A model for the evaluation of graduates' first long-term job on labour market history

The recent theoretical literature on causal inference has built on combined features of earlier work in both econometric structural approach and statistical program evaluation approach. The present work attempts to combine the two approaches proposing a dynamic causal model in the context of a study on the labour market transitions. In such context many statistical articles focus on unemployment and temporary job and their effects on time duration or on probability to get a permanent contract. However in the last decade the concept of job and work stability has changed. The rapid spread of temporary employment and the increased instability of the market has aroused a new concept of work: the work path, which can take place in different sectors and positions and require very different skills and knowledge. In this contest becomes of great interest to define which characteristics are peculiar of a good job and to study the effect of it on the subsequent work path. Having at disposal administrative panel data on both Lombardy labour market and records of the graduates of three biggest University of Milan, I study the impact of the first “stable” job coherent with the university education on the future job coherence. I define stable a job with a duration of at least 540 days. A dynamic logit causal model has been performed as it allows to estimate the dynamic effect of the first stable job distinguishing between true and spurious state dependence. The unobserved heterogeneity between subjects is taken into account by considering a latent variable having a discrete distribution. This model under certain hypothesis is equivalent to a model formulated on potential outcomes. For the estimation of the model parameters I use an EM algorithm computing standard errors on the basis of the numerical derivative of the score vector of the complete data log-likelihood. From the application of the proposed model to the available data I conclude that the first stable job coherent with one's own university degree has a positive causal effect on the future coherence job in the long-term period. The main features that seem to have a significant positive impact on coherence are the subject's ability, measured through the graduation mark, and a short distance from the degree, measured with the number of past experiences.