résumé en français dans l’ouvrage 9!bmgck@:xqoxst!€¦ · tél. : 33 (0)1 40 61 60 00 - fax :...
TRANSCRIPT
l es notes de l’ifri – n° 26
The New American ChallengeLe nouveau défi américain
3.The Transatlantic Exchange of Ideas and Practices:National Institutions and Diffusion
Bruce Kogut
institut français
des relations
internationales
les
no
tes
de
l’if
ri –
n°
26
Th
e T
ran
satl
anti
c E
xch
ange
of
Idea
s an
d P
ract
ices
:N
atio
nal
In
stit
uti
on
s an
d D
iffu
sio
n
les notes de l ’ i fr i
This paper argues that the defining feature of the contemporary trans-atlantic exchange is the political acceptability of institu-tional changes. The diffusion of ideas and methods posesthe problem of complementarities between the existingpractices and the new ones. Complementarities are notsimply determined by technology, but are influenced by cognitive and cultural factors and by politics.Complementarities mean that countries will differ in whatthey can adopt. Sometimes, new practices are not compa-tible with old ones. More interestingly, even when thereare no conflicts in practices, institutions may not be com-patible. These problems can be analyzed as a combinato-rial problem that can be represented by a simple Booleanalgebra. Using these ideas, the paper argues that the cur-rent debate in France over such issues as compensationposes broad political and cultural questions regarding thefuture of French society and the viability of the Frenchnational system. In contrast, the United States is an openinternational system, where its deficits in worker trainingare compensated by a highly crafted immigration policyto attract foreign labor. The French and American systemscan co-exist and each can achieve a set of goals compa-tible with their national aspirations.
Résumé en français dans l’ouvrage
9!BMGCK@:XQOXST!Prix : 45,92 FF - 7 €ISBN 2-86592-094-1
Diffusion :La Documentation française29-31, quai Voltaire75340 Paris Cedex 07Téléphone : 33 (0)1 40 15 70 00Télécopie : 33 (0)1 40 15 72 30www.ladocfrancaise.gouv.fr
The Transatlantic Exchangeof Ideas and Practices:
National Institutions and Diffusion
The “Transatlantic Series” proposes concise analyses in English or French of themain Transatlantic debates and political issues in both the United States andEurope. It benefits from the support of the German Marshall Fund of the UnitedStates.
Ifri is a research centre and a forum for debate on the major international politicaland economic issues. Headed by Thierry de Montbrial since its founding in 1979,Ifri is a non-profit organization.
The opinions expressed in this text are the responsibility of the author alone.
La série transatlantique des « Notes de l’Ifri » propose des analyses concises, enfrançais ou en anglais, des principaux débats transatlantiques et des enjeux despolitiques publiques menées tant aux États-Unis qu’en Europe. Cette série reçoitle soutien du German Marshall Fund of the United States.
L’Ifri est en France le principal centre indépendant de recherche, d’informationet de débat sur les grandes questions internationales. Créé en 1979 par Thierryde Montbrial, l’Ifri est une association reconnue d’utilité publique (loi de 1901).Il n’est soumis à aucune tutelle administrative ; il définit librement ses activités etpublie régulièrement ses travaux.
Les opinions exprimées dans ce texte n’engagent que la responsabilité de l’auteur.
© Droits exclusivement réservés, Ifri, Paris, 2000ISBN 2-86592-094-1
ISSN 1272-9914
Ifri - 27, rue de la Procession - 75740 Paris Cedex 15 - France
Tél. : 33 (0)1 40 61 60 00 - Fax : 33 (0)1 40 61 60 60
E-mail : [email protected] - Site Internet : www.ifri.org
les notes de l’ifri - n° 26
Sér i e t ransa t l an t i que
The New American Challenge/Le nouveau défi américain
3.The Transatlantic Exchange
of Ideas and Practices:National Institutions and Diffusion
Bruce Kogut
Décembre 2000Institut français des relations internationales
The New American ChallengeCorporate Strategies and European Policies
Le nouveau défi américainStratégies d’entreprises et politiques européennes
Note 24The New American Challenge and Transatlantic Technology SourcingFrédérique Sachwald
Note 25Why and How European Companies Reach Out to Silicon ValleyThierry Weil
Note 26The Transatlantic Exchange of Ideas and Practices: National Institutions and DiffusionBruce Kogut
Note 27Changements organisationnels : de la nouvelle économie aux 35 heuresPhilippe Askenazy
Note 28Hegemony Wired: American Politics and the New EconomyEthan B. Kapstein
Note 29Dynamiques industrielles et réglementaires des télécoms : une comparaison États-Unis/EuropeOlivier Bomsel et Gilles Le Blanc
Note 30The Information Society Race: Is Europe a Follower or a Challenger?Michel Catinat
4
Contents
Presentation of the Series:
The New American Challenge p. 7
Le nouveau défi américain p. 9
Abstract p. 11
Résumé p. 12
Introduction p. 13
1. A Boolean primer on institutions and practices p. 18
1.1. Boolean algebra of diffusion p. 24
1.2. Institutional influence p. 26
1.3. Learning to like root beer p. 28
2. Empirical cases of diffusion p. 30
2.1. Diffusion of practices: the multidivisional structure p. 30
2.2. Diffusion of institutions: private enterprise p. 34
2.3. Global systems: skilled labor and the American model p. 38
Conclusions p. 42
References p. 43
5
Bruce Kogut is the Felix Zandman Professor of Management at theWharton School, University of Pennsylvania and is currently a visitingscholar at the Centre de Recherche en Gestion, Ecole Polytechniqueand at Insead business school in Fontainebleau. With Sid Winter, heco-directs the Reginald H. Jones Center at Wharton. His research isfocused on high technology, comparative systems, and privatization.
6
The New American ChallengeEuropean Corporate Strategies and Policies*
“Legions, raw materials and capital are no longer the marks orinstruments of power. And factors themselves are only an externalsign thereof. Nowadays, strength means capacity for invention,i.e. research; and the capacity to incorporate inventions in pro-ducts, i.e. technology. The deposits to be tapped are no longer inthe earth, in numbers or in machinery. They are to be found inminds, more specifically in the aptitude of men to reflect andcreate […]. In this notion of profit, which it is good form to praiseindiscriminately today, […] real profit for both companies andsociety consists of the fruits of innovation”.
Jean-Jacques Servan-Schreiber, Le défi américain,Denoël, 1967, p. 293.
At the end of the 60s, the American challenge that Jean-Jacques Servan-Schreiber and his team had identified was already that of innovation.Moreover, computers were already at the heart of this challenge, that ofthe emergence of the “post-industrial society”. Yet Europe had taken upthe challenge of growth after the Second World War. The major Euro-pean countries, like Japan, had done a remarkable job of catching up,with productivity growing at a faster pace than the US for many years.
The building of Europe was a fundamental element of the catching-upprocess, as was forecast by Servan-Schreiber, who urged the Euro-peans to step up their integration. Moreover, European benefited fromits status of follower; European companies adopted the organizationaland managerial principles that American groups had developed inbetween the two world wars. The presence of US companies in Europefacilitated the transfer of knowledge and practices by both stimulatingcompetition and providing models.
7
* Notes de l’Ifri No. 24 to 30.
In the 80s, the American challenge gave way to the Japanese one,which in turn was also successfully taken up by both American andEuropean companies. Once again, the challenge was organizational innature, and companies were often able to adapt Japanese organiza-tional principles, in order to boost productivity and quality simultane-ously.
The new American challenge appears different in nature. This timearound, as far as companies are concerned, it is institutional rather thanorganizational, and implies changes in the economic and social orga-nization. The aim is to take advantage of the wave of Schumpterianinnovation underlying the innovation-driven growth which the US haswitnessed during the 90s. Thus, the debate on the determinants of com-petitiveness has been enlarged and enriched during the beginning ofthe 80s and the diagnosis of Euroschlerosis which had stimulated theplan for a single market.
The aim of this series of the series of “Notes de l’Ifri” on The NewAmerican Challenge is to better appreciate its nature and to examinethe means which Europe employs, or could employ, to meet this chal-lenge. Some contributions focus on corporate strategies, which differsignificantly from previous decades in the context of globalization.Other papers provide a complementary analysis of public policies. Inparticular, they stress that promotion of innovation and new technolo-gies does not depend solely on innovation policies, but rather on whatone could call the policies of change. The issues tackled, ranging fromtechnological innovation to work organization, make it possible to takeup a major topic in the context of globalization, that of convergence –or lack thereof – with regard to national systems.
Frédérique SachwaldSeptember 2000
8
Le nouveau défi américainStratégies d’entreprises et politiques européennes*
« Ni les légions, ni les matières premières, ni les capitaux ne sontplus les marques, ni les instruments de la puissance. Et les usineselles-mêmes n’en sont qu’un signe extérieur. La force moderne,c’est la capacité d’inventer, c’est-à-dire la recherche ; et la capa-cité d’insérer les inventions dans les produits, c’est-à-dire la tech-nologie. Les gisements où il faut puiser ne sont plus, ni dans laterre, ni dans le nombre, ni dans les machines. Ils résident dansl’esprit, plus précisément dans l’aptitude des hommes à réfléchiret à créer.[…] Dans cette notion de profit, dont il est de bon tonde faire l’éloge indiscriminé aujourd’hui, […] ce qui est le profitréel pour l’entreprise comme pour la société, c’est le fruit de l’in-novation ».
