This article uses archival sources to reconstruct an alternate history of Milton Friedman's A Theory of the Consumption Function, spotlighting the contributions of his collaborators Margaret Reid, Dorothy Brady, and Rose Friedman. Although Milton Friedman offered public credit to his wife and their two close friends, none received formal recognition or reward for their contribution to the permanent income hypothesis. The article documents this hypothesis as an example in professional economics of the well-known “Matilda effect,” in which women's intellectual contributions are systemically devalued, while arguing it is important to distinguish between formal and informal credit. Further, the article connects the lower status of women's consumption economics to broader shifts in the economics discipline across the twentieth century.
To the graduate students she was not a hidden figure but a distant one, “always carrying a pile of books with statistical data.” In winter, she walked with particular care through the snow with a box of meticulously arranged punch cards, headed toward the computer center. To her colleagues she “carried her own weight,” even if she did not head her own workshop. Instead, she was a regular at others' workshops in the University of Chicago's vaunted economics department. “Milton gave her a certain amount of credit in his ‘permanent income hypothesis,’” reflected longtime department chair Arnold Harberger (Harberger and Edwards, forthcoming; Goldin 2021: 46). Indeed, her name surfaces in the introduction to Milton Friedman's A Theory of the Consumption Function, which presented the idea. But few economists of her day knew much about Margaret Reid, despite her nearly half-century tenure as a full professor at one of the nation's leading economics departments. One exception was the Italian American scholar Franco Modigliani, who lauded the “highly imaginative analysis of Margaret Reid (unpublished)” as a “fundamental contribution” to consumption economics. Reflecting on his career after receiving the Nobel Prize, Modigliani (1986) also highlighted the “path-breaking contribution” of two other women: Dorothy Brady and Rose Friedman.
Building on Modigliani's and Harberger's casual reminiscences, this article uses archival sources to reconstruct an alternate history of Milton Friedman's A Theory of the Consumption Function, spotlighting the contributions of Margaret Reid, Dorothy Brady, and Rose Friedman to the work that ultimately helped Milton Friedman win a Nobel Prize. Although Milton Friedman offered public credit to his wife and their two close friends, none received formal recognition or reward for their contributions. Despite Friedman's efforts, disciplinary norms along with the broader culture obfuscated the collaborative and gendered nature of this economic breakthrough. Understood as his accomplishment alone, Friedman's challenge to the Keynesian consumption function was immediately lauded by professional economists and contributed to his growing reputation in the field. Yet it was a joint discovery with origins in conversations, research, and debates among a group of female scholars working in uneasy relationship to academic economics. For detailed discussion of the hypothesis and its reception by economists, see Nelson 2020, 1:192–213.
Restoring this history underscores the centrality of women's contributions to economics, a recovery project still necessary decades after the dawn of women's history. Scholars have documented the Matilda effect—the systemic discounting of women's intellectual contributions—in the natural and social sciences, including economics (Rossiter 1993). The specific history of the permanent income hypothesis adds to this scholarship while also introducing an important distinction between formal and informal credit and recognition. By the middle decades of the twentieth century, female economists had made inroads into professional economics, particularly in government agencies and affiliated research groups. Their male colleagues were willing to rely on their research skills and accord them limited access to graduate degrees and teaching positions. As the case of Modigliani demonstrates, some were even willing to be fulsome in their praise. But all this credit and recognition remained informal, confined to reputation and relationship, and did not translate into publications, research grants, or professional recognition. Attending to this distinction in other cases of the Matilda effect will help scholars determine whether this informal credit should be seen as an intermediary stage along the path to further inclusion or a mechanism for forestalling meaningful reallocation of disciplinary power.
Emphasizing Friedman's engagement with the women's world of consumption economics also offers new insight into the origins and meaning of his ideas. At the very least, it provides a new answer to the question, “Why is there no Milton Friedman today?” (Klein et al. 2013). Considered more broadly, however, rooting the permanent income hypothesis in the work of women economists and the female-dominated world of consumption economics provides a new genealogy for the school of thought known as “Chicago economics.” Scholars have recognized the importance of Friedman's student Gary Becker's exploration of household economics. To some, Becker's interest in pricing the priceless bonds of household labor and familial connection is the paradigmatic example of neoliberalism, an economic and social philosophy said to define the late twentieth century or post-Fordist era (Foucault 2008: 269–70). There has been only minimal investigation, however, of the continuities between Becker's work and that of his female predecessors at Chicago. The case of Friedman's consumption function suggests another place to seek the feminine face of neoliberalism.
The story begins with a tragedy: the stillbirth of Milton and Rose's first child in 1941. The couple had recently moved to Washington, D.C., where Milton took a job at the Treasury Department. They planned to raise their child in the rambling Maryland farmhouse of Aaron Director, Rose's brother. But without a baby, this imagined bucolic playground transformed to a stark landscape of loss. Hastily changing plans, the Friedmans rented an apartment in bustling Dupont Circle. Longtime friend Dorothy Brady offered Rose the ultimate distraction—a part-time job at the Bureau of Home Economics.1
In 1940s Washington, D.C., this was quintessential woman's work, but neither Brady nor Rose Friedman had a typical profile for a government employee. At once steely and sensitive, Brady held a PhD in mathematics and raised her son as a single mother after her marriage to a UC Berkeley economics professor foundered. Rose described her as “closer to perfection than anyone I have ever known,” emphasizing her empathy and emotional intuition (Friedman and Friedman 1998). The two had met a few years earlier, when Brady worked with Milton Friedman on the Study of Consumer Purchases, a massive government survey of consumer behavior (Stapleford 2007).
Rose was midway through her PhD at the University of Chicago, where she had studied with the department's titan of price theory, Frank Knight. Rose was a rarity not only for being a female graduate student in economics but for avoiding consumption economics, that unusual domain where being female was considered an asset. Most other women who studied economics at Chicago came to work with Hazel Kyrk, the department's lone female professor and an expert in household consumption. Rose instead fell in with a cluster of students who idolized Knight, among them her brother and her future husband. Having completed all her coursework, she arrived in D.C. with vague plans for a dissertation in capital theory, Knight's current research interest. In truth, Rose was beginning to lean out when the stillbirth bounced her back into the working world she was ready to leave behind. As a result, she brought an unusual training in neoclassical price theory to the Bureau of Home Economics, whose staff members were typically steeped in the other major economic approach of the era, institutional economics.
