Dispersion of Power as an Economic Goal of Antitrust Policy

Abstract

This chapter surveys the range of objectives ascribed to US antitrust policy from its formative period through the early 1970s. Martin discusses and rejects Robert Bork’s analysis of the legislative intent behind the Sherman Act and considers the Kaldor-Hicks potential Pareto improvement principle, a central element of Bork’s argument. An elementary model shows that social preferences about aspects of market performance not captured by consumer surplus or net social welfare can be included in standard economic models. He further argues that the role of economics as a science in analyzing market performance is limited to characterizing the costs and benefits of pursuing alternative policy objectives and that economics as a science is agnostic concerning what policy goals should be.

These keywords were added by machine and not by the authors. This process is experimental and the keywords may be updated as the learning algorithm improves.

1 Introduction

The Sherman Antitrust Act became part of US law in 1890. It remains the basis of US antitrust policy and has influenced the development of competition policies around the world.

During the 20 years before passage of the Sherman Act, there was broad public debate in the United States about the rise of large business and the appropriate policy reaction to that rise. A wide range of views were expressed, and this range of views was reflected during Senate debate about the Sherman Act.

At first, the new law was ineffective, and public discussion of the trust issue continued. In 1914, the United States adopted two additional pieces of legislation, the Clayton Antitrust Act and the Federal Trade Commission Act, to complement the Sherman Act.

By the mid-1940s, a mainstream consensus had emerged that US antitrust policy aimed to pursue both economic and social goals. Economic goals included competitive market performance, high rates of innovation, and productivity growth. Social goals included the dispersion of economic and other kinds of power and fairness in market processes. It was thought that these goals were mutually consistent and that policies adopted to promote one of them would promote the others as well.

In its formative phase, US antitrust pursued these goals by promoting competition. Section 1 of the Sherman Act promoted competition among active firms by prohibiting agreements not to compete. Section 2 of the Sherman Act promoted potential competition by prohibiting monopolization, which today an economist would describe as the erection of strategic barriers to entry to achieve or maintain monopoly¹ power. The antitrust laws did not object to monopoly that resulted from competition on the merits.²

The original Section 7 of the Clayton Act prohibited mergers carried out by acquisition of shares of stock, “where in any line of commerce …the effect of such acquisition may be substantially to lessen competition, or to tend to create a monopoly”. The legislation was aimed at “the development of holding companies and at the secret acquisition of competitors through the purchase of all or parts of such competitors’ stock”.³ The Congress that passed the Clayton Act in 1914 intentionally excluded mergers carried out by acquisition of the assets of one company by another company from the coverage of Section 7. Such mergers would be public knowledge. As long as the fact of a merger was known, if it created profit opportunities for existing or potential rivals, competition from those rivals would get the best market performance possible. This passive publicity approach avoided any direct government control of market structure.

In 1950, the Celler-Kefauver amendment to Section 7 of the Clayton Act abandoned the passive approach and prohibited mergers carried out by asset acquisition, where “the effect of such acquisition may be substantially to lessen competition”. The extension of Section 7 to asset acquisitions and the requirement to assess the future impact of a merger on competition made economics central to the application of merger policy. With time the enhanced role of economics spread to other areas of antitrust.

In the interwar period, University of Chicago economists, like economists generally, supported antitrust policy as a bulwark against regulation. The 1950 transition from passive to active merger policy accelerated a change, underway since the late 1940s, in the views of Chicago economists toward antitrust policy. From the mid-1950s onward, lawyers and economists working in the Chicago tradition were harshly critical of received antitrust policy.

One element of this criticism was Robert Bork’s rereading and reinterpretation of congressional intent toward the Sherman Act. Bork’s conclusion (1966, p. 11) was that “since the legislative history of the Sherman Act shows consumer welfare to be the decisive value it should be treated by a court as the only value”. It is now understood that when Bork wrote of “consumer welfare”, what he meant was what economists call net surplus, the sum of consumer surplus and producer surplus.

Bork’s analysis of congressional intent is associated with the rise of the so-called “economic approach” to antitrust, which holds that antitrust policy should concern itself with and only with practices that reduce surplus in one or the other of the two meanings given to the term “consumer welfare” and that other alleged purposes of antitrust policy are ruled out, on the ground that they are not “economic”.⁴

But if aggregate welfare depends on market outcomes for which there are no markets, then an antitrust policy that maximizes consumer surplus or minimizes deadweight loss is inefficient in an economic sense, because it ignores the economic consequences of missing markets. I show that social preferences about aspects of market performance not captured by consumer surplus or net surplus can be included in standard economic models. I use preferences about market structure as an example, but the point is general and applies, for example, to preferences about net neutrality, about political contributions by businesses, about genetically modified organisms, about carbon emissions, and other externalities resulting from market activities.

I further argue that the role of economics in analyzing market performance is limited to characterizing the costs and benefits of pursuing alternative policy objectives and that economics as a science is agnostic concerning what policy goals should be.

2 Related Literature

This chapter is related to a large body of work by legal and economic scholars. References to parts of that literature are given at appropriate points throughout the chapter. Here I mention three papers that make contributions particularly related to the conclusions I draw.

Adams et al. (1991) examine the efficiency of equilibrium outcomes in a two-good general equilibrium framework. If both sectors are perfectly competitive, the equilibrium outcome is Pareto optimal, and efficient in the Pareto sense.⁵ Pareto optimality fails if one or both markets is imperfectly competitive or in the presence of joint production by a firm with market power. Whether or not a merger that supports market power improves welfare depends on whether gains in production efficiency, if any, outweigh reductions in consumer welfare.⁶ They also make the point that a firm with market power may profitably choose not to produce some products that consumers would pay for, if the goods were available at competitive prices. This is a missing-market issue of the kind I take up in Sect. 10.8.

Brock and Obst (2009) incorporate preferences about market concentration in a general equilibrium model. I incorporate preferences about market concentration in a partial-equilibrium model (Sect. 10.8). A partial equilibrium version of their result, that a full welfare optimum requires equality between the marginal loss of utility due to reduced output and the marginal gain of utility from decreased concentration, holds in my model.

