Skip to content

Liberalism and the Invisible Hand

Liberalism as a concrete sociopolitical order rests upon a series of invisible hand systems: free competition in explicit economic markets, free competition in the marketplace of ideas, institutional competition among branches of government, and so on. Yet liberal faith in these systems far outruns any of the social-scientific mechanisms or evidence adduced to support them. This is no happenstance, but a consequence of systematic problems that arise when liberal theory dispenses with the invisible hand of Providence in favor of secular mechanisms. First, a dilemma of verification afflicts invisible hand systems, in virtue of their indirect structure.1 By the terms of their creation, it is not generally possible to verify whether they are performing as promised. Second, there are self-undermining mechanisms within liberalism itself that threaten to prevent its invisible hand systems from fully realizing their promise. The main mechanisms are the depletion of preliberal social capital and public choice problems, both of which cause the agents posited by liberalism to act in ways that subvert the invisible hand processes themselves.

Given this structural propensity toward imperfect realization of liberal invisible hand systems, liberalism stands in an unhappy, and chronically unstable, halfway posture, caught between a preliberal faith in the invisible hand of Providence and self-destructive appreciation of its own contingency. Having dispensed with the superintending design of Providence in favor of contingent, indirect mechanisms, liberalism is astonished to find that there is no guiding hand to ensure the fulfillment of its own faith. The predictable reaction is liberal fideism, which insists ever more stridently on the truth of liberalism’s unverifiable and potentially self-defeating claims.


My main interest is in liberalism as an operating sociopolitical regime built around invisible hand systems, not in liberalism as theory. But the two are (partially) connected, and so a few preliminary words are necessary. Liberalism as theory I take to be a clan of doctrines, of which theological, political, and economic liberalism are the main branches. All these are descended more or less directly from the fateful thought that the autonomy of the individual, of the individual’s reason and desires, is of paramount importance.

In a widespread but cartoonish depiction of liberalism, especially of political liberalism, one hears people say that liberalism eschews any theory of the good in favor of a theory or theories of the right. Professional political theory, however, has long since relaxed this constraint, moving into rather interminable debates over “perfectionist” and “anti-perfectionist” liberalism. In the strands of liberalism relevant for my purposes, moreover, it is not true that liberalism even advertises itself as eschewing any concern for the good for human beings. What is true is that liberalism aims to produce human goods indirectly, through and by means of the operation of freely chosen individual actions.2 In this way, liberalism hopes to square its master commitment to autonomy, and its correlative horror of direction by authority, with concern for human goods. Those goods, according to liberal doctrine, will not be directly secured in virtue of authoritative commands, in virtue of ordinances of reason for the common good promulgated by one in authority,3 but instead will emerge indirectly, at the systemic level, out of the freely chosen acts of individuals themselves.

The invisible hand is thus no mere tool, one among others in the toolkit of liberal theory. The shift to an indirect, emergent structure is an indispensable trick that allows liberalism to square the provision of human goods with its radical horror of authoritative direction, on the one hand, and obedience on the other. Its target or antonym exists on two levels. On one level, human positive law and its rational authoritative ordinances directed towards the common good; on a higher level, the Providence of God, which in the classical conception is inherently directive, purposive, and authoritative, for Providence is the very “type of things ordered towards an end.”4

This structural account of the good in liberalism has two critical features for present purposes. First, the goods perceived by liberalism are individual goods, which may become aggregated at the social level but never become truly common. I will not discuss that important set of issues here. Second—this will be my focus—liberalism characteristically seeks to attain these individual goods indirectly, so that they arise through the decentralized action of individuals in spontaneous orders. Goods should in this way emerge at the system level—whether in economic, political, or legal systems—as “the result of human action, but not the execution of any human design.”5 The theologian Henri Grenier puts this succinctly in his outstanding treatment of liberalism:

The advocates of Liberalism imagine that the liberty of man living in society tends to what is best. Hence they hold that the end of civil society is the protection of private rights, from which peace and harmony result. This is the opinion of Quesnay and Adam Smith.6

The end of liberalism is not the protection of private rights, full stop. Rather, the structure of the claim is that liberty itself tends to produce the good indirectly; thus the goal of civil society is to protect private rights so that various goods (peace and harmony) may result from them.

The structure of decentralized individual interactions emerges clearly in Grenier’s account of economic liberalism, the only branch of liberalism he discusses at length. Under this doctrine (1) “private utility” is the “aim of economic life,” because (2) “it is the individual who can best seek, know and promote his own interests or utility,” and (3) “the first law of economics is free competition, i.e., the free play of economic individualities seeking, by any lawful means, the greatest advantages.”7 Each of these points is important. The first says that the liberal invisible hand, although anticipated to produce good(s), is directed at private individual goods (or an aggregate of such goods) rather than a genuinely common good. The second makes clear the connection between economic liberalism and the celebration of private individual judgment, initially expressed in theological liberalism.8 Friedrich von Hayek would later expand this point about the superiority of widely distributed private information into an archetypally liberal account of markets. Carl Schmitt expanded it into a liberal account of parliamentary debate (one that, needless to say, he did not share). The third point makes explicit that the main mechanism for producing these goods is self-interested competition in a decentralized structure, such as a marketplace, a marketplace of ideas, a constitutional system of “checks and balances,” and so forth.

The Invisible Hand: Analytics

The “invisible hand” as a phrase, as a shorthand for this characteristically indirect approach to the good, is of course associated with Adam Smith. Smith uses it first in a treatise on astronomy in an explicitly theistic way, then in the Theory of Moral Sentiments, and finally in The Wealth of Nations.9 The phrase’s background, however, is an enormously rich tradition of providentialist theorizing about politics and society that anticipates many of the ideas and problems of liberal theory.10 It even anticipates some of liberalism’s conclusions. Consider, for example, that Theodoret of Cyrus argued in the mid-fifth century AD that all are better off with a society featuring a division of labor, differences of wealth, and a division into rulers and ruled than they would be in an unspecialized and egalitarian society.11 Theodoret, of course, justifies his conclusions on entirely different grounds than liberal theorists would over a millennium later. For him, the social benefit represents a regulation instituted by divine Providence for the benefit of mankind, rather than a (solely) mechanical system-level consequence of decentralized self-interested interaction. The history of liberalism’s adaptation of the invisible hand, then, is a history of putative secularization, in which the superintending design of Providence is replaced by mechanisms in which system-level goods emerge without anyone necessarily intending that they do so.12 Later I will question the extent to which this arc is best described as “secularization” at all. In Smith, the secularization process is barely even underway.13

