| Summary: | The thesis consists of thee parts. Part one considers firm-union bargain over wages and working conditions (effort). It studies in a partial equilibrium setting, the theoretical relationship between the scope of firm-union bargains and the outcome in terms of wages, employment and effort. In particular, the outcome of a bargain over wages and effort is contrasted with the outcome of a pure wage bargain. A main result is that both effort and wages are lower when effort requirements are negotiable (rather than being determined by the employer). The analysis yields implications for the impact of unions on productivity, and gives an explanation to cross-industry differences in union mark-up on wages. In the second part I study labour contracts under temporarily asymmetric information. Under the assumption that workers are more heavily credit rationed than firms, the standard model of testing and self-selection in the labour market is extended in several directions. First, it is shown that ex post inefficient termination may be used as a self-selection device. This is a new explanation for up-or-out contracts in occupations where workers' productivity is revealed slowly. Second, when risk neutral workers can be of more than two different productivities, only the best worker should be overpaid. Finally, when productivity is non-verifiable, large firms may have an advantage in hiring more able workers. The issue of discrimination in the labour market is addressed in the last part. A model in which firms have incomplete information about workers at the hiring stage is shown to entail discrimination as the unique stable equilibrium outcome, even if no agents have a taste for discrimination. Discriminated groups (e.g., blacks, women) earn lower wages, endure longer unemployment spells, and must satisfy stricter requirements in order to obtain work. The model also offers a new explanation for duration dependence in exit rates from unemployment.
|
|---|
