Personalized pricing—the use of data and algorithms to charge prices based on personal characteristics—is pretty scary, because in theory it could enable firms to charge consumers the maximum they are willing to pay for each product they buy. In Personalized Pricing as Monopolization (Conn. L. Rev. 2019), I argue that the threat to consumer welfare is grave enough to justify antitrust intervention to stop the practice before it becomes widespread. Why antitrust? Because to personalize different prices to different people, a seller must be able to prevent consumers from trading with each other. That is anticompetitive conduct.
But personalized pricing is also a tremendous opportunity for redistributivists, because redistribution with personalized prices is guaranteed to be efficient. Personalized prices need not price willing buyers out of markets—the principal efficiency concern associated with redistributive pricing—because personalized prices can be tailored to ability to pay.
In my working paper, Personalizing Prices to Redistribute Wealth, and in Big Data, Price Discrimination, and Antitrust (Hastings L.J. 2017), I argue that antitrust enforcers and utility regulators today have the power to order the personalizing of socially-just prices—higher for the rich, lower for the poor. That is because antitrust’s consumer welfare standard gives antitrust enforcers authority to take wealth distribution into account, and the authorization to impose “just and reasonable rates” contained in utility statutes does the same for rate regulators.
The Radical Consumer Welfare Standard
As this work suggests, I differ with other antitrust progressives in that I do not see antitrust’s consumer welfare standard as a barrier to law reform, but rather as an opportunity. The consumer welfare standard has made redistribution easier, not harder, because it establishes that antitrust is not just about promoting competition, but about achieving a distributive goal: the maximization of the wealth of consumers.
In The Antitrust Duty to Charge Low Prices (Cardozo L. Rev. 2018), I argue that this change alone should be enough to induce antitrust to treat excessive pricing as a standalone offense, as the European Union already does. The concern that courts will misclassify at-cost prices as excessive prices can be allayed by making the remedy nominal damages, and leaving it to public opinion and reputational harm to enforce judgments.
In The Antitrust Case for Consumer Primacy in Corporate Governance (UC Irvine L. Rev. 2020), I argue that taking the consumer welfare standard seriously would also resolve one of the most important debates in corporate law: whether the purpose of the firm is to maximize profits and turn them over to shareholders or to run the firm to achieve social justice.
The answer is that antitrust requires that firms should earn no profits at all, much less turn them over to shareholders. Instead, firms should charge the lowest possible prices consistent with achieving efficiency in their operations, to the end of ensuring that consumers alone enjoy all of the surplus generated by production.
The consumer welfare standard also plays a starring role in The Efficient Queue and the Case against Dynamic Pricing (Iowa L. Rev. 2020), my critique of the use of data and algorithms to jack up prices in response to unexpected surges in demand. Industry argues that such “surge pricing” attracts new supply, but I argue that selling out at a low price does that too. As we have learned from the pandemic, “sold out” signs scream “opportunity” just as loudly as does a rising price. The only difference is that surge pricing impoverishes consumers.
Technology is also changing the terms of debate in advertising. In The Obsolescence of Advertising in the Information Age (Yale L.J. 2018), I argue that the vast amount of free product information made available by the Internet has eliminated the only major efficiency rationale for advertising: the dissemination of useful product information. Now that consumers can get their product information elsewhere, the only remaining function of advertising is anticompetitive: to manipulate consumers into purchasing products that they do not really prefer, disadvantaging firms that may be better but fail to advertise.
Using the antitrust laws to challenge advertising may sound crazy, but there is actually plenty of precedent. The Federal Trade Commission attacked advertising as anticompetitive conduct throughout the 1960s and 1970s and won a number of cases, including at the Supreme Court.
Finally, in The Hidden Rules of a Modest Antitrust (Minn. L. Rev. 2021, forthcoming), I argue that the Supreme Court’s aversion to blanket rules of per se illegality in antitrust is deeply flawed, because too much nuance in antitrust adjudication is not economical.