Hovenkamp (1982) examines the argument that antitrust policy should have the maximization of efficiency as its unique goal, using three alternative efficiency standards (maximization of consumer welfare, Pareto optimality, and wealth maximization). His conclusion (1982, p. 30) that “Antitrust policy must come to grips with the fact that people may sometimes be willing to pay higher consumer prices to realize certain values, and that these values cannot always be determined in the voluntary market” is an implication of the model I develop in Sect. 10.8. His view (1982, pp. 28–29) that economics can inform the law by determining the costs of alternative policies is much the same as the position I take in Sect. 10.9.

3 The Formative Era

3.1 Economic Changes

The US economy was fundamentally transformed over the 25 years following the end of the American Civil War. Railroads, themselves the first firms to operate at national scale (Chandler 1965), spanned the continent and made possible a single national market. In sectors of the economy that involved economies of large-scale production, markets came to be dominated by large firms that achieved low unit cost and used railroad transportation to supply vast geographic areas.⁷ Farmers, small firms that found themselves competing with large firms, and employees of railroads and other large firms all took exception to the changes that followed.

Agricultural regions depended heavily on railroads, which were often local monopolists or duopolists, to ship grain and livestock to industrial food processors. Railroad rates were high in local monopoly markets, low in local oligopoly markets. Railroad collusion during economic downturns, imperfect though it may have been (Ellison 1994), was bitterly resented by farmers. Farmers were hostile to railroad rate discrimination and supported state legislation to regulate railroad rates (Farmer 1924; Miller 1954).

Local businesses similarly resented competition from distant, lower-cost large-scale firms that could profitably undersell them in their home markets. They lobbied state governments for protection against out-of-state rivals, sometimes with success (Hollander 1964; McCurdy 1978). Had such laws not been invalidated by the US Supreme Court, under the Commerce Clause of the US constitution, there would have been no national US economy, but rather a crazy-quilt of state markets. Local businesses that faced competition from large firms remained unhappy after protective legislation was struck down.

In industries supplied by large firms, workers found themselves on the supply side of monopsonistic markets for labor services. The post-Civil War period saw the rise of organized labor in the United States, a rise sometimes marked by civil unrest.

Thus the creation of a national US market brought forth interest groups-agriculture, small business, and organized labor-that criticized “the trusts”. Segments of the general public looked askance at the development of private firms that were larger, in terms of employees or income, than some state governments. When state antitrust legislation proved ineffective, aggrieved parties turned to the national government for relief.

The topic of the rise of large business was never far from the public eye. Economists, political scientists, sociologists,⁸ lawyers, businessmen,⁹ and reformers all contributed to the ongoing public dialogue about the rise of trusts.¹⁰ Discontent with the ramifications of large firms was widespread.¹¹ It cannot be said that US antitrust, in its formative period, pursued specifically economic goals.

3.2 The Progressive Movement

Enforcement of the Sherman Act ramped up slowly and was hobbled by restrictive court interpretations. The national debate on trusts and their consequences went on, unabated.

Economists were active in this debate both before and after passage of the Sherman Act. They were not enthusiastic about laissez-faire. Nor were they enthusiastic about the prospect of widespread government intervention in the economy, preferring to rely on competition for industries of decreasing or constant returns to scale and inclined toward what would now be called public utility regulation for industries of increasing returns to scale.¹²

A recurring element in the debate about trusts was the speed of entry. If the response to incumbents’ economic profits was swift, large-scale entry, incumbent firms would not be able to exercise market power, and no antitrust policy would be needed.¹³ John Bates Clark, one of the founders of the American Economic Association, came to view predatory price discrimination and restrictive manufacturer-distributor contracts as entry-deterring strategies that could allow incumbent firms to exercise persistent monopoly power.¹⁴ Clark was one of four members of a public service committee that prepared a first draft of the 1914 Clayton Act (Fiorito 2013).

Section 2 of the Clayton Act prohibits price discrimination, and Section 3 prohibits restrictive manufacturer-distributor contracts (in both cases, where the effect may be substantially to lessen competition or tend to create a monopoly). Both provisions followed Clark’s views.

The provisions of the Clayton Act are highly specific. The Federal Trade Commission Act is general: it establishes the Federal Trade Commission as an independent agency within the executive branch, and its operative provision, Section 5, prohibits unfair methods of competition.

The Sherman Act, the Clayton Act, and the Federal Trade Commission Act are the pillars upon which the superstructure of US antitrust rests. There have been amendments to these “big three” laws, one of which is discussed later, and specialized supplementary legislation. Some legislation exempts specific sectors (agricultural cooperatives, insurance) from the coverage of antitrust.¹⁵ Regulatory legislation sometimes assigns responsibility for matters usually thought of as falling in the antitrust sphere to regulatory commissions (such as the Federal Communications Commission). The Webb-Pomerene Act of 1918 and the Export Trading Act of 1982 allow firms to form joint-sales agencies-that is, to collude-for sales on foreign markets, provided there are no repercussions for competition on the US market.

For present purposes, namely examination of the range of goals ascribed to the antitrust laws, the Robinson-Patman Act of 1936 must be mentioned. It amended the Section 2 Clayton Act prohibition against price discrimination with the aim of protecting small grocery stores from the competition of national chains. The legislation was drafted by a lawyer for the United States Wholesale Grocers’ Association (Kintner 1978, p. 2895).

In its 1951 Standard Oil of Indiana decision, the Supreme Court wrote (340 U.S. 231 at 248–249, internal citations omitted, emphasis added)

The heart of our national economic policy long has been faith in the value of competition. In the Sherman and Clayton Acts, as well as in the Robinson-Patman Act, “Congress was dealing with competition, which it sought to protect, and monopoly, which it sought to prevent”. …We need not now reconcile, in its entirety, the economic theory which underlies the Robinson-Patman Act with that of the Sherman and Clayton Acts. It is enough to say that Congress did not seek by the Robinson-Patman Act either to abolish competition or so radically to curtail it that a seller would have no substantial right of self-defense against a price raid by a competitor.