It is probably fruitless—although many have tried—to give necessary and sufficient analytic conditions for a cogent invisible hand account. As with liberalism, the idea is a classic family-resemblance complex, in which there is a kind of unity-in-variety even if there is no unique set of traits shared by all the members of the complex. For my purposes, however, the key hallmarks or notes of liberalism’s invisible hand systems are these: (1) individual agents engage in decentralized interactions, often but not necessarily competitive, such that (2) “as a result of human action, not of human design” (3) some system-level good or goods emerges. Condition (2) implies that the emergence of the system-level good(s) need not be part of the intentions of any of the individual agents, who may be pursuing strictly self-interested aims.14 Condition (3), on the other hand, does not necessarily imply that the goods that emerge are the same self-interested goods for which individuals were aiming.15

Liberalism does not promise that participants in its systems will attain the desires of their hearts, only that the quest for individual goods will cause aggregate—but not genuinely common—goods to emerge. Consider the seller in a perfectly competitive market, whose actions contribute to a system-level welfare maximum even though the seller’s own prices are driven relentlessly down to equal marginal costs. I will return to this point later. Furthermore, condition (3) is open-ended and agnostic about the nature of the relevant goods, which vary across different invisible hand stories. Welfare (preference satisfaction), truth or at least information, and political liberty have all been claimed to arise from invisible hand processes, as I will mention shortly.

Note, finally, that hybrid cases are possible. Condition (2) refers only to the agents within the invisible hand interaction, not necessarily to any agents who may have brought about the interaction. The emergent invisible hand system may itself be set up deliberately, as a result of human action and human design. Markets may be designed, as when corporations deliberately set up internal prediction or information markets to aggregate dispersed information held among their employees. Markets may also be imposed by coercion, as Hayek notoriously desired in the case of Chile.16 The point that markets may be designed or imposed, yet function in an invisible hand way, is frequently missed by those who assume that spontaneously functioning orders must emerge spontaneously.17 Even the hybrid cases are consistent with (what I have suggested is) liberalism’s horror of authoritative direction and obedience, because the designer of the system stops short of using his own substantive judgment about the appropriate goods for the agents, instead leaving them to autonomous action within the designed system. Nonetheless, in what follows, I will put aside the hybrid cases for simplicity.

The Invisible Hand: Examples

Let me now turn to some examples. In the spirit of condition (3), I will begin by lumping rather than splitting, listing a large number of cases that are intuitively in the neighborhood, as it were. I mean to indicate only how the accounts go, but certainly not to affirm their truth. My whole claim will be that structural problems ensure that the accounts far outrun any mechanisms and evidence that can be adduced in their favor, and rest essentially on a fideistic will to believe.

Liberalism in general. Before moving to particular liberal systems, it is important to grasp the striking unity of liberal thought across its different branches. Carl Schmitt argued that the idea of a “balance of opposing forces” with beneficial system-level results—whether by the interaction and competition of market actors, states, estates, or branches of government—“dominated European thought since the 16th century.”18 Schmitt directly linked this obsession to liberalism, arguing that invisible hand mechanisms of balance, competition, and emergent system-level goods were at the heart of the liberal project. The main three-way analogy or parallelism in liberal thought—something like the characteristic reflex of liberalism—becomes visible when we compare welfare-production in economic liberalism, truth-production in the marketplace of ideas, and liberty-production in the constitutional balance of powers. Schmitt writes:

It is essential that liberalism be understood as a consistent, comprehensive metaphysical system. Normally one only discusses the economic line of reasoning that social harmony and the maximization of wealth follow from the free economic competition of individuals, from freedom of contract, freedom of trade, free enterprise. But all this is only an application of a general liberal principle. It is exactly the same: That the truth can be found through an unrestrained clash of opinion and that competition will produce harmony. . . . The belief in public opinion is bound to a second, more organizational conception: the division or balance of different state activities and institutions. Here too the idea of competition appears, a competition from which the truth will emerge.19

This essential unity of liberalism as a concrete sociopolitical order arises, I suggest, because the invisible hand is a device that allows liberalism to reconcile the seemingly irreconcilable—to square the purposive pursuit of social aims, some good or aggregate of goods, with liberalism’s horror of authoritative direction. Jacob Viner correctly observed that “there is no logical conflict between a socially oriented teleology and individualism if the individual either has conscious social ends or by providential design serves such ends without having adopted them as his own ultimate objective.”20 Having made both obedience to the divine and public-spiritedness into matters of private choice, however, liberalism must then find arrangements for pursuing social aims that do not constrain the beliefs or motivations of autonomous individuals in those ways, arrangements that are robust to self-interested individual behavior and, by a kind of substitute miracle, transform those selfish motivations into social goods.

Of course the invisible hand is not the only strategy liberal theory uses to effect this reconciliation. Another vast set of problems involves the attempt to work out the tension introduced by positing that some subjection to authority can itself be freely chosen, self-imposed by the liberal subject, which then yields liberal theories of representation. But the existence of these other strategies does not mean that the invisible hand is a dispensable, contingent tool for liberalism. It promises to reconcile a core tension at the heart of the liberal project.

Explicit markets. “Free” markets are the paradigmatic case for the invisible hand. I will take this to be so familiar that little time needs to be spent on it. Suffice it to say that there are here two variants: the neoclassical general equilibrium model in which markets are taken to maximize aggregate preference satisfaction, and the Hayekian version in which markets are taken to produce coordinating information. In the neoclassical version, all available trades are consummated in perfectly competitive markets, ensuring both allocative and productive efficiency. The economy thereby reaches one or another Pareto-optimal state (which one depends upon initial conditions), such that any reallocation would make at least one person worse off. Nothing in this account addresses the distributive effects of the market; I will return to this point later.

In Hayek’s variant, by contrast, the main benefit of market competition is informational efficiency, which coordinates social plans. The price system impounds widely distributed, tacitly held information about the real cost of resources. In this way the market coordinates individual expectations better than any centralized planner could possibly achieve.21 Competition serves not only as a mechanism for preference satisfaction, but as an epistemic “discovery procedure”22 that reveals objective information about the world, including the plans and expectations of others.

