Thoma, Meet Hayek

The argument in favor of competition does not rest on the conditions that would exist if it were perfect. Although, where the objective facts would make it possible for competition to approach perfection, this would also secure the most effective use of resources, and, although there is therefore every case for removing human obstacles to competition, this does not mean that competition does not also bring about as effective a use of resources as can be brought about by any known means where in the nature of the case it must be imperfect… The practical lesson of all this, I think, is that we should worry much less about whether competition in a given case is perfect and worry much more whether there is competition at all. What our theoretical models of separate industries conceal is that in practice a much bigger gulf divides competition from no competition than perfect from imperfect competition.

— Friedrich A. Hayek, “The Meaning of Competition,” Individualism and Economic Order (Auburn: Ludwig von Mises Institute, 2009), pp. 104–105.

When Austrians complain that several of their neoclassical peers have not internalized some of the more “obvious” lessons of market process theory, the typical response is to deny the charge. At times, the defense is right and Austrians are wrong. But, not always. Mark Thoma proves that the Austrians are right to complain. This post, being from 2007, may not capture the author’s current beliefs, but I believe it productive to challenge it anyways.1 Thoma writes that only if markets approximate perfect competition are they really efficient and only then can we say that government is unnecessary — if you don’t believe me, read the very last sentence in his post. I contend that he’s palpably wrong on both accounts. While I think Thoma is privy to the points I’m about to make, his readers ought to take them into consideration when analyzing Thoma’s comments on the role of markets.

Thoma explains some of the conditions for markets to work their “magical efficiency.” There needs to be perfect information. There needs to be “numerous” buyers and sellers (all agents must be price-takers). Output must be homogenous. Finally, there must be free entry and free exit. He also lists a number of market failures. These latter phenomena will be discussed below, but the rest of them are necessary conditions of perfectly competitive equilibrium. They are not conditions, however, for markets to work well, and to work better than whatever alternative we may have for resource allocation.

Before anything, allow me to quickly discuss the contention that if markets are perfectly competitive we can essentially live without governance. This claim is as erroneous as the one that markets should not be held in such high esteem if they are not perfectly competitive. The market represents a web of institutions, and with institutions comes governance, whether this governance is private or collective. Without the institutions that humans have developed for thousands of years, most of which are private, not only would our markets be less competitive than they are today, but we would be as poor as we were before the eve of civilization. What are institutions? Simply, they are the rules of the game. They help decide the constraints on human choice, and therefore help decide the direction of resource allocation. So, no, perfect competition does not mean that markets trump governance; the conditions of perfect competition include certain rules (constraints) that agents follow (which, in price theory, are abstracted away from).

Now, how important are the preconditions for perfect competition? When judging the true merits of the market, the answer is “not very.” Take the notion of perfect information, for example. The advantage of the market is not that it efficiently allocates resources if there is perfect information. The advantage of the market is in its distribution of information; that is, it’s the search and discovery features of markets that makes one facet of their superiority over rival forms of rationing. Take, for example, Ludwig von Mises’ critique of socialism.2 While Mises was looking to disprove the viability of common ownership of the means of production, he was also looking to make a positive contribution in theorizing on the process of price formation and information dissemination. The problem he faced was that there was no good theory of price imputation (i.e. derived demand and input prices).3 His answer was that there must be a competitive bidding process, based on the institution of private property, for the prices of the means of production to emerge. Only by these means can the information necessary for the relatively efficient allocation of inputs be known.

Hayek later expanded on Mises’ lesson — in “Economics and Knowledge” (1937) and, later, “The Use of Knowledge in Society” (1945) —, making a more general application. In these two articles, Hayek made the case for the market’s role in the distribution of knowledge. I think more important than that, he explained how markets economize on information, allowing a relatively efficient allocation of resources without a complete set of knowledge. In this way, for example, an automobile manufacturer doesn’t need to know why the price of steel has fallen or risen, but what changes in price do tell him is sufficient for him to adjust his production in a way equivalent to that had that manufacturer had more complete knowledge of market changes. Further, Hayek wrote that prices are not the only means of information communication, but it’s through exchange that this information becomes disseminated. Finally, markets are as good at this function as they are because information is decentralized and so are markets, whereas a central authority does not enjoy these benefits.

