Flexible Specialization, Supply-Side Institutionalism and the Nature of Work Systems

M. G. Marshall

University of East London

Review of Social Economy June 1999


From the New Right supply-side perspective, "flexibility" has been seen exclusively in terms of removing or reducing the "institutional rigidities" that prevent the full operation of market forces and the ability of economic agents to respond to price movements. In periods of prolonged economic crisis it is concluded that the market mechanism has been prevented from working efficiently as an allocative and incentive system. Thus the experience of the stagflationary era led many mainstream economists to deduce that "rigidities" and "imperfections" in the labor market had produced excessively high wage costs, restrictive practices and poor productivity. Rigidities in capital markets, allied with excessive regulation, had discouraged investment and risk-taking and high government expenditure had produced high marginal rates of tax that constituted a major disincentive to work, save and invest. In the face of such rigidities, intervention was necessary to: curb vested interests and distributional coalitions; facilitate reallocation of resources between the public and private sectors; reorder the fiscal activities of the state; and, in general, to restructure economic relations and the governmental role in the economy.

Despite the supply-side emphasis on marginal tax rates, at the heart of most neo-conservative analyses of the difficulties of the western industrialized economies was the belief that the above-mentioned "institutional rigidities", primarily those associated with trade union and governmental activity, were the main cause of the high inflation, slow technological change and sluggish economic growth that had characterised the 1970s. By the early 1980s US New Right "liberal productivist" ideas were spreading to Europe in the form of so-called "Euro-sclerosis" theories, most of which have been inspired by, and based upon, the analysis of Olson (1982). It became commonplace, for example, to ascribe the US"s superior employment growth to her more flexible labor market, and the OECD argued that:

[The] Main impediments to better functioning of labor markets arise from specific wage-bargaining institutions, tax and social spending policies, and over-protective legislation. In particular, in many countries the interaction of the income tax and social transfer system, especially unemployment benefits, exacerbated the unemployment problem (Economies in Transition - Structural Adjustments in OECD Countries, OECD, 1989, p. 16).

Alongside neo-classical supply-side analysis, but perhaps rather over-shadowed by it, there was an emergence in the 1980s of a new strand of analysis seeking to develop what Streeck (1991, p. 22) has called "a social-institutional perspective on the supply-side of modern industrial economies." This "supply-side institutionalism" held out the prospect of a superior competitiveness through sophisticated application of new technology, and suggested that a diversified product range and non-price competition policies could be combined with high wages, skilled labor and a flexible non-Taylorist organisation of work.

The two main developments were flexible specialization (FS) and diversified quality production (DQP). According to Piore and Sabel (1984), Fordist mass production was an historically contingent development, related to the pre-eminence of the large US corporations and their ability to generate mass markets for standardized products. Since the 1970s, it is alleged, the system has been in crisis for demand-related reasons. In such a crisis, it is argued, only an alternative technological paradigm can break the deadlock and offer the chance for a "second industrial divide." With the changes that have taken place in the consumer market, smaller-scale production increasingly has a competitive edge over the established vertically integrated firms, and the economies of scope supercede the economies of scale. A reversal of the Fordist fragmentation and deskilling of labor emerges, and with it the prospect of better work in which the versatility and self-direction of the traditional craft worker re-emerges. Associated with these changes is a shift from centralised, hierarchical control to a new strategy of devolved control, and from personnel administration to human resource management.

Underlying the shift to small firms is their alleged greater productivity and flexibility. The nature of the small firm encourages functional flexibility, as a limited number of employees have to fulfil all tasks, but opposes the traditional route to increasing productivity, via the division of labor. It is argued that labor and technical stability are now rigidities, and economies of scale are no longer paramount.

The DQP approach, Streeck argues (1991, p. 28), was influenced by the literature on FS. Certainly there are some obvious and strong similarities between DQP and FS. For example, they are both opposed to the neo-classical free market supply-side approach and argue that there are important market failures, e.g. regarding training. Both FS and DQP reject the view that greater efficiency requires a pared down institutional framework corresponding to the "neoclassical minimum" of unregulated markets and unlimited managerial prerogative. Indeed, certain highly successful production patterns require for their emergence and survival strong non-market institutions that modify, and partly suspend, individual market rationality and unilateral managerial control. High wages, stable employment, worker participation and social equity, it is argued, can be compatible with good economic performance. Also, they are politically motivated in that their guiding interest is the potential of advanced industrial production systems to underwrite a movement towards improving the general welfare via a more egalitarian society with greater autonomy and participation in the workplace and more stable employment. Moreover, they both adopt a view of technology that "emphasises contingency and the scope for strategic choice" (Hirst and Zeitlin, 1991, p. 2). Both argue that new technology does not "determine" the nature of production systems: it does not necessarily lead to a "quality-competitive" approach.

The DQP approach was influenced by the corporatist debate of the 1970s and 1980s in that amongst the most important aspects of this debate was a supply-side dimension of neo-corporatist institutions i.e. a perspective on their actual and potential role in the formation and functioning of production systems. Such insights helped prepare the ground for a reorientation of the corporatist paradigm from the Keynesian concerns of the 1970s to the structural "qualitative" concerns of the 1980s and 1990s. The DQP approach has clearly sought to distance itself from approaches that focus on US and Japanese based analyses of "high technology" or "knowledge-based" manufacturing where trade unions are weak and politics is influential only in the form of MITI-style "high politics". Also, unlike FS, it has not been influenced by the Italian industrial districts. In fact, as shall be seen below, DQP is very strongly based on (West) German experience.

