Review of Keynesian Economics

Workplace relations, unemployment and finance-dominated capitalism

Gary Slater * and David A. Spencer *

Keywords: unemployment, labour productivity, full employment, financialization


This paper considers the interconnections between the nature and organization of work and the level of unemployment. We consider how actions taken at the workplace level can impede as well as facilitate the reduction in unemployment. We also consider how the workplace may be reformed to overcome some of the obstacles, economic as well as political, to full employment. Finally, we examine the impacts of the current finance-dominated capitalism on work organization and workers. Our view is that financialization represents a major barrier to full employment, not only because of its tendency to limit real investment but also because of its negative effects on the bargaining power of workers.

Full Text


Relations at the workplace level have important and far-reaching impacts at the macroeconomic level. The choice of technology within workplaces, on the one hand, can lead to more general shifts in the level and distribution of employment. The polarization of the labour market and the rise of low and high quality jobs, for example, have been linked to technological shifts within and across workplaces. The management of labour, on the other hand, can impact on labour market outcomes. Where employers rely on the threat of unemployment to discipline workers, then this can provide a powerful tendency against the achievement of full employment. Where employers use alternative methods of labour management that do not depend on the threat of unemployment, the scope for achieving full employment may be considered to be much greater.

In this paper, we consider the interconnections between the nature and organization of work and the level of unemployment. We consider how actions taken at the workplace level can impede as well as facilitate policy action aimed at reducing unemployment. We also consider how the workplace reform is central to overcoming some of the obstacles, economic as well as political, to full employment. Finally, we examine the impacts of the current finance-dominated capitalism on work organization. Our view is that financialization represents a major barrier to full employment, not only because of its tendency to limit real investment but also because of its negative effects on the bargaining power of workers.


The link between unemployment and labour productivity has been explored from different vantage points in economics. In this section, we focus on three different accounts from mainstream economics, post-Keynesian economics and Marxian political economy. We draw out their underlying assumptions and ideas, especially in relation to the nature and extent of workplace conflict. We also trace their implications for policymaking.

2.1 Mainstream economics: the ‘shirking’ barrier to full employment

We start with an account of the unemployment–labour productivity relation found in mainstream economics. This account is associated with efficiency wage theory that has become a key aspect of New Keynesian macroeconomics as well as mainstream microeconomic accounts of workplace relations. It can be noted here that efficiency wage theory has been instrumental in opening up mainstream economics to an analysis of the internal organization of work. In the past, as is well-known, mainstream economics consigned the internal affairs of the workplace to a black box. How work was organized and how work was extracted from workers were neglected in mainstream economics. Efficiency wage theory has helped to remedy this longstanding neglect; specifically, it has shown how workplace relations play a key role in determining macroeconomic outcomes, most notably the existence and persistence of involuntary unemployment.

Efficiency wage theory comes in several different forms (see Akerlof and Yellen 1986); here, we focus on one particular example, namely the so-called ‘shirking’ model of Shapiro and Stiglitz (1984). This model takes its name from the fact that it assumes that workers will ‘shirk’ (that is, expend zero effort) if they do not feel threatened by the prospect of job loss. The threat of job loss is assumed to vary positively with the financial penalty incurred by workers in the event of their dismissal for being caught underperforming at work. Employers are then assumed to face incentives to raise wages above outside opportunities to increase the cost of job loss. If all employers act in this way, then wages will be bid-up to a level where there is an excess supply of labour. This excess supply of labour creates the deterrent that employers need to prevent shirking by workers. 1 Although unemployed workers will be ready and willing to work for lower wages than those currently paid to incumbent workers, they will not be hired by employers because their reemployment will eliminate unemployment and hence the cost of job loss. Those workers who face unemployment, in this case, are seen as involuntarily unemployed: they are ready and willing to work, but are not hired because their employment would result in lower labour productivity. The ‘efficiency wage’ aspect arises from the fact that employers are able to reduce unit labour cost by paying higher wages: while it costs employers more to hire workers, this cost can be more than recouped via higher labour productivity from a compliant workforce.

We need to be clear about how the above arguments are arrived at. They are based on several assumptions, some of which are controversial and open to challenge. The most contentious assumption is that workers are ‘shirkers’. It is assumed that workers are born to resist work and will only work if incentivized to do so. This idea is applied indiscriminately, as if all workers are alike in their resistance to work.

