The “negative income tax” as a steering mechanism: the semantic field of the NIT around Milton Friedman in his pre-monetarist period (1939–1948)

Abstract

This paper explores the intellectual history of Universal Basic Income (UBI), specifically focusing on Milton Friedman’s Negative Income Tax (NIT) as a precursor. Examining Friedman’s work in the 1940s, the study reveals the polysemic nature of early NIT proposals and underscores the significance of Friedman and Stigler’s engagement with concepts like Abba P. Lerner’s Social Dividend and Kenneth Boulding’s adjustable tax plan, the latter often overlooked in the literature. Contrary to the prevailing view of NIT as solely a poverty alleviation policy, this research also positions it, at this moment, as a macroeconomic steering mechanism. Drawing on Reinhart Koselleck’s categories, theoretical analysis emphasises the importance of exploring diverse “layers of meaning” in early NIT debates, shedding light on nuances that echo contemporary discussions on UBI.

Keywords:

• negative income tax
• social dividend
• Milton Friedman
• Kenneth E. Boulding

JEL CLASSIFICATION CODES:

• B22
• B30
• E62
• I38

1. Introduction: problems with the intellectual history of universal basic income

In this article, we aim to make a modest empirical and theoretical contribution to the growing literature and scholarship on the development of the Negative Income Tax (NIT) idea by Milton Friedman during the 1940s. Our empirical review emphasises two issues. The first is the polysemic and fragmented nature of the early NIT proposals. The second is the significance of Friedman and Stigler’s engagement with the works circulating at that time, especially in relation to Abba P. Lerner’s Social Dividend and Kenneth Boulding’s adjustable fiscal plan, the latter being completely overlooked by the literature. In this way, we seek to provide a counterpoint to the idea that the NIT was solely designed as a poverty alleviation policy by showing that it was also conceived a steering mechanism for macroeconomic direction. In theoretical terms, the article also demonstrates the need to pay attention to the different “layers of meaning” in the early debates on NIT, drawing on some categories from Reinhart Koselleck. Thus, our aim is not to explain a transhistorical idea of Universal Basic Income (UBI) but to analyse a specific period of debates that prefigures the current discussions on UBI.

Following the convention in the literature on Universal Basic Income (UBI), we begin with a consensus definition that allows us to build our argument: the UBI is “a regular cash income paid to all, on an individual basis, without means test or work requirement” (Van Parijs & Vanderborght, 2017, 1). In this massive field, the history of thought or intellectual history is a discipline that has been present since the beginning of the conceptual development of this proposal but that has only recently begun to generate a place and problems of its own. Since the first proposals, associated with Philippe Van Parijs in the 1980s, the possible past of the UBI proposals has always been present in some way in the literature.

This narrative presents a story that begins with some ideas from “Utopia” (1516) by Thomas More and “De Subventione Pauperum” (1526) by Juan Luis Vives, passing through the reform of the British Poor Laws of Speenhamland and the idea of “right to life” (1796–1835), the proposals of Thomas Paine and Charles Fourier up to the Beveridge Report (1942) and the development of minimum income systems in European welfare states (Vanderborght and Van Parijs [1992] 2005). Subsequently, the debate in the context of the 1960s in the United States focused on the problem of “poverty,” with a reference to the proposal by Milton Friedman in “Capitalism and Freedom” (1963), that is, negative income tax (NIT). The NIT proposal has inspired more recent arguments, such as those of the political scientist and director of the conservative American Enterprise Institute, Charles Murray, in favour of the UBI as a way to replace the welfare state (Murray [2006] 2016).

The idea of the UBI seems to have arisen throughout history in disparate genealogies and geographic and temporal contexts for independent reasons: “The pasts or histories consist rather in apparently isolated and episodic occurrences in quite different contexts over some two hundred years” (Cunliffe and Erreygers 2004, xiii). The historiographic problem of how to generate this narrative in a consistent way or even the problem of whether it is possible or desirable remained in the air as a challenge proper to those who work within so-called “intellectual history.”¹

On many occasions, to study the intellectual history of the idea of the UBI, it is necessary to decompose some of the parts of its definition and extend the analysis to similar ideas in some key aspects. The two best known examples are basic endowment/capital, properly a proposal, by Thomas Paine and the aforementioned NIT by Milton Friedman.² In analytical terms, for instance, basic endowment differs from the UBI in that it is a single endowment of capital at a certain age but is similar in that it is individual, universal and unconditional. The NIT, on the other hand, is a periodic, unconditional and individual payment, but it implies some kind of means test, as it is “ex post” of the payment of taxes.³ However, both, alongside other similar proposals, are part of a general history

In this article, readers will find: first, a theoretical and methodological discussion on the role of the NIT in the general intellectual history of the UBI; second, an examination of the different clues to the appearance of this proposal in the thinking of Friedman and Stigler between 1939 and 1947; third, an analysis of proposals that circulated in their environment under the labels of “negative taxation” or “social dividend,” with special attention to Kenneth E. Boulding’s “Adjustable Tax Plan”; and finally, a discussion of the stabilising role that Friedman believed fiscal policy should have during those years.

2. The NIT and intellectual history of the UBI

The latest significant milestones in the historiographical concern for the UBI are Sloman’s (2019) book Transfer State, the collective publication of Universal Basic Income in historical perspective (Sloman, Vargas, and Pinto 2021) and Welfare for Markets: A Global History of Basic Income (Jäger and Vargas 2023). Sloman (2019), in his own way, recognises from the beginning the same problems that we presented in the previous section. Although formally we can differentiate proposals such as the NIT and the UBI—and some others—in regard to making historiography of the discussions and public debates on this idea, we can only verify that these have been intermingled and their limits become blurred: “Any attempt to understand the intellectual history of basic income must start by recognising the fluidity of the idea and the ambiguity of many real-world proposals” (Sloman 2019, 4). Sloman solves this problem by expanding the object of study of his more general history to the debates on different income transfer policies in British politics since the First World War, having the UBI and the NIT as conceptualizations with which to compare and discuss them.

In the introduction to the book “Universal Basic Income in Historical Perspective,” Zamora Vargas and Pinto identified problems similar to those noted above, but rather than addressing them in the same way, they offer a strictly “contextualist” answer, that is, to think of these problems as derived from “mythological” reconstructions, as the historian Quentin Skinner and the well-known “Cambridge School” have pointed out.⁴ The only way to adequately understand UBI proposals is by analysing the social and intellectual contexts in which the texts related to a historical narrative make specific sense. UBI proposals over time and space have been influenced by different languages and points of reference in each context and have been presented by specific individuals and institutions that participate in public debates and waves of disparate interest.

With this criticism in mind, contextualist authors have argued that the UBI proposal is inextricably related to Milton Friedman’s idea of the NIT as part of a broader “market turn” that extended across various traditions (Sloman, et al. 2021, 14). In a more recent article, Zamora (2022) delves into this idea, presenting a broader analysis of Friedman’s intellectual influences. Thus, the NIT is a way of understanding the needs, poverty, and role of the state within the marginalist and utilitarian analyses of well-being of the first half of the 20th century. For Zamora Vargas, this intellectual context would explain why Friedman’s proposal could later be formulated as a transversal utopia in the 1980s in Belgium at the hands of Philippe Van Parijs and the Collectif Charles Fourier.