This paper, like the other papers in this mini-symposium, addresses aspects of the "post-Fordist" debate. It is not concerned, however, with the overall debate or a comprehensive perspective of alternative views, but is instead focused on competing perspectives of supply-side institutions. It is concerned with the differences between New Right perspectives and those of the proponents of supply-side institutionalism. It contrasts the views of labor market "flexibility" provided by neo-classical Euro-sclerosis analysts and supporters of deregulation with those of the FS and DQP theorists, and assesses the strengths and weaknesses of FS in comparison to its main "institutionalist" rival.


According to neo-liberals, inflexibilities in the labor market have been the key barrier to employment growth, and therefore there has been a need to bring about a fundamental transformation in the relationship between the state and the labor market and to restore freedom of contract as the basis for economic relations. At a policy level, this particular flexibility thesis makes several arguments. Firstly, wages are too high and too rigid, thereby pricing workers out of jobs and creating unemployment. Here powerful trade unions and collective bargaining arrangements are seen as the main problem. Secondly, wage differentials are too small, hindering labor mobility and efficient structural adjustment. Thirdly, legally based labor rights are too extensive, leading to high labor costs. Finally, social security systems encourage voluntary unemployment and act as a disincentive to work. State intervention is held responsible for these excessive legal and financial guarantees to labor. In response to this view of a rigid labor market, governments need to curtail their involvement in the labor market and limit the power of trade unions, to give employers more freedom of action, so that employment will grow.

FS theorists, on the other hand, reject market competitive solutions since they emphasise the importance of social relationships that secure crucial inputs and vital collective services for firms: inputs that cannot be guaranteed by sovereign enterprises purchasing the factors of production in open markets. These include trained labor, market and export information, diffusion of technical information and on-going relationships based on trust with sub-contractors and partner firms. The high costs to firms of low trust relationships, the commodification of information, and the absence of ongoing collaboration with labor involve real competitive disadvantages. Thus, countries like the UK which have been most dominated by liberal competitive policies and antagonist competition between social interest groups over the past two decades are, it has been argued, least likely to be successful in introducing FS strategies. Indeed, according to some FS theorists, the UK constitutes the acid test of whether FS can be consciously encouraged by public policy initiatives since it desperately needs them but has few of the institutional resources necessary to achieve them (Hirst & Zeitlin, 1991, p. 47).

Similarly, Streeck (1991, p. 33) argues that a regime of free markets and private hierarchies is not enough to generate and support a pattern of DQP - even where product demand is favorable. DQP can only be partly, provisionally and precariously sustained on a voluntaristic-contractual basis in a market-based competitive environment. Firms cannot just rely on their own private organizational resources, they need to build on, relate to and invest in a common, public, institutional assets and structures. It is vital, Streeck argues, that individual firms are not constrained by the limits of rational individualism.

Streeck has identified important dimensions of market and hierarchy failure which relate to three key requirements of DQP. These are: "congenial organizational ecology", "redundant capacities" and "collective productive inputs" (ibid., p. 34). Taking these in turn.

Firstly, according to competitive market logic, the prosperity of one firm is based on the achievement of zero-sum gains over competing firms. DQP, however, needs strategic alliances and joint ventures between firms at the same level of the product chain, and trust based co-operation between assemblers and suppliers at different levels. The reasons for this include: high R&D costs due to more rapid product turnover and more specific product customization; the need for high quality standards; and advanced logistical methods requiring suppliers to be both technically competent and highly responsive to the requirements of their customers. Firms need to be both competitors and potential allies and collaborators. Thus institutional mechanisms are required that overcome suspicion among competitors, insure firms against opportunistic defection of partners, and enable firms to invest not only in their own performance but also in that of other firms.

Secondly, greater volatility in markets and a more rapidly changing technology generates a need for fast retooling in order that firms can react to changing demand. A precondition for this is the presence of generalized, unspecific, spare capacity that can be put to different, previously unknown, uses. This spare capacity is not easy to justify to superiors in the firm"s hierarchy. It is, in fact, likely to be perceived as "excess" investment. Small firms in particular find it difficult to build polyvalent spare capacities for production and design and marketing.

Thirdly, empirical research suggests that DQP requires a range of inputs that have significant collective good properties. In these circumstances, prisoners" dilemma problems may often lead firms to choose strategies which are not in their long-term interests. For example, in some circumstances cutting back on institutional efforts to achieve stable industrial relations characterized by trust, or indeed taking opportunistic advantage of temporary bargaining weaknesses of trade unions, may appear rational or even inescapable. Similarly "poaching" workers trained by rival firms may have its short-term attractions. Such a strategy, however, means that there is either a tendency for in-firm training to be reduced or for firms to be interested only in the transmission of firm-specific skills. Unfortunately, DQP requires training in more generalized skills. The uncertainty in internalizing the returns on such training turns it into a collective good; and, as shall be seen below, suggests a need for governmental policy action if the economy is to be provided with a more optimal supply of such skills and training.

Fortunately, amongst policy makers there has been by no means a blanket acceptance of the prescriptions of the flexibility thesis across Europe. In the 1980s and 1990s, in fact, the UK has had the only European government fully committed to the prescriptions of this type of flexibility thesis, and consequently it provides a salutary case study on the effects of zealous deregulation. One of the great dangers inherent in de-regulated economies is that severe foreign competition may lead managers to pursue strategies of short-term opportunism which exploit temporary reductions in the market power of employees - especially as they will think that their competitors will do so. They may resort to "defensive" flexibility strategies that restrict investment in new technology and training and undermine long-term trust relationships between management and employees. It seems that, for most part, defensive motives have predominated in the UK. These defensive strategies by employers have had a detrimental impact on the workforce. Market stagnation and low growth, tight competition in many product markets, and curtailed investment has meant a greater emphasis on increasing labor productivity and reducing labor costs. These practices can be seen as renewed attempts at intensifying labor effort and limiting wage costs within existing production patterns, rather than a move to a new model of economic growth (Grahl and Teague, 1989).

