Liberty Matters

James Buchanan and the Economics of Anarchy

By: Geoffrey Brennan

   

Related links:


It is one of the features of an intellectual’s work that it has a life independent of – and possibly more extensive than -- its creator. In that sense, Jim Buchanan’s death (“after a short illness”) on January 9 is of no particular academic significance – beyond the fact that Jim himself is no longer around to correct misinterpretations of the Buchanan position (as he saw it).[1] On the other hand, it would be unseemly for the occasion to go unmarked. At the very least, Buchanan’s death provides an opportunity to restate and re-assess the “Buchanan position” (at least as I see it); and such restatement and reassessment is my purpose here.
Most economists are of the “have-brain-will-travel” kind. Armed with our kitbag of techniques and our distinctive disciplinary perspective, we look around for varied applications (often finding them in unlikely places). For most of us, our lifetime “contribution” is just the sum of our little inroads (if any at all) into a range of little problems. There is no particular dividing line in the epistemic division of labor that marks off our individualized “contributions”; and the lifetime whole is not much more than the sum of the parts.
In that sense, Buchanan was distinctive -- not just in the size of the contribution, but also in its coherence. There is, associated with his work, an identifiable “research program” – not so much ”Public Choice theory” as such, but what is sometimes called “Virginia School political economy,” or what I prefer to call “constitutional contractarianism,” because the latter term is more descriptive. Of course, that program evolved and became more self-conscious as the work accumulated;[2] but it is uncanny how much of it can be identified in Buchanan’s writings right from the start.[3]
In what follows I will deal first with Buchanan’s contractarianism and then turn to the constitutionalism, trying to indicate how these two c’s are related. I will then try to connect the “constitutional contractarian” project to Buchanan’s credentials as a classical liberal and raise a number of other queries about the intellectual scheme that seem to me worth discussing.
Contractarianism
For Buchanan, economics properly pursued is the “science” (with much emphasis on the “logic”) of exchange.[4] In other words, “exchange” is the lens through which economists should look at the social world, identifying both actual and potential instances of it and its consequences, where present. In this he was a faithful (if originally unconscious) disciple of Richard Whately.[5] Jim used to say that the proper point of departure for an economics principles course was Menger’s account of two horse-traders [auf deutsch], and more generally that the measure of the quality of any principles course was how soon it got to “exchange.” For him, exchange is central.
Of course, he saw markets as the primary institution for the mobilization of exchange possibilities. But his version of the “economic analysis of the study of political processes” (or “Public Choice theory,” as it came to be called) was distinctive in Public Choice circles for its emphasis on “politics as exchange” – a phrase he repeated many times.[6] (As an aside, we should note that for many libertarians/classical liberals it might be more natural to think in terms of politics as coercion – and though Buchanan certainly did not deny the possibility of coercion in politics, he equally certainly denied that all government action is intrinsically coercive.) He believed that exchange via political action is possible and that the role of the economist is to uncover those exchange possibilities. In this, he followed Knut Wicksell (unquestionably one of Buchanan’s intellectual heroes) and Wicksell’s other interpreter, Erik Lindahl – and in this sense was a participant along with Musgrave and Samuelson in the analysis of public goods and associated market failure that dominated the welfare economics/public economics of the 1950s and early 1960s.
That is, Buchanan fully accepted the public-goods argument that markets sometimes fail to exploit all the mutual benefits that are on offer in human society. And he also accepted the (Wicksellian) proposition that collective action could in principle appropriate such “gains from exchange” in public-goods supply. In other words, he thought that there is (in principle) a role for the “productive state” as well as the “protective state” – to use a distinction he developed explicitly in the Limits of Liberty (1975). He explicitly rejected anarchy as a viable option, broadly embracing the Hobbesian argument for government and in the process adding his own kind of ‘rational reconstruction’ of the Hobbesian position.
Indeed, much of Buchanan’s mid-career work involved exploring the details of market-failure argument. Two of his most famous (certainly his most widely cited) papers were his analysis of price-excludable collective goods – his “theory of clubs” (1965) – and his “externality” piece with Craig Stubblebine (1962). And in 1968 he offered a book-length treatment, The Demand and Supply of Public Goods, of the public-goods argument, explicitly trying to connect his own exchange variant to the standard neoclassical version.
Equally, however, Buchanan recognized – again from the earliest point in his career – that the mere fact of market failure to appropriate gains from exchange does not show that government can succeed. The message that Buchanan took from the public-goods literature is rather that the appropriation of possible gains from exchange represents an institutional problematic. Again following Wicksell, he explicitly and forcefully rejected the “benevolent despot” model of government that he saw as dominating standard public economics.[7] Indeed, critique of this “benevolent despot” model became the focus for the entire Public Choice movement over the early decades of its development. One way (Buchanan’s way) to express that critique was to emphasize that efficiency/Pareto optimality[8] i is an “emergent” property, reflecting the institutional features of the process in which different exchanges are made (or fail to be made). The standard neoclassical approach deriving from Samuelson’s treatment, by contrast, was to treat Pareto optimality as a property that could (in principle) be imposed by governmental fiat.[9]
Broadly Buchanan accepted the Austrian critique of the “socialist calculation exercise”[10]– but unlike the Austrians he was not prepared to accept that governments (or social planners) would be motivated to secure Pareto optimality (even for the purposes of the argument)! For him, any such assumption of government “benevolence” assumed what had to be proven: that is, it assumed that agents motivated by the same kinds of desires that apply to market agents would, by the operation of democratic incentives, be led to behave in the public interest. That claim, he thought, requires an argument – not an assumption! And exploring whether any such argument could be sustained became the central agenda for Public Choice analysis. In that exploration, Public Choice theorists took it as obligatory that political agents should be assigned the self-interested motivations ascribed to market agents – to treat the political process as a scramble for rival interests, just as market processes are taken to be. It is in this spirit that Buchanan referred to the Public Choice enterprise as “a theory of political failure” to set alongside the analysis of so-called “market failure”[11] or, to use another of his well-chosen epithets, Public Choice is “politics without romance.”
But at this point Buchanan might seem to face a problem – namely, how this comparison of markets and politics in terms of the exchange norms he endorsed could conceivably have any influence on anything. If markets fail less than politics fails, what possible impact could this fact have on either politics or markets? There simply seemed no point in the exercise where the normative force of the comparison could gain any purchase.
As I see it, it is in the face of this problem that Buchanan’s “constitutional” turn is to be understood. And it is to this I now turn.
Constitutionalism
A central part of the Buchanan intellectual scheme is the distinction between the rules of the game and plays of the game within the rules. Buchanan conceived of both markets and politics as games played within rules. But he thought that there was a “higher” level of choice – the choice of the rules. This level of choice is the “constitutional” level. Constitutions in Buchanan’s sense have, then, two central features:
1. a specification of the proper domains of market and political operation (which entails, in particular, limits on the domain of political decision-making). The notion that it might be left to in-period political processes to determine their own sphere of activities struck Buchanan as totally inconsistent with the constitutional logic. Limited government is the sine qua non of the constitutional approach;

