South Africa's youth unemployment and the employment tax incentive: an empirical re-evaluation

Abstract

Our paper uses a Difference-in-Difference estimator to investigate the impacts of the Employment Tax Incentive after the first year of implementation. We use birth cohort bands to track the difference in employment probabilities of subgroups of targeted workers and untargeted workers. We find evidence of improvements in the employment prospects of young workers in the region of 2 percentage points. Our results were largest for African males. We then perform a placebo and find no evidence that differences in employment probabilities of younger workers and older workers were present before implementation. Possible deadweight loss, displacement and measurement effects limit the extent to which we can attribute changes in employment probabilities to evidence that “new” jobs were “created” by the programme