In its current form, antitrust law is sometimes said to advance consumer welfare and to disregard economic inequality. In fact, because monopoly and monopsony benefit shareholders at the expense of workers and consumers antitrust law redistributes resources from (generally wealthier) shareholders to (generally less wealthy) workers and consumers. Antitrust enforcement agencies seeking to reduce inequality might adjust their priorities and target markets that are disproportionately important for low-income people. Agriculture and health care would be good places to start; food and medicine compose a larger share of the budget of low-income people than of others, and these goods are essential to basic well-being. Regulators should also give priority to labor markets, especially labor markets in which lower-income people participate, and especially where pay gaps based on race or gender are large. In some cases, it is also appropriate to consider sacrificing economic efficiency for distributional goals by introducing distributional weights into antitrust analysis; doing so can increase social welfare. At the same time, antitrust law’s contribution to reducing inequality is subject to substantial diminishing returns.
I. VERY BRIEFLY
If the goal is to reduce economic inequality, public officials have a large number of tools. Does antitrust law have a role to play? Could it be refashioned so as to reduce inequality more effectively? Our answer to both questions is yes. Insofar as it might transfer resources from shareholders to workers and consumers, greater enforcement of the existing antitrust paradigm could, on uncontroversial assumptions, have beneficial effects in terms of equality.1 An explicitly equality-focused approach to antitrust enforcement could have larger effects still. And incorporating distributional effects into substantive antitrust law could do even more, though this approach would raise complex questions. We shall also discuss some possible reforms.
We shall argue that in some cases, it may be appropriate to consider sacrificing economic efficiency for distributional goals by introducing distributional weights into antitrust analysis; doing so can increase social welfare. We do not contend that for those who seek to reduce economic inequality, antitrust law deserves pride of place, but it can make a contribution.
II. INEQUALITIES: INCOME, WEALTH, WELFARE
If antitrust law is to be enlisted to reduce inequality, what, exactly, is it supposed to reduce? When people worry about inequality, they usually consider two kinds: income inequality and wealth inequality. Income inequality refers to differences across annual incomes (resulting from salaries or wages), which people earn from their work. Wealth inequality refers to differences in the stock of wealth owned by people—for example, home equity, possessions like jewelry, and stocks, bonds, and other financial instruments. Income and wealth inequality are usually correlated across lifetimes, but there are exceptions—for example, wealthy heirs who do not work, or highly talented young workers (like top athletes or artists) who are paid vast incomes but have not yet accumulated wealth.
A different and more fundamental concept is welfare, or utility, which refers to a person’s well-being. We can imagine a society in which half the population has a great deal of welfare and the other half has very little. The idea of “welfare” is of course contested, and some people question whether interpersonal comparisons are possible.2 Suppose, however, that in our imagined society, half the population is healthy and happy, with lives full of both pleasure and purpose, and half is unhealthy and unhappy, with lives lacking either pleasure or purpose. It is safe to say that with respect to welfare, that society is highly unequal. In our view, the proper focus, strictly as a matter of principle, should be on inequality of welfare, not inequality of income or wealth.
For various reasons, many people believe that inequality of income, wealth, or welfare may be a serious problem. It may be, for example, that disparities of opportunities, including educational opportunities, have produced inequality in terms of welfare, which means that a morally irrelevant factor has been at work.3 It may be that discrimination on the basis of race and sex has contributed to inequalities, so that one group is systematically below another.4 It may be that some practice—for example, the income tax system—has contributed to, or produced, the relevant inequality, and the practice is itself difficult to justify. Some people are “prioritarians”5; they think that in deciding on relevant policies, it is important to focus on improving the welfare of the least well-off. Prioritarianism has significant implications for regulation, including antitrust law, and those implications have only begun to be explored.6
It is difficult to measure welfare directly,7 and many economists believe (or assume) that welfare is correlated with wealth and income. The extent of the correlation is an empirical question.8 Importantly, economists also assume that welfare increases with money at a decreasing rate. In terms of welfare, a poor person is likely to benefit more than a rich person from a $1,000 check. If you are very rich, that $1,000 check might not benefit you at all; if you are poor, it might produce a significant difference in your life. It follows that if we care only about welfare, and not about inequality per se, we will prioritize policies that reduce (economic) inequality by helping the worst off. We might be less interested in policies that reduce inequality by helping people in the middle catch up to those wealthier than they are or by hurting people at the top without benefiting anyone else.
With regard to both income inequality and wealth inequality, there has been a great deal of concern about changes in the United States over the last fifty years.9 This is not the place for a full accounting, but the actual picture is both complicated and mixed.10 Since 1970, both median incomes and income inequality have increased. In 1970, the median income was $50,200; by 2018, it had grown to $74,600, an increase of 49%. (We use 2018 dollars.) At the same time, the incomes of the wealthy have grown far more than the incomes of middle-class and lower-income households. In the relevant period, the lower-income group gained 43%; the middle-income group gained 49%; and the upper-income group gained 64%.11 And within the upper-income growth, the biggest winners have been those at the very top. As a result, income inequality has been rising for several decades.
