Due to the absence of statistics called “ ethnic In France, researchers do not have data on discrimination possibly produced by the criminal judge. To overcome this lack, Fabien Jobard, researcher at Cnrsexamined the decisions rendered by a Paris region court with regard to offenses of contempt, rebellion and violence against agents posted by the public force. Its study makes it possible to considerably put the widespread idea that the judges would not be insensitive to the skin color in their judgments.
We know little, in France, on discrimination possibly produced by judicial decisions. In Anglo-Saxon countries, studies of “ Race and Sentencing Are plethos. In these countries, researchers have pre-formatted data by registering ethnic characteristics self-declared by people judged or victims. The possibility of having ethnic statistics simplifies the work of objectification of discrimination. This is not the case in France, where the work consisting in determining if there is discrimination in the judgments rendered involves the production of statistics by the researcher himself.
This is what Sophie Nevanen and I are devoted to. We have collected the decisions rendered by the Melun tribunal de grande instance from 1965 to 2005, by focusing on a single type of offenses, infringements to the public authority (Ipdap), either outrage, rebellion, outrage and rebellion, violence (and combinations between these offenses). L’Ipdap Judged without attached offense makes it possible to have a clearly defined set of qualifications, not very polluted by legislative developments as infringements of violence or theft can be, greeted by various aggravating circumstances during the laws which observe their historical readability.
We have thus collected a set of 864 defendants in a Correctional Court of Great Parisian suburbs, which we have distinguished according to their ancestry and their consonance to train three groups (deleted on those in use in judicial police): Group “ European “, band “ Maghreb “, band “ black ». An accused born in France bearing a typically African name or first name comes from the African group ; An accused born in the Maghreb is the matter of the North African group (unless it has a French Catholic name, it would be a black foot-standing from the European group), an accused born in France bearing a name or name Arab or Berber coming from the North African group, etc. This classification is crude because it does not make it possible to distinguish black defendants bearing an Arabic name, nor the black defendants born in mainland France bearing a typically French name and first name (type Thierry Henry).
Notwithstanding all this, the defendants tried to Ipdap Only it is up, over the period considered, for 62% of them in the group “ European “, “, 20% of the group” Maghreb “, 15% of the group” black (And 2.2% other groups). First discovery: these groups are undeniably unequal before pain. Farm imprisonment is pronounced against 13.6% of group defendants “ European “, But it is against 23.6% of the group’s defendants” Maghreb “And 25.4% of the group” black ». In other words, the risk of firm imprisonment is almost twice in group defendants “ Maghreb ” And “ black “That among the group’s defendants” European ». But these groups may be different from each other, made up differently: these “must be examined” Composition effects ” Or “ Structure effects (See the box below), because to stop at the groups thus built, to be only one sorting, there is a risk of hiding a lot and saying too little.
Glossary
*Structure effect : When a population is distributed in subpopulations, it can happen that a quantity evolves in a sense on each subpopulation and in the opposite direction over the entire population. This paradox is explained because the numbers of certain subpopulations increase when others regress: it is the structure effect.
For example, the salary of each profession can stagnate (or increase weakly) while the average salary increases sharply ; This happens if the highly qualified, the best -paid professions are more and more numerous and, vice versa, the unskilled, the least paid, increasingly rare jobs. Conversely, the constant structure variation is calculated as a weighted average of the variations in the averages of each subpopulation, the weights being the masses of grandeur for each subpopulation. (Source : INSEENomenclature)
* Multivariate analysis : Statistical study of relationships that can exist between several variables.
Thus, we must first distinguish the singular offenses which form the whole Ipdap. The penalties provided for violence are greater than those provided for rebellion, etc. In our court, the risk of firm imprisonment is 1/10 for simple outrage, 1/5 for rebellion or for outrage + rebellion, 1/3 for violence. However, group defendants “ European »Are more frequently translated for outrage, less frequently for violence (and vice versa for others). Then the group defendants “ European »Are more frequently present at the hearings than the others ; And this mode of judgment, says “ contradictory », Is always more favorable to defendants. Finally, it is among the groups “ black ” And “ Maghreb “That we find the highest proportions of defendants deemed in a recurrence (most often in immediate appearance): an accused deemed in immediate appearance incurs almost a chance in two to be sentenced to a firm sentence.
The scope of these effects of structure and composition increases when we combine all of these variables to determine the sentence pronounced all other things, by means of a multiple regression analysis, we realize that the variable “ band “Has no influence in itself. In fact, this variable carries in it the other variables, it is only a reflection of the other variables which, themselves, make the decision of the judge. Thus, despite appearances, however spectacular, the judges are indifferent to the color of the skin or the cause of the defendants.
The situation is however different with regard to the police. As a victim, the possibility is offered to them to build civil party (Pc) and request damage to moral damage suffered. However, 37% of group defendants “ European »See police forces Pc against 51% of those in the group “ Maghreb “And 46% of the group” black ». And this time the multivariate analysis (see box) shows that, all other things being equal, the variable “ band It is indeed an influence in itself. If the decisions of the judges seem to not see the defendants, the individual decisions of the police are there, subject to a study on a wider, not insensitive sample.
What to conclude ? First of all that you have to be careful with the statistics. Scientifically answer a political question (“is there discrimination ? ”) Is very complicated: it is not enough to note a gap, as strong as it may be, it is necessary to study its composition, the structure. Then, the statistics records people already sorted, by the prosecution, by the judicial police officer, by the interpeller police. Also, our study does not allow to say that the“ blacks ”or the“ Maghrebis ”are more often committed than the other offenses. “Maghrebis” and “blacks” are more frequently judged for more serious offenses.
The judges know that they can decently reserve the prison only their customers: those who have already known the prison. An accused already condemned incurs a strong risk of being convicted again. In addition, a defendant “ socially inserted “(Which has a non -precarious employment contract, a stable family situation, etc.) will meet the leniency of the judge, who knows that the prison will condemn him as much criminal as socially (this realism of the profession signs to the daily failure of the penitentiary mission). This is what our study highly highlights: “ Black ” And “ Maghreb “Form the bulk of the criminal clientele, this particular population, most of which of socialization, very quickly, is taken by the judiciary.
To go further:
Fabien Jobard articles:
– On young/police relations.
– on the severity of judges in matters ofIpdap
– On the relationship of the police to know :.
The site of Thesedip