The Myth of Declining Competition
The rise of neo-Brandeisian antitrust enforcers under the Biden administration marked a significant victory for a growing progressive movement that sought to reframe the goals of antitrust enforcement toward protecting smaller competitors and to downplay efficiency and economic growth. This movement based its policy goals on claims of rising market power that gained popularity as several concurrent books and academic studies purported to show rising market concentration, price-cost markups, profits, and other metrics as proxies for the growth of allegedly harmful industrial monopolies and evidence of broadly declining market competition in the United States. However, this argument has proved faulty.
In its “Monopoly Myths” series of policy reports, ITIF has systematically debunked many of the claims made by antitrust populists who claimed that a reinvigoration of antitrust enforcement was needed to stop rampant monopoly power from strangling the economy. These reports demonstrated that the proposed connections between corporate concentration and higher markups, prices, and even inflation were erroneous. Contrary to President Biden’s claim that antitrust enforcement over the past four decades had been a “failed experiment,” competition has been alive and well in most industries, including the high tech industry, which has been the hobby horse of the populists.
Now, a recently published research paper by Carl Shapiro and Ali Yurukoglu provides a systematic review of the relevant technical literature and offers a tour de force against the remaining vestiges of the declining competition theory. It lays out in full detail the many limitations of using aggregate statistics, what the authors call “evidence at scale,” to draw economy-wide conclusions when market-level analysis is really needed to say anything useful about competition. In fact, even when overall levels of variables like concentration or markups are increasing, there is no established way to determine whether this is due to a harmful decline in competition or to beneficial competition in action. It could be the former in one market, and the latter in another.
Antitrust populism gained momentum amid a growing consensus around the claim that market concentration overall had been increasing in the U.S. economy over the past several decades. A study by the Obama administration’s Council of Economic Advisers (CEA) initially sounded the alarm about rising industrial concentration using Census Bureau data on concentration in industries defined using the North American Industry Classification System (NAICS). Although this research made a big splash, Shapiro and Yurukoglu note that data based on NAICS industry codes, such as the data in the CEA report, are poorly suited to measuring market power because they group together products that are not close substitutes and are defined around production categories rather than consumer substitution patterns.
For instance, the Census data group all “Software Publishers” under NAICS code 513210, which includes mobile apps as well as desktop operating systems, even though these clearly don’t compete with one another from a consumption standpoint. As Shapiro and Yurukoglu bluntly summarize, “the measures of concentration typically cited are not informative regarding market power. Period.” They further note that research assessing concentration based on more narrowly defined markets of substitutes finds lower or declining levels of concentration. A recent study by ITIF also shows that incorporating data on net imports into the concentration analysis tends to reveal much lower concentration in industrial categories.
Shapiro and Yurukoglu also critique the literature that computes statistics on price-cost markups to show an economy-wide increase in market power, a claim that became a linchpin of the declining competition theory. The headline claim from an influential 2020 paper by De Loecker, Eeckhout, and Unger is that aggregate markups had risen from around 20 percent to over 60 percent in the decades since the 1980s, a staggering increase. However, Shapiro and Yurukoglu highlight serious measurement and interpretation issues in the paper’s methodology. In particular, the observed increase in estimated markups may better reflect factors like operating profits, industry composition shifts toward intangible-intensive sectors, globalization, and the growth of “superstar” firms rather than reduced competition.
Of particular concern is recent work demonstrating that unstated sample restrictions in the De Loecker et al. paper excluded 27 percent of the available observations. Including that missing data shows that markups in the sample are being driven almost entirely by the finance and insurance (F&I) sector. As the cited study highlights, “Excluding F&I, the average markup increases only modestly, from just below 1.30 in the mid-1980s to about 1.35 in 2016.”
Some studies also suggested a link between increasing concentration and profits, which were widely cited as further evidence of growing market power. However, Shapiro and Yurukoglu note that corporate profits as a share of GDP have risen since the 2000s, but much of this increase is attributable to the foreign earnings of U.S. multinational firms, while domestic profit shares are comparatively flat. As they explain, “under the hypothesis that U.S. antitrust policy was too lax over the past 40 years, one would expect to see a significant increase in domestic profits relative to GDP. But the time series of domestic profits over that time period is significantly flatter than the time series of total profits.” This finding conforms to ITIF’s own findings that—when focusing on domestic nonfinancial profits as a share of net value added—profit shares have declined over the long term and remain lower today than they were in the 1960s. Much of the apparent rise in profits reflects foreign earnings or methodological choices rather than weakened competition.
Given claims from prominent commentators that “decades of underenforcement” against mergers have resulted in declining competition, Shapiro and Yurukoglu review the literature on merger retrospectives to see whether mergers have helped or harmed competition. They find that only a small share of mergers has been carefully studied, and the research shows clear problems only in certain sectors, such as hospitals, where mergers have sometimes led to higher prices. In many industries, however, results are mixed—some mergers raised prices, others lowered them, and many had varied effects. The upshot is clear: Merger policy should remain attentive and targeted, but with greater reliance on industry-specific analysis rather than sweeping changes to enforcement rules.
While some commentators attack this new paper as ignoring the reality that exists in certain markets, that stance misses the paper’s central point: Individual market-level analysis is needed to determine whether competition has been lessened. Evidence at scale can be an indicator that deeper analysis is needed in some markets, but it is not dispositive on its own and should not be the sole basis for overhauling antitrust policy.
