On the Political Economy of Labor Market Flexibility
NBER macroeconomics annual
as well as seminar participants at New York University and Northwestern University for helpful comments and suggestions. 1. Considerable empirical and theoretical research has of course been devoted to the European Unemployment problem. The reader can refer to Bean (1993) or Layard et al. (1990). 2. Given that this is a long-run phenomenon, we want to use a neo-classical production function while allowing capital to vary. Under constant returns, the yearly output loss is therefore of the order
... efore of the order of magnitude of the unemployment rate itself, although one would want to correct for manpower's quality and labor supply elasticity. Okun's law is inappropriate here, because it only makes sense for short-run cyclical variations in employment. 152 SAINT-PAUL Figure 1: UNEMPLOYMENT IN EUROPE (Brtain * France among the richest in the world. As a measure of comparison, Marshall aid amounted to around 3% of GDP in recipient countries. Applying a revealed preference logic to this observation, we can conclude that there is not much political concern about persistent unemployment.3 The present paper tries to understand why this is so. The central idea is simple: Given that the unemployed are a minority, political decisions are likely to reflect the interests of the employed rather than the unemployed.4 Hence, a prerequisite for a government being able to fight unemployment is that it reduces the welfare of the employed to a significant extent. Therefore, measures against unemployment are politically viable in certain environments but not others. My purpose is to identify and characterize these environments. The paper is centered around what is perceived to be a key issue in 3. Another interpretation is that governments fear that they cannot do much about it. See Cohen (1988) for a similar view about the rise of conservatism. Sweden, however, seems to have fought it successfully, although this has changed in recent years. It is telling that few attempts have been made to replicate the Swedish experience in high unemployment countries. 4. This logic is the same as that of the well-known insider-outsider model of wage determination (Lindbeck and Snower, 1989). However, the fact that it now works through the political system rather than firm-level wage bargaining radically changes the issues and the analysis. See Burda (1991) for a dynamic voting model of wage formation at the union level. Country Amount (%) Australia 0.27 Austria 0.35 Source: OECD Employment Outlook, 1992. a 1988. the debate over European unemployment: that of labor market flexibility. If unemployment is due to legal arrangements that limit the firms' freedom of choice over employment levels and wages, then getting rid of these arrangements is problematic, because they probably benefit the employed.5 I first set up a simplified model of the labor market under firing costs that gives formal content to this hypothesis. I then show that a natural way to avoid this problem is by setting up a two-tier system, where "flexible" workers coexist with "rigid" ones. The main 5. Otherwise, one would not see why such limitations would exist in the first place. It is possible to write models where the employed actually lose from such limitations (see Atkeson's comment). In partial equilibrium, the employed benefit from firing costs simply because they are fired less often, and these costs are typically payments from the firm to the worker; also, they are likely to increase their bargaining power. In general equilibrium, as long as labor is not the only factor of production, there is scope for the employed to increase their welfare at the expense of firm owners, and also at the expense of the unemployed. If labor is the only factor of production, it is not possible for the employed to increase their welfare at the expense of firms, because they eventually get all the surplus from the match. This, however, may not hold under imperfect competition. Furthermore, in a "renovating" economy with costly labor reallocation, there is still scope for the employed to increase their welfare at the expense of the unemployed by reducing the turnover rate. Firing costs are the natural way of doing so, and it is precisely the mechanism studied in the paper. 6. Of course, this story cannot go all the way through as flexibility increases, because in the limit there is no longer unemployment. 7. See Hibbs (1982) for the empirical analysis of unemployment and other macroeconomic issues in the context of the PBC theories. Alesina, A. (1987). Macroeconomic policy in a two-party system as a repeated game.