In another recent article on the history of the NIT, Coleman (2022) makes this idea more complex, showing how tracing the discussions within the Mont Pelerin Society in the 1940s allows us to observe that the possibility of a guaranteed income is one of the fundamental disagreements on libertarian principles, the obligations and the role of the state between the “neoliberals” and their different currents. In reality, there was never an orthodox neoliberal approach to welfare reform, and debates over the NIT reflected broader disagreements about economic methodology, consumer choice, citizenship, policing, and the moral implications of dependency. Coleman shows that the protagonists of Mont Pelerin sought to destroy the welfare state with strategies that they deemed fundamentally irreconcilable. When Milton Friedman presented his controversial idea of the NIT to his colleagues in Mont Pelerin in 1947, he had already had a trajectory in his own thinking.

From the perspective of theoretical problems, there is a clear evolution in the sophistication of the approach from the initial explorations of Van Parijs, to a proposal aligned in some way with the Cambridge School and the problem of Quentin Skinner’s “mythologies,” especially with the work of Jäger and Vargas (2023). This approach is useful in criticising problems of “anachronisms” and “mythologies” that have been built around the intellectual history of UBI.

However, it does not solve the problem of how to generate a diachronic analysis of the development of a certain object of study or the ambiguity and intertwining of the different concepts/proposals. Here, we propose to return to the theoretical arsenal of the historian of concepts Reinhart Koselleck, specifically to the notion of “temporal strata” (Koselleck 2001). This theoretical metaphor refers to “geological formations that reach different dimensions and depths, and that have been modified and differentiated in the course of so-called geological history with different speeds” (Koselleck 2001, 35).⁵

The usefulness of this metaphor is that it allows us to combine concepts with unique and new meanings of the time and place in which they were formulated and the idea that there are circular structures that give us back other meanings from the past. Understanding the UBI in some way as a “concept” related to those studied by Koselleck in politics, it is then possible to reconcile the different proposals throughout history as part of the “semantic” universe of the UBI concept, assimilating their differences but without forgetting that the UBI is an idea that is still in dispute in its realisation and emerges as a set of texts and languages that try to guide future public policies. We can reconstruct certain semantic networks to try to understand what meanings have been attached to the concept we are studying.

Koselleckian concepts are, by definition, polysemic, and the historian’s job would be to show the history of the tensions they harbour within them. We seek to understand how these particular semantic networks are configured and how the meanings of specific concepts are broadened when relating to each other. This is achieved by analysing the uses different actors attribute to them, observing similarities and differences in comparison with the present and certain narratives until the idea that we know today as the UBI, defined at the beginning of this article, is configured. In this way, we can harmonise the idea that the UBI is not analytically the NIT and, at the same time, accept that the NIT is part of the semantic universe that has structured the meaning of the proposal,³ all without the need to think that this is the only layer of meaning that composes it or to build a deterministic “mythology” around the history of the UBI, unsettling it from a neoliberal heuristic of the fight against poverty. This will allow us to finally show that the NIT proposal is not univocal either and that it has its own contexts and sediments of meaning.

In terms of empirical strategy, we will start from the brief mention in Coleman (2022; Zamora 2022), of the existence of a 1939 draft, archived as “An Objective Method of Determining a Minimum Standard of Living” (Friedman 1939), Milton Friedman’s first alleged mention of some of the ideas related to NIT. Based on the examination and assessment of this clue, the objective of this article is to investigate the intellectual context of the proposal of the NIT in the early years of Friedman’s thought immediately before 1947, during the years Nelson (2020b) has called the pre-monetarist period. To this end, we will extend the investigation to proposals circulating in broader intellectual circles at that time, also operating under the logic of “negative taxation.” This will allow us to observe how the NIT proposal is related to a third proposal close to the idea of the UBI, i.e., social dividend (SD), and to add to the literature the evidence of how the genealogy of the NIT links with this perspective of the stabilisation of macroeconomic cycles with fiscal policy and not exclusively as a social or anti-poverty policy.

A complete conceptual history of the UBI obviously exceeds the possibilities of this space, but it is possible to analyse how the previous semantic network was configured that allowed Milton Friedman to construct the NIT proposal in Capitalism and Freedom in 1963 as a chapter in intellectual history of the UBI and what tensions of meaning it had within it.

3. A minimum standard of living

Since the late 1960s, there has been a specific concern about the history of the “negative tax” proposal. Christopher Green, an economist at North Carolina State University, published his doctoral thesis titled Negative taxes and the poverty problem (Green 1967). This thesis presents a genealogy similar to that discussed previously in the history of the UBI, which is already a sample of belonging to the same semantic universe. The story begins in Speenhamland with the British Poor Laws reforms of 1834 and notes direct antecedents from Milton Friedman and Robert J. Lampman (who suggested the subject of study to Green) in the proposal of Rhys-Williams ([1943] 2004). Although he notes that the idea of a negative tax originated in Milton Friedman’s classes at Wabash College in 1956, Green mentions that the idea had been discussed “informally” since 1940 (Green 1967, 57).⁶ With the information currently in contact, the earliest evidence of this could be the Milton Friedman draft we mentioned above titled “An Objective Method of Determining a ‘Minimum Standard of Living’,” but it has not yet been examined in detail.

In 1939, Friedman was 27 years old and had graduated from Rutgers University with a degree in economics and mathematics six years earlier. He had been admitted on a scholarship and studied for a graduate degree at the University of Chicago from 1933 to 1935. After his initial training, he came to Washington to work on the National Resource Committee (NRC) under the New Deal. It is highly probable that widespread anti-Semitism made it very difficult to hire him at that time, forcing him to look for other options. At the NRC, Friedman worked on a series of consumer studies led by Hildegarde Kneeland in a subordinate role until she left the project before its completion (Nelson 2020a, 38).⁷ In his memoirs, Friedman recalls how his work focused on gathering information to calculate a cost of living index; specifically, he worked on the design of samples and on the preparation of surveys. By the end of 1937, he had become an expert in consumer studies and acquired practical statistics knowledge that supplemented his mathematical training. This knowledge was essential to develop further his theory of the consumption function (Friedman and Friedman 1999, 61–66). In 1937, Friedman moved to New York to work under Simon Kuznets at the National Bureau of Economic Research (NBER).

From 1937 to 1940, Friedman worked with Kuznets as a combination of a paid NBER assistant and a doctoral student. His specific task was a byproduct of the US Census Bureau estimates of official national income accounts. Kuznets detected serious gaps in the data on freelancers. The results of the collaborative work of the NBER was the publication National Income and Its Composition, 1919–1938 (Kuznets 1941), where Friedman appears as part of the research team and in the acknowledgements of the preface. Friedman’s doctoral thesis would thus demonstrate the barriers to access and monopolistic practices in the medical sector (Friedman and Kuznets 1945).⁸

This type of work by Friedman is the cornerstone of Nelson’s (2020a) thesis that this era can be labelled Friedman’s pre-monetarist period. Friedman’s research at this time focused on healthcare costs, meaning that in the 1930s, Friedman was interested in how supply-side cost pressures and monopolistic practices arose in the medical profession. This contrasts with his work in the 1990s and 2000s on health insurance and the costs derived from the demand side. In addition, Friedman explored the income distribution of individuals over time, which allowed him to estimate the long-term income distribution with data from only a few years (Friedman and Friedman 1999, 72). This is what led Coleman (2022) to discover that Friedman’s NIT proposal had originated in his attempt to “technically” solve the problem of people with fluctuating incomes. These people who pay taxes when they have income and receive nothing when they do not have it end up paying more taxes in different circumstances, despite receiving the same income (Coleman 2022, 8).

