A game of mirrors? Economists’ models of the labor market and the 1970s gender reckoning

Written with Cleo Chassonnery-Zaigouche and John Singleton

The underrepresentation of women in science is drawing increasing attention from scientists as well as from the media. For example, research examining glass ceilings, leaking or small pipelines, the influence of mentorship, biases in refereeing, recommendations, and styles of undergraduate education or textbooks are flourishing in STEM, engineering, social sciences, and the humanities. Economics is no exception, as a paper that drew widespread coverage by Alice Wu released in the summer of 2017 exemplified. One thing that nevertheless sets economics and (to greater and lesser extents) its cognate disciplines apart, however, is that research topics such as the gender wage gap, women’s labor supply, and labor market discrimination are phenomena that many researchers in these areas both experience and study. An obvious question raised, therefore, is how the theories, models, and empirical evidence that economists develop and produce in turn shape their understanding of gender issues within their profession. Early debates surrounding the foundation of the Committee on the Status of Women in the Economics Profession (CSWEP) in 1972 are revealing in this regard.

'Women Of The World Unite!'The foundation of CSWEP, which we briefly narrated here, stood at the crossroads of various historical and social trends. One was the growing public awareness of discrimination issues and an associated shift within the US legal context. The Equal Pay Act of 1963 and the last-minute inclusion of gender in the 1964 Civil Right Act brought about a stream of sex discrimination cases, including the famous Bell AT&T case, whose settlement benefitted 15,000 women and minority employees. Phyllis Wallace, a founding member of CSWEP, was the expert coordinator for the Equal Employment Opportunity Commission. Though the 1964 Act excluded employees of public bodies, including governmental and universities hires, legal battles at Ivy Leagues universities resulted in compliance rules and the 1972 Equal Employment Opportunity Act. Professional societies were no exception. Beginning in 1969, the American Historical Association and the American Sociological Association established committees on the status of women in their respective disciplines. Chairing the sociology committee was Elise Boulding, whose husband Kenneth later joined as a founding member of CSWEP. Kenneth Boulding would draft “Role Prejudice as an Economic Problem,” the first part of a paper introducing CSWEP’s first report, “Combatting Role Prejudice and Sex Discrimination.”

Many of the actions pursued by economists were indeed similar to (and inspired by) other professional and academic societies, such as making day care available at major conferences, creating mentorship programs, and developing a roster of women in every field in economics to chair and participate in conferences panels. But other issues were idiosyncratic to economics. In particular, the problems of gender bias in economics were viewed as economic issues from the beginning, as seen in Boulding’s 1973 article on sex Boulding explaining CSWEP creationdiscrimination within the profession (picture on the left). Early CSWEP reports routinely framed their organizational efforts as attempts to study and fix the “supply and demand for women economists,” that is, the labor market for economists. The framing applied in the reports echoes the objectives of the AEA Committee on Hiring Practices to establish better recruitment practices and the preceding work by the Committee on the Structure of the Economic Profession. The reports relied on the logic of basic econ 101 principles at times, but on other occasions, the CSWEP originators, most of whom were trained in labor economics, delved more deeply into ongoing debates on the interpretation of earnings differentials, the determinants of women’s labor supply, the extent of discrimination and causes of occupational segregation, and on ways to fix the labor market for economists. One particularly revealing occasion was a letter exchange between Carolyn Shaw Bell, usually hailed as the driving force behind the women’s caucus that led to CSWEP’s creation, and University of Chicago economist Milton Friedman.

Bell vs Friedman 

29shaw_190Carolyn Shaw Bell (1920-2006) had received her Ph.D in 1949 from the London School of Economics and would spend her academic career at Wellesley College. After war work at the Office of Price Administration with Galbraith, she later did empirical work on innovation and income distribution, and contributed to consumer economics. Bell was convinced to accept the inaugural chairpersonship of the Committee (rather than retire) after the American Economic Association voted to establish the CSWEP in 1971 and launch an annual survey of women economists. In the summer of 1973, she sought to organize session at the December ASSA meeting. She wanted to assemble a panel of economists from various, sometimes opposed, backgrounds to comment on the findings. She therefore asked Elizabeth Clayton, a specialist of Soviet economics at the University of Missouri, leftish labor economist David Gordon of the New School for Social Research, and Milton Friedman to participate. CSWEP members expected “out in the open” controversy from the panel.