There are real and persistent dangers in a competitive free market situation that firms will not act in ways that are to their (and the national economy"s) best long-term interests. Long-term trust relationships are essential if increased competition and more volatile markets are to be tackled in a socially progressive way. Improved worker motivation and co-operation can lead to improvements in productivity and flexibility. Mutual trust and social peace are often based on long-term employment relationships. However, firms will not usually commit themselves to these if they have easy access to external labor markets and can indulge in "negative" and "defensive" flexibility strategies. Constraints are necessary to protect profit-seeking agents from their own individually "rational" choices and to prevent them from looking to short-term fixes. De-Taylorization of work can be a logical, and commercially viable, way of generating greater diversity of product and higher quality: but, it is not a strategy that will necessarily be chosen in the absence of extra-economic pressures on employers for the reorganization of work. Competitiveness can increase as a result of adjustments individual firms, could not, or would not, have made willingly themselves. In certain conditions, the internal flexibility that firms can get in exchange for external rigidities may be more conducive to better performance than the external flexibility associated with unstable employment. Thus, institutionally imposed obligations on management (including, in some cases, legislation) seems to be the key - not de-regulation.

Streeck"s (1987, pp. 455-8; 1991, pp. 51-55) work suggests that a number of institutional rigidities have worked as a stimulus to industrial progress in (West) Germany. These include the following:

1) A system of rigid wage determination operated by strong and well-established trade unions and employers associations. This has meant that wages have been higher, and differentials narrower, than they would have been in a free labor market. This has enforced a movement to product markets where non-price competition is important, to investment in training and retraining in order to increase productivity; and to changes in the organisation of work.

2) High employment stability. The policies carried out in this area by Swedish and German firms have not been paternalistic handouts to undermine trade unions and be withdrawn at will. In both countries, but especially in Sweden, large firms have been under considerable legal, political and economic pressure not to dismiss workers. Under these arrangements there is more redeployment (rather than sacking) and more training and retraining of labor than would be the case under conditions of neo-classical flexibility. In order to compensate for such institutional rigidities, firms have to increase their internal flexibility. Fortunately, this has been possible since workers are encouraged to be more co-operative and more favorably disposed towards technological change. Moreover, in Germany the works council, the extended arm of the trade unions under co-determination, has legal rights to participation in the firm"s manpower policy - including the right to demand that manpower planning be formally undertaken. Medium and long-term manpower planning is a necessary condition of flexibility in an internal labor market closed by "external rigidities."

3) Rules which give workers greater security about long-term "voice" and encourage them to forego short-term advantages and take a longer-term view. Co-determination is a devise which has compelled consultation and given workers greater "voice." It has, according to Streeck, helped guard against short-term opportunistic behavior, but employers would not have adopted it voluntarily - it had to be imposed by law. However, because it was established via legal compulsion, it has had the benefit that workers know it will not be unilaterally withdrawn in times of economic hardship.

4) Training regimes that oblige employers to train more workers in a broader range of skills than are immediately required. The German "corporatist" training system is a system governed by trade unions and employers association under the overall aegis of the state. It has national standardized curricula in which firm"s training activities are closely supervised by quasi-public bodies (Chambers) with compulsory membership and far-reaching powers, and it has helped to produce an "excess" (to current, immediate, requirements) pool of flexible and polyvalent workers that constitute a major advantage in periods of rapid technological change. Such skill acquisition is important for productivity reasons, but is also of value because it provides an institutional underpining for high wages and employment protection. And, finally,

5) Institutional arrangements which create a pressure for improvements in work systems. Through collective agreements, union-inspired governmental programs such as the "Humanization of Working Life" have created pressure for the de-Taylorization of work in Germany. Managements facing pressure from trade unions pursuing the goal of job enrichment are more likely to go for quality production and customized markets. Moreover, employers who are restricted in their ability to fire and recruit externally, become more dependent on worker motivation: and this, in the main, is enhanced by de-Taylorisation of work. At the same time, more motivated workers, with a strong commitment to the firm and their work, enhance the prospects of DQP.

In short then, according to Streeck, the institutional constraints facing (West) German firms have provided a major stimulus to them to move towards DQP.


The flexible specialization model put forward by Piore and Sabel (1984) posits the possibility of a move away from mass production towards smaller units of "craft production" facilitated by information technology. FS is primarily concerned with what is necessary to achieve effective manufacturing performance and it is only interested in macro-economic and social policies in this context. It is not a comprehensive economic and political program as such, although its concepts can be, and have been, integrated into broader advocacy of strategies of reform. Piore and Sabel"s (1984) analysis has not only proved to be influential in the US, but has also generated a large literature in Britain. The analysis has been utilised to explain Britain"s competitive failures (Hirst and Zeitlin, 1989a, 1989b) and has proved attractive to segments of the British left - including radically inclined local authorities seduced by the prospect of self-sustaining "industrial districts" to replace collapsing staple industries. However, much of the literature has been critical, particularly that written from a marxist or quasi-marxist perspective. A number of criticisms of this analysis have emerged.

The fundamental premise of market saturation and fragmentation has been questioned on empirical grounds. The "openness" of the analysis has also been questioned in that critics have argued that it is teleological, with market trends the determining factor. Piore and Sabel"s assertion that "international Keynesianism" could provide renewed prosperity on the basis of mass production, it has been argued, is incorrect in that it would not be able to provide a solution to the market fragmentation alleged by the authors. In effect, it has been argued, the analysis implies that new technology is inevitably bringing in a new era of small scale craft production. This, however, the critics allege, is not the case, and mass production systems, in fact, have proved themselves capable of a great deal of flexibility and product variety (Williams et. al., 1987; Elam, 1990). Further criticism has focused on the apparent ambiguity of FS and its lack of theoretical clarity. Is it an ideal-type analysis? Is it an observable empirical phenomenon, or merely a normative analysis?

