Law enforcement officers in France are often accused of carrying out identity checks on the basis of physical appearance (contrôles au faciès), singling out particular ethnic groups, or "visible minorities". To show that discrimination is taking place, it is not enough simply to demonstrate that police stops affect one group more than another. It also has to be shown that this group is disproportionately targeted in relation to the population present in the relevant geographical area. This is a particularly difficult task. Fabien JOBARD, René LÉVY, John LAMBERTH and Sophie NÉVANEN report on a survey of identity checks carried out in Paris in 2007-2008 and describe the relatively complex methodology that was used. They first conducted a "census" of the population available to be stopped, and recorded some of their visible features. They then unobtrusively monitored police stops and their targets, noting their characteristics based on the same set of features. By this means, the researchers were able to demonstrate that the "stop population" differed from the "benchmark population" not just in terms of visible ethnicity, but also of age, sex and style of clothing.
Contrôle au faciès, or "identity check on the basis of physical appearance", is a familiar yet polemical expression in France, where it refers to what is known in European legal language and the specialist literature as "racial" or "ethnic profiling". It means that instead of stopping individuals on the basis of operational briefings, suspect descriptions, or specific acts leading them to suspect that an offence has been committed or is about to be committed, law enforcement officers decide which people to stop mainly according to their appearance. This practice formed the subject of the survey described here (IC Survey).(1)
Despite the recurrent controversy over profiling, knowledge of this disputed police practice is very limited. This lacuna stems first and foremost from the absence of institutional sources, as identity stops do not generally leave an administrative "paper trail" (unless they trigger criminal prosecutions) and are not systematically included in police statistics. Even if they were recorded, the data would tell us nothing about the circumstances in which they had taken place, as these official sources would only mention the nationality of the people who were stopped. If appearance-based identity checks are indeed a reality, the key factor is apparent ethnicity, not nationality.
Making these police practices visible therefore requires a specific method of investigation, whereby the distinguishing features of the individuals who are stopped can be recorded without disturbing ongoing police operations, and a benchmark population can be established to put the checks in context. This is what we set out to do in Paris between 2007 and 2009, at the instigation of the Open Society Justice Initiative, as part of a programme of research and action on the prevalence or otherwise of "racial profiling" by police forces in different European countries (Open Society Justice Initiative, 2005, 2006, 2007, 2009a, 2009b and 2009c; Miller et al., 2008).(2)
In the first part of this article, we review previous research on stop-andsearch practices and look at how the police conduct identity checks in public spaces in France. In the second part, we describe the survey methodology, and in the third and final part we set out the main survey results.
I. Identity checks in France
The various studies on identity checks available in France are based on observations reported in monographs and declarative data collected from individuals who have been stopped. After providing a brief overview of this research and its flaws, we describe how quantitative studies conducted in other countries have sought to gain an objective picture of the issue. In order to provide all the information needed to understand police practices, we then set out the legal basis for carrying out identity checks in France. …