In an August reply to Bell’s invitation, Friedman declined, as he was not planning to attend the meetings (he was to be replaced by George Stigler on the panel). He did so regretfully, he explained, because he held strong views on the CSWEP report. He especially disagreed with the statement that “every economics department shall actively encourage qualified women graduate students without regard to age, marital or family status.” Though he “sympathize[d] very much with the objective of eliminating extraneous considerations from any judgment of ability or performance potential,” Friedman confessed he “never believed in reverse discrimination whether for women or for Jews or for blacks.” To this list he later added discrimination against conservative scholars, which he believed was strong on college and university campuses.

Whether preferential treatment produced “reverse discrimination” against the majority groups was a key point of contention over affirmative action policies. The social context was politically charged. While ending some of the “Great Society” programs, the Nixon Administration also set up the first affirmative action policy in 1969: the “Philadelphia Plan,” required that federal contractors and unions meet targeted goals for minority hires. The Nixon policy was sold as “racial goals and timetables, not quotas,” but criticisms focused on de facto quotas and applicability. Though formal quotas in US universities did not exist for women and racial quotas were ruled out by a 1978 decision, formal and informal affirmative action were debated in similar ways: Does encouraging women and minority applicant discouraged white men to apply? Was the goal of equal opportunity equal representation? These were recurring questions.

Screen Shot 2018-03-06 at 11.59.32Friedman’s answer, elaborated in his reply to Bell, was straightforward: affirmative action is inefficient and unethical: “should we… encourage men age 65 to enter graduate study on a par with young men age 20?”, he asked “Surely training in advanced economics is a capital investment and is justified only if it can be expected that the yield from it will repay the cost… Individuals trained do not bear the full cost of their training. We have limited funds with which to subsidize such training; it is appropriate to use those funds in such way to maximize the yield for the purpose for which the funds were made available. In the main, those funds were made available to promote a discipline rather than to promote the objectives of particular groups,” he continued. “It is relevant to take into account the age of men or women, the marital or family status of men or women, and the sex of potential applicants insofar as that affects the likely yield from the investment in their training,” Friedman emphasized, in an argument that strongly echoed Becker’s human capital theory: prohibiting the use of criteria such as gender and race in investment decisions was inefficient if they contributed to correctly predicting returns.

Overall, Friedman concluded, equal opportunity would not yield equal representation or “balance”:

I have no doubt that there has been discrimination against women. I have no doubt that one of its results has been that those women who do manage to make their mark are much abler than their male colleagues. As a result, it has seemed to me that a justified impression has grown up that women are intellectually superior to men rather than the reverse. I realize this is small comfort to those women who have been denied opportunities, but I only urge you to consider the consequences of reverse discrimination in producing the opposite effect.

Bell outlined the reasons for her disagreement with Friedman in lengthy response. She insisted that CSWEP favored non-financial “encouragement” over “any preferential financial aid for women.” More generally, while she agreed that the “free market lessens the opportunities for discrimination inasmuch as competition gives paramount recognition to economic efficiency,” she contended that this reasoning only applied to goods, not to human beings. She agreed with Friedman regarding the criteria for investing in professional training, but objected that there was nothing “in [Friedman’s] statement, in the discipline per se or in the existence of scarce resources, to identify those recipients who will, in fact, contribute most to the field.” Instead, she argued, the recipients of investment were selected by those “controlling the awards who learned certain cultural patterns, including beliefs about sex roles.”

Bell went on to admonish economists to re-examine their own biases: like the employers and employees they studied, they had been “brainwashed”: “beginning in the cradle. children… learn over and over again that what is appropriate and relevant for boys is not necessarily appropriate and relevant for girls.” These societal norms were biasing market forces, in that they influenced both supply and the demand of labor, she explained. Career and family expectations, decisions of whether or not to invest in education and training, the choice of education and occupation, as well as the allocation of time are all distorted. “This means that the occupations followed by young men and women do not reflect market considerations,” she concluded, only to add that “until we have a society where little girls are not only able to become dentists and surveyors and readily as little boys but are expected to become dentists and surveyors as readily as little boys we cannot in all conscience rely on the dictates of economic efficiency to allocate human beings.”