Jean-Jacques Servan-Schreiber, Le défi américain,Denoël, 1967, p. 293.
À la fin des années 60, le défi américain que Jean-Jacques Servan-Schreiber et son équipe avaient identifié était donc déjà celui de l’in-novation. De plus, l’informatique était déjà au cœur de ce défi, celui del’émergence de la « société post-industrielle ». Pourtant l’Europe arelevé le défi de la croissance après la Seconde Guerre mondiale. Lesgrands pays européens, comme le Japon, ont opéré un remarquable rat-trapage, avec une croissance de la productivité qui a longtemps étésupérieure à celle des États-Unis.
La construction européenne a constitué un élément fondamental durattrapage, comme le prévoyait d’ailleurs Jean-Jacques Servan-Schreiber qui exhortait les Européens à accélérer leur intégration.L’Europe a d’autre part bénéficié de son statut de suiveur ; lesentreprises européennes ont notamment adopté les principes d’organi-sation et de gestion que les groupes américains avaient mis au pointdans l’entre-deux-guerres. La présence d’entreprises américaines en
9
* Notes de l’Ifri n° 24 à 30.
Europe a facilité les transferts de connaissances et de pratiques, à lafois en stimulant la concurrence et en offrant des modèles.
Dans les années 80, le défi japonais a succédé au défi américain. Luiaussi a été largement relevé, à la fois par les entreprises américaines etpar les entreprises européennes. Il s’agissait à nouveau d’un défiorganisationnel, et les entreprises ont souvent été en mesure d’adapterles principes d’organisation japonais pour accroître simultanément laproductivité et la qualité.
Le nouveau défi américain apparaît d’une autre nature ; non plusorganisationnel pour les entreprises, mais institutionnel impliquant deschangements dans l’organisation économique et sociale. Il s’agit eneffet de réussir à tirer parti de la vague d’innovation schumpeteriennequi sous-tend la croissance fondée sur l’innovation qu’ont connue lesÉtats-Unis depuis les années 90. Les débats sur les déterminants de lacompétitivité se sont donc élargis et enrichis depuis le début desannées 80 et le diagnostic d’eurosclérose qui avait stimulé le projet demarché unique.
L’objectif de cette série de « Notes de l’Ifri » sur Le nouveau défiaméricain est de mieux apprécier sa nature et d’examiner les moyensque l’Europe met en œuvre, ou pourrait mettre en œuvre, pour lerelever. Certaines contributions se concentrent sur les stratégies desentreprises, qui diffèrent sensiblement des décennies précédentes dansle contexte de la mondialisation. D’autres « Notes » analysent de façoncomplémentaire les politiques publiques. Elles soulignent notammentque la promotion de l’innovation et des nouvelles technologies nerelève pas des seules politiques de l’innovation – mais plus de ce quel’on pourrait appeler des politiques de changement. Les questionsétudiées, de l’innovation technique à l’organisation du travail, permet-tent d’aborder un thème majeur dans le contexte de la mondialisation,celui de la convergence – ou non – des systèmes nationaux.
Frédérique SachwaldSeptembre 2000
10
Abstract
The Transatlantic Exchangeof Ideas and Practices:
National Instiutions and Diffusion
This paper argues that the defining feature of the contemporarytransatlantic exchange is the political acceptability of institutionalchanges. The diffusion of ideas and methods poses the problem ofcomplementarities between the existing practices and the new ones.Complementarities are not simply determined by technology, but areinfluenced by cognitive and cultural factors and by politics. Comple-mentarities mean that countries will differ in what they can adopt.Sometimes, new practices are not compatible with old ones. Moreinterestingly, even when there are no conflicts in practices, institutionsmay not be compatible. These problems can be analyzed as a combi-natorial problem that can be represented by a simple Boolean algebra.
Using these ideas, this paper argues that the current debate in Franceover such issues as compensation poses broad political and culturalquestions regarding the future of French society and the viability of theFrench national system. In contrast, the United States is an open inter-national system, where its deficits in worker training are compensatedby a highly crafted immigration policy to attract foreign labor. TheFrench and American systems can co-exist and each can achieve a setof goals compatible with their national aspirations.
11
Résumé
La diffusion transatlantiquedes idées et des pratiques :
le rôle des institutions nationales
L’acceptabilité politique du changement est actuellement l’un desthèmes qui définissent le dialogue transatlantique. La diffusion inter-nationale des idées et des méthodes pose le problème de la complé-mentarité entre les pratiques nationales actuelles et nouvelles. Lescomplémentarités ne sont pas uniquement déterminées par la tech-nologie, mais sont aussi influencées par des facteurs cognitifs et culturels, ainsi que par le contexte politique. L’existence de telles complémentarités implique que les pays diffèrent dans leur capacité àadopter de nouvelles pratiques. En effet, des innovations en matièred’organisation du travail par exemple peuvent être souhaitables, maisincompatibles avec des pratiques nationales existantes. De plus, alorsmême que les pratiques seraient compatibles, les institutions peuventne pas l’être. Ces questions sont analysées comme un problème decombinatoire, qui peut être représenté grâce à une application élémen-taire de l’algèbre de Boole.
Cette « note » s’appuie sur cette approche et suggère que le débat enFrance sur des questions telles que l’évolution des rémunérations posedes problèmes politiques et culturels très larges touchant à la viabilitédu système national français. À l’opposé, les États-Unis constituent unsystème ouvert, où les déficits en matière de main-d’œuvre qualifiéesont compensés par une politique de l’immigration parfaitement adap-tée à l’attraction des travailleurs étrangers. Les systèmes français etaméricain peuvent-ils co-exister et atteindre un ensemble d’objectifscompatibles avec les aspirations nationales ?
12
Introduction
The re-emergence of the US economy in the world economy at theclose of the 20th century is coupled closely with the resurgence of itshigh technology industries. The structural adjustments in the olderindustries of steel, textile, and mining having been made in the 80s and90s, US productivity growth boomed in the new information sectors,especially in computers.1 This growth has served as an engine forincreasing corporate profitability that has, in turn, funded new invest-ments and job creation. In all, the effects of this new economy havefollowed a classic economic pattern common to all countries.
The stunning productivity growth, low unemployment, and leadershipin the new technologies are the features that characterize the newAmerican challenge. Why the US should lead in the developed worldeconomy in these areas is a question that is not at all well explained bythe usual aggregates of labor and physical and human capital. It is notsurprising, therefore, that the discussion of the sources of Americangrowth has led to a search for the institutions, rather than factors ofproduction, that foster high productivity and technological leadership.
There are several institutional facets peculiar to the United States that arewidely believed to have stimulated the growth of the new economy.These facets of flexible labor markets, venture capital, and porousboundaries between universities and research have permitted new indus-tries, largely in technology intensive sectors, to be created in the UnitedStates more rapidly than elsewhere. In France, for example, there havebeen important efforts to increase venture capital and new firm equitylistings and to allow scientists and research centers to commercialize
13
The author would like to thank Rachel Barrett and Nancy Brune for their researchassistance, Pierre Jacquet, Frédérique Sachwald and Thierry Weil for their com-ments, and participants in the workshop held June 2000 at IFRI. He would like toacknowledge the Reginald H. Jones Center of the Wharton School and its financialsupport for this research project.1. See Robert Gordon, faculty-web.at.northwestern.edu/economics/gordon.
themselves their scientific findings. The American resurgence is, farmore so than in the 60s, understood to be an institutional challenge.
This attention on the US resurgence strikes a familiar note sounded inthe celebrated book of the late 60s by Jean-Jacques Servan-Schreiber(1967). Yet, the comparison is also misleading. He proposed thatAmerican economic dominance stemmed from the operating effi-ciency of large multinational corporations. He cited above all thatAmerican strengths derived from efficient practices that the largemultinational corporation brought to Europe. The most radical insightof this book—which is still sorely neglected in current discussions onthe US resurgence—is that organizing principles, not technologies,drove the success of the American company.
This book had the effect of re-enforcing a trend already long underwayin Europe, namely, the move toward the amalgamation of smallerenterprises into large corporations. In fact, in many European coun-tries, the largest firms came to dominate their national industries inlarger proportion than the American case.2 Large firms are, or so wasthe implication, simply a shorthand way of saying better practices.
In retrospect, this Transatlantic diffusion did not challenge the domi-nant institutional frameworks that developed in many European States.It was a diffusion that largely left its legacy in the legitimation of thelarge corporation, on the one hand, and in the adoption and re-workingof practices. These practices consisted of accounting, divisionaliza-tion, pay incentives, and the rationalization of materials, among manyothers. The rise of the American consulting firms, and their inculcationin Europe, rode upon this rising sea of operating practices that swepteastward from the US.3 It were these practices, such as swept under therubric of “Fordism”, that provided firms with the capabilities to com-pete effectively in world markets. Thus, American practices, ifadopted, lead to American capabilities. At least such was the claim.