Antitrust enforcers simply lack the resources to engage in the bespoke application of law to facts that the Court requires today under the “rule of reason.” But this introduces a startling bias into the law as applied. Enforcers cannot save money by unilaterally imposing per se rules of illegality, because only courts can authorize changes in the law. But enforcers can and do save money by creating per se rules of legality, because that just means engaging in less enforcement.
The result is that while the Supreme Court believes that in embracing rules of reason it has made antitrust more accurate, in fact the Court has merely succeeded at reducing enforcement. Unless the Court believes that the low tide in enforcement it has created is a good thing—in which case the Court should decide for itself which rules to eliminate, instead of leaving it to the whim of enforcers to decide—the Court must embrace more rules of per se illegality in antitrust.
More on Tech-Driven Pricing
Technology is allowing firms to charge prices that extract more profit from consumers than ever before. I have written extensively about the legal and economic consequences of these “extractive technologies.” My basic approach has been to characterize the problem of extractive technologies in terms of the endpoint to which most are developing, which is the execution of perfect price discrimination: the charging of the highest possible price for each unit of production sold.
In Big Data, Price Discrimination, and Antitrust, I lay out a roadmap that policymakers should follow in addressing the rise of perfect price discrimination. As an initial matter, I argue that the rise of perfect price discrimination (or approximations thereto) resolves the debate over whether antitrust should adhere to a total welfare standard or a consumer welfare standard because (Posnerian competition-for-the-market arguments aside) the perfectly-price-discriminating monopoly does not reduce total welfare, and so the rise of perfect price discrimination effectively eliminates antitrust liability under a total welfare standard. As technologies that enable perfect price discrimination mature, antitrust will be able to continue on only as a statute focused on using competition to achieve pro-consumer distributions of wealth via the consumer welfare standard.
I argue that there are three approaches that antitrust can take to the rise of these technologies. The first is to ban them. One drawback of a ban is that it would prevent firms from using extractive technologies to appropriate a larger share of the surplus they create for consumers. If firms invest more in innovation the more they can appropriate the gains from such investment, then that might be a problem. But I argue that preventing firms from appropriating a greater share of the surplus is not really a problem, because technological advance was moving fast enough in the decades immediately preceding the Internet revolution. So additional increases in appropriability aren’t needed to promote an acceptable level of innovation.
I also point out that in the perfect-price-discriminatory limit consumers enjoy none of the surplus generated by technological advance, so what’s the point? Growth that doesn’t benefit the public is like the sound of one hand clapping. There is such a thing as too much appropriability, and extractive technologies aim for too much.
A second drawback of banning these technologies is that you also lose their efficiencies–their ability to price consumers into the market who otherwise would be casualties of deadweight loss. The other two possible approaches avoid this problem.
The second approach is not to ban these technologies, but instead to ramp up antitrust enforcement across the board in order to ensure that overall extractive technologies do not give producers a net gain in surplus. Antitrust enforcement can counteract the effects of extractive technologies because competition limits the amount of surplus a firm can extract from consumers, even when the firm engages in perfect price discrimination.
Indeed, we can think of extractive technologies as creating a “second dimension of market power” that is more or less orthogonal to the market-structure-based power–the monopoly power–that antitrust traditionally addresses (I use the “second dimension” language in a related piece, The Bargaining Robot).
Extractive technologies permit a firm to increase the amount of surplus the firm extracts from consumers for any given level of monopoly power that a firm might enjoy. So for example if through uniform pricing a firm operating in a moderately concentrated market is able to extract $100 of surplus from the market, extractive technologies might increase that haul to $150 without any change in the amount of monopoly power enjoyed by the firm.
From this perspective, extractive technologies may be said to magnify the consumer harm associated with any given level of monopoly power. If we think of the current level of antitrust enforcement as representing a national policy regarding how gains from trade are divided between producers and consumers, it is clear that extractive technologies upset that balance in favor of firms.
It would then seem appropriate to try to restore the old balance by ramping up antitrust enforcement across the board, reducing structural power in order to compensate for the tech-driven increase in second-dimension power–pricing power due to extractive technologies. (I recognize that these two dimensions of power are not entirely orthogonal to each other. The introduction of price discrimination enables competition on a per unit basis and so can make markets more competitive.)
