yannick lung

7/29/2019 Yannick Lung

1/35

The Link between the Diversity of Productive Models

and the Variety of Capitalisms

A Review of the Literature and Contextualisation

Using the Car Industry as a Case Study

Yannick Lung

Between the representative firm found in partialequilibrium analysis defined by itshomogeneity/identity hypothesis1 and the radicalheterogeneity characterizing the evolutionistapproach2, many studies have tried to analyse the

diversity of organisational forms, withorganisational theory literature having beenparticularly fruitful in this respect. The presentarticle does not purport to summaries all of thisresearch. Conducted under the aegis of the ESEMKproject, which analyses the variety dynamicsunderlying Europes socio-economic models, itfocuses on a much more limited sub-section withinthis corpus, the one that tries to reflect upon thediversity of firms organisational modes byanalysing them in terms of variety of capitalisms.In this approach, organisational models are linkedto the institutional environments in which firmsoperate. The crux here is the interconnectionbetween the micro and macro levels, an orientationexcluding studies that concentrate on the sectoriallevel alone. Nevertheless, no micro/macro linkage

1 Pure and perfect competition hypotheses imply thatall firms are perfectly identical: all produce the samegood, have access to the same information (using thesame technologies) and factors of production, etc. Infact, as Coase states, economic analysis does not really

incorporate the firm per se.2 Who considers that each firm is different becauseits history is singular, meaning that its learning isspecific.

can do without an intermediary level, to wit theindustry level, which remains the dominant spherein the structuring of a competitive process.

Institutional form diversity approaches that linkthe micro (firm) level to the meso (industry/sector)

and macro (economy/country) levels contain twokinds of stances. The first derives from ahypothesis of isomorphism between the firm andinstitutional environment levels (Di Maggio,Powell, 1983). Such approaches stress theimportance of institutional constraints that arefound in the environment and which generate adominant (if not exclusive) form of organisation.Jacksons formulation for this is as follows:

Institutional constraints often lead to

institutional isomorphism whereby organizationsadapt similar structures and routines (Jackson,2004)

Approaches of this ilk can be dynamic in nature,revolving around successions of varyingorganisational forms (models) throughoutcapitalisms history. The idea here is that changesin firms organisational forms constitute the driverand/or outcome of capitalisms transformations(Chandler, 1990; Langlois, 2003). They are also

synchronic, emphasizing the singular nature offirms organisations in different national contexts:French model (Lesourne, 1998) or German


2/35

110Actes du GERPISA n38

model (Vitols, 2004), etc. are used to describe the development. This involved a diachronic kind ofparticularisms of national economies and/or theirdomestic firms.

In both cases, there is only one dominant (if notexclusive) organisational mode that will be linkedwith the socio-economic environment and selectedfor its economic performances. Of course, thevariety of socio-economic models translates intodiverse firm forms in different countries or sectors- but within a given geographic or historicalcontext, only one way of organising things will beefficient. What we have here is an avatar of theone best way hypothesis.

The second stance is the one underlying theESEMK project (Amable, Lung, 2003). It starts byhypothesizing a diversity of firms organisationalmodels in one and the same institutionalenvironment, that of a given industry in a given

country. This diversity lies at the very heart of theinstitutional dynamics in question.Complementarity/rivalry/otherness relationshipsexist between the models that serve to catalyse, atleast partially, institutional change.

Starting with a productive models analyticalgrid proposed by Robert Boyer and MichelFreyssenet (2000b) based on GERPISA studies ofthe history of the car industry, this paper will try toascertain both the conditions under which said gridcan be applied in other sectors, and also its links to

the meso and macro levels, notably when crossedwith Bruno Amables 2003 social innovation andproduction models analytical grid. The aim here isto specify the interactions and dynamics amongstthese varying levels.

THE ISOMORPHIC HYPOTHESIS:

A FIRMS ORGANISATION AS DETERMINED BY

ITS INSTITUTIONAL ENVIRONMENT

Most approaches to the diversity of institutionalforms seek to analyse institutional dynamics at aglobal level (Boyer, 2004a). Apprehended as aninstitution, the firm is the one (and for someanalysts, the key) component that many analystsuse when trying to characterize a stabilisedinstitutional configuration/arrangement.

This infers the existence of a strict equivalencebetween characterisations of socio-economicmodels (types of capitalism) and specific forms ofcorporate organisation.

An initial set of studies pursued a historicalapproach focusing on the institutional changesassociated with the different stages of capitalist

diversity, with the organisational form that is thefirm evolving in a way reflecting the institutionalenvironments attributes. Conversely, synchronicapproaches try to explain the coexistence ofdifferent types of firms. The focus here is less onan analysis of the dynamics of institutional changeand more on a justification of the contemporarydiversity of institutional configurations that areapprehended either at a macro-economic level(variety of capitalisms) or else at a sectorial level,when the observer is scrutinising theinterconnections between (and the coherence of)the various institutional forms that are beingestablished at the different industries levels.

Historical approaches

Here the idea is that each period of capitalismcorresponds to a dominant form in the coordinationof economic activities. This vision is explicitamongst authors who want to characterize changesin modes of coordination (hence firmsorganisation). It is implicit in studies highlightingcapitalisms periodisation, notably in theRegulation Theory framework. As a flag bearer formodern times, the car industry is the sectorwhose emergent models catalysed mass productionin the 20th century. Hence the legitimacy of startingout by retracing its trajectory, such as many experts

have represented this.

The car industrys different phases andsuccession of one best ways

The car industry offers a remarkable example ofa historical vision geared towards a succession ofdominant forms of production (cf. figure 1): craftproduction, Fordism and Toyotaism, and evenmodular production today (Womack, Jones, Roos,1990).

Craft production would characterise the

emergence of the car industrys late 19th century inEurope (Bardou et al., 1997). Carmakers had lowproduction volumes and offered a broad variety ofmodels to satisfy the expectations of a well-to-doclientele, resulting in non-price competition(Abernathy, 1978). This was an era of carmakerentrepreneurs like Louis Renault, actors capable ofdesigning high performance vehicles (motor racesalready existed at the time). Much of these earlycarmakers production function would besubcontracted (starting with bodywork) and they

had little control over distribution. Medium-sizedfirms (which can be called U-form laWilliamson [1985]) constituted the rule.


3/35


Figure 1. The historical succession of the car industrys dominant modes of production

Craft productionEurope

USA

Mass production (Fordism)

Japan Lean production &

supply (Toyotaism)

Modular assembly &

supply

1880s 1910s 1970-80s 1990-2000s

Source: based on Sako, 1999

The shift to a mass production is thought tohave been the doing of the Ford Motor Company inthe early 1910s. The global auto industrys newcentrality moved to the United States (Detroit,Michigan).

Factors like the single model (Ford T),standardised and interchangeable parts andcomponents, the assembly line, mechanisation(later automation), vertical integration,decentralised assembly (branch plants)internationalisation, increased pay ($5 a day) andcontrolled distribution networks all helped toincrease the scale of production to something like 2million vehicles a year. Reading this long list ofmodifications, we get a sense of how theyrepresented a break from the past, and of their

mutual coherency. Alfred Sloan introducedchanges in this system. Having noted the shift to aproduct renewal type of market, GMs helmsmandeveloped from 1920s onwards an approach thatnot only reconciled models volume and diversity(through the use of multiple brands) but alsoproduct marketing, customer credit facilities andaccelerated obsolescence based on an annualrenewal of models. With the affirmation of unionpower, this model stabilised in the immediateaftermath of World War II, to such an extent that itbecame the benchmark for a wage bargaining mode

that was grounded in a rigorous definition of jobsand functions and based on a productivity gain-driven rise in real wages. In this context, massproduction translated into large vertically

integrated firms, organisations whose managementwas rooted in multidivisional structures(Williamsons M-form) and associated withmanagement control systems. Alfred Chandlers1962 book recapped the history of multidivisional

firms in the United States, with GM constituting alaboratory for this form.

The mass production model that emerged in theUnited States diffused progressively to Europeright after WWII, with French, Italian and Germancarmakers following the same trajectory (a singlemodel produced in large series, assembly linework, etc.) before expanding their product range.Europe would use this period to convert to USfirms managerial virtues. However, this was not atime for the diffusion of mass production in Japan,

where the economic and institutional environmentdid not lend itself to this model. Standard analysisis that the Japanese markets demand for diversityand dearth of available resources (raw materials,space, workforce, etc.) induced carmakers here toseek innovative modes of organisation anchored inthe general principle of a downstream-drivenproduction (pull system). This intimated constantquality throughout the value chain, plus flexibilityto cope with variations in demand (Coriat, 1991).With the Toyota production system, Toyotaism andlean production, this Japanese production model

was widely described and popularised in the late1980s, notably through the MIT InternationalMotor Vehicle Programs lean productionconstruct. Womack, Jones and Roos 1990s


4/35


bestseller tried to characterize a new form of ever since Adam Smith have described by handorganisation; a model attained by implementingbest practices in different areas. Leaner design,production and distribution would become ageneral principle, raising questions about US firmsstructuring.

