1 Introduction

James Buchanan and Vincent Ostrom were astute students of social and political life who, over the course of six decades of scholarly endeavour, made numerous insightful and influential contributions to economics, political science, and public administration. They were founding members of what become the Public Choice Society, both subsequently serving as President, while in 1986 Buchanan was awarded the Nobel Prize in Economics for his contributions to the economic analysis of political decision-making and public economics; Ostrom found fame by association when his wife Elinor (also a President of the Public Choice Society) was the first woman awarded the Nobel in 2009.

Buchanan and Ostrom were personal friends who corresponded over many decades, co-organised several scholarly conferences, and even contemplated jointly establishing an academic journal to be called Constitutional Review.Footnote 1 Buchanan wrote to Ostrom that their shared intellectual position had “two basic articles of faith … (1) Institutions matter” and “(2) Institutions can be constructed”—a proposition to which Ostrom assented.Footnote 2 This shared intellectual commitment, in the words of Peter Boettke and Alain Marciano (2020: 80), led Buchanan and Ostrom “to cultivate a professional conversation on the constitutional level of analysis in political economy”.

A growing body of scholarship has explored the relations between Buchanan’s and Ostrom’s work. Paul Dragos Aligica and colleagues have highlighted their broad similarities, emphasising their common roots in the public choice tradition, shared emphasis on voluntary forms of association, and how both drew attention to the important distinction between pre-constitutional and post-constitutional analysis (i.e., between the choice of rules and decisions made within the framework provided by those rules) (Aligica & Boettke, 2009: 126–31; Aligica, 2019: 61–67, 185–86). Richard Wagner (2018: 24–25) has highlighted Buchanan’s and Ostrom’s shared emphasis on the importance of constitutionally limited government. Others have underlined how Buchanan and Gordon Tullock’s ([1962] 1999) formal analysis of different decision rules shaped Vincent and Elinor Ostrom’s analysis of the optimal size of the political units within which collective action should be organised (Tarko, 2017: 9–10; Duhnea & Martin, 2021: 11–12). Aligica and Boettke (2011: 34–37) helpfully explained how the Ostroms built on Buchanan and Tullock’s analysis of government failure, first of all by developing concepts—such as the notion of the ‘public economy’—that transcended a simple state-market dichotomy and, second, by analysing how more complex institutional arrangements that do not fall readily into the categories of ‘state’ or ‘market’ can improve social outcomes. Further studies have explored Buchanan’s and Ostrom’s views about the importance of constitutional design, insightfully comparing them with Friedrich Hayek’s emphasis on the role of evolution in the development of an institutional framework capable of sustaining individual freedom, as well as their discussion of federalism (Boettke & Lemke, 2018: 55–57, 66–68; Candela, 2021). The relationship between Vincent Ostrom’s work, and the Bloomington School he helped to found, and the wider public choice moment, including Virginia political economy, has been discussed by William Mitchell (1988: 110–13) and by Aligica and Boettke (2009, 2011).

This article seeks to build on, and add to, this body of work by providing a comprehensive discussion of an important set of issues running through Buchanan’s and Ostrom’s work to which the existing literature has not done complete justice, namely their discussions of the importance of the “logical foundations of constitutional democracy” and, in particular, their changing views about how those foundations should be conceptualised. The term “logical foundations of constitutional democracy”, the subtitle of The Calculus of Consent, describes how constitutional structures can be analysed as the outcomes of the decisions of rational economic agents. Indeed, Buchanan’s (1975, 1986a, 1989a) project of modelling “politics-as-exchange” conceptually grounded formal institutions in bilateral exchanges between individuals—so that every institution, from autocracy and majoritarian democracy to constitutional democracy, was conceived as emerging from exchanges between individuals.

We trace how in the 1960s and early 1970s Ostrom took inspiration from The Calculus of Consent, explaining how it shaped both his reading of key texts in political theory and his analysis of public administration (Sect. 2). We then discuss how Buchanan subsequently drew on Ostrom’s notion of artifactual man as he sought to develop his understanding of the “constitutional attitude” (Sect. 3). Next, we examine how Ostrom became increasingly critical of Buchanan’s continued reliance on rational choice theory to analyse constitutional decision-making, an issue about which the existing literature has said little, before examining Ostrom’s views about the appropriate analytical framework for studying the decisions involved in constitutional choice (Sect. 4).Footnote 3 Section 5 concludes.

2 Ostrom on the importance of Buchanan: the significance of the constitutional level of analysis

Ostrom ([1997] 2012: 83) wrote that Buchanan’s work brought about a “major paradigmatic shift” in political science and public administration. The impetus for this shift was provided Buchanan and Tullock’s 1962 book The Calculus of Consent, which made “a critical step” by “shifting the focus of investigation from the practices of normal politics in collective decision making to practices of extraordinary decision making found at the constitutional level of analysis” (Ostrom [1997] 2012: 85). Prior to Buchanan and Tullock’s classic text, Ostrom argued, few twentieth century political scientists had understood the importance of the constitutional level of analysis—i.e., the analysis of how the rules governing people’s everyday interactions are chosen (Ostrom [1997] 2012: 85, 1999: 123–26). It was, as Ostrom later put it, the centrepiece of a “forgotten tradition” in political science (Ostrom [1982] 1999). Only after reading Buchanan’s pioneering work did political scientists, including Ostrom, begin properly to appreciate that constitutions could be analysed, not just as legal and historical documents, but as the product of “rational calculations of reflection and choice about political orders” (Ostrom [1997] 2012: 85; also see Ostrom, 1987a: 242 and Ostrom and Ostrom [2004] 2014: 61):

Perhaps the single most important contribution to my own intellectual development was Buchanan and Tullock’s (1962) The Calculus of Consent ... [which encouraged] a basic shift in the level of analysis from electoral arrangements and deliberative bodies grounded in plurality voting to the constitutional level of analysis: the specification of the terms and conditions of governance. (Ostrom, 2009a: 61.)