Despite her narrow job title—she was officially a “home economics specialist” in the USDA's Bureau of Human Nutrition and Home Economics—Brady's research was anything but a backwater. Prewar Washington was still grappling with a changed economic and political landscape carved by the Great Depression and the New Deal. The administration of Franklin Roosevelt (1932) had started eclectic and experimental, unified by little other than the president's conviction that in an era of mass unemployment, there was a new “duty and responsibility of Government toward economic life.” By the late 1930s, however, leading economists and policymakers had begun to converge on the diagnosis of John Maynard Keynes's General Theory of Employment, Interest, and Money. The problem of the Great Depression, as Keynes saw it, was a fall in aggregate demand, and the solution was government spending to bolster national income.
But what was national income, and how much of it did the United States have? Although economists had long measured output and prices in specific sectors or commodities, only recently had they tried to combine numerous indexes into one aggregate figure of “national income.” In 1934 a comprehensive statistical study, Simon Kuznets's National Income, 1929–1932, became a minor sensation in D.C. Funded by the Commerce Department, it would eventually shape policy at the highest levels. With national income a defining concept for policymakers, the flip sides of income—consumption and savings—were hot topics.
Consumption was also critical because of another Keynesian concept, the multiplier. Drawing on ideas of his student, Richard Kahn, Keynes argued that a dollar spent was not simply a dollar spent but would ramify throughout the economy according to a “multiplier,” as each individual passed a portion of this new income to another through consumption. But how much of this new dollar would be spent, and how much saved? This question was critical to calculating the multiplier and hence the appropriate amount of government stimulus. Keynes (1964: 96) argued that there was a sort of “psychological law” that people would “increase their consumption as their income increases, but not by as much as the increase in their income.” This relationship between income and consumption, quickly dubbed “the consumption function” in the new language economists began to use for parsing Keynes, was assumed to be a stable and predictable relationship.
Yet as researchers threw themselves headlong into testing, refining, and elaborating the Keynesian consumption function, puzzle after puzzle emerged. In 1942 Kuznets published another groundbreaking work analyzing nearly a century's worth of American income and consumption data. He found savings and consumption ratios to be remarkably consistent, even as incomes rose. But fresh data from the Study of Consumer Purchases and other nationwide budget studies pointed in the opposite direction. It turned out that farm families had different savings patterns than urban families; so did residents of cities and villages. Both results could not be true, nor could the problem be resolved in light of Keynes's formulation. All of this had immediate policy relevance. If the consumption function was unstable or unknowable, the basic logic and efficacy of New Deal spending could also be questioned. It was the first shadow of doubt falling across the emergent paradigm.
Rose's new project plunged her into the thick of the so-called Kuznets paradox (Bunting 1989).2 Broadly speaking, Dorothy and Rose were trying to make sense of the mounds of data generated by the Study of Consumer Purchases. The specific challenge they took up was teasing out the relationship between spending and saving in different communities. Their data could be broken down into at least twenty-four different groups, including those identified as farm families, white families, Negro families; west-central middle-size cities, New England villages, southeast villages, Atlanta, Portland, Omaha, Columbus, Providence, Chicago, and New York. Were there any commonalities across these diverse groups? What discernable patterns linked income, consumption, and savings? Their research unfolded within the auspices of the National Bureau of Economic Research, a hybrid academic-private research institution with close links to Columbia University. A few years earlier, Milton Friedman had helped establish NBER's Conference on Research in Income and Wealth, an ongoing project that coordinated the activities of worker bees like Brady with luminaries like Arthur Burns, Wesley Mitchell, and Kuznets.
Rose and Dorothy's research culminated in “Savings and the Income Distribution,” a nineteen-page paper that marked a significant shift away from the social and cultural explanations characteristic of earlier consumption research. The authors began with a nod to the traditional way income differences had been interpreted, “in terms of the income of the individual family, its size, occupational group, race and national origin, region and size of community,” noting that corresponding variations in consumption patterns had been treated “as not yet susceptible of expression in a numerical relation of any form.” By contrast, Brady and Friedman (1947: 247) noted these relationships could also “be expressed in quantitative terms.” Keeping with Brady's statistical bent and the training Rose had received at Chicago, the paper would analyze numerical income distributions rather than digging into the qualitative differences between rural and urban families. Still, their mathematics remained simple, with their major conclusions explained textually and displayed in basic scatterplot diagrams.
Across all twenty-six community categories, Rose and Dorothy reported they had found one key pattern: “Variations in the pattern of consumption and savings among groups of families at given income levels may be explained to a considerable degree by differences in the level and distribution of income” (247). What mattered to savings and spending, the authors explained, was not the absolute income a family received but its relative income, for example, whether the family was in the twentieth or fiftieth percentile vis-à-vis its neighbors. And in turn, this varied from community to community, again with some discernable patterns. In communities with fewer high-income families, savings rates for all families were higher. The denser the population, the lower the savings, across all groups.
Dorothy and Rose's conclusions did retain some intriguing qualitative conclusions. Coming clearly through the data was “the apparent tendency for families, whatever their income, to relate their consumption to the community income situation,” suggesting a residue of cultural analysis remained. Was it norms, traditions, or some other force that accounted for these regularities? In the end, this was not a speculation the authors indulged, other than a brief closing note arguing that black family spending was “doubtless influenced by consumption patterns of the white community as well as by their own social world” (250, 265).3 Yet nothing more could be done with the present information; the paper closed with an unobjectionable call for more data and analysis.