In the light of recent moves towards productionon demand (customization) within the frameworkof increasingly broad product ranges offered byglobalised automobile groups that deal with modelarchitecture design functions by outsourcing eventhe most complex components design andproduction to first tier suppliers, some modernobservers have started talking about a new form oforganisation called modular production, whereinthe automotive industry draws its inspiration fromDellism, an IT model (Sako, 2003) ) or intelecommunication industry (Fleury, Leme-Fleury,

2005).As attractive as our panorama may be, it is

necessarily incomplete. We know that flexibleproduction first appeared in the 1920s (Hounshell,1984); that the Sloanian and Fordian models aredifferent and even incompatible (Boyer,Freyssenet, 2000a); that the original variety ofJapanese automobile demand is a myth (Belis-Bergouignan, Lung, 1993); that the Japanese modelof management never existed (Freyssenet, 2001);and that the car industry has deeper roots in an

integrated architecture than it does in a modularone (Fujimoto, 2001). The pitfalls of seeking anall-encompassing vision become particularly acutewhen people try to make generalisations.

Periodisations in organisational change

The organisational change approach believes inthe existence of an efficient form of organisationfor the firm apprehended as an institution - a formthat depends on the historical context. This stancepervades literature, notably the broader and more

summative overviews. Note two series of studiesthat are directly related to the approach underlyingthe ESEMK project: readings focused on changesin economic agents dominant modalities forcoordinating activities with one another; and theperiodisation offered by Regulation Theory.

Playing different hands: from the invisible tothe visible to the vanishing hand

Arguably, the automotive industrys trajectoryof historically contextualised (i.e., non absolute)forms of optimal organisation at any one momentin time offers a specification of broader changes inthe modalities of efficient coordination that authors

metaphors. These representations are more or lessimplicitly associated with the portrayal of asuccession of industrial revolution

In this view, the First Industrial Revolution,which founded contemporary capitalism, wasmainly rooted in market-based coordination(Smiths invisible hand) amongst small ormedium-sized firms. The Second IndustrialRevolution (late 19th century) gave birth to thelarge modern company, where managers visiblehand (Chandler, 1977) replaced the marketsinvisible one. Managerial theory has broadlydescribed and analysed this rising technocracy,whose need to coordinate activities ex ante (and notjust ex post, as the market does) led to thedevelopment of a managerial science whose aimhas been to promote the efficient management of

large firms. Alfred Chandler accounted for twoforms of managerial capitalism: a competition-oriented form prevalent in the UK and the US; anda more cooperative form in Germany. Thisreplicated Michel Alberts opposition betweenAnglo-Saxon and Rhinean capitalism (see below).

In Business organization and the myth of themarket economy, William Lazonick (1991)extended Chandlers analysis by showing how thehistorical changes that capitalism went throughafter its 19th century British avatar (called

proprietary capitalism due to its unitary featureof family-run companies associating ownership andcontrol) led to a need for greater ex antecoordination, thus providing the impetus for a shifttowards managerial capitalism in the UnitedStates, and more recently towards collectivecapitalism in Japan (with its keiretsu).Thishistorical vision ties into frequent representationsof the car industrys trajectory during the 1990s,when several authors (like Sturgeon, 2002)analysed the Japanese economys problems and therising fortunes of the US (with its ICT-based new

economy, the end of the Solow paradox, etc.) byaverring that centrality had returned to NorthAmerica, or at least to California on its West Coast.

In this view, we now find ourselves in a ThirdIndustrial Revolution, one that revolves aroundinformation and forces us to rethink organisationalmodes. As opposed to the rigidity of managerialfirms requiring perennial restructurings (rangingfrom an 8 year product cycle in automobiles to a20/30 year cycle in aeronautics, both the epitomesof modern industries), what we supposedly have

today is a flexibility constraint in a permanentinnovation regime, as well as an environmentmarked by the waging of strong competitive


5/35


rivalries on a global scale. The crux here is firmsresponsiveness in coping with technologicalinnovation and strategic uncertainty. Electronicgoods life cycles last just a few months (andsometimes just a few weeks). This illustrates theaccelerated temporality associated with the newICT industries, hence the need for a new corporatemodel (Lazonick, 2005).

Furthermore, such responsiveness is said to bebeyond the purview of the large integratedstructures that used to typify the managerial era.Hence the idea that we have entered into avanishing hand phase (Langlois, 2003) in whicheach production problem corresponds to atemporary configuration that brings together a widearray of actors, depending on the project inquestion. Market-oriented coordination betweenindependent economic units becomes central again,

although there is still a need for ex anteconvergence between the activities of thetemporary configurations partners, but thistemporary need for coordination cannot beachieved by forms of coordination based on aauthority principle (hierarchy): the network firm orthe network of firms would be the efficient modeof organisation.

However, several authors (Prencipe, Davies,Hobday, 2003; for a summary, see Frigant, 2005)have shown that the kind of systemic architecture

that a number of industries (including theautomobile sector) have preserved precludes themodular organisational structure. This approach,opposing modular and integral industries(Fujimoto, 2002), offers a framework conducive toa synchronic vision of organisational formsdiversity (see below, 1.2).

Periodisation according to Regulation Theory

Regulation Theory (Aglietta, 1976; Boyer,1983, 2004; Boyer, Saillard, 1995) offers aperiodisation of capitalisms different phases. Thestatus of the firm is not particularly significant inthis conceptual toolbox (Lung, 2005), which startsout by trying to characterize accumulation regimesand modes of regulation as a combination of fiveinstitutional forms: monetary regime; the wage-labour nexus; forms of competition; Stateintervention; and modalities of internationalinsertion.

Regulation Theorys main goal has been toexplain Post-war economic growth, characterizedby strong and continuous increases in productivity

gains, thereby enabling regular rises in wages.Another focus has been how the virtuous circleunderlying a development model labelled

Fordism fell into crisis in the 1960s-70s. In thisview, the Golden Age of Fordism constituted aperiod of reference corresponding to a kind ofregulation that can be called administered ormonopolistic and involving a capitalism of largeunits. This was in opposition to 19th centurycompetitive regulation, marked by thepredominance of small and medium-sized firms.What we have here is a hollow (implied)representation of the Fordian firm associated withthe Fordist development mode a firm thatBenjamin Coriat and Olivier Weinsteincharacterised via five intrinsic attributes:9 a place of antagonism between capital and

labour (),9 a place for the implementation of principles and

protocols (),9 large-sized companies that are usually vertically

integrated (),9 both the expression of and a breeding ground

for a set of contractual practices (),9 a place for establishing standards and norms

(Coriat, Weinstein, 1995, pp.170-2).

The helpful contribution of Michael Piore andCharles Sabels The Second Industrial Divide(1983) will have been to remind people thatfocusing on mass production may cause them to

forget the deployment of other forms of (notablycorporate) production, even during the Golden Ageof Fordism. It highlights the synchronic diversity ofproductions forms of organisation (an ideaextended in Marangoni and Solaris 2004 paper) byinterpreting this in productive model terms. Here,historical perspectives and regulationists muchbeloved periodisation construct act as a catalyst fora diachronic reading focused on a succession ofmodels over time. It was in their vain search for a1990s post-Fordian firm that Regulationists ran outof steam, yielding to the charms of institutionalist

approaches, starting with Masahiko Aokisopposition between A and J firms.

Firms and the diversity of contemporary

forms of capitalism

Debates on the diversity of forms ofcontemporary capitalism became central after thecollapse of the Berlin Wall, which renderedobsolete the opposition between the Socialist andCapitalist systems. There has also been the impactof globalisation, which some see as leading to a

homogenised world, thanks to the diffusion of anefficient capitalist model based both on marketpredominance as a mode of coordination and on the


6/35

glorification of private initiative. In this respect,Michel Alberts book Capitalism against capitalismis quite premonitory since it defends a moresocial form of capitalism described asRhinean, in opposition to an Anglo-Saxonform whose remarkable efficiency is associated

with major inequalities (Albert, 1991). The 1990ssupposedly featured the triumph of this Anglo-Saxon model, with several authors having tried, onvarious levels, to show that it only constitutes oneorganisational form amongst many others (Amable,2003, 2005), and that such forms are a function ofnational or sectorial contexts.