The marketplace of ideas. Justice Oliver Wendell Holmes Jr. famously wrote that “the best test of truth is the power of the thought to get itself accepted in the competition of the market.”23 Here the market is strictly metaphorical, and other metaphors flow freely in its wake, such as the “clash of opinions.” Whatever the literary situation, however, the idea is that truth will emerge from the decentralized interaction of many individuals freely expressing competing ideas. In this way truth, acknowledged to be desirable, is not pursued directly by authorities concerned with the common good. Rather it is taken to arise indirectly out of structured interactions.

The domain of this marketplace of ideas may be either throughout an ill-defined, society-wide sphere of public discourse, or more specifically within a constitutional parliament. Carl Schmitt, in a remarkable parallel to Hayek’s account of widely distributed private information in economic markets, argued that the essence of liberal parliamentarism lay in “the theory of a balance of opposing forces from which truth will emerge automatically as an equilibrium” through open discussion in Parliament, where “particles of reason that are strewn unequally among human beings gather themselves and bring public power under their control.”24

Whatever the precise domain of its application, this invisible hand account of free speech comes in both conceptual and empirical versions.25 In the conceptual version, truth is defined as whatever emerges from a suitably open process of discussion. In the empirical version, usually associated with John Stuart Mill, the claim is that truth tends to emerge on average and in the long run from a suitably open process of discussion. As we will see later, however, the problem with the former is that it is ultimately tautological, while the problem with the latter is simply that there is no particular reason to think it is true.

Checks and balances—international and constitutional. We have seen that Schmitt, in his famous treatise on constitutional theory, declared that “the idea of a counterpoise, of a ‘balance,’ of opposing forces has dominated European thinking since the sixteenth century.”26 One major strand of this, which I will mention but otherwise ignore, involves the “balance of power” as a model in international relations, in which the goods of peace and order in the international sphere are secured indirectly, in equilibrium, by the mutual checking of sovereign powers. Let me focus instead on the domestic analogue, the balance of constitutional powers. For thinkers as diverse as Montesquieu,27 Joseph de Maistre,28 and Schmitt, the crucial example was the (putative) equilibrium of the English unwritten or uncodified constitutional order. As Schmitt put it,

Of special significance for the ideal constructions of the English constitution is an idea of the tripartite balance and of the “equilibrium of powers,” out of which the free government results: king, upper house and lower house; between king and Parliament (specifically, upper house and lower house taken together), between legislature and executive, between the prerogatives of the king and the freedom of the people. All these should be “balanced.”29

The idea of a liberty-protecting equilibrium among King, Lords, and Commons30 was adapted by early modern and Enlightenment theorists, such as Montesquieu, to become a balance among institutions wielding constitutional powers—legislative, executive, and judicial.31

The framers of the American Constitution were in general heavily influenced both by Montesquieu and by the thinkers of the Scottish Enlightenment, including Adam Smith and Adam Ferguson, and theorized the invisible hand as a model for the constitutional arrangements of 1787.32 In Madison’s scheme, one major safeguard—complementing the filtering and selection of elevated characters, in and because of an extended Republic—was an invisible hand system of checks and balances, in which competition among institutions (as “ambition counteracts ambition”)33 prevents any one of them from attaining a monopoly of power, thereby protecting liberty.

Religious pluralism. Madison was also heavily influenced by Smith’s invisible hand account of religious pluralism, in which the absence of any established church meant that a variety of sects would have to compete for adherents. In Smith’s view, this competition would have the beneficial consequence that individual sects would have to moderate and soften their views, especially by acceding to a regime of mutual toleration. From the standpoint of the individual sects, each of which desires a privileged position, this need not at all be a benefit, but Smith believed that individuals generally would benefit from the resulting consolidation of the liberal order. As in explicit markets, the invisible hand system of religious pluralism would produce aggregate goods even though, or because, it constrained and defeated the individual interests of the participants.34

The extended republic and federalism. Finally, in a straightforward extension of Smith’s view of religious pluralism, Madison argued that the extended republic, built around a federal system, would produce a beneficial and mutual jarring of factions, and prevent any one of them from assembling a majority coalition, especially under conditions in which “communication is always checked by distrust.”35

The adversary system. By analogy to free speech, many have argued that an adversarial trial system—as opposed to an inquisitorial one—generates a kind of competition to produce biased information by both sides, resulting, as if by an invisible hand, in more information overall for the tribunal—a result which neither party intends as such. James Fitzjames Stephen, the barrister and judge, argued that while the inquisitorial system of criminal procedure is “beyond all question the true one,” an attainable second-best is

to consider the trial mainly as a litigation, and to allow each party to say all that can be said in support of their own view; just as the best means of arriving at the truth in respect of any controverted matter of opinion might be, to allow those who maintained opposite views to discuss the matter freely and in public. 36

The implicit claim here is that, whereas an inquisitorial system posits a magistrate who aims directly and intentionally at collecting all relevant information, an indirect, competitive, and decentralized system of adversarial production of information will yield more information overall.

The common law. There exists a whole sub-complex of liberal accounts in which the common law, or common-law judging, represents a decentralized process that yields larger social benefits, as if by an invisible hand. I will mention only one version,37 the so-called Cardozo Theorem38—a famous apothegm from Justice Benjamin Cardozo—from his book on common-law reasoning:

The eccentricities of judges balance one another. One judge looks at problems from the point of view of history, another from that of philosophy, another from that of social utility, one is a formalist, another a latitudinarian, one is timorous of change, another dissatisfied with the present; out of the attrition of diverse minds there is beaten something which has a constancy and uniformity and average value greater than its component elements.39

The interaction among judges deciding cases over time leads, without any individual judge necessarily intending it, to a body of law with socially beneficial properties.40

Liberal and Nonliberal Justifications

We have to be clear that an invisible hand account is not, by itself, necessarily a liberal account. There is a crucial ambiguity in Ferguson’s formulation “the result of human action, but not the execution of any human design.” The latter clause is consistent with either of two entirely distinct possibilities: that the emergent system-level good arises as a result of no one’s design, through some discernible but nonintentional mechanism that causes it to arise; or that it arises as a result of nonhuman design, not through human intentions but still through intentions—that is to say, through the divine design for which Providence is a shorthand.

