Neoliberalism,capitalist growth models,and the state: An agenda for industrial relations theory

Debating liberalization

While the fields of comparative political economy and comparative industrial relations were dominated from the late 1990s until the early 2010s by approaches which can loosely be organized around the notion of resilient, nationally distinct varieties of capitalism and thus varieties of industrial relations, the period since the severity and longevity of the financial and economic crisis became clear has widened what was, until then, a niche debate about counter-tendencies to national models. That debate about the extent, form and drivers of liberalization matters for more than just reasons of empirical accuracy and scholarly score-settling. It matters because if the empirical outcomes have not been what the theory predicted, then the theoretical assumptions that underpinned expectations of incremental change along a path, and continued national distinctiveness, are wrong, and that should send us back to a reconsideration of how to understand institutional change, and in particular the relationship between capitalist growth models and the institutions that regulate them.

Yet what is striking to this observer is that while liberalizing tendencies have been acknowledged, their implications have largely not been, nor have they yet changed in any fundamental way how change in industrial relations is theorized. The focus of the vast majority of scholarship in comparative industrial relations remains on explaining varieties of industrial relations; even if lip service is now paid to a generalized tendency towards liberalization, it has been incorporated into some notion of varieties of liberalization. Thelen (2014), for example, shifted seamlessly from acknowledgment of general liberalization to returning to a new version of varieties of political-economic institutions that largely reinforces the old. Her work is among the most nuanced and sophisticated, and it does provide an alternative theorization of institutional variation (relying upon class coalitions) but that is rare. More usual is for scholarship to reference a debate between adherents of the Varieties of Capitalism and ‘liberalizers’ early on, but treat it as an empirical question alone, and continue as if there is no need to rethink the assumptions and models used to explain change.

An insistence on institutional varieties and national or sectoral differences has a powerful hold within the sociology of knowledge production of comparative political economy and comparative industrial relations, privileging deep local knowledge of cases, and constantly reproducing itself in new forms. To acknowledge that what is taking place at the level of our cases is largely the unfolding or playing out of generalized tendencies or a common trajectory is to downplay the importance of that specialized knowledge. But the danger of this understandable temptation is that it becomes an exercise in comparative statics that fails to acknowledge the profound shift in the industrial relations landscape that has taken place over time, and hence the different world faced by workers and unions everywhere. That in turn both sets an unrealistic set of expectations in the union renewal literature (Ibsen and Tapia, 2017) and does not provide strong theoretical foundations for explaining the world.

It is worth noting that in the initial Hall and Soskice formulation of the Varieties of Capitalism (VofC) framework (2001), the institutional infrastructure of capitalist political economies was linked to an argument about compatibility with production regimes. Coordinated market economies were deemed most supportive of a certain kind of ‘diversified quality production’ (Streeck, 1992) organized around high-wage, high-skill, high-value-added forms of manufacture which supposedly benefit from institutions that encourage more ‘patient’ capital, employee voice, industry-specific skills and wage coordination. Meanwhile the institutions of liberal market economies, emphasizing unregulated labor and capital markets and the provision of general skills, were more appropriate for radical innovation and rapid change. Change over time in the production regime was therefore possible within the VofC approach, but in practice nothing in its theoretical armory helped understand how that would take place, the conception of employer and union interests mitigated against conflict over institutions, which were understood to have coordinating and deliberating functions (Hall and Soskice, 2001) rather than powering ones, and very strong mechanisms were anticipated to encourage reproduction of the growth regime (Howell, 2003). Perhaps understandably then, the VofC approach came to focus upon institutions divorced from the growth regimes which they supposedly regulated or coordinated. As Pontusson put it, ‘the VoC literature has a great deal to say about “varieties”, but surprisingly little to say about “capitalism”’ (2005: 164).