In an accompanying footnote, the Supreme Court noted that “It has been suggested that, in theory, the Robinson-Patman Act as a whole is inconsistent with the Sherman and Clayton Acts”. The footnote is more to the point than the text of the opinion, and the frankly protectionist intent with which the Robinson-Patman Act was adopted cannot be reconciled with an efficiency-oriented antitrust policy.

4 From Adoption to Mid-Century

The Sherman Act and the Federal Trade Commission Act are written in terms of broad generality. The operative provisions of the Clayton Act are more specific, but still leave substantial margin for interpretation. Congress delegated to courts the task of fleshing out the framework it had established in adopting the three basic antitrust laws. In the words of Senator Sherman during Senate debate (21 Cong. Rec. 2460 (1890)):

I admit that it is difficult to define in legal language the precise line between lawful and unlawful combinations. This must be left for the courts to determine in each particular case. All that we, as lawmakers, can do is to declare general principles, and we can be assured that the courts will apply them so as to carry out the meaning of the law, as the courts of England and the United States have done for centuries.

Courts have accepted this mandate, and in rendering decisions, have expressed a range of views on the goals of the antitrust laws.¹⁶

4.1 Promote Fairness of Market Processes

In its 1897 decision finding the Trans-Missouri Freight Association in violation of Section 1 of the Sherman Act, the Supreme Court recognized that market processes could disrupt established economic relationships, creating winners and losers and severe hardship for the latter (166 U.S. 290 (1897) at 323):

In any great and extended change in the manner or method of doing business it seems to be an inevitable necessity that distress and, perhaps, ruin shall be its accompaniment in regard to some of those who were engaged in the old methods. A change from stage coaches and canal boats to railroads threw at once a large number of men out of employment; changes from hand labor to that of machinery, and from operating machinery by hand to the application of steam for such purpose, leave behind them for the time a number of men who must seek other avenues of livelihood. These are misfortunes which seem to be the necessary accompaniment of all great industrial changes. It takes time to effect a readjustment of industrial life so that those who are thrown out of their old employment, by reason of such changes …may find opportunities for labor in other departments than those to which they have been accustomed. It is a misfortune, but yet in such cases it seems to be the inevitable accompaniment of change and improvement.

The social decision to use free markets to allocate resources dictates that such disruptions be accepted, when they result from competition in free markets. There is no such presumption if disruption is the result of artificial aggregations of economic power (166 U.S. 290 (1897) at 323–324):

It is wholly different, however, when such changes are effected by combinations of capital, whose purpose in combining is to control the production or manufacture of any particular article in the market, and by such control dictate the price at which the article shall be sold, the effect being to drive out of business all the small dealers in the commodity and to render the public subject to the decision of the combination as to what price shall be paid for the article.

In this view, the Sherman Act seeks to maintain fairness in market processes in order to maintain public support for a market system of resource allocation.

4.2 Maintain Competition

In 1927, in its opinion involving a challenge to a trade-association-mentored cartel, the Supreme Court wrote that the purpose of the Sherman Act was to maintain competition (U.S. v. Trenton Potteries 273 U.S. 392 at 397):

Whether this type of restraint is reasonable or not must be judged in part at least in the light of its effect on competition, for whatever difference of opinion there may be among economists as to the social and economic desirability of an unrestrained competitive system, it cannot be doubted that the Sherman Law and the judicial decisions interpreting it are based upon the assumption that the public interest is best protected from the evils of monopoly and price control by the maintenance of competition.

4.3 Prohibit Unreasonable Restraint of Trade

In the 1933 Appalachian Coals decision concerning a proposed joint sales agreement, the Supreme Court ruled that the Sherman Act prohibited only undue restraints of trade (288 U.S. 344 at 359–360, emphasis added):

The purpose of the Sherman Anti-Trust Act is to prevent undue restraints of interstate commerce, to maintain its appropriate freedom in the public interest, to afford protection from the subversive or coercive influences of monopolistic endeavor. …

In applying this test, a close and objective scrutiny of particular conditions and purposes is necessary in each case. Realities must dominate the judgment. The mere fact that the parties to an agreement eliminate competition between themselves is not enough to condemn it.

It is often said that this decision was an aberration, reflecting the severe economic conditions of the Great Depression. Yet, the decision is cited by subsequent Courts, when they find it convenient to do so.

4.4 Prohibit All Restraints on Trade

The 1940 Socony-Vacuum case involved collusion in the US oil refining industry. The force of the resulting Supreme Court decision is that the Sherman Act rules out any agreement on prices (310 U.S. 150 at 221):

But the thrust of the [per se] rule is deeper and reaches more than monopoly power. Any combination which tampers with price structures is engaged in an unlawful activity. Even though the members of the price-fixing group were in no position to control the market, to the extent that they raised, lowered, or stabilized prices they would be directly interfering with the free play of market forces. The Act places all such schemes beyond the pale and protects that vital part of our economy against any degree of interference.

4.5 Promote Dispersal of Power

For Justice Douglas, dissenting in the 1948 Columbia Steel merger case, the Sherman Act aimed to disperse power (334 U.S. 495 at 536)¹⁷:

In final analysis, size in steel is the measure of the power of a handful of men over our economy. That power can be utilized with lightning speed. It can be benign or it can be dangerous. The philosophy of the Sherman Act is that it should not exist. For all power tends to develop into a government in itself. Power that controls the economy should be in the hands of elected representatives of the people, not in the hands of an industrial oligarchy. Industrial power should be decentralized. It should be scattered into many hands so that the fortunes of the people will not be dependent on the whim or caprice, the political prejudices, the emotional stability of a few self-appointed men. The fact that they are not vicious men but respectable and social-minded is irrelevant. That is the philosophy and the command of the Sherman Act. It is founded on a theory of hostility to the concentration in private hands of power so great that only a government of the people should have it.

4.6 Monopolization

Two opinions by Judge Learned Hand, issued almost 30 years apart, bracket the range of interpretations given to the Sherman Act Section 2 prohibition of monopolization.