This distinction points to two entirely different accounts of how an invisible hand account might work, which I will call the providential account and a liberal or mechanistic account. Accordingly there can be liberal and nonliberal versions of the invisible hand, just as, in many other areas, there can be liberal and nonliberal justifications for seemingly similar political and moral concepts.41 Architecturally similar structures may rest upon very different foundations—in one case a living, providentialist foundation, in another a mechanistic, strictly individualistic and liberal one.

For an example of the contrast, recall the “Cardozo Theorem” mentioned above, in which the biases (Cardozo’s “eccentricities”) of judges making decisions over time somehow yield a body of common-law precedent that is superior to any of the decisions taken individually, in a way that produces a body of law that is socially desirable overall. For this liberal invisible hand account to succeed, there must be some mechanism by which these biases compound or offset one another. The alternative, of course, is simply that decisionmaking according to the biases of successive judges yields an increasingly incoherent, disjointed, politicized body of law. Cardozo was filled with faith that such a result could be avoided, yet offered no real mechanism to explain his faith.

By way of comparison, Cardinal John Henry Newman offered two distinct accounts of invisible hand processes in the life of the Catholic Church, structurally similar to Cardozo’s Theorem. One was an account of the emergent moderation and balance produced by the biases of successive ecumenical councils in the history of the Church:

It would seem as if the Church moved on to the perfect truth by various successive declarations, alternately in contrary directions, and thus perfecting, completing, supplying each other. Let us have a little faith in her, I say. Pius is not the last of the Popes—the fourth Council modified the third, the fifth the fourth.42

In a second account, Newman explicitly compared the liberal account of checks and balances to the endless contest between “authority” and “private judgment” in the life of the Church:

As in a civil polity the State exists and endures by means of the rivalry and collision, the encroachments and defeats of its constituent parts, so in like manner Catholic Christendom is no simple exhibition of religious absolutism, but it presents a continuous picture of Authority and Private Judgment alternately advancing and retreating as the ebb and flow of the tide.43

But there is a crucial difference between the Cardozo Theorem, on the one hand, and the two Newman Theorems, on the other. The difference is certainly not that the former arises “as a result of human action, but not of human design.” That is equally true of both; neither rests on human design, and any suggestion that Ferguson’s formulation uniquely describes the liberal invisible hand account is a confusion. Rather, the difference is that the Newman Theorems implicitly posit the superintending, invisible, and infallible hand of divine Providence, whereas the Cardozo Theorem rests on some hypothetical, unspecified mechanism that causes the desirable system-level result to emerge without any design whatsoever, human or otherwise. The open question for liberalism’s account is, always, whether and to what extent such a mechanism can be supplied, and whether it is in fact supported by evidence. Let us turn now to that question.

Inevitably Imperfect Secularization

I am now in position to complete my thesis: liberalism attempts to replace Providence with individualist, nonintentional accounts of the production of goods, yet this secularization chronically promises more than it can deliver. The mechanisms adduced to yield the system-level emergence of the relevant goods prove—for systematic and not merely happenstance reasons—far more fragile than their liberal proponents suggest. Yet the need for general justifications for liberalism makes admission of this radical fragility threatening. The consequence is a marked tendency to liberal fideism, to ever-more fervent insistence on the scope and power of the liberal invisible hand, even in the face of mounting evidence to the contrary.

I will flesh out the point only with regard to free markets, the marketplace of ideas, and checks and balances, but I believe parallel analyses hold for the other accounts I have mentioned. Whatever the context, the point is that invisible hand mechanisms are intrinsically fragile, holding only under special conditions or highly idealized circumstances. In the following sections, I then turn to some structural mechanisms according to which the liberal invisible hand has self-undermining propensities. Taken together, these arguments show that liberalism’s invisible hand is caught between promise and performance.

Free markets. In a popular lecture series on economics, the session devoted to the invisible hand lists a long series of partially overlapping conditions that must obtain for Smith’s invisible hand to hold in the manner interpreted by the results of neoclassical general equilibrium theory: “an open market,” absence of monopoly power, absence of monopsony power, “fine-grained division of production and consumption,” no producers with “pivotal private technology,” “perfect information, more or less truthful information across the whole market” (!), and government to enforce property and contract rights.44 Needless to say, this bears roughly the same relationship to actually existing economies that Star Trek bears to actually existing navies. It is not even conceptually clear what “perfect information” might mean, in a world of limited attention and computational capacity. To be sure, a main research program of economics is to relax these conditions as far as possible, yet it remains true that the conditions for productive and allocative efficiency in competitive markets are highly fragile, requiring a rather well-defined set of circumstances.45

Hayek’s version of the free market invisible hand avoids some of these difficulties, principally by making more plausible informational assumptions, but encounters other difficulties in its turn. Let me list just a few crucial issues:46

(1) Sometimes prices supply misleading information about the real cost of resources, as in monopoly (in which the monopolist prices above marginal cost). The standard response is antitrust enforcement, which in turn creates the possibility of corruption of the enforcer or political influence, or simply differential success in litigation for the rich. I will return to these public choice problems later.

(2) Coordination problems afflict the choice of “compatibility standards” for products and technology. Such problems occur when there exists more than one mutually beneficial equilibrium for market actors; prices will not tell the actors which equilibrium to adopt.

(3) The Hayekian argument for the informational efficiency of markets notoriously suffers from its reliance upon self-undermining incentives. Roughly, market actors will have incentives to acquire costly information only if they can profit from it. But if they can, then—precisely on Hayekian grounds—their information will be transmitted via the price system to competitors who may profit as well, even though they have not made costly investments in information. Anticipating this free-rider problem, competitors may decide not to invest in information, but if that occurs, then everyone has incentives to do so—and so forth. In other words, the costliness of information creates a free-rider problem that has no general equilibrium solution, implying that markets for information are intrinsically unstable. The more efficient the price system becomes at conveying information, the worse the free-rider problem becomes, the less efficient the price system will be, and so on in a circle. In this way, it has been shown that so long as information is costly, no market can be fully informationally efficient, even in principle.47 Hayek’s appeal to the price system to give an informational rationale for free markets is in this sense self-defeating.