Comparative political economy and comparative industrial relations borrow freely from each other, and yet have different strengths and weaknesses (Doellgast et al., 2018; Vidal and Hauptmeier, 2014). It is certainly the case that comparative industrial relations have been less susceptible to the adoption of equilibrium assumptions about labor markets and approaches in which industrial relations institutions can be understood in terms of coordination problems, not least because class conflict is much more visible in this realm. Marxian analysis has always had greater purchase and remains less marginalized in comparative industrial relations (Hyman, 1975; Kelly, 1998). It is worth noting that labor process scholars have maintained an approach that takes as its starting point the accumulation and valorization needs of capital, and conflict between managers and workers (Adler, 2007; Braverman, 1974; Knights and Willmott, 1990). As such, it is much less sanguine about the stability of varieties of capitalism, and quicker to take note of the broad shift in class power away from workers and unions towards employers. Nonetheless, as from within political economy, industrial relations scholars came to adopt typologies of industrial relations which largely reproduced the logic of varieties of capitalism, albeit with different varieties (Hamann and Kelly, 2008; Milner, 2015), even if they often rejected rational choice causal arguments in preference to explanations that privilege power resources, the ideational, or national traditions (Crouch, 1993; Hyman, 2001). Whatever the typology or the causal explanation for difference, none were good at explaining change, and certainly none anticipated transformational change along a similar trajectory within a comparatively short time.

How then should we measure institutional change in the sphere of industrial relations change in a way that neither is entirely an exercise in comparative statics, nor abstracts institutions from the forcefield of economic and class drivers within which they function? The legacy of the institutional turn within comparative political economy beginning in the mid-1980s was to increasingly detach institutions from the structural conditions that produced them, to attribute independent causal power to them, and to emphasize above all their ‘stickiness’, in the sense of encouraging only limited incremental change along an established path (Hall and Taylor, 1996; Steinmo et al., 1992). Indeed, ‘the idea of persistence is virtually built in to the definition of an institution’ (Thelen, 2009: 474).

Measuring the extent of liberalization in the sphere of industrial relations institutions, therefore, requires specifying what liberalization means and then identifying mechanisms of institutional change that go beyond merely crude exercises in marking the distance from some ideal-typical conception of laissez faire or capturing only the form of institutions and not their function. For industrial relations institutions, liberalization is best thought of as involving an expansion of employer discretion over the management of labor in the firm, which includes pay determination, the ability to hire and fire, and the organization of work (Baccaro and Howell, 2017a), which is to say, the removal of constraints upon employer control, whether external to the firm, such as state regulation or higher levels of collective bargaining, or internal in the form of firm-level bargaining, codetermination and so on. As such, liberalization in practice refers to some combination of deregulation on the part of the state, decentralization and individualization of bargaining, and decollectivization of class organizations, most importantly trade unions.

If we look across the political economies of advanced capitalist societies and use these metrics of employer discretion we can of course identify national (and sectoral) variation at any time point. We should anticipate such variation; countries have different starting points with respect to the degree and kind of constraint upon employer discretion, and thus different institutional legacies and obstacles to liberalization, as well as very different physiognomies of class organization and capacity. We can observe, for example, that liberalization since the late 1970s has moved in waves, each impacting different groups of countries at different times, from the existing liberal market economies in the 1980s to the continental European and Scandinavian political economies in the 1990s and early 2000s (with Australia and New Zealand at the inflection point between these two waves), and a third wave affecting southern European countries since the early 2000s and accelerating since the Eurozone crisis. But in comprehending the scale and scope of liberalization, those differences at fixed points in time, where individual countries stand along a trajectory at any given moment, are less valuable than identifying the common trajectory itself. One can argue whether the range of dispersion of union densities, bargaining levels, degrees of labor market regulation, strike rates and so on was greater in 1975 than 2018, but what is hard to argue with is that in every case there has been movement towards greater employer discretion over the last roughly 35 years.