Macro-micro isomorphism

National systems and forms of the firmThe late 1980s through the early 1990s were a

heyday for literature on Japanese firms. TheJapanese economys international breakthrough,notably in an auto industry that had previouslybeen dominated by the US Big Three, gave manyexperts good cause to try and discover the originsof this competitiveness. Although the States strongrole (MITI ) has been underlined, it was clearly inthe organisation of the firm, both on an internallevel and also in regards to its external (notablysourcing) relations that these experts tried toidentify the specificities of the Japanese firm.Beyond culturalist interpretations stressing the

idiosyncratic nature of any mode of organisationthat is necessarily embedded in a national system(and therefore a product of history), several studieshave proposed analytical interpretations of firms

specific organisational forms. It is mainly out ofthe opposition between the A (H) and the J firmthat Masahiko Aoki built an analytical gridemphasizing the necessary interconnection between(and systemic coherence of) the different attributesof a Japanese firm (Aoki, 1988, 1990, 1995). This

opposition is based on the firms informationalstructure (vertical in the case of the A firm andhorizontal for the J firm) that determines its abilityto react to fluctuations in their environments (Aoki,1986).

The J firms superiority in a turbulentenvironment (with the A firm being better in astable regime or in a chaotic situation) infersflexibility and responsiveness derived fromcollective learning across all hierarchical levels, aswell as an incentives system that is based on ahierarchy of ranks (quasi-lifetime employment,personalised career progression depending on onesinvolvement and efficiency). Bolstered by thepresence of a large of bank within the keiretsu,industrial firms in this configuration find anattentive partner that provides regular monitoringand a posteriori controls, and which supports long-term investments (i.e., there are no short-termconstraints such as the creation of shareholdervalue). Shareholders and employees both superviseexecutives who mediate the firms two essentialcomponents (being the actors who provide funding

and those who offer their labour). The attributes ofthe A (American) or the H (hierarchical) firm arediametrically opposed to this configuration,although they too are coherent (Table 1).

Table 1. The opposition between the A and the J firm

Attributes The J Firm The A (H) Firm

Informational structure Horizontal Vertical

Employment relationships

(incentive)Hierarchy of ranks Hierarchy of functions

Finance Main bank system Financial market

Corporate governance Dual control Shareholders control

Source: based on Aoki, 1995

Largely focused on the A/J opposition whose mainpolitical contours are straightforward, MasahikoAoki has integrated other types of organisation likethe G (German) firm, which differs from the two

aforementioned ones in terms of the relationships

between firms three main actors: its executives(E); employees (W); and shareholders (S) seeFigure 2.


7/35

Figure 2. Characterisation of corporate forms based on the interactions between a firms three partners

S S W

E W S E W E

A firm J firm G firm

(Executives placed under

shareholder control)

(Executives as mediators between

shareholders and employees)

(Executives under the double

control of shareholders and

employees)

Source: Aoki

By introducing other elements like the role ofthe State, we could easily extend this analyticalgrid by choosing one firm model for each systemnational. Some confusion arises between the microand macro analytical levels, whose linkage couldbe a problem. For example, in the case of Japan(Boyer, Yamada, 2000) no consideration is givento the dualism taken into consideration the fabric ofSMEs marked by the uncertain nature ofemployees employment status, nor with a servicesector characterized by very low levels ofproductivity this being the necessary counterpartof a configuration in which large companies offer

quasi-lifetime employment. A macro/microisomorphism may be able to conceive of a varietyof organisational forms, but apparently only byrestricting the breadth of this depiction.

The Variety of Capitalisms School (VOC)The confusion between various levels can also

be found in studies of what Robert Boyer hascalled (2004a) the Variety of Capitalisms (VOC)corpus. This refers to all of the analyses that havetried to build upon Michel Alberts work, with its

opposition of two forms of capitalism, bysolidifying the constructs analytic foundations(Amable, 2005). Including notions like theinterplay between collective modes ofcoordination, notably involving interactionsbetween reticular forms (networks) and the impactof social norms (Hollingsworth, Boyer, 1997), thefocus here is on capitalisms diversity (Crounch,Streeck 1996), apprehended through the prism of a

configuration defined by different institutionalarrangements. Once again we find the oppositionbetween Liberal or uncoordinated MarketEconomies (LME) and Coordinated / organisedMarket Economies (CME).

One interesting version of this approach isfound in studies by Peter Hall and David Soskice(2001). The focus here is on firms (figure 3) whosebehaviour results from the coordinated action ofseveral components:9 The structuring of the financial system: short-

term oriented but risk-friendly in LME due tothe involvement of the financial market, versus

long-term oriented but risk-averse in CMEsince this is based on a stable relationshipbetween banks and industry ;

9 Industrial relations marked by a deregulatedlabour market in LME and by cooperativepractices supported by labour unions in CME;

9 An educational system that stresses generaleducation (LME) or else intensive initialvocational training (CME);

9 Inter-firm relations: very strong competition inLME, versus CMEs strong propensity to

develop cooperative arrangements (notablytechnological ones) inside of networks.

We can see right away that this Variety ofCapitalisms approach is based on a total confusionbetween different levels of analysis, and that itdoes not enable thinking about diversity in areaslike organisation (or even behaviours) within oneand the same type of capitalism.


8/35

Figure 3. VOC: A firm-centred approach

Firm Organization

(Management structure)

(shopfloor organization)

Financial System

Firm

Behavior

Linkages With Other Firms

(clients/suppliers)

(competitors)

Structure

of the

State

and the

Political

System

Industrial Relations System

(organization of unions)

(collective bargaining rules)

Training System

(high/low skills equilibrium)

Source:Hall, Soskice, 2001

Meso-micro isomorphism: the intersectorialvariety of forms of organisation

Certain studies consider the diversity ofproductive configurations within one and the samemacro-institutional environment, with attentionbeing diverted from the national echelon towardsthe meso-level where we find the industry. Byviewing this level as something determinant in thedevelopment of institutional arrangements,organisational models shaped at the level of thefirm reflect the impact of conventions or normsdefined at a sectorial level.

Spheres of production

Conventions Theory (Orlan, 1994; Salais et al.,2000) is what substantiates this kind of analyticalframework. Robert Salais and Michael Storpersspheres of production construct (1993) based ontheir analysis of product identity and qualityconventions between suppliers and customersconcerning the outcome of a production activity.The authors cross two criteria: the nature of theresources being used in production (standard orspecialised), which helps them to apprehend theinteractions between economies of scale and scope;

and whether the demand is certain or uncertain(generic or dedicated products). They distinguishbetween four spheres of production (Table 2):9 The industrial sphere represents the mass

production of standardised goods;

9 The market-oriented (flexible) sphere alsoapplies to standardised products, but includes a

differentiation by type of client;9 The interpersonal (professional) spheremobilises specific types of knowledge so thatproduction can target specific clients;

9 The immaterial sphere is the one where newproducts addressing a generic demand arecreated

Each of these spheres of production ischaracterised by conventions, notably regardingquality, operational logic (how uncertainty ishandled, forms of competition) or deeplydifferentiated registers of action.Different types

of firms can be accommodated here: in theindustrial sphere, large firms will exploit theadvantages of economies of scale, whereas in theinterpersonal sphere, firms will be smaller in sizeso that they can manage interrelationships andbolster the basic competencies driving theircompetitive advantage.

This sort of analytical grid enables reflectionsupon the different spheres of productionssynchronic diversity but precludes anyunderstanding of diversity within one and the same

sphere of production (Coris, 2004). Each sectorrefers to one of these spheres and the typologyproposed is not unrelated to the one established byKeith Pavitt (1986) to apprehend innovationdynamics in varying sectors.


9/35

Systemic environment

Table 2 . Spheres of production la Salais and Storper

Specialised products Standard products

Dedicated products Interpersonal sphere Market-oriented sphere

Generic products Immaterial sphere Industrial sphere

Source: Salais, Storper, 1993

The approach allows us to apprehend inter-sectorial diversity (in different spheres ofproduction) whilst suggesting that each sphere ishomogeneous insofar as it is structured by apredominant quality convention.

Comparative institutional analysis

In more recent studies (Aoki, 2001, 2004)Masahiko Aoki has affirmed that institutionaldiversity is the rule.

He develops a systematic approach(comparative institutional analysis) of theinstitutional change specifying certain concepts

(notably relating to hierarchy and institutionalcomplementarities).Aoki can broaden its analytical framework

beyond the opposition between J and A firms,focusing on the role that informational structureplays in a fluctuating environment, but consideringtwo different units T1 and T2 (Figure 4).

Figure 4. The three generic modes of information circulation

Systemic environment Systemic environment

T1 T2 T1 T2 T1 T1

Hierarchical decomposition Information assimilation Inforamation encapsulation

Hierarchical decomposition and informationsharing correspond respectively to the structuringof vertical and horizontal information.Encapsulated information is a figure generating athird configuration, one corresponding to modulararchitecture in products and organisations. Thisinformational structure corresponds to a SV(Silicon Valley) model that supposedlycharacterizes ICT industries in which the search fornew combinations of modules and the explorationof new horizons translate into organisationaldynamics that do not fit into a binary A/Jframework.