However modest the paper's overt claims, Brady and Rose Friedman were in fact making a bold play for relevance in the emerging debate over Keynesianism. As the Yale economist James Tobin understood, their paper threatened the entire edifice of policy and politics coming to rest on Keynesian concepts of savings and consumption. Tobin recognized that Brady and Rose Friedman's paper, along with the work of the Harvard economist James Duesenberry, established a new “‘relative income’ hypothesis” that challenged the Keynesian “‘absolute income’ hypothesis.” This new hypothesis was appealing, Tobin (1951: 135–36, 152) wrote, in light of Kuznets's findings: “It explains the relative constancy over the last 70 years of the percentage of national income saved. This invariance cannot be explained on the simple Keynesian theory that the saving ratio is a unique function of absolute real income” (Duesenberry 1940).4 In turn, the relative income hypothesis suggested that a quick and easy prediction of national income from family income was impossible. A leading defender of Keynes, Tobin presented his own data in defense of the absolute income hypothesis, at the same time arguing that a fully developed theory must account for wealth and historic growth in asset holdings.
On the surface, Dorothy and Rose had published a basic research report. Considered in the bigger picture, their conclusions spoke to the politically charged question of consumption. Was the paper deliberately framed as an attack on Keynes?5 Both women were dedicated empiricists, and the problem in the data was compelling. At the same time, the solution they came up with dovetailed nicely with each woman's intellectual inclinations. The paper's emphasis on relative income reflected the traditional approach of consumption research that Brady knew well, which often pointed to cultural factors as determining parts of family economic life. Brady's long tenure in reformist D.C. agencies suggests that like most consumption economists, she was probably sympathetic to New Deal social spending. By contrast, although Rose has left little trace of her thinking in this period, she was among the most loyal of Frank Knight's students, and his teachings would have primed her to be skeptical of both the New Deal and Keynesian concepts newly popular among economists.
While it is difficult to parse Rose's distinct contribution, and harder still to detect if Milton played any role in the project, the combination of Rose and Dorothy was uniquely generative. At this stage in her career, Brady was an able technician but not a groundbreaking scholar. Her other publications during this time were applications of the ideas laid out in her paper with Rose or discussions of the inadequacy of various data sets (Brady 1938, 1951, 1952). By contrast, Rose's Chicago training emphasized finding a general principle or relationship within the specific data and moving quickly to the wider implications for policy or economic theory. Working together, they created an empirically rich and theoretically ambitious paper. As one historian of the era notes, their paper was a “first attempt to provide a unifying hypothesis” to explain a widely noted anomaly (Hynes 1998: 29). In this respect, the paper captures both Rose's intellectual potential and the loss to economics that came with her move away from active research.
By the time Dorothy first presented their findings at the Conference on Research in Income and Wealth—nearly three years after their collaboration—Rose had long since moved on.6 However fruitful, working with Dorothy had been at best only a distraction from her dominant goal of starting a family. After about six months, she stopped working in order to focus on her health. The formal collaboration was over, but the conversations and intellectual bonds remained.
Nearly a decade later, Rose had the two children she desired and Milton had secured a coveted tenured professorship at Chicago. Dorothy was still burrowing through consumption data. In 1948, Milton intercepted a paper on food expenditures she had sent to Rose. Building on their earlier publication, the paper attempted to correlate the consumption function to the income distribution. Friedman found the paper “enormously intriguing” but doubted Brady's results could support the hypothesis. He referenced a few points they had discussed in a recent visit and went on to make a number of suggestions about ways to analyze her data.7
Brady responded with a letter accepting some of his criticisms and pushing back against others. Pointedly, she referenced his work on smoothing at the Study of Consumer Purchases, noting many of the same issues he had faced remained unresolved. She agreed, however, that her paper was incomplete, offering a sentiment that almost exactly anticipated Friedman's famous 1953 essay on methodology: “Fundamentally, the only real test of a theory is in reasonably accurate prediction and this experiment did not lead to a ‘formula’ that could be used for prediction.”8 Friedman's next letter to Brady, coming after another conversation, pointed her toward a chapter in his dissertation coauthored with Simon Kuznets, published as Income from Independent Professional Practice. “It's in this chapter that we present the material and breaking down an income of an individual into various components,” he explained.9 Friedman closed with hopes that Brady was settling in to her new position at the University of Illinois, Urbana.
Brady carried Friedman's reference to her new colleague, Margaret Reid. A former student of Hazel Kyrk, Reid's lifework was measuring the economic contribution of household work, previously considered nonproductive activity. After receiving her PhD in economics from Chicago, Reid accepted a professorship at Iowa State College. During the war years, Reid took two leaves of absence from Iowa and accepted federal appointments in D.C., during which time she helped establish nutritional and food-expenditure standards to inform appropriate levels of government support for poor families. Reid was drawn to Washington by a combination of factors, including a reformist desire to shape policy, political controversy at her home department in Iowa, and a rich network of economists interested in similar questions.10 Among these economists was Brady, whom Reid described in glowing terms as “unusually intelligent, clear thinker, high standard of workmanship, marked social conscience, able, systematic, thorough, sensitive to people, sense of humor.”11 Perhaps the most powerful lure of all, however, was the trove of data on consumption that could be found in D.C. and nowhere else.
Here her goals and Brady's goals were particularly symbiotic. Brady had the data, but Reid had the time. Eventually, Reid was able to secure Brady a temporary position at her next employer, the University of Illinois. As Reid wrote in a later grant proposal,
I think you are aware that Dr. Dorothy Brady and I hoped, by coming to the University of Illinois, to have an opportunity to do research in consumption of a type not possible in the day-to-day rush in Washington. We have both felt for a long time that there are data in many Washington, and other files that should be but are not being analyzed. Analysis tends to be neglected because the effort of staffs to such a large extent must go to getting current data which have immediate value, and because of lack of staff to meet both current needs and do long run research.12
Given her base in the Midwest, Reid was keenly interested in patterns of income, savings, and consumption among farm families. Brady was not just a valued collaborator but brought access to data that could help her situate farmers in a broader national context.