Some people might think that the increase in inequality is not terribly disturbing if all groups are experiencing significant income growth. But in terms of wealth inequality, the picture is worse. In 2016, upper-income families had 7.4 times as much wealth as middle-income families and 75 times as much as lower-income families—an increase from 3.4 and 28 in 1983.12 From 2001 to 2016, the wealth of middle-income and lower-income families did not increase at all. Indeed, there was a loss of 45% among lower-income families and of 20% among middle-income families—alongside an increase of 33% among upper-income families.13 A central reason is that the wealthier group was particularly able to gain from a rising stock market because upper-income families have a larger percentage of their wealth in equities and other financial market assets.14 The bottom half of families in terms of net worth owns only 1% of equities, while the top 10% owns 84%.15 (This will be highly relevant to our account here.) One consequence is that from 1983 to 2016, lower-income and middle-income families have had a declining share of aggregate wealth in the United States, falling from 32% to 17% and from 7% to 4% respectively. By contrast, the share held by upper-income families has jumped from 60% to 79%.16
Another statistic that shows the rise of inequality is labor’s share of national income. Over the last twenty years, labor’s share has departed from a relatively stable historical level and plunged several percentage points.17 If everyone owned the same amount of capital, this would not matter for inequality. But because ownership of capital is concentrated in wealthy people, the decline of labor’s share harms the less well-off. While the causes of the decline of labor’s share are complex and contested, part of the explanation is likely the declining bargaining power of labor (which reduces returns to labor) and increasing concentration of product markets (which increases returns to capital).18 A substantial decline in union density since the 1950s coincided with rising inequality and helps explain the divergence in returns for labor and capital.19 Various knock-on effects exacerbate inequality, including reduced intergenerational mobility as the wealthy monopolize educational opportunities for their children, higher housing prices as the wealthy buy second and third homes, and distortion of political outcomes as a result of the influence of political donations. These and other trends have led some commentators to posit the existence of widening inequality of welfare or utility, as famously encapsulated in the claim of rising “deaths of despair” from suicide, drug overdose, and alcoholism.20
For those who are concerned about income inequality or wealth inequality, there can be no doubt that current trends are troubling. And for those who are concerned about economic inequality, the situation was not particularly good in the 1970s or 1980s, when there was also a great deal of such inequality.21 The interest in a more progressive income tax or some kind of wealth tax22 stems from such concerns. Might antitrust law help?
III. MARKET POWER, EFFICIENCY, AND INEQUALITY
A. Conceptual Preliminaries
Antitrust theory is (roughly) concerned with market power—the ability of sellers to charge prices above the competitive rate (“markup”) and the ability of buyers (including employers) to push prices (wages) below the competitive rate (“markdown”). Market power produces two types of effects: efficiency and distributive. The price markup (or wage markdown) causes fewer people to buy things and fewer people to work, resulting in a deadweight economic loss. The price markup (or wage markdown) also transfers wealth from consumers and workers to investors. As we will explain, the distributive effect is normally from poorer to wealthier, though there may be exceptions in practice.
Antitrust law is sometimes said to be governed by the so-called consumer welfare standard. The term is unfortunate and has caused much confusion, even among experts. Start by contrasting the consumer welfare standard with what is sometimes called the “total welfare standard.” In a simple model of supply and demand, “consumer welfare” (or, usually, “consumer surplus”) refers to the difference between price and the amount that buyers at the price would be willing to pay. “Producer welfare” (“producer surplus”) refers to the difference between price and the amount that sellers at the price would be willing to accept. Total welfare is consumer welfare plus producer welfare. Outside antitrust law, total welfare just means efficiency. Antitrust law, as currently understood, advances efficiency only when consumers benefit from the reduction in deadweight loss or, in other words, receive a portion of the surplus that is generated.
But there are further complications. Consumer welfare does not refer only to the welfare of “consumers”; it also refers to the welfare of other buyers such as firms (and thus potentially the investors in those firms); sellers, including workers (who sell their labor); firms and other independent contractors (who sell goods or services); and even investors (if those investors are the victims of anticompetitive behavior).23 Indeed, under current law antitrust will often give remedies to consumer-firms and not to consumer-humans because the firms are intermediaries in a supply chain and thus more likely to be the direct victims of antitrust violators, even though the humans may actually bear the loss.24 Roughly speaking, the consumer welfare standard excludes from antitrust consideration the welfare of the investors of the firms engaging in the allegedly anticompetitive behavior but not the welfare of firms victimized by such behavior. A total-welfare or efficiency standard would incorporate that source of welfare.
Moreover, consumer welfare in antitrust law does not refer to welfare in the sense of utility. It refers to dollar gains and losses; that is to say, consumer surplus. A consumer surplus of $100 will contribute more to the welfare or utility of a poor person than to a wealthy person. Currently, this difference plays no role in antitrust analysis.25
B. The Current Impact of Antitrust Law on Distribution
A simple and straightforward proposition is that insofar as antitrust law is designed to promote consumer welfare, it will also, on standard assumptions, reduce economic inequality as well, essentially as a byproduct.26
To see why, imagine that the firm is a monopolist in the product market and a monopsonist in the labor market. It exercises its market power by raising prices and lowering wages. As a result, customers are hurt (relative to a competitive product market) and workers are hurt (relative to a competitive labor market). Shareholders (the investors) benefit. The overall effect is that input (labor) and output (products) are shrunk, while shareholders obtain a larger portion of the smaller pie, becoming better off on net. Antitrust law that unraveled the monopoly and monopsony, or prevented them in the first place, would thus have two distributive effects: (1) transferring wealth from investors to workers and (2) transferring wealth from investors to consumers. The law would also increase output on net.
It should be immediately clear that antitrust law in this picture would likely—though not necessarily—increase equality and improve social welfare.27 Most consumers are poorer than most investors; thus, the transfer from investors to consumers will reduce inequality and (because a given amount of money will benefit the poor more than the wealthy) increase welfare. Most workers are poorer than most investors; thus, the transfer from investors to workers will also reduce inequality and increase welfare. And because consumers and workers share the surplus, both groups should be better off (or, at a minimum, one group will be better off and the other no worse off). In many cases, the workers and consumers will be the same people: think of a hospital that draws workers from a local market and provides health care in the same local market.
We can think of cases in which the distributive effects would be not merely good but extremely good, from the standpoint of those who care about equality. Imagine that the relevant product (say, food) is bought by many poor consumers. Imagine too that the workers employed by the relevant companies are near or at the bottom of the economic ladder. Imagine, finally, that the investors are quite wealthy, and that the effects of monopoly and monopsony are to enrich them at the expense of consumers and workers. In cases of this kind, antitrust law could have significant effects in decreasing inequality. As we shall see, such cases are not merely hypothetical.