What Masahiko Aoki has done is to expand theframework beyond the national characterisations offirm forms (A, G, J) and incorporate an infra-national diversity dimension the idea being that

although the SV model stems from specificCalifornian organisational/institutional innovations,it can be diffused in other contexts. Lastly, heincorporates supranational configurations byintegrating the global firm (GL) that functions likea set of hierarchised networks. Once again, it is inthe area of ICT and through the use of stylised factsabout certain US multinationals that MasahikoAoki has tried to find elements to explain theemergence of a new configuration.

If we stay within the confines of his twoadditional configurations1, what is clear is that the

1 To underscore the extent of the models diversity,he also included models that are directly inspired fromW (Walrasian) or HM (Grossmann, Hart and More)theory, raising serious questions about said models status.


10/35

new grid incorporates a diversity of firm modelswithin one and the same economy. For example,the US will have type A, SV and GL firms (andeven others). The analysis does infer, however, thatwithin each industry, there exists one firm formthat is efficient, meaning that different

arrangements correspond to specific sectoriallogics. The diversity of firms within one and thesame industry has yet to be satisfactorilyapprehended.

THE DOUBLE DIMENSION OF DIVERSITY:

ORGANISATIONAL AND INSTITUTIONAL

In sum, we should be trying to simultaneouslyapprehend diversitys double dimension, in regardsto socio-economic models but also with respect to

firms forms of organisation, without debasing thelatter by construing it as nothing more than theinterplay amongst different sectorial logics. This istantamount to following an approach la DouglassNorth, who distinguished between institutions andorganisations, with the former defining the rules ofthe game within which the latter operate (being theplayers whose actions are geared towards a specificfinality). Earlier approaches confused the twodiversities (with institutional diversity determiningits organisational counterpart in a univocal manner)

our goal is to understand how diversities that arebased on different foundations interact with oneanother.

The only tool that we can apply towards thisend is the one conceptual framework that sharesthis goal: the productive models analytical grid(Boyer, Freyssenet, 2000a). Note this grid iscurrently being worked on (Freyssenet, 2003) andis likely to evolve somewhat to enable it to connectup our three levels (micro, meso and macro).Introducing a meso level will help us to transcendthe current vision of institutions, and to account

more explicitly for the interactions underlying thedynamics of institutional change in thismicro/macro linkage.

Analysis of productive models diversity

As the outcome of a collective research projectconducted under the aegis of the GERPISAinternational networks 3rd Framework Programme(Training and Mobility Programme), whichquestioned the reality of a Japanese model ofmanagement that was being promoted at the time

as a universal standard1, the Emergence of newindustrial models programme (coordinated byRobert Boyer and Michel Freyssenet) paved theway to the development of an analytical grid thatencourages thinking about the diversity of firmsforms of organisation within any given sector - in

our case the car industry, which was theemblematic sector of Japanese internationalcompetitiveness in the late 1980s.

Several GERPISA researchers have tried toapply this grid to other sectors. Their efforts haverevealed the limitations of its initial version, andthe need to ascertain issues for debate that shouldbe included in a new one.

Presentation of the analytical gridThe grammar underlying the GERPISAanalytical grid

The analytical grid that Robert Boyer andMichel Freyssenet have proposed tries to identify afirms profit strategies so as to apprehend itsconditions of performance. These can be twofold innature, involving, on one hand, the strategysrelevance to the socio-economic environment inwhich the firm operates, and on the other, theprocess driving the achievement of internalcoherence between the varying dimensions of thefirms organisation, based on what the two authorshave called a government compromise. Profit

strategies are central to this vision, since they helpto determine a specification each of the firms inquestion, and to interconnect their external(relevance) and internal (coherence) levels Figure5.

Profit strategies correspond to a priority searchfor one or several profit sources amongst sixfundamental sources thereof: volume; quality;diversity; flexibility; innovation; and permanentreduction of costs. The efficient implementation ofone or several of these sources, hence the creationof a modicum of compatibility amongst varyingprofit sources, is a crucial element in the buildingof a productive model. To be efficient, a strategymust first be implemented profitably in a givensocio-economic context. For Robert Boyer andMichel Freyssenet, a strategys external relevancebasically depends on its ability to cope with two ofthe fundamental uncertainties that firms face:market uncertainty, i.e., whether the firm can find ademand for what it offers; and labour uncertainty,expressing its ability to motivate employees toimplement its productive project.

1 Before the 1990s revealed the weaknesses besettinga host of Japanese firms, and their limitations at amacro-economic level.


11/35

These uncertainties are determined by the modeof growth and national income distribution thatcharacterizes the economy within which the firmoperates mainly, with this mode of growth itselfbeing a function of the international regime, i.e.,the international inclusion modalities of the

economy in question. If we ignore for the momentthe case of so-called rent and shortage modesthat correspond respectively to the situations incertain developing or former socialist countries,industrialised economies can be characterizedbased on two main criteria:9 The more or less egalitarian nature of their

income distribution, which is a de factoreference to Hall and Soskices (2001) VOCopposition between Coordinated MarketEconomies (CME) and Liberal MarketEconomies (LME);

9 The growth driver: internal demand versusexports (price or non-price competitiveness).

We can see that a profit strategy will beinefficient if it is not in sync with its macro-economic context. Correspondence (externalrelevance) is a necessary but insufficient condition,

however. The firm must also be capable ofimplementing its relevant strategy. This harks backto issues relating to the internal coherence creationprocess.

The authors apprehend three structuringelements in regards to this second condition:

9 The product policy, explaining the profitstrategy the firm has adopted;9 The productive organisation (management

of design, sourcing, production methods,distribution, etc.) that the firm hasdeveloped to implement its strategy;

9 The employment relationship negotiatedwith employees and their organisations.

Attempts to achieve complementarity betweenthese three components are rooted in theestablishment of a company governmentcompromise between the various stakeholders:shareholders, employees, creditors, bankers,distributors, suppliers, subcontractors, etc. Theestablishment of this compromise is the mostcomplex element driving a productive model.

Figure 5. - The productive model and its environment

International regime

Mode of growth andnational income

distribution

Market Labour

Profitstrategy

Productpolicy

Companygovernmentcompromise

Productiveorganisation

Employmentrelationship

Source: Boyer, Freyssenet (2000)


12/35

Profit strategy Productive model Automobile group

Volume and diversity Sloanian Volkswagen

Permanent reduction of costs Toyotian Toyota

Innovation and flexibility Hondian Honda

Applying the model in sectors other than thecar industry

This analytical grid was designed as part of acomparative analysis undertaken by GERPISAinternational network members focused on thetrajectories that the worlds carmakers had pursuedover 1973-1993 (Freyssenet, Mair, Shimizu,Volpato, 1998).

It was extended via an exploration of the historyof the auto industry across the whole of the 20thcentury (Boyer, Freyssenet, 2002). Robert Boyerand Michel Freyssenet have deduced from theirstudies that in the recent period only threeproductive models corresponding to different profit

strategies can be identified (Table 3).

Table 3 Automobile group productive models over the period 1973-93

Source: Boyer, Freyssenet, 2000

In a study delving further into the history of thecar industry, Robert Boyer and Michel Freyssenetidentified three other models: Fordian, Taylorianand Woollardian. The former was based on aclearly identifiable volume strategy, notably atFord Motor Company in the 1910s - whereas thetwo others are associated with the diversity andflexibility strategy that dominated the British autoindustry between WWI and WWII.

The three main conclusions of these studiesmerit further emphasis:9 Initially, attempts to achieve some form of

coherence remained the exception, with crisesconstituting the rule. This is because only threeout of the 16 firms under study remainedprofitable over the long run. Now, this modelcan only exist if its two profitability conditionsare satisfied. And yet, the firms that foundthemselves in difficulty did not disappear.

What this means is that the grid does not offerthe kind of functionalist reading that willenable a selection of efficient models: it is theprocess of seeking coherence, grounded in trialand error and experimentation and focused oninstitutional complementarity, which is centralin this analysis.

9 Secondly, the Japanese model never existed(Freyssenet, 2001). On one hand, manyJapanese carmakers (starting with Nissan,which adopted a volume and diversitystrategy) ended up displaying a great deal of

fragility over the course of the 1990s, at a timewhen they were being forced to arrangealliances. Note also the problems faced byMitsubishi and Mazda. As for the countrys

two success stories, Honda and Toyota, closerexamination reveals major differences andincompatibilities between them, concerningtheir strategies but also their efforts to achievea semblance of internal coherence.

9 Thirdly, just like the Japanese firm, theAmerican firm (Regulation TheorysFordism variant) has no intrinsic meaning:Taylorism, Fordism and Sloanism correspondto different models in macro-economicenvironments that ultimately become quitesimilar, especially when their governmentcompromises are based on other foundations.