When George Akerlof, in “The Market for ‘Lemons’,” revisits the problem of information asymmetry and its welfare implications, he notes some of the solutions that dynamic markets bring about in response to information-related market failures. Part IV of the paper (pp. 499–500), titled “Counteracting Institutions,” lists: guarantees (to buyers; e.g. warranties), brand-names, and licensing practices. (Recall the importance of institutions in pushing markets to approach the efficiency of theoretical, abstract models.) These things aren’t products of perfect competition, but market solutions to market failures that come about through the process of competition.

One of Akerlof’s examples of solutions to information asymmetry, brand-names, also speaks to Thoma’s reference to the condition of homogenous products. In price theory, if products are heterogeneous, meaning that related products are imperfect substitutes (e.g. Coca-Cola v. Pepsi), this is evidence of imperfect competition. My intention is not to criticize neoclassical price theory, but to draw attention to the story that these models are less able to tell — although, remember, one of the tenets of monopolistic competition is that the existence of profit will tend to bring about entrants into those industries.4 In the real world, product discrimination and product diversification is evidence of competition, not evidence of a lack of competition.

A interesting example is that of free banking.5 I understand that free banking theory is not without controversy and weak points, but put these aside for the sake of making an example of the positive attributes of “monopolistic competition.” In this case, note brand discrimination — implying heterogeneous, competing monies — is a crucial constraint on banking. In fact, it is the basis of the important institution represented by inter-bank clearing operations. This feature of imperfect substitutes is what puts pressure on banks to restrict money expansion: superfluous notes, over-and-above the demand for money, will circulate back to bank(s) of origin, who will then have to offer some asset(s), as dictated by the contract implicit in the note (its redeemability). Contrast this with a modern fiat currency system, where the money supply is monopolized. This is a perfect illustration of when a homogenous product implies a more monopolized market than the alternative with imperfect substitutes.

The points raised so far reminds me of a comment Ronald Coase makes in his introductory chapter to The Firm, the Market, and the Law (Chicago: University of Chicago Press, 1988),6

All exchanges regulate in great detail the activities of those who trade in these markets (the time at which transactions can be made, what can be traded, the responsibilities of the parties, the terms of settlement, etc.), and they all provide machinery for the settlement of disputes and impose sanctions against those who infringe the rules of the exchange. It is not without significance that these exchanges, often used by economists as examples of a perfect market and perfect competition, are markets in which transactions are highly regulated (and this quite apart from any government regulation there may be). It suggests, I think correctly, that for anything approaching perfect competition to exist, an intricate system of rules and regulations would normally be needed. Economists observing the regulations of the exchanges often assume that they represent an attempt to exercise monopoly power and aim to restrain competition. They ignore or, at any rate, fail to emphasize an alternative explanation for these regulations: that they exist in order to reduce transaction costs and therefore to increase the volume of trade.

— p. 9

In short, whereas Thoma argues that if markets don’t meet the conditions for perfect competition we should doubt their “magic,” the true “magic” of markets is that they help bring about the necessary institutions to push markets towards the theoretical — but unattainable — point of perfectly competitive equilibrium. Markets exist precisely because information is asymmetrically distributed and scarce, individuals and their products are heterogeneous, and because institutions are imperfect. Otherwise, we would no longer have markets, because we would attain a point of perfect satiation of wants. Markets are a product of imperfection; they help human society to overcome imperfection. Where some may interpret product heterogeneity, profit, et cetera as signals of insufficient competition, there’s a strong case to reinterpret them as signals of a working, competitive market process.

The argument in favor of markets is not that they are efficient in the maximal sense. Markets are not optimizing functions (rather, optimizing functions are abstract ways of helping explain markets). Not only will there always be some efficiency loss, but there will be noticeable market failures. Apart from information problems (already discussed), there exist other market failures: externalities. Often, market agents directly related to some exchange will fail to completely internalize the costs and/or benefits of the exchange. For example, a factory that produces the byproduct of contaminating smoke will negatively affect those not directly involved in the manufacturing of that product. These factories will over-produce, because they are not subject to the true costs of their actions. Alternatively, consider most roads. It is difficult to exclude people from using roads, therefore those that bear the costs will, more likely than not, fail to enjoy all the benefits, implying that they will under-invest in road production. These types of failures will probably affect the majority of goods.