In some analyses of institutional change in industrial relations that recognize that models of varieties of capitalism have undergone substantial change but reject the notion of a common liberalizing trajectory, the concept of dualism has been revived (Hassel, 2014; Palier and Thelen, 2010; Thelen, 2014). The argument is that liberalization is occurring in the service sector but that institutions in the core manufacturing sector, particularly of coordinated market economies, remain resilient. But empirical limitations of this claim aside (Baccaro and Benassi, 2017), this is to engage in the same form of comparative statics. After all, Goldthorpe (1984) once argued that dualism characterized the United States and Britain, in contrast to the corporatist countries, and we now know that dualism was just a relatively brief stop along the way to comprehensive liberalization. Applying that same argument now, three decades later, to those once corporatist countries raises questions about why we should anticipate any different outcome this time around. Unless there is some reason to believe otherwise, and absent some reversal in emerging growth models and the balance of class power between labor and capital, dualism is likely to be an intermediate step in the process of liberalization rather than a new equilibrium.

The claim of those who point to substantial liberalization is not one of institutional isomorphism, or a flattening of the landscape of industrial relations, converging upon a single set of liberal market institutions–though labor market deregulation, decentralization of bargaining and declining union density are all but universal. It is instead that institutions are surprisingly plastic, their function shaped by their context, so that the surface appearance of quite different institutions may mask a similar effect, such as when coordinated wage bargaining in Sweden is transformed from promoting solidarism to promoting firm-level flexibility, or concertation in Italy becomes a mechanism for legitimizing austerity rather than political exchange, or works councils in France come to substitute for trade unions rather than support them, or awards in Australia become minimum rights enforced by state agencies rather than mechanisms for extending trade union gains throughout the economy. The argument here is that a fundamental shift in power resources and labor regulation, one that has gone hand in hand with a transformation in capitalist growth models, has largely overwhelmed the stickiness and path-dependent effects of industrial relations institutions; it is as if they have melted enough to become malleable, subject to be re-purposed in ways that enhance employer discretion in the workplace rather than constrain it.

For those scholars for whom the liberalization of industrial relations is the central fact of the last three decades, and the starting point for a reconstruction of industrial relations theory that reconsiders the stability of institutions, the preference ordering of employers, the articulation of power resources, and the role of political regulation, it is the common, indeed universal trajectory of industrial relations institutions that stands out, not continued evidence of national or sectoral difference (Gumbrell-McCormick and Hyman, 2013, and Meardi, 2018, notwithstanding). The point is that in all cases labor is weaker, in all cases employers have more control/discretion over the workplace, and in all cases evidence that it is being reversed is far more conjunctural and weaker than that which it is being accelerated; indeed, since 2008 public sector industrial relations, once relatively insulated from liberalizing trends, have become greater targets with state budget deficits used to justify the attacks, and the sovereign debt crisis has also served to justify externally enforced liberalization of the industrial relations institutions in several southern European members of the European Union (Marginson, 2015). In short, it is simply no longer terribly interesting to identify and emphasize difference, and doing so offers few lessons for challenging the impact of the neoliberal project on class relations.

Back to capitalism

Debates about the extent of liberalization of industrial relations have gone hand in hand with what we might term a turn (or, more accurately, a return) to capitalism in the scholarship of comparative political economy. These two developments are not a coincidence. If the stability and resilience of national models of capitalism are less and less plausible empirical claims, it follows that our theories of capitalist growth models may also require revision. Into this theoretical space have stepped both new scholarship on growth models and a revival of Regulation Theory, which has existed since the late 1970s as an occasional rebuke to the literature on varieties of capitalism in its emphasis upon temporal shifts in the regime of capitalist accumulation. What these old and new approaches share is a recognition that the growth dynamics of contemporary capitalism are being transformed, removing the conceptual basis for an analysis of varieties of institutional form that are predicated upon stable growth models.

This transformation is not well captured by institutional analysis; there is a terraforming quality to the manner in which change in the dynamics of capitalist growth impacts institutions, whether through changing the form or the function of industrial relations institutions. To be explicit about this relationship between the stability of capitalist growth models, class power and industrial relations institutions: while institutions create a stickiness that prevents conjunctural changes in the balance of class power that result from changing growth models from translating into immediate changes in outcomes (both an expansion of employer discretion and greater instability in a growth model), those institutions cannot substitute for a particular balance of class power in the long term. It may be that mid-range analysis is just not that useful for this task, at this point.

