Violence of October 17: Alexandra Gonzalez dismantles the limits of the figure of “75% of Arab-Muslim or African first names” used by the RN and certain media

A spectacular figure, but what does it actually measure? On BFMTV, journalist Alexandra Gonzalez (see video below) detailed the methodological flaws of an analysis by Jérôme Fourquet, invoked by Philippe Ballard, deputy of the National Rally, about the violence of October 17. A sequence which illustrates the eagerness to transform fragile data into arguments for stigmatizing populations designated as “Arab-Muslim or African”.

Jérôme Fourquet is a political analyst and director of the Opinion department of Ifop, an institute specializing in polls and opinion studies. It was from hearing reports published in the press that he put together a corpus of 70 people whose first names were mentioned. According to its classification, 75% have a first name of “Arab-Muslim or of African origin”.

But can we transform this observation into a statistic on all the people involved in the violence? After directly questioning Jérôme Fourquet, Alexandra Gonzalez presented several reasons to contest this generalization. Before brandishing a percentage to designate populations for vindictiveness, the least requirement would be to verify its solidity.

People prosecuted, not necessarily convicted

First problem: according to the journalist, Jérôme Fourquet admits not knowing if the 70 people listed have all been convicted. The outcome of the procedures was not systematically specified in the articles and was not used as a criterion for its analysis.

The corpus can therefore include people who are finally released. Presenting them indiscriminately as “thugs” or “looters” amounts to attributing to them a guilt that this work cannot establish. This shift is particularly serious: when the first name becomes an incriminating argument, the presumption of innocence disappears behind the suspicion of identity. A person should not have to answer for supposed guilt because their first name lends itself to political discourse.

An unrepresentative corpus

Second limit: these 70 people were selected based on the information available in the press, and in particular the presence of their first name. According to Alexandra Gonzalez, Fourquet himself admits that this sample is not representative. The journalist recalls that 917 people were tried in immediate appearance, according to figures that she attributes to Gérald Darmanin. The 70 cases studied therefore represent approximately 7.6% of this total. The other 847 do not appear in the corpus.

The problem is not only due to the number of files: there is no guarantee that the people whose first names the press publishes have the same characteristics as those whose names it does not mention. The percentage obtained cannot therefore be extrapolated to all defendants.

Branding this result as a general truth, while ignoring these limits, is a misleading political use of figures. A percentage does not become proof simply because it reinforces the prejudices of those who repeat it.

Different offenses grouped under the same label

Alexandra Gonzalez also underlines the diversity of the acts prosecuted: violence against the police, possession of fireworks, damage to property, but also contempt. Not all of these offenses correspond to looting or damage.

Finally, it recalls a reservation formulated by Fourquet himself in Le Figaro : the people arrested do not necessarily represent all the participants in the riots, the most organized being able to escape arrest. Bringing everything back to the same category of “breakers” makes it possible to erase these differences, then to associate this category with supposed origins. This amalgam fuels a collective indictment, well beyond just the individuals being prosecuted.

A figure whose scope is largely overestimated

The journalist’s demonstration therefore does not consist of proposing another percentage. It shows why that of 75% does not allow us to affirm that three quarters of the thugs or those convicted have a given origin.

Added to these limits is an essential distinction: a first name is not enough to establish a person’s nationality, religion or family background. Moving from a classification of first names to a general conclusion on the origin of the rioters presupposes elements that this corpus does not provide. Faced with the political resumption of this figure, Alexandra Gonzalez thus puts at the center of the debate what her spectacular presentation tends to erase: the selection of files, the diversity of offenses and the absence of systematic information on convictions.

This eagerness to designate “Arab-Muslims” and “Africans”, according to the favorite terminology of those obsessed with origins, is a mechanism of stigmatization: brandishing a fragile figure to cast suspicion on entire populations. Those who relay this percentage while ignoring its limits are fueling prejudices under the guise of statistical analysis. A first name becomes a piece of evidence, a supposed origin a reason for collective suspicion. This process is unworthy of public debate. The fight against violence requires established facts and individual responsibilities. It cannot serve as a pretext for a trial of origins.

Violence during the demonstration on October 17: Alexandra Gonzalez dismantles the limits of the figure of “75% of Arab-Muslim or African first names” used by the RN and certain media https://t.co/Ec4eUOSs7g

— Oumma.com (@oumma) November 4, 2026