Did French enterprise zones fail poor areas? It’s mainly about jobs

Laurent Gobillon, Thierry Magnac, Harris Selod

24 January 2011



Working class areas on the periphery of French cities developed in a process that spanned two centuries. The distressed areas that have been in the news – popularly known as les banlieues – emerged in the 1950s and '60s with their concentration of migrants, unskilled workers, and social ills.

Probably because of the key role played by French-style housing projects (i.e. the massive concentration of subsidised rental units known as les grands ensembles, which were built to address the post-war shortage of housing and later to accommodate repatriated French settlers from Algeria), it was long believed that the ghettoisation of French suburbs was to be blamed principally on poor urban planning.

This led to the naïve belief that the policy priority was to address housing conditions rather than more fundamental economic issues. This view is at odds with analyses of American ghettos which, as early as in the mid 60s, pointed at the lack of localised economic opportunity as a source of social ills and riots. Furthermore, although a key feature of poor areas in France, unemployment was long perceived to be exclusively caused by a mismatch in skills (a bias of labour demand against low skills) so that policies mostly aimed at retraining laid-off workers and indemnify job seekers. In comparison with the US, it took three more decades in France to discuss and implement policies specifically aimed at addressing the lack of economic activity in distressed areas.

The key justification to try and address spatial imbalances in job locations is based on an abundant, primarily American, literature that shows that it can be very harmful for workers to be physically disconnected from employment locations, essentially because of high job-search and commuting costs. This situation, labelled spatial mismatch by urban economists, is further exacerbated by the very low quality of social networks in segregated areas as unemployed workers cannot rely on their acquaintances to help them find a job.

A menu of policy options

In this context, how can policies seek to improve access to employment for people from the poorer areas? There are basically three broad options at the disposal of policymakers (Ihlandfeldt and Sjoquist 1998).

  • The first option, based on the view that unemployed workers would benefit from relocating closer to job opportunities if they could afford to do so, is to provide money transfers to facilitate residential mobility, typically by granting housing vouchers or financing social housing in places closer to jobs. Such policies can allow residents from distressed areas to move out to more expensive areas (as in the Moving to Opportunity programme piloted in a selection of US cities in the 1990s). However, they have remained experimental and only seem to be implementable on a small-scale basis. Scaling up such interventions would result in a rise in housing prices that would offset any initial improvement in affordability.
  • Rather than trying to oppose or mitigate the sorting of individuals or the residential housing market (a Sisyphean task if any), the second option is to facilitate job search and commuting by improving information flow and transport between places of residence and places of work. This includes investments in infrastructure and subsidies to both private and public transportation.
  • Finally, a third option is to try to attract firms and jobs to designated poor areas through fiscal incentives (mainly through tax exemptions). Such schemes, referred to as “enterprise zones” have received considerable attention in the UK and the US starting in the mid-1980s.

The US experience

In the US, enterprise zones programmes have been implemented both by some states and by the federal government (see e.g. the Empowerment Zone programme started in the mid-1990s). Interestingly, these programmes differ greatly on several aspects, including:

  • the fiscal incentives being provided – which may comprise reliefs on capital taxation, employment taxation or business tax,
  • the eligibility for the tax credit – which is sometimes conditional on the hiring of local workers,
  • and the duration in the phasing out of the exemptions.

These varied experiences are key in appropriately designed policy impact evaluations of what works and what does not. The results however are rather mixed and range from a positive effect on local employment and a negative effect on the local poverty rate (Busso and Kline 2008) to no effect whatsoever in both the short and long run (Neumark and Kolko 2010). There is thus no consensus about the ability of enterprise zones to significantly contribute to poor area economic revitalisation although the nature and intensity of fiscal incentives used to attract firms seem to matter.

