One (program) for all or all (programs) for one: Evaluation of the employment program opportunity for all of the federation of Bosnia and Herzegovina

Abstract

Using administrative data from tax records and public employment services, this paper examines whether the largest wage subsidy program deployed in 2014 in the Federation of Bosnia and Herzegovina, Opportunity for All, was effective at improving employment outcomes. Given the non-experimental design, the paper relies on propensity score matching estimators. It contributes to the literature on impact evaluations of active labor market policies (ALMPs) by exploiting detailed work histories of jobseekers to identify the control group. In the preferred specification, the program was effective in increasing employment among program participants relative to the control by 13 % 12 months after completing the subsidized period. However, the results are highly sensitive to the assumption of the starting date of the job spell in the control group, which carries information about previous work history. When changing the assumptions about the starting date of the subsidized job spell of the control group, the results remain positive in the short run, but turn out to be either larger or even negative for the medium run. In all cases, heterogeneity analysis reveals that the program is most beneficial for jobseekers of about 40 years of age and older, and for low-skilled workers.