A long-term evaluation of the first generation of the French urban enterprise zones

This paper provides new empirical evidence on the long-run efficiency of locally targeted tax incentives in revitalizing distressed areas. We focus on the first generation of the French "Enterprise Zone" (EZ) initiative, implemented in 1997 in continental France. This program provides tax incentives to firms located in designated areas plagued by social and economic difficulties. Compared to EZs in the US and the UK, French Enterprise Zones offer generous tax reductions, but these subsidies are limited to small firms with less than fifty employees. To our knowledge, the first generation of French EZs has not yet been evaluated on a national scale. A very rich georeferenced panel dataset is used to identify the impact of the EZ policy. Our identification strategy relies on the selection mechanism employed by the French government: designated EZs were selected according to a deprivation index based on various socio-economic criteria. Besides, in order to qualify as an EZ, an urban area must have a population of over 10,000. Consequently, a pool of deprived urban areas was left untreated because of this threshold whereas it was similar to selected areas in terms of economic development. We use two different empirical methods: a combination of regression and sub-classification on the propensity score, and a regression discontinuity design method. Both methods yield similar results. We highlight a strong positive impact of this policy on employment and business location during the first years of the policy. Detailed estimates moderate this initial favourable assessment. First, the EZ effect is partially offset after the initial five-year period of full tax exemptions by more frequent business discontinuations. We simulate actual tax reliefs at a firm-level at the beginning of the policy and ten years after in order to explain this waning effect. We highlight the fact that the introduction of generous payroll tax deductions at a national scale in 2003 reduced the attractiveness of targeted areas. Second, while firms already operating also benefit from tax exemptions, no significant impact on their employment level can be detected. Besides, this policy promotes sectors of activity that weakly stimulate local employment and development. Overall, these results suggest that firms are sensitive to tax cuts but call into question the capacity of EZs to improve local economy.