Seen in perspective, we can say that Friedman, in his youth, found himself in the middle of a crucial process to boost state capacity: the development of American national accounting. Although this work had begun before the publication of Keynes’s General Theory in February 1936, as Friedman himself later pointed out, this process was one of the catalysts for the Keynesian revolution and for the shift in emphasis from the economic theory of national income prices (Nelson 2020a, 69). By then, the study of income streams had become much more critical. The draft identified as the first clue to the idea of the NIT in 1939 clearly shows a concern for all these issues that belonged to Friedman’s context before the Second World War.

The draft “An objective method” is filed together with a series of documents that appears to be a correspondence and exchange with Richard Sterner—at that time Gunnar Myrdal’s assistant—and it seems that it may be divided into three parts: a letter addressed to Friedman signed by Sterner, dated June 9, 1939; an unnumbered first draft titled “A plan for Guaranteeing a Minimum Income For All”; and a second numbered draft (1–7) titled “An Objective Method of Determining a ‘Minimum Standard of Living’.” Although the second and third parts may appear to be part of the same draft at first glance, the lack of numbering of the second excludes this possibility. The letter dated 1939 clearly demonstrates that Sterner had read “An Objective Method of Determining a ‘Minimum Standard of Living’” and was in a debate with Friedman about what was the most acceptable method to establish this “Minimum Standard of Living”; however, there does not appear to be any evidence that the paper titled “A plan for Guaranteeing a Minimum Income For All” was part of the discussion at that time, regardless of whether they arrived in this order from the donor.⁹ That is, as Zamora (2022, 3) points out, Friedman himself later noted that the idea of the NIT occurred to him during the early 1940s, but in reality, in this draft, as we will see, we can only discern some elements of its semantic network (the preference for cash over kind), but still cannot qualify as the NIT.

Friedman met Gunnar Myrdal after he visited the National Bureau of Economic Research. In the late 1930s, turmoil in Europe caused an influx of foreign economists to the United States. Myrdal then wrote his famous and influential study on An American dilemma; the Negro problem and modern democracy in the United States (Myrdal 1944). Sterner then became a good friend of Friedman: “At the time, Myrdal was a relatively mild reformer, a New Deal type, not the most extreme proponent of an extensive socialist welfare state that he later became” (Friedman and Friedman 1999, 78). In the letter, Sterner, also involved in calculating the minimum acceptable standard of living, discusses the objectivity of establishing this threshold and the need to show politicians different options so that they know their limitations. Reading this letter suggests that Friedman’s text to which he refers focuses only on technical questions about how to establish this threshold and not on the question of establishing a “plan to guarantee a minimum income to all.”

The unnumbered draft titled “A plan for Guaranteeing a Minimum Income For All” contains, in its first paragraphs, clearly propositional and argumentative content that goes beyond the simple quantitative calculation. Although this text is under construction, several explicitly political concerns are present. Friedman finds in the desire to abolish poverty a legitimate sentiment that, in some way, must be addressed. The problem is that until now, the policies that have been enacted during the New Deal (from the minimum wage, unemployment, old-age insurance, the promotion of unionisation and other types of aid), although widespread, have not yet managed to end poverty. All this aid aims to solve specific problems, but in practice, it leaves large groups without help and has unwanted effects: “It should be attacked by general measures.” Here, the idea that the tax system should be responsible for making these payments in the form of “negative taxes,” a fundamental element of the NIT, does not appear explicitly but, instead, the generic idea that a “general” benefit would be much more efficient than a set of specific benefits.¹⁰

On the subject of how to objectively establish this minimum threshold, in the pages that we can clearly locate in 1939, Friedman discusses the convenience that the nutritional sciences could provide together with the information related to family consumption, with which Friedman, how we saw, had been working on for some years. With this information, two groups of families are distinguished based on their income: those who have or do not have an adequate diet. The next step would be to calculate what Friedman calls the “failure coefficient,” that is, the chances that an increase in family income will not lead to an adequate diet. Here, there is again a relevant argument pertaining to the semantic universe of the NIT. Friedman argues that these calculations are less useful if they were intended for the State to provide aid in kind -food programs- as opposed to cash:

If now the estimate of the minimum standard of living is to be utilized in judging the adequacy of incomes or in reaching some decision as to the amount of money families ought to receive under various circumstances, the case seems far clearer.

Friedman does show here a clear preference for cash aid with a normative argument, i.e., the need to respect people’s decisions:

In a democracy, at least, it is a fundamental premise that, in general, the individual’s choices are to be accepted so that he is the best judge of what he wants and of what is “good” for him. If “we” feel that his choices are faulty, the proper method of changing them is through convincing him that other choices would be preferable, e.g., through education, rather than through coercion. If we wish to judge the adequacy of a given level of income in terms of standards provided by nutritive science, these standards must be combined with the standards set by the individual’s own choices, not substituted for them (Friedman 1939, 5).

Finally, for Friedman, the basic problem of how to establish a “minimum objective” is that this is not only different for each person but that the State should not intervene. In other words, Friedman believes that the state has tools to establish what is a sufficient minimum level of income objectively; from there, it should not interfere in the decisions that people will make.

In short, in these drafts, the idea of “negative taxation” is not mentioned, and the contextual analysis only shows us evidence that he was discussing the issue of how to establish a “minimum standard objectively.” What we can see here is Friedman’s preference for “cash” over food programs due to a normative preference for people’s freedom of choice, one of the fundamental arguments of the semantic network that accompanies the idea of the NIT to the present (Jäger and Vargas 2023). On the other hand, as we will observe below, the specific idea of the NIT is present, with other nuances, in the writings of scholars close to Friedman of the time.

4. Stigler and the NIT

The first specific and published proposal of the NIT, that is, the use of the tax system with the objective of offering a minimum income, was contributed by George Stigler in June 1946 in “The economics of minimum wage legislation.” Despite this, it is also clear that the idea was part of their previous conversations with Friedman, as mentioned in a letter between them in November of that same year (Hammond and Hammond 2006, 42). In his autobiography, Stigler identifies Friedman as responsible for the idea of the NIT, showing an ambiguous position on the matter at the end of his life. Stigler shared the widespread fear among other neoliberals that if applied, such a proposal would add to existing welfare policies and not replace them (Stigler 2003, 156–157).

Stigler and Friedman were together in Chicago during the 1934–35 academic year. Then they met briefly in 1944 in New York at the Statistical Research Group of Columbia University, where they established a close relationship that can be read in their private correspondence. They met again in Chicago in 1947. They began to exchange letters in May 1945, the majority on the correct interpretation of the theory of prices in the work of Marshall’s (1890) Principles of Economics, on the employment situation of both correspondents in a university, or on some of the topics they were working on together. Stigler’s first known collaboration with Milton Friedman was in relation to the regulation of housing prices, published in “Roofs or Ceilings? The Current Housing Problem” (Friedman and Stigler 1946). It is one of the subjects of several of their letters sent between June and November 1946.

Many of these letters focus on a conflict between the director and the Foundation for Economic Education president. The foundation, which would publish Stigler’s work, was concerned that the real estate sector would consider him too egalitarian. The article included a paragraph in which the convenience of reducing income inequality was defended:

For those, like us, who would like even more equality than there is at present, not alone for housing but for all products, it is surely better to attack directly existing inequalities in income and wealth at their source than to ration each of the hundreds of commodities and services that compose our standard of living (Friedman and Stigler 1946, 32).