In other words, I argue that the rise of extractive technologies in and of itself provides a complete basis for an increase in all of antitrust enforcement across the board, quite independently of the question whether concentration has been rising in recent decades.
This is something I emphasized in my remarks at the FTC’s hearings on 21st century competition policy. The advantage of this approach to extractive technologies relative to a ban is that it would permit firms to realize the efficiency benefits of perfect price discrimination but deny them the ability to enjoy any net gain in the share of surplus they take for themselves.
The third approach is to recognize that these technologies can be tools for good, if placed in public, rather than private, hands. In order to operate effectively, extractive technologies must be able to explore both the demand and supply curves of the firm. But any technology that can do that can be programmed to charge the lowest possible prices, rather than the highest possible prices. That is, the technologies are just as good for redistributing wealth to consumers and away from firms as they are for the uses to which private profit maximizers want to put them, which is to redistribute wealth away from consumers and to firms.
But the technologies can only be put to use for the benefit of consumers if government steps in to regulate the way firms use these technologies. What will make this different from traditional government price regulation is that these technologies make price regulation efficient. Government will know supply and demand and be able to target prices within the boundaries they impose to achieve distributive justice without inadvertently pricing anyone, whether firm or consumer, out of the market. (Admittedly, it makes more sense to use technology to impose individualized consumption taxes, as these are the precise dual of income taxes, rather than to use personalized prices to create efficient cross-subsidies in particular markets.)
My subsequent work has been devoted to expanding upon these policy options. In Personalized Pricing as Monopolization, I explore how antitrust might go about banning extractive technologies. I argue that the key here is arbitrage. In order for a firm to engage in perfect price discrimination, the firm must ensure that the low-price buyers do not resell to the high-price buyers. I argue that the various devices used by firms to prevent arbitrage count as illegal refusals to deal for purposes of Section 2 of the Sherman Act and even survive the requirement that actionable refusals to deal be terminations of prior profitable courses of dealing. Moreover, I argue that because extractive technologies are purely redistributive in orientation, they always harm consumers, and therefore no monopoly power test should be required to hold them in violation of Section 2.
In Personalizing Prices to Redistribute Wealth, I explore how government might go about taking control over privately-operated extractive technologies to bring about distributively just and efficient pricing in individual markets. I argue that because antitrust’s consumer welfare standard is explicitly distributive in character–it is widely accepted that the standard would prohibit efficient activities that simultaneously harm consumers, for example–it authorizes courts to take distribution into account in fashioning remedies.
Thus a court could order a defendant to use its personalized pricing technologies to use pricing differentials between rich and poor buyers to redistribute wealth a la AT&T’s mid-century pricing of long distance and local calling, only this time without the efficiency loss because prices would be personalized. I also argue that the “just and reasonable” pricing standard used by most public utility rate regulation statutes authorizes utility regulators to require regulated firms to use personalized pricing to redistribute wealth efficiently.
The paper also makes the case that the rise of extractive technologies should put to rest once and for all the argument that redistribution through the income tax system is more efficient than redistribution via commodity taxation. For personalized pricing in commodity markets eliminates the deadweight loss of commodity taxation.
Finally, in Toward a Per Se Rule against Price Gouging and related work, I depart a bit from this research framework to treat a different class of extractive technologies: those that make it easier for firms to exploit market disequilibrium to redistribute wealth away from consumers. (Note that, due to a snafu during the publication process, a related paper, The Efficient Queue and the Case Against Dynamic Pricing, went to press without having been fully revised and should not be consulted in its present form. The revised version will appear shortly.)
In particular, I argue that technology-driven surge pricing should be treated as per se illegal under Section 2 because it enables firms better to exploit the pricing power afforded them by temporary scarcity. I argue that the ability that surge pricing gives firms to adjust prices much more quickly than supply can adjust in the face of unexpected surges in demand is fundamentally anticompetitive, in the sense that it cuts off the lingering influence over prices of competition immediately prior to the surge. This provides a basis for bringing Section 2 to bear on the practice.