by Wrms who often compete for the same customers and suppliers.100 The consortium claimed in its original mission statement that it would establish a curriculum of emphasis based on a consolidation of the OEMs performance expectations, and would help suppliers to assess training needs. In practice, this has been an uneven—but improving—process. The course oVerings were originally selected by the OEMs to reXect their own supply

It Couldn’t Happen Here?

137

chain management strategies, beginning with a list provided by the Wisconsin Technical College System and the Deere training department. A subset was then agreed upon by the consortium as a whole (changing over time as undersubscribed courses were culled, and others added at OEM request).101 To date, the WMDC by no means embodies all that consortial a model could do to overcome the many relational barriers to OEM–supplier collaboration. For example, the diVerent OEMs have quite varied and even contradictory expectations of suppliers in some areas; likewise, assessment and guidance to suppliers as to OEM needs—while an essential part of the strategic vision—has been somewhat sporadic. Nevertheless, over time, the consortium has worked to improve in these areas, and has taken such important steps as adding supplier representatives to the board (one in 2001, a second in 2003), holding supplier networking meetings, staging a joint program ‘kickoV ’ in 2003 to give nominated suppliers a sense of the governing OEMs’ commonality of purpose, and strongly lobbying as a unit to defend WMEP’s funding in the face of the state’s looming Wscal crisis. I return below to a discussion of the growth and development of the WMDC, arguing that despite its limitations, it is indicative of the possibilities of institution-building in the sphere of interWrm relations even in the American context. First, however, it is important to understand just how the WMDC helps to sustain collaboration in OEM–supplier relationships.

The Basics: The WMDC as Toolbox Even as OEMs moving to a more decentralized production model push suppliers to take on greater responsibilities and to perform complex tasks previously performed in-house, they also absolutely require that these same suppliers meet steadily increasing manufacturing performance standards. Many of these demands, a supplier interviewee explained, are ‘exactly the same’ across diverse end user industries: ‘People are looking for 99.8% or so on time, 500 ppm [parts per million] or less defects and 5–7 percent cost reductions over the year.’102 To help small Wrms restructure their operations to meet this changed reality, the WMDC provides nominated suppliers with an important toolbox, an economical source of ongoing training for both managerial staV and production workers. The variegated nature of individual Wrms’ restructuring processes makes it diYcult to separate out and quantify the eVects of any supplier training program—in the abstract language of statistical methodology, there are too many sources of variation—but suppliers themselves do recognize how their companies have changed and can often assess the impetus of those changes. In interviews, they have provided considerable evidence of concrete and measurable improvements on key manufacturing metrics such as cycle time, productivity, and on-time delivery. Examples can give a Xavor for the sorts of changes that occur. A product specialist reported that training provided by the consortium had been

138

Institutions and Relational Reconstruction

instrumental in an initiative to implement numerous Xexible production cells. In these areas, he said, improvements were substantial, with cycle time ranging from a 35–40 percent reduction in the worst case, to a ninefold reduction in the best. With training and improved setups, they also increased machine utilization in the cell areas (something they began to track as a part of the initiative), moving in some cases from 25–40 to 65–80 percent. Tracking machine downtime allowed them to isolate occurrences in which the cause was insuYcient training of workers in areas relevant to their jobs (like not knowing how to set up a particular machine), so they could correct the observed deWciencies. The result was that where before 20–25 percent of setups presented to quality assurance were correct, a majority were later done entirely by production staV. Another supplier, who emphasized that the WMEP training was but a partial and enabling element in an overall strategic change, reported that workin-progress had been reduced by about 10 percent and inventory turns had improved. Cycle time was halved—from eight to four weeks—eYciency (measured as machine up-time) improved, and they were able to lower prices on about 10 percent of their products. A metal fabricator that had restructured operations with the help of consortium-sponsored training said that inventory was dramatically reduced—by 70 percent—while cycle time also improved from about seven weeks in 1997 to four weeks in 2000. The customer reject rate has spiked up and down, but was generally between 0.25 percent and 0.5 percent, whereas three years ago it was 1.5–2 percent. The improved quality, the company president said, is due to a combination of factors, including training, but the WMDCdriven focus on process has undoubtedly been a factor. He remarked: ‘Three years ago, if you had asked me what our rejects were caused by, I would have said ‘‘operator error’’; but now, rejects are [understood to be] more caused by process problems.’ The company had been reducing job classiWcations, and had doubled—to 50 percent—the percentage of the workforce that was crosstrained.103 This toolbox aspect of the WMDC and the performance eVects generated are extremely important—indeed, as emphasized in Part II, supplier competencies are quite important for fruitful OEM–supplier collaboration—but the simple state provision of training services to small Wrms again do not contrast Hall and Soskice’s descriptions of ‘incentive compatible’ LME industrial policy. There are, however, ways in which the WMDC diVers fundamentally from ‘typical’ LME policy. The consortium does considerably more than simply provide supplier training: it uses the coordinating power of the associational model and its partnership with the public authorities to help align OEM and supplier strategies along collaborative lines by underwriting credible commitment and by providing an external support network for OEM–supplier joint projects.

It Couldn’t Happen Here?

139

The WMDC as a CME-Style Institution I: Strategic Alignment By their very membership, the governing OEMs in the WMDC provide a signal of credible commitment to a collaborative model. As it was explained by the owner of a Wrm supplying two of the founding OEMs (and who later agreed to become a supplier representative at consortium meetings): The idea that two of my major customers would form a consortium with other people to help train their supply base, . . . I saw that as, ‘We’re in a whole diVerent world now.’ This is no longer, ‘We do three quotes and send it to the lowest bidder and every year we go out and rebid it . . . and if things slow up at all, we cancel everybody’s orders and we make it in our own shop.’ That was the paradigm in 1990 [but it is changed today].

Seeing this, he sold oV an unproWtable proprietary product that was intended as a hedge against risk, and substantially intertwined aspects of his Wrm’s strategy to meet the needs of a core of important customers. As was shown in Part II, however, even where there is alignment of strategy, there can be continued relational diYculties and mistrust. This is highlighted, for example, in an extensive study of the eVects of supplier development programs in the agriculture, construction, and industrial equipment industries by Mesquita and Brush (2001: 15). They write that some suppliers ‘see supplier consultants doing more the job of ‘‘policing’’ than ‘‘promotive development’’’, leading suppliers to avoid them because ‘the OEMs may value the help its consultants extend more than suppliers do. In such a case, suppliers have perceived that OEMs initiate a more aggressive bargaining behavior in order to extract ‘‘unfair’’ price discounts.’ Relative to these problems, OEMs in the WMDC are advantaged because they can use the WMEP’s honest broker role to ensure relational continuity even when suppliers are unwilling to divulge to OEMs the strategic information required for a truly joint restructuring of supplier processes. An example is provided by a product specialist. The supplier was in trouble because of large batch sizes and other inXexibilities that were causing delivery and cost problems, and knew they had to make substantial changes in the organization of production if they were to meet increasingly uneven customer order patterns. Although they did work some with a customer’s supplier development engineers, they ultimately were not comfortable sharing ‘the nitty-gritty of all the facts as to how our processes are made and the time it takes, because . . . we know what they’ll do. They’ll use it against us to say, ‘‘Well, we know that you could take this much waste out. Now we want this cost reduction.’’ And we’re doing everything we can right now to keep the cost the same, and to Wnd ways to reduce the cost if we can.’ Instead, the supplier elected to work with WMEP (and use internal resources) on numerous projects to develop a more Xexible manufacturing system that was better able to meet spikes in customer demand (a customized modiWed pull system). The training coordinator commented, ‘You know, for a small company like this, training, huge

140

Institutions and Relational Reconstruction

training budgets are not possible, and so the cost is really good, you know, with the subsidized part. . . . What I’ve told other groups time and time again, is that the WMEP training to me is world class training’ that ‘can bring in the outside knowledge.’ Suppliers interviewed have used the training consortium in almost countless diVerent ways, ranging from general learning about changing customer requirements to starting with coursework and then engaging WMEP manufacturing specialists to help them to substantially reorganize operations. For example, a producer of plastic components working for several consortium members was moving to a build-to-demand organization of production and sent the entire workforce to lean courses because, the general manager said, ‘If we’re going to go this way and we’re going to be doing some things, I wanted everybody to understand [why].’ A job machine shop, on the other hand, began by sending only a couple of people to lean courses. But, liking the results, the supplier then sent approximately 60 percent of the workforce to training oVered by the consortium and has engaged WMEP Weld agents to help push through the creation of several production cells. Using the training program as a means to learn about and use WMEP manufacturing specialists is a relatively common pattern among suppliers who use the program (note, however, that many nominated suppliers never use the training at all). Of the Wrst 75 suppliers to use courses sponsored by the WMDC, 27 also used the services of WMEP manufacturing specialists to help guide them in projects aimed at a more substantial reorientation of production strategy.104

The WMDC as a CME-Style Institution II: External Support for Internal Initiatives A very important second way in which the WMDC is a CME-style institution is that it simultaneously depends on and strengthens existing collaborative eVorts between OEMs and suppliers, combining public and private coordinating capacities to aVect Wrms’ strategic interrelationships. Abstracting and idealizing, and using the same concepts introduced in Part II to explain why Sabel’s neopragmatist learning by monitoring story should not be expected to lead to a self-actualizing collaboration, the model through which the WMDC sustains OEM–supplier collaboration is the following: As collaborationist factions at OEMs use the training program to build competencies at suppliers, this latter group is more likely to have the capabilities to develop joint ‘cost’—not ‘price’—reductions, strengthening the position of those same collaborationist factions at the member OEMs. This can raise their credibility towards suppliers, improving trust as reliability and making it even more likely that suppliers will invest in new competencies under the belief that these are likely to be rewarded—which makes successful joint innovation more likely, in turn further strengthening the voice of collaborationists within OEMs; and so on.

It Couldn’t Happen Here?

141

In this section, I use examples of restructuring at three suppliers to show how the WMDC plays the role of external support network to collaborationist factions. All three stories are about suppliers whose relational and performance problems had left them on the verge of being replaced by a key customer, but, because of good relationships to factions within these customers, they were able to undertake joint restructuring processes that leveraged existing partial collaboration, training, and the WMEP to maintain continuity in the relationship. Two of the supplier Wrms worked with a single OEM, so their cases are presented together, followed by the third case. In the book to this point, I have drawn on descriptions from numerous Wrms of the negotiation of roles between OEMs and suppliers in general, and thus have not used speciWc pseudonyms in my examples. In these next subsections, however, the narrative describes actions between just a few Wrms, so I do use pseudonyms. Rather than inventing company names, I simply use the identifying codes for interviewed Wrms from the list in the Appendix (CM20 and CM25 worked with OE1; CM6 worked with OE5).

Starting with the paternalistic story: OE1 arrives and Wxes everything A maker of engineered components supplying OE1 for over 10 years, CM20 had a history of quality and delivery problems. There was, however, the saving grace of design capabilities, product development support, rapid delivery of prototypes and, most importantly, a good working relationship with OE1 engineers. Nonetheless, as quality and delivery problems worsened when orders increased, in 1999 OE1’s purchasing department began preparations to move the business. However, supplier development engineers from OE1 who had been in CM20 on and oV over the years argued that many of the problems could likely be resolved through improvements in the manufacturing operation, which—if brought up to the level of CM20’s design and engineering—would allow them to retain a valuable supplier. These collaborationist voices prevailed. After some discussion with the supplier, the OEM oVered a long-term deal that mandated yearly cost reductions, but that also promised OEM engineering aid so that these could in fact be ‘cost’ rather than ‘price’ reductions. This was a risky move for both parties. The OEM was banking on its ability to improve quality and delivery at an outside supplier, while the supplier had to hope to be able to meet the agreed upon cost reductions and had to give full disclosure on all their costs, thus revealing margins and damaging future bargaining power. Such information helps customers identify opportunities for collaborative cost reductions, but can easily be misused for short-term gain. OE1 sent a supplier development engineer to the supplier for an extensive period, where she did root cause analysis and discovered (with the aid of supplier personnel) a fundamental welding problem that was then resolved by OEM experts, dramatically improving quality. There was also a ‘spaghetti layout’, with the batched parts moved from machine to machine in shopping carts. The engineer facilitated the formation of a multilevel cross-functional team at

142

Institutions and Relational Reconstruction

the supplier, including the scheduler, a key supervisor, the general manager, a process engineer, and a quality manager. Ultimately, all agreed on a new layout that more closely approximates the desired single-piece Xow production concept. Changes were quite successful, eVectively reducing cost and improving delivery at the supplier. CM25 is a job machine shop that was a problem supplier in the late 1990s, but that improved dramatically since: in 2001, defects were at 400 ppm, down from a 1999 peak of 1800þ, and late deliveries have decreased thirteen-fold. These improvements came in a relatively short period, and coincided with sustained help from OE1’s Supplier Development engineers as well as increased training eVorts. OEM personnel described CM25 as a small family-run business that had decided to grow, but that simply grew too fast and outgrew its procedures. Prior to the turnaround, an OEM interviewee said that the factory was real dingy and dark, with no real rhyme or reason why the machines were where they were, lots of parts piled up between operations. That was part of the problem. You would get parts that by the time they were here [at the OEM] were missing a hole— they jumped an operation, parts shipped not done. . . . Some was worker turnover, not having the workers really empowered, or even caring about their job, part of it was the way the machines were laid out. The computer system was part of the problem, to get info on a delivery, they would have someone walk out into the shop to see what stage it was at.

He added that they wanted to get bigger, but at some point you can’t just keep it all in your head, and mistakes start happening. They wanted to grow, but the computer system wouldn’t let them grow, the way they managed the people and machines, the purchasing guy making sure they had enough steel—they would run out of steel because they weren’t watching. . . . They were getting swamped, we were giving them too many parts and we started to realize that they just couldn’t handle it, they needed a little of everything Wxed.

The Wrst project was simply to improve part tracking, which helped improve quality and spurred them to implement a more contemporary ‘cellular’ manufacturing model. The Wrst cells were kept within particular departments, to make the adjustment easier. Learning from the experience, however, CM25 subsequently became much more conWdent and moved a signiWcant portion of production into work cells. More responsibilities were devolved to shopXoor personnel, especially to the lead people in the cells, with operators handling their own quality and some paperwork. The supplier also increased wages after the reorganization, increased internal worker training, and implemented a formal suggestion system that has been helpful in getting ideas from the shopXoor and in getting workers involved in changes that, in the end, drove measurable improvements in lead times, quality, delivery, cost, and in the quoting process. These in turn combined to make a major diVerence in the metric that generally

It Couldn’t Happen Here?

143

interests most, productivity: CM25 would have been able to maintain the same production levels in mid-2001 (the level was down some because of the downturn) as they had in 1999 with 41 percent less people, even as safety has improved so much that their workers’ compensation insurance risk rating (based on ‘losttime’ accidents) had fallen by 55 percent relative to a peak six years ago. Because of the openness and improvements, OE1 is now increasingly comfortable sourcing parts from CM25. In the past, an OEM buyer said, they would be nervous about some of CM25’s quotes. An initial bid would come in low, but six months later, there would be quality and delivery issues. Such problems no longer occur. Before, ‘If they bid $1.50, in the past I was always apprehensive— are they really making proWt on this, are they just bidding that way to get the business and they really don’t know what the proWt is? Because the other quote is, say, $2.50. Are they really that eYcient at $1.50, or are they really missing the boat?’ Now, he said, he trusts the price, and that they are eYcient. Part of the past mistrust came from unstable prices and fears that CM25 would not be able to generate cost reductions.

Correcting paternalism: the relational perspective It is easy to tell a paternalistic story, with the powerful customer simply imparting wisdom to their suppliers—a variant on hierarchical coordination—but to do so obscures that each of these suppliers has been very actively engaged in reorienting their own operations. It also simply does not square with the story told by OE1 personnel. These engineers explicitly said that the key issues in their jobs lay largely in building trust and helping supplier managers and supervisors to fully understand and believe in the principles behind a single-piece Xow production model. As the head of supplier development at OE1 explained: ‘The engineering part of supplier development is by far the easiest thing to handle. The change management side, on the other hand, is by far the most diYcult’ and is one that is dramatically helped when suppliers have access to third parties that can provide them suYcient knowledge to take ownership of the process and to know that they are not being taken for a ride. This view was strongly reinforced by the lead supplier development engineer on the project at CM25, who said: ‘When you get cells together, 10 percent is getting the equipment moved, the rest is just getting people to do it. The operators have been working there so long, so why do it some other way?’ He emphasized, however, that he did not think he could run CM25 better than the owner. What he could oVer, he said, were useful tools and a steadying presence. Both suppliers tell a similar story. They knew there were problems in their operations, but, as small companies, they beneWted greatly from having an outside engineer who had been through the process before and was there regularly to push the issues and to help them ride out the uncertainties as they began the slow and painful restructuring of their manufacturing processes. The general manager at CM20 contrasted the intensive collaborative model of OE1’s supplier development to a ‘quick-hit’ approach in which the customer would just come in, identify

144

Institutions and Relational Reconstruction

opportunities, give some tools, Wnd a champion and leave, monitoring from afar. An approach that requires the outside party to get their hands dirty, and to build trust and relationships is—while more expensive—much more eVective because it recognizes the resource constraints of a small, Xat organization. Both of these suppliers stridently reject that the improvements were uniquely ‘caused’ by the customer. From their perspective, they have learned many new principles, some from customers and some from WMEP, and are increasingly able to build their own cells, establish their own part families, and generally improve operations. While the customer may have provided a useful ‘kick in the pants’, both suppliers are adamant about their own active role in the process. As should be clear from the above discussion, improving the operation of a factory is not just moving machines around. It requires that people working in the factory do their jobs diVerently. It is here that the WMEP supplier training program adds value. Even in cases in which the OEM invests substantial time and money in improving supplier performance, access to high-level organizational training from an independent and credible third party helps suppliers decide for themselves if the organizational model being pushed by the consortium OEMs is truly viable. CM25 has been extremely active in using the WMEP training, using classes to improve their understanding of cellular manufacturing, cycle time reduction and other ideas being pushed by their customers. The plant manager commented that the WMEP courses aided the joint project with OE1 by giving them conWdence that it would work. They made the ‘whys’ clearer, and gave him the means to train and educate supervisors and lead people. Likewise, employees at CM20 said that the training was very much in line with the principles being espoused by OE1 engineers, so they could use the courses to diVuse the new ideas across supervisors and other key personnel. Most importantly, having the training provided by a third party, such as WMEP, was useful, they said, because there is always a fear that the customer is perhaps too focused; it is much more convincing to have a disinterested outside opinion. Both of these suppliers were very active users of the program, both for speciWc managerial and workforce training as part of the particular transformations described and to learn more generally about changing manufacturing practices. Indeed, the general manager at CM20 said, if the WMEP courses were not teaching similar principles and practices to those espoused by the customer, it would have forced him to question whether the knowledge of the OEM’s supplier development engineers was really up-to-date.

Brokering corporate-plant disjunction CM6 had an extremely diYcult relationship with the subsidiary of a key customer. Both sides thought the other to be at fault. As a manager at the supplier Wrm put it: ‘There was not a lot of mutual respect between us and [the customer]. Neither one of us thought the other had a clue.’ The customer complained that the supplier was constantly late with deliveries, while the

It Couldn’t Happen Here?

145

supplier felt that the source of problems lay in the customer’s extremely irregular order patterns, which were making production programming at the supplier virtually impossible. The ‘running joke’, a key supplier supervisor said was: ‘it’s Thursday, it’s time for the next [subsidiary] crisis’, a reference to constantly having to bring people in on Saturdays (and pay overtime) to handle the erratic needs of this customer. However, CM6 had a better relationship with the OEM’s corporate oYce. The plant manager explained: ‘I think [OE5] has a good feel for what [CM6] can do, and probably [the subsidiary] said to [OE5], ‘‘God, [CM6] is horrible’’ and [OE5] said, ‘‘No, they are not at all horrible’’ because we have done a lot of work for [OE5] over the years.’ Thus, albeit with some trepidation, CM5 agreed when corporate supplier development personnel suggested a joint project, with the goal of reducing the supplier’s response time to make them better able to handle irregularities in the subsidiary’s order patterns. A team consisting of both supplier and OEM personnel was formed, requiring signiWcant investment of human resources by the supplier. In the words of a supplier manager: It was tough. Nobody else was doing our jobs while we sat in these meetings three days a week, every week. . . . There were times where we would skip a week, but always with a reason—homework, projects to get done. . . . When [the OEM team leader] left, people would have their homework. But at the end of the day, you still had the rest of your duties too. . . . There was some great stuV though. Up to this point, we had no clue what a Takt chart was, or a Kanban system.105

The project was not without its conXicts, especially around the release of process information to the customer. The supplier commented: ‘Probably, in the course of the project, we thought we were giving up way too much information, but when the project was completed, and all was said and done, I don’t know that we gave up anything that the customer didn’t already know in most cases, or shouldn’t know.’ Nevertheless, ‘I am sure we gave [the OEM team leader] headaches because we would do our darndest to tap dance around every issue, so we could discuss after he left if we really wanted to give that to him.’ The real problems, as might be expected, came in the arm-wrestling over what was real savings, what was paper savings. . . . The downside, I would say, to the whole project, is that we still did not collect enough data to be really truly accurate. The savings at the end was done on projected savings. And I wasn’t comfortable with that, and I don’t think we ever saw that—but I also don’t think they ever really got the full savings they asked for either. . . . There was some kind of concession at the end of the year.

After this Wrst successful instance, which, one supplier manager said, ‘was the Wrst time that anybody acknowledged, very grudgingly—we had to do it—that something was wrong. And afterwards, it was like, ‘‘Wow, this is not so bad’’ ’ and while ‘there are still issues’ with the subsidiary, things are ‘getting much better’.

146

Institutions and Relational Reconstruction

The supplier deemed the project an ‘exceptional’ success, with machine changeover times reduced to a sixteenth of what they were, lead times halved from three to 1.5 weeks, and productivity in the cell that was the focus of the project improved by more than a third. Importantly, however, as a nominated supplier to the WMDC, CM6 could take advantage of the consortium to ensure that the initial project did not remain an isolated event, serving rather as a sort of ‘kickstart’: the plant manager said that in the Wnal judgement, ‘It may have been an expensive lesson, but we learned a lot, and have recouped it many times over by carrying out [other such] projects [afterwards]’, in part by leveraging their engagement with the consortium and with WMEP to develop the internal capacity to carry out their own continuous improvement projects. Only a few people had been involved in the original project, so generalizing the lessons across multiple plants required the aid of a professional training organization. They developed with WMEP a ‘curriculum for the continuous improvement leaders; it was everything from leadership and facilitation skills to set up reduction, to process mapping, to cycle time reduction. . . . It was six weeks of Tuesday–Wednesday, two days a week for six weeks.’ 6.5. R EG I ONA L I N DUS TRI A L PO LI CY AN D C ON S O RTI A L M O D E L S O F S UP P L I E R T R A I N I N G The WMDC has enjoyed considerable practical success during the Wrst several years of its operation, and even formed the template for a similar policy initiative in the state of Pennsylvania.106 And while it has to date focused primarily on the Wrst condition required for the emergence of a new production paradigm— improving the skills and capabilities of the existing supplier base—and is at best an imperfect solution to the many relational barriers to deeper OEM–supplier collaboration, it nevertheless provides clear evidence that the incremental construction of nonmarket coordinating institutions in the interstices of the American LME is both possible and incentive compatible. In addition to its provision of an external support network for collaborationist factions at both OEMs and suppliers, there are at least four additional ways in which a consortial model helps participants to develop collaborative solutions to common problems.107 Sharing out the costs and beneWts of widely-needed services: Many suppliers work for several competing OEMs. A consortial organization allows OEMs to contribute technical expertise and support to upgrade their suppliers’ capabilities without incurring the high Wxed costs of in-house training operations, and with less risk that competitors will appropriate the beneWts. Public subsidies reduce the cost of training to hard-pressed SMEs, while the watchful eye of the OEM provides an incentive for supplier participation. WMEP can thus devote less of its time and resources to marketing to potential clients and focus more on the provision of core services, while also ensuring that training activities reach a minimum eYcient scale.

It Couldn’t Happen Here?

