In late September, the United States Federal Trade Commission sued Amazon for using a set of anticompetitive strategies to maintain its monopoly in the online retail market. ProMarket asked four antitrust experts —two economists and two law professors —to discuss the foundations and strength of the complaint’s arguments, the history of similar cases, and the potential for a legal remedy.
In late September, the U.S. Federal Trade Commission (FTC) announced a large and multi-faceted complaint against Amazon. The accusation argues that Amazon monopolizes both sides of its business: that it dominates 82% of the market for consumers shopping at online superstores and that the sellers on the platform are beholden to Amazon as a distribution channel.
The stakes are high. As the FTC’s former deputy director of competition, John Newman, said in a recent interview with PBS: “I’m a student of antitrust history… Very seldom in the history of antitrust law — which goes back over a hundred years — has there been one case that could do so much good for so many people.”
Given the importance of this case, ProMarket asked four experts in antitrust — two economics professors and two law professors — to weigh in with their assessment of the FTC’s complaint. They discuss the foundations and strength of the complaint’s arguments, the history of similar facets, and the potential for a legal remedy.
Fiona Scott Morton
Theodore Nierenberg Professor of Economics at the Yale School of Management
The core of the both the FTC’s case and the California case against Amazon is the company’s aggressive use of platform MFNs that block entry by rival platforms, who can’t compete on price, and thus cause higher prices.
“Most Favored Nation” (MFNs) are requirements that a third-party seller on Amazon’s marketplace set prices on its own website, as well as on marketplaces that compete with Amazon, that are at least as high as those found on the Amazon marketplace. These contracts end price competition between marketplaces. A merchant selling on a marketplace with lower fees cannot lower the final price to consumers unless it wants to lower the price on Amazon (which has higher fees) as well. If prices are not equalized, Amazon removes the product from the Buy Box and the merchant loses most of its sales (98% of sales come from the buy box according to the complaint). On its retail website Amazon price-matches to Walmart.com and other rivals. But at the same time it holds the merchants (e.g. Tide) responsible for Amazon’s profit margin. When Walmart.com has a sale, Amazon will match the sale price, Amazon’s profit margin will fall below the guaranteed level and Tide must make Amazon whole with a payment. Obviously, the merchant will work hard to discourage price reductions anywhere online in this environment.
The anticompetitive use of MFNs like these is a well-established antitrust theory of harm that has featured in a number of high-profile cases. The competition treatment of platform MFNs was laid out in 2016 in a paper by Boik and Corts (economic theory) and in 2018 in Baker and Scott Morton (enforcement theory of harm).
Interested readers should focus on the California complaint, both because it is half the length of the FTC document, but also because much less is redacted so you can read more details about Amazon’s behavior. In addition, the California complaint was filed a year ago and survived a motion to dismiss in March 2023, so it is ahead of the FTC’s schedule.
The FTC’s complaint is longer partly because it alleges additional illegal conduct and direct evidence of market power. These include amazon’s refusal to let merchants run their own fulfillment and delivery as part of “Fulfilled by Amazon” (FBA). The lack of independent logistics providers hinders merchant multihoming and therefore entry. Another piece of conduct is Amazon’s algorithm that matches Amazon’s prices to the prices of retail competitors with the goal of disincentivizing price competition and promoting collusive high prices. The conduct that the FTC alleges to be an Unfair Method of Competition, Amazon calls “Project Nessie,” raises prices and checks if rivals imitate those higher prices. If they do not, Amazon brings prices down again. This is classic tacit collusion and it is interesting to see that this theory is one of the first uses of the FTC’S new policy on unfair methods of competition.
Chris Conlon
Associate Professor of Economics, New York University Stern School of Business
The strongest case the FTC may have in their complaint is the allegation that Amazon punishes sellers who list identical products at lower prices on rival platforms by demoting them in search rankings. It is well understood by economists (and the courts) how MFN arrangements could lead to higher prices both on Amazon and rival platforms.
MFNs have been on the FTC’s radar since at least the 1990s and were a significant enforcement push under the Obama administration. Claims that previous leadership wouldn’t have dared to bring a case like this seem hard to take seriously. Indeed a similar case brought by the California attorney general has already survived a motion to dismiss. However, Amazon claims to have ceased many of these practices already, and the California case will likely conclude well before the FTC goes to trial which may impact the result of this case.
There are of course other allegations in the complaint related to advertising and algorithmic matching of competitor prices where it will be more challenging to show these are antitrust violations that harm consumers and reduce output (and not merely reduce the profits of third-party sellers). Perhaps the most interesting allegation is that Amazon ties “Prime” status on listings to using their fulfillment services (for an additional fee), and whether or not these services can really be unbundled.
Ultimately, to prevail, the FTC will also need to convince the court that its definition of the relevant market as “Online Superstores” is appropriate. This requires accepting that Amazon competes with the websites for Target and Walmart, but not their brick-and-mortar stores. Economists often find these “all or nothing” definitions of markets unhelpful, because the truth is that Amazon likely has more market power in some areas (such as books) and faces more competition in other areas (such as consumer electronics and clothing).
