Is Antitrust's Consumer Welfare Principle Imperiled?

• Author: Hovenkamp, Herbert

1. INTRODUCTION 65 II. THE "WELFARE TRADEOFF" 68 A. The Importance of the Consumer Welfare Test to Antitrust Policy 73 B. Finding the "Consumer" in Consumer Welfare; Labor and Other Suppliers 77 C. Measuring Competitive Harm to Suppliers, Including Labor 79 III. ANTITRUST'S LEFT FLANK--REVIVING OLD DEBATES 81 IV. CONCLUSION: TRADING OFF CONSUMER WELFARE 92 I. INTRODUCTION

   Forty years ago, Robert H. Bork published The Antitrust Paradox, which argued forcefully that antitrust policy should be driven by a principle he named "consumer welfare." (1) Bork did not use the term "consumer welfare" in the same way that most people use it today. For Bork, "consumer welfare" referred to the sum of the welfare, or surplus, enjoyed by both consumers and producers. Bork referred to consumer welfare as "merely another term for the wealth of the nation." (2) A large part of the welfare that emerges from Bork's model accrues to producers rather than consumers.

   When economists speak of "welfare," they typically mean Pareto efficiency, Kaldor-Hicks efficiency, total surplus, or some closely related concept of "general" welfare. (3) What these concepts share is that welfare includes the surplus, or wealth net of costs, enjoyed by everyone affected, including producers and consumers as well as others. For example, under Kaldor-Hicks efficiency, sometimes called potential Pareto efficiency, a move is efficient if all gainers gain enough to compensate all losers fully, leaving them indifferent. (4) Actual compensation is not required, but only that the gains be sufficiently large to produce compensation necessary to make everyone either a winner or unharmed. Bork essentially adopted a version of this conception of welfare, except that he misnamed it "consumer welfare."

   By contrast, under the consumer welfare ("CW") principle, as most people understand it today antitrust policy encourages markets to produce output as high as is consistent with sustainable competition, and prices that are accordingly as low. Such a policy does not protect every interest group. For example, it opposes the interests of cartels or other competition-limiting associations who profit from lower output and higher prices. It also runs contrary to the interests of less competitive firms that need higher prices in order to survive.

   Nor does any antitrust goal that maximizes output necessarily satisfy the Pareto principle, which is consistent with the model of perfect competition. The CW principle favors those interests, including consumers, labor and other suppliers, who profit from higher output. It disfavors those who profit from reduced output, even if those who gain from reducing output gain more than the losers lose.

   As a result, if total welfare is to be regarded as the baseline, the CW principle redistributes a certain amount of wealth away from producers and toward consumers. Significantly, however, it does not overtly distribute wealth from wealthy to poor, from employed to unemployed, from capital to labor, or along some other axis that we traditionally associate with redistributive policies. Further, the affected classes--producers, consumers, and labor--are very broad. Everyone who purchases is a consumer, and everyone who contributes something to the economy is a producer, including producers of labor.

   In the perfect competition model, producer gains are competed away over the long run and end up benefitting consumers. (5) This is one of the reasons Bork was at ease with a model that favored producers so strongly. (6) In the economy we actually have, however, that process does not always occur very quickly and may never occur at all. For example, structurally oligopolistic markets produce excessive returns that should induce new entry. In that case prices would be driven back to cost. Many markets have proven quite resistant to new entry, however, even as the firms in them obtain high returns. These persistent suboptimal structures and the practices that facilitate them justify antitrust intervention and make market structure an important factor in antitrust policy. (7)

   John Maynard Keynes put the issue famously and bluntly: "in the long run we are all dead." His context, which is usually omitted, bears quoting:

   [T]his long run is a misleading guide to current affairs. In the long run we are all dead. Economists set themselves too easy, too useless a task if in tempestuous seasons they can only tell us that when the storm is long past the ocean is flat again. (8) That, in a nutshell, was Bork's problem. His excessive confidence in perfect competition and the long run directed antitrust policy toward a useless task.