The vertical disintegration and decentralization of economic power arguments have also drawn heavy criticism. Is it the case that we have entered a "post-Fordist" era which signals the end of vertical integration and centralization of economic power? The restructuring that occured in the 1980s seems to have had mixed effects. In some cases, there was a deepening of the old "Fordist" forms, with vertical integration and the pursuit of scale economies remaining important. In other instances, the experience of the 1970s caused MNCs to review their product strategies and move towards specialization in the high value-added or expanding product markets (Dunning, 1988). However, if some MNCs sold off interests in secondary or peripheral markets and inceased in-house product differentiation and flexibility, this, it seems, was all done without loss of central control over operations along the value-added chain. In fact, it has been argued that there is only limited evidence of decentralization, and there is little suggestion that centralized control has been relinquished in certain strategic areas like research and development (Amin and Dietrich, 1990). There has been a movement towards both integration and disintegration simultaneously. The more operational and functional disintegration within the corporation, the greater the degree of integrated management and control. Whilst the productive process may be becoming decentralized, strategic control has become more centralized (Amin and Robins, 1990; Amin and Dietrich, 1990; Hoggert, 1990).

The changing nature of large firm"s linkages with suppliers, buyers and competitors has given the impression of growing decentralization and anti-monopoly tendencies in the economy. An example of this is the "hollow corporation" in which all but the development, design and distribution of the product is sub-contracted to small specialist firms. This facilitates least-cost and flexible responses to ever-changing market conditions through the externalization of risk, uncertainty and productive capacity, but the "hollow corporation" retains control in all of the strategic areas required for securing profits and market leadership; and, if viewed from the sub-contractors perspective, the problem of centralized control and dependency on one major source of work remains (Amin and Dietrich, 1990). Thus the flattening of the hierarchy between dominant buyers and dominant subcontractors or suppliers need not be understood as decentralized corporate control rather it can be seen as an elaboration of hierarchy: a case of quasi-integration between large players along the value added chain (Amin and Smith, 1991).

There has been considerable debate over the scale and significance of changes in the productive systems of the advanced industrial economies. Several critics have argued that the changes in methods of production have been overstated, and the difficulties facing their implementation and dissemination underplayed. For example, switching between products has either not been sought or proved difficult to achieve, with problems of compatibility between machinery installed at different times proving particularly troublesome (Hyman, 1991). Certainly the spread of computer-integrated manufacturing, the pinnacle of flexible automation, has been exaggerated.

The OECD study (1989) of flexibility concluded implausibly that the prime mover has been "the massive introduction of new technology" which has reintroduced flexibility "into what had been an immutable production law" (p. 58). In fact, contrary to any idea of "immutable production" laws, it is clear that there has been, and continues to be, considerable flexibility within so-called Fordist mass production systems. In car production, the archetypal Fordist activity, there has been clear evidence that the volume producers can handle more customized demand (cabriolets, coupes, etc.) and that within mass production set-ups there can be added a multiplicity of options in terms of engine size, braking and steering systems, four wheel drive etc. (Williams et al., 1987).

Moreover, so-called Toyotaism, whilst it gives a central role to qualified labor, in no way uncouples production from the assembly line. In Germany in the more advanced car plants, the role of skilled workers is perhaps greater, with their work to some extent being removed from the routine of the conveyor belt. Ultimately, however, their work in still constrained within the parameters of the basic mass production system of which they are a part. All in all, the changes that have taken place have constituted only a partial movement from "Taylorism": a modification rather than a fundamental transformation of traditional principles and practices of mass production (Jurgens, 1991). In the US car industry there have been attempts to ape Japanese rivals, but despite the growing emphasis on semi-skilled work teams, the system remains a Fordist one (Jurgens, 1991). Further, Berggren"s (1989, 1992) studies of the Swedish car industry, whilst suggesting that changes have improved the quality of work in some areas and involved a strong trade union participation, have shown that it has suffered from relatively high costs in some areas and also from variable productivity.

More generally, the changes are even more patchy. Appelbaum and Batt"s (1994) analysis of nearly 200 surveys and case studies on workplace change in the US since the 1970s concludes that US firms have implemented innovations on a piece-meal basis and that, in the vast majority of cases, the changes implemented "do not add up to a coherent alternative to mass production" (p. 10). So how far has there been a clear movement away from the assembly line and a reversal of the trend towards the division of labor and in which production is for niche markets? Clearly there have been developments such as the creation of new skilled jobs, teamwork production, growing interest in quality of work programmes: but how far have these spread? And how far do they form a coherent pattern rather than a series of disparate developments? Experiences have been strikingly diverse (Lane, 1988), and certainly, one must be wary of over optimistic visions of the growth of artisan production and expanding self-realization for the labor force in a world in which there are clearly trends in some areas towards deskilling and technological unemployment.


Thus FS, as derived from the work of Sabel (1982) and Piore and Sabel (1984), has been taken to be a pro-small firm, industrial district type analysis with very optimistic implications for the development of non-Taylorist work, and this viewpoint has been subjected to considerable criticism. However, Sabel (1989) has developed and clarified the analysis of FS in the light of these criticisms; and, drawing on this, Hirst and Zeitlin (1991) have disputed many of the inferences drawn, and criticisms made of, the FS thesis. They have also queried the basis of empirically based criticisms. Certainly, some of the features or inferences of the original analysis which caused most concern, for example the exclusive identification of FS with small (rather romanticized) craft production, the failure to recognize the existing degradation of work and the real possibility of the continued generation of "bad work" have been addressed. Also considerable emphasis has been put on the "double convergence" (small and large firm) towards flexible specialization and on the contingent (non-teleological) nature of the analysis. Here it is appropriate to point out that although Piore and Sabel"s analysis has commonly been discussed in textbooks and elsewhere under the generalized heading of "post-Fordism" (along with the work of Freeman and Perez and the French Regulationists), Hirst and Zeitlin take great pains to try and distinguish FS from "post-Fordist" theories. The latter are classified as those theories which see industrial change as a mechanical outcome of impersonal processes. FS theory, on the other hand, emphasises contingency and the scope for strategic choice. As Hirst and Zeitlin see it, much of the debate over FS has missed the mark by construing it to be a similar type to "post-Fordism" in its many variants.

