Selection bias and incentives to firms in development programs: an ex-post evaluation of an aid scheme for Sicily

Abstract

This paper focuses on the evaluation of an aid scheme included in a larger local development program framework implemented in Italy during the 2000-2006 cycle of EU Structural Funds. It follows a previous analysis (Cusimano & Mazzola, 2014) on the ex-post evaluation of the effectiveness of the resources allocated to aid schemes included in the Italian territorial integrated projects (TIPs) and aims at identifying the possible presence of selection bias in measuring the effectiveness of the policy. Previous results showed a partial effectiveness of the program, at least with reference to employment and sales. The objective of this work is to investigate the role of the selection process in influencing the positive effect of the incentive policy. We run an empirical analysis that uses several propensity score matching methods, in combination with a difference in differences approach, in order to check for robustness. With all methodologies we observe that, at least partially, the positive results previously obtained could not indicate an effective success of the policy, but they could be attributed to a distortive effect of the selection process.