This point is relevant because in a generic way, it helps us understand the values of Friedman and Stigler in this period associated with the NIT proposal. In the draft analysed above, we glimpse the Friedman preference for “cash” and, in this case, the evaluative elements in favour of greater equality of income.

In his 1946 article on the minimum wage, in which he presents his NIT proposal, Stigler analyses the effects of the minimum wage on the distribution of resources, family income, aggregate employment and poverty. Stigler argues that minimum wage policies are inadequate regarding the last objective and presents possible alternatives, including the possibility of an NIT. As is to be expected, the problem identified refers to managing the incentives that may arise with the different interventions. For Stigler, the problem of incentives must be treated as a matter of degree because the reduction in specific incentives to improve income

is a price we must pay, just as impairment of incentives is a price we have willingly paid to reduce the inequality of income by progressive income and estate taxes. Society must determine, through its legislators, what minimum income (or addition to income) should be guaranteed to each family (Stigler 1946, 364).

That is, in the face of a shared objective, to reduce income inequality, it is impossible not to upset the incentives in some way; it is about finding the best option. The NIT could be this option.

In this article from 1946, Stigler analyses the need to establish a minimum goal based on the nutritional needs raised by Friedman in 1939, which also shows us the same thread of discussion. Stigler’s first conclusion is providing monetary grants is administratively simpler than providing in-kind aid to alleviate poverty. The second conclusion is that this grant should not be based on the specific situations, but on a general criterion of income deficiency according to family composition. However, direct money transfers can be executed through various administrative strategies. While some of these may pose problems, “There is great attractiveness in the proposal that we extend the personal income tax to the lowest income brackets with negative rates in these brackets” (Stigler 1946, 365). This tax system would allow everyone a payment based on their needs with a minimal administrative cost. Again, although he acknowledges that this could also create some problems and the existence of a trade-off between intervention and incentives, in his argument, there is a clear appreciation in favour of equality: “This is a price we must pay, just as impairment of incentives is a price we have willingly paid to reduce the inequality of income by progressive income and estate taxes” (Stigler 1946, 364). The NIT is the formula that allows for finding a better balance between these variables with respect to minimum wage policies.

Stigler’s proposal had an immediate effect on his surroundings. In a November 1946 letter to Friedman, Stigler mentions how Charles E. Lindblom, another Chicago-trained economist, had sent him a note about the NIT, warning them instead of the problems it would create precisely in relation to work incentives (Hammond and Hammond 2006, 42). In September 1947, the German-American social scientist Fred J. Blum published a harsh critique of Stigler’s article within a broader critique of the marginalist approach to economic policy. Blum’s general argument is that Stigler cheats by focusing his analysis of utility on a specific measure, such as the minimum wage, to eliminate poverty. In reality, the objective of the minimum wage is to eliminate wage poverty, limit downward wage competition and improve aggregate consumption capacity. Furthermore, the adverse effects that Stigler predicts do not all seem reasonable. Blum also dedicates a part of his analysis to commenting on the NIT proposal, raising a question that is still relevant today in the debates about the UBI:

This is neither an “efficient” nor an “equitable” means of eliminating poverty because if applied as an alternative rather than as a complementary measure to minimum wage legislation, it would be nothing but a public subsidy for unfair competition and/or monopolistic exploitation (Blum 1947, 650).

Furthermore, Blum detects a central contradiction in Stigler’s argument, pointed out by many future critics of the proposal and that connects it with the problems of managing economic cycles: he finds no convincing reasons why the NIT will not create inflation problems.

In summary, in Stigler, we see both the problem of preference for cash and the issue of inequality and poverty that we identified in Friedman’s text, but here, there is clearly a specific formulation on implementing a minimum guaranteed income system through the tax system specifically. The criticisms and questions raised show that the discussion is already focused on different income security policies and their effects, including macroeconomic effects. In conceptual terms, we can see many of the uses associated with some UBI and NIT proposals that are beginning to be articulated around discussions about how to end poverty or inequality by reconciling incentives, values and costs. What we will try to show below is that within the semantic universe of the idea of “negative taxation,” there is a broader link that introduces some of the macroeconomic discussions of the moment and the link with another key concept in the history of the UBI, i.e., SD.

5. The SD and the NIT as a steering mechanism

According to the work by Van Trier, it is possible to trace the use of the term “Social Dividend” at least since 1935 (Van Trier 2002). Although in this period economists such as Cole (1935), Meade (1935) and Robinson (1937) mentioned the idea of SD more or less extensively or in detail in those same years, here we will focus on the texts that can show a semantic link with Friedman’s idea of the NIT, focusing our attention especially on Lerner (1944).

Van Trier (2002), in his detailed analysis of the origins of the concept, recognises the question of the polysemy of the meanings of the same concept. Although the concept of SD always refers in some way to a system of benefits paid unconditionally and universally, it is actually used in different senses even within the same work. Within the idea of SD, Van Trier identifies at least four different uses: “Social Dividend as a Redistributive Instrument,” “Social Dividend as a Steering Mechanism,” “Social Dividend as a Societal Framework,” and “Social Dividend as an Institutional Support.” This is clearly an example of the conceptual polysemy with which Koselleck works, and we believe that it is applicable to the study of the NIT herein. As we will see, the use in this period of the concept of “negative taxation” as part of this semantic network also moves between some of these uses.

Meade and Robinson were part of an intellectual network of economists created in the 1930s around the Review of Economics Studies. This magazine became a space for discussion between young academics from Cambridge, Oxford and the LSE who debated proposals for what came to be known as “market socialism.” In 1936, as part of these discussions, the economist Oskar Lange, who would later become Poland’s vice president in 1957, briefly proposed the possibility of paying a “social dividend” proportional to wages (Van Trier 2018). However, Abba Ptachya Lerner, one of RES's co-editors, is the one who developed this idea most deeply.

Lerner, born in present-day Moldova when it was still part of the Russian Empire and raised in London, studied at the LSE and is today considered, along with Lange, one of the greatest exponents of “market socialism.” Additionally, in 1936, Lerner published a critique of Lange’s first idea of an SD, noting that a fixed amount for everyone was better than an amount proportional to wages (Lerner 1936), something necessary so that this policy would not interfere with the incentives of the labour market. That is, Lerner was expanding the concept here towards a more universalist notion, not limited by employment status and the distinction of being employed/unemployed, and therefore beyond proposals such as unemployment benefits or wage supplements. According to Van Trier (2018), Lange incorporated this criticism later without mentioning that it came from Lerner. The topic of SD eventually became a specific and relevant topic for Lerner in his best-known and most discussed contemporary work, The Economics of Control (Lerner 1944), on which he began to work in 1930 (Van Trier 2002, 21). As we will see, Stigler (1945) and Friedman (1947) read and criticised this book.

In Chapter 21 of The Economics of Control, “Interest, Investment, and Employment,” Lerner directly suggests that the idea of SD should be independent of the amount of work one does, and should also be considered a SD in “negative, that is, a tax” when macroeconomic conditions require it. In this chapter, Lerner tries to demonstrate the interrelationships and problems between an interest rate policy and a price control policy that allows controlling inflation and other variables in a collectivist economy within the framework and marginalist logic. All of this is in line with the Keynesian concern for aggregate demand. In his scheme, an SD could be part of workers’ income, in addition to the salary, and the amount of this, although distributed independently of the work, could vary based on the macroeconomic needs and align with other policies. That is, Lerner was announcing the possibility of SD conveyed by the tax system “as a Steering Mechanism.”

