Khalid Maman Waziri
IBD Salle 21
AMU - AMSE
5-9 boulevard Maurice Bourdet
13001 Marseille
Edward Levavasseur : edward.levavasseur[at]univ-amu.fr
Océane Piétri : oceane.pietri[at]univ-amu.fr
Morgan Raux : morgan.raux[at]univ-amu.fr
This paper extends the sample selection stochastic frontier model of Lai (2015) to a Double Hurdle Stochastic Frontier model (DH-SEF) which can take account of sample selection occurring at two different stages. Using the Closed-Skew Normal (CSN) distribution framework and some results of Meng and Schmidt (1985), we derive a general likelihood function which includes the probabilities of passing successfully two selection hurdles before being observed in the frontier estimation sample. Furthermore, we propose an estimator of technical efficiency scores taking account of the double selection information. We apply our proposed DH-SEF model to a sample of young people entering the labor market in France in 2004. The results suggest that when two selection stages exist, taking account of a unique selection process leads to biased estimates of the earnings frontier.