14
2. See table 1.3. See (Kipping 1999, McKenna 1999). Boltanski (1982) describes as well theimportant missions of productivity sent by France, among other countries, to theUnited States following the conclusion of World War II. See also (Djelic 1998,Zeitlin and Herrigel 2000).
In the minds of many commentators, these practices were linked to theinstitution of the large corporation. Only in this regard did Americanpractices become linked to an institutional challenge. In fact, for manycontemporaneous with Servan-Schreiber, the post-war period pre-sented the increasing powers of the State and economic planningacross all industrial countries. If there were to be an equifinalitytowards which all countries converge, it would be the powerful Key-nesian State that regulated and managed the economy.4
This belief in the large firm and State appears anachronistic, and yetthis is perhaps the enduring lesson. It is this ambiguity in the efficacyof institutions and practices that is the outstanding feature of the diffu-sion of economic knowledge. As a consequence, the diffusion ofknowledge is always accompanied by a debate among competing ideas.
The new American challenge poses concretely the problem of identi-fying complex causality and competing explanations. The desire to apethe technological and economic performance of the United States hascalled for the creation of more entrepreneurship, which in turn isbelieved to be the outcome of venture capital and mobile labor mar-kets. Largely independent of this particular challenge, European finan-cial markets have increasingly confronted pressures from large Amer-ican institutional investors to create greater transparency, to restructurelarge holding companies, and to reduce fiscal impediments to devel-oping equity markets.
This new American challenge is then a composite of two forces. It con-sists on the one hand of financial market pressures for institutionalchange that fall under the rubric of globalization. On the other hand,the new American challenge is embodied in the commercial competi-tion driven by new technologies that induces indirectly a call forchanges in labor and educational institutions. In both cases, the debateover practices and institutions is reflected in a debate over ideas aboutthe causal relationship between new institutions and performance andthe political ramifications. Thus, integral to the ideological debate isthe inherent ambiguity over what are the better practices and over thepolitics of change.
15
4. See (Shonfeld 1965, Hall 1989).
institutional analysis. To a surprising extent, the varieties of nationalsystem approaches return to an older body of ideas associated withfunctionalism. What this analysis shows is that ambiguity is inherent indeciphering causal effects. Even in the case of bootstrapping fromwhat is known, many possible evolutions are feasible. Politics and ide-ology are not so much the consequence as the necessary componentsto stabilize national institutions into believed coherent systems. Adap-tation is possible exactly because these systems are coupled not bytechnological givens, but by nominal beliefs.
I illustrate this approach by considering the diffusion of practices,institutions, and people. I show that the United States stands out as aparticularly unusual national configuration of institutions, because itrepresents an “open system” that serves as a source for world innova-tion, while being deficient in some of its own institutions. This analy-sis raises the issue of whether the American model can, and should, infact be imitated by other countries. France, in turn, is currentlyengaged in a deep debate over such practices as stock options. Thisdebate is lodged within the ideology of the importance of adopting for-eign practices to improve performance, but in the peculiar form thatcertain factors, namely managers, clearly benefit. Whether adoptingsuch practices constitute Pareto improvements are very unclear.
� 1. A Boolean primer on institutions andpractices
The literature on the varieties of capitalism varies widely in approach.Some approaches consist of the claim that countries are different intheir business systems and offer a list of factors by which to makecomparisons. Or sometimes an approach simply claims that institu-tions matter and countries differ in their political ideologies and econ-omy in how they achieve social consensus.
More ambitious approaches ride upon notions of complementaritiesand institutional equilibria. In one version of this approach, a countryis characterized by a number of actors who are institutionally repre-
18
sented. They achieve a political balance that also permits effectivecoordination in the workplace. It is this coupling between institutionsthat permit the political entente among corporatist actors that supportthe coupling in better practices.
A good example of this latter approach is the work of David Soskiceon Germany. Soskice (1990) proposes that the institutions of labor,enterprise hierarchies, banking, and business associations characterizeGermany as an ideal type of “corporatist” solution to social coopera-tion. Each of these institutions are represented by formal institutions ordominant actors (the Central Bank and governments also play a role.)These institutional actors bargain to create a “high equilibrium” thatsupports the coordination of work to produce quality export-orientedproducts.
Thus, the national systems approach poses a theory that explains thecomposition of institutions and their relation to economic performance.These theories are often fitted to rich country cases and suggest that theobserved institutions are complements, that is, the bundling of specificinstitutions are sufficient for explaining the economic performance.
To validate such claims, some comparative studies try to isolate aneffect by quasi-experimental design. For example, Bendix (1956) inhis study on the relationship of authority systems to national develop-ment chose Russia, the UK, Germany, and the United States as fourcountries, each which occupied a single cell in his two-dimensionalframework. Since he had two causes and four countries, he effectivelycreated a 2 by 2 factorial design.
The construction of experimental design using a small number of casesis a common template used in comparative research. Charles Ragin(1987) offers the most rigorous treatment of this approach that canserve as a canonical model in comparative international businessresearch. Ragin formalizes and extends the comparative method ofJohn Stewart Mills by using a few simple rules of Boolean algebra.
Consider the argument of Masahiko Aoki (1990) that the strength ofthe Japanese firm relies upon three duality principles that can be sum-marized as the joint presence of the main banking system, vertical
19
ranks for promotion, and horizontal control among workers and sup-pliers.8 He argues that when these three elements are present, they pro-duce a truth condition of a high performing economy. We can repre-sent this by the following claim:
Main Bank Vertical Rank Horizontal ControlTruth Condition:
Performance
1 1 1 1
The ones indicate that these conditions are present. Aoki proposes inother words a 3 factorial design. To consider all the possible cases (thatis, combinations), we would need to look at 23 or 8 combinations. Ofcourse, Aoki does not engage in this comparison and is content to pro-pose his theory as “fitting” the Japanese case. Most of the studies con-ducted in relation to his theory have looked at variations within Japan,especially between firms that belong and do not belong to businessgroups. Of course, these studies run into a selection bias (they arealready chosen to be members or not prior to looking at performance).Nor do they test at all adequately the claim that all three principlesmust be present in order to achieve high performance. Yet these stud-ies are reasonable first approaches to look at how institutions influenceperformance within a country.
However, this approach is not adequate for deriving any solutionsregarding whether these three duality principles have an effect on per-formance in other countries. This point is not minor, as there has beena significant amount of policy advice given to countries to adopt for-eign institutions as a whole because of their jointness and their provenvalue in producing high performance outcomes.
The extrapolation of these findings to other countries is stymied byfour problems. The first one is easily treated and is an example of spu-rious causality as delineated by Simon (1957). This problem is easilytreated by Ragin’s comparative Boolean methodology. In particular,
20
8. We do not have the space here to develop this notion of “jointness” in relation tothe recent work in complementarities. For such an analysis, see (Kogut, MacDuffieand Ragin 1999).
we will utilize the following Boolean rule. A cause A can be present(denoted as “A”) or absent (denoted as “a”). In one case AB are twofactors that are both present and are associated with a truth conditionof high performance. In a second case, factor A is still present but fac-tor B is not, and yet the outcome is still high performance. By Booleanalgebra, given that the truth condition is the same, AB + Ab = A. Inother words, only cause A is causally related to the outcome of highperformance. Indeed, for this comparison, A is sufficient to cause highperformance. In the absence of other causes, A also appears as a nec-essary factor.
Let us reconsider the Aoki formulation more carefully. Using now ourbinary symbols of 1 and 0 rather than upper and lower case letters, wecan represent Aoki’s claim, as we saw, as mapping the combination of{1 1 1} to the condition of high performance. What if we found a sec-ond country, say Korea, that had only two of these conditions but stillhad high performance? Comparing the two cases, we have
Japan: 1 1 1
Korea: 1 0 1
We can now conclude that only the first and third causes are causal;condition two is eliminated by Boolean algebra.
The second problem falls under the label of functional equivalence, asfirst analyzed and studied by Merton (1949). To illustrate this problem,consider a case where we compare Korea to a third country, sayFrance, both of which are high performing (if it seems unfair to com-pare rich France to moderately rich Korea, keep in mind that there aremuch poorer countries than Korea. Sample selection issues remaincritical to this methodology too!). By assumption, these two countrieshave the following configurations:
Korea: 1 0 1
France: 0 1 0
We cannot reduce these expressions further. They represent func-tionally equivalent institutional configurations to achieve high per-formance.
21
Now this conclusion might be troubling to our penchant to want mat-ters to be more precise, such as there is one configuration that domi-nates all others. But as we have learned from the literature on the vari-eties of capitalism (e.g. Berger and Dore 1996), there are many waysby which countries can achieve similar performance outcomes despitedifferent institutional configurations.
The problems of spuriousness and functional equivalence are easilydispatched by the application of a Boolean methodology. However,our third and fourth problems are not fully resolvable. The third is theproblem of insufficient variety in the empirical data. As noted earlier,Aoki’s three factorial design implies 8 distinct configurations. Theseare {0 0 0} {1 0 0} {1 1 0} {0 1 0} {0 0 1} {1 0 1} {0 1 1} {1 1 1}.What if history does not provide all these experiments? Or what if theresearch design did not generate a fully saturated model by the force ofits sampling methodology?