147

Facilitating information Xow: The partnership structure of the consortium allows the WMEP to aggregate the common needs of OEMs and transmit these to suppliers and training providers. The curriculum development and review process of OEM representatives can lead to greater transparency for suppliers about current and potential customers’ quality and service needs, thereby assisting them to adjust their operations to meet the latter’s expectations. Promoting mutual learning: The WMDC is in a strong position to promote mutual learning among the participants above and beyond the speciWc content of the training courses themselves. Although this is an area in which the consortium has considerable room for improvement, the OEMs have discussed developing common supplier qualiWcation and certiWcation procedures. While there is not yet a systematic framework to allow suppliers to learn from one another or to use their responses to training courses and to OEM procurement practices to improve the work of the consortium, there are steps in that direction. The supplier representatives to the board have organized several supplier networking meetings, initially on training issues, and have arranged with supplier representatives on the board to conduct ad hoc supplier forums (with only suppliers present) to get feedback on issues of common concern. Strengthening of ‘business coordinating capacity’: The objection of the VoC approach to the use of coordination-oriented policy in LME is not that such policies are per se objectionable, but rather that they are simply very unlikely to function eVectively. This claim is defended by pointing out that LMEs (or at least the United States) are characterized by weak associational structures in which existing producer groups often focus only on ‘lowest common denominator’ issues (i.e. lobbying for tax cuts, weakening of regulations, lowering the minimum wage, etc.) that are of relatively marginal interest to a large percentage of their membership. It does not necessarily follow, however, that policymakers should not take incremental steps to change that status quo—if they can. The WMDC does suggest that, at least at a regional level, they can, by taking advantage of an existing if limited relational structure in American manufacturing. The consortium has succeeded in organizing members around a common interest, does obtain from those members the information required to eVectively administer coordination-oriented economic development policies, and gives policymakers an interlocutor with the ability to ensure that eVorts to stimulate some nonmarket coordination will not go to naught. It is also large enough (containing some of the most important manufacturers in the state) to exert some ‘structural inXuence’ (Hall and Soskice 2001: 48) in the political arena to pressure the state government to meet its own commitments.

Limitations and Open Questions Accomplishments aside, it is important to recognize the limits of the WMDC. I have not identiWed a panacea for all that ails American manufacturing or even all that aZicts OEM–supplier collaboration.

148

Institutions and Relational Reconstruction

In their survey of state and local experiments in work and welfare policy, Dresser and Rogers stress that the United States has seen a very resistible rise of novel workforce intermediaries, backed by a very weak supporting infrastructure; but they also argue that these incremental eVorts could become something larger. The experience of the WMDC should be similarly interpreted. It, like the WRTP, was constructed in the interstices of the state’s economic development framework. It also Wts with patterns of change in work and welfare policy in that its birth and growth owe a great deal to particular and forceful policy entrepreneurs. Yet it nonetheless clearly shows that it is possible to Wnd or create the relational and business coordinating capacities required of CME-style coordinating institutions even in the prototypical American LME. Most importantly, evidence that the WMDC has successfully garnered business interest and aided OEM–supplier collaboration in incremental-innovation-dependent durable manufacturing contrasts the notion that such policies are not suYciently ‘incentive compatible’. Nevertheless, showing the feasibility and desirability of nonmarket coordinating institutions in the sphere of interWrm relations certainly does not mean that building such institutions is easy, nor does it mean that state policy entrepreneurs and collaborationist factions in business will take advantage of the space available to them. There are three big issues facing policymakers and practitioners if the WMDC or something similar is to become a larger part of state-level industrial policy: its extension, its deepening, and, most importantly, its survival. I present them as open and fundamentally political questions that provide opportunities to rethink and revise the model, rather than as inherent Xaws in a promising—if partial and imperfect—institutional answer to problematic interWrm coordination.

The politics of federalism: Is the consortial model expandable? The most obvious limitation to the WMDC as a template for industrial policy is the need to widen access. Within states, at least, this need not be particularly problematic. Long-term political support for publicly subsidized training limited to a small subset of the state’s supplier Wrms is likely to be tenuous, and premising the delivery of manufacturing extension services on consortial models requires more than seven OEMs and their suppliers (though, recall, this is a large group, combining to employ some 50,000 workers), but there is no reason not to stimulate the formation of additional consortia of locally rooted OEMs with a suYcient commonality of purpose and a substantial shared in-state supply base. This is preferable to simple expansion of the existing consortium, given the importance of the focus and commitment of the governing partners to its eVective functioning. The big issues arise, however, when confronted with the constraints imposed by an American federalism that—while serving the important and useful role of

It Couldn’t Happen Here?

149

lowering the bar for policy experimentation—can force patchwork policy Wxes to the many problems that run across state lines. For example, all of the OEMs in the WMDC have numerous suppliers in other states (most of them also have large production facilities in other states), and virtually all of their suppliers have key customers in other states. Thus, even where there are potential beneWts to greater coordination of public and private resources in an externalized supplier development system, its creation would require substantial intergovernmental coordination in economic development policy. Notably, this barrier is not insurmountable given that there are manufacturing extension partnerships across the Upper Midwest and that the component manufacturing industry is itself heavily concentrated in just Wve states (Illinois, Indiana, Michigan, Ohio, Wisconsin).

The politics of collaboration: Can consortia really help to align performance expectations and supplier development practices across the OEMs? At present, the WMEP and the supplier training program largely serve to complement existing OEM supplier development and supply management resources. But the WMDC could theoretically grow into a more complete externalized framework in which the OEMs increasingly depend on the consortium and the WMEP—eVectively outsourcing the functions they have developed to deal with the consequences of outsourcing. This would necessitate much greater information sharing than currently exists between the WMEP and OEM purchasing departments, requiring, for example, the latter to share substantially more internal data on supplier performance and to better notify the former about problem areas at particular suppliers. It would also require better deWnition of a ‘curriculum of emphasis’ with its implied agreement among the consortium OEMs on the goals of purchasing strategy.108 At time of writing, very important but quite nascent steps had been taken in this direction. In late 2004, a subset of three of the OEMs in the consortium were going forth with a very ambitious project to use WMEP manufacturing specialists to jointly develop a series of key suppliers. This project requires that involved OEMs and suppliers structure operations around a relatively encompassing common metric, ‘manufacturing cycle time’ (Ericksen et al. 2005) that could potentially drive a much more substantial shared realignment of OEM and supplier strategies. This project is being watched by other members of the WMDC as well as by OEMs outside Wisconsin. Access to eVective training and supplier development resources is an important problem, and, as shown above, its resolution can improve collaboration between OEMs and suppliers. Nevertheless, as discussed in Part II, supplier Wrms’ larger problem remains that their customers, even those who talk the talk of open and collaborative supplier relations, often do not walk the walk, partly because of internal organizational barriers. In theory, the subsidies could be used as a ‘stick’ to monitor OEM behavior, but the sums involved are small

150

Institutions and Relational Reconstruction

relative to the sales turnover of these Wrms; hence the OEMs would likely simply walk away from any such enforcement. The ideal solution would perhaps be to somehow stimulate stronger sectoral organization of suppliers, akin to the German Arbeitsgemeinschaft Zulieferindustrie (ARGEZ), an association of component manufacturers that helps their members to enforce minimum conditions for payment, delivery and price changes within contracts.109 This, however, is hard to do (though more likely if the OEMs themselves get better organized). There are nevertheless incremental ways in which a multiWrm public–private partnership might draw on the idea of the consortium as an external support network for collaborationist factions to help OEMs resolve characteristic dilemmas of large bureaucratic organizations in ways that would be beneWcial and cost-eVective to the OEMs themselves. For example, while it would be an extremely ambitious step, the consortium could encourage participating OEMs to draw up a common code of good supplier relations practice, based on member Wrms’ own oYcial procurement policies. Such a code could stimulate the identiWcation and diVusion of good practice among participating OEMs, while also guiding suppliers towards common performance expectations. Along the lines of ISO 9000 quality assurance programs, its implementation within the consortium could be assessed by independent third-party monitors who would also assess the tangible impact of training on supplier performance. This would help to reveal the sorts of systematic short-term-oriented deviations from oYcial policies that force Wrms to hedge against each other’s own behavior. Participating OEMs in breach of the consortium’s code of good supplier relations practice would be asked to submit plans for correcting the problems identiWed by the external monitors within a reasonable time period. In cases of persistent uncorrected breaches of the code, consortium members and the MEP would then consider a range of possible sanctions, culminating in exclusion from the consortium. The third-party monitoring process could itself be harnessed to mutual learning through benchmarking of supplier training practices and related research on OEM–supplier relations, thereby providing a systematic mechanism for generating improvements to the consortium’s curricular oVerings and code of good practice. Third-party reporting on the OEMs’ performance in implementing the collaborative supplier relations policies to which they are formally committed would help mitigate some of the organizational dysfunctions discussed in Chapter 5 and would strengthen the position of reformers in these companies. In assessing such a ‘code of conduct’ (or any such proposal to leverage the consortium to improve OEM practices), it is essential to recognize that it could only work if powerful voices within OEMs can be convinced that it is beneWcial to seek such external reinforcement as a means to enforce adherence to their own oYcial procurement policies across plants.110 And this can only happen if the consortium itself continues to develop and to provide results.

It Couldn’t Happen Here?

151

The politics of survival: Can the model survive the state Wscal crisis and an LME state? At time of writing, the United States is only slowly emerging from a recession that has been particularly unkind to the manufacturing sector. The impact of hard times on Wrms’ interest in consortial supplier development is hard to predict. OEMs experiencing a proWt-squeeze may utilize short-term positional bargaining strategies to play desperate suppliers against each other (or against suppliers in low-cost countries) to salvage proWts now; likewise, in response to tough times, supplier Wrms may simply ‘hunker down’ to weather the storm, investing less in people they are not sure they will be able to keep. Each of these responses threatens the consortium, which depends fundamentally on a commitment from both the OEM partners and their suppliers to working together to improve operations for gains down the line. But there is also an opposite pressure. A recession can also push OEMs to look more closely at their own organizations to identify areas where they can reduce their own Wxed costs, and can make the reliance on external public resources more attractive, even if it requires sharing strategic control of those resources with other Wrms and with state agencies. The economic justiWcation for state investment in supplier training follows from an extremely straightforward market failure argument, the sine qua non of positive-sum industrial policy (Bartik 2003a; 2003b). Given that suppliers work for multiple customers, OEMs are unlikely to be able to fully recover the cost of investments in supplier capabilities. Likewise, the suppliers themselves have relatively little bargaining power and the stringent cost targets they are expected to meet lead them to be wary of investments such as training with relatively intangible and long-term returns—to the detriment of the regional economy. Thus, the knowledge and the training invested in the upgrading of the supply base is a public good for which no private concern in the supply chain is rationally interested to foot the entire bill. The justiWcation for resolving some portion of this market failure through organizations like the WMDC is also quite straightforward. The very existence of the consortium shows that all involved are willing to make some investments in the reinforcement of the state’s shared supply base. By leveraging the existing network of relationships, the state can generate diVuse beneWts across the economy with even a relatively small subsidy. As such, it Wts easily into a ‘high-road’ cluster strategy seeking to enhance human capital and provide other intangible organizational assets necessary to quality-competition strategies (though, as discussed, this strategy should only be followed if there is in fact suYcient commitment by interested Wrms to focus the consortium’s training and development resources to push Wrms along that high-road).111 Interestingly, however, the greatest threat to the consortium’s survival has come not from a lack of business coordination but from a combination of the state Wscal crisis and political disputes between state actors about whether or not

152

Institutions and Relational Reconstruction

to route funding for manufacturing modernization through WMEP or through the technical college system. This dispute, combined with extreme budget pressure caused by a $3 billion structural budget deWcit, led the state to cut even its relatively small outlay to WMEP from $1.5 million in 2002 to just $100,000 in 2003.112 Combined with cuts in federal funding, the survival of WMEP and thus of the consortium became quite tenuous. This, however, led many in the manufacturing community to organize in support of the organization. Though the public mission of the WMEP is to serve small and medium sized Wrms, the large OEMs of the consortium—some of the most prominent manufacturing Wrms in Wisconsin—lobbied both individually and collectively in support of the organization (underlining the importance of ‘structural inXuence’ discussed above). These eVorts were ultimately successful, as the state assembly (97–1) and senate (31–2) voted overwhelmingly in March 2004 to restore $1.5 million over two years to the WMEP, ensuring at least the mediumterm survival of the organization.

6.6. CO NC LUS I O N In this chapter, I have argued that there is space for the building of ‘CME-style’ institutions dependent on substantial business coordinating capacity in the interstices of the highly federalized American economic development apparatus, and that such institutions can help American manufacturing Wrms to overcome some of the barriers to eVective collaboration identiWed in Part II of the book. This analysis draws fundamentally on a theorization of the OEM–supplier relationship as multiplex and contradictory, but also relies on a key insight of the Varieties of Capitalism position: the strategic interactions of Wrms are aVected by the institutional surround even as that institutional surround is itself aVected by the relationships Wrms have to each other. I dispute the path dependency that follows from the VoC’s overly holistic analysis of the American Liberal Market Economy. There is in fact much more space than the VoC Wnds for American policymakers to break with the status quo by taking advantage of contradictions, incongruities and within-system diversities in the relational structure of American manufacturing. Importantly, this is not a claim of radical contingency, but rather a constructivist eVort to identify feasible options that are wrongly—and analytically—foreclosed by other theories.113 To this end, in my descriptions of strategic hedging and contradictory relationships, I identify hooks available for the construction of CME-style coordinating institutions even in the theoretically inauspicious ex ante conditions of the American Upper Midwest. Building directly on the identiWcation of these hooks in Part II of the book— particularly the possibility of depending on and supporting collaborationist factions in OEMs—the example of the WMDC shows where some of this space for nonmarket coordinating institutions in the prototypical American

It Couldn’t Happen Here?

153

LME is to be found. Moreover, the details of its genesis and functioning demonstrate Wrst the importance of recognizing contradiction and multidimensionality in the OEM–supplier relationship, and second that such institutions must be actively constructed (following Granovetter 1992). It is possible in the United States to combine the interest of both OEMs and suppliers in coordinating the improvement of competencies at the latter with mechanisms that draw on and support collaborationist factions at the former. In so doing, they can mitigate some of the uncertainties that force them to hedge their collaboration to a greater degree than serves the regional economy as a whole. Notably, the existence of partial collaboration and supportive factions at key OEMs, the American federal structure, and manufacturing modernization resources at the state level are but enabling conditions for the construction of the sorts of institutions that can push OEMs and suppliers in American manufacturing in the direction of the collaborative new production paradigm outlined in Part I of this book. The WMDC was actively built by individuals who saw an opportunity to leverage outside resources to help them build collaborative relationships they believed would beneWt their Wrms in the long term. It is thus ironic that the greatest threat to its survival does not come from an absence of business organization seeking the tools for a coordinated solution; it comes rather from political disputes in the ‘LME-state’. There are some elements in the United States of ‘CME business’ with long established relationships that can be organized; however, someone has to do the organizing. This is precisely why the claim that nonmarket coordinating institutions are unfeasible and thus a waste of time in the United States is not just wrong but downright counterproductive: if taken seriously, it generates its own unhappy self-fulWllment. It may be diYcult to create the right political coalition in the United States, and policymakers may have to search hard for strong spots in what is undeniably quite spotty collective coordinating capacity, but that hardly seems reason to destine existing counter-examples to a virtually inevitable failure. The point, in short, is that constructing nonmarket coordinating institutions in the American LME is fundamentally a question of political will. There is, in the arena of interWrm relations at least, no irresistible snapback caused by the equilibrating forces of institutional complementarities.

7 Toward the Relational Reconstruction of Regional Political Economy 7.1. S U M MI N G UP: DECE NT R AL I ZE D PRO DUC T I ON , T H REE LITERATURES, THREE UNCERTA INTIES, AND CONTRA DIC TORY C OL LA B ORATION The 1980s are famous as the era when the rust belt—the manufacturing heavy Upper Midwest and Northeast—rusted. But in fact, though there certainly was substantial deindustrialization, those years were characterized even more strongly by deunionization, deurbanization, and especially by a very fundamental deverticalization of production. In the wake of increased global competition and a fragmenting of mass markets in the 1970s, some manufacturing Wrms were forced to close their doors. But many remained and actively engaged the new environment by retrenching to their so-called core competencies in design, marketing, and assembly by subcontracting (‘outsourcing’) other activities to a series of smaller suppliers—some abroad, but very many still in the United States—that now do much of the ‘real’ manufacturing of components. In so doing, they created a post-outsourcing manufacturing economy that is not simply a more global, less urban, less unionized, and less concentrated version of the same thing. Rather, it is substantially reorganized, a new old economy in which most of what matters to manufacturing Wrms no longer happens under roofs they own or control. The severe crisis that struck American manufacturing at the beginning of the twenty-Wrst century must be seen in this light, as occurring in an industry in which the quality and structure of relationships between Wrms is both important and complex. It means that we need to understand this old economy not only by looking at what happens inside large companies; we must also devote substantial attention to how activities are coordinated and governed between companies. This has been increasingly recognized in the social scientiWc literature since the 1980s, and many theories have been generated to explain why and how Xexible production—in various forms—is destined to replace ‘old’ fordist models. In this literature, there are two points of agreement that are especially relevant to the contemporary situation. 1. There are numerous examples worldwide of a normatively attractive new production paradigm, heavily inXuenced by the importation and hybridization of Japanese production systems. Contrary to fears in the 1980s, these practices

Reconstruction of Regional Political Economy

155

have proven potentially replicable, at least in part, in many contexts outside Japan. The new paradigm is premised on the interconnected diVusion of socalled ‘high-performance work organizations’ within Wrms and of collaborative relationships across Wrms, especially as regards the vertical OEM–supplier relationship. 2. However, this highly Xexible new production paradigm comes with a double-edge: there is no guarantee that all Wrms or regions will be able to adequately follow its prescriptive tenets. Deverticalization does not by itself lead to collaborative network production. It too often reXects instead Wrms’ eVorts to chase lower wages both domestically or abroad, and/or the ability of large Wrms to shift risk and to exploit oligopsony positions. And although particular regional or national cultural and institutional conWgurations make it more likely that Wrms will follow the prescriptive tenet of the new production paradigm, this too has an ominous implication: insider Wrms in ‘learning regions’ are relatively more able to compete in high-surplus quality-conscious niche markets because they can exploit relational networks that include other Wrms, associations, unions, and government actors at multiple levels; outsider Wrms struggle with ruinous price competition from the low-wage world. The upshot of this double-edge is that there remain big issues that must be resolved if we are to decipher—and perhaps manipulate—the future of an evidently challenged American manufacturing sector. It is clear not only that the successful governance of decentralized production can help OEMs to compete in global markets, but that it is also vital to the prospects of their more territorially bound small and medium-sized suppliers and to the large percentage of the manufacturing workforce employed at these smaller Wrms. Yet manufacturing regions more accustomed to hierarchical and market governance—like the American Upper Midwest—are thought ill-positioned to generate the mechanisms required to stably sustain a high-collaboration decentralized manufacturing model. Fortunately, as I have argued in this book, there are useful theoretical tools to resolve these issues, most particularly in eVorts to understand how Wrms manage to sustain collaborative relationships in the face of the inevitable prisoners’ dilemma dynamics. Three approaches that feature prominently in these debates—social networks and embeddedness, VoC, and neopragmatist ‘learning by monitoring’—are particularly suggestive. These three positions agree on three fundamental points. First, they all hold that some form of collaborative network production is normatively desirable, at least in some industries (especially those dependent on incremental innovation). Second, all argue that collaborative Wrm networks are possible, citing mechanisms—patterns of social ties, formal and informal institutions, and/or learning routines—that can circumvent problems of incentive alignment and knowledge sharing that more economistic literatures cite as bedeviling interWrm relationships. And third, these underlying coordination

156

Institutions and Relational Reconstruction

mechanisms are self-reinforcing in such a way that relationships between Wrms can be expected mostly to fall into stable and dichotomous equilibrium conWgurations, with a world of hostile arm’s-length contracting on one side and the collaborative sharing of information between weakly bounded Wrms on the other. Intermediate cases are not exactly assumed away, but they are left untheorized: as Crouch and Farrell (2004: 33) explain, for the social embeddedness and VoC approaches, incongruous elements are ‘ ‘‘noise’’ which needs to be disregarded in the interests of an elegant and sharply proWled account’; for Sabel’s learning by monitoring, in-between cases are not necessarily anomalous, but are again essentially noise, a transitional stage in the organizational revolution. The Wrst two points of agreement—that collaboration is often good, and that we can identify mechanisms that allow it to stably occur despite barriers cited by the economics of organization—are useful contributions to the understanding of economic coordination. The third suVers from the worst of problems: it is directly contradicted by the empirical evidence. The case study presented in Part II, based on more than 100 interviews at OEMs and supplier Wrms in quintessential old economy industry—metal manufacturing—in the quintessential old economy region, the American upper Midwest, shows the modal case of the OEM–supplier relationship to be systematically intermediate between the arm’s-length and collaborative poles. Despite this widespread reorganization of production, which has clearly raised the potential payoVs to positive-sum collaboration and that virtually requires that OEMs and suppliers share information and explicitly coordinate operations, these relationships are often quite contradictory and are characterized by strategic hedging along multiple dimensions in a sort of ongoing waltz intermingling competition, conXict, and cooperation. This intermediacy reXects Wrms’ shared negotiation of three fundamental uncertainties that make pure strategies extremely risky: (a) market and technological uncertainties, partially mitigated by the decentralization of production, but inevitable; (b) competence uncertainties, a residue of the legacy of capacity subcontracting; (c) organizational uncertainties, owing to the inability of Wrms to be coherent strategic actors. Market/technological and competence uncertainties are not necessarily problematic for the stories of dichotomous equilibria to which I have objected— indeed, Sabel’s story of the collaborative organizational revolution is actually premised on the claim that they can set oV a virtuous circle in which initial experimentation with collaboration can set the stage for further experimentation (see Section 5.2.). But as the empirical research presented in Part II shows, organizational uncertainty creates a particularly vexing set of barriers to Wrms seeking to build collaborative relationships. Good-faith eVorts to collaborate are consistently undermined by the complexities of staV turnover and general information-transfer diYculties, as well as the oft-times conXicting incentives across diVerent subunits of large organizations—all made worse by the rise and fall of diVerent factions within organizations.

