arts · education · digital
free money, familiar habits
France offered every 18-year-old a €300 culture credit: 84% signed up, but France's audit court found it mostly intensified habits young people already had, only 7% booked live performance other than music, and the app's recommendations barely steered choices
Source: Cour des comptes, “Premier bilan du pass Culture” Open ↗source: primaryLast updated: 2026-11-12
Problem statement
Cost is the most direct barrier to the arts that a government can remove, and France tried to remove it for a whole age group. From May 2021 every 18-year-old in France was offered €300 to spend through a free, geolocated app (the pass Culture) on books, cinema, concerts, theatre, museums, instruments and lessons; younger teenagers were offered smaller amounts. Sign-up was a success: by the end of August 2024, 84% of 18-year-olds were registered, and 4.2 million young people had signed up since 2019. But the Cour des comptes (France's national audit court) concluded in December 2024 that the main effect observed on the first cohorts was "an intensification of cultural practices already well established among young people," with a risk of windfall use by young people who already had more cultural capital at home. Only 68% of young people from working-class families (parents with few or no qualifications, in manual or clerical jobs) had activated their pass; only 7% of young people had booked a live performance other than music; and once the credit was used up or expired, only 38% kept up the activities they had discovered. The unsolved problem is how to design a culture credit so that it changes what young people who are furthest from the arts do, rather than paying for what better-connected young people were already doing.
Why this matters
The pass is a large public programme (the Cour reports budget credits for the individual part rising from €92 million in 2021 to a forecast €244 million in 2024, or €324 million including the collective part, almost all paid by the state), and its design is now a live question: in February 2025 the government cut the 18-year-old credit from €300 to €150, abolished the credit for 15–16-year-olds, and added a €50 supplement for young people from low-income households or with a disability. The Culture Minister described the individual credit as "too often an instrument of cultural consumption and social reproduction." The scheme has real achievements that the redesign should not lose: near-universal reach; high satisfaction, with the culture inspectorate's May 2024 report on a survey of 18–21-year-olds finding that the pass "helps broaden the cultural horizons of many young people" (while also judging its effects on practices "mixed"); books, including literature (40% of book purchases by 2024, as the share of manga fell from nearly 40% to 20%), as the main use; and, according to a culture-ministry statistics study reported in October 2026, young people in priority urban neighbourhoods using the pass about as much as others. The same study found take-up falling from 93% of eligible young people in large urban centres to 66% in very sparsely populated rural areas, where there are far fewer partner cinemas per young person.
What’s been tried and why it hasn’t worked
(1) A universal cash credit with free choice. Eligibility was by age alone, and the operating company was not mandated to steer the quality of offers; the result, in the Cour's words, was that "the objective of registering a maximum number of young people has so far prevailed over the objective of democratising access." The lack of control led to almost €16 million being spent on escape games before 500 such providers were removed after the audit. (2) In-app recommendation. The company built editorialised and algorithmic recommendation tools to diversify choices; they had "only a very limited impact," since almost 90% of bookings started from a search rather than from content the app pushed (the Cour notes "encouraging results" only among the youngest users). Some mediation has also been tried: the operator ran mediation actions for cinema (a "Cinéma Club") and a programme of young "ambassadors" (600 in the third cohort, selected at the end of 2023, trained in cultural mediation, chosen without aiming for representativeness), and the ministry said in December 2024 that work on mediation had begun in June 2024. The inspectorate's May 2024 report records some cultural operators asking for help to build lasting mediation pathways for lack of means; no evaluation of these actions was found. (3) Relying on the offer. Large public live-performance venues were described as reluctant to open to pass holders, both in what they offered and in the seats they released, and only 55% of museums were registered (many already give young people free or reduced entry). (4) Measurement. The ministry did not survey young people's practices before launch or make the pre-registration questionnaire compulsory, so the Cour had to commission a survey of 5,000 young people (jointly with the culture inspectorate); the effect on practices cannot be measured cleanly. (5) The 2025 redesign: a smaller credit, social targeting, and (under an order published on 30 December 2025) narrower eligible offers, which exclude escape games and sharply limit online offers. These address cost and windfall use (the ministry also presented the age change as a way to fit the individual and collective parts together); the Cour's central diagnosis, that individual credits come with too little mediation (guidance by teachers, cultural workers or peers) "to counteract the pre-existing structural inequalities," is not addressed by changing the amount. Maire-info's report of the October 2026 study adds, in the journalist's words, that "the higher the credit, the more purchases diversify" (24% of the €300 credit went to music products, against 12% of the €20 and €30 credits); since the larger credit went to older teenagers, credit size and age are mixed together there, so whether a halved credit will diversify less is an open question. No evaluation of the reformed scheme was found.
What would unlock progress
Mediation built into the credit rather than left to the market: offers that come with a guide, a group or a first visit with a cultural worker, especially for forms young people do not already know (the Cour recommends developing, with cultural organisations, offers that include a mediation component in sectors such as live performance and museums). The collective part of the pass, spent by classes and run by teachers, is the in-house contrast the Cour draws: there, the Cour writes, the collective part "rests on teachers' mediation work." The Cour's own longer-term answer is to join the two parts, so that later cohorts reach the individual credit at the end of an arts-education pathway begun in the first year of secondary school (sixième); that pathway runs through school, so it does not reach young people who have left it, the group the minister asked the operator in May 2024 to target. A mediated offer that works for young people outside school is the open part. Product design is also part of it: since almost 90% of bookings start with a search, any nudge towards new forms has to work at the search step, not in a separate recommendation feed. And the scheme needs before-and-after measurement of practices for the young people it most wants to reach.