It is easy to see that we can make an error. Consider the case in which wehave sampled Japan and Korea as before in the first two rows and thensubsequently consider the case of a third rich country called the US:
Japan: 1 1 1Korea: 1 0 1US: 0 1 0
Clearly, we can no longer decide to eliminate the second cause. In fact,this second cause appears as sufficient, but not necessary in order tohave a rich outcome. We have now two configurations that are sug-gested by Boolean reduction. This analysis reduced the complexity ofthe three cases to two configurations.
Of course, we can in some cases generate “what if” cases based on theempirical data. We know by de Morgan’s law that these two state-ments are equivalent: both France and Germany do not like wine andneither France nor Germany likes wine. We can use this rule on theempirically-observed configurations to generate unobserved cases.
The last problem is the perennial obstacle of omitted variables. It isalways the case in empirical research (and in theorizing) that we haveneglected variables that not only do not matter but also interact with
22
the variables we have chosen. In econometrics, there are statisticaltreatments to eliminate unobserved heterogeneity (as discussed ear-lier), but these treatments are themselves guesses about the distributionof the unobserved error; they do not handle issues of complex interac-tions.
The effect of the problem of complexity is easily represented byRagin’s Boolean approach. Consider a comparison of Japan and Chad;Chad looks the same as Korea but is poor. We could conclude that theabsence of the second factor is causally responsible for poverty by ana-lyzing the following configuration.
Japan: 1 1 1Chad: 1 0 1
But we already know that Korea has the same configuration as Chadand it is relatively rich. The contradiction indicates that there is a prob-lem due to an omitted variable and an incomplete theory. If we add ina fourth condition, we might have:
Japan: 1 1 1 1Korea: 1 0 1 1Chad: 1 0 1 0
Now we see that the fourth variable is causally responsible. But wediscovered this only because there was a logical contradiction, andbecause we expanded our theory to look at a variable in which Japanand Korea agree but Chad differs. It almost looks as if we cooked thebooks. Theorizing country differences that do not permit testing isindeed an exercise in exotic and imaginary cuisines. Yet theorizing isindisputably required in order to guide the choice of variables and toprevent the list-like presentation of country differences that is often tobe found in the literature.
Boolean algebra does not eliminate all inherent complexity in causalrelationships. It does provide, however, a methodology by which tosample countries (i.e. saturate the design) and to characterize the fac-torial combinations as necessary, sufficient, or causally unrelated. Thisapproach has always been implicit in country comparisons, but yetrarely explicit as a template for conducting comparative research.
23
1.1. Boolean algebra of diffusion
The national system literature presents countries as independent exper-iments, sometimes precariously balanced in an equilibrium in whichall actors must continue to agree to perform. However, it is also clearthat practices and institutions diffuse across borders. How can weunderstand the study of Eleanor Westney (1987) on the importation oforganizational forms into Japan if we contend that organizationaleffectiveness is contingent upon rigid institutional configurations?Similarly, the work of John Paul MacDuffie (1993), indicates thatAmerican factories can adopt Japanese practices (in configurations)and achieve high performance productivity in the US despite differentinstitutions?
Diffusion presents thorny issues to national systems. The problem isthat there has been a failure in theory at two levels. The first is to sep-arate the effects of genesis from diffusion. It is perhaps true that certaininstitutional configurations gave rise to particular practices; this is his-torical causality. But once such practices are known, they may diffuseto other institutional conditions.
The second level is a failure to understand that actors are far moreadaptive than implied by these comparisons and that these practicesthemselves undergo radical re-interpretations. Within the corporatistbalance of Germany, practices at the firm or factory level may change,sometimes by diffusion, but in the context of a discursive searchamong actors to adapt these practices. It is surely more complex toadapt when practices challenge existing categorizations of workencoded in an existing division of labor, such as skill categories thatare tied to prestige and to wages. Yet even here, the political balanceamong corporatist actors at the macroeconomic level need not betightly coupled with the changes in work practice adaptations at themicrolevel of work and industrial organization.9
Work on diffusion is still pretty much trapped in fairly crude analysesthat once characterized national comparisons, that is, studies look at
24
9. See (Kogut and Parkinson 1998) for this argument, and also (Kogut 1999) for ananalysis of the adaptation of US work practices in Weimar Germany.
how a practice diffuses and by what channels. There is a growing lit-erature on the difficulty of adopting certain kinds of knowledge (e.g.Zander 1991) and a fledgling body on institutional and political resis-tance to the adoption of these practices (Kogut and Parkinson 1993,Boyer 1996, Berger and Dore 1996). These studies need to be joined,but they are theoretically quite distant.
As a first pass, let us consider the case of the adoption of a practice bya French firm that is institutionally neutral. In this case, the problem islargely cognitive, for the adopting firm needs now only to understandthe right causal combination and to adopt the various elements. TheFrench firm is organized as a hierarchy, with banking investors, andinternal recruitment of top managers. We can characterize this systemthen by the following summary:
Hierarchy Internal RecruitTruth Values:
Performance and Quality1 0 1 0
Consider now an American firm that has a different organizationalstructure of
0 0 1 1
In other words, the American firm produces at the same level of effi-ciency as the French firm but at better quality. What can the Frenchfirm do? It has control, subject to its negotiations with its managers,over the degree of hierarchy and its internal recruitment (indeed, the80s revolution in American corporate life resulted in flatter hierarchiesand more external recruiting [Useem 1996]). Given this, the Frenchfirm can make three changes.10
Hierarchy Internal RecruitTruth Values:
Performance and Quality
1 0 ? ?0 1 ? ?0 0 ? ?
25
10. With two elements to perturb, there are 4 possible configurations; the firm’s cur-rent configuration occupies one possibility.
It becomes transparent right away that there are many experimentaloutcomes, as each configuration can take on four different combina-tions of truth values. This is complicated, so we can make our liveseasy by assuming that productivity does not decline.
Of the three possible changes two cases represent hybridization, verymuch in the sense of Boyer (1998), by recombining American andFrench practices. These two recombinations represent feasible pathsfrom the French system. It is possible, of course, that these hybrids aresuperior to the American configuration, in which case there might bereverse diffusion. The third case (in bold) is Americanization, with thewholesale adoption of the American configuration. If the two firstcases both lead to high quality and high productivity, then they arefunctionally equivalent. If neither works, then the firm is constrainedto choose the American configuration.
It should be clear that even in this easy case, finding the right solutionis not easy. We have restricted the choice to two factors. Of course,there might be more elements. In addition, a firm might be unable torun all these experiments. As a consequence, it might observe otherfirms that have experimented, or it might hire consultants. Of course,even then, borrowing might be too inclusive and practices might beborrowed that have little to do with performance. They are like thehitch-hiking genes in genetics, bits of practices that have no clearcausal outcome.
1.2. Institutional influences
Even in the third case, the new configuration may not do well becauseof interactions with French institutions. This is the problem of omittedvariables discussed above. In this case, the initial configuration mighthave to include the influence of institutions, e.g. the type of externalfinancial market. In effect, we are now combining practices with insti-tutions and are considering two levels of analysis. In this wider con-sideration, the initial configuration is the following:
Hierarchy Internal Recruit Bank Finance Truth Values:Performance and Quality
1 1 1 1 0
26
Here, a value of 0 for bank finance means that the firm relies uponequity markets. If the external financial institution is a fixed constraint,there are three possibilities:1. how a firm finances does not matter to performance;2. finance matters in conjunction with some configurations but not all;3. how a firm finances is a necessary condition (with equity financerequired in order to achieve quality).
The first possibility represents a null hypothesis. The second possibil-ity suggests that French firms can adopt hybrid forms that suit thenational conditions. For different countries, there will be different con-figurations of practices that generate both high productivity and highquality. An obvious point is that there is not a single best set of prac-tices. But the more important point is that what gets diffused, or shouldbe diffused, from a source country (e.g. the US) varies from country tocountry. In Italy, given its small firm structure, the adoption of Amer-ican corporate organization might well decrease productivity. At thesame time, external recruiting of managers might help performance.The idea of core practices is, then, possibly wrong, for it presumes thatthere are necessary practices. As we have seen, diffusion of practicesfrom one country to the next can be compatible with multiple configu-rations.
The last possibility poses the problem of institutional change. For ifFrench firms wish to achieve both quality and productivity, then therewill have to be changes in financial markets. Institutional change isdifferent than adopting practices, for it concerns the social and politi-cal agreements among diverse actors. In this wider consideration, notonly cognition is a problem (that is, do managers and workers recog-nize the evidence for a better practice) along with the internal politicsof the firm, but also the credible commitments made by various partiesto institutions.
Institutional change poses, then, particular problems for diffusion ofpractices. Of course, the example of superior American quality mightseem anachronistic—if it was ever valid. But if we switch the truthvalue to radical innovation, and the financial system under considera-tion to be venture capital (or its absence), we have indeed posed pre-
27
cisely the contentions in the popular press that financial markets arecritical for the new economy success of the United States. And in fact,France, Germany, and other countries have introduced new stock mar-kets to provide incentives (through initial public offerings) for venturecapitalists.