The marketplace of ideas. As we have seen, the marketplace of ideas hovers perpetually and restlessly between a tautology (truth is defined as whatever emerges from the conditions of free and open discussion) and an empirical thesis (conditions of free and open discussion do in fact, on average and in the long run, tend to produce truth). This equivocation papers over the extraordinary weakness of the empirical thesis: the most that can be said is that there is no general mechanism to sustain it and no evidence that it is generally true, even on average and in the long run. As Frederick Schauer observed, “people are not nearly so rational as the Enlightenment assumed, and without this assumption the argument from truth evaporates. . . . [I]f the relevant time period is discrete and observable, history furnishes far too many counter-examples for us to have much confidence in the power of truth consistently to prevail.”48 Since that observation was offered, research programs in cognitive and social psychology and behavioral economics have radically undermined any remaining, naïve faith that truth will automatically tend to “win out” in free and open discussion. It would take a kind of miracle of providential stature for the large and motley assortment of human heuristics and biases to interact in just the right way so as to make the marketplace of ideas into a reliable truth-generating machine.

In the past generation or so we have seen democratic polities with expansive constitutional protections for free speech, emphatically including the United States, make extraordinary mistakes of policy based on distorted information and distorted, ideologically inflected processing of information—Iraq and Libya leap to mind. Domestically, these polities have fallen subject to pervasive conspiracy theorizing, emphatically including the educated elites who process and discuss information, and who partly shape the agenda for public opinion. Those elites, who pride themselves on their discursive rationality and commitment to evidence, are not obviously any less susceptible to the rationalization of political prejudices, nor any more open to evidence, than the public at large. The idea that these groups will shape the public agenda in a way that systematically tends to cause truth to win out has no obvious foundation.

Checks and balances. The liberal faith in checks and balances is powerful, even passionate. It also far outruns any mechanisms and evidence that might be adduced to support it. Three major issues have been identified. First is the gap between individual incentives and interests, on the one hand, and institutional incentives and interests on the other.49 What is good for Congress need not be good for individual legislators, and vice versa. The exact consequences of this gap are unclear; in one version, the gap between individual and institutional incentives is far less for the presidency and executive branch, which thus does a better job of pursuing its long-run institutional interests than does Congress, and tends to accrete power over time.50 Whatever the results, however, the point is that the claimed benefits of the Madisonian invisible hand of checks and balances do not materialize.

A second issue is that there is no external enforcer, standing above the constitutional system of checks and balances, to force the institutions to compete. They may decide instead that various forms of collusion, power-sharing, or delegation make them better off, in ways that undermine the check of ambition against ambition.51 In an idealized ordinary market, antitrust law and other institutions prevent firms from colluding to set prices above marginal cost. (I will relax this assumption below.) This presupposes an enforcing sovereign external to the market mechanism. As there is no such enforcer external to the constitutional order, however, there is no analogue to antitrust enforcement. The hope is that the system of checks and balances amounts to a fully self-enforcing equilibrium, but the conditions under which this will occur are rarely specified and there is no obvious reason to think that they typically obtain.

The final issue is that the analogy to explicit competition in economic markets assumes away a critical issue: there is no price system in the constitutional order. The price system is of course the key mechanism that makes competition in explicit markets conduce to welfare (in the neoclassical version) or to information (in the Hayekian variant). As resources become costlier, the shift in relative prices determines what trades are consummated and what information is impounded in the prices themselves. What exactly corresponds to this in the “market” for constitutional checks and balances? There is no systematic reason to think that the equilibrium level of competition, or that the “balance” of power resulting from this competition, will correspond to any socially desirable state of affairs. If it does, in a particular case, it is a happy accident.

The Dilemma of Verification

In the cases I have just canvassed, the systematic inability to tell whether and to what extent invisible hand systems actually work, or work accidentally, is not itself an accident. It is built into the nature and justification of such systems. If individual goods (however defined theoretically) could be directly identified, generated, and allocated by the political authority, there would be no need for the indirection of liberalism’s invisible hand. In a famous observation about competition as a discovery procedure, Hayek noted that “it would be patently absurd to sponsor a contest if we knew in advance who the winner would be. . . . The only reason we use competition at all has as its necessary consequence the fact that the validity of the theory of competition can never be empirically verified for those cases in which it is of interest.”52

The problem bedevils any number of invisible hand systems. As to the adversary system, barring unusual cases, “[w]e can’t learn directly whether the facts are really as the trier determined them because we don’t ever find out the facts.”53 As to explicit markets, to appoint a centralized bureau to collect all the information needed to determine whether the price system in fact processes information effectively would itself be a conceptual blunder, resting on a misunderstanding of the informational rationale for decentralized prices. In this and the other cases, the very same indirect structure that allows liberalism to reconcile its commitments to autonomy with its horror of obedience also makes it structurally difficult to discern whether the invisible hand works as promised.

Self-Undermining Propensities

So far I have attempted to show that liberalism’s faith in its suite of invisible hand accounts greatly outruns any of the mechanisms or evidence supporting them. We can go farther, however: in important cases, the liberal invisible hand is self-undermining. This can happen in two ways, which I will call the depletion of preliberal social capital and the public choice problem. To the extent either of these occurs, the gap between liberalism’s faith in the invisible hand and its fulfillment will have a systematic propensity to persist for structural reasons.54

The depletion of preliberal social capital. Liberalism’s invisible hand stories malfunction when their own individualistic and mechanistic premises, the premises that enable them to function, undermine the very preconditions that are necessary for the mechanism to work—preconditions that, it is argued, preexist liberalism and that cannot be replaced by it. A variety of hypotheses have been suggested here,55 variously involving the undermining of trust, of unwritten norms and conventions, or of other forms of preliberal or nonliberal solidaristic social capital necessary for the operation of smoothly functioning invisible hand systems. An example from markets is the view that “the costs of contracting and enforcing contracts make it impossible to exclude all forms of literalism, circumvention of promises, and chicanery; norms of good faith are thus necessary to facilitate exchange.”56 When combined with the suggestion that participation in market or quasi-market systems tends to form liberal agents with dispositions that undermine the norms of good faith, the depletion problem arises.

Whatever the details, the larger point is the same across invisible hand systems. Free markets, free discourse in the public sphere, the checks-and-balances system of institutional competition—all these presuppose, even if only implicitly, background conditions that allow competition to occur within stable boundaries and under regulated conditions. The suggestion is that these implicit boundaries on competition improve on the chaotic counterfactual in which competition—economic, rhetorical, or political—becomes unconstrained and mutually destructive. On this view, political competition resembles regulated warfare, which in some times and places occurs within a stable, regulated framework that (it is claimed) makes all better off. An example is the “ethic of partisanship,”57 according to which partisan competition occurs within a framework of tacit internalized norms and expectations that constrain its worst excesses.

