The impact of firm characteristics on the success of employment subsidies : a decomposition analysis of treatment effects

"In this paper we show that firm characteristics have an influence on the success of employment subsidies e.g. wage subsidies and in-work benefits, as they can strengthen positive effects or mitigate negative effects. We consider firm characteristics as post treatment variables, which are realised after the (placement officer's or the unemployed job seeker's) decision regarding programme participation has taken place. Therefore in a first step we estimate pairwise treatment effects by propensity score matching, controlling for pre-treatment characteristics of the treated and control individuals only. In a second step as a methodological contribution we propose a decomposition of the pairwise treatment effects using an Oaxaca/Blinder style decomposition analysis on the matched samples. In this decomposition we include the post-treatment firm characteristics as explanatory variables. Because employment status is a binary outcome variable in our empirical application, we use a generalisation of the decomposition analysis to nonlinear regressions developed by Fairlie (2005). This procedure allows us to distinguish between the part of a treatment effect that is due to differences in firm characteristics between treated and controls (the 'explained' part) and the part that is independent of those differences (the 'unexplained' part)." (Author's abstract, IAB-Doku) ((en))