Yet, real world markets don’t exist in static equilibrium. As Akerlof showed in his paper (and Coase in his “The Problems of Social Cost” [1960]), there are market solutions to market failures. This is another virtue of the dynamic market: solving problems which are crippling in static models. Externalities can be solved by side payments (Coasian solutions). When the pricing process is inadequate, there are alternative institutions, such as the legal system. Markets are much more than simply prices and direct exchange; markets are also entrepreneurs, firms, institutions, et cetera.

None of this is a priori reason to discard the role of the state. The state is another organization with its own institutions, some of which are related to those of the market. As James Buchanan and Gordon Tullock show in The Calculus of Consent, the state can be a force to solving market failures. In this sense, the state is analogous to the firm — it is an organization outside of the pricing process. But, it’s not fair to compare imperfect markets to the perfect state, because if there were really such a thing as a perfect state, markets would have no raison d’etre. Oftentimes, the state is not a good solution to public good problems: it costs resources to measure the extent of market failures, it costs resources to implement solutions, coercive governments impose external costs, and state action is always imperfect in execution. It often is the case that the “public solution” is more costly than the “market failure.” Ultimately, the extent to which the state can make a positive contribution to human welfare is an empirical question.

It’s worth mentioning some non-competitive aspects of public institutions, that makes the process of institutional improvement in the public sphere markedly inferior to the analogous process in the private sphere. This inferiority, one should note, was more obvious before the advent of widespread democracy. Our public institutions of governance have, without a doubt, improved over time. But, some of the problems with democracy are well documented: consider Bryan Caplan’s The Myth of the Rational Voter and Jeffrey Friedman’s critique7 based on the concept of radical ignorance (information problems). Yet, institutional improvements are slow coming. Public organizations often take steps to make the process less efficient and less competitive; I refer the reader, again, to The Calculus of Consent and its analysis of the benefits of log-rolling, and more ideally side-payments (the latter being the equivalent of Coasean solutions). And, of course, the efficiency of the state is severely restricted by transaction costs, often caused by asymmetries in the distribution of power.8 All of these factors should cast doubt upon state action induced by market failure.

Hopefully, the reader walks away with good reasons why one should not hold perfect competition as the standard by which to compare real world markets to. Static price theory models abstract from crucial features of the real world market process, features which are the true virtues of markets. Markets thrive because of imperfect information, heterogeneous products and people, finite resources and talent, not despite of these. I once commented that what makes the theory of comparative advantage so compelling and insightful is that it starts with assumptions of asymmetry and imperfection, something that many modern theories of markets don’t. It is the imperfect of human society that makes markets so necessary, including all the institutions that come with it, including, with high probability, the institutions that guide collective action.



1. I stumbled on this post by Thoma through his more recent “Free, Unregulated Markets are Not Always the Answer.” The timing of my post here is unfortunate. I send my condolences to Professor Thoma for his recent loss.

2. Ludwig von Mises, Socialism (Auburn: Ludwig von Mises Institute, 2009 [1951]). To avoid any misunderstandings, I’m not calling Thoma a socialist, nor do I think that his critique of markets relies on any sympathies towards socialism. I bring Mises’ contribution up only because of its relation to the problem of information asymmetry.

3. Jörg Guido Hülsmann, Mises: The Last Knight of Liberalism (Auburn: Ludwig von Mises Institute,2007), pp. 384–387.

4. Frank H. Knight makes the connection between imperfect competition, markets, and the entrepreneurial function in his short piece, “Profit and Entrepreneurial Functions” (1942).

5. See George Selgin, A Theory of Free Banking (Totowa: Rowman & Littlefield, 1988); Larry Sechrest, Free Banking: Theory, History, and a Laisez-Faire Model (Auburn: Ludwig von Mises Institute, 2008).