How we’d know it’s being solved
Measured on young people least familiar with cultural practices (working-class, rural, not in school): the share who activate the pass and make at least one booking in a form they had not used before; the share who keep up that form a year after the credit expires (38% overall in the Cour's survey); and the share who book at least one live performance other than music (7%). Beware the proxies the Cour identified: registration rate (84%) and total spending measure reach, and the ministry's undefined "diversification" counts a move from books to escape games as success.
Where to start
Facility-free door: using the Cour report's published tables (by social background, territory and type of purchase), design and prototype an app flow that puts a mediated "first time" offer (a guided visit, a group ticket with a host) at the search step, and test it with 17–20-year-olds who rarely attend live performance, measuring the share who choose an unfamiliar form. Analysis door: compare the pass with other youth culture credits (for example Italy's, Spain's or Germany's youth culture vouchers, which a team would need to source and check) on eligibility, credit size, mediation and published outcomes, and produce a design proposal for the targeting and mediation rules. Partner door (design the trial, don't run it): with a youth organisation or a theatre, design a pre-registered comparison of the plain credit versus the credit plus a mediated first visit, with continued attendance after six months as the primary outcome. Access note: pass Culture booking data are not published as open data (the operator's page on data.gouv.fr listed no datasets when checked); the doors use the Cour report's tables and new primary research. Relevant skills: service and interaction design, behavioural economics, evaluation design, French-language source reading.
Genes
Related problems
Stuck the same way, in other fields.
field: the built worldthe wrong ruler for concrete's lifetimeDurability rules judge low-clinker concrete by how fast it carbonates — but carbonation alone does not corrode rebar, 16 competing test standards cannot be converted to real-world rates, and a standard's conservative carbonation rates would write off reusable precast elementsmeasured the wrong thingsource: primaryfield: the built worldfire codes that protect against the wrong fireWildland-urban interface fire codes are largely unproven — building codes address the wrong fire exposuremeasured the wrong thingsource: primaryfield: making & materialsmore recycling collected, worse material outSingle-stream recycling's collection success degrades material quality, crashing the systemmeasured the wrong thingsource: primaryfield: digital & spacesafety-critical ai with few benchmarksFew public benchmark datasets exist for evaluating machine learning in safety-critical applicationsmeasured the wrong thingsource: secondarySources
- Cour des comptes, "Premier bilan du pass Culture," Rapport public thématique, décembre 2024, accessed 2026-10-22. Link ↗
- Service-Public.fr, "Pass Culture: quels changements concernant la part individuelle ?," 6 March 2025, accessed 2026-10-22. Link ↗
- Jean Damien Lesay (Localtis), "Baisse drastique pour la part individuelle du Pass culture," Banque des Territoires, 28 February 2025, accessed 2026-10-22. Link ↗
- Lucile Bonnin, "Pass Culture: un dispositif populaire, mais des usages différenciés selon les territoires," Maire-info, 9 October 2026, accessed 2026-10-22. Link ↗
- Inspection générale des affaires culturelles (Guy Amsellem, Nicolas Orsini), "Les impacts de la part individuelle du pass Culture," Mission d'évaluation, N°2024-15, mai 2024, accessed 2026-10-22. Link ↗
- Ministère de la Culture, "Pass Culture: le ministère de la Culture réagit aux recommandations de la Cour des comptes," communiqué de presse, 19 December 2024, accessed 2026-10-22. Link ↗
Revision notes
This brief gets checked against its sources and contains the following revisions. Future revisions get added here.
Worth knowing
- The figures from the 2026 government study of the scheme come from a news report about it; the study itself was not read.
Full check
Sources read
- Cour des comptes, first review of the pass Culture (2024): full text
- French culture inspectorate (IGAC), report on the pass Culture (2024): full text
- Service-Public.fr, pass Culture changes (2025): full text
- Banque des Territoires, pass Culture reform (2025): full text
- Maire-info, pass Culture use by area (2026): full text
- French Ministry of Culture press release on the pass Culture (2024): full text
What changed
- Numbers: the 7% is the share of young people who booked at least one live performance other than music, not a share of bookings; the share of bookings that start with a search is almost 90%, not 90%.
- Dates: the narrower list of eligible offers was dated from December 2025; the source gives only the date the order was published, 30 December 2025.
Left open
- that no evaluation of the mediation efforts or of the reformed scheme was found (search results, not sources)
Separate check of specific claims: Statements about how two sources were read were brought up to date, and one source address was fixed.
Separate check of specific claims: Two verbs were changed so the page says young people were offered the credit, not that all of them received it.
Earlier full check: Five repairs: the inspectorate report's date, who commissioned the youth survey, an ambassador group's date, which operators asked for help, and a school-year label.
Separate check of specific claims: Both titles were softened, a journalist's sentence was no longer shown as a study result, a quotation was restored to the audit's wording, and good results and earlier mediation efforts were added.