To be sure, hypotheses about the self-undermining of normative capital meet their opposites in the Marxist “feudal shackles thesis” and other arguments that the persistence of preliberal norms or attitudes actually hampers the free operation of liberalism’s invisible hand systems.58 There is no general answer to these debates, which is all I need for my point. The depletion of preliberal capital is a grave possibility that only adds to the debilitating uncertainty about whether liberalism’s accounts capture anything real.

The public choice problem. Earlier, I assumed temporarily that in idealized markets, antitrust rules and other institutions prevent competitive agents from subverting the process of competition itself. As public choice theorists have pointed out, however, the same premises that make the liberal invisible hand story go also suggest that the agents within the market framework will have every incentive to undermine or subvert it, through self-interested action aimed at manipulating the framework itself.59 Where the anticipated economic benefits exceed the expected costs, including legal risk, such agents will use political donations, lobbying campaigns, strategic litigation, or even outright bribery to shape the very rules of the game for enduring advantage, as by corrupting the institutions of antitrust enforcement or by procuring regulation that insulates incumbents against innovation by new challengers. The motivations attributed to the agents within liberalism’s invisible hand accounts ensure that their behavior will escape the storyteller’s control, as it were, to change the parable’s very ending.

This point may also be put in terms of the distribution of benefits from the liberal invisible hand. On close inspection, there is an equivocation in liberalism’s accounts: to whom exactly do the benefits accrue? Sellers in perfectly competitive markets, pursuing self-interest, can price no higher than marginal cost, and profits are driven relentlessly towards zero. The operation of such a market provides immediate surplus benefits to consumers. Nothing in the market setup itself necessarily ensures that the self-interested aims of the participants actually result in promoting their own interests.60 That individuals seek their own benefit need not entail that they do benefit. The self-benefit of participants may be an intended but unrealized consequence, while the system-level benefit to others is an unintended but realized consequence.61

But this creates a dilemma on liberal individualist premises. If others capture all the benefits of my participation in the system, why should that be good for me? Hence the standard neoclassical claim that ex ante free markets are potentially better for all (including producers), in a larger long-run sense, as the surplus generated by markets can be redistributed through the tax-and-transfer system. But as many have pointed out, this Kaldor-Hicks criterion, which identifies a change as desirable if it generates surplus that can be allocated so as to make all better off, is strictly hypothetical comfort. Whether all will in fact be made better off—the very different Pareto criterion62—is a fideistic assertion that outruns all evidence.

The overall point is that the uncertain gap between promise and performance is no mere happenstance, but arises out of a kind of structural propensity. Having denied the superintending role of Providence, liberalism commits itself to trust in indirect, seemingly magical or even miraculous processes for transmuting the dross of human motivations into the gold of system-level goods. But those motivations are not so easily left behind; agents who possess them may either work to undo the framework within which liberalism’s miracles occur (public choice), or else, by participating within liberal systems, lose the traits that were necessary for those systems to work in the first place (depletion of preliberal social capital).

Liberal Fideism

Constitutional arrangements, free markets, and free speech are the core institutions of liberalism. In all of these contexts, liberal faith far exceeds the causal mechanisms or evidence that suggest that liberalism’s invisible hand fulfills its promises. The result is an uneasy halfway posture, in which (1) the invisible hand must succeed so that liberalism’s indirect approach to the good may prevail, yet (2) the providentialist underpinnings of the invisible hand that would yield its success have been abandoned, and (3) it is unclear what if anything has taken Providence’s place as superintendent and guarantor.

Psychologically and analytically, the only way out of this bind is liberal fideism. It must be believed, despite or indeed because of the radical contingency and uncertainty of liberalism’s invisible hand, that the invisible hand will produce its beneficial results sooner or later, in some immanentized eschatological future. Consider the “long run” of the liberal free-speech parable, which promises that open speech conditions will definitely result in the emergence of truth, just a bit farther on—in a long run that is always somewhat longer than any actual time horizon one can observe. Having rejected providentialism, liberalism becomes increasingly insistent, and in liberal societies one encounters lay faith movements and quasireligious orders centered on evangelizing and catechizing the benefits of free markets, free speech, and constitutional liberties, centered on the separation of powers and checks and balances.

Liberalism’s standing gap between promise and performance is filled by a radiant faith, but that faith has already disavowed the premises that would support it. The result is not a rational faith, but a groundless fideism and indeed fanaticism that insists, with ever-increasing fervor, on more of the same. The answer to the problems of free markets is even more free markets; the answer to the problems generated by the operation of the separation of powers system is a renewal and reinforcement of the separation of powers; the cure for the ills generated by free speech is even more free speech. By supposing that the remedy for fire is to add fuel to the fire, this fideistic reaction merely exacerbates the failures of the liberal invisible hand, intensifying the problems it hoped to solve. The result is a regime in which liberalism’s accounts generate ever less allegiance from the populace, while elites insist upon their truth all the more fervently—a condition which, as theorists persuasively documented,63 we indeed observe.

Liberalism as Parable

Let me close not by repeating my claims, but by underscoring what I have not claimed. I have not claimed that the providentialist version of the invisible hand is unproblematic, or that it does not face analogous difficulties. The whole elaborate body of thought that struggles with divine Providence64 in history testifies to both the promises and problems. That the same sorts of arguments that I deploy against liberalism’s invisible hand accounts may also be deployed against the providentialist invisible hand is, indeed, precisely the point. Liberalism’s characteristic obscurantism is to conceal its own character as political theology, as one of the great world religions. The triumphalist version of the liberal faith65 is that, having finally transcended the great fallacy that ordered liberty requires a Designing Hand, we can now see how ordered liberty and its resulting goods arise indirectly, but not by anyone’s design.66 My suggestion is that this is simply to replace one faith and one political theology with another. Which faith is the true one is a separate question, but liberalism’s invisible hand accounts—perhaps it would be better to call them parables—are no less complex and challenging than the providentialist parables they would replace, and the necessary miracles no less astonishing.

This article originally appeared in American Affairs Volume III, Number 1 (Spring 2019): 172–97.


Previously explored in Adrian Vermeule, The System of the Constitution (Oxford: Oxford University Press, 2011).