Lerner is concerned at the time that governments only considered interest rate adjustments as the exclusive tool for managing inflation. A reduction in interest rates could lead to a wage-price spiral, and an increase in rates could result in an even worse situation of deflation and unemployment. To overcome this dichotomy, Lerner proposes a policy of “functional finance,” where the payment of an SD becomes a central element, acting as what today would be called an “automatic stabilizer.” To avoid spirals, the government must have the ability to reduce spending when investment increases and increase spending when investment decreases, keeping aggregate demand at a level sufficient to avoid unemployment but without causing inflation. The alternative to interest rate adjustments is the ability of a “collectivist economy” to act directly on income and, therefore, spending. Consumers receive part of their income from their work, and the rest comes from the government: “This can be considered as the citizen’s share of the earnings of the factors of production other than labour” (Lerner 1944, 267).

Although this redistributive dimension of earnings is introduced, here, the SD is clearly a way of introducing a new instrument of fiscal policy so that the government can intervene in the economy. There is no direct objective interest in inequality or in eradicating poverty:

The distribution of this “social dividend” may follow any principle that pleases the government. The only provision that must be made in the interest of the optimum use of resources is that the amount paid out to any individual should not in any way be affected by the amount of work he does (Lerner 1944, 267).

Lerner emphasises this point because he believes that it is important to keep the salary paid by the employer equal to the marginal product of labour and that the dividend should not affect this:

My personal inclination is for an equal share to be given to each member of society as his right as a citizen, with no questions asked and no exceptions. There could be no better safeguard of the freedom and independence of the individual (Lerner 1944, 268).

However, as we see, the redistributive issue is ultimately considered secondary.

With this framework, an SD could be reduced to zero, as public expenditure, at some point, which would mean that wages and salaries are capable of producing all the available factors of production and could even lead to an increase in “negative social dividend” taxes, reducing the level of demand and acting as a monetary policy tool to coordinate the rise and fall of interest rates in a single coherent policy. The idea of “negative taxation” appears within this framework as one more option that the government has to increase the amount of money in the hands of the population together with the increase in national loans or to print more money:

If the government wishes to increase the wealth or income or expenditure either of particular individuals or classes of individuals or of individuals in general it can bring this about by reducing taxes. Where this is not sufficient to bring about the required results even when taxes have been reduced to zero, negative taxes can be imposed. This means that the government instead of taking money away from people gives it to them. This may take the form of relief payments, old age pensions, bonuses and even a social dividend when it is desired to increase consumption all round (Lerner 1944, 310–311).

Here, there is clearly a correspondence of meaning between the NIT and SD, which fluctuate between meanings as a redistributive instrument and its function as a steering mechanism.

Let us now try to analyse how Friedman and Stigler interpreted these proposals within the uses of the NIT that we saw earlier. In March 1945, a review of Lerner’s book published by Stigler appeared in the Journal of the Academy of Political Science. The economist, who was between the University of Minnesota and Columbia at that time, can be assumed to have read the book during the writing process because his name appears in the acknowledgements. In his later review, Stigler points out his disagreements with Lerner, mainly because he considers his arguments in favour of greater equality in income distribution as “unnecessary” for his argumentation and his economic policy proposals too “vague.” Stigler also briefly analyses the proposal of functional finances, in which he simply does not see what type of macroeconomic indicators he would end up following, and he does not see demonstrated the ability of the monetary authority to influence unemployment and inflation but does not make any specific reference to the SD as such (Stigler 1945).

Friedman goes into slightly more detail in his critique of Lerner’s book in a review published in October 1947 (Friedman 1947). His main criticism is similar to that of Stigler: the imbalance between the formal analysis of optimum welfare and the institutional problems in applying the measures as policies. At this point, some of Friedman’s best-known arguments against economic planning are already evident, which Lerner largely shares: the impossibility of the state being able to coordinate the allocation of resources because it cannot obtain all the current information from the economy and the necessary role of the market price mechanism to generate the appropriate incentives.

The most interesting aspect of this criticism is that Friedman tries to discuss Lerner’s assumption that there is a convergence between income equality and optimum welfare. He criticises Lerner’s formal logic for its failure to account for the ramifications of equally dividing income within the set of mathematical assumptions. An alternative approach proposed by Friedman would involve considering the concept from the lens of what is now referred to as “predistribution” (distribution of resources), rather than “redistribution” (distribution of income):

The difficulty here is that the distribution of income is itself in considerable measure a resultant of the process of satisfying the mathematical conditions for an optimum utilization of given resources. Analytically, therefore, the distribution of income is not an independent “given” that can be manipulated without affecting the rest of the analysis. This difficulty could have been largely avoided by considering instead the distribution of resources. This point is of more than formal interest, since it suggests that measures to reduce inequality by altering the distribution of resources (such as social investment in the training of individuals, inheritance taxation, etc.) may interfere less with the optimum utilization of resources than measures that seek to redistribute income directly (Friedman 1947, 410).

Beyond the formal question of calculating the optimum, Friedman considers that public investment measures or forms of inheritance taxes could be better ways of intervening in the primary distribution of resources, better than directly redistributing income. He then views Lerner’s SD precisely as an “institutional arrangement” just for income equalisation, which illustrates Lerner’s priority for equality over maximisation of satisfaction. It is relevant here because he considers it mainly an institutional technique aimed at intervening in “the primary distribution to individuals for the use of their resources,” analytically separated from the “functional finance” proposal and the cycle problem.

In the following section, Friedman discusses Lerner’s assumptions about fluctuations in economic activity, the “erroneous” Keynesian conclusion that the important thing is to maintain an adequate aggregate demand, and then the conceptual construction of “functional finance.” The main problem is that the government would not have the necessary instruments to determine which is the adequate aggregate demand to be able to intervene. The lags between state intervention and its effects, and vice versa, create a gap of too much uncertainty to capture and intervene in macroeconomic cycles. The corrective measures of “functional finances” on many occasions could, for this reason, end up being counterproductive. Friedman here interprets Lerner’s SD as a redistributive instrument and not as a steering mechanism, as we saw that Lerner actually presents it.

The following is the issue. Despite being able to be considered in practice to be implemented by the same fiscal mechanisms, the SD, as presented by Lerner, is not an automatic correction mechanism when it falls below a certain poverty level but is dependent on political decisions and the macroeconomic situation. Therefore, it can become a “negative” and become a tax. That is, its level is subordinate to later optimal well-being, although this is presented in egalitarian terms of income. Friedman’s reading of Lerner does not show signs of relating the SD to the NIT because he does not recognise that this is carried out by the tax system; he evaluates it instead as a tool to correct income inequality.

Friedman’s transversal criticism of Lerner is precisely that he is not considering, nor does he know as much as Friedman does, the real functioning of the government and the most feasible tools to carry out certain policies. This leads him not to take Lerner’s specific proposals too seriously and to focus his critique on the coherence of the formal analysis, where the SD and NIT do not seem to be proposals aimed at addressing the same issues. However, the explicit semantic connection between these two proposals does appear clearly in another of the economists of Friedman’s environment who was working on these issues at that time: Kenneth E. Boulding.

6. Adjustable tax plan

The SDs we have seen thus far only superficially mention the possibility of considering them in practice as an NIT because they remain largely separated in the formal analysis of macroeconomic dynamics and optimal welfare. Another person from Friedman’s environment who debated proposals similar to the NIT and SD in those years, generally absent in the literature on the subject, was Kenneth E. Boulding. The British economist and philosopher wrote about the idea of the NIT and SD in his book The Economics of Peace (Boulding 1946).¹¹ He came to the United States on a Commonwealth scholarship to study at the University of Chicago in 1932 and met Friedman as a graduate partner, and the two maintained an academic relationship (Friedman and Friedman 1999, 39)¹². Later, in the late 1960s, Boulding created the first organisation the American Economic Association endorsed to study “grants and transfers,” called the Association for the Study of the Grants Economy (Turgeon 2015).