At the same time, it is not hard to think of cases in which distributive effects would be bad or even extremely bad, or simply poorly tailored to the current problems associated with the distribution of wealth.28 Imaginably, a firm could have poor investors and wealthy consumers and workers. If antitrust law broke its monopoly, wealth would flow from poor to wealthy. Another, more interesting example involves labor monopsony. Wage markdowns will sometimes be larger for high-income people than for low-income people. The reason is that high-income people typically have highly specialized talents, and those talents are usually bought by a relatively small number of firms. A pediatric oncologist who lives in a small town may have only one potential employer (the local hospital), while laborers with more limited skills may be able to obtain work from many different employers—fast food franchises, warehouses, factories. Still, the evidence to date shows that on a percentage basis, markdowns do not differ between lower- and higher-income people, possibly because lower-income people are less mobile than higher-income people.29 So a push to use antitrust law in labor markets, if applied broadly, should advance equality; it should also advance well-being across all income classes.
C. Some Limits of Antitrust Law
The discussion so far has assumed an ideal antitrust law, perfectly enforced. A potential concern is that real antitrust law may be over-enforced or poorly enforced, making the situation worse rather than better. Robert Bork, for example, argued that antitrust law, as it existed in the 1970s at least, was used by inefficient small producers to protect themselves from more efficient firms that sought to enter their markets.30 There is now a great deal of skepticism about Bork’s claim, but if it is true, antitrust law will cartelize some markets and potentially transfer wealth from poor to rich. The broader point is that versions of antitrust law that do not advance competitive markets but seek to protect industries will not necessarily produce desirable distributive effects.31 There is a practical concern that aggressive enforcement of the antitrust laws not focused on promoting competition might reduce efficiency and have unwelcome or even perverse distributive consequences.
Even with respect to the narrower question of market power, the scope of antitrust law is limited: even if perfectly enforced, it would not eliminate all markups (or markdowns) created by market power. There are two reasons for this. First, antitrust law tolerates market power when it is a reward for innovation and risk-taking—that is, when the market power occurs “naturally.” So if low-income people cannot afford a particular medicine—one that is cheap to produce—because the manufacturer owns a patent, those people are out of luck. Second, antitrust law generally does not address market power that results from frictions other than concentration or collusion. In labor markets, search costs, job differentiation, and other frictions give employers market power that enables them to suppress wages. Even if antitrust law were able to eliminate all labor market power caused by concentration and collusion, markdowns would still be significant.32 The problem exists in certain product markets as well, such as housing. High housing prices are not due to concentration: nearly all houses are owned by different people. But homeowners have market power deriving from search costs and related frictions.
D. A Note on Race and Gender
Many economists believe that wage and wealth gaps between men and women, and between the general population and historically disadvantaged groups, can be traced to market structure. With respect to the gender pay gap, there is some evidence that women are more tied to location than men are and therefore are less often able to seek jobs that are far away from where they live.33 If this is true, then women will on average live in a smaller labor market area than men do, and that means that fewer employers will compete for their work. With fewer employers competing for their work, women will face larger wage markdowns than men will, even if the employers are not motivated by animus and simply seek to maximize profits. Using German data, Sydnee Caldwell and Oren Danieli found that a quarter of the wage gap between men and women can be explained by differences in commuting preferences and costs.34 Antitrust law aimed at labor markets, which would prevent mergers and other combinations that reduce competition between employers, would help women more than men. This is particularly likely to be true in places (towns rather than cities) where there are relatively few employers. Antitrust theory says that if two employers compete for women and three employers compete for men of equal ability, the wages of the men will be significantly higher than those of the women.
A similar conclusion may be true for groups that have been subject to past and present discrimination or are otherwise vulnerable.35 Black people who live in depressed economic neighborhoods in cities with poor public transportation may be able to find work only with a limited number of employers, compared to other people who live in more prosperous neighborhoods and have access to better public transportation. As Gary Becker pointed out long ago (and as did Joan Robinson even longer ago), employers who harbor animus can more easily discriminate on the basis of race if they face little competition.36 But even if employers do not act on animus and simply maximize profits, they might be tempted to pay differential wages to employees of different races based on differences in employees’ outside options. Those who cannot credibly threaten to quit and work for a competitor will normally be paid less than those who can make such credible threats.
Antitrust law, by enhancing competition, should reduce the wage gaps between men and women and between majorities and minorities. As before, however, there are limits to what it can do. Where a monopolist is benign or susceptible to political pressure, it could use its monopoly power to provide benefits to vulnerable employees that an employer in a competitive market cannot afford to do.37 That means that competition could worsen outcomes.38 And if consumers or workers themselves are discriminators, competitive markets may encourage sellers and employers to accommodate their prejudices (as Becker also pointed out).39 Where prejudice is widespread and reflected in consumer and worker choices in markets, antitrust law can perversely ensure that the market caters to those prejudices more efficiently. Other bodies of law—for example, employment discrimination law—are needed to curtail them.
Taking together the opportunities and limits of antitrust law, one can imagine some possible reforms that would enhance equality. As these reforms have been extensively discussed in the literature,40 we describe them only briefly here.
Enhanced enforcement. If, as we argue, antitrust law can reduce inequality, then one approach is simply to increase enforcement. No special consideration needs to be given to distributive questions. The Department of Justice and the Federal Trade Commission are the main regulatory agencies responsible for antitrust enforcement. Resources available to those agencies for antitrust enforcement have stagnated over the last forty years, even as the size and concentration of the economy has risen and the legal environment in which they have operated has become more challenging, thanks to a series of adverse rulings by the Supreme Court.41 Many commentators have called for larger antitrust enforcement budgets, and Senator Amy Klobuchar has proposed a bill to provide them.42 Larger budgets would lead to greater antitrust enforcement, and greater antitrust enforcement would likely increase equality.