More important then is to link these two sets of theoretical literature that have too often been considered separately (Baccaro and Howell, 2017b): first, the literature on varieties of capitalism and liberalization where, as discussed earlier, a debate has taken place over roughly the last decade about the extent to which distinct national varieties of capitalism continue to survive in the face of powerful economic forces pushing towards institutional convergence, and a reshaping of the power resources available to class actors which has reduced the capacity of organized labor to resist the liberalization of institutions, particularly those related to industrial relations and the labor market; and second, the literature around growth models, or what the Regulation School labels regimes of accumulation. In the initial impulse for the Varieties of Capitalism approach, institutional sets were organized around distinct production regimes, though only two were enumerated. The notion of production regimes focused heavily on the supply-side of the economy, that is, on the institutional conditions allowing key firms to specialize in particular types of innovation (incremental vs radical innovation) and in particular sectors and markets (e.g. Soskice, 1999), neglecting the level and composition of aggregate demand. Over time, even the link of institutions to production regimes faded from most work within the VofC tradition as the focus shifted to the institutions themselves. Linking these two literatures permits a return to an older argument about the relationship between growth models and the institutional sets that regulate them and create the conditions for their stability or crisis.

The last 5 years or so have seen the emergence of a number of heterodox approaches to political economy, all emphasizing the dynamics of capitalism and transformation of capitalism over time, through a succession of capitalist regimes or growth models. They have been the product in part of an evolving critique of the assumptions of the Varieties of Capitalism approach, in part in response to evidence of longstanding and widespread liberalization among all advanced capitalist political economies (not just LMEs), and in part the crisis of the neoliberal, financialized model of growth after 2008. These new approaches to political economy include a ‘neo-Kaleckian’ growth model approach developed by Baccaro and Pontusson (2016), Blyth and Matthijs's notion of growth regimes (2017), Streeck's theory of sequential fixes (2014), and a renewed interest in Regulation Theory (Boyer, 2015).

Regulation Theory provides the link between the two traditions of critical political economy (Aglietta, 2001; Boyer, 1990, 2015). Its two core assumptions – that capitalist economies are inherently unstable and prone to crisis, as left to their own devices they do not spontaneously generate growth and achieve equilibrium outcomes; and that capitalism periodically transforms its growth model, seeking to achieve growth from different sources – generate what is essentially a punctuated equilibrium model in which growth can be stabilized by regulation – in the form primarily of institutions, such as those in industrial relations, financial systems, and welfare regimes – for a period before exhausting that form of growth and lapsing into crisis, at which point efforts to find a new form of growth and construct regulatory institutions appropriate to it begin (Jessop, 1990; Neilson, 2012). The simultaneous focus upon the conflictual dynamics of capitalist growth and the regulatory mechanisms – institutional or otherwise – that seek to stabilize it make Regulation Theory a useful point of theoretical contact between the labor process tradition within comparative employment relations and a more critical comparative political economy (Baccaro and Howell, 2017b; Vidal, 2011, 2014; Vidal and Hauptmeier, 2014).

What these approaches, old and new, have in common is an understanding of capitalism that is more conflictual, anarchic and subject to crisis, and simultaneously less subject to assumptions of stable equilibria, path dependence, peaceful coordination, and pre-strategic rational choice interests which predicted that class actors would defend existing political-economic institutions, all of which were central to understanding the logic of resilient national models of capitalism, and therefore the institutional sets that regulate them.