Yet even the simple idea of introducing a stock market for small firmscan pose complex institutional interactions. We have not, therefore,entirely treated the problem of institutional change. For the problem israrely simply altering a single institution, but rather the considerationof a change on the ensemble of interacting institutions.
Again, an illustration might help. If we return to Soskice’s descriptionof Germany, the corporatist economy is a fragile balance betweencompeting interests. German financial institutions interact with centralbank powers and with national and sectoral unions that bargain forwages. The logic of adopting new practices might require changes inan institution. However, changing a given institution might itself causenational agreements to decay.
Thus, it is not simply an issue of whether a firm should equity finance,but how this affects the strategic behaviors of other economic actors.11
However, the external shareholder representatives of American finan-cial institutions still must sit on supervisory boards consisting of 50%worker and managerial delegates. Whether such piecewise institu-tional changes are possible have yet to be fully observed. Clearly, suchproposals are rejoined with an active debate among the institutionalactors.
1.3. Learning to like root beer
This scenario of the crisis of institutions is a common feature to thecurrent debates in many countries regarding the impact of globaliza-tion and the new economy. In part, this debate is about power; and in
28
11. The Eichel Plan to forgive taxing German companies for restructuring theirholding companies has the appearance of creating more American capital marketpressure on firms.
part it is about persuading each other what needs to be adopted andchanged. Given the causal complexity, these debates will be decidednot on the technical proof of what matters, but by the power of ideas inpoliticized contexts.
However, the characterization of national systems as complementaryinstitutions that support production and innovation within a politicalequilibrium begs the underlying question of when practices and insti-tutions are complements. It is obvious that there are sometimes techni-cal complements. An example is the integer problem in productionplanning: if a steel blast furnace works optimally at 4 million tonscapacity and the iron foundary at 3 million tons, then the optimal con-figuration is 3 blast furnaces and 4 foundaries. Of course, technicalcomplements do not seem to cover many of the practices discussed inthe work practice literature. Hence, complements are sometimes seenas sociotechnical (MacDuffie 1993) or cultural such as in payment sys-tems (Ichinowski, Shaw and Prennushi 1997). Game theory thinks ofcomplements in terms of strategies, or in the credible commitmentsrequired to render credible a course of action.12 Clearly, the term ismore flexible than that suggested by the national systems literature orour Boolean algebra.
In this regard, it is important to remember that complements are oftenissues of tastes in the sense of aesthetics or proper behavior. This con-sideration may seem to stretch the meaning, but it is critical to under-standing the potential for adaptability. In cuisine, we clearly acceptthat complements are cultural. White wine is not drunk with mostcheese in France but is in California; corned beef and mayonnaise ispossible I am told in the US Midwest but not in New York.
Similarly, debates on the virtues of paying stock options are oftencouched in arguments regarding complements (e.g. to motivate ourmanagers in decentralized hierarchies we have to pay them options—the italics indicate the complements). Yet the opposition to stockoptions often argues that they are tasteless or unfair. Accordingly, in
29
12. See the outstanding book by Topkis (1999).
this complex space of culture, technologies, and strategies, there arefar more possibilities for adaptation within national systems than pre-sumed. For surely the attitudes towards money and other economictransactions have changed over the centuries (see the wonderful studyby Zelitzer (1978) regarding life insurance policies in the late 19th cen-tury America). In this regard, Boltanski and Chiapello (1999) are rightto stress the importance of the transformations in the artistic and socialcritiques of capitalism for understanding the adaptive potential of cap-italist economies and their variations and national varieties. To a largeextent, adopting complementary practices is a matter of making cul-turally acceptable what had been anathema before.
� 2. Empirical cases of diffusionThe national systems models provide the starting-point for under-standing diffusion within varying institutional contexts but understatethe adaptive potential of economic and political actors to re-interpretpractices and institutions. The relationship between institutions andpractices is fundamental to understanding the diffusion between theUnited States and Europe. In this exchange of knowledge between thecontinents, the diffusion of practices and institutions are often but notalways linked. When they are linked, the diffusion of practices notonly may be impeded; their adoption may lead to effects quite differ-ent than the home country.
2.1. Diffusion of practices: the multidivisional structure
An excellent illustration for the diffusion of a practice and its relationsto institutions is the diffusion of the multidivisional structure.13 Thehistory of the diffusion of this organizational structure reflects pre-cisely Servan-Schreiber’s observations that the American firm is moreproductive because it is better organized. The multidivisional structurewas invented in the United States just after World War I by Dupontand quickly spread throughout the American economy. The initial dif-
30
13. This discussion summarizes Kogut and Parkinson (1992), while incorporatingstudies subsequent to the publication of this chapter.
fusion showed a strong sectoral pattern, and imitation effects stronglycharacterized its diffusion (Kogut and Parkinson 1998). Teece (1981)showed that its diffusion resulted in increased profitability for earlyadopters. The Armour and Teece (1978) calculated that early adoptionin the oil industry led to over a 2% increase in return on assets, a size-able profit for the capital-intensive oil companies.
As is well known now, the multidivisional structure was taken toEurope not only by American multinational companies but also by theactive agency of consulting companies, especially McKinsey (Chan-non 1970, Kipping 1998, McKenna 2000). Moreover, there had beensome early experimentation with this form in the UK and in Franceprior to World War II. In the case of Japan, interestingly, the multidi-visional form was discovered by indigenous efforts and its diffusionwas quite rapid (Suzuki 1991).
Figure 114 shows the diffusion of the multidivisional structure in someEuropean countries and in Japan15 (the chart is normalized for differ-ences in sample sizes among the countries; exact numbers should notbe taken seriously.) There are two observations to be made from thischart. The first is the more rapid diffusion of the form in the UK and inJapan. One might hypothesize that the differential adoption of the mul-tidivisional structure reflected the larger size and greater diversifica-tion of companies in the leading countries. However, a glance at tables1 and 216 shows that size and diversification do not seem to be relatedto the country patterns in adoption. Differences in adoption speed donot seem to be related to the factors found to be important in the US(Kogut and Parkinson 1996).
The second, and more interesting observation, is that by the 80s, thisform had diffused widely in all of these countries. Why had this formdiffused so widely? Most studies on the diffusion of Taylorism show
31
14. See page 32.15. The European data were collected from archives using the studies of Channon,Dyas and Thanheiser. The Japanese data were provided by Professor Suzuki, drawnfrom his own study; as I cannot check these data to assure comparability, theJapanese data are only indicative.16. See page 33.
remarkable differences across countries and large debates regarding itsintroduction (see Kogut and Parkinson 1993). Clearly, an importantpossibility is that the adoption of the multidivisional structure waslargely institutionally neutral. It did not challenge the institutional bal-ance in the adopting countries. Often, divisionalization did not evenlead to restructuring of assets, merely reflecting changes in the hierar-chical relationships and accounting rules inside the firm. Indeed, incountries such as Germany, the liquidation of companies and theirreintegration into the firm not only posed tax liability issues but alsoremoved labor from supervision over the absorbed companies. Thus, itis not surprising that German divisionalization often maintained thefiscal and juridical identity of the companies, while assigning them toreporting lines within the divisionalized holding structure.
Of course, the other possibility is that the introduction of the multidivi-sional form often reflected a “paper” change rather than deep restruc-turing. There is no question of an element of superficial changes. Andthere was also an element of cultural preference, as revealed by aBritish executive who thought the linking of pay to results was “rathertasteless” (cited in Channon 1973). Yet the historical record of many of
32
0
10
20
30
40
50
60
70
80
90
1932
1947
1948
1951
1953
1955
1957
1932
1959
1961
1963
1965
1967
1969
1971
1973
1975
1977
1979
1981
1983
1985
1987
Britain
Japan
Germany
France
Figure 1Diffusion of the multidivisional structure
in some European countries and in Japan
33
~ 1915 ~ 1950 ~ 1970
France16% 25% 40%
(1912) (1955) (1970)
Germany22% 45.6%
(1949) (1966)
Japan39% 34%
(1950) (1980)
United Kingdom16% 21% 41%
(1909) (1949) (1968)
United States22% 30% 33%
(1909) (1954) (1970)
Sources: For France, Daviet (1988); for Germany and Japan, Suzuki (1991, 1992); forU.K. and U.S., Prais (1976).
Table 1Share of industrial output by 100 largest firms
Single Dominant Related Unrelatedproducts products products
France 20% 28% 43% 8%
Germany 27% 17% 37% 19%
Japan 19% 32% 49% 0%
United Kingdom 6% 30% 58% 6%
United States 7.6% 31% 45.2% 16.2%
Sources: For France and Germany, Dyas and Thanheiser (1976); for Japan, Suzuki(1991); for U.K., (1973); and for U.S., Rumelt (1974).
Table 2Product diversification strategies by country, 1969-1970
these firms shows consistent efforts to reorganize their companiesthrough rather strenuous efforts. In all, the multidivisional structure dif-fused because it was a practice that was institutionally neutral in thesecountries. Where it posed problems—such as in Germany, local adap-tations sufficed to render it politically harmless.
It is, however, a different issue to ask whether the multidivisionalstructure was causally responsible for better performance. Whereas theUS studies are fairly clear in showing increased profits to adoption, theEuropean studies are ambiguous at best. The most intriguing study isthat of Cable and Dirrheim (1984) that showed no effect on profits forGerman firms that adopted.
