But there is something odd in these stories about negative messages, hidden barriers, and gender prejudices. The way of science is to lay out every hypothesis that could account for a phenomenon and to eliminate all but the correct one. Scientists prize the ability to think up alternative explanations, and proponents of a hypothesis are expected to refute even the unlikely ones. Nonetheless, discussions of the leaky pipeline in science rarely even mention an alternative to the theory of barriers and bias. One of the rare exceptions was a sidebar to a 2000 story in Science, which quoted from a presentation at the National Academy of Engineering by the social scientist Patti Hausman:
The question of why more women don’t choose careers in engineering has a rather obvious answer: Because they don’t want to. Wherever you go, you will find females far less likely than males to see what is so fascinating about ohms, carburetors, or quarks. Reinventing the curriculum will not make me more interested in learning how my dishwasher works.61
An eminent woman engineer in the audience immediately denounced her analysis as “pseudoscience.” But Linda Gottfredson, an expert in the literature on vocational preferences, pointed out that Hausman had the data on her side: “On average, women are more interested in dealing with people and men with things.” Vocational tests also show that boys are more interested in “realistic,” “theoretical,” and “investigative” pursuits, and girls more interested in “artistic” and “social” pursuits.
Hausman and Gottfredson are lonely voices, because the gender gap is almost always analyzed in the following way. Any imbalance between men and women in their occupations or earnings is direct proof of gender bias—if not in the form of overt discrimination, then in the form of discouraging messages and hidden barriers. The possibility that men and women might differ from each other in ways that affect what jobs they hold or how much they get paid may never be mentioned in public, because it will set back the cause of equity in the workplace and harm the interests of women. It is this conviction that led Friedan and Clinton, for example, to say that we will not have attained gender equity until earnings and representation in the professions are identical for men and women. In a 1998 television interview, Gloria Steinem and the congresswoman Bella Abzug called the very idea of sex differences “poppycock” and “anti-American crazy thinking,” and when Abzug was asked whether gender equality meant equal numbers in every field, she replied, “Fifty-fifty—absolutely.”62 This analysis of the gender gap has also become the official position of universities. That the presidents of the nation’s elite universities are happy to accuse their colleagues of shameful prejudice without even considering alternative explanations (whether or not they would end up accepting them) shows how deeply rooted the taboo is.
The problem with this analysis is that inequality of outcome cannot be used as proof of inequality of opportunity unless the groups being compared are identical in all of their psychological traits, which is likely to be true only if we are blank slates. But the suggestion that the gender gap may arise, even in part, from differences between the sexes can be fightin’ words. Anyone bringing it up is certain to be accused of “wanting to keep women in their place” or “justifying the status quo.” This makes about as much sense as saying that a scientist who studies why women live longer than men “wants old men to die.” And far from being a ploy by self-serving men, analyses exposing the flaws of the glass-ceiling theory have largely come from women, including Hausman, Gottfredson, Judith Kleinfeld, Karen Lehrman, Cathy Young, and Camilla Benbow, the economists Jennifer Roback, Felice Schwartz, Diana Furchtgott-Roth, and Christine Stolba, the legal scholar Jennifer Braceras, and, more guardedly, the economist Claudia Goldin and the legal scholar Susan Estrich.63
I believe these writers have given us a better understanding of the gender gap than the standard one, for a number of reasons. Their analysis is not afraid of the possibility that the sexes might differ, and therefore does not force us to choose between scientific findings on human nature and the fair treatment of women. It offers a more sophisticated understanding of the causes of the gender gap, one that is consistent with our best social science. It takes a more respectful view of women and their choices. And ultimately it promises more humane and effective remedies for gender inequities in the workplace.
Before presenting the new analysis of the gender gap from equity feminists, let me reiterate three points that are not in dispute. First, discouraging women from pursuing their ambitions, and discriminating against them on the basis of their sex, are injustices that should be stopped wherever they are discovered.
Second, there is no doubt that women faced widespread discrimination in the past and continue to face it in some sectors today. This cannot be proven by showing that men earn more than women or that the sex ratio departs from fifty-fifty, but it can be proven in other ways. Experimenters can send out fake résumés or grant proposals that are identical in all ways except the sex of the applicant and see whether they are treated differently. Economists can do a regression analysis that takes measures of people’s qualifications and interests and determines whether the men and the women earn different amounts, or are promoted at different rates, when their qualifications and interests are statistically held constant. The point that differences in outcome don’t show discrimination unless one has equated for other relevant traits is elementary social science (not to mention common sense), and is accepted by all economists when they analyze data sets looking for evidence of wage discrimination.64
Third, there is no question of whether women are “qualified” to be scientists, CEOs, leaders of nations, or elite professionals of any other kind. That was decisively answered years ago: some are and some aren’t, just as some men are qualified and some aren’t. The only question is whether the proportions of qualified men and women must be identical.