The best treatment is Henri Grenier, Thomistic Philosophy, trans. J. P. E. O’Hanley, vol. 3, Moral Philosophy (Charlottetown, Canada: St. Dunstan’s University Press, 1948). An excellent recent explication of liberal constitutionalism, which clearly grasps the indirect character of liberalism’s approach to the good, is N. W. Barber, The Principles of Constitutionalism (Oxford: Oxford University Press, 2018). Barber’s treatment does suffer from failure clearly to distinguish aggregate good(s) from (the) common good. See, e.g., p. 241 (speaking of “private goods” that are “aggregated into the common good”). But as I mention later, that is just a hallmark of liberalism itself.

Summa theologiae, 2–2.90 et seq.

Summa theologiae, 1.22.1 in corp.

Adam Ferguson, Essay on the History of Civil Society (Edinburgh: Kincaid & Bell, 1767), 183.

Grenier, Moral Philosophy, 353.

Grenier, Moral Philosophy, 454.

See John Henry Newman, Apologia Pro Vita Sua, Note A, Liberalism: “Now by Liberalism I mean false liberty of thought, or the exercise of thought upon matters, in which, from the constitution of the human mind, thought cannot be brought to any successful issue, and therefore is out of place.” Apologia Pro Vita Sua: Being a History of His Religious Opinions (London: Longmans, Green, Reader, and Dyer, 1875), 288.

Paul Oslington, “Divine Action, Providence and Adam Smith’s Invisible Hand,” chap. 5 in Adam Smith as Theologian, ed. Paul Oslington (New York: Routledge, 2011).

10 The classic overview is Jacob Viner, The Role of Providence in the Social Order: An Essay in Intellectual History (Philadelphia: American Philosophical Society, 1972). The providentialist interpretation persists well into the early modern era; consider, for example, Giambattista Vico:

Out of ferocity, avarice, and ambition, the three vices which lead all mankind astray, [society] makes national defense, commerce, and politics, and thereby causes the strength, the wealth, and the wisdom of the republics; out of these three great vices which would certainly destroy man on earth, society thus causes the civil happiness to emerge. This principle proves the existence of divine providence: through its intelligent laws the passions of men who are entirely occupied by the pursuit of their private utility are transformed into a civil order which permits men to live in human society.

Quoted in Albert O. Hirschman, The Passions and the Interests: Political Arguments for Capitalism before Its Triumph (Princeton: Princeton University Press, 1977), 17 (emphasis added).

11 Theodoret of Cyrus, On Divine Providence, trans. Thomas P. Halton, Ancient Christian Writers 49 (New York: Newman Press, 1988), 78–98.

12 Thus Edna Ullmann-Margalit, in a pair of highly influential papers, sees the displacement of providential design by nonintentional secular mechanisms as the key to the development of the invisible hand. Edna Ullmann-Margalit, “The Invisible Hand and the Cunning of Reason,” Social Research 64, no. 2 (Summer 1997): 181–98; “Invisible hand Explanations,” Synthese 39, no. 2 (October 1978): 263–91.

13 Viner, Role of Providence, 79–82. As Richard Whately put it, for Smith, “Man is, in the same act, doing one thing by choice, for his own benefit, and another, undesignedly, under the care of Providence, for the service of the community.” Introductory Lectures on Political-Economy (1832). Quoted in Oslington, “Divine Action,” 61.

14 To be clear, the agents need not be pursuing self-interest strictly understood in a material sense, corresponding to avarice; they may be pursuing other self-regarding aims, such as glory or fame, a pursuit which may in some sense be contrary to their immediate material “interests.” For example:

In fact, the idea of an “Invisible Hand”—of a force that makes men pursuing their private passions conspire unknowingly toward the public good—was formulated in connection with the search for glory, rather than with the desire for money, by Montesquieu. The pursuit of honor in a monarchy, so he says, “brings life to all the parts of the body politic”; as a result, “it turns out that everyone contributes to the general welfare while thinking that he works for his own interests.”

Quoted in Hirschman, The Passions and the Interests, 10.

15 Samuel Fleischacker, “Adam Smith’s Reception among the American Founders, 1776–1790,” William and Mary Quarterly 59, no. 4 (October 2002): 897–924.

16 “Personally I prefer a liberal dictator to democratic government lacking liberalism. My personal impression—and this is valid for South America—is that in Chile, for example, we will witness a transition from a dictatorial government to a liberal government. And during this transition it may be necessary to maintain certain dictatorial powers, not as something permanent, but as a temporary arrangement.” F. A. Hayek, interview by Renée Sallas, El Mercurio (April 12, 1981). See also Corey Robin, “The Hayek-Pinochet Connection: A Second Reply to My Critics,” Crooked Timber (blog), June 25, 2013.

17 This distinction is made helpfully in the papers by Ullmann-Margalit.

18 Carl Schmitt, Constitutional Theory, trans. Jeffrey Seitzer (Durham: Duke University Press, 2008), 221.

19 Carl Schmitt, The Crisis of Parliamentary Democracy, trans. Ellen Kennedy (Cambridge: MIT Press, 1985), 35.

20 Viner, Role of Providence, 60.

21 Hayek, “The Use of Knowledge in Society,” American Economic Review 35, no. 4 (September 1945): 519–30.

22 Hayek, “Competition as a Discovery Procedure,” trans. Marcellus S. Snow, Quarterly Journal of Austrian Economics 5, no. 3 (Fall 2002): 9–23.

23 Abrams v. United States, 250 U.S. 616, 630 (1919) (Holmes, J., dissenting).

24 Schmitt, Crisis of Parliamentary Democracy, 35–36.

25 Frederick F. Schauer, Free Speech: A Philosophical Inquiry (Cambridge: Cambridge University Press, 1982).

26 Schmitt, Constitutional Theory, 221.

27 Montesquieu, The Spirit of the Laws, trans. Anne M. Cohler, Basia Carolyn Miller, and Harold Samuel Stone (Cambridge: Cambridge University Press, 1989), pt. 2, bk. 11, chap. 6, “On the Constitution of England.”

28 Joseph de Maistre, On God and Society: Essay on the Generative Principle of Constitutions and Other Human Institutions, ed. Elisha Greifer, trans. with the assistance of Laurence M. Porter (Chicago: Regnery, 1959); Considerations on France, trans. Richard A. Lebrun (Cambridge: Cambridge University Press, 1994).

