No, the Evidence Does Not Suggest That Competition Has Declined and Antitrust Policies Need an Overhaul

Neo-Brandeisians and anti-corporate activists repeatedly assert that antitrust policies have failed in recent years, leading to a decline in competition as a few firms gain increasing market shares. They often point to studies showing rising concentration, markups, and profit trends to support their assertion. However, a recent paper by Carl Shapiro, an economist who previously served in the President's National Economic Council from 2011 to 2012, and Ali Yurukoglu, a professor of economics at Stanford University, found that despite some empirical studies showing these trends, the effects on competition may be either unknown or positive. As a result, current antitrust enforcement policies may not be a failure after all, and policymakers should thus be skeptical of neo-Brandeisian efforts to radically reform antitrust laws.

First, although some studies have found that concentration has risen when analyzing Economic Census data, related economic literature shows that competition may have nevertheless increased. According to Shapiro and Yurukoglu, a study by Benkard et al. using data that more closely aligned with relevant antitrust markets (products are more likely to be substitutes) concluded that across markets in their dataset, the median market share of the four largest firms fell from 77.5 percent to 72.5 percent from 1994 to 2019-a fall in market concentration-and that concentration in the 50^th, 75^th, and 90^th percentiles of the most concentrated markets also dropped significantly. Relatedly, Benkard et al. also found a reallocation of market shares between incumbent firms in conjunction with an overall decline in the Herfindahl-Hirschman Index (HHI).

Second, Shapiro and Yurukoglu note that even if concentration were to rise, competition and consumer welfare might still increase if the cause of rising concentration is due to firms becoming more efficient. Indeed, the authors asserted that a 2021 study by Sharat Ganapati found "a robust positive correlation over time between concentration and measures of productivity and between concentration and measures of real output." As such, increased concentration may not be due to anticompetitive behavior but procompetitive efficiencies that drive competitive advantage among leading or "superstar firms."

Third, despite critics of modern antitrust claiming that a rise in profit rates is a sign of declining competition, the reality is that domestic profit rates (used to measure domestic competition) have remained relatively unchanged. As Shapiro and Yurukoglu find, a significant portion of the rise in corporate profits as a share of GDP is due to an increase in income earned from foreign operations. Specifically, while domestic profits as a share of GDP were basically unchanged from 1960 to 2020 (moving from approximately 10 to 10.8 percent), foreign profits rose from less than 1 percent to almost 4 percent in the same period. In other words, if increased profits are a sign that competition is lessening, then it's happening abroad and not in the United States.

Fourth, although the neo-Brandeisians and their allies often like to claim that rising markups indicate that competition has declined, the effect of greater markups on competition is difficult to discern, given the challenges of measuring marginal costs. As the Shapiro and Yurukoglu paper highlights, marginal costs cannot be directly observed and have to be estimated either using a demand approach (inferring marginal costs from a consumer demand system) or a production approach (inferring markups based on the observed amount of inputs and outputs).

However, Shapiro and Yurukoglu identify multiple problems with the application of the production approach, which is used by a widely cited paper by De Loecker et al. (DLEU) showing that the average revenue-weighted markup rose from 1.21 to 1.61 from 1980 to 2018, writing that the "application of the production approach to Compustat data departs in important ways from the idealized setting…most firms, especially large ones, sell multiple products…[but DLEU] treat each firm in Compustat as a single product firm." On the other hand, they also found that the demand approach cannot accurately measure markups since estimating a demand system requires unobserved variables, such as unobserved product characteristics.

Lastly, Shapiro and Yurukoglu discuss whether lax merger enforcement may have been a cause of supposed decreased competition, as has been long alleged by Kwoka and others, and relatedly the extent to which mergers actually result in efficiency benefits. On this score, Shapiro and Yurukoglu conclude from merger retrospectives that while some mergers may lead to less competition, other mergers can result in greater competition and lower prices. For example, the authors note that while the Evanston/Highland Park hospital merger increased prices by 20 percent, the Provena/Victory hospital merger decreased prices by 4 percent. Similarly, with respect to airlines, the authors note that although a study by Luo found that the 2009 Delta/Northwest merger raised ticket prices, another study by Das found that the 2013 American Airlines/US Airways merger led to price decreases. At bottom, these findings are consistent with the modern antitrust consensus regarding merger enforcement, which attempts to carefully weigh possible anticompetitive harms against cognizable procompetitive efficiencies, as opposed to the neo-Brandeisian approach of tipping the scales against mergers, as reflected in the new 2023 DOJ-FTC Merger Guidelines.

Shapiro and Yurukoglu's paper is consistent with a 2021 analysis conducted by ITIF. Indeed, an ITIF report analyzing Economic Census data found that concentration remained roughly the same, increasing only by 1 percentage point from 2002 to 2017. But more importantly, the most concentrated industries became less concentrated. Another ITIF report on profit rates showed that domestic profits as a share of GDP have declined from 10 percent to 7 percent from 1949 to 2019, which is consistent with increased competition. This report also found that "much of the modest increase in profits in the last two decades has come from foreign profits of U.S. firms (something that is not a U.S. antitrust issue)." Finally, a third ITIF report noted challenges in measuring markups and found that the top 10 percent of firms with the highest markups are 30 percent more productive, suggesting that rising markups were consistent with increased competition. Moreover, the report noted that the "increase in markups appears to be a reflection of increased fixed costs, especially intangible costs such as R&D and design" rather than anticompetitive practices that lead to less competition.

In sum, policymakers should reject neo-Brandeisian efforts to radically change U.S. antitrust laws under the false pretense that competition has declined in recent decades. As Shapiro and Yurukoglu's paper makes clear, concentration and profit rate trends suggest that competition has increased and benefited America's economy with larger, more efficient firms. Indeed, if antitrust policies are to be reformed, they should arguably be more open to increased size, both to drive innovation and Schumpeterian competition, as well as to support U.S. competitiveness against China.

As a recent ITIF report shows, the United States is already falling behind China in advanced industries. Targeting the practices of large, efficient companies that are keeping the United States ahead will only worsen the competitiveness of these firms against their Chinese counterparts and put the United States at a disadvantage. This is especially critical because the Chinese government is increasingly encouraging their domestic firms to merge and become large and more efficient to outcompete U.S. rivals. In sum, American policymakers should do everything to support antitrust laws that not only benefit U.S. consumers but also do not hinder U.S. companies from vigorously competing against Chinese rivals.

Public link

Please use the above public link if you want to share this noodl on another website.