Even before they have kids, men and women often do different things that may affect earnings. A year out of college, notes AAUW, women in full-time jobs work an average of 42 hours a week, compared to 45 for men. Men are also far more likely to work more than 50 hours a week.

Buried in the report is a startling admission: "After accounting for all factors known to affect wages, about one-quarter of the gap remains unexplained and may be attributed to discrimination" (my emphasis). Another way to put it is that three-quarters of the gap clearly has innocent causes -- and that we actually don't know whether discrimination accounts for the rest.

I asked Harvard economist Claudia Goldin if there is sufficient evidence to conclude that women experience systematic pay discrimination. "No," she replied. There are certainly instances of discrimination, she says, but most of the gap is the result of different choices. Other hard-to-measure factors, Goldin thinks, largely account for the remaining gap -- "probably not all, but most of it."

The divergent career paths of men and women may reflect a basic unfairness in what's expected of them. It could be that a lot of mothers, if they had their way, would rather pursue careers but have to stay home with the kids because their husbands insist. Or it may be that for one reason or another, many mothers prefer to take on the lion's share of child-rearing. In any case, the pay disparity caused by these choices can't be blamed on piggish employers.

June O'Neill, an economist at Baruch College and former director of the Congressional Budget Office, has uncovered something that debunks the discrimination thesis. Take out the effects of marriage and child-rearing, and the difference between the genders suddenly vanishes. "For men and women who never marry and never have children, there is no earnings gap," she said in an interview.

That's a fact you won't hear from AAUW or the Democratic presidential candidates. The prevailing impulse on Equal Pay Day was to lament how far we are from the goal. The true revelation, though, is how close.