For Ostrom, this change in perspective was especially relevant for those interested in exploring the question posed by Alexander Hamilton in The Federalist No. 1, namely “whether societies of men are really capable or not of establishing good government from reflection and choice or whether they are forever destined to depend for their political constitutions on accident and force” (quoted by Ostrom [1997] 2012: 84–85, 1999: 124). In the twentieth century, however, prior to the work of Buchanan and Tullock, Hamilton’s question had become so peripheral as to be deemed “beyond the scope of a political science”, with political scientists focusing principally on choices made within a given set of rules rather than analysing the decisions that brought those rules into being (Ostrom, 1999: 125; also see 1977: 1509). Buchanan and Tullock put this question back on the scholarly agenda; Buchanan’s (1975, 1986a, 1989a) subsequent development of the theory of politics-as-exchange attempted to return to political science the perspective of the American founders who had believed in the reality of self-government by people who framed the rules that governed their collective lives and to show that such rule-making could be the logical consequence of the individual pursuit of rational, self-interest at the constitutional level.

Ostrom set great store by the subtitle of Buchanan and Tullock’s book—“Logical Foundations of Constitutional Democracy”. It signified that Buchanan and Tullock sought to show how constitutional rules, most notably those of the United States at its foundation, could be analysed and explained as the product of decisions made by rational actors (Buchanan and Tullock [1962] 1999: 24–25, 298–99; Buchanan [1986b] 1999: 19; also see Wagner, 2015: 16, 19–21; Meadowcroft, 2020):

Before reading The Calculus of Consent, I had been only superficially aware of the logical foundations of constitutional democracy. In Buchanan and Tullock’s work, I realised the basic paradigmatic significance of, first, methodological individualism. Second, I came to recognise that ‘conceptual unanimity’ or consensus serves as the base rule for constitutional decision making in democratic societies. Third, I realised the importance of a hypothetical cost calculus in a comparative analysis of diverse voting rules ... Fourth and finally, I saw how the general concept of a basic constitution, as formulated by Alexander Hamilton in Federalist 78, can be extended to all patterns of human association and be constitutive of self-governing societies rather than state-governed societies. The choice of rules in rule-ordered relationships would then apply, as Alexis de Tocqueville recognised, to the development of a science and art of association grounded in what can variously be called constitutional economics or institutional analysis. (Ostrom [1997] 2012: 86; also see Ostrom, 2012: 427–28)

For Ostrom, the ideas found in The Calculus of Consent helped to inform two broad, and closely related, strands of research: the first centred on a particular way of reading texts on constitutional issues; the second involved the development of a political economy perspective on the working of the state.

2.1 Reading texts

Becoming acquainted with Buchanan’s emphasis on the logical foundations of constitutional democracy enabled Ostrom to reinterpret classic texts on fundamental constitutional issues “as though they were presentations being made at the constitutional level of analysis” (Ostrom [1997] 2012: 87; also see Ostrom, 2012: 428). Buchanan and Tullock’s work led Ostrom “to attempt to reconstruct the political theory that was being used by Alexander Hamilton and James Madison as they wrote … The Federalist” as embodying the same kind of economic reasoning—the same logical foundations—as Buchanan and Tullock used in their analysis of constitutional design (Ostrom [1997] 2012: 87):

Those familiar with traditional political theory, including those theoretical formulations used in the conception and design of ‘The American decision-making system’, might be expected to find the formulations in The Calculus of Consent to have close parallels to works with which they are familiar. The language of economists addressing the constitutional level of analysis will have a familiar ring to earlier formulations advanced by the American Federalists and the Scottish philosophers, among others. (Ostrom, 1999: 127.)

Ostrom’s rational reconstruction was set out in The Political Theory of the Compound Republic (Ostrom [1971], 2008), a book originally published by the Center for the Study of Public Choice at Blacksburg, whose dedication read, “for Jim. In appreciation to one who challenged” (the Jim in question presumably being Buchanan) (Ostrom, 1971: ii). The book portrayed Hamilton and Madison as analysing the choice of rules to be included in the US constitution in a manner quite consistent with the style of reasoning upon which Buchanan and Tullock’s based their analysis of constitutional decision-making. For Ostrom, the Federalists were engaged in a project in constitutional political economy similar to the enterprise conceptualised by Buchanan and Tullock (Ostrom, 1977: 1509; [1997] 2012: 87–88, 1999: 129–30; also see Ostrom [1973] 2008: 72–80, 146–47, Ostrom and Ostrom [2004] 2014: 6, and Buchanan and Tullock [1962] 1999: 298–99).

Buchanan, of course, explicitly located his work in the Federalist tradition. He wrote that he and Gordon Tullock, “more or less explicitly considered our exercise” in The Calculus of Consent “to be an implicit defense of the Madisonian structure embodied in the United States Constitution” (Buchanan [1986b] 1999: 19). He and Geoffrey Brennan ([1985] 2000: xv) wrote of their approval of “the veneration Americans accord their Founding Fathers [...] James Madison, Benjamin Franklin, Thomas Jefferson”, who were “distinguished by their essential understanding of the reason of rules in political order, an understanding they implanted in the constitutional documents, the ‘sacred’ texts that have, indeed, worked their influence through two centuries”. Moreover, Buchanan (1975, 1986a, 1986c) saw the denigration of the Federalist tradition in American public life as the key to the decline of American democracy in the twentieth century that had left the country exposed to the very real threat of an unconstrained Leviathan.

Constitutions, for Ostrom and for Buchanan, were artifacts, institutions consciously designed by people to achieve certain goals. Such efforts at constitutional design were informed by ideas about how different rules were likely to encourage and enable people to act in ways that led to particular outcomes (Ostrom [1976] 2012: 13–20, 1980; 1982: 151–58, 162–65, 1999: 123–26). On this view, when people such as the American founders engaged in constitutional design, they were “using the logic of theoretical reasoning as a fundamental tool that enables human beings to reason and thus to inform choice. Theory is used both to engage in positive analysis to clarify the implications that follow from particular structural characteristics and to clarify the normative criteria used in the design of a constitution” (Ostrom 1982: 155; also see pp. 153, 162–63 and Ostrom [1973] 2008: 48–49, 56–60).