This challenge is not unique to the Amazon case, and highlights the riskiness of the recently proposed revision to the Merger Guidelines, which seek strict thresholds for market shares and thereby raises the stakes of market definition. The 2010 Merger Guidelines revision placed more emphasis on quantitative measures of competition and less on market share and was motivated in part by the FTC’s investigation of the Whole Foods/Wild Oats merger.
That case also hinged on whether brick-and-mortar superstores like Walmart, Target, and Costco were “in the market.” (In an ironic twist, Whole Foods was acquired by Amazon in 2017, and the proposed definition of “online superstores” would exclude grocery items.)
Herbert Hovenkamp
James G. Dinan University Professor, University of Pennsylvania Carey Law School
The FTC’s complaint against Amazon accuses it of monopolizing two different markets. One is for “online superstores,” which sell millions of noncompeting products. Market power attaches to products, however, not to firms.
For example, Microsoft has substantial market power in its Windows OS (nearly 40%), but not its Bing search engine, which has a 3% market share. In Brown Shoe, the Supreme Court permitted aggregation of noncompeting mens’, womens’, and childrens’ shoes, but only because Brown had roughly the same market share for each. Today, we permit aggregation in some merger cases. The market in Staples was consumable office supplies sold in office superstores, and the thing being purchased was the entire store, not the individual products. That merger would have eliminated competition between the entire lines of two low-price chains. Amazon’s products are sold individually. For most, customers have many choices, both on- and offline. Consumer switching costs are low and sellers’ prices are comparable.
We do sometimes recognize facilities such as hospitals as “cluster” markets, but then the hospital is needed to deliver the service, such as anesthesiology or surgery. You don’t need a superstore to buy a toaster or batteries. For most products Amazon does not have a market share sufficient to make monopolization claims plausible. For example, MFN agreements are competitive for a non-dominant retailer concerned that other retailers might obtain a better deal on a particular product; they can become anticompetitive when used by dominant firms to impose higher costs on rivals. Tying arrangements can be anticompetitive, but only if there is market power in the tying product.
Amazon’s second alleged market, for “Online Marketplace Services,” is more plausible, assuming those services are common across all products. The complaint references such things as the “ability for sellers to set the prices” and to “maintain product detail pages” and display customer reviews. These appear not to be protected by any IP rights and there is no other obvious reason they cannot readily be duplicated. To be sure, Amazon offers these services at a very large scale. Whether this confers advantages sufficient to make it a relevant market remains to be seen.
Harry First
Charles L. Denison Professor of Law Emeritus, NYU School of Law
Remedy, a famed mid-20th century Antitrust Division lawyer once wrote, “is the raison d’etre of the whole lawsuit, for it is the only thing that binds the parties to the litigation and affords relief to the aggrieved public.” Despite this, he wrote, “it was in the past frequently slighted. The exhausted legal warriors, pre-occupied with winning litigious battles, often lost sight of the main economic war.”
Have the government plaintiffs in Amazon similarly lost sight of the goal of this economic war? The short answer is, it’s hard to tell.
The Amazon complaint predictably asks for injunctive relief to stop Amazon from “engaging in its unlawful conduct” or “in any conduct with the same or similar purpose or effect.” It also asks for equitable relief “including but not limited to structural relief.” But the complaint provides no detail on what the terms of such injunctions might be, or, more importantly, what “structural relief” the plaintiffs have in mind.
The opacity of the plaintiffs’ remedy requests is not surprising but also not inevitable. The recent Google search complaint is similarly vague, but the FTC and States’ Facebook complaints call for the divestiture of Instagram and WhatsApp and the Google AdTech complaint specifically seeks the divestiture of important parts of Google’s AdTech business.
The lengthy Amazon complaint points to a number of practices that harm sellers on the “Amazon Marketplace,” particularly involving “fulfillment by Amazon.” Will plaintiffs ask that the fulfillment services unit be spun off? The plaintiffs allege monopolization of the “online superstore market.” Will they ask that this superstore (which consists of Amazon’s sales and third-party sales) be split up somehow? Do the plaintiffs want to separate Amazon’s goods-selling function and its platform-running function, something that has been proposed in Congress?
Greater specificity would help the governments’ case. If structural relief is to be sought, the plaintiffs will need to show that such a remedy is tailored to fit the harm shown at trial, which means that attention to remedy cannot wait until the trial ends.
Government litigators need to win the remedy war, not just the litigation battle. Otherwise, Amazon will not deliver for the aggrieved public.
Professor Scott Morton has consulted for Amazon on antitrust matters in the past. The author’s Amazon engagement ended more than two years ago, and predates the US complaints discussed in this article.
Articles represent the opinions of their writers, not necessarily those of the University of Chicago, the Booth School of Business, or its faculty.