Prioritizing equality. Regulatory agencies could also adjust their priorities in two ways. First, they might target markets that are important for poorer people.43 Agriculture and health care would be good places to start. Food and medicine compose a larger share of the budget of low-income people than of others, and these goods are essential to basic well-being. The two markets are also notorious for the degree of concentration in crucial sectors and the ubiquity of anticompetitive behavior.44 With an equality-focused approach to antitrust law, markets for expensive products like automobiles and jewelry would receive less priority. Second, regulators should also give priority to labor markets, and especially labor markets in which lower-income people participate and where pay gaps based on race or gender are large. As antitrust enforcement has ignored labor markets until recently, anticompetitive behavior in these markets is likely to be common.45
Procedural reform. A consensus is forming that the Supreme Court has excessively weakened antitrust enforcement by raising the barriers to litigation. A string of cases going back to the 1970s has weakened antitrust law by reducing the number of possible plaintiffs to those most directly affected by anticompetitive behavior; raising the threshold for surviving a motion to dismiss; enforcing arbitration agreements that effectively block class actions; raising the standards for class actions; and requiring more proof in a broad range of claims, including conspiracy claims and claims of vertical misconduct.46 Most of these cases were controversial at the time they were decided. In the aggregate, they seem to have seriously depressed antitrust enforcement, especially by private parties. Legal reforms that reversed some or all these decisions would enhance antitrust enforcement and to that extent promote equality.
IV. ANTITRUST LAW, DOING MORE
We see two general approaches to enhancing the role of antitrust law in advancing equality. The first approach would involve strengthening antitrust liability standards so that they advance efficiency and equality. The second approach would involve sacrificing efficiency to equality while retaining the general structure and character of antitrust law. We discuss the advantages and disadvantages of each approach.
Strengthening liability standards. Baker and Salop suggested reconsideration of long-abandoned proposals to use antitrust law to address parallel pricing in oligopolies and monopoly pricing by firms that have obtained their monopolies lawfully, though they were not ready to endorse a change in the law along these lines.47 We agree that antitrust laws should not be available for challenging monopoly pricing. There remains a consensus among economists that the availability of above-market returns is necessary to encourage risk and innovation. And where natural monopolies exist, a specialized regulatory body is the appropriate means for regulating prices.48
However, the long-standing refusal of courts to police parallel pricing should be reconsidered.49 The usual objection is that firms cannot avoid setting their prices by referring to the prices of competitors, so if courts enjoined firms from doing so or awarded damages to plaintiffs, the remedy would be arbitrary and could result in firms underpricing, which can be just as bad as overpricing.50 While that argument might have been a good one when it was first made more than half a century ago, more sophisticated remedies are available today. At least in certain conditions, economists can calculate what prices would prevail in the but-for world in which firms did not price in parallel—for example, by comparing the pricing activity of defendant firms with the pricing activity of similar types of firms in different market areas where a larger number of competitors prevent or reduce the amount of parallel pricing. Firms in concentrated markets would then set prices in anticipation of a sanction if they followed one another’s pricing too closely while charging above the but-for rate and reduce their prices to a level closer to the competitive rate to avoid the sanction.
Note that either approach, if it worked, would simultaneously advance efficiency and equality. As a matter of general understanding, prices sustained by monopoly and oligopoly are inefficient; laws that reduced those prices while preserving economies of scale and other efficiencies would result in higher output. Lower prices and higher wages would effect a transfer of wealth, on average, from rich to poor.
Sacrificing efficiency to equality. To clarify the idea of sacrificing efficiency to equality, imagine that antitrust law is currently perfectly efficient—it generates competition when competition is optimal, maximizing the economic pie. The distributive effects are better than they would be in a world without antitrust law, but they do not achieve the social ideal. The question then is whether antitrust law should be reformed to improve the distributive effects even though the reformed antitrust law would reduce efficiency and, if so, how. To put this point differently, should we want antitrust law to maximize social welfare, accepting reductions in efficiency when the distributional impacts significantly benefit the poor?
As an initial point, the fact that the hypothetical reformed antitrust law would reduce efficiency is not by itself a decisive objection to it. Some well-known social welfare functions, as well as common sense and longtime practice, tell us that efficiency losses can be justified by distributive gains. On utilitarian grounds, we should enthusiastically embrace a reduction in efficiency (measured in monetary terms) if there is an increase in welfare or utility. As an analogy, imagine an occupational safety regulation that costs $900 million but delivers $700 million in benefits. Though the regulation is inefficient, suppose that the $900 million ends up being paid by investors and wealthy consumers and that the $700 million is enjoyed by poor workers. On utilitarian (or welfarist) grounds, the regulation would be a good idea. Some uses of antitrust law could have similar effects. Imagine a merger between two superstores that would reduce appliance prices by an aggregate of $100 million (because of greater efficiencies) but increase food prices by an aggregate of $50 million. It could make sense to block this merger, even though consumers in aggregate benefit, if the food buyers tend to be poorer than the appliance buyers, as is likely the case.
While our claim might seem strange or radical in the context of antitrust law, antitrust already sacrifices efficiency to distribution, as we have already noted. Efficient antitrust law would use the total welfare standard; real antitrust law uses only the consumer welfare standard. As we have explained, the general thrust of this legal approach is to transfer value from (generally wealthy) investors to (less wealthy) consumers and workers, even when a particular action blocked by the law would result in gains for investors that would exceed losses to others. While the consumer welfare standard is often justified on grounds of administrative convenience,51 this seems to us a post hoc rationalization of that standard rather than an adequate defense. Indeed, the legal basis of the standard is suspicion of monopolies, not concern about administration.52
A better objection is that the tax and transfer system redistributes wealth more efficiently than liability rules do.53 This point is hardly unfamiliar, and it can be brought against proposals to address income inequality through antitrust law, though it would require a revision of current law so that the total-welfare standard rather than the consumer-welfare standard is used.54 But that objection is not decisive. Policy makers might rationally decide to seek a better wealth distribution through regulatory law, including antitrust law, if political or practical barriers prevent expansion of the safety net.55 Indeed, the safety net may be subject to diminishing returns. Consider unskilled, poorly housed day laborers who try to avoid bureaucracies because of (justified or unjustified) fear of hostility or prejudice, or who simply cannot navigate the rules.56 Those laborers would nonetheless benefit from enhanced antitrust enforcement that reduced anticompetitive behavior by their employers, as the antitrust law could operate without the laborer’s participation in the legal process.