Ostrom opposed what he saw as the dominant approach taken by American political science since the late nineteenth century, which rejected both the theories that had informed the design of the US constitution and also the idea of federalism (in favour of a unitary state with a single centre of power and authority). Especially influential in this regard was the work of Woodrow Wilson ([1885] 1956, 1887), who argued that by the late nineteenth century the constitution designed by the federalists had become no more than a façade concealing the underlying reality of American government wherein there was a single centre of ultimate power, Congress, and that the fragmentation of authority emphasised by the federalists would only dilute accountability and thereby encourage irresponsibility in the conduct of government (Ostrom [1971] 2008: 181–87, 212–13, [1973] 2008: 20–31, 65–72, 87–92, [1976] 2012: 8–11, 1982: 151–52, 158–62, 1999: 127–28).

Ostrom’s attempt to recover Hamilton’s and Madison’s political theory made him “critically aware that arguments advanced by Woodrow Wilson represented a paradigmatic break from the formulations that had been used in conceptualising the design of the compound republic” ([1997] 2012: 88; also see Ostrom [1973] 2008: 67–68). Ostrom saw Buchanan and Tullock as also challenging this line of reasoning through their efforts to show that rational choice could provide a logical basis for a constitution that would sustain self-government.

Ostrom also drew on Buchanan and Tullock’s work in a second, closely related way, namely to dispute the analysis of the working of the state upon which Wilson relied. This led to Ostrom’s attempt, from the early 1970s onwards, to reconsider the intellectual foundations of the study of public administration.

2.2 Political economy perspective on public administration

From the very beginning of his academic career, Buchanan had criticised the idea that it was appropriate to view the state as a single, unified entity that consciously pursued the public interest. Buchanan argued there was literally no such thing as the public interest or an objective truth that could guide political action; only individual men and women could conceive of their own ends and therefore have an interest that could be pursued (Buchanan [1949] 1999, 1954, 1975). In stark contrast to such ‘organismic’ or ‘truth-judgement’ views of politics, as he termed them, Buchanan argued the state should be viewed as composed of individual people who, in virtue of the rules governing their (inter)actions, stand in certain relations to one another and face incentives to act in particular ways, thereby creating a tendency for certain kinds of outcome to emerge. Viewed from this individualistic standpoint, policy-making is the (not always intended) product of complex interactions between individuals and groups, not the deliberately chosen result of decisions made by a benevolent (or otherwise), omniscient policy-maker sitting atop a unitary government structure (Buchanan and Tullock [1962] 1999: 11–13; Buchanan [1986b] 1999: 20–21, [1986c] 1999: 456).

In a similar vein, starting with his early work on water supply and continuing thereafter, Ostrom analysed government activity, not as if it were conducted by a unified, hierarchically-ordered bureaucratic state, but rather as the (sometimes unintended) outcome of actions taken by people occupying positions in a variegated network of organisations whose rule-governed (inter)actions generated particular kinds of incentives, information, and outcomes (Ostrom et al., [1961] 1999, Ostrom, 1991: 223–26, 1999: 125, 127–28; also see Ostrom, 1977: 1511–12).

Ostrom took inspiration from Buchanan as he sought from the early 1970s to develop a new perspective, informed by public choice theory, on public administration. Explicitly invoking the work of Thomas Kuhn, Ostrom argued that the discipline of public administration was undergoing an intellectual crisis whose resolution required a shift away from portrayals of the institutions and activities of policy-making as centred on a hierarchy composed of professionally-trained public servants and technical experts and possessing a single centre of authority (Wilson 1956; Gerth & Mills, 1946) (Ostrom [1973] 2008: 5–41, 87–115). The failure of the dominant ‘policy science’ model of American public administration had been a concern of some contemporary political scientists for more than a decade—Charles Lindblom’s classic article ‘The Science of “Muddling Through”’ (1959) similarly noted the mismatch between professional claims of technocratic expertise amongst policy-makers and the reality of a federal government that was unable to ameliorate social problems despite the expenditure of huge sums of public money.Footnote 4 The shortcomings of the prevailing Wilsonian paradigm could be resolved, Ostrom argued, through a paradigm shift involving the adoption of a ‘political economy’ approach, inspired by the work of scholars like Buchanan and Tullock who used the tools of economic theory to analyse the public sector (Ostrom, 1964; Ostrom and Ostrom, 1971; Ostrom [1971] 2008: 219–22, [1973] 2008: 42–64, 1999: 126, 129, 2009a: 62–63).Footnote 5

These two strands of research, on interpreting texts from a constitutional perspective and the political economy of public administration, both influenced by Buchanan, were closely related. For Ostrom, the political system devised by the American Founders promised to create the conditions required for the emergence of a system of ‘democratic administration’ in which the public realm was governed through institutions that fragmented authority and divided sovereignty, whose jurisdictions overlapped, and where the exercise of coercion by one arm of government was checked by the countervailing power of the other arms and by an active and engaged citizenry (Ostrom [1973] 2008: 18, 65–86, 96–98, 111–15, 146–47 197 n. 5, [1971] 2008: 140–41, 1977: 1508–10 1991: 227, and 1997: 84, 254).Footnote 6

The challenge that then arose, and was a particular focus of Buchanan’s work from the 1970s onwards, was how a liberal constitution in the Federalist tradition could arise from a political system that incentivised rent seeking and bureaucratic expansion so that political actors were focused on the pursuit of pecuniary advantage via politics rather than on the design of rules that limited the use of the political process for exploitative purposes (Buchanan 1975, 1987, 1998; Buchanan & Vanberg, 1989).

