ABSTRACT
We question whether accessibility to local public employment agencies impacts exits from unemployment. We deal with the potential endogeneity of the residential location of jobseekers by using the unanticipated creation of a new agency in the French region of Lyon as a quasi-natural experiment. We use exhaustive and geo-located individual data on jobseekers and local public employment agencies. Contrary to past evidence based on aggregated data, we find no evidence that jobseekers with improved accessibility to the local public employment services experience an improvement of their probability of exiting unemployment. We however find evidence of transitory organizational effects. These findings strongly question the costly strategy of a fine distribution of local public employment agencies across the territory while suggesting that institutional issues are key.
Disclosure statement
No potential conflict of interest was reported by the authors.
Notes
1 The French local public agency network has 951 agencies for a population of 66.3 million and 2.9 million jobseekers; by comparison, the German network has only 621 local public employment agencies for a population of 81 million and 2.8 million jobseekers. In 2009, 80% of jobseekers could reach their agency in under 30 minutes, versus 96.4% in 2012. Comparatively, the average commuting time was 72 minutes for students and employed workers.
2 For a review of the MAUP problem, see Dusek (Citation2004).
3 This result stemming from experimental data is consistent with the meta-analysis papers presented above, where training policies were shown to be less effective than the ‘services and sanctions’ active labour market policies.
4 Cour des Comptes, who conducts financial and legislative audits of French public institutions.
5 The Pôle Emploi dataset does not provide a finer geolocation of jobseekers.
6 The Villefranche agency also moved in 2013; however, this change was implemented after our period of investigation. Furthermore, it remained within such a small perimeter (less than 500m from its initial location) that we suppose that this move is trivial and will have no impact whatsoever in the future.
7 Other nearby areas north of the zone could also be included in the control group but are located outside the Rhône region, i.e. outside the perimeter of our datasets.
8 The zoning used in the paper is based on the flows of movement from residence to work of active persons observed in the 2006 Census.
9 See companion paper for the presentation of matching propensity scores of the treated versus control groups and before versus after the creation of the agency (Bunel and Tovar, Citation2015).
10 Only year 2009 was kept in this version of the paper since, again, no significant results were found for years 2010 and 2011. Results for years 2010 and 2011 are available upon request.