Kaplow has pointed out that if judges try to redistribute wealth through, say, the common law, they may find their efforts thwarted by legislatures.57 It may also seem inappropriate for judges to use the common law to advance goals that we normally associate with legislation. But antitrust law is based on statutes, and while the law in practice has common-law elements, common law development has pushed antitrust law away from its original goals.58 In a world in which voters do not like the word “tax” but do like redistribution, Congress might (and does) redistribute wealth using other legal means and might (and seems to) tolerate regulatory agencies and courts that do so as well.
Like other commentators who have discussed this issue, we are mindful of the risks of using antitrust law to advance general social values that are not tied to issues of market competition.59 Consider, for example, a frequently made and historically popular argument that small businesses should be protected from competition by large and more efficient firms.60 In some circumstances, this argument makes sense. The commercial areas of many small towns were decimated by the arrival of Walmart and other super-efficient national chains. Prices fell, but it turned out that a bustling Main Street was an amenity that many people valued. Consequently, towns have invested taxpayer money to subsidize small businesses; it is possible that blocking the Walmart store would have been a better strategy. But where small businesses are less efficient than large businesses, they will charge consumers more and pay workers less.61 The owners of those businesses are often middle-class people who are wealthier than the customers and workers. We accordingly do not think antitrust law should protect small businesses in generic fashion; where small businesses are valued, other forms of legal intervention seem more appropriate.
But antitrust law could be used to advance both market competition and the well-being of the least well-off in two ways, which we will designate as “ad hoc” and “systematic.”
For an ad hoc example, consider a recent paper by Christopher Leslie, which investigated the role of anticompetitive behavior in the creation of “food deserts,” areas (frequently inner-city neighborhoods) that lack grocery stores that sell fresh and healthful food.62 Leslie argued that many food deserts are the result of restrictive covenants. Many supermarkets that at one time served poor neighborhoods have moved their operations to wealthier parts of town but have used legal arrangements to ensure that a new grocery store would not be able to move into their vacated space. In some circumstances, the departing supermarkets have sold the property subject to covenants preventing use of the land by another grocery store, often for a very long time, as much as fifteen years. In other cases, the departing supermarkets retained the land but leased it subject to similar covenants. This behavior is potentially subject to antitrust law because it likely was motivated, at least in part, by the desire to prevent a competitor from setting up in the original location and poaching customers who would otherwise drive to the new location.
Leslie argued that the failure of antitrust challenges to supermarkets that create food deserts is due to erroneous antitrust analysis by courts, including the failure to properly define market areas. This may be true, but one could also argue that the courts should have been responsive to the food desert problem even if the courts ruled correctly under the current understanding of antitrust law. To see why, imagine that the supermarket would have stayed in the poor neighborhood if restrictive covenants were not available. The result would be that the wealthier suburban consumers would be forced to pay higher prices (or travel to more distant supermarkets), while the poorer inner-city residents would retain access to fresh food. Efficiency in dollar terms is down, but welfare in human terms is up. We think this outcome is defensible even if a superior outcome could be hypothetically achieved through a never-to-be-implemented adjustment to the tax and transfer system. To implement the ad hoc system, the antitrust agencies could apply higher standards to mergers and other allegedly anticompetitive actions when they directly affect pricing for low-income people. The number of such cases would likely be low, but they could arise in a range of circumstances, when the sellers in question supply basic necessities—food, medical care,63 housing, low-price goods64—or are themselves low-income suppliers of labor.65
The systematic approach would simply generalize the ad hoc example. In debates on cost-benefit analysis and regulation, some commentators have argued that the inputs for cost-benefit analysis should be distributively weighted.66 Under existing law (with an important qualification to be mentioned), regulatory impacts are based on willingness to pay.67 All else equal, a wealthy person will pay more for an improvement in health, safety, or other amenity than a poor person will, so the wealthy person’s interest will be given more weight in regulatory practice. The qualification is that although in theory regulatory impact analyses should make distinctions in terms of wealth, they often do not; in the domain of mortality risk, the willingness to pay for reductions in the risk of death are given equal weight for rich and poor, producing a uniform “value of a statistical life.”68 Whether or not this is the right approach, there is a plausible argument for distributive weights in the context of regulatory policy.69 This approach might sacrifice efficiency, but it would produce equality, and it would also increase social welfare. For purposes of antitrust analysis, where most of the impacts are purely monetary, distributive weighting would result in counting a $1 higher price paid by the poor as more important than a $1 higher price paid by a wealthy person. Once again: such weighting would at once increase social welfare and improve distribution.
Because antitrust law already values consumer impacts and mostly disregards producer impacts, a kind of distributive weighting is already embodied in antitrust law. But it is extremely crude. As noted above, the consumer welfare standard often includes “consumers” who are firms, and thus, in effect, investors. Against a normal social welfare function, antitrust law also gives excessive weight to consumers who are wealthy (including, under the Illinois Brick rule, supply-chain intermediaries rather than final consumers) and insufficient weight to investors who are not wealthy. With this as backdrop, a more sophisticated version of antitrust law that used distributive weighting would be neither radical nor insignificant in its effects.
There are two possible objections to this argument. First, a range of technical and practical problems must be overcome before distributive weighting could be used with confidence. This is true, but the current approach doesn’t so much avoid these problems as assign them to the tax-and-transfer system, which must be based on some social welfare function even if only an implicit one.
Second, and more important, greater emphasis on distributional weights produces some apparent paradoxes. Consider, for example, a group of buyers who agree not to buy goods from certain sellers above a certain price. This would be an illegal cartel under current law. But if the buyers are sufficiently poor and the sellers are sufficiently rich, a distributively sensitive antitrust law would permit this cartel even though it reduces competition (among the buyers). We mention this objection because we expect many readers will make it. However, there is no real paradox here. Antitrust law already allows firms to merge and enter other cooperative arrangements that reduce competition where competition is understood in the lay sense to mean multiple firms supplying the same goods or services. And other areas of law—notably, labor law—allow agents (workers) to aggregate their bargaining power (through unions).70 The goal of antitrust law, as understood today, is not just competition but a version of efficiency that is mostly limited to consumer (or worker) surplus that is threatened by anticompetitive practices.71 Our modest suggestion is that consumer surplus should be replaced with consumer/worker (which is to say, human) well-being, understood in utilitarian terms.