This grid underlines the limitations of aRegulation Theory approach to the firm, and moregenerally of historical overviews that tend to ignoresynchronic diversity and highlight a readinghighlighting changes in those forms that aredeemed to be historically dominant (see above).

Nevertheless, it does evoke serious points ofdebate, especially when one tries to apply it insectors other than the car industry.

Issues for debate

Profit sources and strategiesProfit sources and strategies comprise an initial

set of debate topics. Given the financialisation bywhich many economies have been characterized inrecent years, we can already ask whether financeper se constitutes a source of profit. Still using thecar industry as an example, more than half of allprofits made in the 1990s by the two US carmakers(Ford and GM) came from their financialsubsidiaries (Froud et al., 2002). The Europeanswere also part of this trend, with captive finance


13/35

representing 20% of Renault and PSA PeugeotCitrons profits in 2004. In a sense, GeneralElectric (GE) has been viewed as an efficientbusiness model (Froud, et alii, 2005), inspiringother firms like the Italian Fiat. This is not a newelement in the sense that financial activity has long

been part (although probably to a less extent in the1970s and 1980s) of groups profitability, notablyin the auto sector - another reason to include thisdimension in firms generic profit sources.

Profit sources are generally primarily defined interms of a manufacturing activity, and relationalaspects are ignored. Yet this latter dimension iscrucial for service activities where resources areco-produced within the framework of an interactionbetween the supplier and the customer (du Tertre,2002). Still within the confines of the auto industry,we can see that carmakers have focused largely on

service activities (maintenance, repairs, insurance,rentals, etc.) in their attempts to offset lower profitmargins on new car sales, a consequence of thefierce competition in this market. But arent suchsources of profitability extraneous to the analyticalalphabet we are being offered? The gridstransposition to service activities raises a numberof problems (Lomba, 2002; Coris, 2005).

This problems we have in integrating thoseprofit sources that can occur within a frameworkdefined by the relationship between a firm and itsclients, suppliers (i.e., the quasi-partner rent, c.f.,Asanuma, 1989) or competitors (leadership, inBeffas sense of the term, 2003) indicates that theanalytical alphabet orienting these profit sources isincomplete. This limits the grids potentialapplicability to certain activities. Furthermore, oneof these profit sources is idiosyncratic, to wit thepermanent reduction of costs, defined in anexplicit reference to Toyotas strategy of constantlyseeking to squeeze costs whilst keeping volumesconstant, irrespective of the environment orsituation. There are questions about this profit

source/strategys degree of generality orreproducibility in other sectors of activity. Onepossible avenue would be to opt for anothercharacterisation, fluidity, volume and diversity,an approach viewing fluidity as a source of profit,similar to way in which leadership is treated inJean-Louis Beffas formulation. These are threeadditional profit sources, and they are at odds withthe foundations underlying the models grammar.

Dynamics of the company governmentcompromise

A second point of discussion is the companygovernment compromise. Clearly, the idea ofdecomposing the coherence creation dynamics into

three components appears to be based on solidfoundations: the relationship to the product market(with the product policy acting to reveal the profitstrategy) infers a certain type of productiveorganisation, one that is in phase with a specifiedemployment relationship. This approach would

appear to be transposable, without any majorobstacles, to other industrial sectors, where it couldbe fertile.

Nevertheless, the notion does require finetuning given that the dynamics of the compromisehave yet to be specified (Pardi, 2004a). Forexample, who are its actors? Should a dominantposition be allocated to the employmentrelationship, i.e., the relationship between the firmand its employees? What about finance andrelations with shareholders, be they banks or stockmarket investors? How should this compromise be

established, and how should its stability be definedor its development studied? At what level should itbe apprehended? At first glance, the productivemodels grid appears to stress an approach thathighlights the domestic level, in the firmseconomy of origin. Yet due to productiveinternationalisation, firms are now being asked totransfer their domestic model to transplantfactories, where they are supposed to hybridise it(Boyer, Charron, Jrgens, Tolliday, 1998).

What this infers is the adaptation of product

policy and productive organisation to localcontexts, and the negotiation of specificemployment relationships. Hence the establishmentof a new government compromise at the level ofthe transplant unit (Pardi, 2004b). At an extreme,this could lead to a reconfiguration of the originalmodel, something that at the very least indicatesthe need to set up a number of micro-compromises,and to account for intra-firm diversity.

One of the recent internationalisation phasesother characteristics is the rise of alliances andmergers/acquisitions (Volpato, 2004). The merger

between Daimler and Chryslers, Jaguar andVolvos takeover by Ford or the alliance betweenNissan and of Renault are all examples ofrapprochements between firms with radicallydifferent, and even incompatible, profit strategies(Boyer, Freyssenet, 2000b). Can a diversity of thiskind last?

Given that firms have become veritablemulti/transnationals in many sectors, it is hard tolink the many different spatial levels at which aproductive models internal coherence can be

apprehended, especially since firms interna-tionalisation strategies, in the auto industry at least(Freyssenet, Lung, 2003) are more closely tied to


14/35

regionalisation phenomena than they are toglobalisation. This dimension is just as portentouswhen the aim is to evaluate the external conditionsthat determine a profit strategys degree ofrelevance.

External conditions of relevancePart of the discussion at this level revolves

around two basic uncertainties (even though otherdimensions can be envisaged), notably financialuncertainty and technological uncertainty. Thestrength of the argument about these twouncertainties derives from its Marxian origins,which basically characterize the capitalist mode ofproduction as a market-oriented economy (thesocial validation of private labour to use Marxsterms) and as an economy where the labour force

itself constitutes a commodity. Staying within thesame Marxist-inspired theoretical framework, thisapproach neglects the relationships betweenvarious forms of capital (notably between financialand productive capital), links that are arrangedagainst a backdrop of financial uncertainty (seeBoyer, 2004; Montalban, 2005a). Once again, thefinancial dimension seems to have been poorlyintegrated into the current version of the productivemodels analytical grid. Nor does it account for thedevelopment of productive forces that have

supposedly led to a shift into a new phase ofcognitive capitalism (Azais, Corsani, Dieuaide,2003), thus translating the context of a regime ofpermanent innovation (Hatchuel, Le Masson, Weil,2002)1.

Above and beyond this one dimension, what isat stake is our use of growth and incomedistribution modes to understand an externalenvironment. Transpositions of this grid into othersectors, like steel (Lomba, 2005), aeronautics(Frigant, Talbot, 2005), IT (Coris, 2005) orpharmaceuticals (Montalban, 2005a) have

highlighted the difficulty of using it as a gateway,notably to study issues of relevance within aparticular industry. Although the grid seemssuitable for industries that produce goods targetingend users, or for so-called B-2-C (business-to-consumer) industries like automobiles oragribusiness (Jullien, 2004b), problems arise whenit is applied to other activities, notably industriesthat produce capital or intermediary goods, or elsecorporate services (B-2-B, business-to-business). It

changes in these other industries and their modes ofgrowth (internal demand vs. exports), and evenmore difficult to tie this to income distributionmodes (whether coordinated or liberal). Thiscategory also neglects a few dimensions that areimportant in helping us to understand operative

dynamics and strategies, notably where Stateintervention is involved (Ramirez, 2004a, 2004b),whether this covers intellectual property rightsregime (Coris, 2005), education or public demand,especially military spending.

In a recent contribution, Michel Freyssenet(2005) outlined a research programme aimed atextending the GERPISA approach by applying toeconomies certain analytical tools that werepreviously used to study firms. This has involvedreconstituting the main economies historicaltrajectories to devise a conceptualisation capable of

accounting for the distinction between a strategyand a model. The idea here is that a countrys

growth strategy is based on one of four sources:consumption; investment; exports; and predation. Itis also said to be implemented through growthmodels that are the fruit of socio-politicalcompromises. Figure 6 replaces the upper sectionof the productive models grid (Figure 5),highlighting the relevance of a firms profitstrategy to the growth model of the nationaleconomy in which it finds itself.

The method proposed (described as GERPISA2) has some real strengths, notably insofar as itsapproach is both historical and relativelyexhaustive in nature. It raises real questions aboutthe sources of growth and compromises that areestablished within this framework. However, itseems hard to justify advancing by means of ananalogy between the firm level (companygovernment compromise to implement a profitstrategy) and the national level (socio-politicalcompromise based on a mode of growth). Whereasthe GERPISA 1 approach passed over certain

institutional aspects that were meaningful in sectorsother than the auto industry, the new version seemsto cover all fields, without ranking the variousdomains.

In the absence of a historical approach that willallow us to reconstitute the different socio-economic models emergence2, we will limitourselves to an analysis of the synchronic varietyof the contemporary forms of capitalism that createthe environment in which firms operate, by

is

hard

to

d

evise

adirect

link

between

m

arket

2 There is no doubt that this is a very interesting1 As such, this is an approach that stresses analyses

found in Book 1 of Marxs Das Kapital (1867) whilstneglecting the forms of competition studied in Book III.

approach. It could constitute an extension of the presentproject since it requires the mobilisation of a wide arrayof researchers.


