OVER THE past generation women have made substantial economic gains, even as progress on other measures of social equality has been uneven. Their average level of education has caught up to that of men across rich and poor countries alike. Indeed in much of the rich world the share of young women with a college degree is now above that of men. Income may be divided less equally across the workforce as a whole, but it has become more evenly spread between men and women. In America women account for nearly 30% of the top tenth of earners, up from 5% in the 1960s. That said, progress is far from complete. Gaps in labour-force participation and pay persist. The nature of the obstacles holding back further progress has changed. Although economics ought to be keenly interested in such matters, not least because of gender inequalities in the profession, it has not always been of much help in understanding them. That is changing, however, in ways that could transform the field.
This evolution was apparent in January, in a lecture given by Marianne Bertrand of the University of Chicago. Over the past few decades, gender gaps in the rich world have had ever less to do with overt discrimination, she argued, and ever more to do with women’s decisions. Their choice of degree subject is one. Jobs in science, technology, engineering and maths have smaller gender pay gaps than others. But men are around twice as likely as women to graduate in such fields. More powerful still is the effect of childbirth. The birth of a first child has essentially no effect on a man’s earnings trajectory. By contrast a woman experiences a profound and lasting hit to her pay. The motherhood penalty, suggested Ms Bertrand, is easily the largest remaining contributor to gender gaps in labour markets.
Men and women alike opt to become parents, of course. But the unequal effect on earnings reflects their different responses to childbirth. Women are more likely to leave the labour force or to switch to part-time work. They often choose jobs that allow more flexibility, and accept lower pay as a consequence. Some studies, for instance, suggest that women take jobs with shorter commutes, to make time for their care responsibilities. In France, noted Ms Bertrand, the sacrifice in earnings associated with such decisions is estimated to explain 10-15% of the gender pay gap. It is women’s greater willingness to accept these trade-offs that accounts for diverging labour-market fortunes.
Economists, historically, have let the matter rest there, chalking such choices up to rational self-interest. Perhaps families decide that women have a comparative advantage in child-rearing, and should handle the parenting while men focus on their careers. Gary Becker, the late Nobel prize-winning economist, argued that households specialise in this way. Alternatively, perhaps women’s choices simply reveal their preferences: for subjects other than maths, for instance; or for time spent caring for children, rather than long hours at the office. And such preferences, economists have generally assumed, are to be taken as given. De gustibus non est disputandum, they say: there is no accounting for tastes.
But perhaps there is. As Ms Bertrand noted in her lecture, other social sciences, like social psychology, reckon that preferences are socially determined. In this view, people’s choices are influenced by norms, which specify the roles and behaviours that are appropriate for men and women. Survey evidence shows that, across a broad range of rich and poor countries, both men and women support the view that men should be first in line for a vacancy when jobs are scarce. The level of support varies—for instance it is much higher in Egypt than in Switzerland. But even in Switzerland, roughly a fifth of women agreed with the statement, similar to the share of men. Gender gaps in maths scores are larger in places where gender attitudes are more conservative. This suggests that social influences matter.
Defying social norms is possible but costly. Men who sacrifice their careers to raise children while their partners work may bear emotional costs, if, say, they are seen as being unmanly. Similarly, women who are seen to put their career before their family may face an emotional toll, related to their own guilt or the judgment of others, because of their decision to flout gender norms. The choices made under these pressures are still voluntary, but they reflect the influence of a self-perpetuating gender bias. Human resources may be allocated across the economy in a way that reflects this bias, rather than people’s abilities. Economists know these biases exist. Historically, however, they have tended to regard them as blunt descriptions of the state of the world, rather than evolving social forces that influence behaviour.
Puncturing the patriarchy
Mitigating bias is not easy. Seemingly helpful interventions, such as generous maternity leave, can backfire if they reinforce the norm that women are natural caregivers. Better, Ms Bertrand argued, to favour measures that have no such implications, such as generous support for child care. Her own research suggests that a mother’s working status shapes her children’s perceptions of labour-market norms. Those surrounding men’s behaviour matter too. Survey evidence from Japan suggests that many men feel positively about taking paternity leave. But, because they wrongly believe other men do not feel the same way, they take less time off than they would like. Firms could potentially make their workers better off by, say, choosing to make paternity leave mandatory.
Ms Bertrand’s arguments may not seem particularly subversive. But they carry implications that extend beyond gender discrimination. Her analysis suggests that the decision to participate in a market is not simply about maximising utility given a set of tastes and constraints. Markets, rather, are part of a suite of fluid social forces that shape behaviour. Economists cannot claim to understand the markets until they understand those forces. ■
This article appeared in the Finance and economics section of the print edition under the headline “Progress of a sort”