3 Buchanan and the constitutional attitude: the importance of Ostrom’s notion of ‘Artifactual Man’.

By the late 1970s, the public choice movement to which Buchanan’s work had given such impetus was becoming increasingly well-established, with the creation of research centres and a dedicated journal (Medema, 2000). Nonetheless, Buchanan continued to be frustrated by the continued reluctance of many economists to embrace constitutional political economy:

I have continued to be surprised at the reluctance of my colleagues in the social sciences, and especially in economics, to share the contractarian-constitutionalist research program ... A substantial share of my work over the decade, 1975–85, involved varying attempts to persuade my peers to adopt the constitutional attitude. ([1986a] 1999: 24.)

By a ‘constitutional attitude’, or ‘constitutional mentality’, Buchanan meant the idea that “we create the institutions within which we interact, that we construct the rules that define the game that we all must play.” In his view, the rules of the game had far-reaching consequences, determining, for example, the level of economic development and the distribution of income and wealth ([1979a] 1999: 255, 1989b, 2003). In a paper published in 1979, Buchanan ([1979a] 1999: 255) “sought to understand why the ‘constitutional attitude’ seems so foreign to so many of my fellow economists, to understand why this central aspect of what was a part of the conventional wisdom of our Founding Fathers now seems so elusive”. In attempting to “explain satisfactorily to myself just why attitudes that seemed so natural to me seem so difficult for others”, Buchanan invoked Vincent Ostrom’s notion of ‘artifactual man’.

The term ‘artifactual’ refers to man’s capacity for “becoming different from what he is”. An artifactual person possesses innate characteristics, but also the capacity for self-authorship leading to personal transformation:

I have used the term artifactual here precisely for the purpose of allowing some recognition of the basic constraints of human nature while, at the same time, allowing for wide areas of choice within these constraints, areas within which we can, and do, construct ourselves as individuals, from the base constructed for us by our forebears. (Buchanan [1979a] 1999: 247, 252).

For Buchanan, it was integral to human nature that people were able, within limits, to forge their own character and identity. A repentant smoker “can surely imagine himself or herself freed of the habit, with a transformed set of preferences that would not include any desire to smoke.” If in order to realise that new identity he imposes upon himself a rule that forces himself to quit smoking, then over time “he will find that he does become different from the person he was. His preferences shift; he becomes the non-smoker that he had imagined himself capable of becoming” ([1979a] 1999: 253). On this view, people can gain a measure of critical distance from their preferences and impose rules upon themselves in order to modify their preferences so that they become the kind of person they would like to be. Artifactual man understood the importance of rules in changing outcomes and was willing to engage in (what was effectively) constitutional redesign to change rules and thereby change outcomes.

Buchanan’s account of artifactual man drew explicitly on Ostrom’s work on constitutional construction:

My usage of the word artifactual is borrowed directly from Vincent Ostrom, who has repeatedly emphasised the necessity of considering the political constitution as an artifact, to be categorically distinguished from an evolved legal order. The American experience, perhaps unique in history, has embodied the attitude that we create the institutions within which we interact, one with another, that we construct the rules that define the games we must play. But we can never lose sight of the elementary fact that the selection of the rules, the ‘constitutional choice’, is of a different attitudinal dimension from the selection of strategies within defined rules. (Buchanan [1979a] 1999: 255.)

Buchanan referred to Ostrom’s 1976 essay ‘David Hume as a Political Theorist’ ([1976] 2012), where Ostrom described constitutions as artifacts consciously designed by people with a view to achieving certain goals, in this case “a system of rule where authority is distributed so that every element in the constitution is not only assigned authority to exercise certain powers in the government of a society but each element must reciprocally have the means to resist the usurpations of authority by other elements in the constitution and to procure adequate remedies for defending its resistance” ([1976] 2012: 19).Footnote 7

The problem, Buchanan argued, was that creative behaviour of the kind undertaken by artifactual man cannot be understood using the model of rational choice theory upon which most economists rely. The latter presupposes that people’s preferences, and associated utility functions, are stable, with rational conduct involving the efficient choice of means to satisfy those given ends. This excludes changes of the kind sought by artifactual man, who (as we have seen) seeks to impose rules upon himself in order to cultivate new preferences and thereby forge a new identity. As Buchanan put it, “modern economic theory forces upon us patterns of thought that make elementary recognition of the whole ‘becoming’ part of our behaviour very difficult to analyse and easy to neglect” ([1979a] 1999: 247).

Herein lies Buchanan’s explanation of why his fellow economists struggled to appreciate the constitutional attitude. Most economists rely on a model of human action as a teleological endeavour centring on the satisfaction of given preferences that excludes the kind of creative, non-teleological conduct undertaken by artifactual man:

I was led to ask, however, whether persons who do not and cannot conceive themselves to be artifactual (even if, in fact, they are and must be), can easily conceive of artifactual social institutions, artifactual rules of the game, to be chosen apart from the simple selection of strategies to be played in the complex interaction process defined by the rules of the order. Does the manner in which men model their own behaviour affect, and perhaps profoundly, the way that they model the social institutions under which they live? If individuals conceive themselves in the teleological image of modern economics, can they shift gears to a nonteleological image of a community? (Buchanan [1979a] 1999: 255–56.)

For Buchanan, therefore, constitutional choice could not be reduced to routine utility maximisation, so the constitutional attitude was less readily intelligible to those whose analytical vision was confined to standard choice theory. If, however, man was conceived as an artifactual being who constructs himself through his own imaginative, creative choices, then “it becomes relatively easy for him to envisage changing the basic rules of social order in the direction of imagined good societies” and constitutional construction becomes intelligible (Buchanan [1979a] 1999: 258).

Buchanan (1975, 1987, 1995) did not think that any one individual autonomously chose the rules that governed his or her collective life. Individuals did not select a constitution from the shelf in the way that customers chose tinned goods. Rather, institutions emerged from the interactions of numerous individuals seeking to reconcile their diverse ends and in this sense constitution-making was a process of exchange rather than choice. But in this process of exchange individuals creatively sought mutually advantageous improvements in their positions that led to the emergence of new and perhaps unexpected institutional outcomes (Marciano, 2009; Meadowcroft, 2023).

