We study a simple general equilibrium model in which wages are set by collective bargaining as a mark-up over benefits. The inclusion of taxation and the government budget complicates the relationship between employment and hours worked; hence, we present numerical simulations of employment in terms of hours. There is a range of initial hours from which employment can be increased, or unemployment reduced, by cutting standard working time. Welfare conflicts are explained, but our examples show relatively small profit reductions when hours are diminished below the employers' (collective) optimum, and substantial employment gains.
Persistently high European Union (EU) unemployment has generated renewed debate over the merits of working-time reduction to stimulate employment. A 35-h normal working week was enacted in France and came into force on January 1, 2000 for enterprises with over 20 employees, and 2 years later for smaller firms. Similar legislation is pending in Italy (OECD, 1998). Earlier studies of working time and employment were inconclusive, but generally sceptical concerning the employment benefits of reducing working time Calmfors, 1985, Booth and Schiantarelli, 1986, Hoel and Vale, 1986 and Hart, 1987. Theoretical work, with widely differing models, tends to be somewhat more optimistic Houpis, 1993, Contensou and Vranceanu, 1998a, Contensou and Vranceanu, 1998b, Contensou and Vranceanu, 2000 and Marimon and Zilibotti, 2000. The OECD (1998) emphasises the importance of complementary measures such as wage indexation and flexibility, which together might yield modest employment gains in some enterprises. Such measures and a cyclical upswing have combined to yield rapid employment growth in France since the 35-h week was introduced, but it is difficult to isolate the effect of the hours reduction. Lever (1996) finds significant positive effects of hours reduction on employment in a panel of Dutch manufacturing industries and Rubin and Richardson (1997) find positive employment effects of hours reductions in British engineering. On the other hand, Hunt, 1998 and Hunt, 1999, in detailed studies of German working time, finds no evidence of employment gains. Roche et al. (1996) review international evidence.
Previous modelling of working time has neglected the general equilibrium interaction between unemployment benefits, the payroll taxes that fund benefits and the resulting wage bargain and employment decisions. Hart and Moutos (1995), however, consider efficient bargaining over hours and wages, but unemployment benefits and other exogenous parameters have no effect on employment in their model when the government budget is included. This implausible result, and much other evidence, suggests that the “right to manage” (RTM) model is more appropriate than efficient bargaining. As Teulings and Hartog (1998, p. 144) point out, “explicit simultaneous bargaining on wages and employment is most unusual… The RTM model seems a more accurate description of reality than the efficient bargaining model”. In this paper, we present a first step towards a simple stylish, but consistent, general equilibrium model of the effects of working time and unemployment benefits on employment, wages in collective bargaining, taxes and profits. We adopt strong simplifying assumptions, but numerical simulation are nonetheless required in most cases to obtain quantitative results. Perhaps surprisingly, a substantial range of parameter values suggests employment gains from the reduction of working time.
The plan of the paper is to develop the basic theoretical model in Section (2) progressing from the case of wage and hours setting by a monopoly union, through the regime where employment and hours are determined by individual firms, to the situation where an employers' association determines hours and employment. A computable case of the model is outlined in Section 3 and is illustrated by simulation results. Conclusions are summarised in Section 4.
We have shown how collective bargaining and the relative strength of unions and employers impose natural bounds on the choice of hours. The lower bound is the monopoly union's choice, hMU, which, in general, is close to the general equilibrium employment-maximising level, hm* (the point at which the conflicting effects of productivity gain and utility loss, associated with increasing hours, just balance). Thus, if employers have some power in bargaining, as seems plausible, so that hours are chosen to be greater than both hMU and hm*, then a small mandatory reduction will raise employment. This, perhaps surprising, analytical result is complemented by the illustrative simulations, which show that profit or wage losses from a small cut in working time are likely to be negligible. In some cases, both profit and employment can be increased when hours are cut.
Our general equilibrium model that combines realistic bargaining and the “right to manage” with budget-balancing taxation thus yields a more optimistic evaluation of the effects of mandatory work-time reduction than most previous, partial equilibrium, models. The model is obviously stylish, but, even as a first attempt, it does provide some suggestive hints on these much-discussed policy questions. In particular, countries with strong collective bargaining institutions and high benefits, taxes and wages, such as France and Germany, are less likely to benefit from further cuts if their already relatively short hours than countries such as the UK and the USA. Ironically, there is little public discussion of working time in the latter countries that might well have the most to gain from such measures.