Hirst and Zeitlin (1991) in their long and systematic defence of FS tackle key areas of discussion relating to small firm craft production, the role of demand and contingency. FS, we are clearly told, is not a theory which gives necessary prominence to the small firm. The ideal-type is not to be taken as an empirical generalization. It is not the sole or major form of flexible production. Moreover, it is argued, firms in most countries and periods deliberately mix elements of craft and mass production ("hybridization") because they are acutely aware of the dangers involved in choosing an unalloyed form of either model. They hedge against risks in ways that blur the lines between them. Hirst and Zeitlin argue that ascribing changes in work systems to market factors, implies that strategies are purely reactive. This violates "basic assumptions" of the FS approach - namely that effective strategies are anticipatory. Hirst and Zeitlin, in fact, make clear attempts to downplay the role of demand and suggest that equally as important is the trend for firms organized along FS lines (especially in Japan) to deliberately fragment demand through constant introduction of new speciality products.

Despite Hirst and Zeitlin"s attempts to strengthen and clarify FS analysis, however, important questions remain unanswered, and key issues remain unclear. We are told that hybridization is the norm, but also that the interpenetration of flexible and mass production makes it easier to shift from one pole to the other than abstract considerations of the two models might lead one to expect (ibid., p. 6). But why, if the risk-minimization argument for hybrids is convincing, would firms want to do this? Moreover, would the switching towards the large scale pole not imply a continuing importance for the economies of scale? Further what circumstances favor the small scale craft model of FS? Why has there been such an emphasis on it in the literature if it is a little seen exception to the general rule? Hirst and Zeitlin are concerned to play down the role of demand in causing significant changes in production systems as they seem keen to promote FS as true managerial strategy (active rather than reactive). However, if it is not demand factors, what exactly has been stimulating the choice of FS strategies? We are not told, and constant reference is made to pro-active strategies being utilised "as well as" reactive (to demand) ones. Moreover, confusingly, we are told that the "increasing volatility of international markets and rapidly shifting patterns of demand in the period since the early 1970s favor more specific strategies of response like that of flexible specialization" (p. 41, my emphasis). Apparently we can, despite earlier assertions, have pro-active and reactive FS strategies.

Hirst and Zeitlin claim that although international competition does not impose a single form of productive organisation, there is nevertheless an observable tendency towards the displacement of mass production by flexible specialization as the dominant technological paradigm of the late twentieth century. It is argued that FS, unlike Regulation theory, does not, illegitimately, fall back on the general tendencies of capitalism as a mode of production. Nor does it treat technology as an independent causative agent with its role specified independently of how it is integrated in productive organisation and business strategy. However, what generates these tendencies? What causes the different choices that are made to all tend towards FS and the displacement of mass production?

Moreover, is hybrid FS very different from what others have referred to as flexible mass production? And, if so, how? How far can we actually assess Japanese firms as examples of FS rather than as evidence of increased flexibility within mass production? Apparently (ibid., p. 35), it is not a legitimate objection to FS to emphasise the continued role of large firms, or the continuance of large volume production. Perhaps, but what exactly is the essence of FS? Is it not the absence of the assembly line and mass production? If not, what exactly distinguishes large firm FS from flexible mass production? The same question can be asked of DQP. For example, Streeck"s exemplar of the (West) German car industry illustrates, as we saw above, the sort of changes referred to by Williams et al., (1987) and others as "flexible Fordism."

Hirst and Zeitlin argue that mass production and FS are ideal-type models rather than empirical generalizations or descriptive hypotheses about individual firms, sectors or national economies. Nevertheless, we are assured that "current manufacturing practice is moving in the direction of flexible specialization as a specific model of productive organisation taking account both of the plurality of institutional forms within which it may be pursued and of the possibilities of hybridization" (p. 33). Our ability to ascertain the validity of such claims, however, is clouded by ambiguities about the defining differences between FS and flexible mass production (referred to above), and by conceptual difficulties about the nature of "evidence".

Hirst and Zeitlin argue (pp. 22-26) that we cannot simply test theory by reference to a set of changes occurring "out there" in the "real world", for it is impossible to construct a "theory-neutral domain of evidence". We must test theories in terms of their "plausibility" and their "intellectual productivity" (which "rests solely on its capacity for arguing and showing what it claims is the case with a reasonable degree of probability"). Unfortunately, the micro and macro evidence that would be necessary to assess whether FS theory really "illuminates observable processes of industrial change"(p. 33) is impossible to obtain because of "multi-dimensional complexities"(p. 34) and therefore "the preferred form of evidentialization" for flexible specialization is the analytical case study conducted at the micro level of particular firms, regional economies or industrial districts. Unfortunately, however, as the authors themselves admit, there inevitably arise the problems of interpretation and representativeness with case studies. An example is not proof, and case studies contain an ineradicable element of subjective interpretation.