Is it possible that German firms spent fortunes on restructuring and con-sultancy with little effect on improved performance? This conclusionreflects the possibility that divisionalization had little meaning in theinstitutional context of Germany. With extensive crossholdings amongfirms and banks, divisionalization surely did not play the same role as inthe US in increasing transparency for the shareholder or in increasingshareholders’ representation in corporate decisions. The many studies ondivisionalization in the US never identified the source of the increasedprofitability. Was it improved accounting, better incentives, or bettergovernance? With such blurred causality, it would not be surprising thatthe adoption of the divisional structure should have ambiguous resultsthat depend upon the national location of the adopting firm.
2.2. Diffusion of institutions: private enterprise
The diffusion of practices is a permanent feature of the world econ-omy. With the spread of multinational corporations and their domina-tion of world trade and investment, the diffusion of practices travelthese international arteries more quickly. Whereas their adoptionalways confronts problems of re-interpretation and re-combinationwith existing practices, diffusion is surely better facilitated today bythe increased economic integration.
Institutions, however, do not travel by the arteries of multinational cor-porations. They reflect patterns of behavior that are inscribed in legalcode and in political and economic relationships. Outside the power ofany one actor to change, institutions are social agreements that guideand coordinate the interdependent acts of economic actors in a country.
The thesis of Douglass North and his co-authors is that institutionschange if the costs of maintaining them are too relative to the benefits
34
of change. For example, feudalism collapsed when the plague deci-mated the labor supply, bidding wages up and encouraging workers tobreak their feudal bonds (North and Thomas 1973). Sometimes changehappens by micro-steps, in which institutions corrode by the manydecisions of individuals. Thus, feudalism did not collapse by revolu-tion in most cases, but by the increased power of labor to negotiate itsliberation during a period of labor shortage. Yet in other circum-stances, institutional change occurs through the reconstitution of coali-tional support for their maintenance. It is not surprising, then, thatinstitutional change is often motivated by ideological challenges tonotions of what is best and right.
One of the most important sea changes in the post-war period has beenthe curbing of the power of the State in national economies. In the firstdecades after World War II, the State grew in power and authority in allEuropean countries, as well as in the United States. This growth was duepartly to the military tension of the post-war period that called for largedefense budgets, as well as to meet the demands of social benefits.17
The role of the State, however, varied dramatically by country. Stateownership of key enterprises has been an important influence inFrance, somewhat less in the United Kingdom, and even less so inGermany and Japan. With little prior history of government owner-ship, France experienced two waves of nationalization, just after thewar and with the ascendancy of Mitterrand to the presidency. The UKexperimented for thirty years with direct ownership of steel companiesas well as public utilities, before reversing the policy in the 80s. TheGerman government owned large industrial enterprises inherited fromthe Nazi period, principally Volkswagen and Salzgitter, as well as pub-lic utilities, such as the Bundespost or Lufthansa. Influenced by theAmerican model, Japan did not nationalize enterprises after the war,and has also privatized many of its utilities.
Given these different traditions, it is not surprising that countries dif-fered dramatically in the participation of the government in the owner-
35
17. This literature is too immense to consider to cite. For an incisive study thatreflects the bias of its time, see (Shonfeld 1969).
ship of companies. Shephard (1976) estimates the shares of publicownership in major economic sectors to show Germany at 9% in 1950;Japan at 10% in 1960; the US at 15% in 1960; France at 17% in 1954;and the UK at 25% in 1962. Thus, despite different regulatory regimes,the State took substantial positions in all of these countries.
This role has been greatly curtailed in the past 10 years. Figures 2and 318 show that privatization started in the United Kingdom butquickly diffused to other countries. Thus, in this case, Transatlanticchange started in the UK, even if the ideology has a distinctive Amer-ican flavor.19 Of course, the value of privatizations reflects how muchthe State had to sell, as well as the type of assets being sold. Yet thepicture shows fairly clearly that privatization diffused widely inEurope in the 80s. It has not been a prominent feature of the US, wherederegulation has been the dominant policy to remove the State fromeconomic influence.
The effects of these institutional policies have been quite differentacross countries. The case of France is instructive, because State own-ership has had two consequences. The first is the impact on the careerpaths of the elites. French government policy has been implementednot only through indicative planning and fiscal incentives but alsothrough close personal ties. In 1990, 45% of the CEOs of the 200largest French corporations began their careers inside the government,many of them with degrees from the elite schools (Fridenson 1996) 20
(see also the sociometric study of Kadushin (1995) that confirms theimportance of top school diplomas on subsequent friendship amongfinancial elite. As a consequence of the privatizations, the career pathsof the French elite are less certain now that they have been since theend of World War II.
36
18. See page 37.19. Yergin and Stanislaw (1998) provide an engaging history of the ideological ori-gins of privatization during the Thatcher administration in the UK.20. This pattern of movement from government to industry is known as “pantou-flage”, in reference to house slippers; one need not put on shoes to make the transi-tion. It is interesting to compare this pattern to the Japanese. Schaede (1995) findsthat the Japanese practice for bureaucrats to descend into private firms plays a simi-lar role. More than 50% of the bureaucrats hired by the ministries graduate from the
37
law school of the University of Tokyo. In 1991, of the 177 former governmentbureaucrats that were directors in the 100 largest Japanese firms, 40% were eitherCEO, chairman, or executive director; 5% of all directors were former bureaucrats.
%10
8
6
4
2
0Pre-1984 1985 1987 1989 1991 1993 1995 1997
U.S.
Spain
Portugal
Italy
Germany
France
U.K.
Figure 3
ITALY
PORTUGAL
SPAIN
GERMANY
FRANCE
U.K.
U.S.
35
30
25
20
15
10
5
0
pre .
84
1984
1985
1986
1987
1988
1989
1990
1991
1992
1993
1994
1995
1996
1997
1998
The privatization started in the UK but quikly diffused to other countries
Figure 2
21. It is instructive to note that the evidence for the economic efficiency of priva-tized companies is mixed. See (Coffee 1999) for a terse review.
The second consequence of the nationalizations and privatizationshas been to upset the financial ties between French firms. While theBalladur privatization policy of allocating “noyaux durs” shares toFrench companies insured a core of French ownership, French own-ership links has dissolved in the past 10 years. The deux etoiles ofParibas and Suez no longer are important in French ownership pat-terns. In Germany, the dense intercorporate structures has facilitateda German resistance to the in-roads by foreign equity investments(Kogut and Walker 2000). In France, loss of bank control and pene-tration of foreign equity capital are the highest of the major Europeancountries.
Thus, changes in the French institution of State ownership disequili-brated a national system.21 With decreasing State influence, ownershippatterns, career paths and educational training lack a critical comple-ment to their institutional logic. The growing importance of equityinvestment has opened the door to the discussion of new “complemen-tary” practices, e.g. pay for performance, stock options, that have yetto be re-interpreted in the French context or to be proven in their effi-cacy. No wonder that the arguments for such practices are deeply polit-ical in the tenure of their debate.
2.3. Global systems: skilled labor and the American model
The instability caused by institutional change confirms the claim of thenational systems approach that these systems are tightly coupled sys-tems. Yet the histories of the importation of foreign practices and insti-tutions show a large plasticity in the ability of countries to adapt. How-ever, sometimes adaptation is difficult and blocked. Then whathappens?
One of the most important adjustments made in blocked systems isthe exodus of people to other better performing systems or systemsthat offer these type of individuals better opportunities. The UnitedStates is, in this regard, a predatory system. Lacking the corporatist
38
institutions of Europe, the United States underperforms in manyareas, especially in the education of sufficient numbers of skilledtechnicians and engineers. However, as an open system, it providesa career path for the educated world elite and arguably constitutesa system that innovates for the benefit of the world economy.Predatory behavior and innovation are the two sides of the Ameri-can system.
There are two important elements to this system. The first are the veryclose ties between universities, research centers, and business, often ata regional level. In a fascinating comparative study of the French andAmerican biotechnology industries, Michelle Gittelman (2000) foundthat France and the United States did not differ widely in the quality oftheir basic science. Both French and American start-up companiesshowed similar innovatory records; these companies performed betterin both countries than incumbent pharmaceutical firms. What explainsthe better American record of getting science into patented innovationswas simply that there are very few French start-up companies andmany more American ones. The French scientific system created insti-tutions for science (and for large-scale innovations) but not for radicaltechnological innovations.
The United States has suffered from the shortage of skilled labor due,it is widely believed, to the poor quality of education in many geo-graphical areas. The second important element to the open systemsof the United States is, then, its immigration policy. This policyworks through two mechanisms. The first is the important roleplayed by higher education in attracting the best students fromaround the world, then providing them with work visas to stay in theUS for high-skill jobs. The second mechanism is the special visasallowed for accepting highly skilled labor into the US, either for uni-versity or industrial positions. In recent years, this type of visa hasbeen fully subscribed, with educational institutions complaining thatthey are unable to fill academic posts with the best candidates in theworld as a consequence.22
39
22. Other countries, especially Germany but also Singapore, have adopted similarpolicies in recent years.
Figures 4 and 523 graph the number of skilled labor immigrants into theUS for the past few decades. Because of a change in law, the numberschange in the early 90s, as do the reporting by individual countries.Clearly, we see a rise in Asian migrants, but we also see political con-siderations, such as the number of Polish and Irish immigrants – bothpolitically sensitive ethnicities in the US.