Reconstruction of Regional Political Economy

157

This point, that organizational dysfunction continues to bedevil economic coordination notwithstanding enormous changes in corporate structure, represents theoretical advance through hybridity, and is explicitly intended as a reconstruction and modiWcation of established ideas through their application to a novel situation. SpeciWcally, I begin with the core insight of post-Weberian organizational sociology—as Gibbons (1999) summarizes it, that Wrms are a mess, but, so long as explanation goes beyond simple aggregate interest to examine as well the actions, interests, and decision processes of subunits and individuals, they need not be a mystery—but I show that the implications of this insight must be substantially rethought and made to Wt a radically decentralized industrial organization. Coase showed economists that there is no reason to expect real world Wrms to be oblivious to the conditions that wreck markets; by the same turn, sociologists cannot simply assume that real world networks, despite their theoretical virtues, will be oblivious to conditions known to wreck Wrms. Certainly, that the blurring of organizational boundaries aVects both the internal and the external has not been lost on theorists of the network organizational form. Helper, MacDuYe, and Sabel (2000), for example, explicitly argue that the emergence of pragmatic collaboration requires the contemporaneous reform of both external and internal relationships, a breaking down of traditional hierarchies and re-examination of routines through the federation of production and learning by monitoring. But the story is nevertheless one of a market-driven and seemingly inevitable transition, of initial experimentation causing reforms that despecify assets and render observed instances of hold-up and mistrust uninteresting remnants of the past, mere transitional blips. As such, it is not suYciently attuned to problems in the morphology of the emerging network forms. There have been real and important changes in Wrm structure, but conXictual and centralized intraWrm relationships nonetheless do continue to plague interWrm relationships, owing especially to factional conXict both within and across departments and to the diYculty of aligning particular incentives within organizations to reward behaviors aimed at cementing longterm collaboration. Supplier Wrms react by hedging their own collaboration in turn, even as they recognize that they are engaging in behaviors that are systemically suboptimal and part of a vicious cycle. The result is that American manufacturing is mired in an altogether partial transition to a more collaborative interWrm organization of production, rife with ongoing contradictions that sit uneasily with sociological literatures on network organizational forms. Many large American manufacturers are making real eVorts to follow the prescriptive tenets of the collaborative new production paradigm even as their eVorts are deeply constrained by the need to hedge fundamental uncertainties caused by a history of poor relationships and a lack of institutional support. This combination of mixed motives and occasional abject failure leaves a relational structure that is neither the collaborative production network that Powell (2001) calls the very building block of the

158

Institutions and Relational Reconstruction

twenty-Wrst century Wrm, nor is it an atomistic world of hostile arm’s-length contracting. Rather, relationships between OEMs and suppliers in American durable manufacturing are best described as a complex mix of the two, suggesting that the possibilities are considerably less bifurcated than the existing literature would have them be. We do not see the clean-but-uneven pattern of collaborative alliance capitalism implied by the social embeddedness and/or VoC positions, with insider and outsider positions explained by the oftentimes inhospitable American cultural and institutional conditions. But it is also a contradictory reality that does not Wt the predictions of Sabel’s learning by monitoring: my interviews squarely reject that the federation of production and initial experimentation with collaboration alone enable a stepwise and self-reinforcing move towards a more collaborative new production paradigm. Functionality and market forces alone, it seems, are not enough to drive an organizational revolution; politics matter too. 7.2. T HE PR ACT I CA L MO R AL I N TH E S TO RY The modal OEM–supplier relationship is empirically intermediate between arm’s-length and collaborative, but this need not—in general—mean rejecting theories that predict otherwise; something is ‘noise’ for every theory. But in the particular, the too-facile dismissal of systematic contradiction in these relationships—what Crouch and Farrell (2004: 33) call ignoring ‘incongruities, incoherence, and within-system diversities’ in the quest for an ‘elegant and sharply proWled account’—wrongly obscures important options available to actors in the political economy. Listening to the relational noise can oVer these same actors new tools to escape theorized traps of historical and institutional circumstance. This claim need not—indeed, should not—be intrinsically hostile to the three established theoretical literatures primarily engaged in this book. Rather, mine is a friendly correction that borrows liberally from elements of each, drawing on and contributing to economic and organizational sociology in a relational reconstruction of regional political economy to explain: (a) why eVorts by OEMs and suppliers to build more collaborative relationships are so uneven; and (b) how systematic intermediacy and contradiction can be used in the service of public policy and institution building. SpeciWcally, following Helper, MacDuYe, and Sabel (2000: 475), I agree that the spread of aspects of the Japanese model and learning by monitoring ‘even in a short-term oriented, individual interest-maximizing society like the USA’ contradicts those who would argue that full-blown ‘Japanese-style collective institutions’ and/or a general culture of trust are required to ‘generate the conditions necessary to maintain and nourish collaboration’. But I empirically reject any suggestion that this will lead to a market-driven organizational revolution as Wrms discover a superior set of organizational routines that then become the ‘key to survival in otherwise unmanageably turbulent world’ (Sabel 2004: 2). Contradictory and often failed eVorts by OEMs and suppliers to build

Reconstruction of Regional Political Economy

159

more collaborative relationships are not so much a steady transition as they are the result of a complex but stable mix of hedging strategies. This outcome is driven not only by the need to negotiate market and technological uncertainties and a history of poor relationships, but also by unreliable contracting caused by the failure of organizations to behave as coherent strategic actors. This does not mean, however, that we are condemned to whichever intermediate state of the political economy happens to us, varying perhaps with the relative relational competencies and endowment of recovering Fordist OEMs. Rather, modifying again an established line of argument—this time, ideas prevalent particularly in the political economy of the VoC and the sociological literature on embeddedness—my interviews with OEMs and suppliers conWrm that formal and informal institutional mechanisms do aVect how Wrms negotiate their environments. But, contrary to claims that institutional complementarities and historical speciWcities mean that the mechanisms to sustain interWrm collaboration cannot feasibly be built absent propitious ex ante conditions, there are more options available to policymakers and supportive factions within Wrms than conventional dichotomizing approaches would suggest. Historical Wrm competencies and coordinating institutions depend on past choices and perhaps do not change easily, but they do change. Paralyzing stories of overly strong lock-in sometimes found in the social embeddedness literature notwithstanding, a core tenet of this strain of economic sociology, recalling again Granovetter’s historically rooted point (1992: 7), is that ‘economic institutions do not emerge automatically in response to economic needs [but] are constructed by individuals whose action is both facilitated and constrained by the structure and resources available in social networks in which they are embedded.’ Following this caution, I have in this book been attentive to the potential functionality of network governance of production and exchange on the one hand, and to the barriers to collaboration long-cited by the economics of organization on the other. I have also been thick enough in my descriptions of the OEM–supplier relationship to show that incongruities, incoherence, and within-system diversities form part of the ‘structure and resources available’ to actors in the political economy. It is true that the United States lacks a strong institutional infrastructure to help manufacturing Wrms generate the worker skills, collaborative relationships and network organizational structures required to compete in the rapidly changing and quality conscious markets that can sustain the high wages that must (and should) be paid in the developed world. But it is false that there are not suYcient tools in the United States to build such an infrastructure. There are many examples of coordinating workforce intermediaries at the state and local level in the policy sphere of workforce training and development. And the example of the WMDC, built jointly by OEM collaborationists and state actors, shows that there is space in the interstices of the very decentralized American economic development apparatus for novel solutions, for the incremental construction of institutions that both depend upon and strengthen existing partial collaboration between OEMs and suppliers.

160

Institutions and Relational Reconstruction

The United States indisputably is, on the whole, an LME. But it does not follow that there is something to gain by following the usual liberal market policy prescriptions in all corners of that liberal market economy. And it is simply wrong both empirically and theoretically to think that state actors cannot and thus should not encourage the building of nonmarket coordinating institutions in those sectors—like durable manufacturing—favored by such policymaking. 7.3. T H E M ETA- THEO R ET I CA L MO R AL I N T H E STO RY In Mother Night, Kurt Vonnegut writes that the moral of the story is, ‘We are what we pretend to be, so we must be careful about what we pretend to be.’ The metatheoretical moral of this book parallels—in a sense—Vonnegut’s admonition. To speak about and understand an indescribably complex world, we necessarily generate abstractions, concepts, ideal types and the like—that is, we pretend the world is a certain way so we can talk about it and make decisions about what seems best to do. And there is absolutely nothing wrong with this, it is inescapable; but at the same time, because we use this pretending to decide how to act, the way in which we do it, the level of abstraction, how much causality we infer from our ideal types, and so on, matter. We must be careful, for not all ways of pretending are equally useful: it depends on the problem at hand. Much of the theoretical action in this book follows from this premise. I did not select the three approaches with which I am in dialogue because they are somehow intrinsically ‘wrong’. To the contrary, each struck me as providing essential analytic tools for understanding the complex mix of collaboration and competition I heard so much about when I would talk to people at factories in the American Upper Midwest: Sabel’s learning by monitoring reminded me that apparently anomalous Wndings might well reXect incipient change; the comparative political economy of Hall and Soskice explained why it was so hard for Wrms to sustain vertical interWrm collaboration without deliberative institutions; and Granovetter’s take on social embeddedness made clear that these economic institutions would not come about just because they would be functional—real actors have to build them and they need social and political resources to do that.114 Without this work, I would not have been able to ask the questions that I have asked. But each of these approaches also struck me as somehow incomplete, as silent on aspects of the mix of relationships that, it became clearer to me with every new interviews, matter a great deal. Each has made important contributions to the understanding of decentralized production, bringing to light weaknesses in other, often more economistic (and abstract), approaches that came evident as the global economy changed rapidly in the 1980s and 1990s: both learning by monitoring and the social embeddedness approach have done much to explain the spread of ostensibly anomalous ‘network’ organizational forms and the collaborative relationships on which they are premised; and Hall and Soskice

Reconstruction of Regional Political Economy

161

usefully render dynamic the static typologies typical of comparative political economy by asking how institutions aVect and are aVected by strategic interactions between Wrms. But for the question at hand here—the need to understand and improve the governance of decentralized production—it is not enough to theorize relationships and organizational forms previously seen as anomalous; we must also understand the salience of anomalies within those relationships and organizational forms. This gap in existing approaches risks analytically obscuring the need for, and the possibilities of, constructing the institutions required of a high-wage, highproductivity, high-collaboration manufacturing economy even in the theoretically unpropitious context of the American Upper Midwest. This is the sense in which I hope to have hit on the ‘right’ level of abstraction—the proper sort of pretending—for the problem at hand, sidestepping both the fatalism and paralysis that follow from positing either an errant determinism or a too-radical contingency, and showing why an understanding of systematic contradictions can serve both theory and policy by identifying obscured and hopeful paths available to actors in the political economy.

Appendix A.1. THE I NTERV IEWS This appendix supplements the brief description of the data-gathering process provided in the introduction to Part II and should meet ‘back-of-the-envelope’ standards for the presentation of one’s methodology. That is, it provides the information that someone else would require to reproduce the study (with obvious exceptions made for constraints imposed by the anonymity granted to all interviewees quoted in this book, and of course leaving aside the diYculties of gaining access to informants and convincing them to speak freely). The book is based on a series of 102 qualitative and semistructured interviews with OEMs in key durable manufacturing industries and their suppliers in the component manufacturing sector. As deWned by the Advanced Manufacturing Project (see Introduction to Part II), the component manufacturing sector is made up of the many (often small) Wrms that fabricate and/or assemble molded, forged, formed, and machined goods made of metal and plastic, principally for sale to other manufacturers. Firms in the sector, despite their process variety, share many commonalities, the most important being a market position squeezed between larger raw-material suppliers and Wnal goodsproducing customers. There are also many common intermediary institutions: these include the NIST manufacturing extension partnerships discussed in Chapter 6, trade and professional associations (such as the National Tooling and Machine Association, the Society of Automotive Engineers, Society of the Plastics Industry, Precision Metalforming Association, etc.), and labor unions (i.e. International Association of Machinists, United Steelworkers, Allied Industrial Workers). A listing of all interviews with personnel at OEMs and component manufacturers (CMs) done in the United States is provided at the end of this appendix. In the interviews, I asked about internal organization, about relationships to other Wrms and how these are aVected by the larger market and institutional context. The interview protocols listing the topics and questions asked in these semistructured interviews are also at the end of this appendix. These interviews are the source of the stories used to retheorize the OEM–supplier relationship in Part II. I emphasize, however, that my ability to understand and frame the case has been greatly supplemented by interviews and regular interaction with union personnel and with staV at manufacturing training organizations and at the manufacturing extension. Also, between April 2001 and May 2003 (excepting January–June 2002, when I was doing empirical research in Italy), I regularly attended meetings (approximately every six weeks) of the WMDC, putting me into regular contact with personnel from the participating OEMs and supplier representatives.115 The interviews were carried out in two phases. My Wrst phase of interviews was carried out between August and October of 2000. In addition to a focus on the interWrm relationship, I gathered data to contextualize and assess the eVectiveness of the WMDC.116 This Wrst phase built on work done previously by JeV Rickert. I used summaries of his interviews at the six founding OEMs of the WMDC and returned to ten CMs (all suppliers to these six OEMs, interviewed a year previous by Rickert) to gather further information on their involvement with the consortium and also about

Appendix

163

their changing relationships with their customers.117 The second phase, with interviews carried out between July 2001 and January 2002, considerably extended the investigation. I added generality by interviewing more Wrms in some way connected to the WMDC and by including OEMs and suppliers not connected in any way to that particular policy initiative. This latter phase of the research was particularly bolstered by my aYliation with the Advanced Manufacturing Project, a multistate research consortium dedicated to investigating the determinants and possible policy supports of improved performance in manufacturing in the component manufacturing sector. The principal investigators of AMP were Susan Helper, Gary Herrigel, Daniel Luria, Joel Rogers, and Jonathan Zeitlin. It was funded principally by the Alfred P. Sloan Foundation with supplementary funding from the WMEP. As a part of AMP, I was given access to transcripts and summaries of interviews at OEMs and CMs that provided me a control on my own Wndings (as can be seen in the listing below, I use seven additional interviews at OEMs and ten additional interviews at CMs). Note that for interviews that I did not do personally, I only use those for which I have a full transcript, with the exception of three of the OEMs interviewed in phase one. I have included those cases because I have since met and interacted regularly with the interviewees through our joint involvement with the WMDC. In using quotes from interviews, I do not say which individual actually conducted the interviews referenced (i.e. I write ‘a supplier interviewed said . . .’ even where I was not personally on the receiving end of the statement but have only a transcript). Interviews focused on both internal organization and relationships to customers. They were semistructured, though guided by an interview protocol to ensure that all topics were covered. Interviews generally lasted between one and three hours, and, when acceptable to the subject, were taped and then transcribed; otherwise, I have worked from summaries and interview notes. Subjects were guaranteed anonymity. Thus, in the presentation of the data I have at times substituted identifying words with more general information in square [brackets]. Also, I use expressions such as ‘interviews for the Advanced Manufacturing Project show . . .’ to refer to general synthetic Wndings of the project. Everything cited in this book as a direct quotation is taken from a transcript of a taped interview. All interview data has been coded into various categories and organized using QSR N6 qualitative data analysis software, so I can for any quote easily identify the speaker, the interviewer, and the full context in which a statement was made. Given the assurances of anonymity that have been given, however, I have elected not to footnote quotes to particular Wrms and give only what contextualizing information is necessary to understand their signiWcance. In selecting Wrms to interview, the objective was to construct as complete a picture as possible of the networks of OEM–supplier relationships. I identiWed major OEMs in a range of relevant industries, requested interviews, and then worked down the supply chains of these OEMs by interviewing a series of their suppliers. The listing of interviewed component manufacturers indicates whether they supplied any of the interviewed OEMs. With one exception, every OEM interviewed had over a billion dollars in annual sales. To ensure that suppliers were comfortable speaking about their customers, and to increase the number of points of view on each of the interviewed OEMs’ practices, when possible, I tried to identify suppliers that worked for several of the OEMs interviewed. For

164

Appendix

OEMs 1–3, and 10–12 (see listing), all members of the WMDC, I was able to select suppliers to interview from a full listing of all the suppliers they had nominated for the program, allowing me to cross-reference to see which suppliers worked for multiple participating OEMs. This allowed for multiple perspectives on the practices of particular OEMs and on particular OEM–supplier relationships (i.e. both nodes across particular ties). These interviews down the supply chain were then supplemented with interviews at CMs identiWed in a range of other ways, particularly through ties to associations. At OEMs, when suYcient access was granted, multiple interviews were done, largely with managers responsible for purchasing and supplier development. At smaller supplier Wrms, the subject was often a plant manager or perhaps the company president or owner. In one OEM (OE1), I was able to obtain very substantial access, allowing for multiple interviews at three divisions in three states; and in two supplier Wrms (CM20, CM25), I did full-day visits with numerous interviews to gather detailed information on concrete changes driven by the Wrms’ participation in training provided by the WMDC. The roles of personnel interviewed within the respective Wrms are identiWed in the listing below. The next two subsections contain a listing of all interviews (with identifying information stripped, of course). The listing says who did the interview, Wrm size, and primary industry. After the listing, I indicate in parentheses the phase in which the Wrm is interviewed (1; 2), whether I am working from a transcript or a summary (T; S), and whether a supplier was nominated to the WMDC.

A.1.1. Interviews with component manufacturers in the United States (74 interviews at 44 Wrms) CM1: Josh Whitford and Gary Herrigel interviewed owner. CM1 is an independent 140person urban job machine shot in Illinois, primarily making parts for OEMs in agricultural implements, construction, auto and plumbing industries. CM1 is a supplier to OE1 and OE5. (T) (2) CM2: Josh Whitford interviewed plant manager and training manager. CM2 is a subsidiary of a small conglomerate with 156-employees in a small Wisconsin town making powdered metal parts for a wide array of end-user industries. CM2 is a supplier to OE6, OE3, OE1. (T) (2) CM3: Josh Whitford interviewed owner of small (15-person) urban independent metal spinner in Wisconsin. Shop works for a wide array of end-user industries. (T) (2) CM4: Josh Whitford interviewed controller and owner of independent rotary molded plastics producer in small Wisconsin town. Company makes parts for a wide range of end-user industries, including OE2. (T) (2) (WMDC) CM5: Josh Whitford and Jonathan Zeitlin interviewed the divisional manager of a small conglomerate centered in Wisconsin. Company has several plants in Wisconsin, and mostly produces exhaust systems and related parts for nonautomotive vehicle industries. Company is a supplier to OE1 and OE3. (T) (2) CM6: Josh Whitford interviewed the plant manager, sales manager, and a shopXoor supervisor at a subsidiary of CM5. Tube bending plant employs 350-people in a small Wisconsin town, works for nonautomotive vehicle industries and supplies OE1 and OE3. (T) (2) (WMDC)

Appendix

165

CM7: Josh Whitford interviewed plant manager at independent 140-employee machine shop in small Illinois town. Plant works primarily for Agricultural and construction machinery producers, including OE1 and OE10. (T) (2) CM8: Josh Whitford and Gary Herrigel interviewed sales manager and human resources manager at 52-employee plant making industrial drive shafts in suburban Chicago. Plant is subsidiary of European multinational and makes components for agricultural equipment producers, including OE1 and OE10. (T) (2) CM9: Josh Whitford and Matt Vidal interviewed president and human resources manager of 350-employee independent producer of brake systems for nonautomotive vehicles in suburban Wisconsin. Plant works for OE3. (T) (2) (WMDC) CM10: Matt Vidal interviewed plant manager of subsidiary 60-person metal stamper in suburban Wisconsin producing parts for automotive and agricultural and construction equipment industries, including OE1 and OE3. (S) (2) CM11: Josh Whitford and Gary Herrigel interviewed the owners of a 67-employee independent tube bender in urban Illinois. Plant makes parts primarily for agricultural and construction equipment industries, including OE1. (T) (2) CM12: Josh Whitford interviewed the vice president of manufacturing and the head of sales at an independent midsize (700-employee) metal fabricator in a small Wisconsin town. Company sells primarily to agricultural and construction equipment industries, including OE1 and OE11. (T) (2) (WMDC) CM13: Josh Whitford interviewed quality manager, human resources manager, sales manager, training manager at a 300-employee independent manufacturer of industrial seating in Wisconsin. Company sells to motorcycle, construction, and lawn and garden industries, including OE1, OE2, OE3, OE10. (T) (1; 2) (WMDC) CM14: Josh Whitford and Matt Vidal interviewed the president/owner of a small conglomerate producing injection molded plastic parts for cars and motorcycles, including OE3. Plant in question had 400 employees. (T) (2) (WMDC) CM15: Josh Whitford and Gary Herrigel interviewed plant manager, sales manager and human resources manager at an 800-person metalforming factory in Illinois working primarily for the agricultural and construction equipment industries, including OE1. Company is a member of a small group. (T) (2) (WMDC) CM16: Josh Whitford interviewed the owner of a small (50-employee) independent Wisconsin producer of pulley systems for the lawn and garden industry, including OE1. (T) (2) (WMDC) CM17: Josh Whitford interviewed the plant manager at a 184-person independent Wisconsin producer of wireforms for the lawn and garden industry, including OE1 and OE2. (T) (2) (WMDC) CM18: Josh Whitford interviewed the owner of an independent 148-employee job machine shop in Wisconsin. Plant worked for varied industries (including OE2), was specialized in low volume, including aerospace and medical equipment. (T) (2) (WMDC)

166

Appendix

CM19: Josh Whitford interviewed the president of an independent 120-employee Wisconsin job metal treater working particularly for the lawn and garden industry, including OE2. (T) (2) (WMDC) CM20: Josh Whitford interviewed the plant manager and four shopXoor employees at an independent 120-employee Wisconsin job machine shop working especially for lawn and garden, agricultural and construction equipment industries, including OE1. (S; S) (1; 2) (WMDC) CM21: Josh Whitford interviewed the plant manager at an independent 85-person Wisconsin screenprinter making decals for various industries, but including OE11 and OE2. (T) (2) (WMDC) CM22: Josh Whitford interviewed the plant manager, sales manager, and human resources manager at an independent 300-person Wisconsin producer of industrial seating. Company worked especially for construction, truck and oVroad vehicle industries, including OE1 and OE2. (T) (2) (WMDC) CM23: Josh Whitford and Gary Herrigel interviewed an independent 150-person screenprinter making decals for OEMs in agricultural equipment, lawn and garden, and other industries, including OE1. (T) (2) CM24: Josh Whitford and Gary Herrigel interviewed the owner for an independent 120-employee job machine shop in Illinois. Company worked especially for agricultural and construction equipment industries, including OE1 and OE10. (T) (2) CM25: Josh Whitford interviewed the president and three shopXoor employees of a subsidiary producer of industrial cylinders in Wisconsin. Company supplied varied industries, including lawn and garden and agricultural and construction equipment industries. Company supplied OE1 and OE10. (S) (2) (WMDC) CM26: Josh Whitford interviewed the owner of an independent 160-employee wire harness maker in Wisconsin making parts for the truck industry, including OE4. (T) (2) CM27: Josh Whitford and Gary Herrigel interviewed the VP of manufacturing and head of sales at an independent midsize (360-employee) maker of industrial drive shafts in Wisconsin. Company supplies the agricultural equipment industry. (T) (2) (WMDC) CM28: Josh Whitford interviewed the president of an independent 62-employee Wisconsin company that cut and distributed steel tube, and that supplied OE3, OE1, OE2. (S) (1) (WMDC) CM29: Josh Whitford interviewed a manufacturing engineer at a 115-employee independent Wisconsin metal fabricator that supplied OE2 and OE11. (T) (1) (WMDC) CM30: Josh Whitford and Jonathan Zeitlin interviewed the owner of an independent 80-employee Wisconsin supplier of machined parts to motorcycle, lawn and garden and small engine industries. Company supplied OE3 and OE1. (S) (1) (WMDC) CM31: Josh Whitford interviewed the sales and human resources managers at an 80-person subsidiary making closures for fuel systems in the nonautomotive vehicle industries. Wisconsin company supplied OE1, OE3, OE11. (S) (1) (WMDC)

Appendix

167

CM32: Josh Whitford interviewed the owner and sales manager at an independent 90-employee Wisconsin maker of industrial springs that supplied OE2 and OE12. (S) (1) (WMDC) CM33: Josh Whitford interviewed the owner of an 118-employee independent supplier of large metal fabricated parts to varied industries, including OE4, OE10, OE12. (T) (1) (WMDC) CM34: Josh Whitford interviewed the plant manager and owner at a factory that is a member of a small group centered in Wisconsin. The plant in question employed 70 and made metal fabricated parts especially for the lawn and garden industry, including OE1, OE2, and OE10. (T) (1) (WMDC) CM35: Josh Whitford interviewed the owner at an independent 70-person Wisconsin wire harness maker supplying OE2 and OE5. (T) (1) (WMDC) CM36: Gary Herrigel interviewed the Wnance oYcer at an independent 70-employee Michigan metal fabricator supplying customized parts to machine tool makers, including OE7. (T) (2) CM37: Gary Herrigel interviewed the owner of an independent 30-employee producer of commercial terminals and connectors in Illinois. Company supplies motor vehicle, defense, and electrical industries. (T) (2) CM38: Gary Herrigel interviewed the owner of an independent 80-person stamper and laser cutter in Wisconsin. Company supplied especially lawn and garden and materials handling industries, including OE1. (T) (2) CM39: Gary Herrigel interviewed the owner of an independent 200-employee caster in Illinois. Company supplied the pump and electric utilities industries. (T) (2) CM40: Hyeong-ki Kwon interviewed the sales manager of a Japanese subsidiary in Michigan. Plant employed 40 people making motion control products for the automotive industry. (T) (2) CM41: Gary Herrigel interviewed the general manager of a 20-employee plant in Illinois producing injection molded parts for autos, working at the second tier of automotive supply. Plant is owned by a conglomerate. Plant supplies OE6. (T) (2) CM42: Gary Herrigel interviewed the owner of an independent 100-employee die caster in Illinois producing parts for the automotive and appliance industries. (T) (2) CM43: Gary Herrigel interviewed the purchasing manager, the plant manager, and the head of human resources at a 150-employee Japanese subsidiary making electronic relays, switches and exchanges for automobile applications in Illinois. (T) (2) CM44: Gary Herrigel interviewed the plant manager at a 200-employee subsidiary producer of industrial bearings, especially but not exclusively for the electrical and the construction and agricultural equipment industries. (T) (2)

A.1.2. Interviews with OEM personnel (28 interviews at 12 OEMs) OE1: Thirteen interviews were conducted by Josh Whitford at three diVerent divisions (Gary Herrigel accompanied at one of these divisions) of a Fortune 500 producer of

168

Appendix

Agricultural and construction equipment and lawn and garden equipment. At OE1a, these interviews included: divisional purchasing head, manager of supplier development, supplier development engineers, buyers (Phase 1, manager in purchasing, reinterviewed in Phase 2). At OE1b: divisional purchasing manager, manager of supplier development, buyer. OE1c: divisional purchasing manager, head strategic buyer. (S; T) (1; 2) OE2: Josh Whitford conducted three interviews at agricultural equipment producer, including VP of supply chain, head of purchasing and a buyer in Phase 2. In Phase 1, a purchasing manager had been interviewed. (S) (1; 2) OE3: Josh Whitford and Jonathan Zeitlin conducted interviews at a producer of nonautomotive vehicles. In phase one, a midlevel manager in corporate purchasing was interviewed. The same person was re-interviewed in phase two, as was his direct supervisor. (S; T) (1; 2) OE4: Josh Whitford and Matt Vidal interviewed the vice president of purchasing at a Truck producer. (S) (2) OE5: Josh Whitford and Matt Vidal interviewed a purchasing manager at a small engine producer. (S) (2) OE6: Gary Herrigel interviewed the VP of purchasing at a major Wrst tier supplier to automotive, oVroad vehicle and agricultural equipment industries. (T) (2) OE7: Gary Herrigel interviewed the divisional purchasing manager at a machine tool producer. (T) (2) OE8: Gary Herrigel interviewed the director of business development at a Wrst tier automotive supplier. (T) (2) OE9: Gary Herrigel interviewed the divisional purchasing manager and a supplier development engineer at a producer of heavy and medium trucks. (T) (2) OE10: JeV Rickert interviewed a corporate purchasing manager at a producer of agricultural and construction equipment. (S) (1) OE11: JeV Rickert interviewed a corporate purchasing manager at a small engine producer. (S) (1) OE12: JeV Rickert interviewed a corporate purchasing manager at a producer of heating and air conditioning equipment. (S) (1)

A.2. INTERVIEW MODULES This section contains the interview modules that were used to guide the semistructured interviews with OEMs and suppliers in the United States. A.2.1 contains the questions used to guide the interviews with personnel at OEMs; A.2.2 contains the questions for suppliers; and A.2.3 contains a module that was sent to suppliers prior to the interview to gather descriptive quantitative data.