Antitrust enforcement could advance equality through modest reforms, including increased enforcement, equality-based priority-setting, and procedural changes. More aggressive reforms, including permitting challenges of certain kinds of parallel price and wage setting, would also be justified. All these reforms could be accomplished with little violence to the law—indeed, in the view of many commentators, they would simply help return antitrust law to first principles. More ambitiously, we argue that antitrust reformers should consider reducing the focus of the law on efficiency (as partially embodied in the consumer welfare standard) and see it as a way to improve distributional outcomes even when they involve a sacrifice of efficiency. As we have argued, the law already does that by ignoring producer surplus generated by anticompetitive behavior. A version of antitrust law that addresses this problem in a more rigorous way would produce better outcomes if technical and administrative hurdles could be overcome.
See Jonathan B. Baker & Steven C. Salop, Antitrust, Competition Policy, and Inequality, 104 Geo. L.J. Online 1 (2015); Lina Khan & Sandeep Vaheesan, Market Power and Inequality: The Antitrust Counterrevolution and Its Discontents, 11 Harv. L. & Pol’y Rev. 235 (2017); see also Ioannis Lianos, Competition Law as a Form of Social Regulation, 65 Antitrust Bull. (2020). Additional sources will be discussed infra.
See Jon Elster & John E. Roemer, Introduction, in Interpersonal Comparisons of Well-Being 1–17 (Jon Elster & John Roemer eds., 1998).
See John Rawls, A Theory of Justice (1971).
See Cass R. Sunstein, The Anticaste Principle, 92 Mich. L. Rev. 2410, 2412 (1994).
See generally Matthew D. Adler, Theory of Prioritarianism, in Prioritarianism in Practice (Matthew D. Adler ed., forthcoming 2022).
For one discussion, see Ioannis Lianos, Competition Law as a Form of Social Regulation, 65 Antitrust Bull. 3, 27 (2020).
For some influential efforts, see generally Paul Dolan, Happiness by Design (2014); Daniel Kahneman et al., Back to Bentham? Explorations of Experienced Utility, 112 Q.J. Econ. 375 (1997).
See Ed Diener & Robert Biswas-Diener, Will Money Increase Subjective Well-Being? 57 Soc. Indicators Rsch. 119, 120 (2002).
See Emmanuel Saez, Striking It Richer: The Evolution of Top Incomes in the United States, Pathways Mag., Winter 2008, at 6–7 (illustrating the “explosion of top wages and salaries” since 1970).
See Juliana Menasce Horowitz et al., Most Americans Say There Is Too Much Economic Inequality in the U.S., but Fewer Than Half Call It a Top Priority 12–13, Pew Rsch. Ctr. (Jan. 9, 2020), https://www.pewresearch.org/social-trends/2020/01/09/trends-in-income-and-wealth-inequality/ [https://perma.cc/E73V-HVMA], from which we draw heavily here.
Id. at 16.
Id. at 20.
See Neil Bhutta et al., Changes in U.S. Family Finances from 2016 to 2019: Evidence from the Survey of Consumer Finances, 105 Fed. Rsrv. Bull. 1, 2–3 (2020).
See Robert Gebeloff, Who Owns Stocks? Explaining the Rise in Inequality During the Pandemic, N.Y. Times, Jan. 26, 2021, at B3 (summarizing 2019 Federal Reserve figures).
Horowitz et al., supranote 10, at 20.
See, e.g., Andrea L. Eisfeldt et al., Human Capitalists 1–7 (Nat’l Bureau of Econ. Rsch., Working Paper No. 28815, 2021); Matthew Smith et al., The Rise of Pass-Throughs and the Decline of the Labor Share 1–3 (Nat’l Bureau of Econ. Rsch., Working Paper No. 29400, 2021).
See Yasser Abdih & Stephan Danninger, What Explains the Decline of the U.S. Labor Share of Income? An Analysis of State and Industry Level Data 4–6 (Int’l Monetary Fund, Working Paper No. 17/167, 2017); David Autor et al., The Fall of the Labor Share and the Rise of Superstar Firms, 35 Q.J. Econ. 645, 647–48 (2020); Anna Stansbury & Lawrence H. Summers, The Declining Worker Power Hypothesis: An Explanation for the Recent Evolution of the American Economy, 1–8 (Nat’l Bureau of Econ. Rsch., Working Paper No. 27193, 2020). The question of whether increasing inequality is due to rising market power in capital and product markets has received enormous attention. For some discussions, see Einer Elhauge, Horizontal Shareholding, 129 Harv. L. Rev. 1291 (2016); Filippo Lancieri, Eric A. Posner & Luigi Zingales, The Political Economy of the Decline in Antitrust Enforcement in the United States (Stigler Ctr., Working Paper No. 314, 2021).
See Henry S. Farber et al., Unions and Inequality over the Twentieth Century: New Evidence from Survey Data, 136 Q.J. Econ. 1325, 1326–27 (2021).
Anne Case & Angus Deaton, Deaths of Despair and the Future of Capitalism (2021).
Cf. Cong. Rsch. Serv., R44705, The U.S. Income Distribution: Trends and Issues 7–8 (2021) (charting steady and consistent rise in U.S. income inequality starting in the early 1970s).
See Emmanuel Saez & Gabriel Zucman, Progressive Wealth Taxation, 2019 Brookings Papers on Econ. Activity 437, 438–39.
This is all plain from the case law but has tripped up even sophisticated commentators. For example, Katz and Farrell argued that what they call the “consumer surplus” standard (by which they apparently mean “consumer welfare”) would exclude impacts on workers, and for that reason they prefer a “total surplus” standard. See Joseph Farrell & Michael L. Katz, The Economics of Welfare Standards in Antitrust, 2 Competition Pol’y Int’l 3 (2006). To be fair, courts are not always clear. See, e.g., Nat’l Collegiate Athletic Ass’n v. Alston, 141 S. Ct. 2141 (2021), in which the Supreme Court declined to take a position on this question.
