The Stigler Center’s 2023 Antitrust and Competition conference seeks to answer the question: what lays beyond the consumer welfare standard? In advance of the discussions, ProMarket is publishing a series of papers with proposed alternatives to the infamous consumer welfare standard. This piece is part of that debate.
Over recent years, the antitrust law appears to be returning to its historical standard, the “competition and competitive process” standard, often referred to in the Supreme Court as the goal of “protecting competition.” In this post, I want to defend that trend for rule-of-law reasons as well as present a realistic assessment of the legal system’s capabilities and its limits.
Unlike a consumer welfare standard, the competitive process standard does not aspire to be a standard that governs each individual case on an ad hoc basis. It is, rather, a rule of rules — reflective of statutory intent — that anchors the building of antitrust doctrine. It sets the goal for a set of rules that provide guidance. These rules, whether specified by statute, statutory common law, or rulemaking aim to deter efforts to destroy or disrupt competition.
The standard in operation does require the court to have a pre-existing sense of what healthy competition looks like, which is where the concept of a competitive process is helpful. For example, consider the per se ban on price fixing as established in United States v. Socony-Vacuum Oil Co. The ban could have been justified on welfare terms, but it is actually hard to defend a categorical ban along those lines, given that some price-fixing might increase welfare. It is easier to defend the ban on the grounds that price fixing interferes with the usual competitive process in which competitors gain an advantage through lower prices. Or, consider the conduct in the United States v. Microsoft case,: whether the restraints imposed by Microsoft were good or bad for consumers in welfare terms was unknown and unknowable. The Court found a violation of antitrust laws based on the old rule that forcing another company to freeze out a competitor using a restrictive license is not competition on its merits, but its displacement. These core cases are both difficult to defend using a consumer welfare standard, and illustrate the competitive process standard in action
In this sense, the operation of the antitrust law can be compared to the rules governing fouls in soccer, football, or other sports: protecting the underlying competition. It is easy to see that absent deterrence and punishment, certain conduct (tripping, holding, and so on) will bring out the worst, not the best, of participants and ruin the competition itself. To prevent that outcome, referees, in the usual course, need some sense of what normal competition looks like, and need to assess whether the rules were violated by the conduct in question.
Those are the ordinary situations, but sometimes the system confronts conduct that is not clearly pro- or anticompetitive. An example is a novel tactic, like a forward pass in football (legalized in 1906) or the “rugby scrum” in more recent years. Some (but not all) antitrust cases present these kind of questions, and they necessitate the analysis described well in Chicago Board of Trade: asking whether the conduct “promotes competition, or whether it is such as may suppress or even destroy competition.”
Two things can be seen clearly from this metaphor: first, that it would be folly if referees were asked to calculate whether every individual foul was justified in welfare terms. That would not only be infeasable, but would also offer far too little guidance to teams and not clearly make the competition any better.. Second, a case-by-case approach would invite gamesmanship. That’s what we’ve experienced in antitrust law for the last several decades, yielding little deterrence, and the loss of competition in many industries.
This should give some sense of the competition and competitive process standard. But how does the return to a competition approach really compare with the consumer welfare standard? Given the many meanings of the phrase, it is essential to clarify what we are talking about. We emphatically cannot debate the “idealized” consumer welfare standard, the one that assumes we can measure the costs and benefits of everything we’d need to know. It is, I admit, a beautiful vision. A kind of nirvana for nerds. But it is obviously not the world we live in. Yet it is this version of consumer welfare that proponents often prefer to defend. “Used properly,” they sometimes say, “it can encompass everything.”
Instead, the debate needs to center on the real, in-the-trenches consumer welfare standard, which is far more profane. At its worst, it has become a case-by-case burden on the plaintiff to prove a large and measurable price harm. A first year law student might well ask: what does a price harm burden have to do with the statute’s stated concerns loss of competition or monopolization? Good question.
As if embarrassed by this narrowness, it usually comes with a theoretical openness to evidence on effects that obviously matter, like product quality, effects on innovation or supply chains. But of course, those are all matters unknowable or unpredictable. Hence, the cases usually return to a familiar place: a technical fight over predicted price effects, a free-for-all that costs a fortune and ends up offering no more guidance than the last spin of the roulette wheel.
