Below the fold is the text of a lunchtime talk that I gave today at a conference co-sponsored by the Antitrust Research Foundation and George Mason University’s Scalia School of Law’s Law & Economics Center.
Can Antitrust and Economics Be Reconciled?
Donald J. Boudreaux
– talk delivered at GMU Law for the Antitrust Research Foundation
January 19, 2018
Can antitrust and economics be reconciled? Good question.
The answer depends on the species of economics that one has in mind. If the economics in mind is neoclassical mainstream economics, the answer is ‘yes.’ If it is Austrian economics – or even classical economics – the answer is ‘no.’
I’m here to make the case that, at least when it comes to enhancing our ability to understand market competition, Austrian economics is superior to mainstream neoclassical economics.
On no topic in microeconomics does the Austrian approach differ so profoundly from that of mainstream neoclassical economics as it does on the topic of competition. And because sharp differences in understandings of “competition” – and, hence, of “monopoly” – promote different attitudes toward commercial practices and market arrangements, it’s no surprise that Austrian assessments of antitrust policy differ strikingly from typical mainstream assessments.
Unlike the great majority of economists today outside of the Austrian tradition, Austrians reject the argument that antitrust is needed to keep markets competitive, or that antitrust can be reliably used to increase markets’ competiveness, inventiveness, and efficiency. Three separate reasons justify this deep Austrian skepticism of antitrust, although any one reason standing alone would be sufficient for this policy conclusion.
The first is a reason for skepticism that even many mainstream economists will grant, namely, government will not administer antitrust free of political influences—influences that will often distort its application. I mention this public-choice point only in passing, but its practical importance looms large. Governments have a long history of granting special privileges to politically influential producers. Think of tariffs. The prospect, therefore, of government deploying its antitrust powers to grant favors to politically influential producers ought not be overlooked simply because antitrust regulation is explicitly justified on pro-consumer, pro-competition grounds. And antitrust’s actual history supplies ample reason to fear that antitrust enforcement will indeed generally subvert rather than promote competition.
Second, Austrians are deeply skeptical that even apolitical and highly intelligent government authorities can apply antitrust legislation in ways that improve the operation of markets over time. We can be thankful that the best non-Austrians are sensitive to what Richard Posner describes as “the daunting challenge of designing antitrust remedies that are effective without being anticompetitive.” But Austrians go further. They insist that the challenge is not simply daunting; it is practically impossible to meet. Government officials do not and cannot ever know enough about the countless, ever-changing, and all-important details of markets to intervene in ways that make markets more competitive and better able to satisfy consumer demands over time.
The illusion that helpful intervention is possible is perhaps conjured by mistaking models of the economy for the economy itself. Not only are the two not the same thing, but even the most useful economic models necessarily capture only a razor-thin slice of the individual adjustments and manifestations of creativity that make real-world market economies work.
Unlike, say, a visual model of the solar system, even unmistakably microeconomic models of the economy are constructed mostly of large aggregate concepts (“the price of steel” or “the Hirschman-Herfindahl Index number for the steel industry”) that very much are artifacts of the modeler’s own mind or of conventional classifications that economists or statisticians have come to accept over time. These concepts have neither the objectiveness nor the distinctiveness of the sort that both Jupiter and the sun have as the former orbits the latter.
This observation is no prelude to a call to reject economic models. Far from it. It is, however, a plea for greater appreciation of the limitations of economic models. The extraordinary abstraction from details that is required to construct useful economic models—and the frequent need to rely upon artifactual statistical constructs (such as “the HHI for the steel industry”)—results in models that necessarily ignore the countless on-going individual actions that give rise to the more aggregative phenomena featured as variables in the models.
Yet it is only at this deeply micro level that real-world individuals perceive profit opportunities and act to seize them. Analysts’ knowledge of the vast array of the countless particular facts that are at every moment the ones to which economic actors must adjust is so skimpy and abstract that we must concede that mastery of economic theory is not even remotely identical to mastery of the economy itself or of enterprise. The capacity for models to inform economists and government officials of what are the best ways for firms to meet consumer demands—and of what are the best organizational forms for markets to facilitate the maximization of consumer welfare—is far more limited than mainstream economics suggests.
A third reason for skepticism of antitrust is that markets in reality are much more robustly competitive than mainstream economic theory reveals them to be.