   Antitrust's CW principle is best regarded as taking a "middle run" approach to markets, reacting aggressively to unambiguous harms such as naked price fixing, and more circumspectly to single-firm conduct or other practices that have a significant potential to benefit consumers. The overall goal is clear, however, which is to encourage markets in which output, measured by quantity, quality, or innovation, is as large as possible consistent with sustainable competition. To the extent antitrust intervention furthers this goal it is justified on purely economic grounds.

   Antitrust policy under the consumer welfare principle is currently navigating between two hazards at opposite ends of the ideological spectrum. What these two hazards share in common is that both denigrate the importance of high output and low prices as an antitrust goal. On the right, Bork's general welfare approach would permit efficiency claims as an antitrust defense even when specific efficiencies cannot be proven and the challenged practice leads to reduced output and higher prices that cause consumer harm. On the left is an emergent "neo-Brandeisian" approach that often regards low prices as the enemy, at least when they come from large firms at the expense of higher cost rivals. The neo-Brandeisian approach is also redistributive, tending to redistribute wealth from larger to smaller firms, particularly when larger firms have lower costs. It also redistributes wealth away from consumers and toward these smaller producers.

   The full story is more complex. First of all, few antitrust outcomes have depended on the choice of a welfare test. Much more significant were the ways Bork credited evidence of competitive harm and offsetting efficiencies. He believed that most practices challenged under the antitrust laws produced cost savings or other efficiencies. With the exception of naked price fixing, he also doubted that these practices caused genuine competitive harm. (9) He also argued forcefully that efficiencies are not susceptible to individual proof. Rather, they must simply be assumed. The impact of this position is dramatic. As soon as efficiencies must be proven, efficiency claims become far more tenuous. (10)

   By contrast, a central claim of the neo-Brandeis approach is that markets are fragile, presenting numerous threats of collusion or monopoly. Further, antitrust policy should be driven more by political theory rather than economics. While political voices are diverse, making it difficult to identify a single theme, one clear consequence is greater protection for small business, nearly always at consumers' expense.

   To date, the strongest and most central claim of the neo-Brandeis movement remains untested; that is its assumption that individuals in our society would really be better off in a world characterized by higher prices but smaller firms. Everyone in society is a consumer and consumers vote mainly with their purchasing choices. The neo-Brandeisians still face the formidable task of providing evidence that most citizens believe they would be better off in a world of higher cost smaller firms selling at higher prices, their market behavior notwithstanding. One problem is that these costs have never been calculated, and another is that they have never been effectively communicated. Further, the neo-Brandeis movement at this writing has not provided much in the way of a calculus for determining how these goals should be applied to specific practices, other than highly general ones of the nature that Amazon should be regulated in some fashion. (11)

2. THE "WELFARE TRADEOFF"

   Bork's idiosyncratic definition of "consumer welfare" as including producer profits launched a significant debate about economic welfare tests as goals of antitrust policy. Those favoring a general welfare test believe that antitrust should seek only to maximize aggregate wealth. (12) While such views come close to Bork's, they are not identical: Bork's concept of "consumer welfare" included the sum of welfare enjoyed by producers and consumers, but he paid little attention to the welfare effects on third parties.

   In contrast, consumer welfare focuses entirely on output and, correspondingly, low prices. If consumers lose from a practice, then it is counted as anticompetitive, even if the consumer losses are completely offset by producer gains. In the classic example, suppose a merger of two large firms creates significant market power, raising prices by $1000. This merger also produces savings in production costs of $1200. In this case, producer gains from productive efficiency exceed consumer losses. This merger would be approved under Bork's standard because it produces net gains. It would be unlawful under a consumer welfare standard, however, because it produces lower output and actual consumer losses. The most salient characteristic of this merger analyzed under a consumer welfare test is firm output goes down as a consequence of the merger, and prices accordingly go up.

   These alternative welfare tests have become a kind of holy grail for mainstream antitrust thought today. One advantage claimed for them is they promise antitrust solutions that are free of excessive ideology or bias induced by special interests. They perform as a sort of analogue to the perfectly...