In France, 73% of citizens say they trust the police [1]. Behind this figure lies a choice that 2027 will have to make: continue operating 47 million identity checks per year without recording or evaluating them, or make these checks traceable, public, and subject to legal challenge.
The police hold firm where everything else is collapsing
The gendarmerie enjoys 77% trust, ranking 6th among institutions tested in the February 2026 wave of the CEVIPOF barometer [1]. Justice caps out at 42%, media at 29%, political parties at 15% [1]. Law enforcement occupies a unique position in a landscape of institutional decay.
It is a strong score. It is also misleading, because it conflates two experiences that have nothing in common.
Identity checks are experienced very differently depending on who you are
In 2024, 26% of people surveyed by the Ombudsman said they had been checked at least once in the previous five years, compared with 16% in 2016 [3]. Ten percentage points higher in eight years. The survey, conducted among 5,030 people, is the most precise data available on this point [3].
Young men perceived as Black or Arab are checked four times more often than the rest of the population, according to the Ombudsman’s 2024 survey [3]. They are twelve times more likely to have undergone a thorough check. And 19% of people who were checked say they experienced inappropriate behaviour: informal address, provocations, insults, rough conduct [3]. People perceived as Black or Arab report such abuse twice as often. People identifying as non-heterosexual face it with a 50% higher risk [3].
The first edition of this survey dates to 2016. Ten years later, over-checking remains very high. The gap with other categories has narrowed slightly, but the phenomenon has not been resolved [3].
Under-35s, employees, and manual workers show the strongest warning signals in the CEVIPOF barometer of February 2026 [1]. These are also the categories most exposed to repeated checks.
Forty-seven million checks with no counting or policy framework
In 2021, police and gendarmerie combined conducted approximately 47 million identity checks per year, or roughly 9 per patrol per day [4]. None of these checks is recorded. Their grounds are not documented. Their outcomes are not analysed [4].
In its report of 6 December 2023, the Court of Audit identifies the absence of practical oversight: no standard procedure for officers, no practical guides, no effective monitoring by the general inspectorates of police and gendarmerie [4]. It notes the lack of training in assessing the necessity of conducting a check and the absence of strategic thinking about grounds, which contributes to the risk of abuses [4].
The result is measurable. The National Police General Inspectorate (IGPN) reports a 25% increase in the number of serious injuries resulting from police action between 2022 and 2023 [7]. The same period saw record use of so-called intermediate force weapons. The UN Committee Against Torture has warned of the absence of statistical data on complaints that would allow comparison with investigations opened and prosecutions pursued [4].
Without counting, proving that abuses are marginal is impossible. This is exactly the circularity that the Court of Audit denounced.
On 26 June 2025, the European Court of Human Rights condemned France for discriminatory identity checks: a first [8]. The procedure had begun before French courts in 2012. Thirteen years of proceedings, an international condemnation, and still no systematic recording [8].
What inaction costs, in euros and in efficiency
Combined budgets for the National Police and National Gendarmerie programmes reach 23.4 billion euros in commitment authorizations for 2025, or 0.7 billion more than in 2024 [9, 10]. Part of this effort finances checks that 39.2% of police and gendarmerie officers themselves judge to be of little or no use in guaranteeing territorial security [3]. Nearly two practitioners in five consider their own tool largely unproductive.
Making checks traceable and subject to legal challenge
The 2027 choice concerns the architecture of accountability, not the size of forces or their budget.
Today, no one is responsible for checks collectively. The officer who conducts a check records nothing. The superior does not supervise.
The general inspectorate handles only a tiny fraction of complaints. In 2024, remedies available to people who believe they have suffered a discriminatory check remain underused: a few hundred cases handled each year [3]. Proof of an abusive check is inaccessible to anyone wishing to report it.
The mechanism exists, and it has already been developed. The Interior Ministry itself proposed, in response to the Court of Audit in 2023, to record each check in existing mobility files [4]. The Senate supports a modification of the wanted persons file architecture to clarify, at each mobility consultation, whether it is conducted as part of an identity check [11]. Automatic recording in an already-existing system: no paper forms, no additional burden on the officer.
The next step concerns who can challenge what. Making this data public by territorial unit, in 2026 or 2027, with indicators of distribution by stated profile and by municipality, opens a factual basis for any citizen wishing to exercise recourse through the Ombudsman [3]. This architecture has been developed by the Court of Audit [4], defended by the Ombudsman [3], and confirmed as necessary by the European condemnation of June 2025 [8].
This choice has a real cost. It creates accountability that police hierarchies have never known. It requires political will to assume transparency on figures that may prove uncomfortable. It also makes possible what the average of 73% masks today [1]: trust extended to those for whom this figure means nothing.
Sources
[1] CEVIPOF, “Barometer of political trust, wave 17,” OpinionWay, February 2026, https://www.sciencespo.fr/cevipof/ (accessed 15/09/2026).
[2] CEVIPOF, “Barometer of political trust, wave 16,” OpinionWay, February 2025, https://www.sciencespo.fr/cevipof/fr/actualites/barometre-de-la-confiance-politique-du-cevipof-2025-le-grand-desarroi-democratique/ (accessed 15/09/2026).
[3] Ombudsman, “Survey on Access to Rights 2024: Police-population relations: identity checks and complaint filing,” June 2025, https://www.defenseurdesdroits.fr/sites/default/files/2025-06/ddd_EAD-2024_volume-1_relations-police-population.pdf (accessed 15/09/2026).
[4] Court of Audit, “Identity checks: a widespread practice with purposes to be clarified,” thematic public report, 6 December 2023, https://www.ccomptes.fr/fr/publications/les-controles-didentite (accessed 15/09/2026).
[5] SSMSI, “Experience and perception in matters of security: Survey Report 2024 edition,” Interstats References, October 2025, https://www.interieur.gouv.fr/Interstats/Actualites/Vecu-et-ressenti-en-matiere-de-securite-Victimation-delinquance-et-sentiment-d-insecurite-Rapport-d-enquete-edition-2024 (accessed 15/09/2026).
[6] SSMSI, “Insecurity and delinquency in 2024: statistical review and departmental atlas,” 2025, https://www.interieur.gouv.fr/Interstats/Actualites/Insecurite-et-delinquance-en-2024-bilan-statistique-et-atlas-departemental (accessed 15/09/2026).
[7] IGPN, “2024 Annual Report,” October 2025, https://lessor.org/operationnel/le-rapport-2024-de-ligpn-detaille-lusage-des-armes-par-les-policiers-et-leurs-consequences/ (accessed 15/09/2026).
[8] ECHR, judgment Seydi and Others v. France, case no. 35844/17, 26 June 2025, https://www.defenseurdesdroits.fr/controles-didentite-discriminatoires-la-france-condamnee-par-la-cedh-904 (accessed 15/09/2026).
[9] Senate, report on the 2025 budget bill, Security missions: National Gendarmerie, 2024, https://www.senat.fr/rap/a24-146-11/a24-146-112.html (accessed 15/09/2026).
[10] Senate, report on the 2025 budget bill, Security missions, 2024, https://www.senat.fr/rap/a24-150-12/a24-150-12_mono.html (accessed 15/09/2026).
[11] Senate, bill to restore the bond of trust between police and population, 2024, https://www.senat.fr/rap/l24-587/l24-5873.html (accessed 15/09/2026).
[12] Observatory of Inequalities, “Police checks highly unequal depending on skin colour,” Ombudsman data 2024, https://www.inegalites.fr/Des-controles-de-police-tres-inegaux-selon-la-couleur-de-la-peau (accessed 15/09/2026).
