Note d'introduction sur l'évaluation d'impact d'un programme public par la méthode de régression par discontinuité

This note briefly describes regression discontinuity method which estimates program impact when participation is a function of a selection variable predetermined value (threshold). The program evaluation impact might be done in a 2-step way. First, one graphically analyses the relation between the dependent variable and the selection factor in order to get some information about the functional form and the nature of the discontinuity at the threshold. Next, an econometric analysis is formally done in order to estimate program impact.