Increased Market Concentration Does Not Equal Less Innovation

What is the connection between industrial concentration and innovation? “Neo-Brandeisian” antitrust enforcers under the Biden administration argued that there is a negative relationship: Increasing concentration means less innovation. Their argument is closely analogous to their contention about the proposed connection between market structure and competition, itself a manifestation of the structure-conduct-performance (SCP) theory of industrial organization, which has long been considered antiquated by modern economists. This theory views an industry's conduct and performance—such as prices, output, and innovation—as a direct function of its structure, specifically the underlying levels of concentration or the number of firms in the industry.

In attempting to establish the connection empirically, statistical studies regressed measures of variables supposedly indicative of monopoly power, such as profits or prices, on a measure of concentration, finding mixed, inconclusive results. The economics profession has long moved away from SCP theory, recognizing that concentration is often the result of the lower costs and increased efficiency of larger firms in many markets rather than abusive or anticompetitive conduct. In other words, industry structure is often “endogenous”—it is an outcome of effective market competition.

When it comes to innovation, the neo-Brandeisians’ claim that concentration reduces innovation reflects a fundamental misunderstanding of the relationship between these two variables. Indeed, as economist and UC Berkeley professor David Teece has noted, “Innovation drives competition as assuredly as competition drives innovation,” confirming once again that competition is not necessarily an independent variable in this analysis. Furthermore, Carl Shapiro, another Berkeley economist and former antitrust official at the Department of Justice, has clearly explained how studies of this relationship can yield conflicting results depending on their operating definitions of what competition means.

For example, while neo-Brandeisians did not adopt the classical perfect competition model as an ideal, they strongly advocated for a deconcentration antitrust policy that reflected their core anti-corporate animus. When it came to the innovation debate, they found an ally in Kenneth Arrow, who demonstrated theoretically that a pure monopolist would have little incentive to innovate because any new innovation would simply replace the rents the monopolist already earns from its exclusive access to sell in the market. (This concept is known as Arrow’s replacement effect.) In other words, a pure monopolist is unlikely to engage in significant innovative activity.

According to the view made famous by Joseph Schumpeter, however, market concentration and market power are not the antithesis of competition. Rather, firms with some market power are often the only ones with both the incentive and ability to innovate, whether by developing new products or pioneering production processes that disrupt the status quo of the marketplace. A strong monopoly position is what drives entrepreneurs to innovate by facilitating a firm’s ability to appropriate investments in research and development. In the Schumpeterian framework, more concentrated industries should thus be more innovative. Moreover, in dynamic, technology-driven industries, so-called monopolies are at constant risk of what Schumpeter called creative destruction.

The economic literature provides extensive evidence for an “inverted U” curve describing the relationship between market power and innovation. The seminal 2005 paper by Aghion et al. presented data showing that increasing market power—the extent to which an industry’s output is priced above its costs—is associated with greater innovation until it reaches a maximum point. In the study, innovation is measured by the number of patent filings associated with an industry in the same year. It is broadly consistent with a “weak” Schumpeterian view since firms with greater market power and close competitors have strong incentives to innovate. However, monopolistic firms with secure protection from any close competitors do not.

The paper’s model suggests that innovation declines when one firm dominates the market—particularly when it feels secure from competition, perhaps through government-erected barriers. However, when firms experience close competitors, even in a highly concentrated oligopolistic industry, the rate of innovation intensifies. Moreover, the sort of monopolistic competition favored by Arrowians or neo-Brandeisians is not ideal from an innovation perspective. Indeed, recent additions to the literature confirm the role of large incumbents in driving innovation and economic growth. Sustaining the rate of technological advancement and innovation requires the scale of larger enterprises, which can mobilize the teams and capital integral to research and development.

While the debate between Schumpeter and Arrow will remain ongoing, the inverted U relationship between market power and innovation shows that a simplistic deconcentration merger policy, like that pursued by neo-Brandeisians, will not maximize innovation. If antitrust enforcement is serious about promoting innovation as a key policy goal—as it should be—it certainly should not deter industry concentration for its own sake but instead focus on conduct that explicitly violates the antitrust laws. Hopefully, the new agency officials in the Trump administration will take a more innovation-friendly approach to enforcement, which would be a much-needed boost to the American economy.