The Neo-Brandeisians and Chicago School have employed different statistics to show respectively that markets are both concentrated and not concentrated, leaving Americans in the dark about who really holds corporate power. The evidence only looks contradictory due to the failure of American antitrust scholarship and regulation to understand how conglomerates collect market power across many neighboring markets, through what Paul Friederiszick calls “adjacent market takeovers.” This strategy can make firms more efficient, but it can also raise prices and weaken competition even as each individual market still looks competitive, he writes, drawing on his new paper, “The Conglomerate Power Puzzle.”
Since the Neo-Brandeisian movement emerged about a decade ago, a debate that began with the Sherman Act of 1890 has reignited: Does corporate America have a market power problem? Economists have drawn opposite conclusions when assessing the status of the same economy. One camp points to aggregate numbers, where a shrinking set of giant firms take a growing share of the nation’s sales and profits. This is the world that former Federal Trade Commission Chair Lina Khan and former Special Assistant to the President for Technology and Competition Policy Tim Wu have tried to describe. In this world, the growth of big business has suffocated competition, innovation, and choice and contributed to the decline of workers’ share of income.
The other camp, rooted in the Chicago School that emerged in the 1980s, looks inside individual markets. It finds competition about as vigorous as ever and concludes that what looks like the effect of market power is actually mostly the consequence of the best firms winning fairly in competitive markets. These “superstar” companies are able to pull ahead through technological superiority and efficiencies of scale in what this camp calls “competition in action.”
In my latest research, I argue that both camps are right with respect to their reading of the empirical facts. The contradictory diagnosis forms the heart of what I label America’s “conglomerate power puzzle”: zoom out, and the economy is plainly more concentrated. More than three quarters of U.S. industries have become more concentrated since the late 1990s. However, zoom in to almost any single product market and it looks competitive. That apparent contradiction is a symptom, not a quirk, of looking for corporate power in the wrong place. The antitrust laws and analysis meant to check it do not serve their purpose. I argue that the evidence—high market concentration when aggregated and low market concentration when disaggregated (the conglomerate power puzzle)—can be explained by a situation in which small to medium-sized shares in many different markets are collected by the very same companies. Concentration at an aggregated, industry- or even economy-wide level increases, as a few firms build a larger and larger footprint. At the same time, concentration at the market level remains low. The problem this finding reveals is that market power does not, or at least not exclusively, reside inside any single market, which is why our regulators often keep missing it. The big question is what this paradox means for social welfare and why American antitrust regulation has overlooked it.
The missing focus on non-dominant conglomerates
Antitrust regulation, particularly in Europe, has begun to catch up with firms that leverage their holds in multiple markets to protect themselves from competition. These “ecosystem” theories of harm are particularly interested in Big Tech companies like Amazon and Alphabet that provide services in an array of digital markets. Amazon provides services in online marketplaces, cloud computing, video, and brand-name home goods. Alphabet provides internet browsing and search, mobile phone operating systems, and a digital advertising marketplace. As recent monopoly cases against Alphabet in the U.S. showed, these services incentivize users to use the companies’ other services, making it harder for rivals to enter certain markets.
However, ecosystem theories of harm generally focus on dominant firms that use their large position in a core market to force their way into adjacent markets. For example, critics fear Google may leverage its search data to secure a leading position in the market for foundational artificial intelligence models. I show these anticompetitive concerns also arise among firms that cannot be described as dominant in any market, and which leads both Neo-Brandeisians and proponents of the Chicago School to overlook their activity. Further, this activity is not just happening in digital markets but also pharmaceuticals, healthcare, and basic household goods.
To be fair to the Neo-Brandeisians, they may have seen part of this. They warned about conglomerates and serial acquisitions, and the 2023 U.S. Merger Guidelines, written under their influence, even allow enforcers to examine serial acquisitions. Furthermore, their regulatory activity while in power under President Joe Biden was hampered by the current legal paradigm’s focus on individual product markets (this is the paradigm advanced by the Chicago School). Nevertheless, their attention still went mostly to the dominant Big Tech firms. What they did not focus on are the modest market positions the same large firms hold in many markets at once, where no single share looks dangerous and the problem only appears once you add up who owns what across markets. This is how these powerful firms grow: sideways, acquiring companies across dozens of adjacent products and services, one modest deal at a time. Each target is close enough to strengthen the conglomerate but too far apart to count as the same market. So the authorities wave the deal through. I call this an “adjacent market takeover strategy.”
