There are a lot of things that “explain” inequality. Technology, finance, societal, and cultural changes have all played their part. In a new series, Counterparties takes a look at the various things that correlate with rising income inequality in order to ascertain how we got to this economy and where we might go from here. For story tips/comments/complaints email us at

Matt O’Brien wrote a good post last week on how the classic 1989 romcom When Harry Met Sally explains inequality. I’ll let him explain:

Working men and working women of similar education and income levels bond and marry over similarities—movies, art, paprikash—rather than the differences that defined traditional marriage (he works/she cleans). This new arrangement promises a better world, but it has increased inequality.

O’Brien mostly references a new working paper on assortative mating written by Jeremy Greenwood, Nezih Guner, Georgi Kocharkov, and Cezar Santos for the National Bureau of Economic Research. It looks at marriages from 1960 through 2005, comparing the levels of education of each partner.

The researchers found an increase in college graduates marrying college graduates, and high school graduates marrying high school graduates. Fewer people are marrying across education levels. This corresponds with an increase in the Gini coefficient (a way to estimate inequality) to 0.43 in 2005, from 0.34 in 1960.

Put another way, this rise in income inequality among households in the US is explained by the end of “Cinderella marriages” — the ones where women (traditionally) “marry up” above their parents’ class, James Pethokoukis writes. “Doctors no longer marry nurses, they marry fellow doctors or others of comparable education level,” he says.

Peter Orszag says this is a good example of the complexities that make fighting inequality so hard. He writes, “If income inequality is being driven in part by changes in marriage patterns, what can anyone do about that?”

Further, do we want to do anything about it? The increase in assortative mating also correlates with the number of women going to college and graduate school, an option that really wasn’t available to to them in 1960. Marriages in the mid-twentieth century were about division of labor: the man worked outside the home, the women kept the house and took care of the children.

That’s changed. As O’Brien says, people now marry because they’ve found a compatible life partner, someone they enjoy sitting down and watching Netflix with after a long day at work (for both of them). The correlated inequality (causation is hard to prove here) doesn’t seem worth giving that up.