15/35

National growth.Combination ofsocial relationships

applying an approach delineated in terms of socialsystems of innovation and production (Amable,

Barr, Boyer, 1997; Amable, 2003) or of socio-economic models (Amable, 2005).

Figure 6 . - Strategy and model of growth

Internationalagreements

Internal conditions External conditions

Growthstrategy

ProductionTypes of goods and

servicesType of productivity

Productive fabric

Type of R&D

DistributionPopulation affectedForm of distributionPurchasing powerSocial protection

Socio-politicalcompromise

Market

Regulation Access tointernal marketCurrency-Credit

CompetitionIndustrialrelationsTraining Labour

Source : Freyssenet, 2005

The search for a micro/meso/macro link

It is within this framework that we will indicatea few avenues for an approach that could help tolink up our three levels of analysis the firm, theindustry and the macro-economic level (nationand/or region). These are avenues that the ESEMKproject will be exploring over next few months. Webeing by discussing the constraints that theinstitutional environment characterizing thedifferent socio-economic models imposes uponfirms strategies. This involves crossing twoanalytical grids: the productive models grid at amicroeconomic level; and the SSIP grid at a macro-economic level. The explicit introduction of anintermediary level (i.e., the sector) raises other


16/35

types of question that can be studied later. Lastly,consideration will be given to the reciprocalinteractions that go from the micro to the macrolevels, the purpose being to envisage institutionalchanges endogenous dynamics.

Micro/macro compatibility

Crossing the analytical gridsThe objective at this stage is to amend theproductive models grid and replace a mode ofgrowth and income distribution approach (called agrowth model in its latest formulation) with aSSIP approach so as to apprehend the productivemodels external relevance as well as their creationof internal coherence. Here we discuss both:


17/35

SSIPProduct market

competitionFinancial sector

Wage-labournexus

Socialprotection

EducationProductivemodel

Profit strategy

Product policy

Productiveorganisation

Employmentrelationships

CompanyGovernmentCompromise

9 The compatibility between a firms profitstrategy and its macro-institutionalenvironment; and

9 The possibility of creating a semblance ofcoherence between all of the componentscomprising a productive model, whilstnegotiating a government compromiseenabling the strategys implementation, giventhe institutional constraints that weigh uponactors.

It is hard to conceive of a quality strategy that isnot based on an efficient vocational trainingsystem, and that lacks a wage-labour nexus inwhich labour unions have acquired a strongbargaining position.

Inversely, where professional unions are weakand levels of education minimal, what we find areconditions conducive to the diffusion of a Fordianmodel.

As such, what we should be scrutinising is howthe institutional constraints that are specific to any

given form of capitalism will influence each of ourproductive models. In Bruno Amables analysis(2003), five institutional domains are used todescribe SSIP: competition in the product market;the wage-labour nexus; the financial sector; socialprotection; and the educational system. Thesedomains intervene to a greater or lesser extent todelineate the possibilities available to the variouscomponents of the productive model (cf. Table 4).

Table 4. How institutional constraints weigh upon productive models components

* Each squares degree of greyness represents the institutional domains constrictive strength

Clearly, forms of competition in the product

market have a strong influence on firms profitstrategies and product policies. They also weighupon the productive organisation and theemployment relationship, hence on the governmentcompromise. Indeed, strong competitive rivalriesare associated with market segmentation, due to theinterplay between product differentiation and thesearch for innovation in firms attempts to rebuildprofit niches. This infers an adoption oforganisational forms enabling flexibility oradaptability. Competitive pressures also weighupon the employment relationships by limiting the

room for negotiations, and this affects the searchfor a company government compromise. Onemanifest illustration is the recent phase in theeconomies internationalisation, which has

translated into an opening of markets, strong

competitive pressures (Conway et al., 2005) andfinancial globalisation (Chesnais, 2004). Thesedevelopments have to a large extent weakenedemployees bargaining position and strengthenedshareholders.

The company government compromise feels thebrunt of the major constraints associated with thestructuring of a financial system. It is easy to seehow in a financial market-drive economy,shareholders behaviour and influence onexecutives are not the same as in a configurationwhere stable managerial power is guaranteed bythe alliance between banks and industry. Thisinfluence of the financial sphere weighs upon all ofa firms components, thus on the employmentrelationship due to the balance of power created


18/35

between shareholders and employees. It also has anindirect effect on the productive organisation (i.e.,pressure exerted by institutional investors to getfirms to refocus on their core business andtherefore to outsource activities considered non-essential, or to spin them off if the firm has several

core businesses). Lastly, it weighs upon the productpolicy since financial norms of this ilk impactfirms strategic choices, notably their search fornew profit sources in areas like finance or services.All is which is tantamount to saying that therelevance of a profit strategy is largely determinedby the characteristics of this institutional domain.

The same applies to the different dimensions ofthe wage-labour nexus, which refers explicitly toone of the two fundamental uncertainties, helpingus to apprehend the external relevance of aparticular profit strategy (see above) and

constituting a major component in theestablishment of a company governmentcompromise. This institutional domain exerts itsstrongest influence on the employment relationshipinasmuch as it is here that the balance of power and(meta-) rules of negotiation (irrespective of thenegotiations degree of decentralisation, which canvary) are defined at a national level and find aspecific breakdown within the firm, or even theunit. Nor should one forget its impact on productpolicy (particularly issues relating to theemployees demand) and on the productiveorganisation, notably via a labour flexibility thatstrongly affects organisational forms.

The two other institutional domains have less ofan impact. The forms of social protection can playa major role in the employment relationship,depending on whether such protection is firstconsolidated at the level of the whole nationaleconomy, or else internalised within the firm, and ithas an impact on the market in B-2-C industries(indirect income stabilises the households demand).Lastly, questions about the educational system play

a not insignificant role in the productiveorganisations definition and employmentrelationships negotiation.

Automotive industry lessons for the productivemodels

The car industry is the only sector whereGERPISAs analytical grid was more or lessimplemented in its entirety. Returning to RobertBoyer and Michel Freyssenets findings, we can tryto see how the environment in which firms operate,apprehended in SSIP terms, takes the interactionsbetween certain institutional constraints and tries toenable (or inversely, to restrict) profit strategiesand the productive models that are associated with

them1. Starting out with a reasoning grounded inthe different elements found in the previous table,we come up with the following findings (cf. Table5).

Social Democratic capitalism would appear atfirst glance to create a context that is conducive to

quality and volume and diversity strategies,what with the existence of a qualified workforce,the strong role devolved to unions and, inversely,shareholders long-term stability (not to forget theimportance of indirect wages that level outdisparities in remuneration and stabilise incomes).At the same time, these very same characteristicsmean that this form is not very conducive tostrategies based on strong market polarisation(diversity and flexibility, innovation andflexibility and volume) despite significantworkforce flexibility, inasmuch as this infers afragmentation of demand and/or a focus on basicmodels or on niches geared towards extreme socialcategories. The strong pressures on workorganisation that are found in the volume andpermanent reduction of costs strategies (a factorthat fluidifies processes and creates greater controlover the work organisation), seem scarcelycompatible with the autonomy that employeeswould otherwise enjoy. Encouraged by NorthernEuropean economies degree of internationalopenness, the Swedish carmakers Volvo and Saab

have based their strategies on quality - eventhough the increased importance of economies ofscale helped Ford and GM, respectively, to grabthese two brands. Despite the importance ofexports (70% of total output), small economiesreduced size seems to be a major obstacle to anystrategy including volume as a source of profit.

1 Theoretically, it is possible to conceive of differentproductive models that implement one and the sameprofit strategy. However, in the case of automotiveindustry history, the only example thereof has been thediversity and flexibility strategy (Boyer, Freyssenet,2000).


19/35


Table 5. Crossing SSIPs and Productive Models

SSIP

Profit strategy /Productive model

Social LiberalDemocratic Marketcapitalism Capitalism

Mediterr- Continental Meso-anean European corporatist

capitalism capitalism Capitalism

Volume / Fordism + + +Diversity and flexibility /Taylorism

+ +

Volume and variety /Sloanism

+ + +

Innovation and flexibility /Hondaism

+/ + + +

Permanent reduction ofcosts / Toyotaism

+ +

Quality ** + + +

* the markers + / correspond to a favourable / unfavourable institutional environment

** Quality is a profit strategy, without a model having been formally identified by the authors, due to thelack of any in-depth studies on top-of-the-range European carmakers.

