出版社：Dep. of Statistical Sciences "Paolo Fortunati", Università di Bologna
摘要：In western countries national surveys of the labour markets usually provide discrete time observations. Individuals are interviewed at some distinct dates, so that we can observe their status (employed, unemployed, ...), but nothing is known about what happened between observation dates. This kind of data does not permit the estimation of multi-states duration models, which are usually employed to study individual transitions. On the contrary, econometric tools based on continuous-time Markov processes can be applied. In this case, it is very important to take into account as much as possible individual heterogeneity. In this paper, we put forth a new approach (based on the Kalbfleisch and Lawless (1985) algorithm) which completely exploits the sample information on observable heterogeneity. I estimate two models: a reduced form, and a structural form derived from Burdett et alii (1984) dynamic stochastic utility model. I analyse a sample from the Italian survey "Rilevazione Trimestrale delle Forze di Lavoro" on the period January 1985-April 1986.
Loading...