This must influence our understanding of how and along what trajectory industrial relations institutions and systems change because the liberalization of industrial relations is part and parcel of a general crisis of the Fordist wage-led growth regime in advanced countries and of the uncertain search for alternative growth models that has ensued (Boyer, 2015; Lavoie and Stockhammer, 2013). The growth models that have replaced the Fordist/wage-led one are all inherently unstable and crisis-ridden because, unlike the Fordist growth model, they lack a well-functioning institutional mechanism ensuring that aggregate demand grows in tandem with aggregate supply. The collapse of the collective industrial relations institutions which stabilized Fordist growth has meant that the growth models which replaced Fordism are chronically unstable and crisis-ridden. The result has been dysfunctional growth models or accumulation regimes (Vidal, 2014).

This is not the place for an examination of the growth models that have replaced Fordism, but what they share is the absence of wage-led growth, or rather its diminished role, and its replacement by alternative sources of demand, of which debt-led and export-led growth are the most developed. Baccaro and Pontusson (2016) have outlined the characteristics of these models in more detail, but they are consistent with a range of other contributions to an emerging heterodox and Marxian-inspired political economy, including Harvey's work on flexible accumulation (1989) and Durand's on financialization (2017).

Only two brief implications of these forms of growth for industrial relations institutions will be noted here. The first is a negative implication: neither debt-led nor export-led growth requires institutions of collective employment relations, or institutional linkages between productivity and consumption (Baccaro and Howell, 2017b); to the extent that they survive it is because of the power resources of organized labor, not an accumulation imperative, and states are likely to feel less pressure to maintain them. The second implication is that a strong case can be made that both of the more recent forms of growth encourage a liberalization of employment relations. Export-led growth turns the old Fordist logic on its head and represses domestic wage growth in order to spur export growth, an effect that can be negotiated but can also be achieved through a weakening of collective institutions of employment relations. Debt-led growth, and the financialization that drives it and is driven by it, encourage what Umney et al. have called a return to ‘class discipline’ through state policies of wage restraint and welfare reform that have ‘subverted and marginalized regulatory labour market institutions’ (2017: 1). Similarly, Westcott et al. (forthcoming) have pointed to the manner in which financialization has been an important driver in the growth of employer strategies that shift risk from business to labor and contribute to the growth of low-paid and precarious work that is often insulated from regulation by either states or trade unions, a process nicely captured by Weil's account of the ‘fissured workplace’ (2017).

What the experience of Fordism suggests for the current period is that the logic of a growth model can leave latitude for multiple forms of industrial relations institutions. The logic of wage-led growth could be achieved through industrial or pattern bargaining as in the United States and Germany, corporatist wage solidarism as in Sweden, state substitution for collective wage-setting as in France, or the extension and generalization of trade unions gains through an awards system as in Australia. Different sets of national industrial relations institutions were capable of producing similar outcomes. In the neoliberal era, the trajectory is in all cases towards greater employer discretion, towards ‘flexibility’, but this too can be achieved in different ways, through industrial relations institutions that are shaped by particular historical legacies, configurations of class power, and state roles. To the latter, this review now turns.

Bringing the state back in (again)

In the 1970s theoretical analysis of the state as an actor was inseparable from discussion of the dynamics and regulation of capitalist economies. Influential debates within Marxism sought to explain the relationship between capitalism and the state (Miliband, 1969; Poulantzas, 1978), a revival of the Weberian tradition contemplated a more autonomous and purposeful state willing and able to drive economic development (Evans et al., 1985), comparative political economy looked to state policy to explain different responses to economic crisis (Katzenstein, 1978), and British industrial relations scholars, faced with a set of ambitious state projects of labor law reform, corrected their longstanding pluralism by incorporating the state as an actor (Clegg, 1979). Hyman connected the insights of Marxist analyses of the state to industrial relations in his seminal text (1975).