As Hirst and Zeitlin claim, showing that FS strategies have not been generalized, that they exist only in certain cases, and that they do not exist in a pure ideal-typical but in a hybrid form, does not constitute a refutation of FS as advocacy. All the advocate of FS as a normative approach has to do, they suggest, is to show that such strategies are possible and that they can be expanded beyond given cases, even if in hybrid form. FS as an ideal-type and normative proposal can be seen as having a lot to recommend it and, of course, it retains this force even in the face of empirical evidence that there is no trend towards it as yet. However, even on the level of advocacy there are ambiguities. We are told quite clearly (ibid., p. 7, fn. 10) that FS is not an optimistic general theory of the labor process which can be counterposed to Braverman"s deskilling thesis (although this is how many observers seem to have interpreted Piore and Sabel"s analysis). There may be technologically innovative forms of both mass and craft production, but there are also "stagnant" versions in which firms compete through squeezing wages and working conditions and product quality - practices as common in large declining enterprises as in small sweatshops. Certainly the FS theorists Christopherson and Storper (1989) suggest that the movement to FS in the movie-making industry in the US has been associated with an increase in part-time working, greater casualization of work, and indeed the emergence of a core-periphery labor force. New institutions are needed, they argue, to ensure that flexibility does not mean insecurity and exploitation. Trust relationships and a system of rules are necessary before we can even have normative advocacy for FS it seems. However, unfortunately, FS theorists have little tangible to say about how "progressive" FS can be brought about.

Since FS is not a theory predicated on preferring small firms (large multi-product, multi-national companies can adopt FS strategies as well as small workshops) there is no bias within FS industrial policy, as there is within traditional neo-classical theory, towards small firm competitive models. Successful FS firms, however, may well become the target of asset rich large firms and therefore need government protection from take-over. Equally, however, Hirst and Zeitlin oppose the policy objective of promoting concentration to secure economies of scale. Many large firms which have grown rapidly via a process of state-indulged merger and acquisition are not efficient they argue. Moreover, the Japanese model is also rejected on the grounds that it is heavily dependent on close and informal links between senior civil servants and corporate executives that would not be acceptable in the West. Also, Hirst and Zeitlin attribute much of the Japanese success to orthodox company-led innovation and manufacturing strategies characteristic of FS, rather than MITI-style intervention whose record, allegedly, is not as good as some of its Western enthusiasts suggest. The French Regulationist writers are criticised because there is little in the way of (narrowly defined) industrial policy in their work, but what emerges out of the above in the way of specific, concrete proposals for policy? Most of it is just telling us what FS theorists do not support.

Trust needs to be institutionalized and co-operation presupposes forums in which it can be developed. This, Hirst and Zeitlin argue, is the most important lesson for policy makers. It is the socio-political conditions in which manufacturing is embedded that form the core of the FS approach to policy. There are two main routes, we are told, for developing and institutionalizing FS:

a) building up - linking up firms with collaborative institutions to form and cement industrial districts (and seeking to generalize and link such districts to form the dynamic core of a national economy),

b) building down - reorganizing major multi-national firms into constellations of semi-autonomous sub-units that can co-operate with one another or other firms.

But who does this and how? What could the role of government policy be here? FS policies place considerable emphasis on training, since broadly trained workers are a core component of such a strategy. However, as has been seen, there are problems of achieving the right levels of training because of the free-riding problem of poaching. There is a need for co-operation between firms and with labor (to set mutually acceptable standards and employment policies). How are these problems to be resolved? Only in the DQP analysis of Streeck do we get some definite proposals. Further, as discussed above, FS is subject to destabilizing effects and can take regressive forms e.g. low wage, high-intensity regimes promoted by competition from unscrupulous (or desperate) employers. Is not some sort of framework of law or regulation required? Again we have to turn to Streeck for a discussion of such issues.

As was noted above, there are a number of important similarities between DQP and FS. However, they are also clear differences, and in examining these we see the areas in which DQP is superior to FS. As we have seen, there is a certain ambiguity concerning the differences between "large firm" FS, DQP and flexible mass production. DQP analysis acknowledges the importance of the economies of scope. However, unlike FS advocates, DQP theorists argue, perhaps more realistically, that there continues to be significant economies of scale in manufacturing. Also they recognize that there are still important advantages associated with vertical integration and large scale production.

The incompatibility of supply-side institutionalism with neo-liberalism, noted above, is because FS and DQP require for their long-term success an irreducible minimum of trust and co-operation amongst economic actors, for example between managers and workers within the same firm, and between firms and their external subcontractors. Such co-operation depends, in turn, on the establishment of rules limiting certain forms of competition such as sweated wages and conditions. Trust and co-operation does not rule out conflict, of course. The reproduction of social consensus can only be sustained in the longer term through the creation of institutional mechanisms for the resolution of disputes. However, only Streeck provides us with an analysis of what these mechanisms might look like. He also has a view (as has been seen) about how perceived conflicts can be institutionalized in ways that may stimulate profit-seeking entrepreneurs to take a longer-term view and to move in socially desirable directions.

The more speculative chapters of Piore and Sabel"s book see FS conjure up a high technology version of Jeffersonian "yeoman democracy" (communitarian voluntarism). DQP, however, eschews utopian visions, and requires merely the containment and counter-balancing, rather than the elimination of, corporate hierarchies. As Streeck argues, the most important difference between the two approaches is that between the communitarian voluntarism of FS and the "harder" approach of DQP which is based on an analysis in which organised social conflict, trade unions, employers" associations, the law and the nation state continue to play a major part. Streeck"s Durkeimian concern with the regulatory, interventionist, and non-voluntary elements of social institutions corresponds to a strong emphasis on prisoners" dilemma-type co-ordination problems, and on the insufficiency and indeterminacy of rational individual calculations vis-à-vis complex, long-term decisions on the development of advanced productive capacities. The latter are present in FS theory, but the "softer" cultural mechanisms which Piore and Sabel rely on to contain competition and enforce obligations in the industrial districts are unlikely in reality to be able to perform that function. They appear to be insufficient to promote the successful long-term reproduction of a socially progressive system of FS.