The opposite applies to Liberal Marketcapitalism, which is conducive to all strategiesbased on volume (volume, volume anddiversity) but also on diversity and flexibility,given labour market fluidity and competitivepressures that lead to a differentiation which causesmarket segmentation in turn. Reciprocally,permanent reduction of costs and quality

strategies inferring a modicum of long-termstability in employment (so that people can benefitfrom learning effects) appear inappropriate. Theinnovation and flexibility strategy provides anambiguous example. Although a very segmentedand flexible market seems to lend relevance to thisstrategy, questions can be raised about itsimplementation since it assumes a financialautonomy that is difficult to demand in a financialmarket system. Chryslers problems illustrate thisambiguity, with the firm having been incapable of

developing a government compromise allowing itto implement such a strategy durably, culminatingin its takeover by the German, Daimler. Thevolume and diversity strategys two signaturefirms (Ford and General Motors) are also the oneswho invested heavily in captive finance, addingthis as a new source of profit. Rovers annihilationin England sanctions the abandonment of itsdiversity and flexibility strategy, translating theBritish carmakers inability to stabilise itsgovernment compromise.

Because of its financial structure,Mediterranean capitalism does not suffer fromthis shortcoming. As a result, it is free to pursueinnovation and flexibility or diversity and

flexibility strategies. The latter seems moresuitable than volume and diversity, since themarket here is relatively polarised between twoextremes, creating room for a volume strategygeared towards entry-level vehicles. Neither apermanent reduction of costs nor qualitystrategy would appear to be in sync with thecharacteristics of the wage-labour nexus and

educational systems typifying this construct although they do seem to be present, at least to acertain extent, in Italy, where one extremecombines small series, top-of-the-range specialistslike Ferrari and Maserati with engineering firms(thereby illustrating the automobile marketspropensity towards fragmentation), whilst the otherbrings Fiat (unable to abandon its seminal Fordianmodel to transmogrify into a Sloanian model multi-brand group) together, at an intermediary level,with Alfa Romeo, which was incapable of

consolidating its quality strategy positioning. Theremarkable breakthrough of Spain, which in threedecades became Europes third largest carproducing country, is largely (but not entirely, c.f.,Layan, 2004) a volume-based phenomenon, aswitnessed by its specialisation in entry-level carsand small vans (light commercial vehicles).

Continental European capitalisms profile is theopposite of Liberal Market capitalism. The marketand wage-labour nexus characteristics found heredo not create conditions conducive to a volumestrategy (except in the context of an initialequipment market) or to diversity and flexibility.The relatively egalitarian distribution and largelysocialised nature of incomes in this configuration


20/35


means that the volume and variety strategy diversity and flexibility strategy is not verypredominates, with all European carmakers beingone way or the other present in the markets fourmain segments (lower range, midrange, uppermidrange, top-of-the-range). With theseeconomies openness (a reflection of the marketsregional integration), abundant middle classes andhighly skilled workforce, conditions are rife for thedevelopment of top-of-the-range and nichevehicles, and this enhances the relevance ofquality and innovation and flexibilitystrategies. Moreover, institutional shareholderswield less influence, something that encourages theestablishment of compromises conducive to thelatter strategys implementation. Finally, thesekinds of compromises can ostensibly be negotiatedat a local level so that a permanent reduction ofcosts can be obtained. With its numerous brands

(Audi, Seat, Skoda, Volkswagen, Bentley andLamborghini) and platform strategy, VW is the flagbearer of the Sloanian model (Boyer, Freyssenet,2000a), followed by PSA Peugeot-Citron and notforgetting Ford and GMs European subsidiaries,which have developed Sloanian strategies on aglobal scale (Bordenave, Lung, 2003). Brands likeBMW and Mercedes, and even Porsche, aresynonymous with German quality and dominate themarket for top-of-the-range cars, even if recentchanges (downscaling of BMW Series 3 and 1 orMercedes Classes A and B, DaimlerChryslermerger, small car makes like Mini and Smart)indicate a shift towards a volume and varietystrategy. Lastly, with the monoboxes Renault haslaunched (Space, Twingo, and Scenic) it hasbecome one of the worlds more innovativecarmakers a strategy enabled by the stablecompromise it established during the 1990s.

As for meso-corporatist capitalism1 this ischaracterized by a highly decentralisedmanagement of the wage-labour nexus in largefirms. As a result, there is room for a number of

negotiated configurations: volume where aninitial equipment market is involved, but alsoquality at the top-of-the-range or else apermanent reduction of costs when the whole ofthe product range is in question. Strong job securityand limited external flexibility means that the

1 Called Asian capitalism in Bruno Amableswritings. Meso-corporatist (as it is sometimes called)seems preferable to avoir any confusion: he term Asiancapitalism is ambiguous as we can identidy very

different types of socio-economic models in Asia (India,China being different from Japan and South Koreawhich are the main industrialised countries for thisconfiguration).. But this would be a ASEMK project

operative here - unlike the innovation andflexibility strategy, which in order to occupyprofitable niches in an increasingly segmentedautomobile market mobilises internal flexibilityresources. This extreme segmentation makes thevolume and diversity strategy inefficient due tothe fact that consumers expect specifications thatdo not mesh with a commonalisation system. InJapan, for example, asides from Honda and Toyota(the two firms to have founded an originalproductive model), the most notable events areNissans problems in implementing a volume anddiversity strategy similar to the ones Koreancarmakers faced when trying to abandon avolume strategy in favour of volume anddiversity.

Although grid crossing of this kind seems to

lead to interesting results at an analytical level, twoproblems do remain. On one hand, there arequestions about the robustness of this method andstability of its outcomes, as it is not at all certainthat the same reasoning, when applied by otheractors, would lead to an identical interpretation,given the residual areas of ambiguity in theinteractions between institutional constraints(although such grey areas may well be a breedingground for the emergence of new configurations[see above]). On the other hand, the grids shouldalso be crossed in sectors other than the carindustry.

The meso-economic dimension

Sectorial approaches in the two analyticalgrids

Despite the lack of any firm-focused analysis ina SSIP-based approach, the construct is explicit inits incorporation of the sectorial level insofar as itdoes account for the way in which institutionalconstraints affect different industries dynamics(macro-meso determinism). De facto, this does notinvolve talking about industries per se but moreexactly about the way in which a socio-economicmodel impacts the technological innovationprocess by means of the interactions betweenScience, Technology and Industry (STI).

The STI system was central in the initial versionof the SSIP analytical grid (Amable, Barr, Boyer,1997), forming the core thereof (cf. Figure 7). Inthis grids second version (Amable, 2003),technology is not used to characterize the types ofcapitalism what is stressed is the way that

institutional arrangements work downstream tocreate or else to block opportunities fortechnological development, using economies


21/35


scientific competencies as a launchpad for theiractions. Despite being built on varyingmethodologies, the two grids do converge as theirfindings focus on the relationship between differenttypes of capitalism, on one hand, and scientific,technological and commercial specialisation, on theother (with each socio-economic model generatinga comparative institutional advantage). Forexample, Liberal Market capitalism is conducive toresearch in general biology, and to its knock-oneffects in biotechnology, because of the close tiesbetween the academic world and industry, theavailability of venture capital and the presence ofstock option-based remuneration systems. Markedby the strong presence of the State, ContinentalEuropean capitalism is geared more towards largetechnological programmes (hence Airbus andArianes performances in the aeronautics and space

sectors); Meso-corporatist capitalism is said to bemore predisposed towards electronics technologiesand mechanical industries, etc. Many studies(Pavitt, 1984; Malerba, Orsenigo, 1996, for asummary, see Bergouignan, 2005) have highlighted

innovation dynamics sectorial diversity. In theSSIP approach, it is very indirectly (via an analysisof institutional constraints effects on theinteractions between science, technology andindustry) that a particular form of capitalism is saidto create a specialisation covering a whole range ofactivities. Such an approach has two limitations: onone hand, the institutional dynamics that arespecific to each sector are excluded from analysis,here due to the univocal determinism of the macrotowards the meso level; on the other, intra-industrial heterogeneity is neglected since theanalysis ignores firms organisation. One couldimagine, at least theoretically, specialisations thatconverge in identical sectors whilst being part ofproductive models that diverge from one economyto the next, or from one specialisation to the next inthe same country, thus translating the

predominance of a given type of model (cf. insert).Hence the approach needs to be able tosimultaneously reflect upon diverse forms ofcapitalism, sectors and productive models.

Figure 7. Social Systems of Innovation and Production (SSIP1)

Education

Training

Social system of

innovation Economic

Science

Technology /Innovation

Humanresources

IndustryFinancialsystem

Innovation and Production System

Source : Amable, Barr, Boyer, 1997, p.127

An approach expressed in productive modelterms will view a sector as a space of competitiverivalry (in a manner redolent of Porter, 1980).When apprehending a profit strategy, emphasis willbe placed on the goods market, and discussionsabout the relevance of the strategy and thecompany government compromise will focus


22/35

129Actes du GERPISA n38almost exclusively on carmakers. Yet the carindustrys performance is not limited to theseactors alone, even if their position is a strategicone. The whole of a filire, value chain orautomotive system must be apprehended toconsider the systemic nature of competitiveness,what has been developed in the CoCKEAS project


23/35

Sa

MP2MP3

C

SSIP II

(Lung, 2003). By ignoring vertical relations,notably sourcing relations, a productive modelapproach will apprehend an industry as all of thefirms that are in direct competition in a market, towit, the car market.

