After all these years I am still amazed at the persistence of people in believing things that are unproved or that are more complicated than they realize. My most recent example is Pay Equity Day, called to protest the fact that women's pay is only 74 percent of men's. Pay Equity Day was April 8. Perhaps you missed it.
Why make pay equity the subject of a protest? Surely Pay Equity Day's organizers don't think that everyone should get the same pay. I'd guess that my housekeeper's pay is less than 74 percent of the pay of the organizers. Why don't they protest that fact? Presumably they would explain that they are more productive than my housekeeper. But the only evidence they have of their superior productivity is the fact that someone is willing to pay them more than anyone is willing to pay my housekeeper.
One might think that would be the end of the matter. Men get paid more than women because someone is willing to pay them more, just as Pay Equity Day's organizers are paid more than my housekeeper because someone is willing to do it.
But, of course, that isn't the end of the matter. If we accept productivity as a proper measure of what people should earn, we have to consider the possibility that employers' willingness to pay is a wrong measure of workers' productivity. It could be wrong in either of two senses. Employers may have an incorrect estimate of the relative productivities of men and women. Or, having a correct estimate of the relative productivities, they may want to pay women less than their productivity merits because they have some prejudice against women in the workplace.
In a perfect world we could compare the relative earnings of women and men with their relative productivity. But there is no good way to measure relative productivity--at least, no better way than looking at relative earnings, which only leads us back to where we started. So, students of the subject approach it indirectly, comparing the earnings of men and women who are similar in the respects that contribute to productivity: They compare the incomes of men and women of the same age, the same years of work experience, the same years of education, and in jobs of the same stress, riskiness, and difficulty. When they do, they generally find that the gender gap in earnings remains but is smaller than the gap for women and men in total. But the results are difficult to interpret. The number of years devoted to education and the number of years of experience, for example, do not make the same contribution to productivity. And it is never possible to be sure that you have taken account of all the factors that determine productivity.
Suppose you compared the earnings of men and women who are 35-year-old lawyers, all childless, and all Law Review graduates of Harvard Law School, and found that the men's earnings are higher than the women's. (I'm making this up.) Would this mean that there is discrimination against women? We would also find that all the women in this category do not earn the same salary. There is something, some X Factor, other than the conditions I have listed and other than gender, which explains the difference in salaries among the women. But if this X Factor is unequally shared by men and women, there will be a difference in the average earnings of men and women that does not reflect discrimination.
Maybe the X Factor is height. A taller lawyer can reach the books on the top shelf without a ladder. If all women were paid the same as all men of the same height, the average pay of men would be higher than that of women because the average man is taller. But there would be no discrimination.
Much of what we know about the economic status of women is summarized in an excellent monograph by Diana Furchtgott-Roth and Christine Stolba, Women's Figures: The Economic Progress of Women in America. One striking fact presented there is that childless women at age 30 earn 95 percent as much as men at the same age, whereas mothers earn only 75 percent as much as men. This suggests that the relatively low earnings of many women are related to their relatively low productivity because child-bearing and child-rearing interferes with their productivity in the marketplace. But another force may be at work. Women who can command earnings that are 95 percent as high as those of men may decide not to have children rather than forgo those earnings. We don't know how the earnings of these women compare with their productivity. If their productivity is 10 percent higher than that of men, the fact that they are paid only 5 percent less would not show that there is little discrimination.
There are three gaps involved in this discussion:
1) The gap between the earnings of men and women.
2) The gap between the earnings of women and their productivity.
3) The gap between the productivity of women and the productivity of men.
There is plenty of evidence that the first gap has been declining--fairly rapidly by historical standards. There is no good way to measure the gap between women's earnings and their productivity, but it is reasonable to say that their earnings have risen pretty much in line with their productivity. At least, it seems clear that the earnings of the total labor force have risen pretty much in line with productivity (output per hour of work) when measured correctly. Women are so large a part of the labor force that it is hard to believe that this could be true of the total if it were not also true of women. If the gap between the earnings of women and men is declining, and if the earnings of women are rising in line with their productivity, it follows that the productivity of women has been rising relative to the productivity of men. That would be consistent with what we know about changes in the character of women's education and their distribution among occupations.
Whether this combination of facts and speculation is grounds for demonstrations of protest is a matter of taste.
I think there is a problem lurking here, but it is not the one the protesters have been protesting. In an earlier age, when incomes in the market were lower than they are now, the cost to a parent of forgoing market employment in order to stay home and care for a child was also lower than today. And when the men-women wage gap was lower, the clear choice was for the woman to stay home and look after the child. But given higher market incomes, having and rearing a child is more expensive in terms of forgone income. And with the men-women wage gap narrowed it has become less clear how the staying home with the child should be divided between the mother and the father. This problem of family life is not a result of incomes being too low or the wage gap being too big. It is rather the reverse.