One of Reid's specific goals was to analyze consumption across multiple years.13 Here the Friedman and Kuznets reference was apposite. The chapter Friedman cited to Brady discussed the concepts of permanent and transitory income. To better compare the incomes of dentists and doctors, Friedman and Kuznets had drawn a distinction between onetime “windfall” or transitory income, and more regular permanent income. Reid was focused on an entirely different group—farmers—and she was interested in consumption. But consumption was, after all, the obverse of income. And farmers were beset by an extreme version of the dynamic that Friedman and Kuznets had detected. Vagaries of weather, prices, and the need for new equipment could make farm consumption swing wildly from year to year. Permanent income suggested a way to smooth out these cycles and gain a more robust estimate of income over the long term. Thus Brady brought more than data to Reid: she introduced her to an important theoretical framework.
Brady also drew Reid firmly into the Friedmans' social orbit. Reid became a regular visitor to Rose and Milton's New Hampshire summer home, “the Hideaway.” Located near the Connecticut River in Orford, the Friedmans' getaway was a modest cabin surrounded by a thick grove of pine trees. Arthur Burns had first introduced them to the area, which was popular with New York economists due to the proximity of the Dartmouth College library. After buying the Hideaway in 1949, the family spent every summer there. When Friedman was not battling mice, porcupines, or raccoons, Orford became the site of his deepest thinking. Near the house—but not too near—was an artists' studio, with a large glass window looking out to the woods and perfect light (Friedman and Friedman 1998: 164).
Splendid isolation worked for Friedman, but what got Rose thinking was her friends. When Dorothy and Margaret visited, the evenings became an impromptu seminar on consumption research. They were sometimes joined by Ruth Mack, an economist neighbor who had worked with Friedman at the Treasury Department (Shoup, Friedman, and Mack 1943).14 As the children drifted off to sleep in a balcony above the firelit living room, they heard below the steady hum of adult voices.
At the crux of their conversation was the Keynesian consumption function. According to Keynes, there was a fairly straightforward relationship between income, savings, and consumption. As income rose, people would increase their spending. Their savings would also increase. This relationship, which economists termed “the consumption function,” was critical to the multiplier, which in turn determined the amount of additional investment needed to stimulate a moribund economy or maintain growth. Yet the empirical puzzle continued, as detailed surveys of income and spending revealed a variety of contradictory patterns. Almost a decade earlier, Rose and Dorothy had been among the first economists to suggest a rival theory, now called the relative income hypothesis. Other economists, notably James Tobin, had suggested an alternate explanation that seemed to redeem the original Keynesian idea: the wealth-income hypothesis. Yet as Dorothy and Margaret explained what they were finding, neither explanation really seemed to fit.
Slowly, over the summers of 1949 and 1950, yet another hypothesis began to evolve through the group's conversations. The scholars added Friedman and Kuznets's concept of permanent income to the consumption data Brady and Reid were analyzing and began to think through the implications. What if spending was not the result of income, per se, but of “permanent” income? What if consumers—like Reid's farmers—based their spending and savings on a forecasted average of income over several years, or even their lifetimes? If transitory or windfall income was factored out in favor of permanent income, would the data make more sense?
Taking such a perspective did not mean setting aside entirely Rose and Dorothy's first insight on relative income. Comparative spending could still be a factor, even if it was less important than the crucial elements of time and anticipation. It was even possible to integrate Tobin's argument that wealth played an important role in savings. Some consumption—for example, of durable goods—could be understood as a form of savings. But as their conversations progressed, the scholars became convinced that their permanent income hypothesis, as Milton Friedman (1957: 6) claimed in the finished work, was an idea that was “potentially more fruitful and is in some measure more general” than the alternatives.
An important spur to discovery was the sudden possibility Reid and Brady might secure faculty positions at the University of Chicago. Hazel Kyrk was on the cusp of retirement, and while there was no rule her replacement need be a woman, both department tradition and the nature of the appointment—joint with Home Economics—opened a rare opportunity. Simultaneously, Friedman was coming into his own as a force within the department, fighting a guerrilla war against a rival group of economists at Chicago, the Cowles Commission. In Friedman's view, the Cowles Commission econometricians were not really economists but glorified mathematicians. Rather than building predictive mathematical models, economists should develop and test theories by closely examining data from the real world—just the sort of work Brady and Reid were doing. Friedman saw the chance to supplant his intellectual enemies with two women he admired as scholars and knew as personal friends.
In March 1951, Reid and Brady sent department head T. W. Schultz a long memo outlining a program for consumption research. Landing amid Friedman's maneuvering against the Cowles Commission, and no doubt written at his urging, the memo suggested an alternate focus for the Chicago department. Along the way, it boldly defended and celebrated the unsung work of consumption researchers. Consumption had moved to “the forefront of economic discussion” due to the Keynesian revolution, the memo noted. Yet economists had ignored a century-old tradition of empirical work. “The early efforts to discover the consumption function through the use of family data were carried on almost as if nothing had previously been done in this field,” Brady and Reid declared tartly, leaving unspoken the truism that women had done most of the prior research. Because it ignored the field's pioneering scholarship, the new work “has been spotty and much of it shoddy,” the two women declared. They went on to propose a program in consumption research, centered at a university.15
The department took up the hiring question in May. With Friedman operating behind the scenes, the department's chair, Shultz, led discussion. The retirement of Kyrk, Schultz noted, opened up the possibility of “undertaking to put resources into the field of consumption research.” The department should even consider “a full scale workshop directed by a team consisting of Miss Reid and Mrs. Brady.”16 For now, though, they had funding for one appointment, joint with the Home Economics Department. Reid soon emerged as the lead candidate. Unlike Brady, who held a PhD in mathematics, Reid had already established an academic career in economics. She was also known to Schultz from their time as colleagues at Iowa State College, and perhaps most important, she had the invaluable credential of a Chicago PhD.
In a second meeting, Schultz pressed for a detailed resolution stating that the department would hire Reid “as the first move in the establishment of the research enterprise, with the hope that Miss Reid soon could be joined by Mrs. Dorothy Brady.” The motion to hire Reid passed unanimously, but the department supported only a tepid “continuing interest in consumption economics as a field of research,” declining to record a more detailed resolution.17 Any future moves would depend on outside funding and the department's willingness to offer Reid a workshop akin to those run by her colleagues.