These sectorial norms of production and

exchange are necessarily multiple since otherwisethe productive models would lack diversity. This isgenerated endogenously (endometabolism) sincethe competitive process itself creates conditionsfavourable to the emergence of alternative orcomplementary productive models (Boyer,Freyssenet, 2002). Because of the first moveradvantage, it is more efficient to adopt an originalprofit strategy rather than to imitate the insiderfirms one, as long as this new profit strategy isrelevant and the firm can ensure the strategyscoherence with its various components. It was to

transcend the Fordian model that had proved soefficient in a volume strategy that Alfred Sloandeveloped a new strategy on behalf of GeneralMotor, one that integrated diversity and appliedadapted organisational changes (multidivisionalstructures, management control, etc.). Theinnovation and flexibility strategy, and theHondian model in general, were the only ways forthe models eponymous carmaker to compete withthe firms (Nissan and Toyota) that already hadsolid positions in its domestic market. This also

explains its immediate orientation towards externalmarkets (exports).

Insert: Complex combinations of the threelevels: productive model, sector, SSIP

If we try to explain economies international

specialisation by the interconnected diversity of theinstitutional arrangements characterizing the threelevels of analysis, combining them will createcomplex configurations, at least at a theoreticallevel. Consider various industrial sectors S (a, b, c,) that can be an object of specialisation foreconomies A, B or C. The firms epitomising thiscompetitive advantage in a particular sector mighthave adopted diversified productive models MP1,MP2, MP3. The economies belong to differentsocio-economic models SSIP I, SSIP II, etc.

First exampleOne and the same international specialisation in

a sector (Sa) in three economies A, B and C cancorrespond to different productive models that arepart of different socio-economic environments -with different productive models, MP1 and MP2respectively, for economies A and B in one and thesame SSIP I type institutional environment, or elseanother productive model MP3 in another SSIP IItype institutional environment for economy C.

Sector

MP1Productive model

(firm)

A B Economy / Country

SSIP ISSIP / socio-

economic model

llustration: The car industrys international competitiveness in Japan (Hondian and Toyotian models) and Germany(Sloanian model and quality strategy) in the 1980s.

Second example

The same industrial specialisation (or Sb)

translates the same productive model MP1 indifferent institutional environments (SSIP I or SSIPII) or else (Sb) another productive model in another

institutional environment (economies B and C).Combining the institutional environment and firms

strategies (SSIP II and MP1 for economy B) willbenefit several industries (Sa and Sb).


24/35

MP3Productive

model

A

SSIP I

B

Sa Sb ScSector

MP1 MP1MP2

C Country

SSIP II SSIP IIISocio-economic

model

Illustrations: The volume and diversity strategy adopted by VW in Germany and Nissan in Japan in the firstinstance; but Japan and Koreas car and electronic specialisations cannot be found in Germany in the other.

Conversely, note that this approach totallyneglects any process that tries to specify anindustry in terms of other industries, since this wasnot its original objective. At the same time,attempts to transfer this approach from the carindustry to other sectors have underlined the needfor a better characterisation of such sectorsinstitutional environment.

An integration that has yet to be builtIn short, it behoves us to include sectorial

approaches of an institutionalist inspiration, likeRegulation School studies, notably thoseundertaken by the Rgulation Secteur TerritoiresGroup. These studies apprehend the sectorial levelas a complex social construct for the productivesphere, historically identifiable [], where specificinstitutional mechanisms are deployed (Bartoli,Boulet, 1990, quoted by du Tertre, 1995, p.313-4).

We find the same problematic identification of anindustrys institutional dynamics in studies byJullien (2004a, Jullien, Smith, 2005), who offers ananalytical framework where the sector isapprehended as a stabilised form for the creationof coherence between four fundamentally institutedrelationships (4IR) corresponding to the relationsfirms in a given sector entertain with fourcategories of resource providers they coordinate intheir midst: employees, financiers, suppliers andclients (cf. Figure 8).

We can see right away that within one and the

same type of SSIP, configurations of these 4RIswill assume different forms depending on thesector involved (i.e., automobile as opposed toaeronautics, textiles or agribusiness), even if eachof these fundamental relationships is being

instituted in a way that reflects the strategies beingpursued by firms in the branch, or the constraintsand opportunities its institutional environmentoffers. There is an immediate link between the 4IRand SSIP analytical grids, translating the directinfluence that the different institutional domainshave on three of the relationships underlying thesectors dynamics: the employment relationship;

the funding relationship; and the commercialrelationship. But the determinism is not particularlystrict at this juncture, since the sector is anintermediary space marked by the establishment ofa specific socio-economic compromise betweenactors, as exemplified by branch agreements onwages, arrangements relating to competitiondynamics (a logic rooted in both conflict andcooperation) or financial standards accepted byfirms and by analysts, bankers and even publicfunding providers. This kind of compromise can

also be found in the sourcing relationships, notablyin the norms governing client-supplier relations.Absent per se in our two grids (SSIP and MP),even if a trace thereof subsists in some componentsor institutional domains, these relationships play akey role in sectorial dynamics. If we use thisanalytical grid to integrate the intermediary(sectorial) level, we will have to explain how theorganisation of vertical relationships affects thedynamics of the other two levels. However, thisgrid has only been used to think about thesingularity of certain sectorial logics1 (Jullien,

2004a; Jullien, Smith, 2005) and up until now ithas not offered the kind of generality that would

1 Notably developments linked to the redefinition ofautomobile distribution and services, and PGI in Europe.


25/35

allow us to reflect upon a typology of sectorialconfigurations such as the one we find in othermeso-economic approaches and in other gridsbeing used at a micro or macro level (Montalban,2005b).

In addition, incorporating the sectorial level

means that we can transcend Norths distinction.

between organisations and institutions, which hadseemed legitimate at the micro and macro levels.This helps to expand the institutional construct inan approach that is more geared towardsinstitutional change (Aoki, 2001; Amable, 2003).

IR n1

Employmentrelationship

IR n2

Purchasingrelationship

Firms and industries: linksbetween the four relationships

instituted within theorganisations and competition

IR n3

Financia

l

IR n4

Commercialrelationship

Figure 8. Industry as a mode for linking four fundamental instituted relationships (4IR)

Source : Bernard Jullien (2004)

The grid is not so much designed to analyse thecompetitive process per se as it is to apprehend thedynamics of institutional change through theemergence (or non-emergence) of an industry; itsdevelopment; and mutations and crises. By sodoing, it highlights an approach revolving aroundthe construction process, and around the break-upof the various actors existing compromises. In the

emergence of an industry, there is a long process oftrial and error, before relationships can beinstituted between the actors participating in afinalised activity (the productive project) and alsobefore a set of shared representations can developconcomitantly, starting with an identification of theindustry itself and conceivably leading to aconstitution of formal institutions (like professionalassociations or labour unions). It remains that thisconfiguration, which is likely to emerge throughthe establishment of a compromise consolidatingwhatever relationships have been instituted (and by

ensuring that these relationships are coherent andinterconnected) is condemned because of inter-actor interactions to constant criticism, due to theongoing transgression (hence renegotiation orbetrayal) of these relationships. New configurationscan arise, overturning those relationships that have


26/35

been instituted (hence their linkage), withoutraising doubt about the perimeters of the industryin question. In other circumstances, this can happenper se and also due to the industrys connections toother sectors. What results are deeperrestructurings, at both the organisational andinstitutional levels, due to the fact that thesechanges are likely to lead to transformations at amacro-social level. Examples include the risingpower of component makers (destabilisation of thepurchasing relationship), the restructuring of thedistribution function and positions taken in theservice sector (going as far as mobility services),all events that could be harbingers of areconfiguration of the auto industry.

From micro-change to institutional change:

hybridisation and contamination

There is no strict macro or meso determinism

towards the micro-economic level: the rules aresusceptible to different interpretations revealing theambiguities of the institutions involved. Moreover,actors are constantly trying to ensure a modicum ofcoherence amongst these various components. Inshort, what we have here are interstices, spaces ofcreation where a local emergence can be born out


27/35

of the hybridisation of older forms (Boyer, 1998,2004b) with new configurations. These are likely tobe diffused throughout the socio-economic systemand therefore, by contamination, to produceinstitutional change at a global level.

The hybridisation of modelsA productive model affirms itself in its externalrelevance and internal coherence in one or the othertype of socio-economic environment. It necessarilyemerges within a given institutional architecture(SSIP) that constitutes a fertile breeding groun

yannick lung

Documents