Vincent Ostrom similarly emphasised the importance of the creative human agency of many individuals in constitutional design. At the start of The Intellectual Crisis of American Public Administration, Ostrom argued that structural conditions mattered, because the rules and organisational arrangements people used to govern themselves shaped the outcomes that arose. Social scientists should therefore “be able to specify the consequences that will follow from different organisational conditions. To assert that consequences follow from conditions is to say that effects have their causes. We should be able to indicate the conditions and consequences that derive from the choice of alternative organisational arrangements if theories of organisation have scientific warrantability” ([1973] 2008: 2; also see p. 159). However, like Buchanan, Ostrom was adamant that, because the causal mechanisms constituted by such rules were animated by, and depended for their continued existence upon, people’s actions, and because people had the capacity to (re)conceptualise and respond creatively to their circumstances, structural conditions did not determine people’s actions in a strict causal sense:

We must, however, distinguish between a determinate causal ordering and a quasi-causal ordering. In a determinate causal ordering a cause impinges directly upon and determines an effect. A quasi-causal ordering depends upon the intervention of human actors who are capable of thinking, considering alternatives, choosing, and then acting. The one is determined, the other is constituted. In such circumstances we are required to take account of how individuals view themselves, conceptualise their situation, and choose strategies in light of the opportunities available to them. Analysis in the social sciences requires recourse to strategic thinking in quasi-causal orders. The rule-ordered relationships that are constitutive of human organisation function as soft constraints that are themselves subject to choice. (V. Ostrom, 2008: 2.)

Like Buchanan, therefore, Ostrom emphasised the importance of creative human agency as well as social structure.

As the passage just quoted makes clear, Ostrom—like Buchanan—was a subjectivist who maintained that in order to explain why people act as they do, it is important to understand the ideas that help them interpret and understand their circumstances (Ostrom 1997: 89, 92, 98, 105–7, 112–14). “Ideas are always the basis for action,” Ostrom ([1973] 2008: 172–73; also see Ostrom, 2008: xxvi) wrote, so that “Patterns of order in human societies depend … on shared communities of understanding about how human beings relate to one another in ordering the[ir] ways of life.” Ideas matter, therefore, making a difference to how people choose to govern themselves and, as Ostrom noted, people “are never confined to only one way in formulating patterns of order” (p. 173). They may, for example, rely on conceptions that encourage them to construct monocentric, top–down systems characterised by command and control under a single centre of ultimate power; or they may, if their efforts to engage in constitutional design are informed by a different set of ideas, strive to develop a polycentric system of federal governance with no commanding centre (Ostrom [1979] 2014: 227–35, 1997: 100–01, 112–14, 180). In other words, the possibility of self-governance requires that citizens themselves believe that they can, and indeed should, play a role in co-producing the system of rules under which they live. That is, they need what Buchanan referred to as the constitutional attitude that we, the people, create the rules that govern our daily life ([1979a] 1999: 255, 1989b, 2003).Footnote 8

Buchanan and Ostrom came, therefore, to agree that doing justice to the constitutional attitude required understanding people as artifactual beings with the capacity to create the rules that govern their conduct. But how precisely should such decisions be conceptualised? What are their logical foundations? It was in answering these questions that Buchanan’s and Ostrom’s views diverged.

4 Diverging views on the logical foundations of constitutional democracy: Ostrom’s critique of Buchanan and the role of the framework

In several papers written from 1997 onwards, Ostrom returned to Buchanan’s work, considering its significance and also engaging in a critique of Buchanan’s commitment to standard models of rational decision-making as the logical foundation for constitutional choice (Ostrom [1997] 2012, 1999, 2012). Near the outset of his critique, Ostrom wrote that Buchanan “explicitly defends the use of the model associated with Homo economicus, vigorously emphasising the importance of using a single model applicable to different types of choice. This reduction is an issue requiring serious reflection” ([1997] 2012: 92; also see Ostrom, 2012: 428). While Buchanan’s exchange approach might be satisfactory for understanding market transactions, it “stands apart from important aspects of human experience”, including those involved in choosing rules to govern our interactions:

The criteria for making choices are likely to vary among different types of choice. A pecuniary choice of goods offered at a price in exchange for money is different than a choice of rules about how to order relationships in the conduct of human affairs. ... A choice of rules is ... not a monetised exchange relationship. ([1997] 2012: 93, 95; also see p. 97 and Ostrom 1997: 90, 2009a: 63–66, 2012: 429.)

Such constitutional choices, which “turn[.] on how ideas expressed through the language of rules are used to constitute patterns of human relationships” (1997: 26), involve motivations that are irreducible to individual preference satisfaction and utility maximisation:

Maximising utility … does not apply to epistemic choice or constitutional choice in the same way that such calculations might be thought to apply to the choice of substitutable alternatives in one-to-one comparisons of distinguishable but similar items. To rely on a single specifiable criterion of choice, such as Utility, is to treat human societies as one-dimensional realms in which the forest cannot be seen for all the trees that obstruct one’s view. (Vincent Ostrom 1997: 279; also see pp. 98, 102, 293).

More specifically, the process through which groups of people agree to abide by certain rules involves them engaging in what Ostrom refers to as “covenantal reasoning”. This involves people recognising an obligation to consider the interests and goals of others in deciding how to act that transcends the self-interested preference-satisfaction and utility-maximisation found in rational choice theory of the kind to which Buchanan subscribed (Ostrom 1997: 188; also see Ostrom, 1991: 62–66, 252–53, 1997: 12–18, 93–96, 100–02, 184, 279–81, 292–95; [1997] 2012: 94):

The emphasis on maximising ‘utility’ … means that primary attention is being given to preference orderings; other aspects of the political economy of life are excluded from the focal attention of inquiry and swept into the background … [T]he place of a moral order as constitutive of fiduciary relationships, the place of law and the requirements of justice, and the requirements of intelligibility in human artisanship are treated as outside the focus of inquiry … If attention is given only to preferences, there is a danger that ‘the whole moral and intellectual condition of a people’ will be reduced to ‘intellectual dust’, as Tocqueville asserted. (Ostrom 1997: 99).