As these deliberations unfolded, Friedman decided to write up the ideas he called “our tentative hypothesis.”18 He sent a copy to Brady, Schultz, and Reid. He also rewrote the proposal Brady and Reid had already submitted and forwarded a copy to Schultz. This document was generous with credit. “Dorothy Brady and Margaret Reid have concluded from their recent work that no one of these is acceptable. Together with Milton Friedman, they have developed an alternate hypothesis,” he wrote.19 In another letter to Brady, Friedman referenced “the consumption hypothesis we have so far arrived at.”20 Brady wrote back encouraging Friedman to develop it into something publishable. “I believe that you included all of the logic as I remember your oral summary late one night a month ago,” she noted.21 With Reid now heading to Chicago and leaving Illinois, Brady returned to her government job. Friedman continued to lobby for Brady's hiring at Chicago, even suggesting her as a potential director of the Cowles Commission.22 He would eventually succeed in securing her a temporary position.
When the permanent income hypothesis appeared six years later in A Theory of the Consumption Function, Friedman gave credit to both Reid and Brady. It was Reid, he claimed, who had first tested the theory “with characteristic enthusiasm, persistence, and ingenuity.” She had then pressed him “to write up the underlying theory so that she could refer to it in a paper presenting her conclusions” (Friedman 1957: ix). Reid was working on a paper, “The Relation of the Within Group Transitory Component of Incomes to the Income Elasticity of Family Expenditures,” that did reference Friedman's theory.23 This paper was never published, but portions of it were presented at the 1953 Annual Meeting of the American Economic Association (Reid 1953). But from his correspondence at the time, it appears Friedman spontaneously wrote up the hypothesis as part of broader efforts to secure Reid and Brady faculty positions at Chicago. Moreover, it appears that Brady was the first to press Friedman to write something for publication. Separately, Reid may have done the same, and she may have urged him in person rather than through a letter.
But why would Reid have wanted Friedman to take this next step, rather than write it up herself? She may well have believed herself to lack the technical training and knowledge to do the hypothesis justice. This had not, however, been an obstacle in their conversations.
Perhaps she understood that her work would go farther, faster if it rested on a theory developed by a well-known male economist.24 Indeed, at the very same moment Friedman's other major collaborator, Anna Schwartz, was struggling to secure a doctorate despite her prodigious research output.25 Reid's memo to Schultz indicates she knew well how economic research by women, no matter how relevant, could be systematically overlooked.
Regardless, it was Friedman who ultimately took ownership, turning it from “our tentative hypothesis” into “my hypothesis.”26 After Reid's hiring, the ideas languished until the summer of 1953, when Friedman went on a tear. “I have been meaning to write to you all summer, for you have been much in my thoughts,” he told Reid. “I finally got around to writing up the theory of consumption, and naturally drew much on your paper.” The result was “a beast” and “a mammoth” of a manuscript, running to 140 handwritten pages. It had turned out better than he dared even hope, Friedman wrote, for much of the evidence conformed to the hypothesis. “I am beginning to believe we really have something,” he reported excitedly.27 His second letter, sent ten days later, slipped between “my paper” and “our approach,” as he referred to the completed paper, which Brady was having typed up and would send to her.28
In the fall, Friedman headed to Oxford for a year abroad, leaving the hypothesis to be tended by Reid and Brady. In his absence, a thunderstorm descended on Brady. She was suspended from her government job in early 1954 due to a loyalty investigation. Brady had been caught up in the McCarthyite dragnet most likely due to her ex-husband, a known political radical. Although they were long divorced, decades-old associations with suspected Communists could still end careers.29 Or possibly she was a victim of the lavender scare; Dorothy had a female partner, and this relationship could have been deemed a security risk. Unable to work until her case was resolved, Dorothy was stressed and haggard. “She looked dreadful,” Anna Schwartz reported to Friedman, passing on Dorothy's thanks for a letter he had written on her behalf.30 A month later, Brady's case was closed and she was able to resume work. But she remained “whipped” by the experience, as Schwartz saw it.31 She returned reluctantly to her former job at the Bureau of Labor Studies.
At this stage, with Friedman overseas, Reid's link to the emerging hypothesis was common knowledge in Chicago. Director of Graduate Studies Gregg Lewis sought out Reid to discuss a possible appointment of James Tobin as director of the Cowles Commission. After Tobin and Reid had a lengthy discussion, she reported back to Friedman that he was not inclined to support a major grant for research in consumption. “However Dorothy feels and I do that he can be educated,” Reid wrote. The dream of a consumption workshop was still alive. She also noted that Tobin's visit, along with her AEA presentation, had raised the interest of a visiting scholar, Sigbert Prais. A Cowles affiliate, Prais wanted to “translate my hypothesis into algebra.”32 Prais wrote an eight-page paper analyzing Reid's longer paper alongside Friedman and Kuznets's book.33 Reid also shared her extended take with Franco Modigliani, her AEA copanelist. Twenty years later he cited the never-published paper, along with the article by Rose and Dorothy, in his Nobel Prize address.34
If field specialists and the Chicago department knew of Reid's involvement in Friedman's latest project, he declined to share this information when corresponding with Tobin. This was a curious omission, given that Tobin had publicly commented on the very first iteration of the hypothesis, Dorothy and Rose's now obscure 1947 paper. Arriving back in the United States the following summer, Friedman felt confident enough to alert Tobin, whom he identified with the absolute income hypothesis, to his criticism and seek a response.35 By August 1954, Friedman was preparing the manuscript for submission to the National Bureau of Economic Research. It had now definitively become, in the words of NBER director Solomon Fabricant, “your manuscript.”36 And when Friedman debuted the completed manuscript to the field at a Princeton conference, Brady, Reid, and Rose had faded away entirely. In a sad irony, NBER hired Rose to check the footnotes on the finished work. The experts at the conference may have known about Dorothy and Rose's original paper. But their remarks addressed Friedman and offered him credit and blame for a hypothesis they found at once “splendid and stimulating” yet not entirely convincing.37
Ultimately, the book would do much to revive Friedman's flagging reputation in the field. Unlike his work in monetary theory, which was not then a high-prestige research area, A Theory of the Consumption Function addressed critical issues in the mainstream of the field. And while there was a political tint to the work, economists need not share Friedman's policy preferences to recognize the contribution. The Yale economist James Tobin, a frequent Friedman sparring partner, laid down a favorable verdict: “Research on the consumer behavior will be different in the future than it has been in the past as a result of this work. The kinds of phenomena that will have to be investigated in any future study have been changed as a result of this work.”38 It was the ultimate academic compliment. Even if they did not know it, Friedman had forced the field to contend with the woman's world of consumption economics.