4.6.1 Corn Products Refining

The Corn Products Refining Co. was a dominant firm, formed by merger, that used deferred loyalty rebates, bogus independents, and below-cost pricing to maintain its market position.¹⁸ Judge Learned Hand’s 1916 opinion in a US government monopolization case made clear that the Sherman Act had no objection to monopoly obtained by competition on the merits (U.S. v. Corn Products Refining Co. et al. 234 F. 964 at 1015)¹⁹:

The national will has not declared against elimination of competitors when they fail from their inherent industrial weakness. On the contrary, it has declared with great emphasis against any methods by which such weaknesses might be concealed; in so doing it has assumed a positive purpose toward industry, has established a norm to which competition must conform. This purpose the Corn Products Refining Company has persistently and ingeniously endeavored to thwart from the outset. Its constant effort has been to prevent competitors from that test which would in the long run discover whether they could manufacture as well and as cheaply as itself.

4.6.2 Alcoa

Judge Hand’s 1945 opinion in the Alcoa monopolization case found both economic and social goals behind the antitrust laws.²⁰ The economic goals related to the good market performance expected to result from the stimulus of actual competition or the threat of potential competition (148 F.2d 416 at 427):

Many people believe that possession of unchallenged economic power deadens initiative, discourages thrift and depresses energy; that immunity from competition is a narcotic, and rivalry is a stimulant, to industrial progress; that the spur of constant stress is necessary to counteract an inevitable disposition to let well enough alone. Such people believe that competitors, versed in the craft as no consumer can be, will be quick to detect opportunities for saving and new shifts in production, and be eager to profit by them. In any event the mere fact that a producer, having command of the domestic market, has not been able to make more than a ‘fair’ profit, is no evidence that a ‘fair’ profit could not have been made at lower prices.

Judge Hand also found broader social purposes in the Sherman Act (148 F.2d 416 at 428–429; footnote omitted; not set off as a list in the original):

• We have been speaking only of the economic reasons which forbid Monopoly, but, as we have already implied, there are others, based upon the belief that great industrial consolidations are inherently undesirable, regardless of their economic results. In the debates in Congress Senator Sherman himself …showed that among the purposes of Congress in 1890 was a desire to put an end to great aggregations of capital because of the helplessness of the individual before them.

• That Congress is still of the same mind appears in the Surplus Property Act of 1944… and the Small Business Mobilization Act… Not only does § 2(d) of the first declare it to be one aim of that statute to ‘preserve the competitive position of small business concerns,’ but § 18 is given over to directions designed to ‘preserve and strengthen’ their position. …

• Throughout the history of these statutes, it has been constantly assumed that one of their purposes was to perpetuate and preserve, for its own sake and in spite of possible cost, an organization of industry in small units which can effectively compete with each other.

Regarding the first bullet point, the remarks of Senator Sherman to which Judge Hand refers deal among other matters with public concern about inequality in the distribution of wealth that was thought to follow on the heels of trusts.²¹ In the second bullet point, Judge Hand notes that the Congress of 1944 was concerned to promote small business.²² In the third bullet point, he recognizes the implication of similar sentiments expressed in the Sherman Act debate in 1890. Those legislative sentiments are inconsistent with the view expressed in Corn Products Refining (and in Senate debate before passage of the Sherman Act) that US antitrust does not object to dominant market positions based on competition on the merits. They cannot be reconciled with the hypothesis that Congress conceived of the goals of antitrust policy exclusively in terms of the maximization of consumer surplus or the minimization of deadweight loss.

4.7 Résumé

Courts, following the congressional mandate to “carry out the law”, have imputed a variety of purposes to the Sherman Act and companion legislation. Writing after this period, Areeda described the possible goals of antitrust policy (1983, p. 534):

Let me begin by stating summarily the other possible goals of antitrust beyond maximizing consumer welfare. They include the political and social values of dispersed control over economic resources, multiple choices for producers and consumers free of the arbitrary dictates of monopolies or cartels, equal opportunity, equitable income distribution, and “fairness” in economic dealings. As a general proposition, such goals are attractive to many citizens and perhaps most of them.

It should also be obvious, however, that these very goals are widely served by that effective competition which maximizes consumer welfare.

5 Celler-Kefauver

Martin (1959) and Bok (1960) discuss the legislative history of the Celler-Kefauver Act.

The shift from the passive, publicity approach to merger control of the original Section 7 of the Clayton Act to the active antimerger policy of the 1950 Celler-Kefauver Act was motivated by concern about the social and political consequences of the rise of large firms, not its economic causes or effects²³:

The control of American business is steadily being transferred … from local communities to a few large cities in which central managers decide the policies and the fate of the far-flung enterprises they control. … Through monopolistic mergers the people are losing power to direct their own economic welfare. When they lose the power to direct their economic welfare they also lose the means to direct their political future.

In his next remarks, Senator Kefauver invoked the lessons of recent history, that the concentration of economic power in the hands of the few led to either Fascism or Socialism. In the words of Bok (1960, pp. 236–237):

To anyone used to the preoccupation of professors and administrators with the economic consequences of monopoly power, the curious aspect of the debates is the paucity of remarks having to do with the effects of concentration on prices, innovation, distribution, and efficiency. To be sure, there were allusions to the need for preserving competition. But competition appeared to possess a strong socio-political connotation which centered on the virtues of the small entrepreneur to an extent seldom duplicated in economic literature.

In its first opinion applying the amended Section 7, the Supreme Court summarized as follows congressional intent in amending Section 7 (370 U.S. 294 (1962) at 315, footnotes omitted):

The dominant theme pervading congressional consideration of the 1950 amendments was a fear of what was considered to be a rising tide of economic concentration in the American economy. …Other considerations cited in support of the bill were the desirability of retaining “local control” over industry and the protection of small businesses. Throughout the recorded discussion may be found examples of Congress’ fear not only of accelerated concentration of economic power on economic grounds, but also of the threat to other values a trend toward concentration was thought to pose.

The Supreme Court noted the explicit statement in a Senate Report that the amendment aimed to give antitrust a tool to stop a trend to increased concentration before it became irreversible (370 U.S. 294 (1962) at 317–318, emphasis added): “The intent here … is to cope with monopolistic tendencies in their incipiency and well before they have attained such effects as would justify a Sherman Act proceeding”. If a merger would not worsen market performance, but could be expected to be the first of a series of mergers that would, in total, worsen market performance, congressional intent was to block the merger.