But the important point of this chart is the long tradition of the US inaccepting highly skilled labor into its economy as part of its techno-logical strategy. As Saxenian has shown in her recent studies, expatri-ate labor head over 25% of new start-ups in the Silicon Valley (Saxe-nian 1999). The US national system is global in its premeditateddesign to attract the best-educated and able, a design particularlyattractive given the high variance of primary and secondary education.
In this regard, the US succeeds not because it imports practices moreefficiently than other countries or because its institutions create thehigh-skill domestic equilibrium that Soskice claims is the Germanachievement. The US system succeeds because it identifies the knowl-edge held by individuals to be the asset that it wishes to import fromthe global economy. The United States is not a national system, it is aglobal one.
It may well be hard for other countries, given their institutionalizededucational and labor markets – not to mention the frequent culturaloutbursts against immigrant populations – to replicate this aspect ofthe US system. There is, clearly, a collective action problem, insofarthat the United States can be seen as predatory and free-riding on thehuman capital investments of other countries. The term “brain drain”appears to be a peculiarly unfair tax on the educational investmentsof developing countries. Yet it can also be argued that this drain hasled to innovations that, though originating in the US, drive the inno-vative frontier of the world economy through diffusion and higherproductivity.24
40
23. See page 41.24. My thanks to Pierre Jacquet for this observation.
41
United Kingdom
Germany
R.o. Europe
Netherlands
IrelandFrance
Poland
1982 1985 1987 1989 1991Year
Per
sons
1,600
1,400
1,200
1,000
800
600
400
200
0
Figure 5Skilled European worker entry into US
N. America
R.o.W.
Europe
Asia
1985 1987 1989 1991 1993 1996 1997Year
28,000
26,000
24,000
22,000
20,000
18,000
16,000
14,000
12,000
10,000
8,000
6,000
4,000
2,000
0
Per
sons
Figure 4Skilled worker entry into US
� Conclusions
These short histories of the diffusion of practices and institutionsacross borders point to the importance of understanding the politicaleconomy of national systems. Given the high ambiguity in causal rela-tions among practices and institutions, the adoption of new ways ofdoing things is rarely determined by transparent efficiency. Rather,adoption of institutions or practices that impinge upon institutions areaccompanied by an ideological and cultural reconfiguration amongeconomic and political actors.
There is a cautionary inference from this analysis. It is hard to isolate acausal factor in the success of another country which, if moved byitself, would have the expected consequences. An example is thebelieved importance of finance and venture capital in the technologicalsuccess of the US. Yet it is likely that functional equivalents exist inother countries. The current importation of venture capital also pre-sents institutional conflicts that may impede its efficacy. No doubtthere are interested parties to the diffusion of venture capital. It isexactly in this microdynamic by which economic actors seek to decou-ple and recouple institutions that national systems change, but rarely inthe direction that any actor expected.
42
References
Albert, Michel, 1993, Capitalism versus Capitalism: How America’s Obsession withIndividual Achievement and Short-term Profit Has Led it to the Brink of Collapse,New York: Four Walls Eight Windows.
Aoki, Masahiko, 1990, “Toward an Economic Model of the Japanese Firm,” Journalof Economic Literature, 28: 1-27.
Armour, H. and D. Teece, 1978, “Organizational Structure and Economic Perfor-mance,” Bell Journal of Economics, 9: 106-22.
Bendix, Reinhard, 1956, Work and Authority in Industry; Ideologies of Managementin the Course of Industrialization, New York: Wiley.
Berger, Suzanne and Ronald Dore (eds.), 1996, National Diversity and Global Capi-talism, Ithaca: Cornell University Press.
Boltanski, Luc, 1982, Les cadres : la formation d’un groupe social, Paris: Editions deMinuit.
Boltanski, Luc and Eve Chiapello, 1999, Le nouvel esprit du capitalisme, Paris:Gallimard.
Boyer, Robert, 1996, “The Convergence Hypothesis Revisited: Globalization but stillthe Century of Nations?”, in Suzanne Berger and Ronald Dore (eds.), National Diver-sity and Global Capitalism, Ithaca: Cornell University Press.
Boyer, Robert, 1998, “Hybridization and Models of Production,” in Robert Boyer,Elsi Charron, Ulrich Juergens, and Steven Tolliday (eds.), Between Imitation andInnovation. The Transfer and Hybridization of Productive Models in the InternationalAutomobile Industry, Oxford: Oxford University Press.
Brannen, Mary Yoko, Jeff Liker and Mark Fruin, 1999, “Recontextualization and Fac-tory-to-Factory Knowledge Transfer from Japan to the U.S.: The Case of NSK,” withJeffrey K. Liker and W. Mark Fruin, in Jeffrey K. Liker, W. Mark Fruin, and Paul S. Adler (eds.), Remade in America: Transplanting and Transforming Japanese Pro-duction Systems, New York: Oxford University Press, 1999: 117-154.
Cantwell, John, 1989, Technological Innovation and Multinational Corporations,Oxford: Blackwell.
Chandler, Alfred, 1990, Scale and Scope: The Dynamics of Industrial Capitalism,Cambridge, MA: Belknap Press of Harvard University.
Channon, Derek F., 1973, The Strategy and Structure of British Enterprises, Boston:Division of Research, Graduate School of Business, Harvard University.
Coffee, John C., 1999, “Privatization and Corporate Governance: The Lessons fromSecurities Market Failure,” working paper, Columbia University Law School.
43
Daviet, Jean-Jacques, 1989, Une multinationale à la française. Saint-Gobain 1665-1989, Paris: Fayard.
Djelic, Marie-Laure, 1998, Exporting the American Model: The Postwar Transforma-tion of European Business, Oxford: Oxford University Press.
Dosi, Giovanni and Bruce Kogut, 1993, “National Specificities and the Context ofChange: The Coevolution of Organization and Technology,” in Bruce Kogut (ed.),Country Competitiveness, New York: Oxford University Press.
Dyas, Gareth P. and Heinz T. Thanheiser, 1976, The Emerging European Enterprises:Strategy and Structure in French and German Industry, London: Macmillan.
Fridenson, Patrick, 1996, “The Relatively Slow Development of Big Business in 20thCentury France,” mimeo.
Gittelman, Michelle, 1999, Scientists and Networks: A Comparative Study of Coope-ration in the French and American Biotechnology Industry, Ph.D. thesis, WhartonSchool, University of Pennsylvania.
Hall, Peter A. (ed.), 1989, The Political Power of Economic Ideas: KeynesianismAcross Nations, Princeton: Princeton University Press.
Kadushin, Charles, 1995, “Friendship Among the French Financial Elite,” AmericanSociological Review, 60: 202-221.
Kipping, Matthias, 1999, “American Management Consulting Companies in WesternEurope, 1920 to 1990: Products, Reputation, and Relationships,” Business HistoryReview, 73: 190-220.
Ichniowski, C., K. Shaw and G. Prennushi, 1997, “The Effects of Human ResourceManagement Practices on Productivity,” American Economic Review, 87: 291-313.
Kogut, Bruce, 1992, “National Organizing Principles of Work and the ErstwhileDominance of the American Multinational Corporation,” Industrial and CorporateChange, 1: 285-325.
Kogut, Bruce, 1999, “The Recoupling of a Somewhat Coupled System: Weimar Ger-many and American Work Practices,” The Wharton School, working paper.
Kogut, Bruce, J.P. MacDuffie and Charles Ragin, 1999, “Prototypes and Fuzzy WorkPractices,” Reginald H. Jones Center Working Paper Series, WP99-08.
Kogut, Bruce and David Parkinson, 1993, “The Diffusion of American OrganizingPrinciples to Europe,” in Bruce Kogut (ed.), Country Competitiveness: Technologyand the Organizing of Work, Oxford: Oxford University Press.
Kogut, Bruce and David Parkinson, 1998, “Adoption of the Multidivisional Structure:Analyzing History from the Start,” Industrial and Corporate Change, 7: 249-273.
Kogut, Bruce and GordonWalker, 2000, “The Small World of German Ownership;the Durability of National Networks,” forthcoming American Sociological Review.
Kogut, Bruce and Udo Zander, 1992, “Knowledge of the Firm, Combinative Capabi-lities, and the Replication of Technology,” Organization Science, 3: 383-97.
44
MacDuffie, John Paul, 1993, “Human Resource Bundles and Manufacturing Perfor-mance: Organizational Logic and Flexible Production Systems in the World AutoIndustry,” Industrial & Labor Relations Review, 48: 197-221.
McKenna, Chris, 1999, “Economies of Knowledge: A Theory of ManagementConsulting,” mimeo, ESSEC.
Merton, Robert, 1949, Social Theory and Social Structure: Toward the Codification ofTheory and Research, Glencoe, IL: Free Press.
North, Douglass and Robert Thomas, 1973, The Rise of the Western World. A NewEconomic History, Cambridge: Cambridge University Press.