Appendix

169

A.2.1. Advanced Manufacturing Project: Interview Question Guideline for OEMs These questions are indicative of the areas that will be asked about in the interviews. The interviewer will also follow up on answers given by the interviewee. I. Strategy 1. Could you brieXy outline your philosophy of OEM–supplier relations? (a) What is behind procurement strategy—what is the relative importance of speed, quality, delivery, design, etc. and why? II. Organizational structure (vis-a`-vis purchasing) 1. Who are the key Wgures in deciding Wrm purchasing strategy? 2. How much autonomy do plants have in making purchasing decisions? (a) Is this autonomy contingent on meeting targets set by corporate? 3. Is supply chain management strategy decided at the corporate or plant level? 4. Are suppliers chosen at the corporate or plant level? III. Product design, marketing, and purchasing strategy 1. Which functional departments are involved in product design? (a) How are relations across these departments? 2. At what point are the purchasing and manufacturing people brought into the design process? Marketing? 3. How often do people in purchasing meet with people in engineering? In marketing? 4. At what point are suppliers brought into the product design process? (Will return to this later) 5. Do marketing, engineering, and purchasing give a consistent message to suppliers (in your opinion)? IV. Make/buy issues 1. Roughly what percentage of the product is made from purchased materials/parts? (a) How/why has this changed over time? 2. What operations are internal? (a) How/why has this changed over time? (b) Are some operations partially internal (i.e. some work done here, overXow at suppliers)? 3. How is the decision to outsource a particular operation made? (a) Who makes the decision? (i) What level (corporate/plant) (ii) What functional divisions are involved? (b) What are key criteria in the decision? (i.e. investment in capacity, labor costs, specialization, desire to take advantage of specialized assets?) (c) Have parts ever come back in after having been outsourced? Why? 4. Has outsourcing been the source of negotiations with the union at this plant? (a) If jobs are eliminated due to outsourcing, are workers retrained? V. The supply base and purchasing strategy 1. What—if anything—ties you to this region? (i.e. labor force, customer base, supply base, sunk costs, schools, history, etc.).

170

Appendix 2. How many suppliers do you have? (a) How has this changed over time? 3. How large is the purchasing staV? (a) How many suppliers per staV person, responsibilities of buyers, etc. (b) How is OEM buyer performance assessed? 4. Roughly what percentage of suppliers is in the upper midwest? Abroad? (In terms of both number and value purchased materials) (a) Change over time? (b) What are the advantages to having suppliers nearby? 5. What percentage of suppliers are considered ‘strategic’? (a) How do you assess supplier performance? 6. What percentage of suppliers are ‘sole source’? (a) When multiple sourcing, what is the typical number of suppliers? 7. Relationship with key suppliers (b) How frequently do you communicate with your key suppliers? (i) What form does that communication take? (ii) Do any suppliers keep staV ‘on-site’? Do you keep staV on-site at suppliers? (c) How much lead do you give your suppliers on orders? (d) Do you maintain ‘long-term’ contracts with key suppliers (or Master Supplier Agreements)? On what terms? (e) Has ‘trust’ with suppliers improved or worsened over time? Explain ( f ) What types of information do key suppliers provide about the process used to make their products? Has this changed over time? (g) How are price increases/changes addressed? (i) Do you demand annual price reductions? (a) Are these across the board? Do you help to meet particular cost targets? (ii) If you are able to engineer out costs, are savings shared? (a) Do suppliers dispute the amount of cost reduction justiWed by particular changes? (iii) If a competing supplier underbids your long-time supplier, promising equal quality, how do you react? (iv) When there are changes in material prices to your suppliers, are they allowed to pass these costs along? (v) If suppliers oVer ideas that would allow them to reduce costs if you slightly modify procedures, how do you react? Examples? (h) Do suppliers commit speciWc assets or resources to you? (speciWc examples) (a) Are they able to use these investments—at least in part—for other customers? 9. Have your suppliers been ‘moving up’ the supply chain? (again examples) (a) Are some of them adding additional production processes or steps (e.g. machining or assembly in the case of foundries or stamping plants) to increase value added or to satisfy your requests? (b) Have some taken on responsibility for supplying complete systems, modules, or subassemblies, including parts procured from other companies?

Appendix

171

(c) Has your Wrm taken the lead in attempting to ‘tier’ by asking particular suppliers to take direct responsibility for subsuppliers that previously reported to you? (i) If so, do you indicate the subsuppliers they must use? 10. Do you expect your suppliers to provide more product engineering and design than in the past? (a) Are your suppliers able to do this? (b) What impact do such added services have on costs? (c) Impact on your ability to change suppliers if costs run out of control? 11. Do you rely more on suppliers for technology, new ideas, information than you did in the past? 12. Do you expect your suppliers to work more to your own schedule than they used to? (a) Does this amount to them ‘eating’ cyclical Xuctuation, or are they able to balance your needs with other customers? 13. As you reduce your own inventory, are your suppliers able to reduce theirs as well? Or do they feel like you have just pushed inventory carrying costs down the supply chain? VII. Supplier development/continuous improvement 1. Do you have a formalized supplier development or continuous improvement program? 2. What does ‘supplier development’ consist of? (more than simply certiWcation?) (a) Do you provide suppliers with training? (b) Do you work with suppliers to reduce their costs? (i) How are the beneWts shared? (c) Examples, of both success and failure? 3. Does your supplier development program require suppliers to share sensitive information? (a) Are they resistant to sharing this information? (b) How do you assure them that this information will not be ‘misused.’ 4. Are there external training providers or other sources of market or technological information that are particularly important to your supply base?

A.2.2. Advanced Manufacturing Project: Supplier Interview Guide I.

The basics of the Wrm 1. A brief history of the Wrm? 2. What are your main products? 3. Why are you here (geographically)? What—if anything—ties you to this region? (i.e. labor force, customer base, supply base, links with MECs, schools, etc.). 4. What is the ownership structure of the company? (a) (If a division) What kinds of decisions are made at this plant? (i) Are you expected to meet particular proWt targets? 5. Sales? Sales three years ago? 6. How many employees? (a) What percentage are production workers? (b) How many engineers/highly skilled trades/design staV, etc.

172

Appendix

II. Products, markets, and strategy 1. What end-user industries do you supply? 2. Who (and how many) are your main customers? (a) What percentage of your overall output goes to your top few customers? 3. For your key or representative product please describe the key steps in producing it. (a) What are typical production volumes for a representative product? (range of variation?) 4. How is the design for your products generally developed? (i.e. co-design, buildto-blueprint?) If you design your own, how did you come to have this capability? 5. Do you sell proprietary products? What percentage? 6. Do you design your own production processes? If so, who is involved (i.e. operators, skilled trades, engineers, etc.)? 7. Do you make modiWcations to your products (rather than work to customer prints) or processes on your own initiative? Do you provide any ‘unusual’ services (e.g. fast delivery, customization) that make your Wrm distinctive? 8. What are your most important sources of new ideas about your product and processes? (tech staV, operators, links to WMEP, to customers, etc.). Give examples. 9. Why are you here (geographically)? What—if anything—ties you to this region? (i.e. labor force, customer base, supply base, links with MECs, schools, etc.). 10. In this region, are there external training providers or other sources of market or technological information that are particularly important? III. Performance measurement 1. What are your key performance criteria and how are they monitored? (a) What measures are used? 2. Have they been improving/stable/falling? IV. Technology and quality 1. Over the last three years have you (a) introduced substantially more productive new equipment? (e.g. computercontrolled machines) (b) made major changes in your plant layout? 2. Have you been constrained in your ability to buy new technology because of lack of access to credit? 3. Do you receive assistance from your customer in implementing new technology, improving quality, and/or reducing inventory? 4. Have you changed the way that information Xows through the plant, for example by adopting ERP (Enterprise Resource Planning), JIT, etc.? 5. As a result of adopting new technology, have you changed the way you organize work in this plant? V. Work organization 1. Over the last 3 years have you (a) given production workers new responsibilities? (i.e. maintaining own production schedules, communicating directly with customers or suppliers, stopping a production run under new circumstances, increased cross-training)

Appendix

173

2. Created (or changed the role of) oV-line teams, such as quality circles or labormanagement committees? 3. Do you have mechanisms for workers to contribute their ideas (suggestion program, quality circles)? Are these active? What are some examples of ideas gained through these programs? 4. Over the past three years, has there been an increase in the skills, knowledge, and abilities you expect from your production workers? In what ways? 5. Have any of these changes (new technology, skill requirements, work org) been in tension with pre-existing job classiWcation or compensation systems or created other diYculties? (a) Have you (either as a consequence or not) (i) reduced the number of job classiWcations? (ii) introduced a variable compensation system? (gainsharing, pay-for-skill, pay-for-knowledge, proWt sharing, pay for performance, etc. (b) What has been the overall eVect on wages? (c) Has this generally created problems in terms of labor relations? VI. Workforce and training 1. Describe types of training and frequency (and who is trained)? (a) Who gets trained? Who decides (i.e. employee option, or manager decision). (b) How much of the training is speciWc to your company and how much more general? (c) How much training is given to new employees? (d) Would the training help employees Wnd another job if they quit or were laid oV? (e) Are their wage incentives for workers to take training? 2. Describe your experience with the Supplier Training Program? 3. How many courses have you sent people to in the last year? How many people in total? 4. How do you select the courses you use? (a) Have your OEM customers pointed you to any courses, or otherwise given guidance? 5. Have you been satisWed with the quality of instruction? 6. Have you been able to use the WMEP training to improve performance for customers who are not members of the STC? 7. Problems and suggestions for improvement? VII. Relations with Customers 1. Do customers in diVerent industries make very diVerent demands? 2. What advantage—if any—do you get from working for customers in multiple end-user industries? 3. Relationship with key customers (a) How frequently do you communicate with your key customers? (b) Do any of your customers have formal supplier development programs? (i) What do these programs consist of ? (Are they more than simply supplier certiWcation?) (ii) Do any of your customers provide you with training? Describe?

174

Appendix

4. 5. 6.

7.

(c) How would you rate the level of trust between you and your customers? How has it changed over time? (d) What types of information does your business unit provide to your customer about the process that you use to make their products? Has this changed over time? (e) Do your customers dictate which subsuppliers (including materials) you are to use? (f ) How are price increases/changes addressed? (i) Do customers demand annual price reductions? (a) Do they help to meet cost targets? (ii) When they (or you) engineer out costs, how are savings shared? (iii) How would your customer react if one of your competitors oVered a lower price for a product of equal quality? (iv) How would your customer react if your material suppliers raised their prices? (v) Suppose your business unit has an idea that would allow you to reduce your costs, but would require your customer to make a slight modiWcation in its procedures. How would your customer react? (g) Has your Wrm dedicated resources to a key customer, resources that you could not reuse if you lost their order? Have any of your customers re-bid your parts through online auctions? (a) If so, what happened? Do you ever ‘Wre’ important customers? Why? Have you been moving up the supply chain? (a) Have you introduced additional production processes or steps (e.g. machining or assembly in the case of foundries or stamping plants) to increase value added or satisfy customer requests? (b) Have you taken on responsibility for supplying complete systems, modules, or subassemblies, including parts procured from other companies? Are OEMs asking you to do more product engineering or design? Are you able to do it? (a) Do you get better margins as a result? (b) Who owns half-Wnished designs on joint projects? (c) Do you feel you learn to trust your customer more by engaging in joint quality control and product design activities? (d) Does OEM desire to reduce inventory lead suppliers to hold it? Have you reduced your own inventory over the last three years? If not, why not?

A.2.3. Advanced Manufacturing Project: Preinterview Facility Data ProWle This questionnaire solicited some initial information from suppliers. The questions are listed here. It began: This questionnaire solicits information that helps us make the best use of our limited interview time at your facility. It will also allow us to benchmark you against a national sample of similar businesses. You will be provided with a copy of a customized, conWdential benchmarking report. Please try to report all of your data (Wnancial and other) for the same 12-month period, preferably calendar 2000 or a

Appendix

175

Wscal year ending on or after December 31, 2000. Data on your company will be pooled with information on other companies and not released publicly. Thank you in advance. 1. What do you make at this plant? Using what processes? What materials? For what types of customers? (You must provide a written description.) 2. Please enter your 4 digit primary SIC code or 6-digit NAIC, if you know it. 3. Is this facility part of a larger corporation that also operates other facilities? 4. Approximately what percent of your sales fall into each of the following categories? (a) Engineer-to-order (one-of-a-kind items: dies, prototypes, custom machines, etc.) (b) Job shop products and services (short-term or nonrepeating orders) (c) Make-to-order jobs run regularly (long-term, repeating orders) (d ) Make-to-stock work 5. Many shops are encountering pressure from some or all of their customers to reduce prices. Over the past year, have you had to reduce your prices, on average: (check only one) (a) Not at all? (have been able to raise prices) (b) Not at all? (have been able to hold the line) (c) Somewhat? (less than 5 percent) (d) SigniWcantly? (5–10 percent) (e) A great deal? (more than 10 percent) 6. In the past year, approximately what percentage of your sales were from jobs where you designed the part or assembly? (Makers of large, one-to-an-order products should enter the percent of sales from design-and-build jobs.) 7. Approximately what percent of your sales were accounted for by products shipped as assemblies of two or more parts? 8. Approximately what percent of sales were to Wnal consumers, wholesalers, or retailers (as opposed to other manufacturers)? 9. Approximately what percent of your sales were accounted for by products not oVered three years ago? 10. Approximately what percent of your sales were to customers not served three years ago? 11. What were your total sales at this location in the past year? 12. Roughly what percent of your sales (from Q11) went to cover your outside purchases—material, parts, shop supplies, utilities, outside contracting (excluding temp agencies), rent and insurance, freight, and sales commissions paid to reps? 13. Of your suppliers of material, parts, and components, roughly what percentage are local (shipping from less than 50 miles away), what percent are elsewhere in the United States and Canada, and what percentage are in Mexico or oVshore? 14. What percent of your sales (from Q11) were accounted for by orders received over the Internet? What percent of your purchases (from Q12) were made over the Internet? 15. How much inventory (raw, WIP and Wnished) did you have on hand at or close to the end of the year? 16. Roughly what percentage of that inventory was work-in-process? 17. On average in the past year, how many individuals worked at this location? Please include part-time and contract labor. (If your employment counts

176

Appendix

18. 19.

20. 21. 22. 23.

changed signiWcantly over the past year, please be careful to report AVERAGE, not year-end counts.) How many of your personnel (from Q17) were contractors or temps, rather than people on your own payroll? Over the past year, how many diVerent people did you have on your payroll for at least one day? That is, how many W2’s did you generate? (Note: Your answer should not be less than Q17 minus Q18.) How many of your personnel (from Q17) were hourly people who worked on the shop Xoor? On average in the past year, how many hours per week did these shop workers (from Q20) work? How many of your shop Xoor workers (from Q20) were represented by labor unions? How many of your shop Xoor workers (from Q20) are skilled, and how many production? Roughly how much per hour could an employee in each group expect to earn with three years’ service at this facility? What would be a typical entry wage for a new production worker?

This page intentionally left blank

Notes 1

2

3

4

5

6

Data in this paragraph are from the Bureau of Labor Statistics. 1965 represents the postwar peak in terms of percentage of manufacturing jobs, and would be an alternative turning point in the deindustrialization of the economy. That is about the time that the rate of female labor participation began to increase rapidly, concurrent to an increase in employment in the service sector (between 1960 and 1965, the female participation rate increased by 4.2 percent; between 1965 and 1970, it increased by 10.1 percent). From my perspective, however, 1979 is a somewhat more interesting place to mark the ‘beginning’ of deindustrialization: not only is it the year in which there was the greatest number of manufacturing jobs, it is the year after which the percentage of the workforce employed in manufacturing sector has unflinchingly decreased (there was an increase in manufacturing as a percentage of American employment between 1978 and 1979). Rowthorn and Ramaswamy (1999: 34) estimate that north–south trade ‘has, on average, contributed less than 20 percent to the relative decline in manufacturing employment in the advanced economies’, and that ‘the impact of north–south trade on deindustrialization has been mainly through its effect in stimulating labor productivity in the manufacturing sector of advanced economies’. As Atkinson (2003: 2)—though openly a political progressive—writes: ‘Some argue that the real economy is about making things, while providing services is a second class activity; the reality is that our economic well being is determined by the productivity of the entire economy, not any one sector.’ Likewise, manufacturing does not, as some argue, necessarily pay more than services. The point is to hold onto those manufacturing jobs that do, and to recognize that the spatial concentration of manufacturing means that rapid declines can be very hard to cover with service sector expansion. As Atkinson also explains (with a reasoning I elaborate in the introduction to Part I), the health of manufacturing matters a great deal for the trade and the current account balance. For reasons that are explained later in the text, I at times treat the large first-tier automotive suppliers as one with the ‘OEMs’, as they are large enough and do enough of their own entirely original design that they have that role vis-a`-vis smaller suppliers. Author’s elaboration from BLS publication, June 23, 2004: ‘Comparative Civilian Labor Force Statistics, Ten Countries, 1959–2003’. Data for Korea from OECD, ISIC revision 2, ratio of MþF employees in mfg (ISIC sector 3) to MþF employees in all sectors. 1980 is used as cutoff year as Germany did not release sufficiently disaggregated industrial employment statistics in the 1970s. Korean manufacturing employment for 1980 is actually all industrial employment (manufacturing employment is a subset of all industrial employment) because manufacturing employment was not disaggregated in the OECD statistics for Korea until the mid-1980s. For subsequent years, only manufacturing employment is counted, meaning that the rise between 1980 and 1989 was in fact even more striking. These central cities—which employed as of 1979 less than 30 percent of manufacturing workers (BEA regional accounts data)—are: Akron; Allentown–Bethlehem;

Notes

179

Baltimore City; Buffalo; Chicago–Gary–Hammond; Cincinnati; Cleveland City; Detroit–Flint; Indianapolis; Milwaukee; New York City–Newark–Jersey City; Philadelphia; Pittsburgh City; Providence–Warwick; St. Louis; Toledo. 7 My elaboration of data from the 1977 and 1997 Economic Censuses (in 1977, called the ‘Census of Manufactures’; establishment size data from the 2002 economic census were not available at time of writing). The 1977 data are aggregated using SIC classifications, whereas by 1997 the Census Bureau had switched to NAICS. This shift did result in some establishments moving between classification groups, but the differences within manufacturing sectors are quite minimal for the aggregate levels at which I am comparing here. Even if the data to make adjustments were readily available, they would not affect the analysis in any substantive way. Note also that this trend has occurred across developed economies (see, e.g., Loveman and Sengenberger 1991). 8 For the most famous statement of the roots of the crisis of mass production, see Piore and Sabel’s (1984) The Second Industrial Divide. It has been well documented that there has been a general worldwide trend to smaller production units and that this is driven to no small degree by a relative deverticalization of production. On the general trends and a highly informative debate over its historic meaning, see the 2004 special issue of Enterprise and Society, with contributions by Langlois, by Lamoreaux, Raff, and Temin, and by Sabel and Zeitlin. Examples of econometric studies include Brynjolfsson et al. (1994) and Feenstra (1998). 9 On the rural/urban split, see Manufacturing and Technology News, May 2, 2003, as well as Wilkerson (2001). The claim about union job losses is drawn from the Bureau of Labor Statistics, Current Population Survey. An adequate accounting of the employment performance of firms of different size classes must await the release of establishment size data from the 2002 economic census (unavailable at the time of writing). However, data from County Business Patterns on the number of establishments by size class does suggest a relatively widespread and secular decline. If one makes the not unreasonable assumption that the average firm size within each reported size class (1–19, 20–99, 100–499, 500þ) will have remained roughly constant since the 1997 economic census, one can then infer the degree to which the job losses have been distributed across size classes. Performing this calculation with data from the 2002 County Business Patterns suggests that 52 percent of the jobs have been lost due to the disappearance of establishments with less than 500 employees. Firms in this size class employed 61 percent of manufacturing employees in 1997. Compared to the previous two downturns, this is a trivial reapportionment across firm size classifications. 10 These two terms are often used in confusing ways in public discussion, with ‘outsourcing’ coming to refer to the movement of service production abroad (whether or not this involves their externalization from a previously integrated process). In this book, when it is relevant to make clear that intermediate goods are purchased outside the United States, I will write offshoring or offshore outsourcing. When I write outsourcing, I hew to its older and more common usage in manufacturing industries. I mean the moment when particular operations are first subcontracted, so long as the intent is that the change be long-term. Note also that the distinction between outsourcing and subcontracting can be both loaded and confusing. Strictly speaking something is ‘outsourced’ only when the outsourcer

180

11

12

13

14

15

16

17

Notes retains the ability to bring it back in without too much trouble, but in manufacturing in the last two decades operations have for the most part moved in one direction only, leading to a substantial and long-term shift of productive capacity and responsibilities from large vertically integrated manufacturers to supplier firms. For an effort to estimate the degree to which firms themselves have relocated, see Bronfenbrenner and Luce (2004: 47) who argue that the ‘overwhelming majority of companies that shifted production out of the US between January–March 2004 were ultimately owned by extremely large, profitable, US-based, publicly-held multinationals’. A full analysis of the degree to which the escalating goods trade deficit is due to increasing imports of intermediate goods is beyond the scope of this introduction, where my primary purpose is simply to make the prima facie case that there are real changes in the organization of American manufacturing and that we should look closely at those changes. The examples given are drawn from industry sector data provided by the American International Trade Administration (http://ita.doc.gov/td/ industry/otea/industry_sector/tables_naics.htm). The dollar’s value relative to the yen and euro fell dramatically. Its value relative to the Chinese Yuan remained stable, as the exchange rate is maintained by the Chinese government (at least as at the time of writing). I do not mean to suggest that the relative value of the dollar is unimportant for manufacturers. This is a vast literature. A representative but inexhaustive list of well-known works mixing sociology and political economy (as I do in this book) might include Piore and Sabel (1984), Sabel (1989), Hirst and Zeitlin (1991), Streeck (1991), Pyke and Sengenberger (1992), and Herrigel (1996). See also Harrison (1994) and Amin and Thrift (1992) for a critical counterpoint; they agree that the conceptual apparatus in the debate derived heavily from Italy, Germany and Japan, but dispute the others cited here as to the implications and proper characterization of the spread of flexible production models. Particularly notable contributions in economics and business studies include Milgrom and Roberts (1988; 1990), and Prahalad and Hamel (1990). Even the counterclaims of those who dispute the extent of change in the United States do not suggest that there has been no change, but are rather claims about the relative ability of different systems to accept particular institutional modifications (see especially Hall and Soskice 2001). This issue will be discussed at length in Chapter 2 and in Part III. See also Smith (1997: 318): ‘Japan’s economic success inspired widespread debate about the feasibility of and limitations of importing into the American context, the Japanese lean production techniques.’ This book is about changes in relationships between manufacturing firms in the last quarter century have important implications for both social theory and economic development policy. This is hardly the only approach one can take, and I readily concede that a concentration on the interfirm relationship is not the best approach for all questions, premised as it is on a characterization of organizations—made up of people, but not people per se (with an ensuing inattention to issues of class)—as embedded in social and institutional contexts. Examples of related work responding to changes in late twentieth-century capitalism that take different approaches include especially Smith’s (2001) Crossing the Great Divide, and Osterman’s (1999) Securing Prosperity. They tackle more directly than do I the theoretical and policy

Notes

18

19

20 21 22

23

24

181

implications of changes internal to firms and their meaning for workers making arguments and recommendations that pivot on issues relevant but not core to my story. Peck’s (1996: 264) Work-Place takes a more macro-systemic approach attentive to issues of class, and speaks to the changing—and very spatial—dynamics of labor-market regulation under conditions of neoliberal hegemony, arguing that the contrast between globally mobile capital and locally rooted labor has undermined the ‘regulatory advances made by and conceded to labor during the ‘‘golden age’’ of Fordist–Keynesian expansion’ but also that ‘global deregulation is a political project requiring political responses’. And Dore’s (2000) Stock Market Capitalism: Welfare Capitalism and Fligstein’s (2001) The Architecture of Markets show the importance of changing patterns of ownership and control, aspects of the reorganization of the economy which are not central foci of this book (though they are certainly referenced when relevant). This is supported by Helper and Sako’s (1995; 1998) findings of a partial convergence of American customer–supplier relations on the ‘voice’ model associated with Japanese subcontracting, though they emphasize that this is partial, and that there remain significant differences. Many also hold that the various aspects of the ‘full’ Japanese model can be decoupled from each other and the cultural context, as firms see them as tools to solve their own problems. Sabel (1996) comments, for example, that Japanese main-bank corporate governance with its patient capital seems an unlikely arrival on American shores, a claim verified by Fligstein and Freeland’s (1995) review of the corporate governance literature denying any convergence on a single type. See, for example, Dyer (2000); Helper (1991); Helper and Sako (1995); Helper and MacDuffie (1999); and Helper et al. (2000). See especially the September 2004 special issue of Enterprise and Society (with contributions by Langlois, by Lamoreaux Raff and Temin, and by Sabel and Zeitlin). The best known of these are the ‘Lean’ system associated with the International Motor Vehicle Project (MacDuffie 1995; Womack et al. 1991) and the more unionfriendly ‘High Performance Work Organization’ (Appelbaum et al. 2000). There are substantial and interesting debates over what should count as ‘high performance’ and the degree to which they really require worker empowerment—but there is consensus on the basic point I am making here, which is that there is a better way, but that better way is, at best, only partially realized (see Babson 1995). Kelley and Harrison (1992: 250), in a relatively early and well-designed study, find evidence that employee participation programs in nonunion settings do not lead to improvements. However, ‘by contrast, among branch plants of more complex organizations, unionized operations are significantly more efficient than operations (plants) that use the same technology, are organized similarly, make the same product(s), pay the same average wages, and require the same skills—but have no unions.’ They suggest this is primarily due to better job flexibility and mechanisms for negotiating job redesign. Note that Florida (1996: 317, italics added) argues—as will I later in the book using my own empirical findings—that the industrial upper Midwest has ‘been inaccurately portrayed as a region undergoing long-term secular and irreversible economic decline, based to a large degree on its outmoded Fordist organizational configuration and inability to inculcate new and more advanced forms of production.’

