No, Digital Platforms Competing With Third-Party Services Is Not Anticompetitive

The recent iRobot debacle has stirred up considerable controversy, and rightfully so. Had the company’s planned transaction with Amazon in 2022 not been blocked, it would have been able to scale and challenge the growing dominance of state-subsidized Chinese competitors in the robotics industry. This story is a lesson for antitrust populists who believe that digital platforms like Amazon should be treated as utilities and therefore should not be permitted to enter separate product and service markets. This kind of thinking will needlessly hobble some of our most dynamic companies, leading to less innovation and less competition.

Opponents of the Amazon/iRobot deal believed that Amazon would favor iRobot products on its e-commerce platform, thereby harming competition. This concern about vertical conduct is nothing new. To antitrust populists, such as former Federal Trade Commission Chair Lina Khan, it invariably leads to harm in digital markets and should therefore be banned outright. In a Columbia Law Review article, Khan argued for the “separation of platforms and commerce,” calling for digital platforms to be prohibited from offering consumer products and to be regulated like utilities. In other words, she believes vertical integration in digital markets should be per se unlawful.

On the contrary, antitrust enforcers have long recognized the pro-consumer benefits of vertical integration. For example, automotive companies install radios and floor mats in cars, which is a form of vertical integration and allows for what is known as the “elimination of double marginalization.” When a car manufacturer produces its own radios instead of buying them from a separate company, it can eliminate the need for both firms to earn a profit markup. In e-commerce, Amazon supplying its own Amazon Basics products eliminates double marginalization in the same way that supermarket store brands do when competing with premium brands, regardless of whether the seller is a digital platform. Vertical integration is extremely common and tends to increase as technological advancements allow firms to incorporate and manage more steps in the supply chain. In theory, antitrust authorities could increase competition by requiring separate firms to provide radios and cars, but this would likely raise prices and harm consumers—the opposite of antitrust policy’s intent.

A platform offering complementary third-party services can also generate economies of scope. A firm can increase efficiency by using its resources across different lines of business, such as a restaurant that uses the same equipment and labor to provide both dine-in and takeout services. Economies of scope exist in digital markets, too. For many internet users, Google’s search engine is a natural place to quickly find reviews of restaurants and other establishments, as well as directions through Maps. Because Google is already serving users through its search engine, it can take advantage of economies of scope to provide additional convenient services that draw on the resources it already employs and save consumers time. This makes Google’s platform even more valuable to users.

In pathbreaking work on vertical integration, economist Oliver E. Williamson detailed how it can reduce transaction costs by preventing opportunistic behavior between separate firms in the supply chain. This insight is particularly important in digital markets, where platforms invest heavily to maintain both high-quality services and strong relationships with customers. Third-party suppliers have an incentive to free ride on these investments, which diminishes investment incentives. For instance, as ITIF has explained, Apple’s “closed” mobile ecosystem is distinguished by privacy and security features that require continuous investments to maintain. The incentive to sustain and enhance these features is greatly diminished if third-party developers can free ride without paying commissions to Apple or following its ecosystem guidelines. Apple’s vertical integration of consumer services and restrictions on third parties help internalize the cost of platform maintenance.

Enforcers have also raised concerns about vertical exclusionary behavior, notably self-preferencing, which involves a platform promoting its own competing services in ways that it does not offer to third parties. For example, the European Union’s Digital Markets Act (DMA) targets American digital platforms such as Alphabet, Amazon, and Apple by designating them as “gatekeepers” and makes certain practices, like self-preferencing, off-limits through per se bans. However, self-preferencing is often procompetitive, increasing interbrand competition and offering low-cost alternatives alongside premium brand services. Now that the DMA has been in force for over a year, we know what a ban on self-preferencing looks like in practice. ITIF has documented significant drawbacks, including the fact that “search traffic has been directed to Google’s large intermediary competitors often at the expense of the small businesses (e.g., restaurants and hotels) that EU competition policy is ostensibly designed to protect—potentially resulting in as much as a 50 percent reduction in their online traffic and millions of euros in revenues.”

To be sure, self-preferencing can indeed be harmful to competition and should not always be permissible. The European Commission has successfully challenged such conduct in the Google Shopping case, where Google was found to have artificially downgraded comparison-shopping competitors in generic search results. This conduct could reduce consumer benefits from the platform and diminish incentives for third-party entry. However, self-preferencing is a general and common business practice that produces procompetitive benefits so often that an outright ban makes little sense, especially when existing antitrust enforcement tools are already capable of challenging its anticompetitive forms.

Despite the harm caused by misguided EU competition policies, some U.S. lawmakers seek to follow the same path by passing the American Innovation and Choice Online Act, which would implement DMA-style regulation in the United States. Although the bill previously failed, it could likely be reintroduced. As ITIF has noted, such sweeping legislative interventions are warranted only in the face of persistent market failure, which is clearly absent in dynamic U.S. digital markets. Furthermore, bans on practices like self-preferencing would restrict extremely common procompetitive business conduct in ways that would radically reshape the digital economy. Congress should avoid the same misguided thinking about vertical conduct that not only led to the bankruptcy of iRobot—an American company in a strategically important industry now being sold to a China-based entity—but is also wreaking havoc in Europe under the DMA.