The neo-classical supply-side approach ignores or downplays non-market behavior and market failures. From this perspective, flexibility involves the removal of all institutional "frictions" that slow down or prevent reaction to market stimuli. Such views, however, are simplistic and ignore the fact that certain "institutional rigidities" are essential to the working of markets. They are needed to reduce uncertainty, guide expectations and generate a necessary minimum of trust (Hodgson, 1988; Neilson, 1991b). Flexibility in some areas requires stability ("rigidity") in others. FS and DQP offer radical supply-side policies that are vital in the context of the failures, and social costs, associated with New Right supply-side policies. Their significance is further enhanced by the constraints and limitations facing Keynesian demand management policies in a world of increasing trade liberalization and more intense competition in world manufactures. As Hirst and Zeitlin (1991, p. 41) have argued, meeting the conditions of international competition requires appropriate supply-side policies and business strategies. Ceteris paribus, stimulating domestic demand may simply promote the import of manufactures and accelerate de-industrialization.

DQP, in particular, points the way towards an alternative non-neoclassical supply-side response to the employment problem which would make productive use of "institutional rigidities." The strategy is one based on finding ways of increasing productivity in the context of "high" and "rigid" wages, rather than one which matches the distribution of incomes to the current productivity profile. The goal is to design an institutional infrastructure on the supply-side that can sustain socially benevolent industrial change. However, it must be remembered that institutions that are no more than mere constraints (on firms, management or workers) will choke off economic growth. Indeed, such institutions may well become strongholds of Olsen"s rent-seeking organizations pursuing "distributional coalitions". On the other hand, institutional arrangements that merely present "opportunities" for enterprises (like de-regulation strategies) will favor exploitation of workers and the public by private sector producers.

Of course, there are difficulties and constraints facing supply-side institutionalism. The creation/redesign of supply-side institutions is more difficult than deregulation and neo-classical "institutional minimalism." Moreover, given the duration of political cycles, there will always be a temptation for policy-makers to go for short-termism. Also, of course, there will be political opposition. The idea of greater prosperity coming in joint supply; of collective investment taking precedence over private investment; of institutional regulation curbing the freedom of firms to do what they want, may seem too collectivist a strategy for many. Moreover, in certain circumstances, there may be the possibility of firms migrating to escape institutional constraints.

Finally, it must be emphasised that countries cannot simply follow "models" (be they German, Swedish, Italian or Japanese). Work systems are affected by the socio-economic framework that shape and support them. These vary from country to country and the scope for their reform similarly varies both over time and between countries. The "one best way" after all is a Taylorite chimera. This is not to say, however, that there are no lessons that can be learned from specific cases of, for example, DQP, nor that there are not certain general principles that can be derived from the analysis. The analysis of FS and DQP, and the evidence provided by important, empirically based, studies such as Appelbaum and Batt"s (1994) have proved to be illuminating. These suggest, in particular, that public policy directed towards: optimizing job training, promoting employee participation and inter-firm co-operation, and restricting the ability of firms to indulge in short-termism will be most productive in promoting the cause of socially progressive industrial production and "good work".



Appelbaum, E. and R. Batt, The New American Workplace: Transforming Work Systems in the United States. Ithica, New York: ILR Press, 1994.

Amin, A. and M. Dietrich, "From Hierarchy to "Hierarchy": The Dynamics of Contemporary Corporate Restructuring in Europe," Towards a New Europe? Stuctural Change in the European Economy, edited by authors. Aldershot: Edward Elgar, 1990.

Amin, A. and K. Robins, "The Re-emergnce of Regional Economics? The Mythical Geography of Flexible Accumulation," Environment and Planning D: Society and Space, v. 8, (1990), pp. 7-34.

Amin, A. and I. J. Smith, "Vertical Integration or Disintegration? The Case of the UK Car Parts Industry," Restructuring the Global Automobile Industry, edited by C. Law. London: Routledge, 1991.

Berggren, C. "New Production Concepts in Final Assembly," The Transformation of Work?, edited by in S. Wood. London: Unwin, 1989.

Berggren, C. Alternatives to Lean Production: Work Organization in the Swedish Automobile Industry. Ithica, New York: ILR Press, 1989.

Braverman, H. Labor and Monopoly Capital: The Degradation of Work in the Twentieth Century. New York: Monthly Review Press, 1974.

Child, J. "Organization Structure, Environment and Performance: The Role of Strategic Choice," Sociology, v. 6, no. 1, (1972), pp. 1-22.

Christopherson, S. and M. Storper, "The Effects of Flexible Specialization in Industrial Politics and the Labor Market," Industrial and Labor Relations Review, v. 42, (1989), pp. 331-47.

Dunning, J. " International Business, the Recession and Economic Restructuring," Strategies in Global Competition, edited by N. Hood and J. E. Vahlne. London: Croom Helm, 1988.

Elam, M. "Puzzling Out the Post-Fordist Debate: Technology, Markets and Institutions," Economic and Industrial Democracy, v. 11, (1990), pp. 9-37.

Freeman, C. and C. Perez, "Structural Crises of Adjustment, Business Cycles and Investment Behavior," Technical Change and Economic Theory, edited by G. Dosi, et. al. London: Pinter 1988.

Grahl, J. and P. Teague "Labour Market Flexibility in West Germany, Britain and France," West European Politics, v. 12, no. 2, (1989), pp. 91-111.

Hirst, P. and J. Zeitlin "Flexible Specialization and the Competitive Failure of UK Manufacturing," Political Quarterly, v. 60, no. 2, (1989a), pp. 164-178.

Hirst, P. and J. Zeitlin Reversing Industrial Decline? Industrial Structure and Policy in Britain and Her Competitors. Oxford: Berg, 1989b.

Hirst, P. and J. Zeitlin "Flexible Specialization Versus Post-Fordism: Theory, Evidence and Policy Implications," Economy & Society, v. 20, no. 1, (1991), pp. 1-56.

Hodgson, G. M. Economics and Institutions. London: Polity Press, l988.