182 25

26

27 28

29

30

31

32

Notes The 2004 piece in Rationality and Society is a slightly narrower version of the argument first published by the Max Planck society in 2002. I use quotations in their more recent version for the most part, using only one quote from the longer version as it offers a succinct summation of a point that I want to draw on in my own argument. The ‘polya-urn’ process used to model path dependence is a simple game in which players draw balls of varying colors randomly from an urn. To model increasing returns and path dependence, ‘every time a ball of a particular color is pulled from the urn, it is returned, and a further ball of the same color is added to it’. Hypothesizing a Bayesian actor, Crouch and Farrell (2004: 9) show that minor changes to the game are sufficient to permit actors to escape the chains of lock-in. The example modifications they use are intended to model ‘such processes as: the use of past or redundant institutional repertoires; transfer of experience across action spaces; or from other agents, through networks of structured relationships; the emergence of perceived ‘‘one best’’ solutions’. For an excellent short critical summary of contractual perspectives, see Foss (1993: 129–32). My characterization here relies heavily on his summary of the issues. A definition of asset specificity is provided by Chandler (1992: 85): ‘The specialized physical facilities and human skills that, because they can only be used in the production of specific products and services, lose value if deployed in other activities.’ Williamson defines at least six types of asset specificity: site specificity; physical asset specificity; human asset specificity; brand names; dedicated assets; temporal specificity. This is by the ‘folk’ theorem (found in any microeconomics text, such as, for example Mas-Colell et al. 1995). It shows that in a repeated prisoners’ dilemma, ‘both cooperate’ is a possible equilibrium given sufficient future discounting. Williamson (1985) famously wrote that the distribution of organizational forms should be expected to be ‘thick in the tails’, dominated by the relatively pure dichotomous types of market and hierarchy, with networks a hybrid form. Williamson (1991) later essentially accepted that the intermediate form is more common than he had previously thought, but held that such forms were a poor fit for highly uncertain environments. Both of these assertions have been disputed as a poor fit for the empirical reality. This is shown in detail below. See in particular Holmstro¨m and Roberts (1998) and Powell (1990). The evolutionary perspective is not always in direct contradiction with the contractual literature, but it consistently holds that the exclusivist logic of TCE misses that the real action on firm boundary decisions is found in the distribution of productive knowledge (Winter 1993). For the classic statement on the evolutionary perspective, see Nelson and Winter (1982). Foss (1999: 732) makes it clear that he intends all the standard theories, writing: Whatever their differences may be, one heuristic is characteristic of the various contribution streams that constitute the modern economics of organization: an overriding emphasis on conceptualizing virtually all problems of economic organization as problems of reducing incentive conflicts through ex ante contractual alignment and through ex post governance mechanisms.

Notes 33

34 35

36

37

38 39

40

41

183

While all are relatively hostile to TCE, the competence literature is accepted by Hall and Soskice, who explicitly incorporate it into their attempts to marry political economy to business studies; it also fits easily with the social embeddedness view. Sabel and collaborators, on the other hand, are critical of what they believe its toostrong emphasis on bounded rationality and the fixity of routines. On exit and voice, see the classic statement by Hirschman (1970). Note that Granovetter, without distancing himself from the concept of embeddedness per se, has distanced himself from some of the ways in which it has been used since his 1985 article (see Krippner et al. 2004). A more recent formulation of the same basic query is in Carruthers and Uzzi (2000: 487). In an article delineating what they see to be the big questions that economic sociology should confront in the ‘new millennium’, they ask: as ‘relationships become more rather than less important, how are the capabilities to build relationships, trust, reciprocal obligation, private information transfer, and other features of embedded ties developed?’ Amable (2000: 661–4) cites seven distinct approaches: (a) Vintage re´gulation (Aglietta, Boyer, Lipietz, Coriat, Mistral, Petit); (b) Social structures of accumulation (Bowles, Gordon, Weisskopf); (c) National systems of innovation (Lundvall, Nelson, Freeman); (d) Social systems of production (Streeck, Hollingsworth, Sabel); (e) Varieties of Capitalism (Soskice, Hall); ( f ) Comparative institutional analysis (Aoki); (g) Social systems of innovation and production (Amable, Barre´, Boyer). I rely in my exposition largely on the introductory essay by Hall and Soskice, as that is the clearest general statement of the approach. See also Soskice (1999). Hall and Soskice (2001: 3–5) recognize their indebtedness to previous perspectives on institutional variation in the study of comparative capitalism—from the modernization theories of the 1960s to the neocorporatism of the 1970s and especially to the social systems of production (SSP) approaches of the 1980s and 1990s. The second (neocorporatist) of these directed attention to the organization of society, though overemphasizing the trade union. The SSP emphasized ‘the movement of firms away from mass production toward new production regimes that depend on collective institutions at the regional, sectoral, or national level’ and brought a ‘wider range of institutions into the analysis and [adopted] a more sociological approach, stressing the ways in which institutions generate trust or enhance learning within economic communities.’ Following North, they define institutions as ‘a set of rules, formal or informal, that actors generally follow, whether for normative, cognitive, or material reasons, and organizations as durable entities with formally recognized members, whose rules also contribute to the institutions of the political economy’ (Hall and Soskice 2001: 9). By focusing on how these impact strategic interactions, Hall and Soskice (2001: 5) believe they can generate ‘an analysis that focuses on some of the same institutions others have identified as important but construes the impact of those institutions differently as well as one that highlights other institutions not yet given enough attention in studies of comparative capitalism.’ The countries they argue to be easily classified as LMEs are: Australia, Canada, Ireland, New Zealand, United Kingdom, and the United States. CMEs would be: Austria, Belgium, Denmark, Finland, Iceland, Germany, Japan, Netherlands, Norway, Sweden, and Switzerland. Italy, France, Spain, Portugal, Greece and Turkey are

184

42 43

44

45

46

Notes in an ‘ambiguous position’, perhaps constituting a third ‘Mediterranean’ capitalism ‘marked by a large agrarian sector and recent histories of capacities for non-market coordination in the sphere of corporate finance but more liberal arrangements in the sphere of labor relations’ (Hall and Soskice 2001: 21). The distinction between radical and incremental innovation and its implications is discussed in Chapter 3. Hall and Soskice (2001) continue, giving examples: ‘These include regional development schemes based on tax incentives, vocational programs focused on formal instruction in marketable skills, and government subsidies for basic research. Because of the bluntness of the instruments available to states and the importance of markets to these economies, deregulation of various kinds is often the most effective way to improve coordination in LMEs.’ See especially Helper and Sako’s (1998) empirical work derived from their surveys of automotive suppliers, along with Helper and Sako (1995); Helper and MacDuffie (1999); and MacDuffie (1997). Sabel (2004b) goes well beyond the auto industry for his examples, describing also organizations making computer parts and metal components, organizations delivering public services, ‘high reliability organizations’, and more. For example, when an OEM gives up the internal capacity on a particular process, they may fall behind technologically and become unable to design new products without the aid of suppliers. Helper, MacDuffie, and Sabel (2000: 468) write: Benchmarking, simultaneous engineering and error detection methods like the ‘five why’s’ are procedures for doing just what the standard view of effective action given bounded rationality says cannot be done: routinely questioning the suitability of current routines. Whether in the initial specification of new designs (benchmarking) the concretization of these approximations (simultaneous engineering) or in the course of their practical application (error detection), this disciplined inquiry of routines occurs at just those times when self-interrogation seems most valuable but most difficult.

47

48

This is not to say that functionality cannot be part of the story, especially insofar as actors often actively seek to establish organizational forms that they perceive to be functional. I mean quite precisely that it should not be assumed without a causal explanation of either why the market in question was a good natural selector, or of how actors built the organizational form in question. In Part II, I will devote considerable space to the argument that there are considerable constraints to the diffusion of even functional network forms. Podolny and Page also mention as a possibility the effect of power on the diffusion of network forms, but cite only a 1990 study of relationships between corporations and investment bankers by Wayne Baker. Baker’s study finds an association between a transactional orientation and greater corporate power, and that efforts to exploit network ties are tendentially the province of weaker corporations. Although Podolny and Page’s claim, in citing only Baker, is not obviously applicable to the context of production discussed here, the underlying idea that powerful actors with easy exit options are less likely to find networks to their benefit is in fact quite consistent with the argument of Helper et al. that firms become more likely to turn to the innovative potential of decentralization and learning by monitoring as market

Notes

49

50 51

52

53

54

55

56

57

58

59

185

competition reduces the final market rents available to them (a proxy for market power). On this, I find the arguments of Hall and Soskice to be especially problematic, since they believe their descriptions to have strong prescriptive (policy) implications. This aspect of their position and my rejection of it are developed in Part III. See Wired Techology News, September 1997. In the interests of full disclosure: Herrigel (2004) argues Whitford and Zeitlin (2004) to be guilty of a similarly ‘hard’ dichotomy. I am now much more careful to speak only of a relative shift. And perhaps more importantly, understanding the determinants of that systematicity can reveal levers for the stimulation of more positive sum collaboration in manufacturing through focused public policy, a normatively desirable goal for many stakeholders beyond the firms’ owners themselves. For purposes of these statistics, I am defining durable metal manufacturing as primary metal, metal fabrication, machinery, electrical equipment and appliances, and transportation equipment. The data is drawn from the Bureau of Economic Analysis Regional Accounts series for 2003. See Luria (2002). See also proposal to the Alfred P. Sloan Foundation (2000): ‘Component Manufacturing: Creating an Advanced Manufacturing Sector’. The statistics in this subsection are using SIC codes 308 (plastic products), 346 (forgings and stampings), 359 (pistons, valves, small motors), and 3714 (motor vehicle parts) as illustrative of the sector in question. Industrial classifications do not perfectly capture all the firms that ought to be included in this sector, but they do give a reasonable approximation for purposes of providing aggregate numbers. All of the data in this paragraph are drawn from work done by Dan Luria of the Advanced Manufacturing Project. Data sources for the quantitative material are the US Bureau of the Census, County Business Patterns and 1992 input–output tables. To contextualize the passage: ‘Market to book ratios continue to increase; the relationship between financial performance and stock price is decreasing; physical assets have become a liability.’ Veloso adds that ‘This is particularly relevant for complex systems or modules such as an ABS, where it is assumed that the supplier is able to spread its development cost across several clients (assemblers).’ Another major trend in the automotive industry is for production plants to follow the market geographically, with assembly done relatively close to the eventual point of sale. Obviously, this strategy is not without risks, however, and there remains variation in the degree both to which assemblers have devolved full responsibilities to suppliers, and to which they have been willing to utilize ‘sole-source’ supply relationships. Assemblers that devolve full production responsibilities give up significant power over the supply chain and are likely to fall behind in the production knowledge that would allow them to bring component production back inside should they become dissatisfied with their suppliers (Veloso 2000). As a general stylized fact, it is easy to speak of the supply structure as completely tiered, entirely organized through coordinating first tiers. This is, however, only an approximation. The structure is much more tiered than twenty years ago, but there are many firms that operate at both the first and second tiers as well as firms

186

60

61 62

63

64

65

66

67 68

Notes producing single parts that are not integrated into subsystems by other suppliers that deliver directly to the assemblers. The assemblers also often maintain a strong voice in the selection of lower tier suppliers, complicating the claim that these are the supplier to the first tier etc. I will reference these complications only when they are relevant to my argument, but recognize that as a matter of empirical fact, the industry is not a perfectly rationalized pyramid of upward flows. This is not a marginal drop in the number of direct suppliers. Interviewed OEMs had reduced their supply bases to as small as a quarter the size of ten years previous (i.e. from 1200 to 350 in one case, with an expectation of reducing that further; from 750 to 250 in another case, from 1100 to 250 in still another). This is a very consistent pattern. This argument is fully developed in Part III. My characterization of modularization and what it means for component manufacture borrows very heavily from the analysis of Gary Herrigel (2004). Note also that Herrigel and I are in many cases working from the same empirical data. We did several interviews together for the Advanced Manufacturing Project, and, also as a part of that project, have shared several raw transcripts of interviews that we have done individually or with others. In an analysis of Fiat’s modular strategy, Becker and Zirpoli (2003) argue that Fiat ceded its ability to integrate the knowledge of suppliers, thinking it would be enough to be able to integrate systems. Becker and Zirpoli argue that this has been disastrous, a claim bolstered by Fiat’s current difficulties and the relatively weak embrace given the recently released Stilo model by even Italian consumers. Innovation, by definition, involves uncertainty, but is also affected by the investment decisions of the interested parties. Once an OEM decentralizes production, it must negotiate its flow even in rough terms. For example, an interviewed supplier of plastic components that operates at both the first and second tier of the automotive industry, though increasingly at the second tier, states its primary growth strategy to be edging out competitors through innovation and performance; the former through quality design and engineering out cost and the latter through adding new products and processes. They consider their design capabilities essential to maintaining business, and said, for example, that even as they are losing direct share to Daimler-Chrysler, they actually have a larger overall portion of the business, through sales growth to first tier suppliers to Daimler-Chrysler. Note also that one OEM’s ‘widget’ supplier or other process specialist can be another OEM’s very important specialist, depending on the characteristics of the final product. Speaking of a particular sort of die casting, an OEM purchasing manager commented that ‘There are some cases where there just aren’t [many suppliers]. . . . When it comes to [specialized process], there are only one or two in the world that do that, and that is a centerpiece of our product. So there are some areas of technology where we don’t have any choice.’ For another treatment of supplier strategy in the current production regime based on many of the same interviews, see Herrigel (2004). A pure specialization strategy is sustainable only if the suppliers’ customers continuously grow or if they are still in supply-chain trimming mode, so that the best specialists acquire competitors’ market shares.

Notes 69

70

71 72 73

74

75 76

77

78

187

The major categories (cost; quality; design and technology competency) in relation to which the needs and expectations of supplier firms in manufacturing have increased substantially in recent years can be measured separately (and are in OEM auditing of supplier performance). However, they are interrelated, and in many cases, improvements in each of the three areas will go hand in hand. For example, upgrading process technology or suggesting design improvements to render particular parts more easily manufactured (‘design for manufacture’) are a frequent means to both production cost reduction and to quality improvement. For example, the more integrated suppliers might be expected to keep engineers onsite at the customer, even as the same suppliers may have ongoing disputes with the purchasing division. And there is a wide range between letting your customer have full information on your costs, partial information, and not letting them in the door. The practice spread most famously through the auto industry, but is used in many industries. For the metaphor of the waltz to describe the process by which OEMs and suppliers search for joint cost reductions, I am indebted to Gary Herrigel. In auto, the general impression is that a combination of selection at the first tier and the earlier uptake of new manufacturing processes in the industry has left a relatively capable supply base, or at least one that usually has the financial and managerial resources to knowledgeably buy the help that they need (the second tier is another story). For a good study of the supplier development at the Japanese transplant assembler in the United States, see Sako (2004). On supplier development and its motivation and effectiveness in mid-volume durable manufacturing, see Mesquita and Brush (2001), and Rickert (1999). Note that suppliers do sometimes give up this information. Another supplier also laughed and said ‘Not if we can help it!’ to my question asking if they gave full costing data on every operation to customers. He said that his firm had given this sensitive information to a key customer as a part of a very successful joint process redesign to reduce production cycle times on a key subsystem the OEM was purchasing.’ This particular project is discussed in detail in Section 6.4. As an interviewed OEM manager noted: ‘Substantially, I think everyone’s perception was that you couldn’t trust a complex part in Mexico. You can.’ Source for Figure 2: the database of the Performance Benchmarking Service of the Michigan Manufacturing Technology Center (a partner in the Advanced Manufacturing Project). Figure prepared by Dan Luria. The shape of the curve for wages, from the same database, is almost identical. On the idea that transactions are embedded in relationships, see Granovetter (1985) for the classic sociological refutation of ‘undersocialized’ economic analysis. Note that the claim is that all economic action, not only relational contracting, is socially embedded. Accepting this general claim—which I believe common to more or less all of economic sociology—does not, of course, require accepting the full Granovetterian program. This should not surprise, as Gary Herrigel was also a part of the AMP project. We have done many interviews of component suppliers, both together and separately, and have shared several transcripts of interviews. I am even drawing on some interviews that he did without me in this text (see Appendix for details). We are in full agreement as to what sort of relationship is to be explained, and while there

188

79

80

81

82

83

84 85

86

Notes are differences in the factors we cite as pushing relationships off the arms-length and collaborative poles, our positions are complementary rather than competing. The differences are essentially questions of relative emphasis. Examples would be: Wall Street pressures on OEM share prices; the changing price of money; sudden changes in steel tariffs created by title 201; other regulatory changes (i.e. EPA rules); radical technological change; and so on. For a fuller typology and discussion of inter-organizational trust—using the categories of contractual, competence, and goodwill trust—see Helper and Sako (1998). Their concepts map roughly onto mine. These general categories describe the sorts of beliefs that would lead contracting parties to predict greater returns to collaboration (i.e. they are definitions). I do not mean a trust that is ‘in the air’ (with apologies to Alfred Marshall), but rather categories that enable us to talk about what sorts of things are likely to lead to more or less collaborative relationships between OEMs and suppliers. Some of Powell’s more recent work has taken a similar tack to Sabel, focusing on the likelihood that we will see the spread of a ‘new logic of organizing’. See Powell (2001) for a relatively general rendering of the argument. ‘Build-to-forecast’ (BTF) and ‘build-to-demand’ (BTD) strategies should be arrayed on a continuum. A pure BTF strategy would fix the number of units that would come off the line with some agreed upon advance notice, while a pure BTD strategy would produce only the units that had actually been sold. A modification of BTD that moves slightly towards a BTF would be to build to a relatively small buffer inventory that would then cushion market swings to some degree; the inventory would then be replenished regularly, which would allow for relatively accurate forecasts over short periods. A pure BTF with long advance forecasting makes planning production easy, but virtually guarantees an output that poorly predicts market demand, while a pure BTD risks an inability to react to market swings if production cycle times or capacity cannot be adjusted as quickly as demand fluctuates. On the roots of ‘post-Weberian’ conceptions of the organization, it makes little sense to cite lifetimes of work that have by now spawned massive literatures. Most famously in historic terms, I refer of course to Simon’s (1976) Administrative Behavior; to Cyert and March’s (1992) A Behavioral Theory of the Firm; and to March’s (1962) ‘The Business Firm as a Political Coalition’ in the Journal of Politics. For a reflection by March on ‘Continuity and Change in Theories of Organizational Action’, see his 1996 piece in the Administrative Science Quarterly. The relevant ‘buyer’ position is variously named. I am referring to the lead purchasing person managing the relationship. For reviews of the available business literature verifying that purchasing plays an increasingly strategic role and is increasingly incorporated into cross-functional teams, see Cousins and Spekman (2003) and Chen and Paulraj (2004). He also said that ‘I think there is probably a lot of competitive pressure out there right now in the industry. And, there’s a lot of overcapacity. And the rate of change in the manufacturing is so quick and so fast that there’s a lot of pressure to come to market and come to market with more features, with higher quality, and with lower prices. And I think maybe that sometimes your ethical standards may suffer to some degree

Notes

87 88

89

90

91

92

93

94 95

96

189

because of it.’ He said explicitly that this sort of shopping of ideas was unethical in his view. On this vicious cycle, see especially Luria (1996b; 2000a). By ‘existing sociological literature’, I mean particularly writers in the embeddedness tradition as well as institutionalists such as Hall and Soskice. These claims are established in Chapter Two. Likewise, see that chapter for my discussion of Sabel’s model of Learning by Monitoring in American manufacturing, as it is outlined in Helper et al.’s article (2000) on ‘Pragmatic Collaborations’. A variant of my views on Helper et al. can be found in Whitford and Zeitlin (2004). This is not to say that firms do not exercise influence; but it is a disorganized influence that is more likely to push only for particularistic favors or for deregulation, rather than for the maintenance of active state programs. On the possibilities of state and local institution building taking advantage of the fragmentation of the American federal structure, see especially Rogers and Luria (1999). On the underestimation of American federalism in general, see Lowi (1985). On the fragmentation of the American economic development policy arena, see Eberts and Erickcek (2002). I use these two works because they are very recent, are by prominent scholaractivists with a hands-on knowledge of the experiences discussed, and review a wide range of local-level experimentation with new institutions aiming to better coordinate the American labor market. For further treatment of the subject, see especially Osterman (1999) and Osterman, Kochan, Locke, and Piore (2001). Bartik (1995) notes that ‘many location and expansion decisions are unchanged by incentives. The cost of incentives to businesses whose location decisions are unchanged exceeds the taxes from businesses whose location decisions are changed. The average net governmental cost per job created by incentives is around $4,000 annually for the life of the plant.’ Bartik (2003a) estimates that some $20–30 billion is spent annually in the U.S. for economic development, most of which simply goes to tax incentives to attract branch plants. He strongly argues that more of this money should be focused toward ‘positive-sum’ economic development. Clusters, Porter (2000: 15) claims, ‘represent a new way of thinking about national, state, and local economies, and they necessitate new roles for companies, for various levels of government, and for other institutions in enhancing competitiveness.’ States that have consciously formed cluster initiatives (using that term) include Arizona, California, Connecticut, Illinois, Massachusetts, Minnesota, New York, North Carolina, Ohio, New Mexico, and Oregon—with many others in formation. See Porter (2000), Waits (2000), and Lashinsky (1992). These figures are current for 2001 fiscal year, from interview with personnel at WMEP responsible for the program. The MEP centers have consistently been shown to help the firms they serve to improve their productivity and were in 2004 named a finalist for the Kennedy School of Government’s Innovations in American Government Award (see Manufacturing and Technology News, May 4, 2004; p. 2). His ‘associative’ model is the Wisconsin Regional Training Partnership, discussed above. The ‘difficulty in promoting’ business/public sector cooperation in the WRTP case refers to struggles over regional training standards.

