To develop a meaningful understanding of living standards and the distribution of income in capitalist economies, it is important to consider the length of the working day. Since Karl Marx's discussion, in part 4 of Capital's first volume, of relative surplus value, there have been debates over the intensive utilization of labor. More recently, these debates have extended to considering the application of neoclassical rational-choice methodology in explanations of labor extraction. (1) However, there have been historical periods during which the extensive utilization of labor has predominated, manifest in the form of increases in the length of the working day and/or reductions in holidays (see Hobsbawm 1968; Bienefeld 1972). There have also been periods when working hours have fallen, and the trend for nearly all advanced capitalist economies, including Britain, for most of the twentieth century was downward. In the 1980s, however, the downward trend in average hours in Britain and the United States halted, and simultaneously work effort began to be intensified (Green 2001; Schor 1991). It is important to explain why and how working hours increase, decrease, or remain constant in particular periods of history, and this paper intends to explore a number of perspectives on the determination of working hours. It is further suggested that Marxism and institutionalism may both contribute to our understanding and indeed have elements in common.
Economists working within neoclassical economics, and a number of heterodox traditions, have offered alternative theories identifying the determinants of working hours. Neoclassical approaches have tended to emphasize individual choice subject to budget constraints. Institutionalist economists have frequently rejected such forms of analysis, suggesting this emphasis on maximizing behavior is misplaced. Instead, they argue, there emerge norms or ideals, which are in part determined by the institutional structure (one example might be the establishment of a normal working day). Finally, Marxian approaches explain the determination of working hours as an outcome of a class struggle between collective capital and collective labor (one such explanation is offered by Philp 2001). In such approaches, it is power and conflict, rather than individual rational choice in a harmonious environment, that are key in explaining the duration and dynamics of working hours.
Recent interest in the hours people work has been stimulated by a number of factors. First, the high and persistent levels of unemployment exhibited in many European Union economies has prompted interest in policies which redistribute work from the employed to the unemployed. Recent work by Gerhard Bosch and Steffen Lehndorff (2001) evaluates the use of working-hour reductions as an employment policy, suggesting that this is potentially an effective instrument. John Maynard Keynes also identified three elements in maintaining full employment, namely, a sound investment policy, increased consumption, and reductions in average working hours. He suggested "[h]ow you mix up the three ingredients of a cure is a matter of taste and experience, i.e. of morals and knowledge" (1980, 384). In spite of this recognition, however, Keynes was skeptical of short-time working at the industry level since this entailed idle machinery for long periods of the working week (see Cameron and Ndlovu 1999). Thus Keynes maintained, "The cotton industry ... has ruined itself by organized short-time extending over five years, which, by increasing overhead expenses, has raised its costs of production above the competitive level" (1981, 578). As a short-term policy Keynes was thus critical of short time, but in the long run he suggested reduced working hours could be one way to maintain full employment. In fact, recent proposals to limit working time in countries such as Germany have addressed this issue by seeking to the working-time reductions to increased flexibility on the part of workers. It is this factor which has provoked left criticism, which has noted that what is being posited is a thirty-five-hour working week, not a seven-hour working day.
For economists not committed to fundamental redistribution of income the issue of how reductions in working hours are to be funded is problematic. Bosch and Lehndorff have suggested that working time should be reduced as part of a long-term package linking wages, conditions, and increased productivity through time: "In this way, cuts in working time and wage increases can be implemented simultaneously ... without causing a rise in real unit wage costs, which are particularly important in determining international competitiveness" (2001,239). (2) The effect of rising wage costs has also concerned Franco Modigliani et al., who have suggested that the claim that working hours should be reduced while maintaining the weekly wage amounts to "little more than demagogy" (1998, 339). They have suggested that increased labor costs could hardly be expected to come out of profits and that the outcome of reduced working hours without an equivalent reduction in weekly wages would be higher prices, reinforcing the inflationary spiral. The emphasis on avoiding inflationary pressure through limiting the increase in wage costs has also been advanced from within post-Keynesian economics. In advocating incomes policies Paul Davidson has suggested, "One of the most important functions of government in any anti-inflationary struggle is to educate the public that any ongoing income distribution struggle is ultimately costly for all" (1991, 124). In contrast Marxist economists are, of course, not committed to maintaining the prevailing distribution of income between classes and would consider working hour reduction as one element in the class-based conflict over the distribution of power, wealth, and work.
A further policy issue concerns health and safety. Indeed Morris Copeland (1931) has argued that such concerns are principal reasons behind factory legislation aimed at curbing excess working hours. Some evidence can be brought to bear to illustrate this point. Studying workers in California, P. Buell and L. Breslow (1960) found that men working more than forty-eight hours per week had double the incidence of coronary heart disease compared to those working below this total. The extensive meta-analysis conducted by K. Sparks et al. (1997) also supports the view that there is a link between long work hours and ill health. There is thus a body of evidence suggesting that some workers are suffering adverse health symptoms, such as death, connected to overwork. Such policy concerns have prompted the recent interest in the economic and social impact of longer/shorter working hours. However, it is important to reflect on what factors determine the hours of work and to consider at what level working hours are determined. This relates to the philosophical issue of microreduction, and in the following section the relevance of this will be demonstrated before considering a number of competing perspectives on the determination of working hours.
The paper is structured as follows. Initially we consider, from the perspective of the philosophy of science, the microfoundations approach, upon which the construction of the neoclassical labor-supply function is based. Since neoclassical economists emphasize acts of constrained optimization by individual agents, with given preferences, whereas Marxists emphasize the determination of working hours as an outcome of class conflict, such concerns are far from peripheral. In the next section, we consider some explanations from institutional and Marxist economics. In particular there emerge similarities in the analysis of the determinants of working hours offered by John R. Commons and Karl Marx. Before concluding we consider remaining differences between Marxist and institutionalist analyses of working hours.
Microfounded Explanations of Working Hours
The key question posed in this paper is what determines working hours. A number of alternative explanations may be offered. One microexplanation is that a given agent chooses to work particular hours because of his or her preferences for income and leisure time, respectively. Such an approach assumes preferences are determined exogenously, and this approach underpins the theory of labor supply in neoclassical economics. An alternative, macroexplanation, might be that the political power and economic circumstances (including technology) prevailing in the economy determines the hours people work. Marxian economics, in emphasizing class power, is one approach that adopts the macroexplanation of working hours. These alternative approaches raise philosophical questions with respect to levels of explanation. Neoclassical economists, and recently analytical Marxists, have challenged the validity of social scientific explanations that do not provide rigorous microfoundations. (3) In addition, new institutionalism has utilized an approach which adopts neoclassical rational-choice methods.
Objections to the microfoundations approach can be raised from the perspective of the philosophy of science. On a practical level the microfoundations approach requires a considerable amount of detail, the analysis of which may be intractable. Microfoundations in social science usually refers to explaining the state and dynamics of social groups in terms of individual human beings. However, in the philosophy of science there are further reductive levels such as cells, molecules, atoms, and sub-atomic particles. Michael C. Howard and John E. King (1991) suggested it is always possible to reduce to yet lower levels and so explaining in terms of individual human beings itself involves some violation of the principles of microreduction.
As well as the practical problems of the microfoundations approach, Alan Garfinkel (1991) has raised philosophical objections. It...