2. a specification of how in-period politics is to operate. The most familiar illustration of this latter concern is, of course, the determination – in the Calculus of Consent, written with Gordon Tullock – of the “optimal decision rule” (or rules more accurately, since different kinds of decisions would predictably call for more or less inclusive majorities). Of course, the Calculus contains many other interesting arguments -- about bicameralism and the separation of powers, and fascinating suggestions about the role of institutions that are not strictly either markets or political processes, but something else – institutions of “civil society” perhaps. But the issue that most readers take away from the Calculus is that of whether simple majority rule would be the “appropriate” rule for collective decisions – with “appropriate” here taken to mean “unanimously chosen at the constitutional level.”
In an important sense the significance of “constitutional choice” lies in its capacity to instantiate the normative authority of exchange. In a move redolent of Rawls’s famous “veil of ignorance,”[12] Buchanan’s thought was that uncertainty about where one would lie in relation to the rules over the long horizon that were expected to be in place would moderate the effects of self-interest and thereby permit application of the unanimity rule (universal veto). In this way, Wicksell’s ideal of unanimity – the rule that Buchanan and Wicksell both saw as the “exchange ideal” – could be applied. Such a rule could not be applied for in-period collective decision-making because of “decision-making costs”: This was the central claim in the Calculus analysis. In that sense, unanimity at the in-period level is self-defeating: One would expect (virtual) unanimous consent to the proposition that unanimity be abandoned as the decision-rule for ordinary politics.
In-period politics could never be expected, then, to be fully “efficient” – to appropriate all possible gains from exchange – any more than markets could. The objective is not to eliminate market or political failure, but to optimize across institutions so that “failures” are minimized. That is, within the limits of what is institutionally feasible, one could specify an optimal mix of decentralized (market) and centralized (political) decision-making institutions – and it is that mix that would emerge in a unanimously chosen constitution behind a suitably thick veil of ignorance.
Questions and Queries
Now that I have outlined Jim’s main contributions, I would like to raise some points about his work that seem worth discussing.
1. Buchanan as a Classical Liberal: Buchanan is a self-declared classical liberal. By this, I take it that he means that he places a high value on liberty (understood as something like “noncoercive social relations”) and that he is a minimalist about the appropriate role of governments. His Public Choice analysis can clearly be viewed as providing the reasoned grounds for that minimalism, as his description of Public Choice theory as the “theory of political failure” suggests. Yet he is an unusual classical liberal. For one thing, whereas most classical liberals take as their point of departure some kind of conception of the individual’s moral rights and derive their conception of liberty in terms of rights violations (or coercion), Buchanan’s normative point of departure is in the intrinsically collective exercise of jointly working out the rules of the social/economic/political game to which citizens are to be subject. In that latter exercise all individuals hold a virtually complete right of veto over what those rules will be (including the specification of the personal and property rights that the individuals will possess and be subject to). Many libertarians have thought that this collectivist point of departure is inconsistent with true liberal individualism. Buchanan was insistent that social outcomes are not chosen because they are efficient (or fair) – they are efficient (or fair) because they are chosen (in the appropriate unanimous setting). In that sense, the foundational liberal element is embodied in the fact of unanimous constitutional choice – whatever the outcome of that choice process may turn out to be.

