French Firm Tests Colorblind Hiring

"The French people think the immigrants should always be at the service of the French," said Akremi, 37, who works at a Parisian cybercafe. He said he had sent out 1,000 résumés but never been called for a single interview, even though he speaks French, English and Arabic. The only jobs he has held he acquired through the Arab community in Paris, and most were minimum-wage, Akremi said.

It was France's refusal to confront discrimination or even discuss it that led the Montaigne Institute, a private Paris research center founded by Bebear five years ago, to study the issue. The goal was to "help solve it by describing accurately what it is, and breaking the taboos and the mythology about the integration of individuals regardless of their national origin," said the institute's director general, Philippe Maniere. "We live in a world of prejudice and myth. We wanted to show that discrimination exists and help France correct itself."

A report the institute published in 2004 cited a study showing that a job applicant with a French-sounding name was five times more likely to be called for an interview than someone with an identical résumé but an Arabic- or North African-sounding name. That bias contributed to unemployment rates as high as 40 percent -- four times the national average -- among young men in low-income communities outside Paris, according to the report.

Armed with such data, Bebear and the Montaigne Institute have persuaded about 300 companies -- including the Total energy group, the car manufacturer Peugeot Citroen, the steel giant Arcelor and SNCF, the national railway -- to sign a charter pledging to oppose discrimination and make their companies "reflect the diversity of France."

Among the recommendations in an accompanying report were that companies use anonymous résumés in hiring, produce annual reports charting progress in ending discrimination, and create internships to start young minority people on career paths.

Bebear said the proposals were meant to correct inequities without American-style affirmative action programs, which are illegal under French law and viewed by many people here -- Bebear included -- as antithetical to French culture and society.

"If I hire someone not because he's competent but because he's black, immediately everyone in the company is going to think that blacks are unqualified," Bebear said. To him, affirmative action should principally mean better education, so that minorities will have "the same level of confidence as the others."

AXA adopted the use of anonymous résumés in January 2005. Peugeot began using them about eight months ago. Other companies have expressed interest in the idea and are working toward implementing it, but they are having problems devising software and other systems to ensure anonymity across large, diversified businesses, officials said.

Another problem is that, under French law, companies are not allowed to build databases that link the names of employees with their race or ethnic group. That, said Maniere, the Montaigne Institute director, is a deeply rooted reaction against the roundup of people from different ethnic groups during World War II.

Mehaignerie, the aide to the equal opportunities minister, said anonymous résumés were "a good tool," but she cautioned: "It seems to me we won't be able to go very far if we can't have data. If you can't know who's being recruited and how they're being promoted, how can you validate who's doing good and who's not?"

Samuel Thomas, vice president of the anti-discrimination group SOS-Racisme, said there were many ways to gauge whether a company discriminates in hiring without fancy software programs and extensive databases.

"So many people want to analyze more than they want to do," he said, calling anonymous résumés "a very important solution" to the problem of achieving colorblind hiring. "The first problem is not to study discrimination," he said. "The first problem is to fight discrimination and stop it."

Special correspondent Caroline Hout contributed to this report.