The problem, of course, is that the preferences for work of workers are not fixed; rather they are influenced by the social and institutional context of work. If workers resist work, this is likely to reflect on the harsh conditions they face at work. Such resistance may also be a response to the exploitation of employers. The focus on the alleged shirking proclivities of individual workers tends to mask these other sources of resistance to work; indeed, it tends to paint workers as the reason for conflict at work, thereby detracting from the potentially exploitative behaviour of employers.

The argument about involuntary unemployment being a consequence of the shirking behaviour of individual workers gives the impression that full employment is denied because of a problem inherent in human nature (or at least the nature of workers). It appears that full employment is an unachievable goal, owing to the natural tendency for workers to shirk when involuntary unemployment is zero. Society, it seems, must resign itself to involuntary unemployment as a worker disciplinary device. Involuntary unemployment, it is argued, is necessary and indeed desirable to achieve high labour productivity.

The shirking model ultimately offers a counsel of despair. Invoking the assumption of universal shirking by workers, it presents full employment as an unrealizable policy goal. Ideologically, the model becomes a tool for justifying the existence and persistence of involuntary unemployment.

2.2 Post-Keynesian economics: Kalecki and the political obstacles to full employment

In post-Keynesian economics, the emphasis on deficient aggregate demand as a barrier to full employment is well-known. Keynes emphasized that, without an adequate level of aggregate demand, capitalist economies would languish in a state of involuntary unemployment. It was the job of the State to run budget deficits and to reflate the economy where the ‘animal spirits’ of the private sector were low. Yet, there are other aspects to the post-Keynesian account of full employment. One aspect of interest to us is the emphasis given to the political obstacles to full employment. This aspect is more associated with Kalecki than Keynes. Indeed, the link to Kalecki means going beyond Keynes, in that the explanation for the obstacles to full employment requires consideration of the class politics of capitalism. Following Kalecki, it can be argued that even if aggregate demand policies could gain the support of the State, their implementation would be resisted by capitalists out of a fear of their adverse effects on the balance of power in the workplace. Kalecki's key contribution was to show how full employment policies would be blocked by capitalists due to their desire to maintain their power over workers in the workplace.

Kalecki's arguments on the political obstacles to full employment are laid out in an article he wrote in 1943. He argued that capitalists favoured unemployment because it helped them to keep ‘discipline in the factories’. The threat of unemployment made incumbent workers more unwilling to resist the demands of capitalists. It not only suppressed calls for higher wages but also quelled resistance to long work hours and high work intensity. Unemployment, Kalecki argued, suited the economic as well as political interests of the capitalist class.

Capitalists would clearly benefit from the achievement of full employment. The latter was associated not only with buoyant demand but also with higher profits that could be used to fund further investment and growth. But the negative political implications of the maintenance of full employment presented too much of a risk to the capitalist class. The prospect of workers gaining greater power and using this power to challenge the prevailing social order made the capitalist class resist the achievement and then maintenance of full employment. In practical terms, this meant that capitalists would press the State not to enact aggregate demand policies that would help to achieve full employment. It also meant that if full employment happened to be achieved, capitalists would press the State to deflate the economy in order to increase unemployment. 2

Kalecki's argument is profound in the sense that it implies that full employment will be prevented by the force of class politics. Thus, even though full employment may be possible and achievable in practice, it will be rendered impossible by the interventions of the capitalist class whose interests are better achieved by the maintenance of mass unemployment. Needless to say, we are presented here with a very different vision of the world to that found in efficiency wage theory. The problem of persistent unemployment is linked not to the shirking behaviour of individual workers but to the influence of naked class power. The thwarting of full employment at root is about capitalists wanting to maintain their power over workers, not some problem of natural slothfulness on the part of workers. Achieving and maintaining full employment, in this case, requires that we consider how society is governed – more specifically, it challenges us to think of new institutional arrangements that are able to broker a class accord between capital and labour without the back-up threat of unemployment. Kalecki's message is that, without ‘fundamental reform’ to tackle entrenched class power, capitalist society is doomed to suffer unemployment. What form such reform might take is addressed below.

