FACULTY AND RESEARCH

Previous literature has documented a large short-term earnings gap for entrepreneurs that return to the wage sector. Using matched employer–employee data from the Belgian Labor Market & Social Protection Database, we document how this initial gap is remarkably persistent. Former entrepreneurs earn 27% less than propensity-scored matched controls 5?years after returning to wage work, which is almost unchanged from the short-term earnings drop. About 60% of this gap results from reductions in hours worked while the remaining 40% is due to reductions in the wage rate. We offer evidence that the decline in hours worked is a choice of former entrepreneurs and therefore reflects a compensating differential, while the decline in the wage rate is a penalty resulting from statistical discrimination by employers.