There are three reasons why building a doctrine based on the competitive process standard is preferable to this. The first is the limits of the legal system and human capabilities. There is a reason Jonathan Kanter described consumer welfare as a “central planning standard.” For what Fredrick Hayek said of economic planning can also be said of a complex antitrust case:
“If we possess all the relevant information, if we can start out from a given system of preferences, and if we command complete knowledge of available means, the problem which remains is purely one of logic. … This, however, is emphatically not the economic problem which society faces. The reason for this is that the “data” from which the economic calculus starts are never for the whole society ‘given’ to a single mind which could work out the implications and can never be so given.”
As already alluded to, in all but easy antitrust cases, the legal system faces ambitious and often impossible task of assessing measurable but also immeasurable harms and benefits; and in merger cases, the further complication of needing to predict them in the future. We are all vain creatures, but who are we kidding?
The competition and competitive process standard, on the other hand, acknowledges that economic activity and competition are highly complex processes involving much that is unknown and unknowable. The standard punishes attacks on competition: it does not aspire to and is not keyed in to the impossibly ambitious task of assessing the full welfare effects of any individual conduct or transaction.
Even when it can be conducted, the consumer welfare standard is still an indefensible and embarrassing deviation from Congressional intent, as Eric Posner points out well in this series. Congress did not create an open-ended welfare regime but was more specific: it banned monopolization and enumerated forms of conduct whose effect “may be substantially to lessen competition, or to tend to create a monopoly.” The consumer welfare standard, on its face, would allow a merger to monopoly if predicted to lower prices or raise output. That cannot be squared with the text of either Section 7 or Section 2.
Lastly, the consumer welfare standard is starkly incompatible with rule-of-law goals. Any good legal regime, to the degree it can, should seek to offer ex ante guidance and deterrence. An ex poste welfarist standard is incompatible with that goal.
Consider how absurd it would be if the criminal law tried to assess measurable welfare effects in individual cases. Murder need be punished without trying to figure out if the particular death might have actually benefited society. When a poor man robs a wealthy man, it is still a crime, even if the defense attorney might show, based on a marginal utility analysis, that the wealth transfer was efficient. As with the competitive process, there is presumed to be an underlying system of reasonable human interactions from which gross deviations are punished. It is be hard to do much better, especially if you want to tell the world that “stealing is a crime,” as opposed to ‘stealing may be a crime if its costs exceed the benefits.”
A system where everything is up for grabs every time is no system at all. That said, protecting competition and the competitive process also requires constant work – that of building and maintaining the rules and categories of conduct that are presumed, perhaps in the twinkle of an eye, to be illegal. Fortunately, as antitrust grew up as a legal system, not a welfarist system, it still retains many of the foundations of such a system. We need to work on improving and building the antitrust law and its doctrines, not turning every case into a welfarist free for all.
Antitrust is part of the legal system. It should offer guidance, deterrence, and rule of law. The various iterations of consumer welfare offer none of these and have instead led to an indefensibly expensive, indeterminate system that has strayed so far from the laws themselves as to raise major questions of democratic legitimacy. The goal of rigor was laudable, but there is nothing predictable in what has come out of the project. That is why the retreat is on. The consumer welfare era is ending.
* * *
In the remainder, I wish to address two important previous posts on this site. In a much earlier post, Einer Elhauge, responding to a speech by Jonathan Kanter, argued that seeking to protect the competitive process is too ambiguous or conclusory. But Elhauge’s critique turns out to be largely a difference over language.
Elhauge does not offer a pure defense the current consumer welfare standard, as he agrees that antitrust law should protect “competition and the competitive process.” Instead, he is a critic: he says that the “current way of applying the consumer welfare standard not only creates uncertainty but also undermines deterrence. It imposes enormous litigation costs on plaintiffs. It also frequently requires econometric proof of matters that cannot be established given data limitations.” Well said.