2.3 Marxian political economy: the ‘reserve army of labour’

The ideas of Kalecki on unemployment were shaped by those of Marx, particularly his notion of a ‘reserve army of labour’. According to this notion, a surplus or reserve army of unemployed workers is required by capitalists to moderate and curb the desires and pretensions of the active or employed workforce. Marx argued that unemployment was functional to capitalism; without the threat of unemployment, capitalists would not be able to police the workforce in ways that meet their class interests.

For Marx, unemployment was a result of capital accumulation in general and of the centralization and mechanization of production in particular. The uneven nature of economic development over time and across industries, together with the emergence of new branches of production, led to a fluctuating reserve army of unemployed workers. But unemployment remained an ever-present feature of capitalism, enabling the expansion of production as well as the emergence of new industries. Unemployment, in essence, was functional to capitalism; it provided the means necessary for capitalists to reorganize and restructure production in ways that generate surplus value.

A key reason why unemployment existed and persisted under capitalism, according to Marx, was because of developments in technology. Marx argued that capitalists would use technology to extract more surplus value from workers and the outcome of this would be job losses. Technologically induced unemployment then added to the pressure on the employed to expend more unpaid labour time. In this sense, Marx viewed unemployment as both a systemic outcome of the capitalist mode of production and a necessary foundation for it; it was both the ‘product’ and ‘lever’ of capital accumulation (Marx 1976, p. 784).

The resistance of workers to low wages, shorter working hours, and better working conditions had to be overcome by the threat of unemployment. Marx saw that capitalists were in conflict with workers and they would use the reserve army of the unemployed to tip the balance of power in their favour. As we saw with Kalecki above, Marx believed that capitalist class conflict resulted in unemployment. But, unlike Kalecki, Marx was pessimistic that much could be done to reform capitalism in ways that could facilitate the achievement of full employment. Any ‘fundamental reform’ under capitalism was doomed to failure. Unemployment was an endemic problem under capitalism and the idea of full employment capitalism was an oxymoron. Only by ridding society of class conflict and thus going beyond capitalism could we envisage achieving an outcome where everyone could be fully and meaningfully employed.


The previous section highlighted the nexus between workplace relations and unemployment. The management of conflict at work, in particular, can lead to the maintenance and reproduction of unemployment at the aggregate level and can be an active force preventing the attainment of full employment. This raises the issue of how work is organized and whether work can be reorganized to reduce the necessity for unemployment as a threat to workers. Can workplaces be reorganized in a way that allows for lower unemployment without lower labour productivity? This nexus between unemployment and labour productivity has not been well explored in the economics literature; much more attention has been given to the unemployment–inflation nexus. One contribution of this paper is to highlight the need for further exploration of the unemployment–labour productivity nexus. Such exploration can help in addressing the broader workplace requirements for sustainable full employment, a topic we address below.

Taking efficiency wage theory first, it gives the impression that policies aimed at reforming work are limited in scope. Such reforms are effectively reduced to better monitoring systems and stronger incentive schemes. Because all workers are assumed to be shirkers, stress is placed on limiting the discretionary power of workers and on incentivizing workers not to slack off at work. The suggestion that reforms can be used to increase the discretion and autonomy of workers over work is dismissed on the basis that such reforms will encourage workers to shirk. We are presented with a vision of the world in which the use of ‘carrot and stick’ is the only option for achieving high labour productivity.

Efficiency wage theory effectively asks us to accept the existence and persistence of involuntary unemployment as the price for combatting shirking by individual workers. Note it blames workers for unemployment, not employers: unemployment is at root a problem linked to the alleged indolence of individual workers. In turn, it blames workers for preventing the reform of work in ways that could potentially improve the quality of work. Workers must accept tightly controlled work and the imposition of hard incentives – two features of low-quality work – as the price for their own indolence.

But what if the effects of unemployment on labour productivity are not always positive? What if unemployment can limit labour productivity below potential levels? There are dangers in an excessive focus on the functional role of unemployment as a threat to spur worker effort. There may well be some relationship, but this is too crude a reading of how work motivation is managed and determined in the real world. The factors that generate high labour productivity in actual workplaces are more complex than abstract carrot and stick models would suggest, and excessive use of such threat tactics may actually undermine labour productivity.