29 Schmitt, Constitutional Theory, 221.

30 The idea descends ultimately from Polybius’s account of an equilibrium among Senate, consuls, and people, which of course was justified on entirely different grounds. Polybius, The Histories, trans. W. R. Paton, Loeb Classical Library 188 (London: Heinemann, 1927), bk. 6.

31 M. J. C. Vile, Constitutionalism and the Separation of Powers (Oxford: Clarendon, 1967).

32 See, among others, David Prindle, “The Invisible Hand of James Madison,” Constitutional Political Economy 15, no. 3 (September 2004): 223–37.

33 See The Federalist; with The Letters of “Brutus”, ed. Terence Ball (Cambridge: Cambridge University Press, 2003), no. 51 (James Madison).

34 Fleischacker, 911–12.

35 Federalist, no. 10 (James Madison).

36 James Fitzjames Stephen, “Of Criminal Procedure in General,” in Selected Writings of James Fitzjames Stephen: A General View of the Criminal Law, ed. K. J. M. Smith (Oxford: Oxford University Press, 2014), 173.

37 For others, see Vermeule, Law and the Limits of Reason (Oxford: Oxford University Press, 2009).

38 Nicola Gennaioli and Andrei Shleifer, “The Evolution of Common Law,” Journal of Political Economy 115, no. 1 (2007): 43, 44–45.

39 Benjamin N. Cardozo, The Nature of the Judicial Process (New Haven: Yale University Press, 1921), 177.

40 Gennaioli and Shleifer offer economic models that interpret the Cardozo Theorem and attempt to supply mechanisms showing evolution of the common law towards informational efficiency. For citations, and for my (qualified) skepticism about such models, especially as to constitutional adjudication, see Vermeule, Law and the Limits of Reason, 115 et seq.

41 Adrian Vermeule, “‘According to Truth,’” The Josias, July 19, 2018.

42 The Letters and Diaries of John Henry Newman, vol. 25, The Vatican Council: January 1870 to December 1871, ed. Charles Stephen Dessain, CO, and Thomas Gornall, SJ (Oxford: Clarendon Press, 1973), 310.

43 Newman, Apologia Pro Vita Sua, 277.

44 Douglas W. Rae, “Counting the Fingers of Adam Smith’s Invisible Hand,” April 5, 2011, in Capitalism: Success, Crisis and Reform, Open Yale Courses, podcast, MP3 audio, 45:42.

45 As argued at length in Warren F. Samuels, Erasing the Invisible Hand: Essays on an Elusive and Misused Concept in Economics, with the assistance of Marianne F. Johnson and William H. Perry (Cambridge: Cambridge University Press, 2011).

46 Some material in these paragraphs is adapted from Vermeule, “Local and Global Knowledge in the Administrative State,” chap. 13 in Law, Liberty and State: Oakeshott, Hayek and Schmitt on the Rule of Law, ed. David Dyzenhaus and Thomas Poole (Cambridge: Cambridge University Press, 2015).

47 Sanford J. Grossman and Joseph E. Stiglitz, “On the Impossibility of Informationally Efficient Markets,” American Economic Review 70, no. 3 (June 1980): 393–408.

48 To be clear, Schauer also considers democratic and expressive rationales for free speech protection (see Schauer, Free Speech, chaps. 3–5). Those are outside the scope of my concerns here, although I believe that the argument from democracy at least is subject to parallel criticisms.

49 Daryl J. Levinson and Richard H. Pildes, “Separation of Parties, Not Powers,” Harvard Law Review 119, no. 8 (June 2006): 2311–86.

50 Curtis A. Bradley and Trevor W. Morrison, “Historical Gloss and the Separation of Powers,” Harvard Law Review 126, no. 2 (December 2012):

51 Sometimes this story is told through an emphasis on political parties, spanning institutions, as the mechanisms that produce cooperation rather than conflict. See, e.g., Levinson and Pildes, “Separation of Parties.” But this isn’t essential to the point.

52 Hayek, “Competition as a Discovery Procedure,” 9–10 (italics in the original).

53 David Luban, Lawyers and Justice: An Ethical Study (Princeton: Princeton University Press, 1988), 68.

54 I don’t say “will necessarily persist” for methodological reasons, discussed in Vermeule, “Some Confusions about ‘Classical Liberalism,’ Progressivism, and Necessity,” Mirror of Justice (blog), June 15, 2018.

55 A classic discussion, both of this class of hypotheses and their opposite, is Albert O. Hirschman, Rival Views of Market Society and Other Recent Essays (Cambridge: Harvard University Press, 1986).

56 Vermeule, System of the Constitution, 81. The classic treatment is Kenneth J. Arrow, “Social Responsibility and Economic Efficiency,” Public Policy 21 (1973): 303–18.

57 Nancy L. Rosenblum, On the Side of the Angels: An Appreciation of Parties and Partisanship (Princeton: Princeton University Press, 2008).

58 Hirschman, Rival Views.

59 For an overview, see Dennis C. Mueller, Public Choice III, 3rd ed. (Cambridge: Cambridge University Press, 2003).

60 Fleischacker, 911–12.

61 For this distinction, see Amartya Sen, foreword to Albert O. Hirschman, The Passions and the Interests, twentieth anniversary ed. (Princeton: Princeton University Press, 2013), xvi.

62 Strictly speaking, in the conventional definition, the weak form of Pareto requires only that at least one person be made better off and no one worse off.

63 Patrick J. Deneen, Why Liberalism Failed (New Haven: Yale University Press, 2018); see Vermeule, “Integration from Within,” American Affairs 2, no. 1 (Spring 2018): 44–55, endorsing Deneen’s diagnosis but not his prescriptions.

64 See, e.g., Thomas Aquinas, Summa 1.22; Joseph de Maistre, St Petersburg Dialogues; or, Conversations on the Temporal Government of Providence, trans. Richard A. Lebrun (Montreal: McGill-Queen’s University Press, 1993); Reginald Garrigou-Lagrange, Providence (St. Louis: Herder, 1937).

65 On display in, e.g. Ullmann-Margalit, “The Invisible Hand.”

66 Subject as always to the caveat that an invisible hand system may itself be deliberately put into place, as we have discussed above (“The Invisible Hand: Analytics”).

Sorry, PDF downloads are available to subscribers only.
Subscribe Now
Already a subscriber? Log In