190 97

Notes

Besides John Deere, the founding OEMs included Harley-Davidson, Trane, CaseNew Holland, Ariens, and Mercury Marine. Oshkosh Truck joined the consortium in 2002, Kohler Corporation joined in 2003, and Mercury Marine left in 2003. 98 Small and medium sized enterprises, as defined by NIST, must employ less than 500 people (at the establishment in question). 99 Figures announced by the WMDC at the 2003 ‘kickoff’ meeting in Milwaukee. 100 This ‘honest broker’ role of public agencies has elsewhere been found to facilitate the success of collaborative partnerships among competitors in related areas of common interest such as joint research and development (Tripsas et al. 1995). 101 WMEP schedules the courses in numerous areas around the state and contracts with instructors to teach the courses, using personnel from the technical colleges and other independent instructors. Courses are often cancelled when enrollment is too low, but suppliers are not limited to the scheduled times and can arrange instead with the WMEP to hold a (nominated) course on-site. To increase program uptake, on-site courses are offered for a flat rate, permitting the supplier to bring as many people as desired (and allowing them to divide costs with other nominated firms by sharing classes). Perhaps unsurprisingly, on-site courses have proved the more popular option. For example, in the 1999 fiscal year (July 1999–June 2000), 36 classes were held as scheduled, while 106 took place on-site. Overall, 50 different companies sent 1,586 students (1,135 unique students) to 2,244 eight-hour training days, worth over $250,000 at market rates, though suppliers are charged only about half the market rate. The remainder was subsidized by a grant from the state budget, which stipulated that no supplier receive more than $20,000 in subsidies, while the sum total of suppliers nominated by a given OEM was not to exceed $100,000. This latter constraint was somewhat loose, because many of the suppliers work for multiple consortium OEMs, allowing their nomination to be shifted around when the cap became a problem. At the close of the first grant, the state began administering the money directly to WMEP, which then in turn provides the training subsidy directly from its own budget, benefiting as well from NIST’s 33 percent match for every dollar spent by the manufacturing extension programs. They have dropped the spending caps, but retain the requirement that subsidies go only to SME suppliers (WMEP, like all NIST manufacturing extensions, is required to serve the SME population). As of May 2003, some 192 different suppliers had used training, sending 7,144 students (as reported at the WMDC 2003 kickoff meeting). 102 Parts per million (ppm) is the typical unit of measure for parts that are defective or delivered early or late (both are considered delivery problems). 103 In policy writing published by the Center on Wisconsin Strategy, I have written extensively on the concrete effects and usages of the training program, especially as regards concrete improvements on manufacturing metrics and also direct workforce effects. Because my focus here is particularly on the relational aspects of the consortial model (presented in detail below), I have chosen not to fully document these effects here, and limit myself to assertion and a few examples. For much more data on concrete performance effects that can be ascribed to the WMDC, see especially Whitford et al. (2000) as well as Rickert et al. (2000).

Notes 104 105

106

107

108

109

110

111 112

113

191

Using $228,410 worth of services, as of January 18, 2002 (from an email from WMEP staff to the author). ‘Takt’ comes from the German term for musical meter, but, via Japan, refers to the cycle time for each operation. Kanban is also from Japanese manufacturing practices, and refers to a card that is used to signal when parts are to be transferred from one department to another and to automatically reorder products using minimum and maximum inventory levels. Harley-Davidson spearheaded the formulation of another training consortium (with different OEMs) around its operations in York, PA in cooperation with MANTEC, an MEP in that state. Each of these four ways in which the consortial model provides a collaborative resolution to problems in economic development applies as well, even more strongly, to the more established ‘parallel’ case of the Wisconsin Regional Training Partnership discussed in Section 6.2 (with the particulars modified to fit the case, of course). It would also likely require a greater focus of, and financing by, the state’s economic development infrastructure towards strategies aimed specifically at fomenting inter-firm collaboration. See Ericksen et al. (2002) for a proposal on how to deepen the impact and widen the reach of the consortial approach to supplier upgrading as part of a broader reorientation of Wisconsin’s economic development policies. I am grateful to Gary Herrigel for making me aware of this particular German association of component manufacturers. Again, I emphasize that my arguments regarding the space for institutional construction in the LME context are feasibility claims. I hold neither that the United States begins from an ex ante state of strong business associations or that building such associations is easy or inevitable, even if they would be functional. Indeed, the WMDC has developed in very small increments. But it has developed. I am indebted to Jonathan Zeitlin for the idea of a ‘code of supplier relations practice’. It is also discussed in our joint paper (Whitford and Zeitlin 2004). For a related approach to the improvement of domestic and international labor standards through third-party monitoring and certification of corporate codes of conduct, see Fung et al. (2001). On ‘high-road’ economic development strategies, see especially Rogers and Luria (1999). The $1.5 million was then matched by 33 percent from NIST, with the combined total making up about 25 percent of WMEP’s budget in FY2002 (from an email from a WMEP board member, Feb. 2003). Note that only a small portion of this money directly subsidizes the supplier training program. The bulk of it goes into the general operation of the organization. Recent and fascinating work by Berk and Schneiberg (2004: 1) is suggestive in this regard. They show that in the first twenty-five years of the twentieth century, many former price cartels evolved into ‘developmental associations’ which, they argue, ‘institutionalized collaborative alternatives to mass production in the midst of an economy typically thought to be dominated by markets and corporate hierarchies’. Their findings, they write, ‘revise conventional institutionalist assumptions about order and agency to make room for institutional diversity and actors’ capacities for reflexivity and learning’.

192 114

115

116 117

Notes I am attaching the names of particularly prominent authors to identify the approaches—but the approaches per se do certainly go well beyond the writings of just the particular authors named here (see Chapter 2 for a full discussion). Over the course of the years I was doing this field research, my understanding was informed by so many varied sources, it is impossible to list them all. But an ‘understanding’, in any case, is independent of reproducibility of data, and all the quotes and examples from my fieldwork that are then used to defend my claims are from the interviews listed at the end of the appendix. Some of this data is published in two policy reports published by the Center on Wisconsin Strategy (COWS). See Rickert et al. (2000) and Whitford et al. (2000). I have summaries of Rickert’s earlier interviews. Thus, all CMs interviewed in phase one were interviewed at least twice.

This page intentionally left blank

References Alichian, Armen and Harold Demsetz (1972). ‘Production, Information Costs, and Economic Organization’, American Economic Review, 62(5) 775–95. Amable, Bruno (2000). ‘Institutional Complementarity and Diversity of Social Systems of Innovation and Production’, Review of International Political Economy, 7(4): 645–87. Amin, Ash and Nigel Thrift (1992). ‘Neo-Marshallian Nodes in Global Networks’, International Journal of Urban and Regional Research, 16(4): 571–7. Appelbaum, Eileen, Thomas Bailey, Peter Berg, and Arne Kalleberg (2000). Manufacturing Advantage: Why High-Performance Work Systems Pay OV. Ithaca, NY: Cornell University Press. Arthur, W. Brian (1994). Increasing Returns and Path Dependency in the Economy. Ann Arbor, MI: University of Michigan Press. Atkinson, Robert (2003). ‘The Bush Manufacturing Crisis’. Washington, DC: Progressive Policy Institute. Babson, Steve (ed.) (1995). Lean Production and Labor: Empowerment and Exploitation. Detroit, MI: Wayne State University Press. Baker, Wayne (1990). ‘Market Networks and Corporate Behavior’, American Journal of Sociology, 96(3): 589–625. Bartik, Timothy (1995). ‘Economic Development Incentive Wars’, Employment Research, 2(1), published online at www.upjohninst.org/publications/empres.html. —— (2003a). ‘Thoughts on American Manufacturing Decline and Revitalization’. Kalamazoo, MI: W.E. Upjohn Institute for Employment Research. —— (2003b). ‘Local Economic Development Policies’. Kalamazoo, MI: W.E. Upjohn Institute for Employment Research. Becker, Markus and Francesco Zirpoli (2003). ‘Organizing New Product Development: Knowledge Hollowing-Out and Knowledge Integration—The FIAT Auto Case’, International Journal of Operations and Production Managements, 23(9): 1033–61. Berk, Gerald and Marc Schneiberg (2004). ‘Varieties in Capitalism, Varieties of Association: Collaborative Learning in American Industry, 1900–1925’, Unpublished paper, University of Oregon Department of Political Science and Reed University Department of Sociology. Bivens, Josh (2004). ‘Shifting Blame for Manufacturing Job Loss: EVect of Rising Trade DeWcit Shouldn’t be Ignored’. Washington, DC: Economic Policy Institute. Boyer, Robert, Elsie Charron, Ulrich Ju¨rgens, and Steven Tolliday (eds.) (1998). Between Imitation and Innovation: The Transfer and Hybridization of Productive Models in the International Automobile Industry. Oxford: Oxford University Press. Bronfenbrenner, Kate and Stephanie Luce (2004). ‘The Changing Nature of Corporate Global Restructuring: The Impact of Production Shifts on Jobs in the US, China, and Around the Globe’, Report submitted to the US–China Economic and Security Review Commission. Brynjolfsson, Erik, Thomas Malone, Vijay Gurbaxani, and Ajit Kambil (1994). ‘Does Information Technology Lead to Smaller Firms’, Management Science, 40(12): 1628–44.

References

195

Burroni, Luigi and Carlo Trigilia (2001). ‘Economic Development through Local Economics’, in Colin Crouch, Patrick Le Gale`s, Carlo Trigilia, and Helmut Voelzkow (eds.), Local Production Systems in Europe. Oxford: Oxford University Press. Cap Gemini Ernst and Young Center (2002). ‘Measuring the Future: The Value Creation Index’. Cambridge, MA: Cap Gemini Ernst and Young Center for Business Innovation. Caprile, Marı´a and Clara Llorens (2000). ‘Outsourcing and Industrial Relations in Motor Manufacturing’, European Industrial Relations OnLine. Carruthers, Bruce and Brian Uzzi (2000). ‘Economic Sociology in the New Millennium’, Contemporary Sociology, 29(3): 486–94. Casper, Steven (2001). ‘The Legal Framework for Corporate Governance: The InXuence of Contract Law on Company Strategies in Germany and the United States’, in Peter Hall and David Soskice (eds.), Varieties of Capitalism: The Institutional Foundations of Comparative Advantage. Oxford: Oxford University Press. Chandler, Alfred D., Jr (1992). ‘Organizational Capabilities and the Economic History of the Industrial Enterprise’, Journal of Economic Perspectives, 6(3): 79–100. Chen, I. J. and A. Paulraj (2004). ‘Understanding Supply Chain Management: Critical Research and a Theoretical Framework’, International Journal of Production Research, 42(1): 131–63. Coase, Ronald (1937). ‘The Nature of the Firm’, Economica, 4(16): 386–405. Cooke, Philip and Kevin Morgan (1998). The Associational Economy. Oxford: Oxford University Press. Cousins, Paul and Robert Spekman (2003). ‘Strategic Supply and the Management of Inter and Intra Organisational Relationships’, Journal of Purchasing and Supply Chain Management, 9(1): 19–29. Crouch, Colin and Henry Farrell (2002). ‘Breaking the Path of Institutional Development? Alternatives to the New Determinism’, MPIfG Discussion Paper, Cologne. —— (2004). ‘Breaking the Path of Institutional Development? Alternatives to the new Determinism’, Rationality and Society, 16(1): 5–43. Culpepper, Pepper (1998). ‘Rethinking Reform: The Politics of Decentralized Cooperation in France and Germany’. Boston, MA: Harvard University. Cyert, Richard and James March (1992). A Behavioral Theory of the Firm, 2nd edn. Cambridge, MA: Blackwell. Deavers, Kenneth L. (1997). ‘Outsourcing: A Corporate Competitiveness Strategy, Not a Search for Low Wages’, Journal of Labor Research, 18(4): 503–19. DiMaggio, Paul (ed.) (2001a). The Twenty-First-Century Firm: Changing Economic Organization in International Perspective. Princeton, NJ: Princeton University Press. —— (2001b). ‘The Futures of Business Organization and Paradoxes of Change’, in Paul DiMaggio (ed.), The Twenty-First-Century Firm: Changing Economic Organization in International Perspective. Princeton, NJ: Princeton University Press. Dore, Ronald (1983). ‘Goodwill and the Spirit of Market Capitalism’, The British Journal of Sociology, 34: 459–82. —— (1986). Flexible Rigidities: Industrial Policy and the Structural Adjustment in the Japanese Economy. Stanford: Stanford University Press. —— (2000). Stock Market Capitalism: Welfare Capitalism: Japan and Germany Versus the Anglo-Saxons. New York: Oxford University Press.

196

References

Dresser, Laura and Joel Rogers (2003). ‘Part of the Solution: Emerging Workforce Intermediaries in the United States’, in Jonathan Zeitlin and David Trubek (eds.), Governing Work and Welfare in a New Economy. New York: Oxford University Press. Durand, Jean-Pierre, Paul Stewart, and Juan Jose` Castillo (eds.) (1999). Teamwork in the Automobile Industry: Radical Change or Passing Fashion? London: Macmillan Press. Dyer, JeVrey (2000). Collaborative Advantage: Winning Through Extended Enterprise Supply Networks. New York: Oxford University Press. Eberts, Randall and George Erickcek (2002). ‘The Role of Partnerships in Economic Development and Labor Markets in the United States’. Kalamazoo MI: Upjohn Institute StaV Working Paper No. 02-75. Erickcek, George (2001). ‘ ‘‘Smart’’ Economic Development’, Employment Research, 8(2), published online at www.upjohninst.org/publication/empres.html. Ericksen, Paul, Joel Rogers, Josh Whitford, and Jonathan Zeitlin (2002). ‘Leveraging Manufacturing Excellence: A Supplier Strategy for Wisconsin’. Madison: White Paper for the Wisconsin Economic Summit III. —— Rajan Suri, Bash’shar El-Jawhari, and Aaron Armstrong (2005). ‘The Gap in the Toyota Production System: Rethinking Supply Management in the Age of Global Sourcing and Lean’, APICS—The Performance Advantage, 15(2), published online at www.apics.org/Resources/Magazine/Past/February2005/default_February2005.htm. Feenstra, Robert (1998). ‘Integration of Trade and Disintegration of Production in the Global Economy’, Journal of Economic Perspectives, 12(4): 31–50. Fligstein, Neil (2001). The Architecture of Markets: An Economic Sociology of TwentyFirst-Century Capitalist Societies. Princeton, NJ: Princeton University Press. —— and Robert Freeland (1995). ‘Theoretical and Comparative Perspectives on Corporate Organization’, Annual Review of Sociology, 21: 21–39. Florida, Richard (1995). ‘The Industrial Transformation of the Great Lakes Region’, in Philip Cooke (ed.), The Rise of the Rustbelt. New York: St Martin’s Press. —— (1996). ‘Regional Creative Destruction: Production Organization, Globalization, and the Economic Transformation of the Midwest’, Economic Geography, 314–34. —— (2000). ‘Competing in the Age of Talent: Quality of Place and the New Economy’: Report prepared for the R.K. Mellon Foundation, Heinz Endowments and Sustainable Pittsburgh. Foss, Nicolai (1993). ‘Theories of the Firm: Contractual and Competence Perspectives’, Journal of Evolutionary Economics, 3: 127–44. —— (1999). ‘Research in the Strategic Theory of the Firm: ‘‘Isolationism’’ and ‘‘integrationism’’ ’, Journal of Management Studies, 36(6): 725–55. Fung, Archon, Dara O’Rourke, and Charles Sabel (2001). Can We Put an End to Sweatshops? Boston, MA: Beacon Press. Gibbons, Robert (1999). ‘Taking Coase Seriously’, Administrative Science Quarterly, 44: 145–57. —— (2001). ‘Firms (and Other Relationships)’, in Paul DiMaggio (ed.), The TwentyFirst-Century Firm: Changing Economic Organization in International Perspective. Princeton, NJ: Princeton University Press. Granovetter, Mark (1985). ‘Economic Action and Social Structure: The Problem of Embeddedness’, American Journal of Sociology, 91(3): 481–510.

References

197

—— (1992). ‘Economic Institutions as Social Constructions: A Framework for Analysis’, Acta Sociologica, 35: 3–11. Hall, Peter and David Soskice (eds.) (2001). Varieties of Capitalism: The Institutional Foundations of Comparative Advantage. Oxford: Oxford University Press. Harrison, Bennett (1994). Lean and Mean. New York: Basic Books. —— and Maryellen Kelley (1993). ‘Outsourcing and the Search for ‘‘Flexibility’’ ’, Work, Employment and Society, 7(2): 213–35. Helper, Susan (1991). ‘Strategy and Irreversibility in Supplier Relations: The Case of the U.S. Automobile Industry’, Business History Review, 65: 781–824. —— and David Hochfelder (1997). ‘ ‘‘Japanese-Style’’ Supplier Relationships in the American Auto Industry, 1985–1920’, in Takao Shiba and Masahiro Shimotami (eds.), Beyond the Firm: Business Groups in International and Historical Perspective. Oxford: Oxford University Press. —— and Mari Sako (1995). ‘Supplier Relations in Japan and the United States: Are They Converging?’, Sloan Management Review, 36(3): 77–84. —— (1998). ‘Determinants of Trust in Supplier Relations: Evidence from the Automotive Industry in Japan and the United States’, Journal of Economic Behavior and Organization, 34: 387–417. —— and John Paul MacDuYe (1999). ‘Creating Lean Suppliers: DiVusing Lean Production through the Supply Chain’, in JeVrey Liker, W. Mark Fruin, and Paul Adler (eds.), Remade in America: Transplanting and Transforming Japanese Management Systems. New York: Oxford University Press. —— —— and Charles Sabel (2000). ‘Pragmatic Collaborations: Advancing Knowledge While Controlling Opportunism’, Industrial and Corporate Change, 9(3): 443–83. Herrigel, Gary (1996). Industrial Constructions: The Social Foundations of German Industrial Power. Cambridge: Cambridge University Press. —— (2001). ‘Memo on AMP Wndings’, University of Chicago internal memorandum to the Advanced Manufacturing Project. —— (2004). ‘Emerging Strategies and Forms of Governance in the Components Industry in High Wage Regions’, Industry and Innovation, 11(1/2): 45–80. Herrigel, Gary and Volker Wittke (2004). ‘Varieties of Vertical Disintegration: The Global Trend Toward Heterogeneous Supply Relations and the Reproduction of DiVerence in US and German Manufacturing’, in Richard Whitley, Glenn Morgan, and Moen (eds.), National Business Systems in the New Global Context. Oxford: Oxford University Press. Hirschman, Albert (1970). Exit, Voice and Loyalty: Responses to Decline in Firms, Organizations, and States. Cambridge, MA: Harvard University Press. Hirst, Paul and Jonathan Zeitlin (1991). ‘Flexible Specialization Versus Post-Fordism: Theory, Evidence and Policy Implications’, Economy and Society, 20(1): 1–56. Holmstro¨m, Bengt and John Roberts (1998). ‘The Boundaries of the Firm Revisited’, Journal of Economic Perspectives, 12(4): 73–94. Kelley, Maryellen and Bennett Harrison (1992). ‘Unions, Technology, and Labor-Management Cooperation’, in Larry Mishel and Paula Voos (eds.), Unions and Economic Competitiveness. Armonk, NY: Sharpe. Kenney, Martin and Richard Florida (1993). Beyond Mass Production. Oxford: Oxford University Press.

198

References

Krippner, Greta, Mark Granovetter, Fred Block, Nicole Biggart, Tom Beamish, Youtien Hsing, Gillian Hart, Giovanni Arrighi, Margie Mendell, John Hall, Michael Burawoy, Steve Vogel, and Sean O’Riain (2004). ‘Polanyi Symposium: A Conversation on Embeddedness’, Socio-Economic Review, 2(109). Kristensen, Peer Hull and Jonathan Zeitlin (2004). Local Players in Global Games: The Strategic Constitution of a Multinational Corporation. Oxford: Oxford University Press. Lamoreaux, Naomi R., Daniel M. G. RaV, and Peter Temin (2004). ‘Against Whig History’, Enterprise and Society, 5(3): 376–87. Langlois, Richard (2004). ‘Chandler in Larger Frame: Markets, Transactions Costs, and Organizational Form in History’, Enterprise and Society, 5(3): 355–75. —— and Nicolai Foss (1999). ‘Capabilities and Governance: The Rebirth of Production in the Theory of Economic Organization’, Kyklos, 52(2): 201–18. Lashinsky, Adam (1992). ‘DCCA Training Plan Aims Small: Eager Response to Job Funds for ‘‘Clusters’’ ’, Crain’s Chicago Business, Chicago. Levy, Jonah (1999). Toqueville’s Revenge: Dilemmas of Institutional Reform in Post-dirigiste France. Cambridge, MA: Harvard University Press. Liker, JeVrey, W. Mark Fruin, and Paul Adler (eds.) (1999). Remade in America: Transplanting and Transforming Japanese Management Systems. New York: Oxford University Press. Loveman, G. and Werner Sengenberger (1991). ‘The Reemergence of Small-Scale Production: An International Comparison’, Small Business Economics, 6(1): 1–37. Lowi, Theodore (1985). ‘Why Is There No Socialism in the United States’, Society, 22(2): 34–42. Luria, Daniel (1996a). ‘Toward Lean or Rich? What Performance Benchmarking Tells Us About SME Performance, and Some Implications for Extension Center Services and Mission’. Atlanta, Conference on Manufacturing Modernization: Learning From Evaluation Practices and Results. —— (1996b). ‘Why Markets Tolerate Mediocre Manufacturing’, Challenge, 11–16. —— (2000a). ‘A High-Road Policy for U.S. Manufacturing’, in Candace Howes and Ajit Singh (eds.), Competitiveness Matters: Industry and Economic Performance in the U.S. Ann Arbor, MI: University of Michigan Press. —— (2000b). ‘Good Manufacturing Jobs: Recipe Known, Outlook Uncertain’, What Future for Manufacturing: Trade Unions and the Challenges of Change in Manufacturing. Cambridge, MA: Harvard University. —— (2002). ‘US Component Manufacturing at a Crossroads: Region-Loyal Production & Global Manufacturing DeXation’. Madison: Presentation at UW-Madison conference on Supply Chain Governance and Regional Governance in the Global Economy, September 10. MacDuYe, John Paul (1995). ‘Workers’ Roles in Lean Production: The Implications for Worker Representation’, in Steve Babson (ed.), Lean Work: Empowerment and Exploitation in the Global Auto Industry. Detroit, MI: Wayne State University Press. —— (1997). ‘The Road to ‘‘Root Cause’’: Shop-Floor Problem-Solving at Three Auto Assembly Plants’, Management Science, 43(4): 479–502. March, James (1962). ‘The Business Firm as a Political Coalition’, Journal of Politics, 24: 662–78.

References

199

—— (1996). ‘Continuity and Change in Theories of Organizational Action’, Administrative Science Quarterly, 41: 278–87. Mas-Colell, Andreu, Michael Whinston, and Jerry Green (1995). Microeconomic Theory. Oxford: Oxford University Press. Mesquita, Luiz and Thomas Brush (2001). ‘Relationship Management in Vertical Manufacturing Alliances, Supplier Development and Supplier Performance’. Purdue University. Milgrom, Paul and John Roberts (1988). ‘Economic Theories of the Firm: Past, Present, and Future’, Canadian Journal of Economics, 21(3): 444–58. —— (1990). ‘The Economics of Modern Manufacturing: Technology, Strategy and Organization’, American Economic Review, 80(3): 511–28. Nelson, Richard and Sidney Winter (1982). An Evolutionary Theory of Economic Change. Cambridge: Belknap Press. Nelson, Richard and Sidney Winter (2002). ‘Evolutionary Theorizing in Economics’, Journal of Economic Perspectives, 16(2): 23–46. Nishiguchi, Toshihiro (1994). Strategic Industrial Sourcing: The Japanese Advantage. New York: Oxford University Press. NIST (2003). ‘An Overview of MEP’. Osterman, Paul (1999). Securing Prosperity: The American Labor Market: How It Has Changed and What to Do About It. Princeton, NJ: Princeton University Press. —— (2003). ‘Organizing the US Labor Market: National Problems, Community Strategies’, in Jonathan Zeitlin and David Trubek (eds.), Governing Work and Welfare in a New Economy. New York: Oxford University Press. —— Thomas Kochan, Richard Locke, and Michael Piore (2001). Working in America: A Blueprint for the New Labor Market. Cambridge, MA: MIT Press. Peck, Jamie (1996). Work-Place: The Social Regulation of Labor Markets. New York: Guilford Press. Piore, Michael and Charles Sabel (1984). The Second Industrial Divide: Possibilities for Prosperity. New York: Basic Books. Podolny, Joel and Karen Page (1998). ‘Network Forms of Organization’, Annual Review of Sociology, 24: 57–76. Porter, Michael (1998). On Competition. Boston, MA: Harvard Business School Publishers. —— (2000). ‘Location, Competition, and Economic Development: Local Clusters in a Global Economy’, Economic Development Quarterly, 14(1): 15–34. Powell, Walter (1990). ‘Neither Market Nor Hierarchy: Network Forms of Organization’, in Barry Staw and L.L. Cummings (eds.), Research in Organizational Behavior, Vol. 12. Greenwich, CT: JAI Press. —— (2001). ‘The Capitalist Firm in the Twenty-First Century: Emerging Patterns in Western Enterprise’, in Paul DiMaggio (ed.), The Twenty-First-Century Firm: Changing Economic Organization in International Perspective. Princeton, NJ: Princeton University Press. Prahalad, C.K. and Gary Hamel (1990). ‘The Core Competence of the Corporation’, Harvard Business Review, 68(3): 79–91. Putnam, Robert (1993). Making Democracy Work: Civic Traditions in Modern Italy. Princeton, NJ: Princeton University Press.