The French experience: New evidence

France did not pilot enterprise zones before a national programme was adopted and implemented in 1997. It originally concerned 44 poor areas of at least 10,000 inhabitants (usually a portion of a municipality, and sometimes across two adjacent municipalities), among which nine (the biggest) were located in the Paris region. The creation of these zones, labelled Zones Franches Urbaines, aimed at attracting firms thanks to a series of significant tax reliefs on property holding, corporate income, and in particular on wage taxes. An important measure was to condition the wage tax exemptions on firms hiring at least 20% of their labour force locally. It was believed that conditioning on local employment would strongly boost local hires. The number of designated zones was later extended and the intensity and duration of the proposed subsidies were intensified in 2004 and in 2006.

In recent research (Gobillon et al. 2010), we conducted an impact evaluation of the programme by looking at the effect of the first wave of enterprise zones on the propensity of unemployed workers to find a job. Contrary to the previous literature which usually focuses on employment growth or on the local creation of firms, our choice of outcome of interest was motivated by the fact that a main objective of the policy had indeed been to help locals move out of unemployment (not just create or displace jobs, which may only have an indirect effect on the local population).

This impact evaluation was carried out for the Paris region, using an exhaustive dataset on job applicants registered at the French National Unemployment Agency, and resorting to a varied toolkit of statistical methods. We assessed whether unemployed workers in municipalities with a newly created enterprise zone improved their chances of finding a job compared with unemployed workers living in similar municipalities but where no enterprise zone was created.

We found a small positive effect of the enterprise zone programme. In each municipality in our sample, while on average about 300 unemployed workers find a job every semester, enterprise zones help an additional group of 10 workers to find a job over the same duration. It could be argued that this figure represents a lower bound of the effect of tax exemptions since out-of-the-labour-force residents may also have reacted to these new opportunities. Because of missing information, some exits to a job may also have been attributed to other types of exits from unemployment that were disregarded in the analysis.

Even if the true impact on job creations for residents is substantially larger than the direct effect on exits from unemployment, the overall impact is likely to be small and this is consistent with findings of a modest effect on firm and job creation (Rathelot and Sillard 2009).

These job effects are also small given the huge cost associated with the policy. In 1997, the first year of programme implementation, it is estimated that the total cost of the policy for the whole France amounted to €141 million. The wage tax exemption was granted to 26,000 jobs for a total of €53 million but only 6,000 of these jobs were held by residents of enterprise zones (Ministère Délégué à la Ville 2001 and 2002).

This means that for each job held by an enterprise zone resident, almost €9,000 were granted in wage tax exemptions. The cost associated with the hire of an unemployed worker residing in an enterprise zone is even greater.


Busso M and P Kline (2008), “Do Local Economic Development Programs Work? Evidence from the Federal Empowerment Zone Program”, Yale Economics Department Working Paper 36.

Gobillon L, Magnac T and H Selod (2010), “Do Unemployed Workers Benefit from Enterprise Zones: the French experience”, CEPR Working Paper 8084.

Ihlanfeldt K and D Sjoquist (1998), “The spatial mismatch hypothesis: a review of recent studies and their implications for welfare reform”, Housing Policy Debate, 9:849-892.

Ministère Délégué à la Ville (2001), Bilan des Zones Franches Urbaines. Rapport au Parlement, 40 pages.

Ministère Délégué à la Ville (2002), Bilan des Zones Franches Urbaines. Rapport au Parlement, 61 pages.

Neumark D and J Kolko (2010), “Do enterprise zones create jobs? Evidence from California’s enterprise zone program”, Journal of Urban Economics, 68(1), 1-19.

Rathelot R and P Sillard (2009), “Zones Franches Urbaines : quels effets sur l’emploi salarié et les créations d.établissements?”, Economie et Statistique, 415-416, 81-96.



Topics:  Labour markets Poverty and income inequality Taxation

Tags:  France, Enterprise zones, tax incentives

CNRS Senior Researcher, Paris School of Economics; Research Fellow, IZA and CEPR

Professor of Economics, Toulouse School of Economics

Senior Economist, Research Department, World Bank