6 Bork on Legislative Intent

From the early post-World War II period, University of Chicago economists developed views inimical to what they sometimes referred to as “collectivism” (Brown et al. 1955).²⁴ This included a reversal of Chicago economists’ interwar support for antitrust policy.^25,26 One element of this program was Bork’s analysis (1966, 1967, 1978) of congressional intent behind the Sherman Act, based on the debates reported in the Congressional Record. Bork quotes extensively from congressional debates. He gives 21 instances in which senators or representatives referred to the trust problem as one of raising price to the consumer, advancing cost to the consumer, and the like, which suggests speakers’ concern with the impact of monopoly pricing on purchasers. He gives eight instances in which speakers referred to restraint of trade or the like, which to a modern economist might suggest an interest in minimizing deadweight loss.

Much of Bork’s discussion makes sense only if “consumer welfare” is taken to mean “the welfare of purchasers”. To give one example of what could be many, writing of an antitrust rule against mergers leading to monopoly, Bork makes a clear distinction between consumer welfare and producer welfare and argued that Congress gave priority to consumer welfare over producer welfare (1966, p. 11, emphasis added):

The argument for this rule in Congress, however, shows that it derived in large measure from a desire to protect consumers from monopoly extortion. Insofar as other classes, such as small producers who sold to or bought from monopolists, were to be benefitted, that benefit was not seen as conflicting with the consumer-welfare rationale but rather as reinforcing it. Where producer and consumer welfare might come into conflict … Congress chose consumer welfare as decisive.

If “consumer welfare” means “protecting purchasers from monopoly extortion”, it does not include the welfare of the owners of firms that exact monopoly prices.

The only hint that what Bork intends by the term “consumer welfare” is the combined welfare of consumers and producers is an oblique turn of phrase in his introduction (Bork 1966, p. 7, emphasis added):

My conclusion, drawn from the evidence in the Congressional Record, is that Congress intended the courts to implement (i.e., to take into account in the decision of cases) only that value we would today call consumer welfare. To put it another way, the policy the courts were intended to apply is the maximization of wealth or consumer want satisfaction.

Bork (1967, p. 242) is more explicit: “My thesis is that existing statutes can be legitimately interpreted only according to the canons of consumer welfare, defined as minimizing restrictions of output and permitting efficiency, however gained, to have its way”. In The Antitrust Paradox, after defending a consumer welfare goal for antitrust for 110 pages, Bork explains that the surplus that purchasers lose by paying high prices is not a loss of consumer welfare, but an income transfer (1978, p. 110):

Those who continue to buy after a monopoly is formed pay more for the same output, and that shifts income from them to the monopoly and its owners, who are also consumers.

Bork’s turn of phrase has hopelessly muddled these waters, with courts citing Bork in support of the position that antitrust policy aims to maximize consumer welfare, then making arguments that conceive of harm to market performance in terms of the welfare of purchasers, not net surplus.

Nor have legal scholars found Bork’s argument convincing; law journals are strewn with alternative interpretations of the goals of US antitrust policy.

Fox writes (1981, p. 1154) that “[T]he claim that efficiency has been the goal and the fulcrum of antitrust is weak at best. The values other than efficiency that underlie the commitment to power dispersion, economic opportunity, and competition as market governor demand equal attention”. For Lande (1982, p. 150) “Each antitrust law grew in part out of a desire to define and protect consumers’ property rights, an antipathy toward corporate aggregations of economic, social, and political power, and a concern for small entrepreneurs”. Rowe (1984, p. 1560, in footnote 295) observes that “Not unexpectedly, the 1890 debates reveal a mix of social, economic, and political concerns”.

Brodley regards the economic and noneconomic goals of antitrust as mutually compatible (1987, p. 1022, footnote omitted):

[T]here is a unity between the pragmatic substance of antitrust-its economic goals-and the law’s animating spirit-its social and political foundations. Thus, the pursuit of the correctly defined economic goals of antitrust will generally advance the social and political objectives of the law as well.

Posner (1987, pp. 209–210, footnotes omitted, emphasis added) denies Bork’s principal conclusion:

The legislative history makes clear that the [Sherman] Act was aimed at the great trusts (cartels and monopolies) of the time, but is not single-minded concerning what aspect of the trusts was reprobated. Some members of Congress wanted to punish the trusts because the trusts restricted output and raised price, and thus hurt consumers. Others believed that the trusts, whether through economies of scale or other efficiencies, produced a greater output at lower price, thus helping consumers (in both the short and long run) but hurting inefficient competitors. Still other members of Congress relied on both reasons for supporting the Act, believing that it would both help consumers and help the trust’s competitors. The modern economic analysis of monopoly has made the inconsistency of these two reasons transparent. But no one in 1890 understood the economic concept of efficiency; it hadn’t been developed yet.

Adams and Brock regard the economic goals of antitrust as subsidiary to its political goal, maintenance of a free society (1987, p. 1116)²⁷:

The primary purpose of antitrust is to perpetuate and preserve a system of governance for a competitive, free enterprise economy. Efficiency and consumer welfare constitute ancillary benefits that are expected to flow from a system of economic freedom. …Antitrust calls for a dispersion of power, buttressed by built-in checks and balances, to guard against the abuse of power and to preserve not only individual freedom, but also more importantly, a free system. Antitrust is founded on a theory of hostility toward private concentration of power….

Hovenkamp writes of Bork’s work (1989, p. 22, footnotes omitted)

[Bork] concluded all too quickly that because some members of Congress knew that demand curves slope downward (i.e., that output is reduced as prices rise), that they also had a modern conception of allocative efficiency and the social cost of monopoly. Not a single statement in the legislative history comes close to stating the conclusions that Bork drew.

In sum, we can agree with Ginsburg (2014, p. 947) that Bork’s analysis of the legislative intent underlying the Sherman Act has been rejected by the Academy.