200

References

Pyke, F., G. Becattini, and W. Sengenberger (eds.) (1990). Industrial Districts and InterFirm Co-operation in Italy. Geneva: International Institute for Labor Studies. Pyke, Frank and Werner Sengenberger (eds.) (1992). Industrial Districts and Local Economic Regeneration. Geneva: International Institute for Labor Studies. Rickert, JeVrey (1999). ‘Supplier Development at John Deere–Horicon Works’, Master’s thesis, Department of Sociology; University of Wisconsin at Madison. —— Joel Rogers, Darya Vassina, Josh Whitford, and Jonathan Zeitlin (2000). ‘Common Problems & Collaborative Solutions: OEM–Supplier Relationships and the Wisconsin Manufacturing Partnership’s Supplier Training Consortium’. Madison, WI: Center on Wisconsin Strategy. Rogers, Joel and Daniel Luria (1999). Metro Futures: Economic Solutions for Cities and Their Suburbs. Boston, MA: Beacon Press. —— (2003). ‘Babies, Bathwater, and American Manufacturing: What’s Worth Saving and How’. Madison, WI: Center on Wisconsin Strategy. Rowthorn, Robert and Ramana Ramaswamy (1997). ‘Deindustrialization—Its causes and implications’. Washington, DC: International Monetary Fund. —— (1999). ‘Growth, Trade, and Deindustrialization’, IMF StaV Papers, 46(1): 18–41. Sabel, Charles (1989). ‘Flexible Specialization and the Re-emergence of Regional Economies’, in P. Hirst and J. Zeitlin (eds.), Reversing Industrial Decline? Industrial Structure and Policy Britain and Her Competitors. Oxford: Berg. —— (1994). ‘Learning by Monitoring: The Institutions of Economic Development’, in N. Smelser and R. Swedberg (eds.), The Handbook of Economic Sociology. Princeton, NJ: Princeton University Press. —— (1995). ‘Intelligible DiVerences: On Deliberate Strategy and the Exploration of Possibility in Economic Life’. Florence, Paper presented to the 36th Annual Meeting of the Societa´ Italiana degli Economisti. —— (1996). ‘An American View of the Novel Universalism of Japanese Production Methods and Their Awkward Fit with Current Forms of Corporate Governance’. Tokyo: Conference on Socio-Economic Systems of the Twenty-First Century. —— (2004a). ‘Theory of a Real Time Revolution’: Revised copy of paper presented at 19th EGOS colloquium, July 2003; paper forthcoming in Organization Studies. —— (2004b). ‘Pragmatic Collaborations in Practice: A Response to Herrigel and Whitford and Zeitlin’, Industry and Innovation, 11(1/2): 81–8. —— and Jonathan Zeitlin (1985). ‘Historical Alternatives to Mass Production: Politics, Markets and Technology in Nineteenth-Century Industrialization’, Past-and-Present, 108: 133–76. —— —— (2004). ‘Neither Modularity nor Relational Contracting: Inter-Firm Collaboration in the New Economy’, Enterprise and Society, 5(3): 388–403. Sako, Mari (2004). ‘Supplier Development at Honda, Nissan and Toyota: Comparative Case Studies of Organizational Capability Enhancement’, Industrial and Corporate Change, 13(2): 281–308. —— and Fiona Murray (1999). ‘Modules in Design, Production and Use: Implications for the Global Automotive Industry’. Oxford: Said Business School, University of Oxford. Saxenian, AnnaLee (1994). Regional Advantage: Culture and Competition in Silicon Valley and Route 128. Cambridge, MA: Harvard University Press.

References

201

Schwartz, Michael (2000). ‘Markets, Networks, and the Rise of Chrysler in Old Detroit, 1920–1940’, Enterprise and Society, 1: 63–99. Scott, Allen J. (1998). ‘The Geographic Foundations of Industrial Performance’, in Alfred D. Chandler, Jr, Peter Hagstrom, and Orjan Solvell (eds.), The Dynamic Firm: The Role of Technology, Strategy, Organization and Regions. Oxford: Oxford University Press. Shapira, Philip (1998). ‘Manufacturing Extension: Performance, Challenges, and Policy Issues’, in Lewis Branscomb and James Keller (eds.), Investing in Innovation: Creating a Research and Innovation Policy that Works. Cambridge, MA: MIT Press. Simon, Herbert (1976). Administrative Behavior: A Study of Decision-Making Processes in Administrative Organization. New York: Free Press. Smith, Vicki (1997). ‘New Forms of Work Organization’, Annual Review of Sociology, 23: 315–39. —— (2001). Crossing the Great Divide: Worker Risk and Opportunity in the New Economy, Ithaca, NY: Cornell University Press. Smitka, Michael (1991). Competitive Ties. New York: Columbia University Press. Soskice, David (1999). ‘Divergent Production Regimes: Coordinated and Uncoordinated Market Economies in the 1980s and 1990s’, in Herbert Kitschelt, Peter Lange, Gary Marks, and John Stephens (eds.), Continuity and Change in Contemporary Capitalism. New York: Cambridge University Press. Stark, David (1996). ‘Recombinant Property in East European Capitalism’, American Journal of Sociology, 101(4). Storper, Michael (1995). ‘The Resurgence of Regional Economies, Ten Years Later: The Region as a Nexus of Untraded Interdependencies’, European Urban and Regional Studies, 2(3): 191–221. —— and Robert Salais (1997). Worlds of Production: The Action Frameworks of the Economy. Cambridge, MA: Harvard University Press. Streeck, Wolfgang (1991). ‘On the Institutional Conditions for DiversiWed Quality Production’, in Wolfgang Streeck and Egon Matzner (eds.), Beyond Keynesianism: The Socio-Economics of Production and Employment. London: Edward Elgar. Sturgeon, Timothy (2002). ‘Modular Production Networks: A New American Model of Industrial Organization’, Industrial and Corporate Change, 11(3): 451–96. Teubner, Gunther (2001). ‘Legal Irritants: How Unifying Law Ends up in New Divergences’, in Peter Hall and David Soskice (eds.), Varieties of Capitalism: The Institutional Foundations of Comparative Advantage. Oxford: Oxford University Press. Tolliday, Steven and Jonathan Zeitlin (1991). ‘Employers and Industrial Relations Between Theory and History’, in Steven Tolliday and Jonathan Zeitlin (eds.), The Power to Manage? Employers and Industrial Relations in Comparative Historical Perspective. New York: Routledge. Trigilia, Carlo (1999). ‘Capitale sociale e sviluppo locale’, Stato e Mercato, 57. Tripsas, Mary, Stephan Schrader, and Maurizio Sobrero (1995). ‘Discouraging Opportunistic Behavior in Collaborative R&D: A New Role for Government’, Research Policy, 24: 367–89. Turner, Robert (1999). ‘Public Policies for Manufacturing Revitalization: Competing Models in Three American States’, Political Science. Madison, WI: University of Wisconsin.

202

References

Uzzi, Brian (1996). ‘The Sources and Consequences of Embeddedness for the Economic Performance of Organizations: The Network EVect’, American Sociological Review, 61: 674–98. —— (1997). ‘Social Structure and Competition in InterWrm Networks: The Paradox of Embeddedness’, Administrative Science Quarterly, 42: 35–67. —— (1999). ‘Embeddedness in the Making of Financial Capital: How Social Relations and Networks BeneWt Firms Seeking Financing’, American Sociological Review, 64: 481–505. Veloso, Francisco (2000). ‘The Automotive Supply Chain Organization: Global Trends and Perspectives’. Cambridge, MA: Massachusetts Institute of Technology. Vidal, Matt (2000). ‘Temporary Employment Relations and the Organization of Work in Manufacturing Establishments: Flexibility, Productivity and the Appropriation of Surplus labor’, Master’s Thesis, Department of Sociology, UW-Madison. Madison, WI: University of Wisconsin. Waits, Mary Jo (2000). ‘The Added Value of the Industry Cluster Approach to Economic Analysis, Strategy Development, and Service Delivery’, Economic Development Quarterly, 14(1): 35–50. Westney, Eleanor (2001). ‘Japanese Enterprise Faces the Twenty-First Century’, in Paul DiMaggio (ed.), The Twenty-First-Century Firm: Changing Economic Organization in International Perspective. Princeton, NJ: Princeton University Press. Whitford, Josh (2001). ‘The Decline of a Model? Challenge and Response in the Italian Industrial Districts’, Economy and Society, 30(1): 38–65. Whitford, Josh and Jonathan Zeitlin (2004). ‘Governing Decentralized Production: Institutions, Public Policy, and the Prospects for Inter-Firm Collaboration in US Manufacturing’, Industry and Innovation, 11(1/2): 11–44. Whitford, Josh, Jonathan Zeitlin, and Joel Rogers (2000). ‘Down the Line: Supplier Upgrading, Evolving OEM–Supplier Relations, and Directions for Future Manufacturing Modernization Policy and Research in Wisconsin’. Madison, WI: Center on Wisconsin Strategy. Wilkerson, Chad (2001). ‘Trends in Rural Manufacturing’, The Main Street Economist: Commentary on the Rural Economy. Center for the Study of Rural America: Federal Reserve Bank of Kansas City. Williamson, Oliver (1985). The Economic Institutions of Capitalism. New York: Free Press. —— (1991). ‘Comparative Economic Organization: The Analysis of Discrete Structural Alternatives’, Administrative Science Quarterly, 36: 269–96. —— (1993). ‘The Logic of Economic Organization’, in Oliver Williamson and Sidney Winter (eds.), The Nature of the Firm: Origins, Evolution, and Development. Oxford: Oxford University Press. Winter, Sidney (1993). ‘On Coase, Competence, and the Corporation’, in Oliver Williamson and Sidney Winter (eds.), The Nature of the Firm: Origins, Evolution, and Development. Oxford: Oxford University Press. Wolfe, David (2003). ‘Social Capital and Cluster Development in Learning Regions’, in J. Adam Holbrook and David Wolfe (eds.), Knowledge, Clusters and Learning Regions. Kingston, Ontario: School of Policy Studies, Queen’s University (electronic copy of forthcoming chapter). —— and Meric Gertler (1999). ‘Globalization and Economic Restructuring in Ontario: From Industrial Heartland to Learning Region’. Paper prepared for the NECSTS/

References

203

RICTES-99 Conference on Regional Innovation Systems in Europe, Donostia-San Sebastian, Spain. Wolman, Harold and David Spitzley (1996). ‘The Politics of Local Economic Development’, Economic Development Quarterly, 10(2): 115–50. Womack, James P., Daniel T. Jones, and Daniel Roos (1991). The Machine that Changed the World: The Story of Lean Production. New York: HarperPerennial. Zeitlin, Jonathan (2000). ‘Introduction: Americanization and its Limits: Reworking US Technology and Management in Post-War Europe and Japan’, in Jonathan Zeitlin and Gary Herrigel (eds.), Americanization and its Limits: Reworking US Technology and Management in Post-War Europe and Japan. Oxford: Oxford University Press. —— (2003). ‘Introduction: Governing Work and Welfare in a New Economy: European and American Experiments’, in Jonathan Zeitlin and David Trubek (eds.), Governing Work and Welfare in a New Economy. New York: Oxford University Press. Zeitlin, Jonathan and Gary Herrigel (eds.) (2000). Americanization and its Limits: Reworking US Technology and Management in Post-War Europe and Japan. Oxford: Oxford University Press.

This page intentionally left blank

INDEX Advanced manufacturing Project (AMP) 55, 163–4 Alfred P. Sloan Foundation, Advanced Manufacturing Project and 163–4 Amable, Bruno 39–40, 184n37 American deindustrialization, not matter of relative currency values alone 13 American durable manufacturing, diversity in 2, 160 American economic development, decentralization, space for local experimentation 132 American federal structure, creates space for local and regional experimentation 126, 190n90 American LME 42 institutions for nonmarket coordination and 130, 132, 146, 190n91 nonmarket coordinating institutions question of political will 153 American manufacturing, American supply base, unevenness of forces hedging on both sides 93–4 changes in organization 3, 5, 57–60, 95, 117, 154 essential to move beyond static stories of inevitable decline 47 failure of production networks to sidestep condition that wreck firms 119 increase in specialized subcontracting (1980s) 19 interfirm cooperation difficult to build in 19–21, 24–5, 28, 97 job losses ascribed to poor trade performance 12 massive declines in employment in 9–11

new economy being built with pieces that came before 111 realism if we are to understand ‘new production paradigm’ 53–4 regional concentration 54 relationships are multiplex 48 sector in deep crisis 2, 11–2 American Upper Midwest, case study 51, 54–56, 160–1 CME-style coordinating institutions in 152 disingenuous and failed collaboration and 99–100 interviews suggest specialization and diversification 67, 187n67 lacks necessary ex ante social and institutional conditions 25 metal manufacturing , interviews in 2, 51 American welfare and labor market policy, systemic changes and 130 APPENDIX, Advanced Manufacturing project: preinterview facility data profile 175–77 interview modules 169 interview process 55–56, 163–69 interview question guideline for OEMs 170–5 list of firms in 141 Applebaum, Eileen 21–22 appliance industry, manufactures as little as possible 58 arm’s-length bargaining tactics, ‘way the market is right now’ 88, 100 Arthur, W. Brian 25 automotive industry, assemblers 11, 58–9, 62, 65 changes in American and global manufacturing 57 contracts can be canceled at any time 86–87 modularization 61–63

206 automotive industry (cont.) no historic expertise in electronics 52 reorganization has left first tier suppliers in charge of design of new products 65 tiered structure needs collaboration of component manufacturers 66, 187n65 Automotive News 65 ‘bad waltzing,’ defined 101–2 Bartik, Timothy 129, 132, 134, 151, 190n92 Bayesian learning, modifications in Polya-urn games and 30–1, 183n26 Behavioral Theory of the Firm 110 benchmarking 28, 46, 88, 90–1, 118, 150, 185n46, 188n76 Berk, Gerald 191n113 Bivens, Josh 12 blockages, difficulty of reforming institutionalized behaviors 32 blurring of organizational boundaries 110, 157 Bolster, Harald (Mercedes-Benz) 65 ‘Boundaries of the Firm Revisited, The’ 35 Boyer, Robert 15 brainstorming phase, better point to get supplier input 81 Brush, Thomas 139 Bureau of Economic Analysis 12 Burroni, Luigi 24 buyers, at OEMs 87, 102, 105–6, 109, 112–4, 116, 143 buying ‘higher in the billables’ (i.e. buying subassemblies) 60 Caprile, Maria 58 Carruthers, Bruce 16 Case new-Holland, ‘millions of different product configurations’ 59 Casper, Steven 41 Chandler, Alfred D. 36 The Visible Hand 11

Index changes in production, not supply chain paradigm 17 China, global competition from 2, 19, 87 Chrysler 65 SCORE program 77, 85 classical organizational theory, blurring of organizational boundaries 110 Clinton administration, industrial extension programs 134 ‘cluster’ development 133–4, 190n93 CMEs 41–2, 48, 123, 126, 153 coordinating institutions even in the protypical American LME 148, 152 governments work together with business associations 42, 124 Coase, R. 26, 33, 157 ‘code of conduct’ 150, 192n110 collaboration in practice 76 hedging in the waltz 91–4 limits of the waltz 84–86 positional bargaining, big economic hammers and five percent letters 86–89 supplier competencies in historical context 89–91 multidimensionality and the components of successful waltz 81–83 waltzing at its best 77–81 what is target costing? 76–77 collaborative network production, theoretical approaches to 26 collaborative production 25, 31, 44 comparative institutionalism ‘Varieties of Capitalism’ 27–8, 39–42, 123–4 competence trust, definition 96 competence uncertainties, residue of legacy of capacity subcontracting 156 component manufacturing sector, defined 54, 185n54 component manufacturing firms, suburban and less likely to be unionized 54

Index component manufacturing industry, heavily concentrated in just five states 54, 149 component supply, compared to contract manufacturing in electronics 61–3 ‘concentration without centralization’ 21 ‘conflictual partnership’, combination of collaboration and wariness 92 cooperative equilibria, external interaction to punish defection or continued interaction between parties 74 Coordinated Market Economies see CMEs coordination-oriented policies, to improve competencies of firms 124 core workforce, flexibility and a casualization of labor relationship 21–22 corporate governance 40 cost distinguished from price 72 ‘cost’ as organizing frame learnt from OEMs and suppliers 76 ‘cost reduction’, relational intertwining of incremental innovation and interfirm collaboration 76 cost reduction waltz, defined 76–7 eliminating redundancies as a step in 106 how OEMs and suppliers develop innovations 56 incremental innovations can add up to real money 84 involvement of multiple divisions at suppliers and OEMs 81 limits 84–86 multidimensionality in 81–3 partiality of collaboration and 95 reason cost reductions now often require collaboration 78–9 cost reductions, reason they often require collaboration 77–81 Crouch, Colin 29–30, 126, 156, 158 Culpepper, Pepper 125 customer, flexible production system only if subcontractors flexible 21

207

Cyert, Richard 37, 110 Daimler-Chrysler, five percent letter after Daimler takeover 86 Deavers, Kenneth L. 19 decentralization of production 9, 20, 37, 43, 45, 56, 133 state level economic development and 132–3 decentralization of American manufacturing 57–8 automotive as lead sector in reorganization of American durable manufacturing 58–9 between segmentation and hedging 73–5 changing role of mid-level component supply beyond auto 66–67 changing role of mid-level component supply in auto 64–66 characterizing post-outsourcing American productive model 60–3 component manufacturer strategy in decentralized production regime 67–72 from managing final market uncertainty to negotiating innovation uncertainty 63–4 modularization versus the persistence of integrated design 61–3 raised profile of interfirm relationship even at mid-level 72–3 similar patterns across durable manufacturing 59–61 Deere, John, Horicon Works 135 supplier training infrastructure administrative services 136 training department 137 deindustrialization 1–2, 9–13, 154 design for manufacture and/ or value engineering 69, 77, 82, 103 see also cost reduction waltz design, simultaneous engineering and 44 design modularity, obstacle to is pressure for rapid change 62

208

Index

Detroit, cooperative product development (1920s and 1930s) 19 deunionization 9, 54, 154 deurbanization 9, 54, 154 developing world, transfer of laborintensive manufacturing to 1, 20 deverticalization of production 11, 53, 61, 154–5 DiMaggio, Paul 15, 39 Twenty-First-Century Firm, The 18 diversification 68, 71 Dore, Ronald 38–9, 97, 181n17þA231 Dresser, Laura 130–1, 148 durable manufacturing industries 2, 58–60, 66, 160 Durand, Jean-Pierre 15 Dyer, Jeffrey 51, 98, 111 Eberts, Randall 129 economic development, regional see industrial policy, regional Economic Policy Institute 12 ‘Economic Theories of the Firm: Past, Present and Future’ 34 economics of organization 32–6, 183n27 embedded ties, embeddedness 26, 27–8, 36–38, 46–7, 48 , 97, 158–60 183n33, 189n88 Crouch and Farrell, lock-in, and 29–30, 155–6 employee involvement programs 21 engineering, OEM 59–60, 65, 67–8, 70, 72, 77–9, 81–2, 84, 87, 89, 103, 106, 109, 114–5, 118 engineering, supplier 58, 65–70, 72, 77–9, 81–2, 87, 100, 118 Erickcek, George 129, 132 Ericksen, Paul 135–36, 149 Ernst and Young, OEM strategies and 57, 185n55 externalization of production, manufacturers in developing world and 20 ‘selective adaptation’, requirements of local circumstances and 16

fairness 100 Farrell, Henry 29–30, 126, 156, 158 federated firms 43–4 Feenstra, Robert 12 Fine, Charles 61 firm, theory of coordinating incentives and acquiring tacit productive knowledge 32 difficulties of transferring tacit knowledge and 35–36 how they are continuously reconstituting and revising relationships 51 integrate vertically to rein in opportunism and/or to share tacit knowledge 36 long-term cooperation between, repeated prisoner’s dilemma 34, 183n29 maximizers and nonmaximizers in current market conditions 46 profit-maximizing black box production cannot explain why they exist 33 reorganizing under learning by monitoring principles 45, 185n45 standard economic theories , vertical integration and conflicts 34–5 ‘five percent letter’ 86, 100, 109 ‘flexible specialization, Italian districts and 15 Fligstein, Neil 181n17, 181n19 Florida, Richard 15, 17, 61, 181n24, ‘high-performance economic organization’ 24–25 Ford 65 Fordist mass producers, systemic rather than cyclical difficulties (1980s) 15 Fordist period 19, 60 Foss, Nicolai 35–36, 182n32 French policymakers, difficulty in schemes for regional or technical development 125 functionality, diffusion and 46, 185n47

Index General Motors (GM), suppliers refused to do business with 101 Germany, Arbeitsgemeinschaft Zulieferindustrie (ARGEZ) 150 coordinated market economy (CMEs) 41 ‘diversified quality production’ 15 firms can take a long-term view 28 workforce employed in manufacturing fell 31 percent 9 Gertler, Meric 24 Gibbons, Robert 34, 110, 157 Global Positioning Systems (GPS) 52, 186n50 globalization 1, 2, 23–24, 47, 87–9 Granovetter, Mark 37–8, 153, 187n77 ‘economic institutions do not emerge automatically’ 32, 126, 159 rejection of ‘undersocialized’ Williamsonian actor 27 social embeddedness and 27, 32, 160, 183n35 Hall, Peter asset specific investment and CMEs 42 business and ‘structural influence’ 124, 126, 147, 152 CME vs. LME capitalism 41–2, 48, 97, 123, 180n15 complementarity and institutional path dependency 41, 61, 127 explaining why business opposes neoliberal restructuring 47 incentive compatibility in LME policy 124–5, 134–5, 138, 184n43 institutions and strategic interactions between firms 27–8, 40–1, 123– 5, 160–1, 183n40 national vs. regional or sectoral level 41, 126 radical vs. incremental innovation 63, 123 sphere of interfirm relations 40, 130–1 theory of the firm 183n33

209

Varieties of Capitalism 27, 40, 183n37–9, 183n41 Hamel, Gary 17 Harrison, Bennett 20, 24, 180n14, 181n23 Lean and Mean 20 Helper, Susan, Advanced Management Project 164 blockages to collaborative relationships and 48 changed relationships between firms 15, 181n18 cooperation and trust 29–31, 34, 46, 110, 112, 157, 188n80 Japanese-style customer-supplier relationships and 19, 158 learning by monitoring 44–5, 97, 184n46 neopragmatist approach and 28 nonstandard firms in United States 43 limits of American supply base 90 Herrigel, Gary, Advanced Manufacturing Project 80, 164, 187n78 conflictual partnership 92 cost reduction waltz 79–80, 91, 187n72 deverticalization 61 interviews with component manufacturers in United States 165–69, 186n62 Japanese production model and 15 limits of hard dichotomy between pragmatic collaboration vs. old style standard firm 53, 186n51 modularization vs. the persistence of integrated design 61–3, 185n51, 186n62 High Performance Work Organizations 21–22, 155 Hochfelder, David 19 Holmstro¨m, Bengt 35 horizontal diversification strategy, capacity subcontractor 71 horizontalizing firms, industrial crossfertilization and 71 Hyeong-ki Kwon 168