On this view, the reduction of constitutional decision-making to considerations of preference-satisfaction and utility-maximisation implies that considerations, pertaining to the covenantal nature of the choice of rules, that are central to constitutional decision-making are excluded. Hence Ostrom remarked that while “[t]he cost calculus introduced by Buchanan and Tullock (1962) … goes some distance in taking account of factors that would enter a method of normative inquiry appropriate to the formulation of rules … it does not provide the basis for formulating a fair set of rules as such” (1997: 101; emphasis added).Footnote 9

Ostrom elaborated on the idea that there may be different aspects of human conduct, not all of which are reducible to models of people as utility-maximisers, by distinguishing between three different kinds of concepts: frameworks; theories; and models (Ostrom 1997: 104–05, 1999: 133; Ostrom and Ostrom [2004] 2014: 66–67). The most abstract was the framework, which sets out the key features of the social world that must be specified in any form of analysis. For Ostrom, these key features were twofold: the people, who qua framework are conceived simply as purposeful actors (without a more specific account of their conduct being given); and the ‘action situation’, or account of the rule-governed social space within which those people interact. In developing a theory, social scientists select particular elements from the framework, make more specific assumptions about them, and examine how they inter-relate to form a causal mechanism that, when animated by human actors, generates particular kinds of outcome. The development of a model involves making still more specific assumptions about the elements so as to make it possible to deduce specific predictions about how people will behave and what outcomes will ensue (Ostrom 1997: 105).

Ostrom’s argument was that while the model of people as self-interested utility-maximisers is suitable for analysing some kinds of decision, such as those made in competitive markets, it is less appropriate for analysing constitutional decision-making, for which a different account of choice involving covenantal reasoning is appropriate:.

The model of man used in neoclassical economic theory does not exhaust the efforts to conceptualise how ‘representative individuals’ might relate to the various possibilities for human choice and the conduct of human artisanship. The paradigmatic significance of The Calculus of Consent is by no means confined to economic reasoning narrowly construed. (Ostrom [1997] 2012: 91; also see Ostrom, 1999: 133.)

In Ostrom’s view, what was most valuable about Buchanan’s work—its “paradigmatic significance”—was his theoretical emphasis on the need to consider rules as the subject of individual decisions, not the specific model of self-interested utility-maximisation and preference satisfaction used in The Calculus of Consent. Covenantal reasoning and the instrumental rationality characteristic of standard choice theory are different theories of rational decision-making that form part of the same over-arching framework but are nevertheless suitable for understanding different domains of social life. Ostrom therefore rejects the idea that a single model of rational choice (as utility-maximisation/preference satisfaction) can be used to analyse both market behaviour and political exchange (i.e., he rejects Buchanan’s presumption of behavioural symmetry) (cf. Allen, 2012: 82–83, Bish 2014: 233, 235; Gaus, 2018: 139; Lewis, 2021: 624–26).Footnote 10

Ostrom observed that Buchanan sometimes acknowledged the need to transcend simple models of rational conduct as preference satisfaction and utility-maximisation if constitutional politics was to be properly understood, but did not do so consistently:

Buchanan ... is frequently emphatic about the fundamental importance of relying on a rigorous model of Homo economicus in the analysis of rules and their place in rule-ordered relationships. Yet he is distinguished for having advanced the thesis that human beings have reflective capabilities that enter into the development of their personal character structure. In other words, character structure is an artifactual creation (see Buchanan ... ‘Natural and Artifactual Man’). (Ostrom, 1999: 132.)

Ostrom welcomed Buchanan’s recognition in ‘Natural and Artifactual Man’ of the artifactual nature of human beings (Ostrom 1997: 89, [1997] 2012: 93–94, 97, 1999: 132). However, he also observed that Buchanan was still “subject to some ambiguity about his ‘model of man’” and that “Buchanan’s openness to meta levels of analysis is not well served by his close adherence to the language of neoclassical economics theory” ([1997] 2012: 92, 97).Footnote 11

Ostrom was arguably right to be concerned because, notwithstanding Buchanan’s criticism of orthodox choice theory as being unsuitable for analysing creative decision-making of the kind undertaken by artifactual man, Buchanan resorted to it once again in his subsequent analysis of how people decide what preferences, and what character, to cultivate. Buchanan (1994) was insistent that moral behaviour could be placed within a rational choice framework—that it was in people’s rational self-interest to behave morally, by, for example, adopting and internalising a powerful work ethic, even when such behaviour ostensibly appeared to impose costs on the individual.

In a constitutional context, Buchanan assumed that people have a set of stable, well-defined meta-preferences—that is, (higher-order) preferences over their (everyday, lower-order) preferences—that guide and inform such decisions (Brennan and Buchanan [1985] 2000: 80–81; Buchanan [1989c] 1999: 447). Buchanan (1975: chapter 2, 1998) assumed, for example, that individuals would have a consistent meta-preference against rules that allowed unconstrained majority rule; individuals would always fear unconstrained majority rule, even if they did not know whether a winning coalition existed, who would be its members if it did exist, and how those members might exercise their power. Far from departing from the notion of choice of utility-maximisation, therefore, Buchanan continued to rely on the notion of utility-maximising choice in his analysis of personal constitutional preferences.Footnote 12

Indeed, discussing the American foundation from a public choice perspective, Buchanan and Viktor Vanberg (1989) explicitly located the motives of the founders within a rational choice framework, positing that their seemingly public-spirited investment in securing the public good of the constitution was in fact an attempt to obtain the personal benefits of influence and esteem from participation in elite-level national politics. It was also rational for others to defer to their wisdom and knowledge, rather than making similar personal investments to become well-informed, thereby confirming the special role of figures like Madison and Hamilton in constitutional construction. According to Buchanan and Vanberg, the whole American constitutional enterprise fit neatly within a standard rational choice framework.