Did Friedman unfairly claim singular credit for A Theory of the Consumption Function? In the book's introduction, he went so far as to call it “a joint product of the group,” humbly claiming “my hand held the pen” (Friedman 1957: ix). But there is no evidence he seriously attempted to publish the book as a coauthored work, even though much of his early work was jointly authored.39 In the informal world of intellectual give and take—where the actual thinking and discovery happened—Friedman was generous with his praise and credit. In the formal world of publication and peer review, Friedman defended his own interests.
This was less an indictment of Friedman and more testimony to the power of norms. In fact, Friedman's willingness to champion the work of Brady and Reid set him apart from the field. Not only did Friedman want to hire his friends, he wanted them to have real institutional power, a position his peers did not share. In his work with Anna Schwartz, Friedman not only published as coauthor but successfully pushed Columbia faculty to grant her the PhD. In both cases, however, women economists found their options constrained by formal structures of power and recognition. Reid could become a valuable member of the Chicago department, recognized within that sphere for her tangible contributions. But never would she get her own workshop, the place where the real work of the department occurred. Schwartz could be given operational control over the NBER study on monetary cycles that culminated in A Monetary History of the United States, but the credential of a PhD came only grudgingly.
What if A Theory of the Consumption Function really was a solo work, with Friedman doing all the heavy analytic lifting and the women acting as passive sounding boards, cheerleaders, or simply suppliers of unprocessed research data? In other words, was Friedman justified in excluding Reid, Brady, and his wife from formal authorship or recognition? Given the difficulty of pinning down the process of intellectual discovery, this question is tricky to answer. Since Friedman took it upon himself to write up the hypothesis in extended form, and since the general concept had first appeared in his dissertation, no doubt he appeared in his own mind to deserve full credit. (Perhaps Kuznets got stiffed too.) Before accepting this interpretation of his female collaborators' reduced role, however, we should reflect on how commonly these framings have been used to sideline women whose substantive contributions can be thoroughly documented. We should also listen to Friedman himself, who in private correspondence conceptualized the project as joint until the final stages when it assumed formal shape.
Another approach is to ask, Would Friedman have hit on the permanent income thesis by himself? Certainly one tributary came out of his dissertation research with Kuznets. Yet the more fundamental context was the group of woman economists he had known for decades who brought him a steady stream of ideas and provocations—not to mention painstakingly gathered data. Rose's role was critical; as their son reflected, “she was part of a cluster of people and ideas out of which the consumption function came.”40 Without her, there would have been no long summer evenings, no letters from Dorothy Brady, no plots and plans to create a consumption workshop at Chicago, and no reason for Milton to divert his thinking away from money. One element of Friedman's profile that particularly impressed the Nobel committee decades later was his versatility. He was not simply a monetary theorist, but a consumption researcher and a historian. Without the women in his life, he would simply have been the first.
Beyond Rose, another influence was the Chicago tradition in household economics. Since the 1920s, the department had distinguished itself by keeping at least one woman on the faculty. The title of Friedman's book was a hat tip to Hazel Kyrk, whose first book was A Theory of Consumption. Friedman may have been the first to realize the possibilities of this fact, but he was not the last. A decade later, Gary Becker would make waves with audacious claims to measure in dollars and cents the value of children and the price of housework—the very sorts of calculations household economists had been doing for years.41 Friedman had started a new twist on an old tradition, laundering women's economics for consumption by a male-dominated profession.
What is at stake in understanding the true history of A Theory of the Consumption Function? First, it clarifies that women were important contributors to the fundamental disciplinary debates surrounding Keynesianism in the twentieth century. Friedman, for one, saw A Theory of the Consumption Function as a direct blow against the Keynesian consensus, claiming that “acceptance of the permanent income hypothesis removes completely one of the pillars of the ‘secular stagnation’ thesis.” A key part of New Deal thinking, the idea of secular stagnation held that higher incomes would lead to higher savings, less investment, and continued economic sluggishness. One solution was taxing and redistributing high incomes. Yet the permanent income theory suggested higher income did not necessarily lead to higher savings; the relevant concept was permanent income. Friedman spelled out the implications: “It destroys the case for one proposed remedy,” he argued. Reducing inequality would not automatically increase the propensity to consume and stimulate the economy. This was not merely Friedmanism grafted onto the text; surviving letters show he and Dorothy regularly discussed income inequality, how to measure it, and its social and political implications—even referencing the ideas of Alvin Hansen, a leading American Keynesian.42
Friedman (1957: 237) went on to argue that the theory downgraded the importance of the multiplier, thereby revealing “an inherently cyclically more stable system.” This went to the crux of the matter: did capitalism need government intervention to function? A Theory of the Consumption Function reminds us that monetarism was not Friedman's only challenge to Keynesianism. Indeed, while the Keynesian-monetarism debates petered out by the end of the 1980s, permanent income remains a fundamental economic and sociological concept; only its roots in the female networks of consumption economics remain obscure.
Second, A Theory of the Consumption Function touches on the boundary-crossing drive of neoclassical economics—so-called economics imperialism. Attending to consumption and household economics shows that this male-driven colonization project drew in fundamental ways on a protofeminist impulse to understand, value, and price the unpaid household labor of women.
Finally, while most critics of twentieth-century economics do not share Friedman's politics, they do share his judgments that economics became too theoretical, too arid, too wedded to a formalist interpretation of human behavior. In short, economics became a science of the artificial.43 The recent emergence of behavioral economics as a corrective shows the wide resonance of this critique. But what drove this turn to formalism, and how can its mistakes be avoided in the future? One overlooked factor is how an earlier generation of academic economists stigmatized and sidelined the corner of their field engaged in extended study of how households and individuals actually made economic decisions in real life, not in simulations. It was women who gathered and analyzed critical information on economic behavior, from income to consumption to savings. In an age of economics driven by big data, it is worth remembering that the original computers were women (Grier 2005).