210

Index

Illinois Industrial Training Program 134–36, 190n94 ‘incentive compatible’ policies 42, 124, 134, 138, 185n43 increase in scale, greater specialization of division of labor 34 ‘industrial policy’ 129, 132–5 industrial policy, regional Alabama, Mercedes plant built (1990s) cost the state $168,000 per job 132 brokering corporate-plant disjunction 144–46 clusters 133–4 consortial models of supplier training 146–47 correcting paternalism: the relational perspective 143–4 economic development, linking entities that create regional competitiveness 133 Illinois Industrial Training Program 134–36, 190n94 incentives 32–3 limitations and open questions of consortial models 147–8 MEP 132, 134–5, 190n95 paternalistic story: OE1 arrives and fixes everything 141–3 politics of collaboration: can consortia really help to align performance expectations and supplier development practices across the OEMs? 149–50 politics of federalism: is the consortial model expandable? 148–9 state investment in supplier training, market failure justification 151 Tennessee, jobs in new Nissan plant went for $11,000 each (1980) 132 Wisconsin manufacturers’ development consortium (WMDC), as 135–37, 295 Wisconsin Regional Training Partnership (WRTP) 131–2

WMDC as a CME-style institution I: strategic alignment 139–40 WMDC as CME-style institution II: external support for internal initiatives 140–1 WMDC as toolbox 137–8 workforce intermediaries and coordination in American labor market 130–2 zero-sum economic development, makes less sense in decentralized production 132 industrial relations 21–3, 40, 131–2 innovation, the radical and the incremental contrasted 63, 123–4 innovations, incremental 21, 54, 56, 63–4, 117 changed role of buyers and 112–3 cost reduction as, claim defended 76 influenced by negotiations between OEMs and suppliers 61, 73, 76, 78, 81, 84–6, 93, 112 LMEs and CMEs 28, 41–2, 123–4 need for mutual adjustment and 73–4 uncertainty and 63–4, 74, 90, 94, 96–9, 119 institutionalism 27–8, 39–43, 123–7, 158–60 institutional framework, interlocking complementarities between parts 39–41 institutions, importance in economics acknowledged nowadays 39, 183n40 ‘several institutions taken together reinforce each other’ 39–40 International Monetary Fund 1 International Motor Vehicle Program, studies by 58 international trade, growing across advanced manufacturing economies since (1970s) 12 interviews conducted 54–5

Index intrafirm relationships, plague interfirm relationships 56, 112–16, 157 ISO 9000 quality assurance programs 150 issues of organizational change as issues of institutional change 32 Italian literature on industrial districts 39 Italian regions, social embeddedness and 97 Japan 9, 16, 47 auto industry and target productivity 77 CMEs 123 ‘coordinated market economies’ 28 collaboration based on long-term interpersonal relationships 39 interfirm cooperation in 44 Japanese manufacturing, social network theorists and 97 ‘‘Japanese-style’’ customer-supplier relations, American auto industry (1920) 19 ‘Japanese production model’, importation and hybridization of 15, 154–5 ‘lean production’, Japanese producers and 15, 60 target costing techniques 76 JIT 17–18, 58, 68, 107–8 job rotation 21–22 Jones, Daniel T. 15 just in time see JIT Kelley, Maryellen 20 Kenney, Martin 15 Kristensen, Peer Hull 111–2 Langlois, Richard 11, 35–36 Learning by Monitoring 26, 28–9, 43–6, 96–9, 111, 155, 160, 184n46 learning regions 24–25, 155 Levy, Jonah 125 Liberal Market Economies see LMEs

211

Liker, Jeffrey 15 Llorens, Clara 58 LMEs 46, 48, 125–26 comparative institutional advantage’ in radical innovation 123 effective business coordination capacity cannot be built ‘‘spontaneously’’ 124 ‘incentive compatibility’ 134, 138 institutional support for market and hierarchical coordination 41 MEP ‘blunt’ policies that are feasible in 135 policymakers and comparative advantage in radical innovation and 42 weak associational structures in which producer groups focus on ‘lowest common denominator’ 147 Lopez, Jose Ignacio (purchasing at GM 1991 and 1993) 101 ‘Lopezing’ 101 Luria, Daniel 2, 10, 88, 90, 164 MacDuffie, John Paul 21, 29, 43–46, 112, 157–8 Machine that Changed the World, The 15, 24 manufacturers, contemporary troubled by competence uncertainty 96 manufacturers in developing world, externalization of production and 20 manufacturing extension partnership see MEP manufacturing job loss (1980s), firms restructuring due to global pressures 12 manufacturing modernization, WMEP or through technical college system 152 March, James 37, 110 ‘market incentive’ policies, use incentives to induce actors to perform more effectively 124

212 mass-customization, shortened product life-cycles 59 MEP 132, 134–5, 190n95 decentralized nature of addresses regional issues 135 Mesquita, Luiz 139 Metal Spinning supplier in Milwaukee, negotiation with OEM 117–18 metal stamper working for mid-volume OEMs, story of online auction 108–9 methods of error-detection 28, 185n46 Mexico 19, 89 metal fabricator, meeting changing demands and help in product design 118 Michigan Manufacturing Technology Center (MMTC) see Performance Benchmarking Service Midwest manufacturers, shift towards ‘high-performance economic organization’ 17 Milgrom, Paul 34 misleading forecasting, systemic problems and 107 model-of-the-decade, foreign observers flocked to rising industrial power 16 modifications of product or process, not always cost reductions 105 modularization 18, 58, 61–63, 117 ‘moral hazard’ 33 multiplex relationships, often require explicit cross-firm coordination 81 Murray, Fiona 62 National Establishment Surveys (1992 and 1997) 22 National Institute of Standards and Technology (NIST) 134 Nelson, Richard 35 network ties, information sharing and voice before exit 37, 184n34

Index ‘networked firm’, essence of new production paradigm 17–19 network organization 17–19, 26–27, 36–9, 96, 110–1, 155 networks, noise and institutional change, conclusion 46–48 economics of organization: from black box to contract and competence theories 32 contractual perspective: peeking inside the black box 33–4 inadequacy of pure incentive focus: orthodox and heterodox critique 34–36 ‘hard institutional’ explanation: comparative institutionalism and the ‘varieties of capitalism’ 27–8 introduction 27–8 ‘organizational’ explanation: neopragmatist ‘learning by monitoring’ 28–9 ‘social’ explanation: social networks and embedded ties 27 new determinism vs. the contradictory middle ground 29–32 sociology of economic organization 36 ‘hard institutional’ explanation: comparative institutionalism and the ‘varieties of Capitalism’ 39–42 ‘organizational’ explanation: neopragmatist ‘learning by monitoring’ 43 learning by monitoring and the organization of cooperation 44–6 ‘organizational’ explanation: neopragmatist ‘learning by monitoring’, federated production and the obviation of vertical integration 43–4 ‘social’ explanation: social embeddedness and the network organization 36–39 ‘new determinism’ 30

Index new old economy, defined 3 ‘new’ paradigm, meaning of 19 new production paradigm, collaborative relations between firms and 17–21 ‘lean’ or ‘high performance’ work systems and 21–3, 182n22 new production paradigm for new old economy, crisis and hybridization in productive models 15–17 double-edge of productive decentralization 19–21 double-edge of the learning region 24–25 locating new production paradigm: geography, learning and lock-in 23–24 networked firm as crux of new production paradigm 17–19 new production paradigm inside firm: flexibility in work organization 21–23 new work systems, contradictions of 22 New York Garment manufacturing, patterns of cooperation in 27, 97 Nishiguchi, Toshihiro 21, 77 Nobel Prize, North (economist 1993) 39 nonstandard firm, core element is the work group 43, 110 ‘outperforms hierarchies in volatile environments’ 45 normative challenge 48 North European CMEs 123 OEM-supplier relationships, cost reduction in 76–81 collaborative versus arm’slength 17–19, 31, 52 contradictoriness of 100–1 effective training and supplier development resources can improve 149

213

increased salience of 51–3 incremental innovation 63–6 more variation in than captured by simple binary choice 53 multidimensionality in 81–83, 153 possible to reduce costs without ‘absorbing someone else’s margins’ 78 supplier competence important for 89–91, 138 OEMs 5–6 ability to get their way with contracts and suppliers 87, 91 ability to impose their will not unbounded 89 adjusting operations to accommodate changes suggested by supplier 53 asking suppliers to provide full breakdowns of all their process costs 81 buffeted by pressures beyond their control 92 ‘build-to-demand’ strategies 107, 189n82 buyers 87, 102, 105–6, 109, 112–4, 116, 143, 189n84 buying more complete subsystems 70 by decentralizing production, transfer of uncertainties and 92 cautious about relying on supplier firms 31–2, 89–91 collaborative relationships with suppliers partial and contradictory 47 common supplier qualification and certification procedures 147 ‘component manufacturing’ suppliers and 54 culling supply base to work with fewer suppliers 60, 187n60 decentralized production as means to compete in global markets 17–19, 57–61, 155 decentralizing crucial elements find price strategy unsustainable 93

214

Index

OEMs 5–6 (cont) decentralizing depends on disciplined exchange of information 45 dependence on suppliers’ ability to keep up with technological change 17, 83 developing teams that include members of multiple departments 114 don’t have design horse-power anymore 68–9 engineering, OEMs 59–60, 65, 67–8, 70, 72, 77–9, 81–2, 84, 87, 89, 103, 106, 109, 114–5, 118 expect suppliers to deliver goods on very short notice 106–7 expect that suppliers help them meet targets for ‘cost reduction’ 72–3, 76 experimenting with online auctions to drive down price 108 factional conflict and 99, 112–16 focus on limited set of core competencies 67–8 forced to reorganize relationships 58 global reach to move business to suppliers in lower-wage regions 12, 87 hedge collaboration with suppliers 86–94, 117–119 implications of reliance on supplier technology and design 72–3 incorporation of electronics and new materials in key metal industries 52 incremental innovation risk to their decentralizing production 63–4, 187n64 inter-departmental difficulties and 99, 112–16 interest in moving to collaboration conditioned by uncertainties 94 interviews with personnel, question of ‘with whom’ to collaborate 64 long-term contracts and 87

‘market-price-minus’ principle rather than ‘cost-plus’ 77 market power advantage , profitability of exit option and 89 maximize know-how and cost gains from suppliers at aggregate level 91–2 may have interest in refusing to waltz 53, 86, 151 may seek to retain operations beyond core competencies 62–3 misuse of private information 103 ‘need their costs to go down’ 85 need to be flexible and reactive to improve designs 21 purchasing managers and purchasing departments 57–60, 64–6, 81, 82, 85, 89, 101, 114–5, 141–2, 149 rotate people through buyer position quickly 113 treat subcomponents as commodities and use arm’s-length transacting 100 try to convince suppliers to deliver price cuts before savings secured 105 wary of getting caught over a barrel in negotiations 74 offshoring and outsourcing not equivalent 12 oligopoly, hardly precludes brutal fights over market share 89–90 ongoing supply-chain revolution, led by automotive industry 60 open communication, important in early stages of product development 81 opportunism 26, 28, 36, 38, 98, 100–10 organizational sociology 110–112 organizational screw-ups, interactions between internal OEM departments and suppliers 109 organizational uncertainty, defined 109–10 ‘original equipment manufacturer’ see OEM

Index explanation of terminology 5–6 Osterman, Paul 22–3, 130–1, 180n17, 189n91 innovative work systems, higher chances of layoffs in (1997) 23 workers, prefer higher levels of responsibility and training in new skills 22–23 outsourcing 1, 2, 3, 11–12, 17, 19–20, 58, 154, 156 of costs that had previously been assumed by the OEM 106 worth doing only if supplier does all the engineering work 58 Page, Karen 15, 26, 37–8, 46, 184n48 Peck, Jamie 181n17 Performance Benchmarking Service (MMTC), proprietary database of 88 using value-added-per-full-time-empl 188n76 ‘permanent uprising against routine’, ‘key to survival in turbulent world’ 29–30, 45–46, 158 ‘physical assets have become a liability’ 57, 186n55 Piore, Michael 24, 43 Podolny, Joel 15, 26, 37–8, 46, 184n48 Porter, Michael 133 positive-sum cost reductions, effective only as open and collaborative practice 77 post-Weberian synthesis in organizational sociology 110, 157 Powell, Walter 3 deverticalization as network governance 37 distinguished collaborative network from markets and hierarchies 20, 26 importance of space in network organizational forms 23 ‘new logic of organizing’ and network 18, 27, 37, 98, 110, 157–8, 188n81

215

political-institutional conditions for networks 39–40 production networks ‘Neither Market Nor Hierarchy’ 20, 23, 26, 37, 111, 182n30, Prahalad, C.K. 17 principal-agent models, closest to black box neo-classical theories of firm 33 productive and exchange behavior, social and institutional contexts 23 public policy, regional institutions and interfirm collaboration 129–30 conclusion 152–3 regional industrial policy and consortial models of supplier training 146–7 limitations and open questions 147–8 politics of collaboration: can consortia really help to align performance expectations and supplier development practices across the OEMs? 149–50 politics of federalism: is the consortial model expandable? 148–9 politics of survival: can the model survive the state fiscal crisis and an LME state? 151–2 state and local industrial policy in world of decentralized production 132–5 Wisconsin manufacturers’ development consortium 135–7 basics: WMDC as toolbox 137–8 WMDC as a CME-style institution I: strategic alignment 139–40 WMDC as CME-style institution II: external support for internal initiatives 140–1 brokering corporate-plant disjunction 144–46 correcting paternalism: the relational perspective 143–4

216

Index

public policy, regional institutions and interfirm collaboration 129–30 (cont.) paternalistic story: OE1 arrives and fixes everything 141–3 workforce intermediaries and coordination in American labor market 130–2 purchasing large proportions of value as competitive necessity 57–8 purchasing managers and purchasing departments 57–60, 64–6, 81, 82, 85, 89, 101, 114–5, 141–2, 149 Putnam, Robert 97 QS9000 (quality certification standards) 69 QSR N6 qualitative data analysis software 164 quality circles 21–22 radical and incremental changes, product and process and 73 Ramaswamy, Ramana 1 reconstruction of regional political economy 154–58 meta-theoretical moral in the story 160–1 practical moral in the story 158–60 relationships define products 93 reliability begets competency begets reliability, virtuous circle 111 ‘resistible rise of workforce intermediaries’ 131–2 restructuring as inevitable in some form or other 10 Rickert , Jeff 163, 169 Roberts, John 34–5 Rogers, Joel 2, 130–1, 148, 164 Roos, Daniel 15 Rowthorn, Robert 1 Sabel, Charles 24 American manufacturers and Japanese techniques 18, 28–9, 181n19

blockages of efforts of OEMs and suppliers for collaboration 48 blurring of organizational boundaries 157 customer-supplier relations not selfstabilizing 125–26 effective collaboration and 112 emergent new logic of organizing around ‘non-standard’ firms 100 explanation for emergent collaboration 29–30, 44–46, 98 Learning by Monitoring 26, 28–9, 43–6, 96–9, 111, 155, 160, 184n46 more collaboration in American productive model 18 ‘‘mutually ignorant specialists’’ and preexisting game 111–12 neopragmatism, and 30–1, 47, 97, 140 organizational revolution, inadequacy of market and hierarchy 29, 43–4, 110, 125–26, 179n179 path dependency, opposition to 30 risks of decentralization 45–46 United States nonstandard firms ‘federated’ not ‘centralized’ 43 virtuous circle and 111, 156 Sako, Mari 15, 62, 90, 181n18, 187n73, 188n80 Salais, Robert 32 Saxenian, Annalee 97 Schneiberg, Marc 191n113 Schumpeterian change 30 Schwartz, Michael 19 Second Industrial Divide, The 24 segmentation 52, 64, 73–5 Shapira, Philip 134 sheet metal process specialist, how ideas are developed 79 shift from arms-length, partnership relationships and 19 shift from capacity to specialized subcontracting 17 Silicon Valley 52, 97, 123 Simon, Herbert 37 Administrative Behavior 110

Index simultaneous engineering 28, 60, 185n46 small and medium-sized enterprise see SMEs SMEs 91, 134, 136, 146, 152 Smith, Adam 34 Smith, Vicki 15, 21–22, 180n16–7 Smitka, Michael 77 social science, place of analytic dichotomies in 53 Soskice, David 61, 63 asset specific investment and CMEs 42 business and ‘structural influence’ 124, 126, 147, 152 CME vs. LME capitalism 41–2, 48, 97, 123, 180n15 complementarity and institutional path dependency 41, 61, 127 explaining why business opposes neoliberal restructuring 47 incentive compatibility in LME policy 124–5, 134–5, 138, 184n43 institutions and strategic interactions between firms 27–8, 40–1, 123–5, 160–1, 183n40 national vs. regional or sectoral level 41, 126 radical vs. incremental innovation 63, 123 sphere of interfirm relations 40, 130–1 theory of the firm 183n33 Varieties of Capitalism 27, 40, 183n37–9, 183n41 South Korea, changes in manufacturing employment 9 specialized suppliers, learning curve to resist pressure tactics 69 reduce cycle times 68 Spitzley, David 132 standardization and codification of component interfaces, uncertainty and 62 standardization and modularization, render market coordination unproblematic 61

217

Stark, David 111 Storper, Michael 23, 32 strategy space for component suppliers 67–73 influenced by national systems of self-reinforcing coordinating institutions 28 strengthening of ‘business coordinating capacity’ VOC approach and 147 ‘structural influence’ 147, 152 structure of argument 4, 5 Sturgeon, Timothy 11, 18, 61–2, 73 subcontracting see outsourcing specialized vs. capacity 17 suppliers, adjusting strategies to respond to new commercial environment 67 aware there are limits to protections of co-development 88 cautious about investing in new competencies and opening books 112 cheaper processes, opportunity to ‘take’ parts 80 co-development only gives limited power 88 collaborative incremental innovations and 97 difficulties they have in meeting needs of collaborative model 91 diversifying vertically adding new process capabilities 70 engineering, supplier 58, 65–70, 72, 77–9, 81–2, 87, 100, 118 expected to meet target productivity improvements 77 firms’ design improvements get shopped to competitors 117 hedge by maintaining arm’s-length relationships when necessary 84–5, 99, 117, 157 investment in capital equipment 68 less productive than their OEM customers on average 90 may ‘hunker down’ to weather the storm of recession 151

218

Index

‘muddy the waters’ as they pass on information 103–4 nonnegotiable enforcement of annual targets and 102 online auctions and 108–9 ‘phantom’ orders and 107 reject assistance from customers for fear of compromising information 104 relationships with customers help them direct the design process 81–2 ‘rules, norms and procedures’ required to invest in new competencies 111 some ‘see supplier consultants doing more the job of ‘‘policing’’ than ‘‘promotive development’’ 139 specialization with diversification and 68 strategy space for component suppliers 67–73 supplier confusion due to internal OEM miscommunication 115 ‘tear-downs’ of OEM products and 80 unilateral restructuring of terms, not exclusive to supplier-provided ideas 105–106 used WMEP training consortium in countless different ways 140 ‘value engineering’ and ‘design for manufacture’ and services 77 wary of OEM declarations of ‘partnership’ 119 whipsawed by mixed messages from fragmentary OEMs 114 withholding costing information to get leverage to keep market 119 work for several competing OEMs 146, 151 target costing 76–77, 188n71 TCE 33, 36, 183n33 transactions, frequency, uncertainty and relative asset specificity 33–4, 183n28 under fire even among economists 34

technology, point of entry for new suppliers 59–60 Teubner, Gunther 41 Tolliday, Steven 46 total quality management 22 Toyota’s supplier relationships 98 trade deficit 12 Transaction Cost Economics see TCE transportation costs and communications, reduced, price competition from China 19 Trigilia, Carlo 24 trust 96, 104–5 trust as reliability 96, 140 Turner, Robert 135 twenty-first-century firm, made from, not in place of, twentieth century firm 112 uncertainty 28, 33, 62, 92, 94, 96, 100, 117–20, 156–57 competence 96, 119, 156–7, 159 market and technological 95, 96, 156–7, 159 organizational 99–100, 117, 125, 153, 119, 156–7, 159 uncertainty and contradiction in the new old economy 95–96 contradictory collaboration and the new old organization 110–12 misuse of inside information 103–4 abuse of trust and ‘classic’ hold-up 104–106 big question: where does contradictory collaboration come from? 109 deceptive use of online auctions 108–9 shifting unremunerated risk and cost to suppliers 106–8 organizational roots of contradictory collaboration 112–16 systematic implications of organizational uncertainty 117–20

Index uncertainty, trust and the stability of the intermediate case 96–100 waltzing badly: contradictory collaboration 100–2 non-negotiable cost reduction targets: when the waltz is all lead and no follow 102–3 unionized establishments, high production work systems in 22–3 United States, cannot effectively hybridize CME-style institutions 125 core limitations on ‘coordinationoriented’ policymaking political 127 deindustrialization an uneven process 9 effect of decentralization of production 10–11, 51, 57–61 elements of ‘CME business’ with long established relationships in 153 enforceable formal contracts 41 enormous changes in organization of manufacturing 15 inhospitable cultural and institutional conditions for collaborative manufacturing in 24–5, 119, 123 lacks institutional infrastructure to help generate worker skills 159 LMEs 41–2, 123, 160 manufacturing employment 10 no real national policy but 5,000 industrial policies 123 OEM-supplier collaboration not new phenomenon in 19 organization of production more collaborative 18, 125 regional policy to build coordinating institutions in 127 resistible rise of novel workforce intermediaries 148 trade deficit 12–13 wage-cost driven outsourcing a problem in 19 Uzzi, Brian 16, 27, 38–9, 97, 183n36

219

Varieties of Capitalism: The Institutional Foundations of Comparative Advantage 40, 184n38 Varieties of Capitalism see VOC Veloso, Francisco 57–8, 63, 76 vertical diversification, not easy to manage and diseconomies of scope 71 vertically integrated manufacturers, decentralizing production to smaller suppliers 11, 20, 26, 57–61 Vidal, Matt 21–22, 166, 169 virtuous circles 99, 156 reliability begets competency begets reliability 111 VOC 27–28, 39–42, 46, 123–27, 131, 155, 159–60, 183n37–9, 183n41 asset specific investment and CMEs 42 business and ‘structural influence’ 124, 126, 147, 152 CME vs. LME capitalism 41–2, 48, 97, 123, 180n15 complementarity and institutional path dependency 41, 61, 127 explaining why business opposes neoliberal restructuring 47 incentivecompatibility in LME policy 124–5,134–5,138,184n43 institutions and strategic interactions between firms 27–8, 40–1, 123–5, 160–1, 183n40 national vs. regional or sectoral level 41, 126 radical vs. incremental innovation 63, 123 sphere of interfirm relations 40, 130–1 theory of the firm 183n33 vocational training and education 40 Vonnegut, Kurt, Mother Night 160 Westney, Eleanor 39 Williamson, Oliver 26, 33–36, 38, 183n30 Williamsonian synthesis 110 Winter, Sidney 33–36, 46 Wisconsin Manufacturers’ Development Consortium see WMDC

220

Index

Wisconsin Manufacturing Extension Partnership (WMEP) Wisconsin Regional Training Partnership see WRTP Wisconsin Technical College System 137 Wittke, Volker 60, 91–2 WMDC 129, 163, 165, 193n115 all OEMs in have suppliers in other states 149 CME-style institution 140 could grow into more complete externalized framework 149 garnered business interest and aided OEM-supplier collaboration 148 helps to sustain collaboration in OEMsupplier relationships 137 individuals who saw opportunity to leverage outside resources to build collaborative relationships 153 limits of 147 need to widen access 148 provides nominated suppliers with toolbox 137 space for nonmarket coordinating institutions in American LME 152, 159–60 strong position to promote mutual learning among participants 147 success during first years of operation 146, 192n106 threat from political disputes in the ‘LME-state’ 153 ways it differs from ‘typical’ LME policy 138 WMEP 135–36, 149, 164, 191n97, 191n101 courses aided the joint project with OE1 by giving them confidence 144

curriculum for continuous improvement leaders 146 enhancing supplier-OEM business relationships and 136 facilitating information flow 147 focus more on the provision of core services 146 ‘honest broker’ 136, 191n100 mission to serve SMEs and large OEMs of consortium 152 supplier working with 138–9 Wolfe, David 23–25 Wolman, Harold 132 Womack, James P. 15, 24 workforce intermediaries, can form basis of systemic change 131 WRTP, constructed in interstices of state’s economic development framework 148 focuses attention on modernization in member firms 131–2 Zeitlin, Jonathan, Advanced Manufacturing Project 164 effect of Japanese production model across the globe 15–16 interviews with component manufacturers in United States 165, 167, 169 new systems of welfare provision and labor-market governance 130 OEMs and suppliers and collaborative relationships 111–112 reform of the multinational corporation and 111 uncertain markets and selection processes 46