Boulding, in a chapter in which he analyses the proposals and solutions to unemployment in The Economics of Peace, also analyses the capacity of the State to stabilise the monetary income of families. While suggesting that stimulating investment remains critical in the short term, the British economist believes that increasing consumption is the most crucial problem facing Western societies in the long term. Boulding believes that what he calls the classic “steady state” is approaching, that is, the moment when capital will have already occupied all the necessary spaces to revalue itself, and no further investment will be necessary (Boulding 1946, 172–173).¹³ For Boulding, this will come after the expansive effects of reconstruction after the devastation of the war in Europe and after a possible expansion of investment in Asia. In this situation, courageous management of monetary and fiscal policy will be necessary under technical capacities to avoid massive unemployment and impoverishment. For Boulding, a system left to itself produces meaningless fluctuations in monetary income that can be avoided with government deficits.

The first mention of SD appears in the book at the end of Chapter 6 dedicated to analysing “justice in distribution.” In the section where he is analysing the public finance system as the main tool for income redistribution, Boulding wonders about the difficulty of determining how much redistribution should be provided in cash and how much in kind and who should be the beneficiaries. There, he presents the idea of SD as a redistributive instrument, as the universalist commitment to cash:

Every member of a society, rich and poor alike, might be regarded as sharing a potential property right in the economic surplus, and therefore entitled to a money income from the state—an income which would, of course, be taxed away again in the case of the rich. There is a certain attractive simplicity about this idea of a “social dividend” (Boulding 1946, 119).

The problems that an SD could bring in the form of a direct universal benefit regarding the incentives to work makes Boulding here seem to lean towards the view that “the alternative solution of a complex system of social security benefits is probably more desirable, though less elegant.”

However, the SD also reappears later when Boulding is instead discussing monetary policy strategies and increasing the velocity of money in the form of “depreciating stamp money.”¹⁴ The main problem that Boulding identifies in the “depreciating stamp money” scheme is its lack of flexibility. Although the concept of making money lose value over time can be useful to increase the velocity of circulation in times of unemployment or deflation, the scheme does not adequately consider substitutes for money, such as bank deposits. If it is implemented only as a currency with depreciation stamps without affecting bank deposits, people would simply switch to bank deposits or other substitutes. Furthermore, the scheme is not well adapted to periods of inflation because it would promote increased spending when the opposite would be desirable. This is the same problem that he predicts that the SD would have (equated to Major Douglas’s concept of “Social Credit”), although the idea of providing a certain amount of money to each individual on a regular basis can have benefits in times of severe deflation; in periods of inflation, it becomes inadequate (Boulding 1946, 153–154).¹⁵ In addition, he mentions as a similar proposal the “Townsend Plan,” a quasi-universal pension proposal financed with a transaction tax, which was very popular during the 1930s in the United States.¹⁶

In Chapter 9, dedicated to the “Full Employment Policy,” Boulding instead discusses an alternative that would not have these problems “The possibility of a negative taxation”:

There is no particular reason why tax rates should stop at zero, if it is necessary to counteract a large deflationary movement of incomes. A negative tax rate would mean, of course, that the government would pay money to the taxpayer instead of taking money from him (Boulding 1946, 164).

Although Boulding admits that this idea could be surprising to his audience and that it may never be necessary precisely because of the expectations that automated price stability would generate, he presents it, like the SD of Lerner and Lange, simply as a mechanism for generating automatic deficits of the State and sustaining the income of the population in periods of deflation. The tax rate fluctuates between positive and negative levels depending on the existence of the risk of inflation or deflation.

The name Boulding uses for his proposal is “The adjustable tax plan.” Like Lerner’s proposals from SD, Boulding’s proposal from the NIT is focused on influencing monetary policy:

Once we get out of our heads that the primary purpose of taxation is to “raise money for the government to spend,” we can see the true purpose of taxation, as the most powerful weapon in our hands for the stabilizing of money prices and incomes (Boulding 1946, 165).

However, Boulding then analyses the distributional effects that his system of “adjustable taxes” would have. This mechanism would alter a progressive income tax system, and in times of deflation, when taxes were lowered, the rich would also benefit. Boulding does not see many possibilities of this happening frequently, but above all, he thinks that this would be rewarded in the opposite periods where the rich would pay more. However, in any case, he thinks that his system could work from a progressive standard tax scheme, based on the volume of normal public spending, on which his NIT scheme could be built on, where that progressive basic tax structure would not be touched.

That is, through Boulding, we have a different combination of the semantics of these concepts. SD is understood only as a constant universal benefit and not dependent on income level or another macroeconomic perspective—that is, more similar to the contemporary idea of the UBI—and the possibility of an NIT instead with macro stabilisation objectives because it can also act automatically to restrict or expand consumption capacity—that is, more like Lerner’s idea of SD.

7. Friedman and the role of fiscal policy in economic stability

Until the early 1950s, Friedman also believed that fiscal policy should be a fundamental instrument of public policy for macroeconomic stabilisation, growth, and inflation (Greenwood 2022). In the first article of 1942 on the subject of inflation, his classic monetary question was not even mentioned (Friedman 1942)¹⁷. By 1953, he had completely changed his mind; fiscal policy played practically no role in stabilising the business cycle, and the focus was on monetary policy (Friedman 1953b). In this pre-monetarist period, Friedman reflected on the same problems that Boulding faced regarding the state’s role in intervening in the economic cycle in an article published in 1948 titled “A Monetary and Fiscal Framework for Economic Stability” (Friedman 1948).

This article is, in many ways, a continuation of his comments to Lerner and Boulding. In this article, Friedman begins by pointing out that he knows that what he presents is not very original but is rather a common denominator and space for consensus between various proposals that existed at that time, including that of Boulding. In addition, here, he clearly maintains the idea that the objective of substantial equality is part of the objectives that any economist should have: “The basic long-run objectives, shared I am sure by most economists, are political freedom, economic efficiency, and substantial equality of economic power” (Friedman 1948, 246). In addition, while we wait for a truly competitive market to be in place, Friedman is clear that fiscal measures are the best way to attack it:

While a truly free market in a “competitive order” would yield far less inequality than currently exists, I should hope that the community would desire to reduce inequality even further. Moreover, measures to supplement the market would need to be taken in the interim. For both purposes, general fiscal measures (as contrasted with specific intervention) are the most desirable non-free market means of decreasing inequality (Friedman 1948, 246).

The centre of Friedman’s analysis is to analyse what type of responses the government should give to changes in the economic cycle. Friedman’s proposal includes proposals for reforms to the monetary system, public spending on goods and services, public transfers, and the tax structure. He refers to the transfer programs - to the social security system - and to the need for these to be maintained regardless of the phase of the cycle and that these payments will simply be high when unemployment is high and vice versa, in a footnote:

These transfer payments might perhaps more appropriately be regarded as negative revenue. […] It may be hoped that the present complex structure of transfer payments will be integrated into a single scheme co-ordinated with the income tax and designed to provide a universal floor to personal incomes. But this is a separate issue (Friedman 1948, 248).

