Earnings of young doctorates in private jobs after participation to post-doctoral programs

In this paper, we address the question of destination of post-doctorates for young French PhD graduated in exact sciences in 1996. The creation of post-doctoral program was firstly designated to PhD willing to embark on a public career. However, an important part of post-doctorates rather get to the private sector, particularly outside research positions. The question that occurs is that of the value of post-doctoral experience. We propose here to estimate earnings of young doctorates in private jobs had they participated to a post-doctoral program. To control for selection bias arising in the case where unobservable elements are correlated between participation and earnings, we estimate a treatment effect model. The main finding is that when selection bias is not control for, post-doctoral participation increases earnings, however, when controlling for selection bias, there is no more a positive effect of the participation. As regards to this finding we point out that post-doctoral program play much more the role of a signal in the first-stage career. This finding is also strengthened when using a bivariate selection rule. In that case, we also control for the endogenous nature of having been recruited in the private sector.