On the morning of September 29th, students at the César Baggio high school, in Lille, stacked bins against the gates and declared their school closed. Within three weeks, the blockades had spread across the country, and by October 17th hundreds of thousands of people were in the streets: lycéens, university students, and, marching beside them, older protesters furious about the cost of living. Organizers spoke of a convergence des luttes, a phrase with deep roots in French political memory. The pictures that travelled furthest, though, were of fire. Earlier in the autumn, in Nantes, the Nelson Mandela high school, recently rebuilt for several million euros to meet environmental standards, had been set alight.
There is an obvious irony in teenagers burning a school to protest the state of schools, and commentators have been quick to point it out. But irony is a cheap way of not listening.
France has a habit of treating its youth revolts like weather: storms that blow in every generation or so and blow out again, leaving a few slogans on the walls. May 1968 is the template, and every uprising since has been measured against it. This one has a more prosaic target. Its enemy is, in large part, a website.
Parcoursup, the national platform that matches high-school graduates to higher-education programmes, launched in 2018. It replaced an earlier system, APB, which had become notorious for settling oversubscribed courses by drawing lots. The promise was to swap chance for judgment. Each spring, seniors list their choices, programmes rank the applicants, and an algorithm sorts out the rest. What the algorithm cannot do is create new seats.
That is the part of the story Thomas Piketty insists on in a recent column for Le Monde. Between 2017 and 2026, the government's higher-education budget rose by about seventeen per cent in nominal terms, a figure ministers like to repeat. Inflation over the same period was roughly twenty-two per cent. The student population, meanwhile, grew from 2.6 million to more than three million. The upshot, Piketty calculates, is that spending per student has fallen by about a fifth since Emmanuel Macron took office. Hundreds of thousands of teenagers work hard for their baccalauréat, he writes, yet "nothing is planned to welcome them into higher education." The squeeze is worst for graduates of technical and vocational tracks, whose numbers have risen while places in the short, practical programmes built for them have not.
Someone has to absorb that shortage, and Parcoursup decides who. Gabriela Ramos, a former senior official at UNESCO and the OECD, has laid out the results in an essay for Project Syndicate. In last year's main admissions round, 97.2 per cent of general-track graduates received at least one offer, compared with 82.9 per cent of vocational graduates. The platform ranks students on records already shaped by postcode, parental education, and private tutoring, then presents the ranking as merit. Its criteria are not public.
It is worth dwelling on what that opacity accomplishes. A minister who announced that ten thousand fewer seats would be funded next year would face a press conference and probably a strike. A platform that simply issues no offer faces a seventeen-year-old alone with a phone. The algorithm is not a neutral piece of plumbing. It is the mechanism by which a political decision about money becomes thousands of private disappointments, each of which looks like a verdict on the individual. "Questions of fairness are a societal discussion, not a technological one," Ramos writes. The students, in their way, are insisting on that discussion.
None of this was hard to foresee. In February 2018, as Parcoursup was being rolled out, Piketty published a blog post warning that the reform risked deepening both the inequality and the opacity of French admissions. Rather than the crude quotas the law proposed, he suggested a points system that would weigh family background gradually, as some Indian universities do. Eight years later, his new column reads like the sequel.
The sorting begins long before anyone logs in. Research by the economist Asma Benhenda, which Piketty cites, finds that contract or beginner teachers make up about a tenth of secondary staff in Paris and its wealthiest suburbs, but around half in the poorest départements, such as Seine-Saint-Denis. Average teacher pay there is roughly twenty per cent lower as a result. The education minister has argued that Seine-Saint-Denis is in fact better endowed than Paris, pointing to smaller classes. Piketty's answer is that slightly smaller classes taught by much less experienced teachers are no bargain.
Parcoursup has its defenders, and their case is not frivolous. Capacity is finite, they argue, and some form of selection is unavoidable; a ranking, however imperfect, beats a lottery. A 2026 parliamentary report proposed fixes rather than abolition: anonymize applications and remove the applicant's school of origin. Ramos adds another. When a scholarship student declines an offer, the place could pass to another scholarship student instead of the next name on the list, a change that by one estimate would help about ten thousand more students. None of these changes would create a single additional seat.
Ramos ends her essay with a warning that reaches well beyond France. Governments that hand decisions about education and other vital services to automated systems, without transparency or accountability, will widen inequality, and the consequences will not stay inside institutions. In Lille, the students did not blockade a website; you can't. They blockaded the building. The machine has been saying no for eight years. This autumn, a generation started saying it back.