I devote the remainder of my remarks to explaining why this robustness is real and how mainstream scholars have been misled—and, hence, mislead—by their failure to appreciate this robustness.
For the past century, mainstream economists have defined “competition” as an equilibrium state of affairs. In the mainstream view, market competition is an outcome—or, alternatively, competition is a set of equilibrium conditions in which each seller of some given good or service maximizes its profits by producing that volume of output at which marginal cost equals price. In the Austrian view, market competition is entirely different. Competition is a process. Competition is a time-embedded complex of activities and experiments, many of which are incompatible with each other, in which entrepreneurs continually—and often quite creatively—vie to raise the net present values of their firms.
Unlike in mainstream economics in which prices greater than marginal costs, and above-normal profits, are unambiguous evidence of inefficiencies, in the Austrian view, prices above marginal costs and above-normal profits play important and positive roles. They both signal to entrepreneurs just where greater productive effort is most useful and supply incentives to redirect resources to these more-productive efforts.
If no special protections or privileges are available from the state, the competitive market process plays out exclusively in the form of entrepreneurs struggling to increase their efficiencies and to raise the attractiveness to consumers of their product offerings. No equilibrium is ever reached, but the process is continually weeding out error and experimenting with new products, as well as with new ways of producing and distributing products to consumers.
As F.A. Hayek and other Austrians have noted, no actual competition—as that term is popularly understood—occurs in perfectly competitive markets. To attract more customers, no perfectly competitive firm must cut prices, advertise, or build a better mousetrap. Each firm that manages to keep its unit cost low enough to be covered by the market price of its output can sell as much as it wants. All that each firm in perfectly competitive markets must do is to choose its level of output. And even that choice is mechanical: expand the rate of output produced per period of time up to, but not beyond, the rate at which marginal cost is made equal to the externally determined, fully known, and given market price.
Matters are similar also for consumers in perfectly competitive markets. These consumers simply are assumed to be fully – or at least always adequately – informed about prices, product quality, and product availability. Also, consumers’ demands simply are assumed somehow to prompt firms to produce that mix of outputs which optimally satisfies those demands.
There is in the theory of perfect competition no organization, agent, or process that discovers what consumers do, or might, demand. There are no active economic agents who catalyze inchoate consumer preferences into economically meaningful, concrete consumer demands. And there are certainly no agents who create, intensify, reduce, refine, distort, or otherwise change those demands.
Consumers and firms in this theory are mechanical and utterly artless computing devices. Consumers are mere vessels of utility functions that, when mixed with consumers’ incomes, are (somehow) transformed into demand schedules for various goods and services. Firms are nothing more than devices for transforming inputs into outputs that satisfy these given and fixed demands. For consumers and firms alike, then, all demand under perfect competition exists prior to market activity. Consumer preferences that give rise to demand exists prior to market activity.
Likewise for costs. By assumption, firms are fully informed about production functions and about input prices. Nothing need be discovered, for there is nothing to discover.
In short, because in a perfectly competitive world there is no error, no misinformation, or no yet-to-be exploited, or even perceived, profit opportunities, there is no entrepreneurship or economic growth. There’s not even any recognizable human activity that adjusts prices upward or downward to the equilibrium levels identified in the model. In a perfectly competitive world, all that is done by real-world entrepreneurs—even something as simple and as necssary as actually adjusting prices—is already done.
Of course, every theory is unrealistic in that each one abstracts from many features of reality in order to focus attention on those few features judged to be most relevant for the purposes at hand. Unrealism, in this sense, is indispensable to any useful theory. The theory of perfect competition is no exception. As a tool for sharpening our insight into the likely consequences of certain exogenous changes—for example, how increases in consumer demand for soybeans affect the price and output of soybeans if soybeans are sold by many different sellers—this theory works well enough.
Trouble arises, instead, out of a confusion borne largely of its name. As the great Harold Demsetz and a few others have observed, the theory of perfect competition is not a theory of competition. We learn nothing from this theory about how firms actually compete. Beyond the positive relationship between the number of competitors and the intensity of price competition—a relationship that is simply assumed rather than demonstrated—this model is silent about the kinds of competitive activities that real-world firms might practice under different market conditions. And it sheds no light on the many different modes of competition that we actually do observe in reality except that none of these modes will ever happen under the conditions assumed in the theory of perfect competition.