In a rejoinder to Bell’s reply, Friedman reprised one of his most famous arguments: market solutions should be preferred because alternatives systematically lead to the tyranny of the majority:

Screen Shot 2018-03-06 at 12.18.38

In short, Bell was advocating for institutional changes to the professional training of and labor market for economists (procedures in a very concrete way, cf. premise of JOE), while Friedman was arguing political philosophy. In her final response, Bell rejected the notion that actively countering “the existing system of brainwashing” through affirmative action would be useless, arguing instead that present discrimination resulted in capital investment “which may reduce the mobility of other resources in the future” and therefore was inefficient. Finally, Bell insisted that the CSWEP report advocated for voluntary participation in affirmative action plans, and that the “mild suggestions” proposed were far from the “dictatorial imposition of power” that worried Friedman.


Conflicting models of the labor market

The exchange quoted above reveals how much thinking about the status of women in economics and what should be done about it was embedded in wider economic debates on how to model the role of women in the economy. The initial focus of Bell and Friedman’s exchange was the plan advanced by CSWEP. They both tacitly considered it a special case of the larger debate on whether affirmative action would advance the status of women in the US economy, the disagreement deriving from their respective visions of the labor market, of how agents make economic decisions, and the extent to which gaps in outcomes and other phenomena reflected discrimination. Moreover, the 1970s were a decade in which the field was pervaded with thorny debates, some which reflected rapid changes in US labor market themselves.

13342748403Friedman’s arguments drew on a vision of the labor market that was becoming dominant at that time and that he had contributed to shaping through his famous 301 Price Theory course at the University of Chicago. As he was jousting with Bell, Becker’s 1957 Economics of Discrimination had just been republished with a fanfare that contrasted sharply with the resistance the book had encountered 15 years earlier (Friedman had to put considerable pressure on Chicago University Press for them to publish his former student’s Ph.D. dissertation). The main thread of Becker’s work, one partly inspired by Friedman himself, was to model some employers as rational maximizers with a taste for discrimination. That is, they used “non-monetary considerations in deciding whether to hire, work with, or buy from an individual or group.” Those employers, he contended, were disadvantaged, as their taste acted as a tariff in a trade model. Discrimination was thus an inefficient behavior that would be pushed out of competitive markets in equilibrium. The notion that the labor market where employees, being rational about their human capital investment and their work/leisure tradeoff, meet cost-conscious employers, was efficient, pervades his exchange with Bell.

But Friedman’s vision exhibited an additional characteristic historical and political twist. The proof that markets were, in the long run, efficient was historical: they had brought improvements in the living conditions of Jews, African-Americans, and Irish people throughout decades and centuries. And the market did so by protecting them from the tyranny of the majority, so that any attempt to fiddle with the market to accelerate the transition was bound to failure. This argument had come to maturity in Capitalism and Freedom (1962). In the fifth chapter, Friedman took issue with Roosevelt’s 1941 Fair Employment Practice Committee (FEPC), tasked with banning discrimination in war-related industries. He wrote:

If it is appropriate for the state to say that individuals may not discriminate in employment because of color or race or religion, then it is equally appropriate for the state, provided a majority can be found to vote that way, to say that individuals must discriminate in employment on the basis of color, race or religion. The Hitler Nuremberg laws and the laws in the Southern states imposing special disabilities upon Negroes are both examples of laws similar in principle to FEPC.

Like Becker, Friedman believed this general framework applied to any kind of discrimination: against Jews, foreigners, women (his correspondence with Bell echoed Capitalism and Freedom almost verbatim), people with specific religious of political beliefs, and blacks. In a lengthy interview with Playboy the previous year, for instance, he again explained that “it’s precisely because the market is a system of proportional representation [as opposed to the majority rule in the political system] that it protects the interests of minorities. It’s for this reason that minorities like the blacks, like the Jews, like the Amish, like SDS [Student for Democratic Society], ought to be the strongest supporters of free-enterprise capitalism.