Piore, Michael and Charles Sabel, 1984, The Second Industrial Divide: Possibilitiesfor Prosperity, New York: Basic Books.
Prais, S.J., 1976, The Evolution of Giant Firms in Britain: A study of the Growth ofConcentration in Manufacturing Industry in Britain, 1909-70, New York: CambridgeUniversity Press.
Ragin, Charles, 1987, The Comparative Method, Berkeley, CA: University of Califor-nia Press.
Rumelt, Richard P., 1974, Strategy, Structure and Economic Performance, GraduateSchool of Business, Harvard University: Boston.
Sabel, Charles and Jonathan Zeitlin, 1985, “Historical Alternatives to Mass Produc-tion: Politics and Technology in Nineteenth Century Industrialization,” Past and Pre-sent, 108: 134-176.
Sabel, Charles and Jonathan Zeitlin (eds.), 1997, World of Possibilities: Flexibility andMass Production in Western Industrialization, New York: Oxford.
Saxenian, Annalee, 1999, “Silicon Valley’s New Immigrant Entrepreneurs,” PublicPolicy Institute of California.
Schaede, Ulrike, 1995, “The ‘Old Boy’ Network and Government-Business Relation-ships in Japan,” Journal of Japanese Studies, 21: 293-317.
Servan-Schreiber, Jean-Jacques, 1967, Le défi américain, Paris: Denoël.
Shepherd, William, 1976, Public Enterprise: Economic Analysis of Theory and Prac-tice, Lexington, MA: Lexington Books.
Shonfield, Andrew, 1965, Modern Capitalism, the Changing Balance of Public andPrivate Power, New York: Oxford University Press.
Simon, Herbert A., 1957, Models of Man: Social and Rational; Mathematical Essayson Rational Human Behavior in a Social Setting, New York: Wiley.
Soskice, David, 1990, “Wage Determination: The Changing Role of Institutions inAdvanced Industrialized Countries,” Oxford Review of Economic Policy, 6: 36-61.
Stark, David, 1996, “Recombinant Property in East European Capitalism,” AmericanJournal of Sociology, 101: 4, 993-1027.
Suzuki, Yoshitaka, 1991, Japanese Management Structures, 1920-80, New York:St.Martin’s Press.
45
Teece, D. J., 1981, “Internal Organization and Economic Performance: An EmpiricalAnalysis of the Profitability of Principal Firms,” Journal of Industrial Economics, 30:2 (December), 173-199.
Topkis, Donald, 1999, Complementarities and Supermodularity, Princeton: PrincetonUniversity Press.
Useem, Michael, 1996, “Corporate Restructuring and the Restructured World ofSenior Management,” in P. Osterman (ed.), Broken Ladders: Managerial Careers inthe New Economy, 23-54, New York: Oxford.
Westney, D. Eleanor, 1987, Imitation and Innovation: The Transfer of Western Orga-nizational Patterns to Meiji Japan, Cambridge: Harvard University Press.
Yergin, Daniel and Joseph Stanislaw, 1998, The Commanding Heights, New York:Simon and Schuster.
Zander, Udo, 1991, Exploiting a Technological Edge - Voluntary and Involuntary Dis-semination of Technology, Doctoral Dissertation, Stockholm School of Economics.
Zeitlin, Jonathan and Gary Herrigel (eds.), 2000, Americanization and its Limits,Oxford: Oxford University Press.
Zelitzer, Viviana, 1978, “Human Values and the Market: The Case of Life Insuranceand Death in19th-Century America,” American Journal of Sociology, 84: 591-610.
46
les notes de l’ifri
Sér i e t ransa t l an t i que
US and ESDI in the New NATOLes États-Unis et l’IESD dans la nouvelle OTANPhilip GordonN° 4, 56 pages, juillet 1998, 35 F, 5,34 €.
Du cinéma au multimédia.Une brève histoire de l’exception culturelleLaurent Burin des RoziersN° 5, 96 pages, septembre 1998, 45 F, 6,86 €.
Où en est l’Alliance atlantique ? L’improbable partenariatFrédéric BozoN° 6, 64 pages, novembre 1998, 30 F, 4,57 €.
Where Does the Atlantic Alliance Stand? The Improbable PartnershipFrédéric BozoN° 6 bis (version anglaise du n° 6), 64 pages, février 1999, 30 F, 4,57 €.
The United States, Iran and Iraq: Containment or Engagement?Les États-Unis, l’Iran et l’Irak : endiguement ou dialogue ?Geoffrey Kemp, Phebe Marr et Raymond TanterSous la direction de Bassma Kodmani-DarwishN° 7, 96 pages, novembre 1998, 45 F, 6,86 €.
The United States and the Arab Gulf MonarchiesLes États-Unis et les monarchies arabes du GolfeJoseph A. KechichianSous la direction de Bassma Kodmani-DarwishN° 8, 64 pages, janvier 1999, 30 F, 4,57 €.
Anglo-American Relations and Crisis in Yugoslavia (FRY)Les relations anglo-américaines à l’épreuve de la crise yougoslaveJane M. O. SharpN° 9, 84 pages, mars 1999, 40 F, 6,10 €.
47
NB : Les nos 4, 5, 8, 9 et 10 sont en vente uniquement à l’Ifri. Les nos 6 et 7 età partir de 11 sont diffusés par la Documentation française (29, quaiVoltaire, 75344 Paris Cedex 07, Tel. 33 (0)1 40 15 70 00).
Le partage du fardeau dans l’OTANVolume 1 – Enjeux et réalitésNicole GnesottoN° 11, 68 pages, juin 1999, 30 F, 4,57 €.
Burdensharing in NATOVolume 2 – The US and Transatlantic BurdensharingStanley R. SloanSous la direction de Nicole GnesottoN°12, 56 pages, juin 1999, 30 F, 4,57 €.
Burdensharing in NATOVolume 3 – The German PerceptionKarl Heinz KampSous la direction de Nicole GnesottoN°13, 50 pages, juin 1999, 30 F, 4,57 €.
Les États-Unis et la France. La puissance entre mythes et réalitésPhilippe Moreau DefargesN° 14, 72 pages, juin 1999, 35 F, 5,34 €.
Banalisation et restructuration des industries de défenseFrédérique SachwaldN° 15, 62 pages, juin 1999, 45 F, 6,86 €.
Defence Industry Restructuring: The End of an Economic ExceptionFrédérique SachwaldN° 15bis, 62 pages, octobre 1999, 45 F, 6,86 €.
Le triangle atlantique. L’émergence de l’Amérique latinedans les relations Europe-États-UnisAlfredo G. A. ValladãoN° 16, 64 pages, juin 1999, 40 F, 6,10 €.
The Euro and the Dollar: An Agnostic ViewClaude Bismut et Pierre JacquetN° 17, 54 pages, juillet 1999, 40 F, 6,10 €.
La mondialisation, l’euro et le système monétaire mondialÉdouard BalladurN° 18, 52 pages, mars 2000, 45 F, 6,86 €.
Globalization, Euro and the World Monetary SystemÉdouard BalladurN° 18bis, 48 pages, mars 2000, 45 F, 6,86 €.
48
Russia’s Security Relations with the West After Kosovo and ChechnyaDmitri TreninN° 19, 48 pages, mai 2000, 45 F, 6,86 €.
Irak et Serbie. Les sanctions économiques au cœur du débattransatlantiqueJacques BeltranN° 20, 60 pages, mai 2000, 45,92 F, 7 €.
Evaluating NATO’s Efficiency in Crisis ManagementKori SchakeN° 21, 48 pages, août 2000, 45,92 F, 7 €.
Le recours à la force dans l’affaire du Kosovo et le droit internationalSerge SurN° 22, 40 pages, septembre 2000, 45,92 F, 7 €.
L’Amérique et l’AllianceÉtienne de DurandN° 23, 64 pages, octobre 2000, 45,92 F, 7 €.
The New American ChallengeLe nouveau défi américain
The New American Challenge and Transatlantic Technology SourcingFrédérique SachwaldN° 24, 68 pages, septembre 2000, 49,20 F, 7,5 €.
Why and How European Companies Reach Out to Silicon ValleyThierry WeilN° 25, 80 pages, octobre 2000, 49,20 F, 7,5 €.
The Transatlantic Exchange of Ideas and Practices: National Institutionsand DiffusionBruce KogutN° 26, pages, décembre 2000, 45,92 F, 7 €.
Changements organisationnels : de la nouvelle économie aux 35 heuresPhilippe AskenazyN° 27, à paraître
Hegemony Wired: American Politics and the New EconomyEthan B. KapsteinN° 28, 60 pages, novembre 2000, 45,92 F, 7 €.
49
Dynamiques industrielles et réglementaires des télécoms : une comparaison États-Unis/EuropeOlivier Bomsel et Gilles Le BlancN° 29, 70 pages, novembre 2000, 49,20 F, 7,5 €.
The Information Society Race: Is Europe a Follower or a Challenger?Michel CatinatN° 30, à paraître
Thèmes des ouvrages à paraître
National Missile Defense (NMD)Michael Nacht
Défense antimissile (NMD)Thérèse Delpech
50
Mise en page et impressionbialec, nancy
Dépôt légal n° 53007 - janvier 2001