As in many other topics related to human nature, people’s unwillingness to think in statistical terms has led to pointless false dichotomies. Here is how to think about gender distributions in the professions without having to choose between the extremes of “women are unqualified” and “fifty-fifty absolutely,” or between “there is no discrimination” and “there is nothing but discrimination.”
In a free and unprejudiced labor market, people will be hired and paid according to the match between their traits and the demands of the job. A given job requires some mixture of cognitive talents (such as mathematical or linguistic skill), personality traits (such as risk taking or cooperation), and tolerance of lifestyle demands (rigid schedules, relocations, updating job skills). And it offers some mixture of personal rewards: people, gadgets, ideas, the outdoors, pride in workmanship. The salary is influenced, among other things, by supply and demand: how many people want the job, how many can do it, and how many the employer can pay to do it. Readily filled jobs may pay less; difficult-to-fill jobs may pay more.
People vary in the traits relevant to employment. Most people can think logically, work with people, tolerate conflict or unpleasant surroundings, and so on, but not to an identical extent, everyone has a unique profile of strengths and tastes. Given all the evidence for sex differences (some biological, some cultural, some both), the statistical distributions for men and women in these strengths and tastes are unlikely to be identical. If one now matches the distribution of traits for men and for women with the distribution of the demands of the jobs in the economy, the chance that the proportion of men and of women in each profession will be identical, or that the mean salary of men and of women will be identical, is very close to zero—even if there were no barriers or discrimination.
None of this implies that women will end up with the short end of the stick. It depends on the menu of opportunities that a given society makes available. If there are more high-paying jobs that call for typical male strengths (say, willingness to put oneself in physical danger, or an interest in machines), men may do better on average; if there are more that call for typical female strengths (say, a proficiency wi
th language, or an interest in people), women may do better on average. In either case, members of both sexes will be found in both kinds of jobs, just in different numbers. That is why some relatively prestigious professions are dominated by women. An example is my own field, the study of language development in children, in which women outnumber men by a large margin.65 In her book The First Sex: The Natural Talents of Women and How They Are Changing the World, the anthropologist Helen Fisher speculates that the culture of business in our knowledge-driven, globalized economy will soon favor women. Women are more articulate and cooperative, are not as obsessed with rank, and are better able to negotiate win-win outcomes. The workplaces of the new century, she predicts, will increasingly demand these talents, and women may surpass men in status and earnings.
In today’s world, of course, the gap favors men. Some of the gap is caused by discrimination. Employers may underestimate the skills of women, or assume that an all-male workplace is more efficient, or worry that their male employees will resent female supervisors, or fear resistance from prejudiced customers and clients. But the evidence suggests that not all sex differences in the professions are caused by these barriers.66 It is unlikely, for example, that among academics the mathematicians are unusually biased against women, the developmental psycholinguists are unusually biased against men, and the evolutionary psychologists are unusually free of bias.
In a few professions, differences in ability may play some role. The fact that more men than women have exceptional abilities in mathematical reasoning and in mentally manipulating 3-D objects is enough to explain a departure from a fifty-fifty sex ratio among engineers, physicists, organic chemists, and professors in some branches of mathematics (though of course it does not mean that the proportion of women should be anywhere near zero).
In most professions, average differences in ability are irrelevant, but average differences in preferences may set the sexes on different paths. The most dramatic example comes from an analysis by David Lubinski and Camilla Benbow of a sample of mathematically precocious seventh-graders selected in a nationwide talent search.67 The teenagers were born during the second wave of feminism, were encouraged by their parents to develop their talents (all were sent to summer programs in math and science), and were fully aware of their ability to achieve. But the gifted girls told the researchers that they were more interested in people, “social values,” and humanitarian and altruistic goals, whereas the gifted boys said they were more interested in things, “theoretical values,” and abstract intellectual inquiry. In college, the young women chose a broad range of courses in the humanities, arts, and sciences, whereas the boys were geeks who stuck to math and science. And sure enough, fewer than 1 percent of the young women pursued doctorates in math, physical sciences, or engineering, whereas 8 percent of the young men did. The women went into medicine, law, the humanities, and biology instead.