And what of our opening question—was informal credit a stepping-stone or a dead end? Vexingly, it was both at once. It would take broad shifts in social norms to open traditional markers of professional success to women in economics. And there was no making up for lost time. Yet when these shifts finally began under the pressure of feminism and the rights revolutions, informal credit could sometimes transform into something more.44
After her children departed for college, Rose Friedman turned once again to her husband's work. “With the assistance of Rose D. Friedman” proclaimed the cover of Milton Friedman's 1962 popular work, Capitalism and Freedom. Once again, there were hints her role had been essential. By Milton's account, she “pieced together the scraps of the various lectures, coalesced different versions, translated lectures into something more closely approaching with written English” (Friedman 1962: xvi). In fact, just as A Theory of the Consumption Function would never have appeared without Rose, so did Capitalism and Freedom owe its existence to her presence. Rose consistently pushed Milton toward political engagement, sharpening his political identity and encouraging him to take up public-facing opportunities like his Newsweek column. By 1982 she appeared as a coauthor of Free to Choose, a best-selling book based on Milton's PBS documentary, of which she was a coproducer (Friedman and Friedman 1980). In the documentary itself, she remained always offstage.
Rose's ideas, be they economic or political, remained subsumed by her husband. “My husband never wrote anything without my reading it over and talking about it,” she bragged to an interviewer in Singapore. “Now you can't tell who wrote what, the style is the same throughout the books. I always tell people we work as one; we are one.” Rose was not bothered that all the attention went to Milton, she maintained. “I'm not a competitive person, or I would have ordered my life differently.”45 Still, at times there was a burn. Whenever she was introduced as Dr. Friedman, Rose confessed to another interviewer, “I feel upset inside when that happens. It's hard to come out and say at the time, ‘I'm sorry, but I'm not a doctor,’ but I don't want to fly under false colors.” In 1986, that problem was solved when Pepperdine University awarded her an honorary doctorate. A picture of glowing Rose, surrounded by family at the commencement ceremony, was the last photograph included in the couples' joint memoir (Ullrich 1976; Friedman and Friedman 1998).
In the end, Dorothy Brady found her place in academe, but not as a consumption expert. Nonetheless, it is entirely possible her work with Friedman made a difference. After a brief stint as a visiting professor at Chicago, in 1958 she was hired as a professor of economics at the Wharton School, University of Pennsylvania. Had Friedman pushed for her hiring there, as he had at Chicago? Brady's papers do not survive, making it hard to know for sure. But doubtless a year at Chicago burnished her résumé as an economist. At Wharton, Brady's interests began to shift toward economic history, where she drew on her statistical studies to depict economic life and relative prices in the nineteenth century. By the time of her death in 1977, she was a beloved figure to former colleagues and students (Easterlin 1978).
During most of her decades at Chicago, Margaret Reid was actually an emeritus professor, assuming that rank as required by mandatory retirement in 1961. Still, she remained a fixture on campus, regularly attending economics workshops and seminars. Undergraduates knew her as the lady in the library “taking up two whole tables with her stuff.” Graduate students were slightly more informed, telling one another, “There goes one of the ancients. She was important and, amazingly, she is still doing research!” Decades later, the Harvard professor Claudia Goldin reflected that Reid was “the only female economist I knew as a graduate student.” Reid maintained an active research program into her eighties, passing away in 1990 (Hershfield 1985; Goldin 2021: 49, 266n46).
Had she lived five years longer, Reid would have witnessed an upwelling of interest in her work, symbolized by the special issue of Feminist Economics devoted to her life and career. In 2021, Goldin hearkened back to Reid as a pioneer. “I wish I had possessed enough foresight to have struck up a conversation with Margaret when I was a graduate student,” she lamented. “How naive I was for not recognizing her importance to the field of economics!” Goldin summarized major themes in Reid's work beyond the permanent income hypothesis, including her work designing the consumer price index. She highlighted as particularly relevant Reid's unsuccessful push to include women's unpaid labor in measures of national income (Goldin 2021: 47).
Before Reid passed away at the age of ninety-four, she was able to enjoy a formal marker of disciplinary prestige beyond her Chicago job. In 1979 she was elected a Distinguished Fellow of the American Economic Association, the field's major professional organization. It was small beans compared with the Nobels and Clark Prizes of her colleagues, but nonetheless a triumph to be savored. As the first woman so honored, Reid was determined to get credit at last. At her request, the traditional wording of the citation was changed, with an updated pronoun that recognized “her contribution” to the field of economics (Hershfield 1985).
I am grateful to John Singleton and Cléo Chassonnery-Zaïgouche for the invitation and encouragement to write on this topic. I received helpful comments on earlier versions of this article from attendees at the 2021 History of Political Economy Conference, a 2018 Economic History Seminar at the Stanford University Department of Economics, a 2018 Gender History Workshop at the Stanford History Department, and the 2015 Women in Intellectual History Workshop at the Stanford Humanities Center. This research was supported by Stanford University, the National Endowment for the Humanities, and the Hoover Institution Library and Archives.
By contrast, J. J. Thomas (1989) argues that the impact of both Kuznets’s statistics and Keynes’s consumption function have been overstated, emphasizing that earlier budget studies had already made economists aware of the complexities of consumption relationships.
Brady may also have been influenced by the Berkeley economics professor Jessica Peixotto, a colleague of her former husband, who developed what one historian describes retrospectively as a “relative-income theory of consumption” in her 1927 study of college professor budgets (Peixotto 1927; Cookingham 1987: 47–65).
Despite Tobin’s critique, Brady and Friedman’s paper is often omitted even from detailed discussions of the topic. For example, Thomas 1989 closely examines early progenitors of the relative income hypothesis but cites only Duesenberry. Duesenberry did not cite Brady and Friedman (he did cite two of Brady’s statistical papers).