Bell’s vision of labor markets, on the other hand, was not a straightforward reflection of a stabilized research agenda. The field was buzzing with new approaches in these years and her work stood at the confluence of three of them: attempts to understand the consequences of imperfect information on labor supply and demand; new empirical evidence on the wage gap and on the composition, income, and occupation of American households; and the challenges brought to mainstream economic modeling by feminist economists.

For example, her letters and the solutions that she pushed to fight the poor representation of women in economics betray a concern with the consequences of imperfect information on access to employment opportunities, expectations, employers’ and employees’ behavior, and in the end, the efficiency of market outcomes. In this regard, some of her statements closely echoed Arrow’s theory of statistical discrimination, which he had elaborated in a paper given in Princeton in 1971. His idea was that employers use gender as a proxy for unobservable characteristics: beliefs on average characteristics of groups translate into discrimination against individual member of these groups. As Bell and Friedman were corresponding (a correspondence that Bell circulated to the other CSWEP members and to Arrow), Arrow was refining his screening theory, in which preferences were endogenous to dynamic interactions that may create self-selection, human capital underinvestment, segregation and self-fulfilling prophecies. Arrow’s awareness to information imperfection hadn’t prevented him from telling Bell that, as AEA president in charge of the 1971 conference program, he couldn’t find many qualified women to raise the number of female presenters and discussants. Bell quickly came up with 300 names, 150 of whom expressed interest in presenting a paper. Bell’s advocacy for formal procedures producing information on jobs but also on “qualified women” in economics contributed to the establishment of a CSWEP-sponsored roster of women economists and of the JOE soon afterwards.

Bell also participated in the flourishing attempts to document women and household behavior empirically. While there is no evidence that she was then aware of Blinder and Oaxaca’s attempts to use data from the Survey of Economic Opportunity (which eventually became the PSID) to decompose the wage gap, she was involved in commenting on the 1973 Economic Report to the President, in which the Council of Economic Advisors had included a chapter on “the economic role of women.” Her familiarity with the economic and sociological empirical literature on earnings differentials subsequently led Bell to gather a substantial body of statistics on US families. She summarized her results in “Let’s Get Rid of Families,” a 1977 article she decided to publish in Newsweek rather than in an academic journal. That she aimed to challenge the notion that the typical US family was built around a breadwinning father and a stay-at-home mother was already seen in her insistence that economists and citizens alike are “brainwashed” by social norms and beliefs.

Finally, Bell’s letters to Friedman suggest that she wasn’t merely looking for a microeconomic model alternative to the Beckerian theory of discrimination. Her whole contribution was embedded in a more radical theoretical criticism of economic theory, one she would later carefully outline in a 1974 paper on “Economics, sex and gender.” The way that economic decisions are modeled does not allow an accurate representation of how women make economics choices as producers and consumers, she argued. Agents are making a choice between work and leisure, yet “leisure” in fact covers many types of work between which women need to allocate their time. Likewise, women usually did not consume an income they had independently earned, as standard micro models assumed. Not taking into account this variety of decisions in fact reinforces the social model economists tend to take as given, in which women are primarily caregivers who don’t exercise any economic independent choice. “Both economic analysis and economic policy dealing with individuals, either in their roles as producers or consumers, have been evolved primarily by men,” she concluded.


Bell and Friedman’s divergent takes on which actions the newly-established CSWEP should implement were thus inextricably intertwined with their theoretically and empirically-informed views of the labor market. Their exchange further reveals that their views were also tied to their respective methodological beliefs and personal experiences. Their arguments exhibited different blends of principles and data, models and action. Friedman was primarily focused on foundational principles about markets and the free society and explicitly connected them to the discrimination he had experienced as a Jew and a conservative. When the first letter he had sent to Bell was reprinted in a 1998 JEP issue celebrating the 25th anniversary of CSWEP, he further explained that “the pendulum has probably swung too far so that men are the ones currently being discriminated against.” Bell, by contrast, grounded her defense of the CSWEP agenda in economics principles, economic facts, beliefs and prejudices found in American society, as well as in her interactions with AEA officials and decades at Wellesley of tirelessly mentoring women in economics.

Note: Permission to quote from the Friedman-bell correspondance was granted by the Hoover Institution


Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s