But such an approach arguably falls foul of Buchanan’s own critique of the standard model of rational choice on the grounds that, because it conceptualises people as seeking to satisfy given (meta-)preferences, it is overly-deterministic and under-estimates the capacity of individuals to creatively envisage new rules of social organisation.Footnote 13 Moreover, such an approach, resting as it does on a notion of rational conduct as preference-satisfaction and utility-maximisation, is at odds with Ostrom’s view that an exclusive focus on those aspects of human conduct fails to do justice to the convenental aspects of constitutional agreement. It is unclear, therefore, that—as it is deployed by Buchanan—the notion of meta-preferences provides the conceptual resources needed to develop his account of artifactual man in a way that does justice to Ostrom’s concerns about covenantal reasoning, human creativity and sources of motivation beyond self-interested preference satisfaction and utility-maximisation.Footnote 14

5 Ostrom and Buchanan on the conditions of self-governance: capabilities, a rich associative life and moral order

Given Buchanan’s and Ostrom’s scepticism about deductive models of constitutional decision-making, the question arises of how they analysed the behaviour of artifactual man and the conditions that made possible self-governance. Ostrom’s answer involved tan attempt to develop a theory of the capabilities required for constitutional decision-making (where the term ‘theory’ is used in the sense, defined above, of an account of the causal mechanism that makes possible a certain kind of behaviour) (Ostrom 1997: 104–05, 1999: 133). Rather than developing formal models that make it possible to deduce that people will take specific actions, Ostrom focused instead on a different kind of theoretical or explanatory endeavour by seeking to identify the underlying conditions that make possible self-governance. This involved an account both of the attributes decision-makers need to engage in self-governance (in particular, their capabilities) and also of the social-structural conditions that foster the development of these capabilities (1997: 30, 273, 290–91, [1999] 2012: 394).Footnote 15

For Ostrom (1997: 276–77), if people are to engage in constitutional construction they require the ability to practice “the art of association” (i.e., to create, deploy effectively, and then maintain the rules that facilitate self-governance and mutually productive relationships). There are at least two aspects to this. First, as we have seen, people require the intellectual resources—the ideas and the imagination—to conceptualise the possibility that governance can indeed be a matter of reflection and choice rather than force; they need to “think of themselves as citizens working with other citizens to build enduring patterns of association in which the community of persons involved achieves self-governing capabilities … Democratic societies cannot achieve long-term viability if democratic processes are viewed only as a struggle to win and gain dominance over others” ([1973] 2008: 168; also see pp. 163–72 and Ostrom [1971] 2008: 151, 1997: 98, 2009b: 51–52).Footnote 16

However, ideas alone are not enough for people to develop an effective constitutional attitude. Secondly, people also need the capabilities required to put those ideas into practice (Ostrom 1997: 284; Ostrom & Allen [1994] 2012: 463–64, 476). “A player who invents a new game,” Ostrom ([1973] 2008: 198; also see pp. 168–69) states, “and expects that game to be successful must make certain assumptions about the capabilities of potential players”:

[D]emocratic societies cannot be maintained without the knowledge, moral integrity, skill and intelligibility of citizens ... Self-governing democratic societies exist only under conditions in which individuals become their own masters and become capable of governing their own affairs and working with others in mutually productive relationships. (Ostrom 1997: 3, 288.)

The requisite conditions of the possibility of self-governance are several: people need the courage to assume responsibility for managing their interactions; they need to be skilled in discovering, articulating and striking compromises over their (diverse) interests as well as in constructing institutions that appropriately reflect them; they need to be able to identify, agree upon, and resolve conflicts over the requisite rules; they must be skilled at using language to express those rules; and, once the rules have been developed, people need the capacity to use them to hold public officials and politicians to account. A citizenry in possession of such capabilities is one (agentic) condition for the possibility of self-governance (Ostrom [1971] 2008: 139–42, 146–47, [1973] 2008: 148–50, 153–55, 173, 1997: 3, 114, 271–73, 276–77, 286–91, 294–303; 2009b: 51–52; Ostrom & Allen [1994] 2012: 459–60, 490–91).Footnote 17

Ostrom went on to argue that direct, practical experience with self-governance in civil society can help people to learn the art of association and thereby develop the capabilities for full democratic self-governance. Ostrom quoted Tocqueville’s Democracy in America on the importance of institutions that help people to develop the capabilities needed to work together to overcome collective challenges: ‘“Municipal institutions constitute the strength of free nations. Town meetings are to liberty what primary schools are to science: they bring it within the people’s reach, they teach men how to use and enjoy it’” ([1973] 2008: 85; also see Ostrom [1971] 2008: 140–44, [1999] 2011: 342, [1999] 2012: 401 and 2009b: 44). It is through the experience of being involved in running various institutions in civil society—that is, through being involved in the civic activities central to what Ostrom referred to as res publica, the open public realm—that citizens-artisans acquire both the habit of looking to each other, rather than to the state, to solve their problems, and also the civic knowledge and practical skills needed to bring their shared endeavours to a successful conclusion (Ostrom [1971] 2008: 140–41, 1991: 199–221, 1997: 273, 290–91, [1999] 2011: 342, 2009b: 48–49; Ostrom and Ostrom [2004] 2012: 86).Footnote 18

Ostrom’s point was that formal constitutional structures alone are necessary, but not sufficient, for a functioning polycentric system. The latter also requires those structures to be aminated by a citizenry with the requisite capabilities: “Federal systems of government can only be expected to work when those who use instrumentalities of government know how to make proper use of them” (Ostrom, 1987b: 164; also see Ostrom [1982] 1999: 158 and Fotos, 2015: 76–77). If people lack the requisite ideas, and so tend to look to the state as the only solution to their problems, or if they lack—or lose—the capabilities needed to make a polycentric system work effectively, perhaps because they fail to exercise and thereby sustain them, then the scope for engaging in successful self-governance will decline: “If citizens in a democratic society do not have a conscious awareness of the essential importance of the logical conditions of constitutional democracy and do not develop a civic culture for relating responsibly to one another and to those who act as agents on their behalf, democratic societies are placed at risk” (Ostrom, 1999: 131; also see Ostrom, 1987b: 209–10, [1999] 2011: 344–47).Footnote 19