7 Kaldor, Hicks, and Potential Compensation

It is in the nine pages of Chapter 5 of Bork (1978) that Bork explains his expansive use of the term “consumer welfare”. The issues involved in that usage are the subject of a large economics literature, to which Bork does not refer, and can be presented with reference to Fig. 10.1.

Fig. 10.1

Consumer surplus vs. net surplus (constant marginal cost c, entry-inducing price p_E)

It depicts a market supplied by a single firm, a monopolist in the antitrust sense, that can set a price p_E that is greater than its marginal cost c without inducing entry by higher-cost potential competitors.

If the market were perfectly competitive, price would equal marginal cost c. Output would be Q_c. The area of the triangle cDF is consumer surplus under perfect competition. Like Hicks (1942, p. 126), Bork follows Marshall (1920, p. 125) definition of consumers’ surplus as “The excess of the price which [consumers] would be willing to pay rather than go without the thing”. Thus consumers’ surplus is a monetary amount, not a measure of utility. Δπ is not a measure of the utility the owners of the firm gain when they spend their additional income in other markets. ΔCS is not a measure of the utility consumers lose when they reduce spending in other industries.

If the market is supplied by a single firm that sells Q_E units of output at price p_E, the firm earns an economic profit in the amount of the shaded rectangle cP_EEG. Relative to perfect competition, this economic profit is a transfer of purchasing power from the consumers who purchase the good at the price p_E to the owners of the firm. It is also a monetary amount.

Consumers’ surplus at price p_E is the area of the triangle p_EDE. Consumer surplus under monopoly is reduced by an amount equal to the area cp_EEF. If the increase in owners’ income is set against the lost consumers’ surplus, what is left, the area of the triangle EFG, is the deadweight loss due to monopoly power.

Consumers who remain in the market under monopoly reduce spending in other markets by an amount equal to the area of the rectangle cP_EEG. The owners of the firm increase spending in all markets in which they buy by an identical amount. That the changes in income of the two groups are of the same monetary amounts does not imply that the changes in satisfaction of the two groups are the same. What one can say from Fig. 10.1 is that consumers would spend less for the privilege of being able to continue to buy Q_E units than they would be willing to spend to buy Q_C units of output, and that the owners of firms have economic profit $\left({p}_E-c\right){Q}_E$ to spend under monopoly that they would not have under competition. The figure tells us nothing about the implied changes in satisfaction.

We are thus led to the issue of interpersonal welfare comparisons, as were Robbins (1938), Kaldor (1939), and Hicks (1939). Robbins entered into a discussion of the welfare consequences of relaxing nineteenth-century restrictions on the importation of corn into England. The resulting lower price of corn would make consumers better off and farmers worse off, just as the price increase shown in Fig. 10.1 would make consumers worse off and the owners of the firm better off. Robbins’ view was that it was not possible to measure the net welfare impact of the change without making the assumption that consumers and farmers had “equal capacity for satisfaction”. He argued that there was no basis in economic science for making such an assumption.

In the analysis of the welfare impact of monopoly power, if lost consumers’ surplus and increased economic profit represent comparable units of satisfaction, then deadweight loss represents the net welfare loss due to monopoly power. But, Robbins’ position implies, there is nothing in economics as a science that allows us to say that changes in identical monetary amounts of income represent identical changes in satisfaction.

Kaldor (1939), seconded by Hicks (1939),²⁸ responded with what is now called the Kaldor-Hicks potential Pareto improvement welfare criterion. A reallocation of income is a Pareto improvement if it makes some individuals better off and no individuals worse off. A reduction in the tariff on corn is not a Pareto-improving policy change: it makes consumers better off and farmers worse off. If, however, the government were to levy lump-sum taxes on consumers and make lump-sum distributions to farmers in an amount just sufficient to compensate farmers for lost income, and consumers were still better off with a lower price of food, there would be a Pareto improvement.²⁹

Kaldor proposes to look at what the net welfare impact of a policy change would be if these kinds of lump-sum transfers from winners to losers were made, whether or not the transfers actually are made; whether compensation should be given or not (1939, p. 550) “is a political question on which the economist, qua economist, could hardly pronounce an opinion”.

With this approach, Kaldor sought to avoid interpersonal comparisons of satisfaction (1939, p. 551, footnote 1). The consensus of the literature is that this does not work (Winch 1965, p. 406)³⁰:

The original version of the compensation principle, now generally referred to as the Kaldor-Hicks criterion …, under the assumption that full compensation will in fact be made, is uniquely free from either interpersonal comparison or value judgment. …Under the assumption that all changes will be offset, the only question is whether the amount of money collected from gainers will be enough to compensate losers; all comparisons are in hard, homogeneous dollars. It is only when the possibility, or desirability, of compensation is questioned that problems arise. Then some people might be worse off, and it becomes necessary to compare changes in satisfaction.

In the context of the income transfer impact of monopoly power, we can write a weighted sum of economic profit and lost consumer surplus as

$$ {\theta}_1\varDelta \pi +{\theta}_2\varDelta CS.\kern1.00em $$

(10.1)

In a move from competition to monopoly, the change in profit Δπ is positive, and the change in consumer surplus ΔCS is negative. If θ₁ = θ₂ = 1, (10.1) becomes deadweight loss, the statistic is implied by the Kaldor-Hicks potential Pareto improvement criterion and championed by Bork under another name. If θ₁ = 0, θ₂ = 1, (10.1) becomes lost consumer surplus, which is what is relevant under a consumer welfare standard. But there are no theorems in economics that tells us what the weights should be. Weighting changes in income equally, θ₁ = θ₂ = 1, does not avoid interpersonal comparisons: it is simply one of many possible choices about the way the income changes of the two groups might be summarized in a single statistic.

Stigler’s reaction to the potential Pareto improvement principle was that whatever weightings are used should be indicated up front (1943, p. 359): “There is grave danger in leaving the value judgments unspecified except by implication…”.³¹ After a careful review of the economics literature on the topic, Chipman and Moore agree (1978, p. 581):

After 35 years of technical discussions, we are forced to come back to Robbins’ 1932 position. We cannot make policy recommendations except on the basis of value judgments, and these value judgments should be made explicit.