Buchanan was an unusual classical liberal in other ways. He believed rather passionately in confiscatory estate and gift duties: He reckoned that inherited wealth (though not self-made or first-generation wealth) violates basic equality of opportunity, and his enmity towards dynasties was notable. Hence the antipathy to John F. Kennedy mentioned in footnote 12. Buchanan though Papa Joe had bought the White House for his boys, and it infuriated him. However, one might think that individual sovereignty should extend to gifts and bequests, and that totally confiscatory gift/estate duties are unlikely to emerge from unanimously approved political rules.

2. The Double Role of Exchange: A related aspect of the Buchanan construction is the double work that the notion of exchange plays. At one level, individuals behind the veil of ignorance will predictably assess alternative institutions according to the mutual gains those institutions give rise: Clearly, certain basic facts about the operation of markets and the operation of democratic politics under various specifications will predictably be taken into account by the constitutional contractors. And it is specifically the role of economics (and Public Choice analysis as part of that enterprise) to reveal those facts in relation both to markets and to democratic political process. But within Buchanan’s scheme the ultimate exercise of “exchange” occurs in and through the constitutional contract itself, and the ultimate test of markets and politics lies in the constitutional endorsement they receive. In this way, the determination of the truth of claims about markets and/or of Public Choice seems to be assigned to the judgment of actual constitutional contractors. Buchanan could sometimes make such subjectivist gestures towards truth claims; but it seems bizarre to allow claims about the exchange properties of markets (say) to be determined by constitutional contract.

3. Market Operations: In the late afternoon of his life, Buchanan became intrigued by the significance of “increasing returns” in the operation of markets. In essence, this involved a recapturing of Adam Smith’s account of the fundamental forces making for the wealth of nations and recognizing its distinctiveness (as say from Ricardo and the modern mainstream tradition). One notable feature of this work is its “objectivist” qualities, that is, the “general opulence” distributed across all classes of society that was the explanandum for Smith is an externally observable phenomenon – a brute fact about human progress and not something that exists merely in the mind of the observer/evaluator.