No competitive activities of the sort that we routinely observe in the real world occur in the model that economists – the experts! – call perfect competition. From the confines of that model, activities such as advertising, price discrimination, and product differentiation are naturally viewed as suspect—as evidence either of existing monopoly power or of attempts to secure monopoly power. What else can such activities be from the perspective of the model of perfect competition?
In perfectly competitive markets, consumer welfare by assumption is maximized because firms are price takers that secure maximum profit exclusively by adjusting their rates of output in response to observed exogenous changes in the market prices or costs of production that these firms confront. Any other activities violate the conditions of perfect competition. No other activity need be done because by assumption everything that needs to be known about the market is already known.
From this neoclassical vantage point it’s a short leap to the conclusion that the real world is chock-full of markets that are not ideally competitive. The verdict seems clear: real-world markets, because they differ so starkly from perfectly competitive markets, are infused with elements of monopoly power and, hence, generate imperfect outcomes that are at least potentially correctable by policies that make real-world markets more closely resemble perfectly competitive ones.
Theories of imperfect competition do nothing to change this assessment. The standard – the ideal – remains the equilibrium outcome generated under “perfect competition” – most importantly, of course, maximum price competition such that price equals marginal cost.
The assumptions and logic of neoclassical economics do indeed carve out a role for antitrust.
Austrians reject this neoclassical theorizing about competition. They do so not because such theorizing abstracts from some features of real-world markets, but because it abstracts from the very features of real-world markets that are most in need of being explained by any theory of competition. In the Austrian understanding, the discovery of consumer demands—discovery not only by producers, but also by consumers themselves—is an important function of real-world markets. Likewise the discovery of lower-cost methods of production. Likewise the discovery of information about often rapidly changing prices, product qualities, and availabilities of products and inputs. Likewise the potential for producing and selling entirely new products. And, importantly, likewise the discovery of ‘optimal’ firm sizes, of ‘optimal’ industry structures, of optimal contractual arrangements, and of optimal pricing policies.
In reality, none of this knowledge is ever given or fixed. It must be discovered. And competition—real-world competition, the struggle among producers to increase their profits by better appealing to consumers, or by lowering their costs, in any ways that they can—is chiefly a process of such discovery.
Clearly, this entrepreneurial discovery process differs greatly from the “competition” that occurs in neoclassical models. In Austrian accounts, types and qualities of outputs are never given. Nor are demands. Nor are prices. Nor is knowledge of prices. Nor are production functions, costs, and knowledge of input availability. And nor are firm sizes and industry structures. These economic phenomena are all understood to be, at least in large part, discovered—or even created by—entrepreneurial actions of the sort that are assumed away in the model of perfect competition.
As I suggested a moment ago, one result of this difference is that many real-world activities that either do not occur in the theory of perfect competition or that are plainly at odds with the assumptions of that theory are, in the Austrian view, revealed as being at least potentially pro-competitive. Many of the “monopolistic” elements or “imperfections” that mainstream economists see in real-world markets are, through Austrian lenses, seen as manifestations of well-functioning and creative competition.
For Austrians, competitive markets exist as long as there are no artificial barriers to production and exchange. The range of actions available to entrepreneurs and consumers in a market is, in fact, open-ended, and therefore the range of observed arrangements and market ‘outcomes’ that are consistent with competition is also open-ended. What might well appear to an economist trained only in mainstream models to be evidence of monopoly power is perhaps, in reality, evidence of the market’s creative way of groping toward greater efficiencies.
Obviously, a question is raised by the statement “competitive markets exist as long as there are no artificial barriers to production and exchange.” What, exactly, is meant by “artificial barriers”? The answer—or, my answer—is a barrier to production and exchange is artificial if it blocks or penalizes an economic activity that is consistent with the common law of property, contract, and tort. Such barriers include, but aren’t limited to, legislation and regulation that gives differential advantage to a particular product, person, firm, industry, or region.
In the absence of government favoritism, and of activities proscribed by the common law – activities such as vandalism, theft, and breach of contract – a firm can over the long-run increase its profits only by achieving greater efficiencies in production and distribution, or by enhancing the attractiveness of its product in the minds of consumers. The resulting continual and creative struggle among entrepreneurs for maximum profits will result in a variety of experiments—some successful, some not—across the spectrum of possible ways to organize firms and industries. For Austrians, the test of whether or not markets are competitive is not how well markets conform to some external criteria imposed by economists, courts, legislators, or government administrators. Rather, that test is whether or not artificial barriers exist. Period.