The essence of Friedman’s proposal is that it seeks automatic adaptations outside the government’s discretionary control so that the flow of income can compensate for the problems of aggregate demand in certain areas and modify the money supply. Public deficits and surpluses control the amount of money in the economy. According to Friedman, surpluses destroy money; therefore, he proposes to combine monetary and fiscal policy and use the budget to control the currency issue in a countercyclical way to stabilise the economy. Finally, Friedman’s proposal of 1948 actually shows the basic elements of consensus with Lerner’s proposal for “functional finance” and for Boulding’s “adjustable fiscal plan.” This shows a certain dominant consensus space in the discussions at that time, in which Friedman participated, on using fiscal policy to stabilise business cycles.

In this text, for Friedman, there are two conditions for the economy to return to full employment with these mechanisms: spending must be countercyclical, and taxes must be pro-cyclical. This implies a strong social safety net for spending on cash transfers to rise sharply in a recession:

Deficits or surpluses themselves become automatic consequences of changes in the level of business activity. When national money income is high, tax receipts will be large and transfer payments small; so, a surplus will tend to be created, and the higher the level of income, the larger the surplus. This extraction of funds from the current income stream makes aggregate demand lower than it otherwise would be and reduces the volume of money, thereby tending to offset the factors making for a further increase in income. When national money income is low, tax receipts will be small and transfer payments large, so a deficit will tend to be created, and the lower the level of income, the larger the deficit. This addition of funds to the current income stream makes aggregate demand higher than it otherwise would be and increases the quantity of money, thereby tending to offset the factors making for a further decline in income (Friedman 1948, 251).

As mentioned, the political focus of Friedman’s argument is that the government would ultimately not intervene in a discretionary way but rather that intervention would be automated in the tax system. Although the effects of this type of system on full employment do not seem to be guaranteed automatically—the transmission can be delayed in time, and the rigidity of certain prices must be taken into account—it would only be necessary to adjust to make them work. He criticises approaches such as Boulding’s because Friedman does not believe in the “stationary state” doctrine. It is in this climate of discussion that the idea of “social dividend,” “the adjustable tax plan” and “negative taxes” appear somewhat interrelated and are consistent with Friedman’s views and discussion framework.

This can also be seen in the well-known intervention in the April 8 session at the Mont Pelerín Society meeting in 1947, in which Friedman proposed the possibility of an NIT to his colleagues. Although the emphasis of his intervention was on the “problem of poverty” and his colleagues discussed the issue of unconditionality, he also considered relevant the argument for the role of the NIT in stabilising macroeconomic cycles:

Purely in terms of the Elimination of Poverty. But it also has an anticyclical effect. If you manage to have a liberal society with this flexibility, it would be very good from a cyclical point of view. […] Regarding the business cycle, a progressive income tax is a form of automatic stabilizer which a negative income tax further strengthens. It supplements incomes in downturns and collects more in taxes in booms, thus reducing recessionary or inflationary pressures at the appropriate points of the cycle (Taylor 2022, 188).

In 1963, there is no trace of this allusion to the “steering mechanism” in the last chapter of “Capitalism and Freedom” (1963). Possibly, this has led to the adoption of an excessively narrow view of the literature from the 1940s on the subject. This transition aligns with theses such as Ben Jackson’s (2010), suggesting a clear differentiation between the early theorists of neoliberalism in the 1930s and 1940s and those in the 1960s and 1970s. “The most striking feature of the ideological map in the late 1930s and early 1940s is therefore that, in some respects, the gap between market liberals and the moderate left actually narrowed.” (Jackson 2010, 150). Just as Timothy Shenk (2023) suggests that removing the “mythological” lenses of neoliberalism in these decades allows us to see more clearly the “Keynesian ghost” of discretionary economic management, we could say the same regarding the “war on poverty” lenses of the 1960s. Friedman was part of a generation that agreed on the need for macroeconomic stabilisation allowing for a certain closeness between market socialists and neoliberals.

8. Conclusions

As noted in the introduction, investigating the intellectual history of the UBI requires letting go of the idea of a linear and progressive evolution of the proposal as well as the connections and coherence between the authors who have written about it and about the use of specific concepts. This idea appears fragmented and dispersed, and the semantics that are part of its conceptual universe have their own internal developments and tensions. The idea of the NIT that appears in the last chapter of Capitalism and Freedom in 1963 by Milton Friedman can be constructed from the conditions of possibility that give a series of antecedents and specific contexts that have been generating sediment. In this process, some of its meanings are retained and reinforced, while others fade away; however, nothing prevents them from resurfacing. When we remove the neoliberal lens and focus our attention on its intellectual environment and the emergence of the NIT idea in those “pre-monetarist” years, we see discussions arise around welfare economics and market socialism, fiscal and monetary policy, and the possibility of a “social dividend” or “adjustable tax plans,” with formulations sometimes interchangeable or not entirely clear. The idea of “negative taxes” also emerged as a form of intervention in these discussions and to establish a distinctive standpoint.

The Koselleckian formula of thinking in terms of “layers of meaning” allows us to understand and organise the reading of these texts, starting from the problems related to the UBI and the way of constructing and thinking about its intellectual history. This geological metaphor incorporates the idea that the concept of the UBI has specific meanings in each context and even in linguistic formulations (such as the NIT or SD) but recognises that there is a certain repetition of some of the arguments, which we are trying to reconstruct. In this way, talking about a “semantics” of the UBI and of the universe of concepts accompanying it facilitates and clarifies the research parameters on this topic.

Between the 1930s and 1940s, Milton Friedman was immersed in developing the capacity of national accounting and state techniques within the extraordinary framework of the New Deal. This directly affected his views, which reflected the dominant consensus on the need to increase income equality and expansive intervention in depressive business cycles but also gave him practical knowledge about the real tools that governments could use to intervene beyond formal calculations and formulations.

The contextual analysis shows us that in 1939, Friedman still did not mention the idea of “negative taxation,” but hints at the concept within a broader discussion about the possibility of objectively establishing a “minimum standard” and the preference of cash over food programs. Stigler, in 1946, first expressed in writing the idea of the NIT in his closest environment in clearer terms as a form of minimum income implemented through the tax system. Stigler intervenes in the discussion regarding the best programs to avoid inequality and working poverty in contrast to minimum wages. The debates that Stigler’s article generated show that the discussion is already focused on different guaranteed income policies, their incentives, and their macroeconomic effects.

The semantic link between the idea of “negative taxation” and the discussions on how to intervene in economic cycles becomes more evident when we observe its relationship with another protagonist of the intellectual history of the UBI, i.e., the idea of SD within Friedman’s intellectual space, especially when we focus on the discussions he had with Abba P. Lerner. In The Economics of Control (Lerner 1944), the SD proposal is not a mechanism for eliminating poverty, or at least this is not its main role. Its amount ends up depending, for Lerner, on formal political decisions on the macroeconomic situation, so much so that it can become negative, that is, a tax, adopting, therefore, the form of an NIT when the SD is positive. Friedman at the time did not distinguish this parallelism in Lerner’s proposal, mainly because Lerner maintains himself above all in a formal analysis of “functional finances” and does not develop the real implementation process that Friedman himself is capable of imagining.