8 Missing Markets and the Goals of Antitrust

Under stringent assumptions that include existence of a complete set of competitive markets, the First Theorem of Welfare Economics tells us that equilibrium is Pareto optimal (Koopmans 1957, Essay 1). This result fails if some markets are not competitive. It fails if some goods are public goods. And it fails if there are missing markets (Newbery 1989).

There are many consequences of market activities for which there are no markets. In some such cases, as, for example, carbon emissions, governments may attempt to create a market that would not otherwise exist. But there may be preferences about other consequences of economic activity for which this is impossible. Individuals may prefer net neutrality over differential pricing for internet service/product suppliers. Individuals may prefer that there be information sources expressing a range of viewpoints, even though they disagree with some of those viewpoints. Individuals may prefer that the corporations should not be able to make unlimited political contributions, or have preferences about absolute firm size. Individuals may prefer that markets be less concentrated rather than more concentrated, even markets for products they do not buy, because they object to the consequences of the rent-seeking expenditures they expect to occur in concentrated markets. It is straightforward to include preferences about the outcomes of market activity for which there are no markets in standard economic models. I illustrate this by comparing two models of a market with linear demand, one in which there are no preferences about market structure and one in which there are preferences about market structure. In the second model, welfare is reduced if market concentration, an indicator of market and other dimensions of power, exceeds a threshold level.

8.1 Demand

8.1.1 No Preferences About Market Structure

As is well known, a linear demand curve can be derived from a micro-level model of N individuals, each purchasing 1 or 0 units of the homogeneous good and each with a reservation price that is uniformly distributed on the interval $\left[0,a\right]$. With such a specification, individual utility in the market for the homogeneous good is

$$ {u}_i=\Big\{{\displaystyle \begin{array}{ll}{x}_i-p& {x}_i\ge p\\ {}0& {x}_i<p\end{array}},\kern1.00em $$

(10.2)

where the individual reservation price x_i is uniformly distributed on the interval $\left[0,a\right]$. The implied inverse demand equation is³²

$$ p=a- bQ.\kern1.00em $$

(10.3)

As is also well known, the inverse demand equation (10.3) can also be derived from a quadratic aggregate welfare function for the good in question,

$$ U(Q)= aQ-\frac{1}{2}b{Q}^2.\kern1.00em $$

(10.4)

8.1.2 Preferences About Market Structure

A specification of individual utility in the market for the homogeneous good if individuals have preferences about market structure is

$$ {u}_i={h}_i\left({y}_i-p\right)+\left\{{\displaystyle \begin{array}{ll}{x}_i-p& {x}_i\ge p\\ {}0& {x}_i<p\end{array}}-\right\{{\displaystyle \begin{array}{ll}\frac{1}{2}\beta {\left({m}_D-m\right)}^2& m<{m}_D\\ {}0& m\ge {m}_D\end{array}}, $$

(10.5)

where m is the number of firms in the market, m_D is the individual’s preferred number of firms, and β > 0 is a parameter that determines the strength of preferences about market structure.

The assumptions made in the following section about the supply side of the market imply that firms produce identical amounts in equilibrium. The number of firms m is then a complete description of equilibrium market structure: the smaller is m, the more concentrated is the market.³³ Preferences of the form (10.5) imply that if the market is more concentrated than an individual prefers, m < m_D, individual utility is reduced.

If the market is less concentrated than the individual prefers, m ≥ m_D, individual utility is determined as in (10.2).

For simplicity, I suppose all individuals have the same preferred number of firms.³⁴ This specification implies that individual welfare falls as the number of firms falls short of the preferred number of firms.

The implied aggregate welfare function is

$$ U\left(Q,{m}_D,m\right)= aQ-\frac{1}{2}b{Q}^2-N\Big\{{\displaystyle \begin{array}{ll}\frac{1}{2}\beta {\left({m}_D-m\right)}^2& m<{m}_D\\ {}0& m\ge {m}_D\end{array}}. $$

(10.6)

In this specification, individuals have preferences over market structure, but there is no market for market structure. Inverse demand in the market for Q is again given by (10.3).

8.2 Supply and Cournot Equilibrium

Since the alternative specifications of individual utility yield the same aggregate demand, Cournot equilibrium values are the same for preferences of both types.

There are m firms. Let constant marginal cost c and fixed cost F be the same for all firms, so the firm cost function is

$$ C\left({q}_i\right)=c{q}_i+F.\kern1.00em $$

(10.7)

In this linear inverse demand constant marginal cost model, firm i’s objective function is

$$ {\pi}_i=\left(a-c- bQ\right){q}_i-F.\kern1.00em $$

(10.8)

Cournot equilibrium output per firm is

$$ q(m)=\frac{1}{m+1}\frac{a-c}{b}.\kern1.00em $$

(10.9)

Equilibrium price is

$$ p(m)=c+\frac{a-c}{m+1},\kern1.00em $$

(10.10)

and equilibrium profit per firm is

$$ \pi (m)=\frac{1}{b}{\left(\frac{a-c}{m+1}\right)}^2-F.\kern1.00em $$

(10.11)

We follow common practice and treat the number of firms as a continuous variable. Then the equilibrium number of firms m_LR makes profit per firm equal to 0:

$$ {m}_{LR}=\frac{a-c}{\sqrt{bF}}-1.\kern1.00em $$

(10.12)

If we substitute (10.12) into (10.9) and (10.10), we get the corresponding long-run expressions:

$$ {q}_{LR}=\sqrt{\frac{F}{b}}\kern1.00em $$

(10.13)

$$ {p}_{LR}=c+\sqrt{bF}.\kern1.00em $$

(10.14)

8.3 Minimum Number of Firms

Suppose the government specifies a minimum price,³⁵

$$ {p}_{\mu }=c+\frac{1}{2}\left(a-c\right)-\frac{1}{2}\sqrt{{\left(a-c\right)}^2-4\mu bF},\kern1.00em $$

(10.15)

where

$$ \mu >{m}_{LR}\kern1.00em $$

is a target number of firms.