But slowly and unevenly, certainly with plenty of exceptions, analyses that centered the role of the state receded from view in the course of the 1990s and early 2000s. Optimism about a return to market-led economic growth and pessimism about the capacity of states to manage national economies increased the influence of neoliberal discourse on the Right and a globalization Third Way narrative on the Left. The field of comparative political economy made a number of theoretical moves, including moving away from power resource theory toward a rational choice approach to the coordination strategies of economic actors, and a delinking of markets from capitalist development, culminating in the Varieties of Capitalism theoretical framework, which had implications for how to think about conflict, change, and stability, and hence the need for political regulation. As noted above, industrial relations scholars were more resistant to these shifts, and there have remained influential accounts of the role of the state in employment relations (Clark, 2000; Hyman, 2008; MacKenzie and Lucio, 2014; Meardi, 2014, 2018; Meardi et al., 2016). Nonetheless, comparative industrial relations scholarship has tended to be better at empirics than theorizing; its multidisciplinarity has made it hard to develop its own mid-range theories, which has in turn encouraged a reliance on the latest theoretical innovations within comparative political economy, which includes a more peripheral role for the state in the regulation of class relations; and even critical labor process scholars, for whom there is little debate about the continued centrality of conflict in the workplace, have rarely looked to the state to explain its regulation.

Yet the state has never receded from the regulation of class relations, and only rarely has become secondary to regulation by class actors. Indeed a strong case can be made that states have become more interventionist in industrial relations in the neoliberal era, not less, with major projects of reform multiplying across the advanced capitalist world over the last 30 years (Howell, 2017). There is some irony that the heyday of collective laissez faire in industrial relations went hand in hand with the Fordist era of political regulation while the anticipated retreat of the state in the neoliberal period has seen ever more ambitious state reform efforts. This review is not the place for a full reconstruction of the theoretical and empirical bases for an argument about the state, neoliberalism, and industrial relations (for some initial formulation, see Howell, 2016, 2018), but in light of the two previous sections of this review, which pointed to widespread liberalization of industrial relations institutions and a resurgence of interest in capitalist growth models, some speculative conclusions are possible.

What these developments highlight are two essential starting points for any theorization of the role of the state in the regulation of industrial relations: first, that the inherent instability, perpetual dynamism and deeply conflictual nature of capitalist growth make political regulation essential; and second, that moments of liberalization, of market-making in the Polanyian sense, when resistance to (re)commodification is greatest and existing regulatory institutions are in need of reconstruction, are also precisely those moments when the state is most necessary. Indeed, the scale and scope of the liberalization of industrial relations over the last three decades could only have taken place with a more active role for the state. What follows is space for the development of a range of state strategies of labor regulation, all with the dual goal of stabilizing capitalist growth and permitting liberalization, but each emerging out of the specific conditions, historical legacies and configurations of class relations of its time and place.

Debates about contemporary post-Fordist capitalist growth models open up space for rethinking and re-centering the role of the state in the regulation of capitalist political economies, and particularly industrial relations. This is in part because of the distinctive instability of financialized capitalism (Jessop, 2014; Vidal, 2014), and in part because much of this recent literature focuses upon the demand-side of growth – in contrast to the supply-side emphasis of the Varieties of Capitalism approach – for which the need for extra-market regulation is far more persuasive. It is striking that a common feature of these more heterodox approaches to political economy is that they all acknowledge the need for political regulation, and therefore for a permanent role for the state. Baccaro and Pontusson (2016), for example, identify a state role as essential to reproducing both debt-led and export-led growth models through financial deregulation and wage repression, and even where wage-led growth remains important, it increasingly comes courtesy of the public sector. Boyer has also moved the state to a more central role in Regulation Theory, arguing that the state is ‘a vector of class compromise’ in assembling any mode of regulation (2015: 43).

The argument for the central role of states in the liberalization of industrial relations has two related components. The first is that states have a set of unique capacities, not available to private class or non-class actors, when it comes to institution-building. As Fordist growth first went into crisis in the early 1970s, showed signs of exhaustion by the early 1980s, and then gave way to non-Fordist growth, its attendant regulatory labor market and industrial relations institutions proved increasingly dysfunctional. The process of reconstructing those institutions required the state, in part because private actors are often timid, concerned with short-term interests and reluctant to abandon existing institutions, and in part because states have capacities–particularly related to legitimation and the generalizing and embedding of institutions–that employers and trade unions do not. These include the coercive powers of the state, the ability to solve collective action problems for class actors, permitting derogation from labor law in return for specific behavior, a privileged position in the narration of crisis and hence the range of politically viable responses (Hay, 1996), and the legitimation of new industrial actors (Howell, 2009). In practice, even the most self-confident, organized employers are unable to reconstruct an industrial relations system alone, even if they are able to undermine an old one.