See Illinois Brick Co. v. Illinois, 431 U.S. 720, 745–46 (1977).
For one of many helpful discussions of this tangle, see Steven C. Salop, Question: What Is the Real and Proper Antitrust Welfare Standard? Answer: The True Consumer Welfare Standard, 22 Loy. Consumer L. Rev. 336 (2010). A further absurd wrinkle is that Robert Bork and some of his followers have used the term “consumer welfare” to refer to total welfare (in the sense of aggregate efficiency). For a discussion, see Barak Y. Orbach, The Antitrust Consumer Welfare Paradox, 7 J. Comp. L. & Econ. 133 (2011).
See Daniel A. Crane, Antitrust and Wealth Inequality, 101 Cornell L. Rev. 1171 (2016). For a rebuttal of Crane’s argument, see Elhauge, supranote 18, at 1291–97.
José Azar, Steven T. Berry, & Ioana Marinescu, Estimating Labor Market Power (Sept. 18, 2019), https://ssrn.com/abstract=3456277.
See generally Robert H. Bork, The Antitrust Paradox: A Policy at War with Itself (1978).
See Herbert J. Hovenkamp, Antitrust Policy and Inequality of Wealth, Antitrust Chron., October 2017, at 1, 1–2.
See Suresh Naidu & Eric A. Posner, Labor Monopsony and the Limits of the Law, 57 J. Hum. Res. S284 (2021).
See, e.g., Thomas Le Barbanchon et al., Gender Differences in Job Search: Trading off Commute Against Wage, 136 Q.J. Econ. 381 (2021). For a discussion of the literature, see Alan Manning, Monopsony in Labor Markets: A Review, 74 ILR Rev. 3, 20 (2021).
Sydnee Caldwell & Oren Danieli, Outside Options in the Labor Market (2021), https://sydneec.github.io/Website/Caldwell_Danieli.pdf.
The evidence so far is suggestive rather than conclusive. See Lisa Dettling et al., Recent Trends in Wealth-Holding by Race and Ethnicity: Evidence from the Survey of Consumer Finances, FEDS Notes (Sept. 27, 2017); Mark Stelzner & Kate Bahn, Discrimination and Monopsony Power 2–3 (Equitable Growth, Working Paper No. 071320, 2020); Gary A. Hoover et al., The Impact of Economic Freedom on the Black/White Income Gap, 105 Am. Econ. Rev. 587 (2015). For recent discussions, see Aida Farmand & Teresa Ghilarducci, Monopsony Power, Race, and Gender, Oxford Research Encyclopedia of Economics and Finance (2022); Josh Paul Davis et al., Antitrust as Antiracism: Antitrust as a Partial Cure for Systemic Racism (and Other Systemic “ISMS”) (2021), https://ssrn.com/abstract=3816202.
See Gary S. Becker, The Economics of Discrimination 39 (1971); see generally Joan Robinson, The Economics of Imperfect Competition (1933). As noted in the text, there are important qualifications. If, for example, customers favor men over women, competitive pressures might lead to more discrimination, not less.
This was the case for a range of postwar industries in the United States, including automobiles and steel. The oligopolized industries shared their rents with workers, albeit with the informal prodding of the U.S. government. See generally Samuel Milner, Robbing Peter to Pay Paul: Power, Profits, and Productivity in Modern America (2021).
As suggested by some recent papers. For example, a study showed that when private equity firms took over nursing homes, health outcomes worsened—presumably because the new owners were more attuned to maximizing profits. See Atul Gupta et al., Does Private Equity Investment in Healthcare Benefit Patients? Evidence from Nursing Homes (Nat’l Bureau Econ. Rsch., Working Paper No. 28474, 2021).
For interesting new evidence, see Ellora Derenoncourt & Claire Montialoux, Minimum Wages and Racial Inequality, 136 Q.J. Econ. 169 (2021). The authors found that the extension of the federal minimum wage law in 1967 reduced the Black-white wage gap, likely because it helped Black workers trapped in back-of-the-house jobs perhaps designed to accommodate the prejudices of customers that white workers refused to take because of the stigma.
See Baker & Salop, supranote 1.
For a summary of the literature, see Lancieri, Posner & Zingales, supranote 18, at 1–3.
This seems to be the view of Khan & Vaheesan, supranote 1, though they also deny that the goal of antitrust law should be efficiency, and Baker & Salop, supranote 1, at 4–5, though some of their proposals go beyond the efficiency standard. See also Carl Shapiro, Antitrust: What Went Wrong and How to Fix It, 35 Antitrust L.J. 33, 34 (2021). See generally Jonathan Baker, The Antitrust Paradigm: Restoring a Competitive Economy (2019).
On agriculture, see David Baumer et al., Curdling the Competition: An Economic and Legal Analysis of the Antitrust Exemption for Agriculture, 31 Villanova L. Rev. 183, 185–86 (1986). On health care, see Khan & Vaheesan, supranote 1.
As suggested by the literature on noncompetes, see Evan Starr et al., Noncompete Agreements in the U.S. Labor Force, 64 J.L. & Econ. 53, 80–82 (2021); on franchise no-poaching agreements, see Alan B. Krueger & Orley Ashenfelter, Theory and Evidence on Employer Collusion in the Franchise Sector 4 (Nat’l Bureau of Econ. Rsch., Working Paper No. 24831, 2018). For a discussion, see generally Eric A. Posner, How Antitrust Failed Workers (2021).
See Maurice E. Stucke & Ariel Ezrachi, The Rise, Fall, and Rebirth of the U.S. Antitrust Movement, Harv. Bus. Rev. (Dec. 15, 2017), https://hbr.org/2017/12/the-rise-fall-and-rebirth-of-the-u-s-antitrust-movement.
See Baker & Salop, supranote 1. Another group of reforms center on allowing certain vulnerable sellers (e.g., farmers and ride-for-hire drivers) who face powerful counterparties to form buyers’ collectives—effectively unions for sellers. See, e.g., Sandeep Vaheesan, Gig Workers Need Antitrust Reform, Dissent (Dec. 18, 2020), https://www.dissentmagazine.org/online_articles/gig-workers-need-antitrust-reform.