Profit per firm is

$$ {\pi}_i=\frac{1}{2}\left[a-c-\sqrt{{\left(a-c\right)}^2-4\mu bF}\right]{q}_i-F.\kern1.00em $$

If there is free entry, the number of firms adjusts until π_i = 0. Then output per firm is

$$ {\displaystyle \begin{array}{c}{q}_{\mu }=\frac{2F}{\left(a-c\right)-\sqrt{{\left(a-c\right)}^2-4\mu b F}}\\ {}=\frac{1}{2}\frac{1}{\mu b}\left[a-c+\sqrt{{\left(a-c\right)}^2-4\mu b F}\right].\end{array}} $$

Total output is μq_μ, the market-clearing price for which is

$$ a- b\mu {q}_{\mu }={p}_{\mu }.\kern1.00em $$

Thus with free entry, the market is supplied by μ > m_LR firms at a price p_μ > p_LR.

p_μ is below the monopoly price. It rises as μ rises. The difference

$$ {p}_{\mu }-{p}_{LR}\kern1.00em $$

is the cost to purchasers, per unit of output, of having a less concentrated market structure.

8.4 Welfare

In this section, I use a numerical example to illustrate the welfare consequences of a minimum-number-of-firms policy.³⁶

Let a = 10, 000, b = 1, and c = 0. Then the inverse demand equation is

$$ p=10,000-Q.\kern1.00em $$

If F = 4, 000, 000, the long-run Cournot number of firms m_LR is 4, each producing output q_LR = 2000. The market-clearing price is p_LR = 2000.

If there is a minimum price p_μ = 4000, then with free entry the market is supplied by μ = 6 firms, each producing q_μ = 1000 units of output.

The welfare consequences of a minimum-number-of-firms policy depend on the structure of preferences.

8.4.1 Preferences Independent of Market Structure

If preferences are independent of market structure, the subaggregate welfare function is

$$ \widehat{U}(Q)=10,000Q-\frac{1}{2}{Q}^2.\kern1.00em $$

(10.16)

In Cournot equilibrium, firms break even. Since profit per firm is zero, aggregate welfare is simply consumer surplus,

$$ C{S}_{LR}=\frac{1}{2}{Q}_{LR}^2=32,000,000. $$

(10.17)

If government sets price p_μ = 4000 and there is free entry, consumer surplus is reduced:

$$ C{S}_{\mu }=18,000,000. $$

(10.18)

Increasing the number of firms in the market reduces welfare if the population is indifferent toward the number of firms that supply the market.³⁷

8.4.2 Preferences Dependent on Market Structure

Alternatively, let preferences depend on the number of firms in the market,

$$ \widehat{U}\left(Q,{m}_D,m\right)=10,000Q-\frac{1}{2}{Q}^2-N\Big\{{\displaystyle \begin{array}{ll}\frac{1}{2}\beta {\left({m}_D-m\right)}^2& m<{m}_D\\ {}0& m\ge {m}_D\end{array}}, $$

(10.19)

and let β = 700.

For the example, consider the case that m < m_D. Then

$$ \widehat{U}\left(Q,{m}_D,m\right)=10,000Q-\frac{1}{2}{Q}^2-3,500,000{\left({m}_D-m\right)}^2. $$

(10.20)

using a∕N = b = 1 to get N = 10, 000.

The long-run Cournot equilibrium values are as before, m = m_LR = 4, p_LR = 2000, q_LR = 2000.

If m_D = 6, Cournot equilibrium aggregate welfare is

$$ \widehat{U}\left(Q,6,4\right)=18,000,000. $$

If the government sets price p_μ = 2763.9, then μ = 5 with free entry. Each firm produces q_μ = 1447.2 units of output, and aggregate welfare is

$$ \widehat{U}\left(Q,6,5\right)=22,680,340. $$

This is greater than net social welfare in long-run Cournot equilibrium. If preferences extend to market structure, then some reduction in concentration increases welfare.

9 Conclusion

From its inception, the mainstream view of US antitrust was that it pursued a range of mutually consistent goals. Closer-to-competitive market performance, for example, promotes economic efficiency in the sense of minimizing cost and effectively allocates resources across markets. It also sustains public belief in the fairness of market processes and so maintains support for a market system of resource allocation, despite the fact that a dynamic market system creates winners and losers.

From the post-World War II period onward, legal and economic scholars associated with the Chicago School of antitrust analysis³⁸ have sought, with much success, to narrow the focus of antitrust, arguing that it should promote allocative and productive efficiency, and that other possible goals are inadmissible because they are not “economic”.

My argument in this chapter is that what is called an “economic” approach is no such thing, and by its nature, could not be. Economics does not teach that antitrust policy should seek to minimize deadweight loss due to monopoly power, or to maximize consumer surplus, or to protect small firms, or to reduce market or aggregate concentration, to the exclusion of other goals. Economics provides tools to measure the costs and benefits of such policies; the choice among them, if choice there must be, is outside the realm of economics (Robbins 1932, p. 129, footnote omitted)³⁹:

[T]here are no economic ends. There are only economical and uneconomical ways of achieving given ends. We cannot say that the pursuit of given ends is uneconomical because the ends are uneconomical; we can only say it is uneconomical if the ends are pursued with an unnecessary expenditure of means.

Bork, as Ginsburg (2014) says and as noted earlier, has been rejected by the Academy. But as he also notes (Ginsburg 2014, p. 949), the Academy has failed to persuade the judiciary: landmark antitrust decisions often cite Bork in support of an “economic approach” that sees antitrust as maximizing net surplus.⁴⁰ The continued adherence of US courts to policy prescriptions for imperfectly competitive markets generated by a school of thought that insists most markets can be treated, most of the time, as if they were perfectly competitive (Reder 1982) presents a challenge to mainstream economists. The adherence of the Supreme Court,⁴¹ if in a unclear way, to a reading of legislative intent that antitrust scholars reject as indefensibly narrow creates a quandary for those scholars when they seek to formulate antitrust policy advice.⁴²

Bork himself, however, explained why it is worthwhile to put forward mainstream economic results (Bork 1967, pp. 242–243):

Antitrust policy is determined, far more than most people realize, by the Supreme Court. Reform is as likely to come through change in the intellectual world which ultimately reaches the Court as by any other means.