Duesenberry (1940: 1, 116), for one, admitted his book “began as a critique of the Keynesian consumption function,” while hastening to underscore, “It is not, of course, intended to argue that nothing needs to be done about depressions.”
Brady first presented the paper in November 1945 at the Conference on Research in Income and Wealth.
Milton Friedman to Dorothy Brady, May 4, 1948, Milton Friedman Papers, Hoover Institution Library and Archives, Stanford University, box 21, folder 18.
Brady to Friedman, May 13, 1948, Friedman Papers, box 21, folder 18. This comment, along with discussion of the desired features of a scientific theory in the introduction to A Theory of the Consumption Function, suggests the 1953 essay was cross-fertilized by the group’s work.
Friedman to Brady, November 8, 1948, Friedman Papers, box 21, folder 18; Friedman and Kuznets 1945.
Margaret Reid, typed diary entry, July 19, , Margaret Reid Papers, University of Chicago Regenstein Library, box 53, folder 1.
Margaret Reid to Kathryn V. Burns, Acting Head, Home Economics Department, January 19, 1949, Reid Papers, box 1, folder 15.
Reid to Burns.
A Columbia PhD and longtime staffer at NBER, Mack appears to have joined the group occasionally and offered comments on a late-stage draft.
Margaret G. Reid and Dorothy S. Brady to T. W. Schultz, “Memorandum: Proposed Research Project on the Consumption and Savings of Families,” March 12, 1951, p. 4, Friedman Papers, folder 18, box 21. Friedman later rewrote this memo, so we know this version was written without his substantive input.
“Minutes, Meeting of the Department, May 23, 1951,” University of Chicago Department of Economics Records, Regenstein Library, University of Chicago, box 41, folder 2.
“Minutes, Meeting of the Department, May 28, 1951,” Department of Economics Records, box 41, folder 2.
Milton Friedman to Brady, June 14, 1951, Friedman Papers, box 21, folder 18.
Milton Friedman, “A Program for Consumption Research,” June 11, 1951, Friedman Papers, box 232, folder 12.
Milton Friedman to Brady, July 15, 1951, Friedman Papers, box 21, folder 18.
Brady to Milton Friedman, July 17, 1951, Friedman Papers, box 21, folder 18.
Milton Friedman to T. W. Schultz, March 29, 1954, Department of Economics Records, box 42, folder 10. Friedman’s energetic advocacy can be contrasted to Paul Samuelson’s tepid support of his student Margaret Garritsen de Vries (see Laskaridis, this volume).
Margaret Reid, “The Relation of the Within Group Transitory Component of Incomes to the Income Elasticity of Family Expenditures,” unpublished paper, Reid Papers, box 10, folders 8–9.
The sociologist Robert Merton even suggested this as a strategy that lesser-known scientists might deliberately employ to spread their ideas (Rossiter 1993: 326).
This episode is covered in Jennifer Burns’s forthcoming biography of Milton Friedman.
Cf. Milton Friedman to Brady, June 14, 1951, Friedman Papers, box 21, folder 18, and Milton Friedman to Ruth Mack, August 31, 1954, Friedman Papers, box 109, folder 8.
Milton Friedman to Margaret Reid, August 7, 1953, Reid Papers, box 6, folder 1.
Milton Friedman to Margaret Reid, August 17, 1953, Reid Papers, box 6, folder 1.
Brady’s name appeared as belonging to a “Red Front” in a report of the Dies Committee, likely the original source of the investigation. See “Government Employees Listed by Dies as in Red Front,” New York Herald Tribune, October 26, 1939, 8.
Anna Schwartz to Milton Friedman, January 19, 1954, Friedman Papers, box 90, folder 7. David Friedman, interview with author, Stanford, Calif., October 16, 2014.
Milton Friedman to Schwartz, February 1, 1954, Friedman Papers, box 90, folder 7.
S. J. Prais, “An Exegetical Note on Some Recent Work on Income Variations,” eight-page typescript, Friedman Papers, box 232, folder 13.
Margaret Reid to Milton Friedman, February 7, 1954, Friedman Papers, box 232, folder 13.
Margaret Reid to Milton Friedman, February 7, 1954, box 232, folder 13; Modigliani 1986.
Milton Friedman to James Tobin, July 24, 1955, Friedman Papers, box 109, folder 10.
Solomon Fabricant to Milton Friedman, October 26, 1954, Friedman Papers, box 232, folder 12. The book was published by Princeton University Press as part of an NBER series.
Friedman was not able to attend the conference in person, but two attendees, Eugene Lerner and Warren Nutter, sent a transcript of remarks. Lerner to Milton Friedman, October 28, 1955, Friedman Papers, box 232, folder 13.
James Tobin, transcript of “Conference on Consumption and Economic Development,” Princeton University, October 21–22, 1955, 6, Friedman Papers, box 232, folder 12.
Friedman (1945) wrote most of Income from Independent Professional Practice, which listed Simon Kuznets as coauthor. The book served as Friedman’s Columbia University doctoral dissertation.
David Friedman, interview with author, Stanford, Calif., October 16, 2014.
Kyrk’s career is covered in Kiss and Beller 1999. Reid’s relationship to Friedman and Becker is discussed in Yi 1996. For Becker’s role in the development of New Household Economics, including an analysis of how the field became less hospitable to women, see Grossbard-Shechtman 2001. A revised and shortened version of the article presents Becker in a more positive light. See Grossbard 2006.
See, e.g., Brady’s undated Friday letter to Friedman, which references past conversations on the topic and hopes for continued discussion (two-page handwritten letter, Friedman Papers, box 21, folder 18).
For a historically rigorous version of this thesis, see Mirowski 2002.
Outside the scope of this article is the example of Schwartz. Friedman’s intervention secured her a doctorate, and she was recognized as coauthor of A Monetary History of the United States. Nonetheless, Friedman’s Nobel citation identified A Monetary History as his work alone. By the 1980s, Schwartz had become more widely known to economists for her work on the Shadow Open Market Committee, and to a wider public for her service on the US Gold Commission in 1981.
“And Friedman Chose a Rose . . . ,” Straits Times, Saturday, October 18, 1980.