The shirking model of efficiency wages and involuntary unemployment discussed above (Shapiro and Stiglitz 1984) assumes that unemployment is an effective device for disciplining workers, given the assumption that workers possess a universal proclivity to shirk. Such an approach takes a one-sided view of production and of human nature, leading to a failure to consider the down-sides of the threat effect in terms of reduced labour productivity. The possibility of job loss may well assist firms in gaining the compliance of workers, but what about the ingenuity and creativity of workers? Contrary to the economist's standard view of production, the release and harnessing of workers' ingenuity and creativity plays a crucial role in determining labour productivity (see, for example, Burawoy 1979; Edwards 1990).

There is a crucial difference to be drawn here between ‘consent’ and ‘cooperation’. Consent to work entails an acceptance of the authority of employers and compliance to any set rules and procedures. Both can be secured by issuing threats of dismissal. Cooperation, by contrast, entails a deeper source of commitment to work and a willingness to ‘go the extra mile’ for employers. It means that workers seek out better ways of doing their work; such experimentation can aid employers in revealing better ways of doing work. It can also mean workers coming forward with new ideas about how to change work in ways that can enhance labour productivity and allowing workers to use their initiative to find the best way. Where workers are encouraged to use their voice at work, labour productivity can be improved. Much better in this case that workers use their ‘voice’ than show their frustration and discontent with work by exercising their right to ‘exit’ their current jobs. The point here would be that while unemployment helps to secure the consent of workers it does not help – and indeed hinders – the elicitation of their cooperation. An unintended and hidden cost of higher unemployment is lower cooperation that can translate into lower overall labour productivity. The loss of labour productivity may occur where the threat of unemployment by employers is perceived by workers as a breach of trust and a loss of confidence in workers' previous contributions. In short, far from helping to achieve high labour productivity, high unemployment may well thwart its achievement.

The grain of truth in efficiency wage theory is that the threat of unemployment can be and often is used by employers to discipline workers. There is some evidence of a link between unemployment and workplace performance. An early study of the efficiency wage model in the UK by Wadhwani and Wall (1991), for example, found some evidence of a threat effect: higher unemployment raised workplace productivity. More recently, Lazear et al. (2013) have explored the impact of the 2007–2009 recession on worker productivity in the US. Based on a detailed analysis of one national technology-based services firm, they found that worker productivity (as captured by the computers used by workers) rose in the recession as retained employees worked harder. 3

However, the above study focuses only on one firm. Other studies using a broader set of data have shown that the threat effect of unemployment varies according to the nature of the industry and employment relations within it (in line with post-Keynesian and Marxian analysis). From an aggregate, cross-national perspective, Weisskopf (1987) found that the relationship between unemployment and labour productivity varied across economies according to the nature of the industrial relations system in place. In the adversarial systems of the US and UK, a generally positive association was found, in contrast to a negative relationship in the more corporatist systems of Sweden and Germany. Such variations have also been found in within-country studies. Green and Weisskopf (1990) examined differences in the impact of unemployment on work intensity across US manufacturing industries between 1958 and 1981. In over half their sample, they found evidence of a positive and significant ‘worker discipline effect’ from unemployment. Yet, importantly, they also reported significant differences in the nature of this effect across industries, according to their structure and employment practices. In those characterized by large firm size, concentrated product markets, high wages and high union density – all indicators of primary labour market status – the threat effect from unemployment was found to be relatively weak. In contrast, unemployment was found to have the strongest impact in industries dominated by small, rivalrous, labour-intensive firms operating in secondary labour markets. Further evidence that threat effects are more prevalent within secondary labour market firms, where unions are poorly represented, is reported by Green and McIntosh (1998) for Britain and by Rebitzer (1988) for the US.