On the other hand, the relationship between these two proposals and the two possible functions is much more explicit in the case of the work of another of Friedman’s colleagues, Kenneth E. Boulding, in The Economics of Peace (Boulding 1946). Boulding understands SD only as a system of universal benefits with all the characteristics of what we now call the UBI, that is, a benefit oriented exclusively in terms of the egalitarian distribution of a portion of the collective surplus. An NIT would instead be, for him, a way to avoid the macro imbalances that a constant SD would generate, thus understanding it directly as an automatic stabilisation mechanism that restricts and increases the consumption capacity (the way in which Lerner understands the SD). This same idea is defended by Friedman (1947, 1948), in which the NIT is not only an anti-poverty mechanism but also an alternative to the social security system with automatic adjustments to the macroeconomic situation.

When we look at their specific linguistic uses, in some way, the NIT and the SD are two sides of the same coin. In reality, the meaning of both changes depends on the emphasis placed on their redistributive aspect or as an economic management mechanism, and in many cases, their definitions are confusing or incoherent. The political or normative objective (equality of income, resources or the end of poverty; or the automatic stabilisation of economic cycles) and the specific tool (a benefit or a tax) are mixed in the readings that both make use of the same concepts, causing them to fluctuate systematically throughout this semantic network.

Koselleck resorts to another category to explain the usefulness of his theory of “strata,” the “simultaneity of the non-simultaneous”: the presence at a given moment of categories and concepts typical of other contexts that coexist simultaneously. A concept such as the UBI in our present contains different meanings that coexist, even when they can be perceived as contradictory and in tension. The idea that it is necessary to distribute the fruits of private property equally among the entire population because that implies access to certain political rights is how Thomas Paine thought of the UBI at the end of the 18th century. This idea coexists with that of the UBI as a way of freeing individuals from the constraints of a network of public policies designed by the state to end poverty. In the same way, the NIT, both in the conception of Milton Friedman and in his intellectual environment, can be considered a proposal strategically positioned against other types of policies, such as minimum wages, and is thought of as an SD and a form of economic planning to generate state deficits automatically through fiscal policy. This “macro” perspective of the UBI and NIT proposal has traditionally received less attention than its approach as a tool to combat poverty. Currently, this aspect of the idea continues to be discussed at different levels, linking the idea of financing the UBI as a “Monetary Helicopter” (Mencinger 2017) or directly as an alternative monetary policy in the form of “stabilisation grants” (Wehner 2021, 37–48). The study of the intellectual history of UBI allows us to discern the genealogies that link all these ideas to the same semantic universe of debates. Moreover, it enables us to interpret the meaning of UBI as a battleground shaped by different worldviews that are still contending to be defined today.

Acknowledgments

This article is the result of the Postdoctoral Fellowship Program at Universidad Nacional Autónoma de México (UNAM), Ciudad de Mexico, Fellow of the Instituto de Investigaciones Económicas, advised by Dr. Oscar Ugarteche. I would particularly like to thank Dr. Ma. Teresa Aguirre Covarrubias for their comments, as well as the discussion space of the AISPE 2023 Colloquia doctoralia and the suggestions of Dr. Cosma Emilio Orsi. Finally, thanks to the anonymous reviewers who have provided comments that have helped substantially to improve this work.

Disclosure statement

No potential conflict of interest was reported by the author(s).

Additional information

Funding

Programa de Becas Posdoctorales, Universidad Nacional Autónoma de México (UNAM).

Notes

1 Although less known, for being unpublished, the first important antecedent of the concern for the properly historiographical questions about the UBI was the doctoral thesis of Walter Van Trier, directed by Philippe Van Parijs, titled Everyone a king: An investigation into the meaning and significance of the debate on basic incomes with particular reference to three episodes from the British interwar experience (Van Trier 1997). This work constitutes the first attempt to address the essential epistemological questions and problems related to the possibility of historicising the UBI in its own terms.

2 The compilation of texts that speak of Basic Capital and Basic Income is made in parallel in the anthology by Cunliffe and Erreygers (2004) that we mentioned above is an expression of this problem.

3 For the discussion on Thomas Paine’s proposal and the history of the UBI, one can read Camporesi (2023). To explore the question of the conceptual and practical difference between NIT and the UBI (Van Parijs and Vanderborght 2017, 32–40).

4 British historian Quentin Skinner headed the so-called Cambridge School. He renewed the history of these political ideas in the 1960s, arguing that classical historians of ideas got caught up in a kind of “mythification” of one’s own ideas in history. This “mythification” led them to create teleological and linear narratives from a presentist perspective and fall into a persistent anachronism. Although it is an approach that is still discussed today, since then it has been necessary to question the possibility of a “history of ideas” and always consider how different contexts determined contingent ideas at each moment (Skinner 1969).

5 Although it is a historiographical tradition that is little used for the history of economic thought, there are fruitful and inspiring examples of using some of Koselleck’s categories in this field; see Eich (2022) or Tribe (2015).

6 Friedman, in his lecture notes on economic cycles, mentions having read ‘Looking Backward’ by Edward Bellamy (Friedman, 1940). This book, an egalitarian utopian novel, has sometimes been pointed out as a precursor to the idea of UBI. However, in line with our hypothesis, it seems that Friedman was interested in Bellamy precisely as an example of a vision of a world with the ability to avoid major economic cycles rather than poverty itself.

7 The publication resulting from the work with Kneeland was Consumer Expenditures in the United States: Estimates for 1935-36 (United States National Resources Committee Industrial 1939).

8 Friedman’s income comparisons between different professions also allowed him to suggest that barriers to entry into the medical profession increased the income of physicians. This research generated a storm of criticism from the American Medical Association when it was made public.

9 I thank the Hoover Institution and the clarifications that Daniel Keough provided me regarding the way in which these documents had been filed.

10 Although it is difficult to determine where exactly this page belongs without numbering its first sentence, “The poor are always with us” argumentatively recalls the content of his speech at Mont Pelerín in 1947 “Even if we had completely free access to different employments, and to capital, there would still be the problem of poverty” (Taylor 2022, 184).

11 Although not essential to the interests of this article, interestingly, Boulding at the time was a member of the Quaker organization Religious Society of Friends. Other proponents of ideas similar to the UBI, such as Thomas Paine in the late eighteenth century (Camporesi 2023) and E. Mabel and Dennis Milner (Sloman 2018), were also Quakers, as was Fred J. Blum, mentioned above.

12 In 1953, he published with Stigler a selection of texts on price theory in which Friedman also participated (Boulding and Stigler 1960).

13 The possibility of a “steady state” as an idea of classical economists and its possibilities was being discussed at the time (Pigou 1943). This is one idea that serves as the basis of the current conception of “secular stagnation” (Backhouse and Boianovsky 2016).

14 The main idea behind this scheme is that money loses value as it is retained. To achieve this, money holders would be required to affix a stamp to the back of banknotes at regular intervals. Boulding mentions that this was put in place in Alberta, Canada; the banknotes had 52 spaces on the back, one for each week of the year. A two-cent stamp was to be affixed to each space until the week the money was spent. If someone held the money for four weeks, they would have to buy 8 cents worth of stamps to stick on the banknote before they could spend it. At the end of the 52 weeks, the fully stamped banknote would be redeemed by the state government.

15 To elaborate on Douglas’s idea in relation to these discussions, one can read (Van Trier 2005). He does not quote Lerner when he talks about the social dividend, but he does quote The Economics of Control on a couple of occasions (Boulding 1946, 201, 211).

16 For more detail on Towsend’s proposal, read (Buck 2022).

17 Friedman republished this article in 1953 and added a specific comment on this omission to the original article (Friedman 1953a, 251). It seems that the change in stance was mainly guided by the empirical evidence to which he subsequently had access, which, from Friedman’s perspective, made him see that there was no point in concentrating on fiscal policy (Greenwood 2022).