Buchanan (1975, 1987, 1997) was also concerned with the relationship between individual attributes and constitutional outcomes. For Buchanan, the outcome of political exchange was determined by the capacities of individual participants, broadly defined in terms of their material resources, skills and values. Buchanan thought that when individuals possessed very unequal capacities, those with greater resources, superior skills and willingness to exploit others would use their advantage to self-interestedly frame constitutional rules to oppress those weaker individuals. For instance, the constitutional order created at the American foundation reflected the particular circumstances of revolutionary America in which those engaged in constitutional design were relatively equal and homogeneous because the most vulnerable members of society were excluded from constitutional decision-making (and were indeed oppressed and exploited) (Congleton, 2014; Meadowcroft, 2020, 2023; Munger, 2020). But whilst there are some hints in Buchanan’s writings that he perceives the importance of the kinds of capabilities Ostrom identifies as prerequisites for constitutional decision-making, as for example when he states that people need to be educated so as to cultivate their powers of imagination and so that they acquire the capacity to evaluate outcomes ([1979] 1999: 254–55), those hints are arguably under-developed in comparison to what is seen in Ostrom’s work (Lewis & Dold, 2020: 1172–73).

Similarly, Buchanan ([1981] 2001: 187) did argue that a constitutional order of self-governing individuals depended on the existence of “‘the social capital’ that provides the basic framework for our culture, our economy, and our polity”. It was within a framework of mutual respect that “the ‘free society’ in the classically liberal ideal perhaps came closest to realization in our history”. This moral order of social capital did not describe moral consensus or shared deep values, but rather a thin moral order of reciprocal toleration that enabled every individual to pursue their self-conceived ends. Shared moral values might serve an important purpose in uniting a small group as a moral community, but a polity predicated on moral consensus would be counterproductive for producing a constitutional agreement that divided power and imposed institutional checks and balances. Accordingly, Buchanan saw a liberal constitutional order as arising from a social context of weak ties, not from a moral community of strong bonds. This marks a fundamental point of divergence between Buchanan and Ostrom—a divergence that became increasingly apparent as their work developed. For, as we have seen, Ostrom came to believe that constitutional choices were covenental in nature and therefore involved motivations that were irreducible to individual preference satisfaction and utility maximisation. Buchanan was resolutely sceptical of such an enterprise and instead wanted to show that even absent equality, homogeneity and civility, it was logical for self-interested, utility-maximisers to construct a constitutional order.

This tension between Buchanan and Ostrom was also a tension within Buchanan’s work. Buchanan ([1981] 2001) acknowledged that a moral order of civility was a prerequisite for a liberal constitution, but his account of that moral order was almost certainly under-developed in its analysis of the requirements for people to develop an effective constitutional attitude (Dold & Petersen, 2021; Lewis & Dold, 2020: 1172–73). Comparison of Buchanan’s and Ostrom’s work brings to the fore this tension within Buchanan’s constitutional political economy.

6 Conclusion

This article has traced the development of James Buchanan and Vincent Ostrom’s views on the logical foundations of constitutional democracy. As we have seen, Buchanan and Tullock’s (1962) classic analysis of how constitutional structures are the product of the decisions of rational economic agents influenced Ostrom’s reading of key texts in political theory and shaped his analysis of public administration. The direction of influence was not, however, entirely one-way; Buchanan subsequently drew on Ostrom’s account of people as artifactual beings in his attempt, in ‘Natural and Artifactual Man’, to understand the ‘constitutional attitude’ (i.e., the idea that people create the rules governing their interactions).

Ostrom and Buchanan therefore shared a good deal of common ground; but they also had differences of opinion. As Ostrom (2012: 427) observed in one of his final essays, whilst he recognised the “paradigmatic significance” of Buchanan’s work, in particular The Calculus of Consent, “[t]hat does not mean that I agree with everything that Buchanan or Tullock have to say any more than I expect them to agree with … what I may have to say in this assessment. There is substantial room for continuing discourse and inquiry.” Ostrom became increasingly critical of Buchanan’s continued emphasis on models of rational choice, arguing that a satisfactory understanding of how people make constitutional decisions required theorists to move beyond assumptions of self-interested preference satisfaction and utility maximisation.

Instead of focusing his analytical attention on models of utility maximisation designed to make it possible to deduce people’s decisions, Ostrom sought to develop a theoretical account of conditions that would make such constitutional choices possible in the first place. As Ostrom wrote, “I’ve dedicated a large part of my work to understanding the conditions affecting the way human beings relate to one another and those that generate a functioning social order” (quoted in Aligica & Boettke, 2009: 146). These conditions are, as we have seen, twofold: people need the capabilities—the powers of imagination, the capacity to use language to express rules, and the skills at negotiation needed to settle disputes—required to play the role of citizen-sovereigns and engage in constitutional decision-making; and they also need a rich associative life, involvement in which will enable them to acquire those capabilities.Footnote 20 In focusing on the pre-requisites for constitutional choice, Ostrom complements Buchanan’s account of moral order which, as noted above, was almost certainly under-developed in its analysis of the capabilities required for people to develop an effective constitutional attitude (Dold and Peterson 2021; Lewis & Dold, 2020: 1172–73). As Richard Wagner (2015: 25) has written, “Ostrom … strongly advances a program of Virginia political economy through the centrality he assigns to a science of civic association, recurring to Tocqueville in doing that.” In this respect, Buchanan and Ostrom turned out to have complementary talents, interests, and emphases, with Ostrom’s focus on the conditions of the possibility of self-governance building on and complementing Buchanan’s emphasis on politics-as-exchange and constitutional agreement.