We analyse what determines the incidence of unemployment among Danish employees by estimation of a logit model for becoming unemployed. Our data is incomplete in the sense that we do not observe whether a transition was caused by the person quitting or being laid off, so we apply the...

We discuss the normal vairance-mean mixture model from a semi-parametric point of view, i.e. we let the mixing distribution belong to a non parametric family. The main results are consistency of the non parametric maximum likelihood estimator in this case, and construction of an asymptotically...

Equilibrium search models require unobserved heterogeneity in productivityto fit observed wage distribution data, but provide no guidance about the location parameter of the heterogeneity. In this paper we show that the location of the productivity heterogeneity implies a mode in a kernel...

In this paper, we compare two different types of models for the duration of unemployment spells: the multiple phase duration model and the competing risks model. Both models allow for non-proportionality of the effect of the explanatory variables on the hazard function. The two models are...

In this paper we extend a job search-matching model with firm-specific investments in training developed by Mortensen (1998) to allow for different offer arrival rates in employment and unemployment. The model by Mortensen changes the original wage posting model (Burdett and Mortensen, 1998) in...