The mainstream economist objects, along with advocates of active antitrust enforcement: “This Austrian definition of ‘artificial barriers’ simply assumes away the possibility that such barriers can arise in free markets.” But this objection misses the larger picture. Economic theory— mainstream as well as Austrian—has at its foundation the assumption that entrepreneurs are forever searching for opportunities to earn profits as large as possible and that consumers are forever searching for opportunities to increase their utility as much as possible. With these plausible assumptions, economists readily recognize that, say, an unexpected increase in the demand for bananas relative to that for papayas will prompt producers to supply fewer papayas and more bananas. In this case no one frets that consumers will not be supplied with more bananas, or that consumers will long pay exorbitant prices for bananas. Freedom to enter the banana market is recognized as sufficient to ensure that the current higher-than- ”competitive” price for bananas (and the current lower-than-”competitive” quantity of bananas supplied) will be corrected by the increased production of bananas.
Importantly, in such a case as that of an increased demand for bananas, no mainstream economist worries if told that the production functions of each existing banana producer make it unprofitable for these producers to increase their production of bananas even at the higher price of bananas. The reasonable assumption in this event is that, absent government-imposed restrictions on entry into the banana industry, the higher demand for bananas would then be met exclusively by new entrants.
This same set of assumptions and train of reasoning that apply in the case of an exogenous increase in consumers’ demand for bananas should apply also in the case of the most widely condemned violations of antitrust legislation, including horizontal collusion and horizontal mergers. If in the absence of government-imposed restrictions on entry a rise in the price of bananas caused by increased consumer demand for bananas attracts new entrants into the banana market, why suppose that in the absence of government-imposed restrictions on entry horizontal collusion or horizontal mergers among banana producers that restrict output and raise prices to ‘monopoly’ levels will not do the very same—namely, attract new competitors? No good reason exists.
It will not do to insist that a relevant distinction is found in the fact that horizontal collusion or consolidation is initiated by producers while rising consumer demand (as in the banana example) is not. If new entry is assumed possible and effective when the latter occurs, it must be assumed possible and effective when the former occurs. In both cases, new entrants are self-interestedly seeking profits in response to current patterns of prices above costs and not in response to the motives or reasons that give rise to these patterns.
Nor will it do to assert that, unlike rising consumer demand, horizontal agreements that restrict output and raise prices serve no legitimate economic function and, therefore, ought not be tolerated.
First, even some neoclassical scholars have identified plausible situations in which successful collusion promotes consumer welfare over the long run. Second and more importantly, as long as there are no government-erected barriers to entry, the most reliable test for what arrangements best promote long-run consumer welfare is the market test. If a particular horizontal arrangement survives in the face of entry, or the possibility of entry, we are not scientifically entitled to assume that that arrangement is undesirable. Our presumption must run in the other direction. The same intellectual humility that obliges us to regard the continuing supply of vanilla ice cream, in a market free of artificial barriers, as serving consumers’ best interests also obliges us to regard a horizontal merger or even a successful collusive agreement, in a market free of artificial barriers, as serving consumers’ best interests.
Put differently, the potential for entry that is free of obstruction by artificial barriers is sufficient to discipline producers to continually experiment with organizational arrangements and contractual practices that improve their abilities to serve consumers. Arrangements and practices that serve consumer interests poorly relative to other arrangements and practices will over time be displaced in competition with those other arrangements and practices.
To most economists and legal scholars this Austrian position seems extreme. When the complexity and dynamism of the economy is reckoned properly, however, the Austrian position is seen to be more realistic than the position staked out by the mainstream. What F.A. Hayek famously called “the particular circumstances of time and place” are all-important. It is these details that must be reckoned with, moment to moment without end, by people on the spot. It is these details that furnish hints only to close-in observers—only to those people on the spot—for how resources might be reallocated, or how organizational forms and practices might be altered, to generate more profit.
Using antitrust legislation to prevent experimentation with organizational forms and practices short-circuits competition among organizational forms and practices. Even if antitrust enforcement results in more intense price competition – and, hence, in lower prices – it will do so by weakening other forms of competition and likely over time causing consumer welfare to be lower than it would otherwise have been.