We differ in this respect: he believes we might still renovate the consumer welfare brand. While admitting current practice is bad, Elhauge would hold onto the phrase “consumer welfare” but have it understood as a method for improving the rules and deciding hard cases. To avoid confusion, I believe it better to drop the phrase but agree that using economic methods” is important for hard cases. But calling the use of “economic methods” “consumer welfare” is in my view too confusing.
Imagine a merger in an industry with many firms. You need economic methods to figure out if, in fact, the merger would substantially lessen competition or tend to create a monopoly, as the statute demands. But the statute doesn’t require a burden to prove price effects as is implied by phrase consumer welfare standard. That’s what makes its retention confusing.
In my view the phrase ‘consumer welfare standard’ has simply become too tainted. It implies case-by-case welfare analysis and weakens deterrence. It is very hard to control how language is used – and if adopted by officials, retaining the phrase would constantly risk being deliberately misunderstood and lapsing back into a price burden of proof. The whole debate actually reinforces Kanter’s point that there are too many meanings of the phrase “consumer welfare.” Some brands aren’t worth saving.
Another reason Elhauge gives for holding onto “consumer welfare” is the fact that the Supreme Court has used the phrase in several cases. But the Supreme Court has, for much longer and in far more cases, said that the goal of the antitrust laws is to protect competition. For example, “the Sherman Act was designed to be a comprehensive charter of economic liberty aimed at preserving free and unfettered competition as the rule of trade. . . . [T]he policy unequivocally laid down by the Act is competition.” As Barak Orbach has pointed out, the goal of protecting competition is by far the older and more pedigreed. Consumer welfare is a recent add-on, not unlike an unfortunate 80’s addition to an old stone house. There is no need to demolish the main structure — it is still there, set in stone, anchored in more than a century of caselaw. The real question is what we should do with that 1980’s addition.
* * *
Second, unlike Elhauge, Herbert Hovenkamp and Fiona Scott-Morton are defenders of an open-ended, case-by-case consumer welfare standard, with their proposed demand that plaintiffs demonstrate a “measurable effect of a challenged practice on output or price.” They suggest that courts need to take measurable output as well as prices into account in individual cases like Ohio vs American Express, in which it failed to do so.
Hovenkamp and Scott-Morton make a good-faith effort to show how many concerns, like effects on labor markets, can be captured by focusing on output as well as price. But the approach has the broader deficiencies stated above: it offers little guidance or deterrence and relies very heavily on the availability of measurable data. Hovenkamp and Scott-Morton do admit that “[m]easurement issues are significant.” However, they say that “for nearly all purposes the measurements are ordinal, not cardinal.” But what if there is no data at all? Or no reasonable prospect of measuring essential but immeasurable matters? Back to Elhauge, whose attack on current consumer welfare practice is damning:
“[D]ata limitations are sometimes inherent, but are often worsened when courts do not order the needed production of data from defendants, other market participants, and often other markets. All too often, defendants successfully fight the production of data, and then win on the grounds that the lack of data prevents the needed econometric proof, much like Abraham Lincoln’s lament about the “man who murdered his parents, and then pleaded for mercy on the grounds that he was an orphan.”
Hovenkamp and Scott-Morton believe this is okay: “we need to show that a practice tends to reduce output, but calculating an exact amount is unnecessary to make this point.” In other words, sometimes effects don’t actually need to be measurable. But when? Antitrust experts may think this is meant to leave room for “the twinkle in the eye” analysis – some system of presumed harm. But if the real goal is to create a series of presumed harms – close to a set of rules – Hovenkamp and Scott-Morton ought say so.
Even on its own terms, the formula offers little in terms of the rule of law or deterrence. It is a more complicated version of the case-by-case welfarist analysis that costs a fortune and is prone to gamesmanship. Its message to defendants is this: there is no message. That’s an unacceptable state of affairs for a law “designed to be a comprehensive charter of economic liberty aimed at preserving free and unfettered competition as the rule of trade.”
Articles represent the opinions of their writers, not necessarily those of the University of Chicago, the Booth School of Business, or its faculty.