What the above research shows is that the threat of job loss is not required to secure high labour productivity and that alternative, often unionized, arrangements can provide a substitute to underpin high labour productivity. However, in many workplaces, employers do rely on unemployment, but this would appear to be more for reasons of power – a conscious choice to follow an insecure, intensive secondary labour market route – rather than an efficient response to some universal problem inherent in human nature. That is, it is possible for employers to raise labour productivity by relying on other (non-coercive) means to motivate workers; however, the challenge that the loss of the unemployment threat potentially poses for their power in the workplace leads them to insist on its maintenance, even if this means foregoing productivity gains. 4

The productivity loss here is relative to some other type of work organization that does not rely on the threat of unemployment to raise labour productivity. Here we can think of types of work organization that elicit workers' cooperation by granting them greater autonomy over work and greater participation in decisionmaking. The starting point for this approach is not the image of the ‘lazy worker’ who will use any discretionary power to avoid the exertion and responsibility of work. Rather, it is the idea that workers' effort is endogenous and positively related to the autonomy they have over work and their involvement in the workplace. Workers thus may respond to having more autonomy over work by seeking out innovations in the way they do work; they may also feel more confident to put forward their ideas about how to improve labour productivity. Cooperation, though, will only be forthcoming where the right conditions hold. Conditions where workers and employers are at loggerheads, or in dispute, will obviously not be conducive to the extension of workers' autonomy over work. In such conditions, consent by threat may represent the only option open to employers. But this does not mean that alternative conditions could not be created. Indeed, the argument here would be that such alternative conditions need to be created if unemployment is to lose its status as a worker disciplinary device. Policies for sustainable full employment, in short, must include policies to reform the workplace.

The burgeoning literature on the connections between workplace structures and practices and performance in the fields of industrial relations and human resource management provides evidence of the positive impact that participatory practices such as worker involvement and job autonomy can have on labour productivity (Strauss 1998; Ichniowski et al. 2000; Pfeffer 2007). This literature is suggestive of the type of reforms required to achieve high labour productivity without high unemployment. Yet so-called ‘high performance’ practices remain relatively limited in actual workplaces. This seems puzzling. If such practices can raise labour productivity, why are they not more widespread? Following our previous line of argument, one reason why employers may reject these practices is because of fear of losing control of the workplace. Pfeffer (2007) suggests that this might be the case, when he refers to the distorting influence of financial stakeholders on labour management. These stakeholders have biased investments towards short-term cost-cutting exercises and away from long-term investment in employee involvement schemes. Hallock (2009) has also pointed to the increasing impact of cost control on US firms' employment decisions. Thompson's (2003) ‘disconnected capitalism thesis’, in addition, suggests that short-term pressures from finance prevent the adoption and spread of high-performance management practices. In sum, whilst it appears that more progressive ways of achieving and sustaining labour productivity exist and may be implemented in practice, it is clear that analysis needs to look beyond the level of the workplace to understand the scope for reform. Issues of power, in particular, remain important in understanding the barriers to the reform of work in ways that would help facilitate the pursuit of full employment.


This section considers, in brief, the extent to which developments in modern finance have provided a fetter to improved work conditions and macroeconomic outcomes. We focus in particular on the rise of a financialized capitalism. The process of financialization has transformed the investment strategies of firms. It has also led to shifts in the organization of work that have had specific detrimental consequences for labour, as we will show below.

The term ‘financialization’ originates in radical political economy and is used to describe in a systematic way the dramatic rise of financial activities and financial institutions within economy, society and culture. Financialization has been a secular and global process over the past 30 years or so, recently encompassing the global financial crisis and ensuing period of austerity in capitalist societies. Different meanings are attached to it and this section cannot review all these different meanings (for such a review, see Lapavitsas 2011). Some broad characteristics of financialization, however, can be identified. One key aspect of financialization concerns the enormous increase in the role and importance of finance and financial activities in the economies of mature capitalist nations, particularly the US and UK (Epstein 2005). This increase has not been cost-free, but instead has come at the expense of a decline in the industrial sectors of these nations. The spread and consolidation of speculative behaviour and a short-termism culture has ultimately placed limits on real investment, output and growth. Large corporations, in particular, have undergone a process of financialization as they have looked to engage more in financial activities (for example, use of hedge funds, lending to households). These corporations have also funded investment more by internal sources than by loans from banks. This fact has meant that banks have had to adapt their behaviour, focusing more on lending to households as a means to secure revenue (Lapavitsas 2011). The everyday lives of people, more generally, have become increasingly ‘financialized’ (Martin 2002). A combination of factors – for example, declining or slowly growing real wages, the privatization of public services, and the entrenchment of a consumer culture based on mass advertising – have meant that peoples' lives have become more bound-up in the financial system. Borrowing to fund consumption and investment in financial assets (for example, housing, pensions, insurance) have become a more normal and accepted part of everyday life (Lapavitsas 2011, p. 620). The stagnation of real wages coupled with the acceleration in incomes of financiers has also resulted in rising levels of inequality (Glyn 2006). Financialization, finally and critically, is linked to the global economic crisis that began in 2007: the rise of finance and financial speculation is seen to have created an unsustainable bubble that has burst with devastating effects in the post-2007 era. The process of financialization has been uneven across capitalist nations; it has been most evident in nations such as the US and UK and has been less apparent in other nations like Germany. The recent global economic crisis, however, has meant that all nations have been exposed to the effects of financialization.

