One of my first columns dealt with a 1989 report accusing Wellesley College of racial discrimination. The report contained no evidence of anything remotely close to real bias. Instead it focused on discomfort among non-white students — some cafeteria food was unfamiliar, posters in the bookstore featured Bavarian castles but no Third World settings. All of this was said to add up to subtle, "unconsciously white" bias that hobbled minority students.This was my first exposure to the modern evidence-free, feelings-based bias report. Perhaps the most successful of these reports was the 1999 report on gender bias at the Massachusetts Institute of Technology's School of Science. Gender bias operates in "a stealth-like way," said biology professor Nancy Hopkins, the driving force behind the report. "Stealth-like," "subtle" and "institutionalized" biases are the kinds you needn't document or even describe. They are just there.Hopkins was allowed to form her own panel, which conducted its own inquiry and — lo and behold — produced the report that found her charge accurate. The committee made a stab at gathering evidence, measuring offices and counting heads, but nothing convincing was put on the table. The panel said it looked at quantitative measures of academic achievement, but refused to make its data public.

Professor Judith Kleinfeld of the University of Alaska-Fairbanks, who did an analysis for the right-of-center Independent Women's Forum, said the MIT study "presents no objective evidence whatever to support claims of gender discrimination." Younger women teaching at MIT seemed to agree. Buried in the back of the report was the admission that "untenured women faculty feel that men and women faculty are treated equally."

No matter. America's diversity machine threw itself fully behind the report, shaky as it was. The New York Times gave it uncritical Page One coverage. The Ford Foundation donated a million dollars to see if other universities needed the MIT treament. Hopkins was invited to the White House. MIT capitulated, accusing itself of deep (but "subtle" and vague) gender bias.

Now the gender bias analysts at MIT are back in the news, this time with reports on the four MIT schools not mentioned before. Again, the text is awash in feelings-based prose, with "women more frequently reporting negative experiences." Women "often feel undervalued" and "expressed feelings of exclusion." The word "marginalization," mandatory in bias reports, appears relentlessly. Subtle, hard-to-get-at sexism was again found to be the culrpit: "The discrepancies in treatment of male and female faculty have much more to do with small unconscious biases than blatant sexism."

Why do some MIT women feel marginalized? According to a report co-authored by James Steiger, a statistician and professor of psychology at the University of British Columbia, MIT senior women in biology, on average, published fewer scientific articles, were cited less in other scientists' work and brought in less grant money than the male scientists. All the women at MIT are very good. All would be stars elsewhere. But MIT is a very fast track, so some are below average for a campus that sets standards that high. Men in this predicament can't chalk it up to gender. Women can.

One MIT report tosses in good news then quickly paddles past it. It mentions that the engineering school has a history of giving tenure to a higher percentage of women than men, and quicker promotions to full professor as well. But then the report lurches back to chatter about marginalization. Even substantial pay raises for women are disparaged on the grounds that if they had been given earlier, more pension or retirement money would have accumulated.

The 200-odd pages of the reports are laid out in a way that makes it hard for journalists to encapsulate what is being said. One example, though: At one point the management report attempts to measure the comparative discomfort levels of female and male professors. It's a wildly unscientific effort. In general, said Steiger, "any undergraduate in a research methods course could debunk these reports."

The sad truth is that MIT, one of the world's great centers of scientific education, has now produced and accepted two astonishingly unscientific studies of its own administrative behavior. In response to these studies, nobody on campus has spoken out. Worse, the culture of MIT is being changed. Gender equity has replaced scientific merit as the value administrators will be judged by. As always in affirmative action schemes, women on the faculty will now come under suspicion as people who wouldn't be there except for politics. And all without any real discussion or open debate. Amazing.