The second component is a recognition of Karl Polanyi's essential insight into the political construction of markets (1944), and its application to the present. Markets do not emerge spontaneously, and certainly not labor markets. The commodification of ‘fictitious’ commodities is always likely to bring resistance as the removal of forms of insulation from the labor market – whether once provided through collective bargaining, labor law or social rights – exposes those who have to sell their labor to higher levels of risk, and that in turn encourages a ‘counter movement’ on the part of workers to protect themselves. It is why Polanyi anticipated that the state would have to take a central role in market-making, and why Gamble argued that the construction of the free economy requires a strong state (1988).

In the first half of the 19th century, that state role was about constructing a labor market in the first place, and in so doing overturning laws that limited labor mobility and wage labor itself, and overcoming expectations of access to poor relief and common land. That alone required an illiberal state. 150 years later, the liberalization of labor markets, dismantling collective industrial relations institutions, and expanding employer discretion involve overcoming the crystallized form of more than 30 years of broadly social democratic expectations, the politicization of the political economy (in the sense in which the long postwar boom undermined the notion of a natural economy and replaced it with the expectation that economic outcomes were the result of political choices), and myriad forms of working-class power. If liberalization means an expansion of employer discretion, that in turn requires a shift in the balance of class power; it is not an abstract exercise in simply setting the market free. And because class power is at stake, it is likely to require a more active, more interventionist use of state power to bring it about. As such, evidence of continued state regulation, intervention and management of the labor market and industrial relations tells us little about the extent of liberalization (contra Meardi, 2018).

It is here that a conception of neoliberalism that is materialist and historical, and that situates neoliberalism within the history of capitalism and its transformations, is useful. Neoliberalism can be thought of as the political, institutional and ideational infrastructure of contemporary capitalism, or, in Regulationist language, as the mode of regulation of the post-Fordist regime of accumulation. Cahill and Konings argue that neoliberalism is ‘authored by the state’ (2017). Unlike classical liberalism, neoliberalism is not about limiting state intervention, but rather about using state power in new ways to institutionalize a market order. It is a response to both the crisis and collapse of the Fordist growth model, and the failure of classical liberalism to legitimate a market society, a failure that was as much political and ideational as economic (Dardot and Laval, 2014; Peck, 2010).

One can distinguish two distinct elements of this neoliberal mode of regulation: first, neoliberalism in support of a regime of accumulation, or a family of growth models, which require a set of institutions to stabilize growth and permit markets to function; and second, neoliberalism as a constructivist project of reshaping society to allow a market order to function. As Cahill and Konings put it, neoliberalism recognizes that a market order ‘needs to be actively constructed, institutionally and politically’ (2017: 14). In this regard, neoliberalism should be understood as a political project to re-shape society so as to make a market order possible, indeed to institutionalize it. And given the obstacles to that project, the state will be a central actor.

Understanding why states have become more interventionist in the regulation of class relations in the service of expanding employer discretion does not tell us how they have intervened, or why the form of intervention differs from one time period and country to another. There remain varieties of industrial relations, and varieties of state intervention. It might be useful to think in terms of state strategies of labor regulation, though strategy may imply greater intentionality and coherence than is often the case. These are the forms of state policy for regulating industrial relations, specific to time and space, different by virtue of different institutional legacies, different configurations of class power, and different variants of post-Fordist growth models. Thelen has identified pathways to liberalization (2014), and while one need not adopt her typology of pathways or her explanation for their emergence, it is a fruitful approach. Here we can think of variations in the strategies of labor regulation that states adopt in order to bring about liberalization.