For our purposes, the interesting issue concerns the consequences for work and work relations of financialization. Here we attempt to develop and apply some arguments made elsewhere (O'Sullivan 2000; Glyn 2006; Lapavitsas 2011; Thompson 2011). Four issues can be stressed. First, financialization has impacted directly on the employment relationship. Specifically, it has enabled employers to increase their power over workers. As workers have accumulated financial assets and taken on greater amounts of debt, they have become less able and willing to push for higher wages and better working conditions. To the contrary, their weakened economic position has made them more vulnerable to real wage cuts, longer work hours and more precarious forms of employment (Glyn 2006). Higher personal debt ratios increase the vulnerability of those in work and the desperation of those without.

Second, financialization has shifted corporate governance in the direction of a ‘shareholder value’ model. The interests of shareholders have been elevated above those of other stakeholders, including workers. Pressurized by financial markets to maximize short-term profits and to raise dividends for shareholders, managers have looked to reduce wages, lay off workers, and downsize production (Lazonick and O'Sullivan 2000; Froud et al. 2006; Thompson 2011). They have also entered into mergers that invariably have led to reduced levels of wages and employment (Glyn 2006, pp. 64–65). The simultaneous growth of private equity investors that have bought and sold productive assets, in addition, has resulted in poorer employment outcomes for workers (Clark 2009).

The third issue relates to the impact of the global financial crisis on work. The crisis has its origins in the financialization process and among its outcomes has been a decline in employment, job security, wages and working conditions. Already weakened by financialization, workers have faced a further reduction in their bargaining power due to rising unemployment and fiscal consolidation. As mature capitalist economies confront a future of slower growth rates, there is little chance of any quick turnaround in the fortunes of workers. Indeed, the responses to the crisis made at the national and international level by governments and governmental organizations are making life more difficult for workers (Heyes et al. 2012). In the name of ‘austerity’, all kinds of regressive policies have been implemented, from reductions in the availability of out-of-work benefits, through cuts in public services, to the repeal of labour laws. These policies have been pursued with the backing of business and against the demands of organized labour.

Fourth, and relating to the discussion in the previous section, financialization impedes the adoption and sustainability of forms of participatory work practices that rely on the elicitation of commitment and flexibility from workers. Despite their perceived economic benefits, as mentioned above, these practices have not been adopted to any great extent (Godard 2004), in part because they represent too high a cost for firms, but also because they pose a challenge to employers' ‘right to manage’. Where they have been implemented, employers have found it hard to keep their side of the ‘bargain’ (Thompson 2003). Thompson (2003; 2011) has pointed to the fragility and non-sustainability of ‘high performance’ practices under conditions where managers are under pressure from financial stakeholders to maximize shareholder value. Managers demand commitment and flexibility from workers but their drive to maximize shareholder value means that they cannot maintain the conditions required to secure the continued cooperation of workers. Financialization, in this sense, undermines the ability of employers to sustain ‘high performance’ practices and produces tendencies towards harsher forms of human resource management based on coercion. The repeal of labour laws in the era of austerity has further undermined the conditions of security and commitment needed for workplace practices of involvement and commitment.

The process of financialization has clearly weakened the bargaining power of labour. In terms of the discussion of this paper, it has added to the tendency for employers to use short-term measures to control workers and has reduced the scope for the adoption of alternative methods of labour management that rely on eliciting the cooperation of workers. Financialization, in short, has increased rather than diminished the role of unemployment and the threat of job loss as regulators of wages and labour productivity. Its ultimate effect has been to entrench insecurity in the workplace, backed by unemployment in the wider economy.


