NORDSTRÖM SKANS, Oskar, 2002, Labour Market Effects of Working Time Reductions and Demographic Changes; Department of Economics, Uppsala University, Economic Studies 66, 118 pp, ISBN 91-87268-73-6.
This thesis consists of three self-contained essays.
Essay I studies the impact on actual hours worked of a 5 % working time reduction for one class of shift workers in Sweden using individual level panel data from firms’ payroll records during the second quarter each year. The main result is that actual hoursonly decreased by approximately 35 % of the reduction in standard hours. Quantile regression results show that the effect was relatively homogeneous over the distribution of hours worked. Much larger effects are derived by studying the effects of the individuals’ scheduled hours. This indicates that a low rate of actual implementation may explain the results and suggests that using variation in self-reported, rather than contractual, standard hours may have biased the results of previous studies.
Essay II extends a model of equilibrium unemployment showing that a general working time reduction will reduce equilibrium unemployment unless the firms have fixed costs for workers. A counteracting effect exists if firms have substantial fixed costs. A testable implication is that a partial working time reductions preferred by the workers should reduce hourly wages unless firms have substantial fixed costs. A 5 % working time reduction for shift workers in Sweden is studied using register-based panel data. The results show that hourly wages increased as a result of the working time reduction. Such an increase in wage demands is consistent with fixed costs and would tend to increase equilibrium unemployment if working hours were reduced for all workers.
Essay III studies the effects of changes in the age structure on aggregate labour market performance using a panel of Swedish local labour markets. The methodology of Shimer (2001) is used for studying the effects of youth cohort size and is extended to include the full age distribution. The results show that young workers benefit from belonging to a large cohort. This is in line with previous results for the US. Furthermore, it is shown that most of the positive effect for young workers is due to an inward shift in the Beveridge-curve. In contrast to the US experience, older workers in Sweden do not benefit from large youth cohorts. Further results show that large numbers of 50 to 60 year old workers have an adverse effect on the labour market. This is consistent with negative externalities from well-matched individuals.