There are many such strategies of labor regulation, of which the most obvious, in terms of its direct assault upon organized labor, is probably the decollectivizing strategy utilized by the Thatcher government, where a combination of labor law reform, privatization and a willingness to fight strikes successfully undermined the deeply entrenched workplace power of British unions. Three additional strategies of labor regulation can also be briefly outlined to illustrate the point. The first is a type specific to center-left parties in political economies where collective industrial relations institutions have largely collapsed. Here the goal is to offer some protection for workers in the absence of trade unions capable of doing do, but to do so in such a way as to still permit the liberalization of labor markets. This involves the substitution of minimal individual rights at work, enforced through state agencies, for collective rights. One might term this a Third Way (Crouch 2001) strategy of labor regulation because it was exemplified by the New Labour government in Britain after 1997, but it can also be seen in the Australian Labor Party's Fair Work legislation and elements of the lois Auroux in France.

A second state strategy of labor regulation has been the use of social pacts with the goal of facilitating austerity and the liberalization of labor markets and industrial relations. Dating from the mid-1980s and often tied to the need to meet criteria for Economic and Monetary Union, this is concertation of a different kind from the heyday of corporatism, often in countries other than those traditionally associated with social democracy and corporatism in the Fordist era, such as Ireland and Italy. What these more recent forms of concertation have in common is an effort on the part of states to use peak-level bargaining in order to gain acquiescence to neoliberal macroeconomic and social policies and to overcome entrenched labor strength at the firm level or within parts of the public sector.

A third state strategy of labor regulation is specific to countries where organized labor is stronger, or there is a fear that a frontal attack upon collective industrial relations institutions of the Thatcherite kind will produce unacceptable political or industrial backlash. Here a strategy of derogation, where formal legal or contractual constraints remain but states permit industrial actors to bypass or ignore them, can be employed. As such, it tends to require a more active role on the part of the state. An industrial relations system in which actors are allowed to ignore institutional rules with impunity is de facto subject to deregulation. Derogation permits liberalization of industrial relations without having to formally end or replace existing institutions. It was often more palatable than a frontal attack on institutions and could be justified as an emergency measure, or as institutional change under carefully controlled conditions. It can be thought of as a form of ‘neoliberalism as exception’ (Ong, 2006). It was widely used in French industrial relations legislation from the mid-1990s onwards, most fully in the 2004 loi Fillon, but also in Germany, Sweden, and Italy in the past two decades.

Conclusion

This review has argued that the generalized trajectory towards the liberalization of industrial relations and the expansion of employer discretion over work, workers and the workplace poses questions for industrial relations theory. Drawing on debates within comparative political economy and comparative industrial relations, I have suggested that scholars need to rethink the role of industrial relations institutions, specifically their ability to shape, mediate, and mitigate the impact of broad transformations in types of economic growth and the balance of class power. Further, recent work to develop more heterodox approaches to the study of capitalist growth models has important implications for understanding what has replaced the Fordist wage-setting institutions inherited from the past, but also for explaining the chronic instability of these newer post-Fordist growth models. And both the near-universal liberalization of industrial relations and the instability of emerging capitalist growth models highlight the centrality of political regulation and the role of the state in market-making and institution-building, particularly with regard to industrial relations. The review then argued that refocusing attention on the state permits the articulation of what we can term state strategies of labor regulation that seek to stabilize growth and permit liberalization under conditions specific to time and place.

These developments within industrial relations theory may also help to make sense of growing interest in recent years, often from the most unlikely of sources including central bankers (Lowe, 2017) and international economic organizations (Organisation for Economic Co-operation and Development (OECD), 2018), with the dangers of wage stagnation, or to use Summers’ term ‘secular stagnation’ (2014). In light of the failure of post-Fordist growth models to find stable mechanisms for responding to demand constraints, this public interest should be unsurprising. It also suggests that state responses, should they be forthcoming, will have to take the form of new strategies of political regulation which pay attention to the linkages between industrial relations institutions and capitalist growth models.