4. The Supply of Versus the Demand for Rules: It is one thing to establish the “reason for rules,” and even what rules agents might choose in the hypothetical constitutional setting, and another entirely to explain how those rules will be enforced at the in-period level. As Jeremy Bentham famously put it in relation to rights, the demand for rules are no more rules than hunger is bread. In the treatment in the Calculus, where the agenda is to discuss modest modifications of rules that are already in play (the size of the majority), the assumption that the modifications will be enforced can be carried perhaps by the uncontestable fact that simple majority rule seems to be pretty robust. But as the agenda is generalized to include the entire template for rules governing social and political and economic life, the problem becomes acute. It needs to be explained just why agents who know their positions and who are presumed to be predominantly self-interested will find it in their interests to enforce and/or comply with the provisions previously agreed. To the extent that we look to courts to make decisions on the rules, and to the police to enforce court decisions, do we not need a “theory of legal failure,” alongside market failure and political failure, to sustain the entire project? Buchanan seems to have had a blind spot about this issue. But without some response to the quis custodet ipsos custodes challenge, it is by no means clear that the whole elegant intellectual edifice can get off the ground. And to the extent that the necessary response involves some modification of the extreme homo economicus motivational hypothesis, may we not be required to carry that modification into the analysis of markets and in-period politics on exactly the same generality grounds that Public Choice mounts its attack on the benevolent despot?

5. Chosen versus Inherited Rules? It is a critical feature of Buchanan’s constitutional paradigm that citizens choose the rules by which they are to live: Those rules have to be products of explicit consent. That fact explains why the market is an “efficient” institution only to the extent that it is constitutionally endorsed. Many observers (including F. A. Hayek and in another sense David Hume) are inclined to respect “evolved rules” in themselves and to doubt the intellectual pretension that Buchanan’s kind of constitutional constructivism involves. Buchanan himself explicitly rejects that kind of “respect”: He thought it invokes a kind of quietism towards the institutional status quo that is ultimately servile. Middle ground is presumably available here – but one would certainly want some principled way of discerning which established rules ought to be treated with piety and which ought to be interrogated and perhaps ruthlessly overturned. There may be a tension between American vigor and European traditionalism in play here.