The relation between unemployment and labour productivity has received less attention from economists than it should. As this paper has argued, this relation is central to understanding whether unemployment is achievable or not. The nihilistic view found in efficiency wage theory is that unemployment is necessary and inevitable to combat the shirking behaviour of workers. On this view, high unemployment is always needed to secure high labour productivity. A more progressive perspective contained in heterodox economics is that high unemployment is just one way to secure high labour productivity. If high unemployment is being relied upon to secure high labour productivity, then this reflects the action of employers and means the foregoing of other methods of achieving high labour productivity that do not entail high unemployment and its associated economic and human costs.

The ideas of Kalecki fit this second view. Only by undergoing ‘fundamental reform’, argued Kalecki, could capitalism accommodate the goal of full employment. ‘Full employment capitalism’, Kalecki (1943, p. 331) wrote, required the development of ‘new social and political institutions which will reflect the increased power of the working class. If capitalism can adjust itself to full employment, a fundamental reform will have been incorporated in it’. Capitalism, in essence, had to find a way to accommodate the competing claims of capital and labour without the necessity for unemployment.

Kalecki, while vague on the details, hinted at several reforms that could help to achieve and maintain full employment under capitalism. These included a coordinated system of wage bargaining based on the establishment of consensus over the distribution of income between wages and profits: this had the particular advantage of accommodating inflationary pressures at full employment. Other reforms included the reorganization of work, including the introduction of some form of worker democracy. By creating a more democratic system of work, workers and employers could reach a compromise over the distribution of income; they could also agree on ways to secure high labour productivity without the back-up threat of unemployment. ‘Discipline in the factories’ thus could be maintained at full employment under a reformed and more democratic work regime.

Innovations in worker involvement have shown that there are different routes to sustaining workplace productivity. However, this paper has also indicated how broad the reform agenda needs to be to achieve Kalecki's aim. Not only does the industrial relations system need reform, so too does the system of employment law and, crucially, the relations between finance and industry. The present dominant pressures of financialization – undimmed despite the global financial crisis – are a fundamental block to any reform agenda. With a globalized, finance-dominated capitalism, this task is all the harder. It does not, however, make it any less urgent. Whether, though, capitalism can accommodate full employment remains to be seen. In this sense, Marx may be right once again.

  • 1

    Efficiency wage theory sees unemployment as a supply-side problem and retains the same labour supply and labour demand framework as standard neoclassical theory. This is in contrast to heterodox perspectives which see wages as determined by political, historical and social factors (see Botwinick 1993 for a detailed exposition).

  • 2

    Kalecki (1943, p. 326) argued that capitalists would be particularly opposed to welfare spending and subsidies of consumption goods: ‘The fundamentals of capitalist ethics require that “You shall earn your bread in sweat”’, which ruled out such policies from their perspective. By contrast, Kalecki noted that not all public investment would necessarily be opposed; only those parts competing with private investment. The idea that full employment policies might be tolerated by capitalists as long as they do not generate labour ‘indiscipline’ and hence undermine profitability is developed by Moudud and Martinez-Hernandez (2014).

  • 3

    The accumulated empirical evidence for the so-called ‘wage-curve’, a negative trade-off between local unemployment and wage levels, is also often taken to be supportive of efficiency wage theory (for example, Blanchflower and Oswald 1995). From this perspective, a higher level of local unemployment would mean that firms have to rely less on wage incentives to maintain the cost of job loss. However, at this level of analysis, the observed relationship could also be explained in Marxian terms, reflecting an imbalance of power in the labour market.

  • 4

    The US retail sector provides some telling examples: Cascio (2006) contrasts Walmart's low-wage, high turnover business model with that of Costco, where employee benefits are higher and conditions better, without adverse performance consequences; Guy (2003) concludes that the failure of other retailers to follow Safeway's adoption of high involvement work practices is likely rooted in a desire to maintain control; firms sacrifice efficiency based on the perception that to do otherwise would reduce profits. See also Pfeffer (2007).



Gary Slater - Senior Teaching Fellow in Economics, University of Leeds, UK

David A. Spencer - Professor of Economics and Political Economy, University of Leeds, UK