See generally Stephen Breyer, Regulation and Its Reform 4 (1982).
See generally Louis Kaplow, Competition Policy and Price Fixing (2013).
See Donald F. Turner, The Definition of Agreement Under the Sherman Act: Conscious Parallelism and Refusals to Deal, 75 Harv. L. Rev. 655, 665–66 (1962).
See Hovenkamp, supranote 31, at 2.
See United States v. Philadelphia Nat’l Bank, 374 U.S. 321, 372 (1963).
See Louis Kaplow & Steven Shavell, Why the Legal System Is Less Efficient Than the Income Tax in Redistributing Income, 23 J. Legal Stud. 667, 669 (1994).
See Louis Kaplow, On the Choice of Welfare Standards in Competition Law 3–4, in The Goals of Competition Law (Daniel Zimmer ed., 2012). Kaplow has recently published a paper that offers a more qualified view; see Louis Kaplow, Market Power and Income Taxation, 13 Am. Econ. J.: Econ. Pol’y 329 (2021); see also Farrell & Katz, supranote 23; Carl Shapiro, Antitrust Law in an Age of Populism, 61 Int’l J. of Ind. Org. 714, 716–17 (2018).
Cf. Zachary Liscow, Redistribution for Realists, 107 Iowa L. Rev. 1 (2022); and Daniel J. Hemel, Regulation and Redistribution with Lives in the Balance, 89 U. Chi. L. Rev. 649 (2022).
See Pamela Hurd & Daniel Moynihan, Administrative Burden (2019).
He makes this argument against redistributive antitrust law as well. See Kaplow, supranote 54, at 10–11.
See Barak Orbach, How Antitrust Lost Its Goal, 81 Fordham L. Rev. 2253, 2276 (2013).
Various proposals have been made to allow businesses to avoid antitrust liability when they cooperate to achieve social values like sustainability. In Germany, for example, these proposals have had impact on the law. For a discussion, see Sustainability and Competition – Note by Germany, OECD (2020), https://one.oecd.org/document/DAF/COMP/WD(2020)63/en/pdf. The risk here is that businesses would be able to use these defenses to rationalize anticompetitive behavior and escape liability. For example, all cartels reduce output: to allow polluting companies to argue that their cartel thus helps the environment would either immunize them from antitrust liability or embroil courts and regulators in impossibly complex controversies that involve tradeoffs that are more appropriately handled with different bodies of law (e.g., environmental law).
See, e.g., Thomas E. Kauper, Influence of Conservative Economic Analysis on the Development of the Law of Antitrust, in How the Chicago School Overshot the Mark: The Effect of Conservative Economic Analysis on U.S. Antitrust 43 (Robert Pitofsky ed., 2008).
See Autor et al., supranote 18, at 647–48.
See Christopher R. Leslie, Food Deserts, Racism, and Antitrust Law, 110 Cal. L. Rev. (forthcoming 2022).
See, e.g., Indictment, United States v. DaVita, Inc., No. 21-cr-00229-RBJ (D. Colo. July 14, 2021); Mike Scarcella, DaVita Loses Bid to Dismiss DOJ’s Criminal Antitrust Charges, Reuters (Jan. 28, 2022, 6:21 PM EST), https://www.reuters.com/legal/litigation/davita-loses-bid-dismiss-dojs-criminal-antitrust-charges-2022-01-28/. But note that the poorest consumers received medical care through Medicaid.
See Press Release, Fed. Trade Comm’n, FTC Requires Dollar Tree and Family Dollar to Divest 330 Stores as Condition of Merger (July 2, 2015), https://www.ftc.gov/news-events/news/press-releases/2015/07/ftc-requires-dollar-tree-family-dollar-divest-330-stores-condition-merger.
See, e.g., Deslandes v. McDonald’s USA, LLC, No. 17-C-4857, 2021 WL 3187668 (N.D. Ill. July 28, 2021); Daniel Wiessner, Judge Rejects Nationwide Class in McDonald’s No-Poach Case, Reuters (July 29, 2021, 2:30 PM EDT), https://www.reuters.com/legal/litigation/judge-rejects-nationwide-class-mcdonalds-no-poach-case-2021-07-29/.
Matthew D. Adler, Cost-Benefit Analysis and Distributional Weights: An Overview (Duke Env’t & Energy Econ. Working Paper. No. EE 13-04, 2013), https://ssrn.com/abstract=2467673; Matthew Adler, Measuring Social Welfare: An Introduction (2019).
See generally Matthew D. Adler, supranote 5.
See Departmental Guidance on Valuation of a Statistical Life in Economic Analysis, U.S. Dep’t Transp. (Mar. 23, 2021), https://www.transportation.gov/office-policy/transportation-policy/revised-departmental-guidance-on-valuation-of-a-statistical-life-in-economic-analysis.
See generally Hemel, supranote 55; Chris Hope, Discount Rates, Equity Weights, and the Social Cost of Carbon, 30 Energy Econ. 1011 (2008).
15 U.S.C. § 17.
As Farrell and Katz point out, a welfare standard requiring antitrust law to maximize welfare in all circumstances would eliminate the distinctive features of antitrust law, notably, its focus on competition. Farrell & Katz, supranote 23, at 7. But that is a generic problem with all types of economic analysis, including economic analysis that seeks to maximize efficiency. Domain restrictions are required, and in some contexts may be tricky, but that is inherent in the exercise of law-related policy analysis.
Kirkland and Ellis Distinguished Service Professor, University of Chicago Law School.
Robert Walmsley University Professor, Harvard University. Many thanks to Einer Elhauge, Steve Salop, and Larry Summers for helpful comments; none should be held responsible for what we say here. We are grateful to Eric Xu for superb research assistance. After this paper was substantially completed, Posner took a position in the Department of Justice; the views expressed in this paper do not necessarily reflect those of the Department of Justice or U.S. government. While this paper was being written, Sunstein served as Senior Counsel in the Department of Homeland Security, which does not focus on the issues explored here.