6. Expressive Constitutionalism? I cannot forbear to mention, by way of conclusion, an anxiety that arises out of work of my own on voting.[13] That work is an extension of the idea of rational ignorance attributable to Anthony Downs in the sense that it takes as its point of departure the asymptotic irrelevance of each individual voter in determining the electoral outcome. This means that the relation between interests and behavior has a character in markets quite different from that in the ballot box. The individual voter is subject to a veil of insignificance not unlike the Rawlsian/Buchanan veil of ignorance, in that agents are distanced from their interests by the circumstances of choice. This fact very much blunts the distinction Buchanan draws between constitutional and in-period levels of choice in two senses: first because interests are attenuated in both settings; and second because individuals are in a large-number setting in making their constitutional agreements and hence a significant element of expressive behavior is likely to enter at the constitutional level. That is, people can quite rationally “cheer” for democracy or “trial by jury” or whatever, even when such institutions would not deliver better outcomes for them. This is not just a matter of rational ignorance – though there will predictably be plenty of that. It is also a matter of giving rational assent to all kinds of nostrums that “have strong expressive appeal” even when one knows they are silly or worse. Consider for example the vast extension of “rights” that seem characteristic of most modern (popularly endorsed) constitutions.
It is possible that in raising such queries I have been unjust to Buchanan’s intellectual scheme or placed the emphasis inappropriately -- or perhaps “buried” where I should have “praised.” My own view is that the way to honor Jim is to take his ideas seriously. That is what I hope I have done here. And I take confidence from the fact that Jim himself never drew back from a good robust argument!
Endnotes
[1] There is an interesting question here as to whether authors’ intentions are definitive in “proper interpretation.” In fact, Jim was never much preoccupied with what others made of his work: He was content to let it speak for itself (admittedly, in many iterations!) In any event, the fact of Jim’s absence gives interpreters (like me here) a certain license – and a corresponding responsibility – that we didn’t have in the same measure when he was alive.
[2] And of course not quite everything can be neatly shoehorned into a single frame. In over 20 volumes of work (the Collected Works published by Liberty Fund amounts to 19 and includes almost everything published up until 1998) there are bound to be somewhat independent pieces. But the claim of a general conceptual unity stands.]
[3] See for example his 1949 Journal of Political Economy article “The Pure Theory of Government Finance” – one of his very first major pieces. This article can be found in vol. 1 of The Collected Works of James M. Buchanan, pp. 119-32. [Not available online.]
[4] He drew a distinction between “logic” and “science” falling pretty much along the lines of the Austrian distinction between the a priori and the empirical. And like the Austrians, he thought that the a priori element could take one a long way.
[5] Whately, as the second incumbent of the Oxford Drummond chair in the 1830s, engaged a stirring (but obviously unsuccessful) battle to have the name of the discipline changed from “political economy”/“economics” to “catallactics.”
[6] Elsewhere I try to indicate what I see as distinctive in Buchanan’s emphasis on “politics as exchange” and in particular contrast his perspective with more conventional interpretations of the ”Public Choice paradigm.” See Brennan (2012).
[7] Despite protestations to the contrary, I reckon it still does. In most public-economics texts, “Public Choice” analysis is relegated to a separate independent chapter – as if it were a rather eccentric preoccupation that some scholars regard as important, but which can be excluded if the lecturer prefers. A distinguished notable exception to this generalization is Arye Hillman’s (2003) text and especially the second edition thereof.
[8] And, indeed, any normative feature of the social order against which market outcomes might be assessed (so “distributive justice,” for example). Buchanan and I discuss this issue explicitly in our Reason of Rules (1985) ch 7.
[9] In this, there was a kinship between the Public Choice approach and the earlier Austrian critique of “socialist calculation.” But whereas the Austrians emphasized the epistemic constraints involved in recognizing optimality, the Public Choice scholars emphasized the motivational aspects: If governments had the power to impose outcomes, why would they be motivated to choose the one that maximizes the “public interest” as they see it?
[10] He agreed, that is, that social planners would lack the information required for any such imposition.
[11] I say “so-called” here since the existence of “failure” is any relevant sense presumably implies that it is feasible to have something better.
[12] Buchanan always recognized the affinity between his approach and that of John Rawls, and often remarked that his project and Rawls’s are very similar, even though “they have been interpreted differently.” On one notable occasion at a Liberty Fund conference, Anthony Flew was mounting an all-out attack on Rawls’s “procrustean” scheme and was astounded at the severity of Buchanan’s response. Buchanan was as defensive of Rawls as he was enraged by John F. Kennedy (though the Kennedy issue is another story.)
[13] In various collaborations: Brennan and Lomasky (1993), Brennan and Hamlin (1998), and notably Brennan and Buchanan (1984).
ADDITIONAL READING
Online Resources
The Collected Works of James M. Buchanan, Foreword by Geoffrey Brennan, Hartmut Kliemt, and Robert D. Tollison, 20 vols. (Indianapolis: Liberty Fund, 1999-2002). The following 9 volumes are available online: <oll2.libertyfund.org/titles/280>.
The Intellectual Portrait Series: A Conversation with James M. Buchanan (2001)

Copyright and Fair Use Statement

“Liberty Matters” is the copyright of Liberty Fund, Inc. This material is put on line to further the educational goals of Liberty Fund, Inc. These essays and responses may be quoted and otherwise used under “fair use” provisions for educational and academic purposes. To reprint these essays in course booklets requires the prior permission of Liberty Fund, Inc. Please contact oll@libertyfund.org if you have any questions.

Current Posts

Contrasting Views of Liberty in Hume and Rousseau (November 2024) (Nov 12, 2024) Oakeshott and Liberal Education (October 2024) (Oct 08, 2024) Classical Liberals on Race (August/September 2024) (Aug 20, 2024) The Rebirth of the Austrian School and the South Royalton Conference: Marking the Fifty-Year Anniversary (June 2024) (Jun 04, 2024) Archived Posts…

Part of: James Buchanan and the Economics of Anarchy: Introduction