Birds of a feather flock together, so it shouldn't be surprising that, when given the choice, intelligent and successful women tend to marry intelligent and successful men.
But until recently, most women in the U.S. didn't have that choice. They had far fewer educational opportunities than men. Those who excelled academically often were unable to translate their accomplishments into remunerative careers: Only about 35 percent of women aged 25 to 54 worked outside the home in 1950, compared with about 75 percent now. Economists now tell us that this social progress is responsible for a significant portion of the rise in household income inequality.
They start by tracking whether men and women have become more likely to pick marriage partners on the basis of their education level. They find that "the number of matches between husband and wife with the identical education level is larger than what would occur if matching was random" and that the tendency for men and women to engage in "assortative mating" has increased over time.
Here's the kicker: Assortative mating can explain the entire increase in U.S. household income inequality as measured by the Gini coefficient. (Click the link for a longer explanation. Short version is that a coefficient of 1 means one person holds all the wealth, while 0 means equal distribution.)
The reason is that women are more likely to work than in the past, and the women most likely to work are the ones with the best educations and the highest salaries. Women in poorer households are much less likely to work. Moreover, the gap between the labor force participation rates of poor married women and high-income married women was much wider in 2005 than in 1960.
According to the economists, had marriage patterns remained unchanged from 1960, today's empowered career women would have pumped up the incomes of households with men who were professionally beneath them and kept the U.S. Gini coefficient constant.
The economists test this by comparing the Gini coefficient against their model of what the Gini coefficient would have been if men and women married randomly. In 1960, the actual Gini coefficient in the U.S. was 0.34, almost indistinguishable from the Gini coefficient produced by their random model (0.33). That suggests there was very little assortative mating back then. (Think of all the executives who ended up with their secretaries.)
By 2005, the Gini coefficient was 0.43. That's a big increase. What's more, the economists found that "if people matched in 2005 according to the 1960 standardized mating pattern there would be a significant reduction in income inequality; i.e., the Gini drops from 0.43 to 0.35."
Intelligence, education and incomes tend to go hand in hand in a free society. Until quite recently, the ambitions of half the population were suppressed by law and culture. That made the distribution of income more equal than it would have been under a less-restrictive social system. Now that tens of millions of women have greater opportunities to live up to their potential and marry men who are their equals, household income inequality has increased. If that's the case, lowering inequality will be a lot harder than anyone imagines.
This column does not necessarily reflect the opinion of Bloomberg View's editorial board or Bloomberg LP, its owners and investors.