Essays on Temporary Work Agencies and the Economic Analysis of Law

Paper [I] adds to the theoretical literature on the incentives of Temporary Work Agencies (TWAs). Using a principal-agent model with hidden action to model two main types of contracts between a TWA and a Client Firm (CF), the TWA is shown to potentially act against the best interest of the CF when helping to fill a vacant position. The results also suggest that the adverse effect of the incentive misalignment is larger when the worker is going to be leased instead of hired by the CF. However, this effect could potentially be offset by introducing a sufficient level of competition among the TWAs. <p> Paper [II] uses individual-level data on young adults to estimate how the probability of being employed in the Swedish temporary agency sector is affected by whether a partner or other family member has experience of temporary agency work. The results show a significant effect from all peer groups of a magnitude that correspond to the other most influential control variables. We also find that this cohort of the agency sector has a relatively high education level compared to the regular sector, and that there are predominately men working in this sector. <p> Paper [III] analyses possible effects on total employment, and the distribution between agency work and regular contracts as a consequence of the implementation of the EU Temporary and Agency Workers Directive in Sweden. The analysis is based on changes in the compensation to agency workers in a calibrated extension of a Mortensen-Pissarides search model. Even though the results suggest a negative net effect on total employment, the implementation is shown to increase (utilitarian) welfare, and an increased transition probability from the agency sector into regular employment will increase welfare even further. <p> Paper [IV] focuses on settlement probabilities for different types of representation within the Swedish Labour Court. Empirical estimates on a set of unjust dismissal cases show that private representatives are generally less likely to reach a settlement than their union counterparts. The settlement probabilities converge following court-mandated information disclosure, which suggests that information asymmetry is an important factor in explaining differences in settlement behaviour. Privately instigated negotiations are therefore in general insufficient for making cases with non-union representation reach the same settlement rate as cases with union representation.