La fabrique des statistiques pénitentiaires : comment une enquête ethnographique peut éclairer une recherche économétrique

This article intends to relate a specific research experience in which ethnography complements an econometric work. First, I published some results based on prison statistics and relative to electronic monitoring and recidivism; then, I tried to find out how these figures were constructed on a daily basis inside prisons. I thus decided to lead some ethnographic observations inside two Parisians prisons, where I was able to witness the encoding process in the penal and correctional services. This article shows that there exists an important cognitive dissonance between prison warders, who never speak about the statistics they are constructing during their duties and are hardly aware of what is at stake, and members of the French Department of Prison, who use these data to publish official statistics. Ethnographic observations show in particular that sociodemographic variables are not properly filled by prison warders; these data should be treated with caution by econometricians.