The Dismal Science

Stuck in the ’50s

Men still don’t like it when their wives out-earn them.

Do we still have a problem with women in the workplace?
Do we still have a problem with women in the workplace?

Ingram Publishing/Thinkstock.

Nearly half a century ago, Betty Friedan’s Feminine Mystique helped kick off the modern feminist movement. Friedan described the widespread melancholy of women who felt trapped by the notion that they could find fulfillment only as wives and mothers. Women have come a long way since then—more women are working and fewer getting married, just as Friedan might have hoped. Yet many working women seem to suffer from a despondency similar to the one that Friedan attributed to a life as a homemaker. And despite an initial closing of the gender wage gap, differences persist between what men and women with similar skills and education earn.

A preliminary study by University of Chicago and Singapore economists argues that there may be a single explanation that lies behind all of these trends, one related to the very slow-moving social identities of men and women and how these identities collide within a marriage or romantic partnership. Though in 2012 women are better educated and higher-skilled than men, in identity terms we’re still stuck in a 1950s world in which “men work in the labor force and women work in the home.” These ’50s era gender identities prevent many men from partnering with women who out-earn them and creates friction for the men who do. The study’s findings suggest that a focus on gender discrimination in the workplace won’t be nearly enough to create gender equality in America.

Economists Marianne Bertrand, Emir Kamenica, and Jessica Pan have imported social identity theory from social psychology to study the evolution of gender relations over the past few decades. Social identity theory holds that individuals define themselves as members of groups, each of which has a set of behaviors that are expected of their members. Deviating from these expectations is costly, so if, as Friedan argued, the identity of women is defined by housework, taking a job outside the home would be a source of angst and torment. And if a man’s identity is defined as breadwinner, seeing his wife bring home a bigger paycheck is going to create problems for him, and in turn for the marriage.

The authors document some striking patterns in who marries whom, how husband and wife divide their household responsibilities, and which marriages survive, all of which suggest that the gender identities that Friedan documented 50 years ago continue to cast a very long shadow.

They begin by analyzing the earnings of young married couples in 2010—the men and women who should be least subject to attitudes and values of bygone eras. Using data from the American Community Survey, they compare wives’ and husbands’ incomes within a marriage. In the absence of any male insecurities about their spouse’s pay, there should be about the same number of families in which the wife earns a little more than her husband as those in which she earns a little less—which partner earns more in these marriages of near-equals should be a matter of random chance, depending on who got the most recent raise, who lucked into a slightly better job, and so on. Yet the data reveal a very different story. The authors estimate that in about 13 percent of families, the husband barely out-earns his wife, while the wife earns slightly more than her husband only 8 percent of the time. And in looking at couples’ earnings data since 1970, the researchers found that while women’s wages have gone up overall, the gap between the number of families in which men earn slightly more than their wives and those in which they earn slightly less has stayed more or less constant.

This increase in the overall earning capacity of women makes it that much harder for men to find partners they can safely out-earn. The authors predict that this could have two implications: First, fewer women may get married, and second, high-skilled women may take jobs they’re overqualified for to satisfy the roles imposed by gender identities—or they may not work at all. To test the first of these predictions, the authors examine marriage rates between 1970 and 2010 using Census data and find that in regions and in periods of time when women’s wages go up relative to men’s, there are correspondingly fewer weddings. They estimate that nearly a third of the drop in marriage rates since 1970 may be attributed just to the rising wages of women relative to men. To test the second hypothesis, they generate a prediction for what a given woman should earn, based on attributes like education and job training as well as local economic conditions. They then compare this predicted wage to her husband’s actual reported income. They find that women who, based on their skills and backgrounds, would be expected to out-earn their husbands are more likely to exit the labor force—a very costly way of maintaining gender identities and restoring familial harmony. Among those women who remain in the workforce, their actual earnings fall short of what the economists’ wage model would have predicted for them, especially if their predicted income was higher than their husband’s. While no individual piece of evidence provides dispositive proof that gender identity lies behind these trends—for example, women may marry less often when their earnings go up simply because they don’t need to rely on husbands economically—together, they present a compelling argument for the authors’ claims.

The authors find that high-earning women may also compensate for adopting “male” identities by putting in more hours cooking dinner, doing laundry, and otherwise taking care of the home. Using data from the American Time Use Survey, the researchers find that wives who earn more than their husbands actually report doing a greater fraction of household work. Standard economic theory would predict the opposite: The family’s high-earner should spend her time generating labor income, not cooking dinner. But when viewed through the lens of gender identity, it’s a natural outcome of trying to compensate for acting too much “like a man.”

And so, we end up hearing about women making cookies for the school bake sale at 4 a.m. before heading to Wall Street, and reading stories questioning whether it’s really possible to “have it all.” “Having it all”—being the perfect homemaker and employee—makes it hard to find time for intimacy, stresses both parents out, and generally puts an enormous strain on a marriage, and indeed the study’s closing finding is that couples in which a woman earns more than her husband are less likely to report that they’re happy in their marriage—and their unions are more likely to end in divorce.

This bleak picture suggests that the feminist revolution may have stalled not just because of continued workplace discrimination—though that’s surely a part of it—but rather because of gender identities at home, which lag well behind advances in women’s educational attainment and employment opportunities. In the end, it has proven much easier to change employment laws and workplace practices than to change ourselves.