This asymmetry is writ large in massive surveys of job-related values and career choices, another kind of study in which men and women actually say what they want rather than having activists speak for them.68 On average, men’s self-esteem is more highly tied to their status, salary, and wealth, and so is their attractiveness as a sexual partner and marriage partner, as revealed in studies of what people look for in the opposite sex.69 Not surprisingly, men say they are more keen to work longer hours and to sacrifice other parts of their lives—to live in a less attractive city, or to leave friends and family when they relocate—in order to climb the corporate ladder or achieve notoriety in their fields. Men, on average, are also more willing to undergo physical discomfort and danger, and thus are more likely to be found in grungy but relatively lucrative jobs such as repairing factory equipment, working on oil rigs, and jack-hammering sludge from the inside of oil tanks. Women, on average, are more likely to choose administrative support jobs that offer lower pay in air-conditioned offices. Men are greater risk takers, and that is reflected in their career paths even when qualifications are held constant. Men prefer to work for corporations, women for government agencies and nonprofit organizations. Male doctors are more likely to specialize and to open up private practices; female doctors are more likely to be general practitioners on salary in hospitals and clinics. Men are more likely to be managers in factories, women more likely to be managers in human resources or corporate communications.
Mothers are more attached to their children, on average, than are fathers. That is true in societies all over the world and probably has been true of our lineage since the first mammals evolved some two hundred million years ago. As Susan Estrich puts it, “Waiting for the connection between gender and parenting to be broken is waiting for Godot.” This does not mean that women in any society have ever been uninterested in work; among hunter-gatherers, women do most of the gathering and some of the hunting, especially when it involves nets rather than rocks and spears.70 Nor does it mean that men in any society are indifferent to their children; male parental investment is a conspicuous and zoologically unusual feature of Homo sapiens. But it does mean that the biologically ubiquitous tradeoff between investing in a child and working to stay healthy (ultimately to beget or invest in other children) may be balanced at different points by males and females. Not only are women the sex who nurse, but women are more attentive to their babies’ well-being and, in surveys, place a higher value on spending time with their children.71
So even if both sexes value work and both sexes value children, the different weightings may lead women, more often than men, to make career choices that allow them to spend more time with their children—shorter or more flexible hours, fewer relocations, skills that don’t become obsolete as quickly—in exchange for lower wages or prestige. As the economist Jennifer Roback points out, “Once we observe that people sacrifice money income for other pleasurable things we can infer next to nothing by comparing the income of one person with another’s.”72 The economist Gary Becker has shown that marriage can magnify the effects of sex differences, even if they are small to begin with, because of what economists call the law of comparative advantage. In couples where the husband can earn a bit more than the wife, but the wife is a somewhat better parent than the husband, they might rationally decide they are both better off if she works less than he does.73
To repeat: none of this means that sex discrimination has vanished, or that it is justified when it occurs. The point is only that gender gaps by themselves say nothing about discrimination unless the slates of men and women are blank, which they are not. The only way to establish discrimination is to compare their jobs or wages when choices and qualifications are equalized. And in fact a recent study of data from the National Longitudinal Survey of Youth found that childless women between the ages of twenty-seven and thirty-three earn 98 cents to men’s dollar.74 Even to people who are cynical about the motivations of American employers, this should come as no shock. In a cutthroat market, any company stupid enough to overlook qualified women or to overpay unqualified men would be driven out of business by a more meritocratic competitor.
Now, there is nothing in science or social science that would rule out policies implementing a fifty-fifty distribution of wages and jobs between the sexes, if a democracy decided that this was an inherently worthy goal. What the findings do say is that such policies will come with costs as well as benefits. The obvious benefit of equality-of-outcome policies is that they might neutralize the remaining discrimination against women. But if men and women are not interchangeable, the costs have to be considered as well.
Some costs would be borne by men or by both sexes. The two most obvious are the possibility of reverse discrimination against men and of a false presumption of sexism among the men and women who make decisions about hiring and salary today. Another cost borne by both sexes is the inefficiency that could result if employment decisions were based on factors other than the best match between the demands of a job and the traits of the person.
But many of the costs of equality-of-outcome p
olicies would be borne by women. Many women scientists are opposed to hard gender preferences in science, such as designated faculty positions for women, or the policy (advocated by one activist) in which federal research grants would be awarded in exact proportion to the number of men and women who apply for them. The problem with these well-meaning policies is that they can plant seeds of doubt in people’s minds about the excellence of the beneficiaries. As the astronomer Lynne Hillenbrand said, “If you’re given an opportunity for the reason of being female, it doesn’t do anyone any favors; it makes people question why you’re there.”75