Hoggert, P. Modernisation, Political Strategy and the Welfare State: An Organisational Perspective. School for Advanced Urban Studies, University of Bristol, Studies in Decentralization and Quasi-Markets, no. 2, 1990.

Hyman, R. "Flexible Specialization: Miracle or Myth?," New Technology and Industrial Relations, edited by R. Hyman and W. Streeck. Oxford: Blackwell, 1988.

Hyman, R. "Plus ca change? The Theory of Production and the Production of Theory," Farewell to Flexibility?, edited by A. Pollert. Oxford: Basil Blackwell, 1991.

Jurgens, U. "Departures from Taylorism and Fordism: New Forms of Work in the Automobile Industry," The Politics of Flexibility: Restructuring State and Industry in Britain, Germany and Scandinavia, edited by B. Jessop, H. Kastendiek, K. Nielson and O. K. Pederson. Aldershot: Edward Elgar, 1991.

Lane, C. "Industrial Change in Europe: The Pursuit of Flexible Specialization in Britain and West Germany," Work, Employment and Society v. 2, no. 2, (1988), pp. 141-168.

Lash, S. and J. Urry, The End of Organized Capitalism. Cambridge: Polity, 1987.

Lee, E. "Globalization and Labor Standards: A Review of Issues," International Labor Review, v. 36, no. 2, (1997), pp. 173-89.

Mahon, R. "From Solidaristic Wages to Solidaristic Work: A Post-Fordist Historic Compromise for Sweden" Economic and Industrial Democracy, v. 12, (1991), pp. 295-325.

Marshall, M. G. and P. Arestis. "Reaganomics and Supply-Side Economics: A British View," Journal of Economic Issues, v. 23, (1989), pp. 965-975.

Marshall, M. G. and P. Arestis. "The Myths and Realities of Conservative Economic Policy-Making in the US," Review of Social Economy, v. XLIX, no. 2, (1991), pp. 218-242.

Marshall, M. G. "Restructuring, Flexibility and the New Right in the US: The Political Economy of Plutocracy," The Political Economy of Full Employment: Conservatism, Corporatism and Institutional Change, edited by P. Arestis and M. G. Marshall. Aldershot: Elgar, 1995.

McLoughlin, I. and J. Clark, Technological Change at Work. Guildford: Open University Press, 1994.

Meegan, R. "A Crisis of Mass Production?," The Economy in Question, edited by J. Allen and D. Massey. London: Sage, 1988.

Nielson, K. "Towards a Flexible Future - Theories and Politics," The Politics of Flexibility: Restructuring State and Industry in Britain, Germany and Scandinavia, edited by B. Jessop, H. Kastendiek, K. Nielson and O. K. Pederson. Aldershot: Edward Elgar, 1991a.

Nielson, K. "Learning to Manage the Supply-Side: Flexibility and Stability in Denmark," The Politics of Flexibility: Restructuring State and Industry in Britain, Germany and Scandinavia, edited by B. Jessop, H. Kastendiek, K. Nielson and O. K. Pederson. Aldershot: Edward Elgar, 1991b.

Olson, M. The Rise and Decline of Nations. New Haven: Yale University Press, 1982.

Organisation for Economic Co-operation and Development (OECD) Labour Market Flexibility: Trends in Enterprises. Paris: OECD, 1989.

Organisation for Economic Co-operation and Development (OECD) Economies in Transition - Structural Adjustments in OECD Countries. Paris: OECD, 1989.

Piore, M. J. and C. Sabel, The Second Industrial Divide. New York: Basic Books, 1984.

Pollert, A. "Dismantling Flexibility," Capital and Class, no. 34, (Spring, 1988), pp. 42-75.

Pollert, A. "The Orthodoxy of Flexibility," Farewell to Flexibility?, edited by A. Pollert. Oxford: Basil Blackwell, 1991.

Rose, M. Industrial Behavior: Research and Control. Harmondsworth: Penguin, 1988.

Sabel, C. Work and Politics: The Division of Labour in Industry. Cambridge: Cambridge University Press, 1982.

Sabel, C. "Flexible Specialization and the Re-emergence of Regional Economies," Reversing Industrial Decline? Industrial Structure and Policy in Britain and Her Competitors, edited by P. Hirst and J. Zeitlin. Oxford: Berg, 1989.

Smith, C. "Flexible Specialization, Automation and Mass Production," Work, Employment and Society, v. 3, no. 2, (1989), pp. 203-20.

Streeck, W. "Neo-Corporatist Industrial Relations and the Economic Crisis in West Germany," Order and Conflict in Contemporary Capitalism: Studies in the Political Economy of Western European Nations, edited by J. Goldthorpe. Oxford: Clarendon Press, 1984.

Streeck, W. "Industrial Relations and Industrial Change: the Restructuring of the World Automobile Industry in the 1970s and 1980s," Economic and Industrial Democracy, v. 8, 1987, pp. 437 - 462.

Streeck, W. "On the Institutional Conditions of Diversified Quality Production," Beyond Keynesianism: The Socio-Economics of Production and Full Employment, edited by E. Matzner and W. Streeck. Aldershot: Elgar, 1991.

Williams, K., T. Cutler, J. Williams and C. Haslam "The End of Mass production?" Economy and Society, v. 16, no. 3, (1987), pp. 405-39.

Wood, S. (ed) The Transformation of Work? London: Unwin Hyman, 1989.

Woodward, J. Industrial Organization: Theory and Practice. Oxford: Clarendon Press, 1980.

click here to return to the Materials and Publications index

Produced and Hosted by the Center for Digital Discourse and Culture     © Center for Digital Discourse and Culture, Virginia Tech. All rights reserved. The physical campus is in Blacksburg, Virginia, U.S.A. For more information, please contact the Center at cddc@vt.edu