In France, the Paris Institute for Political Studies—informally known as Sciences Po—has long been known for its snobbery: As one of France’s famed grandes écoles, it counts among its students the last four French presidents (though Nicolas Sarkozy did not graduate) and a large share of the French political and business elite. It’s also a gateway to the Ecole Nationale d’Administration (ENA), the ultra-competitive antechamber of the ruling class.

The ENA was created in 1945 to democratize the senior civil service and put an end to old nepotistic hiring practices. But by the turn of the century, it became clear that these schools did not exactly train the elite so much as renew their privileges and ensure that the next generation benefited from them, too.

So in 2001, Sciences Po began an effort to admit students from less privileged backgrounds. The new policy was the brainchild of the late Richard Descoings—a brilliant, eccentric, and eventually infamous man who was as admired as he was demonized for his efforts to “Americanize” the institution. Making the school more “American” in this case meant bringing affirmative action to a country that still balks at acknowledging, let alone discussing, ethnic differences in public. It caused an uproar. The French republican ideal of equality could only allow for a “territorial” approach to affirmative action.

Since the early 80s, the French policy to combat inequalities in education has mainly consisted of drawing “priority education zones,” mostly in the suburbs, and accommodating their needs. This reasoning assumes location is a solid indicator of privilege, class, and wealth. Sciences Po signed “contracts” with seven high schools from these disadvantaged areas that would filter their best students into a less strenuous track that replaced the dreaded written admission test with two oral examinations.

At the time, Descoings was considered a dangerous disrupter, with the original test seen not only as the best (and only possible) measure of merit. Critics cried foul, invoking the image of neocolonialism towards the suburbs. Then, echoing contemporary debates in the United States, a right-wing student union filed suit for “breach of equality between students” that led the school to slightly modify the process toward more transparency.

The results of the program were considered largely positive. Out of 37 students admitted in the first year, five had ranked in the top 10 on the end-of-year exams. To date, nearly 100 high schools in France have signed contracts, and 1,929 students have benefited from the scheme. Dropout rates turned out to be the same for students admitted through either track, and employment rates after graduation are comparable. This seems to validate Descoings’s objective of “diversity in excellence.”

At the same time, a recent investigation led by the left-leaning newspaper Libération in 2017 found that 40 to 45 percent of the beneficiaries of this “back door” are now students whose well-off parents who intentionally placed their children in struggling high schools. This might look positively benign compared to American college-admissions scandals, but it suggests that the territorial approach was too loose to target the right students.

Indeed, according to the think tank Observatory of Inequalities, in elite schools, “the share of children from the lower class who represent 38 percent of young people aged between 18 and 23, has not changed since the late 1990s.” The prestigious Ecoles Normales Supérieures count 20 times more children of senior executives than those of factory workers (who represent 20 percent of the workforce). In Polytechnique, another prestigious institution, the factor is 50. In spite of these efforts to be more inclusive, and even though the French “Ivies” are relatively affordable—a year at Sciences Po costs 0 to 10,000 Euros on a sliding scale—the problem of social homogeneity remains.

Questions surrounding privilege, education, and mobility resonate singularly in France today, as the Yellow Vest protests each Saturday make the rift between the elites and “the people” impossible to ignore. In fact, the debate has grown so heated that in his response to the protests, President Emmanuel Macron came out for the abolition of the ENA. (Macron—nothing if not the product of this establishment—has probably read the Italian sociologist Vilfredo Pareto, who theorized the notion of the “circulation of elites” and observed that “history is a graveyard of aristocracies.”)

The ENA’s recruitment channels “are no longer meritocratic channels, where, when you come from a family of workers, farmers, or craftsmen,” he said at a press conference. When Macron graduated from the ENA in 2004, fewer working-class children passed the entrance exam than in the 1970s. In France, diversity among the elites isn’t significantly increasing: not geographically, not socially, not economically. And when it comes to race and ethnicity, there’s no way of really knowing, as France does not collect this kind of data.

What this picture shows is that by addressing inequalities individually and territorially, with a very small number of students getting “rescued” from their class, French diversity policies legitimize the idea of an effective meritocracy, all while doing little to diversify elite institutions.

This reasoning misses the fact that the cause of social inequality runs much deeper than going to a bad high school. To make it, one has to have survived the grueling series of eliminations that mark public-school curricula practically from birth. Because, in spite of its reputation for generous social welfare, France has very poor social mobility: French students whose parents have graduated are 14 times more likely to have a higher-education diploma (compared to four in Canada). According to the OECD, it would take six generations for a descendant of a poor French family to achieve an average income as opposed to two or three in the Scandinavian countries.

As the sociologist Pierre Bourdieu has noted, the amount of cultural capital received (or not) by an individual ultimately has an enormous impact on their social destiny. Free education, then, does not cancel inequalities; bootstrapping is blind to the subtleties of family and social environment and its corollaries, self-censorship, lack of information, guidance, and social networks.

Three years ago, when I worked as an editor at Condé Nast, I took a 15-year-old girl on a one-week internship through a nonprofit that places ninth graders from disadvantaged schools in offices they would not otherwise have access to. This is a crucial year in France, when the weakest students are typically shunted into vocational fields, to learn butchery, hairstyling, or mechanics. It is by this process that preexisting inequalities intensify.

As the teenager’s guide, I understood at once that the internship week was not going to teach her how a newsroom works or what an article is. The urgency was elsewhere: the girl, who spelled like an 8-year-old, had no idea what to expect from her life the following year. Her father had left, and her mother was illiterate. She struggled just to envision what her options were, what it meant to go to a regular high school or to choose a vocational path—choices that were nevertheless decisive for her future.

I could see that the initiative to compensate for lack of opportunity was rewarding for those on the giving end and totally inadequate to the reality of the situation. Still, she told me she had the best week of her life, and we agreed to meet a few weeks later, after meeting with her high school’s career counselor. I waited; she didn’t show up.

Then, a few years into the experiment, the organizers of the program began demanding proof that students in fact come from disadvantaged backgrounds. Here, too, some middle-class parents had found ways to get their children a cool internship. Petty exploitation like this confirms what Charles de Gaulle once observed: “The desire for privilege and the taste for equality [are] dominant and contradictory passions of the French of all times.” Given that, it will take a long time before any type of affirmative action will bring about more equality.

Next September, the commission on social diversity appointed by Macron will return its conclusions. It will address the best way to make the brilliant but disadvantaged latchkey kids from the banlieues or remote rural areas succeed—but it shouldn’t stop there. It should also address diversity in every sense, including diversity of thought. We should think of what is being taught: values, ambitions, morals—as well as who’s being taught, and by whom.

Diversity should be an objective both at the beginning and at the end of the curriculum. Only then can France’s top schools accomplish what a previous incarnation of Sciences Po sought to achieve: to come up with more than “conservative platitudes” to oppose the “revolutionary platitudes” of the crowd.