Abstract:
In recent applications of discrete choice models of labour supply considerable attention has been devoted to strategies to increase the flexibility of models for a better fit to the data. These include the introduction of random parameters, fixed cost of work or flexible functional forms of preferences. Based on estimates of models of recent studies this paper compares these different modeling strategies. Results for Swiss data show that the traditional way to interpret fixed cost of work is ad hoc. Furthermore our results indicate that care should be taken when using very general function forms of preferences