6. Speaking of Coase presents a great opportunity to briefly point to his argument for the existence of the firm, which is exactly that the price mechanism is imperfect and the market creates alternative solutions to resource allocation. Also from “The Nature of the Firm” (1937), is this nugget: “In the rest of this paper I shall use the term ‘entrepreneur’ to refer to the person or persons who, in a competitive system, take the place of the price mechanism in the direction of resources” (p. 36). This, in turn, necessitates referencing Israel Kirzner’s Competition and Entrepreneurship.

7. See also J. Friedman’s critique of Caplan’s book, “The Irrelevance of Economic Theory to Understanding Economic Ignorance.” An interesting byproduct of Friedman’s research is the comparison of information economizing between public and private institutions, where the latter are much superior. This ought to put the market failure of information asymmetry in new light, because it pushes us to accept that the scarcity and decentralization of knowledge is not so much a market failure, but a problem inherent in human society.

8. See Daron Acemoglu, “Why Not a Political Coase Theorem? Social Conflict, Commitments, and Politics.”

  • Roman P.

    I think that Hayek is more on target than neoclassicals. His views have the kernel of truth, while the neoclassical views on the desirability of the perfect competition are fragile: it takes MC<0 cases to flip the argument on its head, and as we know from Blinder that's about half of all firms. My humble opinion: the markets' magic is not that they efficiently allocate resources amongst some firms within some stage of production, but rather that they coordinate between the different stages of production.
    It's possible to plan the economy in its entirety (a la Barone), just not easy for practical purposes – USSR can attest to it. Market process splits this up into many little plans of different firms and then glues them up with a monetary system. The resultant system is largely stable because firms don't usually screw up their planning too much and they have buffer stocks to deal with inevitable failures of their local planning.

    • JCatalan

      I think growth in a complex economy is also a probabilistic phenomenon: the more firms there are, the more likely a successful micro plan can be found (and once it’s found it can be copied until a better one is discovered). Regarding socialism, the overbearing problem is knowing how to allocate resources, and it’s not clear how this information would become available without prices, profit and loss, and competition. This has something to do with the probabilistic approach to capitalism, but I’m not sure if anybody has followed this route in expanding the theoretical argument against socialism (I know it’s briefly touched upon in Engineering the Financial Crisis, although the main topic of that book is much narrower).

      • Roman P.

        I think that states with central planning have a higher chance to hit the high rates of growth (relatively higher, that is), though maybe market economies are more robust overall. I don’t understand the conceptual problem of allocating resources: as long as commodities are produced by the means of commodities and we know the technological maps, allocating resources ex ante for the purposes of production is theoretically doable. You set some arbitrary number of what you want to get (like, say, a hundred ships), then you know the amount of steel and machinery you need to produce, the amount of workers, their equipment and so on.
        Of course, it is the practical problems of central planning that make it impossible to implement in a pure way. Even in USSR it was far from Barone’s Ministry of Production: first it was a mix of private industry and state enterprise, with the main role played by the nationalized banking sector. Then it was an unholy mix of central planning through Gosplan, indicative planning through Gosplan and state departments and firms buying stuff from each other with funds (non-cash rubles) directly provided by the government, with a black market playing some major role.

  • Roman P.

    I think that Hayek is more on target than neoclassicals. His views have the kernel of truth, while the neoclassical views on the desirability of the perfect competition are fragile: it takes MC<0 cases to flip the argument on its head, and as we know from Blinder that's about half of all firms. My humble opinion: the markets' magic is not that they efficiently allocate resources amongst some firms within some stage of production, but rather that they coordinate between the different stages of production.
    It's possible to plan the economy in its entirety (a la Barone), just not easy for practical purposes – USSR can attest to it. Market process splits this up into many little plans of different firms and then glues them up with a monetary system. The resultant system is largely stable because firms don't usually screw up their planning too much and they have buffer stocks to deal with inevitable failures of their local planning.

  • Pingback: Quoting ‘Thoma, Meet Hayek’ by Jonathan Finegold | Who Plans Whom?

  • tr00pa

    Fantastic piece!

  • Pingback: Market Power and Entrepreneurship | Economic Thought

  • Pingback: Economists Do it With Models | Economic Thought