How non-dominant conglomerates harm competition
To see how these non-dominant conglomerates impact social welfare, consider the two separate civil antitrust actions filed in 2019 and 2020 by coalitions of U.S. attorneys general against major generic drug manufacturers, including Teva, Sandoz, Mylan, Taro and Perrigo. The complaints allege that these companies fixed prices, divided up customers, and rigged bids for more than 100 generic drugs sold throughout the U.S. This alleged conspiracy is among the largest price-fixing cases in American history. Just this month, Sandoz agreed to pay $400 million to settle the states’ claims. The first trial against the other drug manufacturers is set for February 2027. What makes this alleged cartel unusual is how it was built. Many of those firms had grown by acquiring adjacent product lines until the same handful of producers ended up facing each other in many markets at once (often referred to as multimarket contact). When Sandoz, one of the alleged cartelists, bought the dermatology producer Fougera in 2012, it subsequently overlapped with the same rivals in dozens of creams, ointments, and lotions. According to the complaint, competitors then approached Sandoz as a “strategic opportunity to collude on overlapping products.” Sandoz even built an internal database to track where winning customers for one drug might provoke retaliation in another. Undercut a rival on one drug, and you invite retaliation across all the others you both sell. As one executive from one of the alleged cartelists put it: “We have a lot of products crossing with Mylan [a rival firm] right now, I do not want to ruffle any feathers.”
This is partly a kind of collusion that is referred to as “tacit collusion,” meaning that no explicit communication on prices was needed by the alleged cartelists. Rather, the unique industry structure where the same conglomerates consistently came in contact with one another enabled them to cooperate rather than compete. The adjacent market acquisitions that built this structure passed unnoticed through merger review because each, on its own, barely moved any single market’s shares. Yet, according to an economic study by Amanda Starc and Thomas Wollmann, the prices of cartelized generic drugs increased by about 50 %, making it even more difficult for patients to afford necessary medication.
Collusion is not the only harm caused by adjacent market takeovers, and the harm from adjacent market takeovers does not even require multiple firms. Bundling and tying are one of the oldest concerns (“theories of harm”) in antitrust. Consider Regeneron v. Amgen, in which a federal jury in Delaware found in 2025 that Amgen had violated the antitrust laws. In this case, Amgen did not develop its two must-have drugs for inflammatory diseases, Enbrel and Otezla. It bought them, each in unique markets separate from those in which it currently operated, and so each acquisition passed merger review. Only afterwards did Amgen turn them into a weapon. It told pharmacy benefit managers that they would only receive full rebates on these must-have drugs if they favored Amgen’s own cholesterol drug. Regeneron’s competing cholesterol drug was effectively shut out. Regeneron could not fight back, because it had no large portfolio of its own to retaliate on different markets.
Regeneron v. Amgen in particular shows how these non-dominant conglomerates can leverage the unique market power gained from adjacent firm takeovers. This and the previous pharmaceutical price-fixing case show why the “competition in action” camp, right as far as it goes, cannot settle Neo-Brandeisian concerns about the competitive harms of big business. That the best firms do win more, and that technology rewards scale, is a claim about what happens inside a market. It says very little about a firm that grows across neighboring or adjacent markets.
On the other hand, the Neo-Brandeisians never articulated the framework to show how conglomerates that did not dominate any single market could still lead to rising market concentration and opportunities for anticompetitive behaviors like tacit collusion or unfair practices. Without that framework, their concern remained a conviction, and in court, convictions under current U.S. law tend to lose to market definitions. Perhaps the biggest example of this was their case against Meta, although the case was initially filed and later lost under the first and second Trump administrations, respectively. The case collapsed in November 2025 largely because the judge drew the market differently than the agency had.
Conclusion
For forty years we have asked a single question: Does market concentration raise prices? It is the consumer welfare test at the heart of the Chicago School. But the danger of a few firms amassing power across the economy is not reducible to any single price. Competition becomes more fragile, new entry difficult. Economic power builds up due to adjacent market takeover strategies and translates into political power. An economy can look competitive in every market a regulator examines and still slide toward a handful of firms with large portfolios across many adjacent markets. This is the intuition the Neo-Brandeisians reached for but could not operationalize. Adjacent market takeovers may be the mechanism they were missing. The United States, I conclude, suffers from a conglomerate enforcement gap.
An economy without adjacent market takeovers carried out by the same conglomerates that buy up all the neighboring markets would not be an economy without big firms, or even without (temporary) monopolies. A company with a better product would still win, sometimes win everything, and earn monopoly profits for a while. Then someone new with a better idea would come for it. Monopolies would rise and fall. That is what competition and capitalism are supposed to look like.
However, this is not the economy we have. The same few conglomerates hold significant market shares in hundreds of markets. If one of them loses a market, it barely matters to its leverage, because it still sits in all the others. This would not be a problem if those non-dominant market positions were held by different firms. One company with 10% market shares in a hundred neighboring or adjacent markets is a different economy from a hundred companies with 10% market share each. In the economy we have, there are more barriers to entry and weaker incentives to innovate and develop new products rather than acquire them. Prices tend to be higher, as business partners are at a disadvantage in terms of negotiating power. Creative destruction slows as temporary monopoly power stops being temporary. The way we measure market concentration in our economy needs to shift from looking only inside single markets to also studying how firms reach across multiple markets. Enforcers should track concentration across whole sectors. Every merger review should not only measure the change in concentration within a single market, but also the change across the whole sector (the “sector delta”), even when the deal creates no dominant position in any single market. Adjacent market takeovers and the market-forming activity of conglomerates, both dominant in single markets and non-dominant, should be included in the merger guidelines. They should go further than the current 2023 U.S. Merger Guidelines and spell out when adjacent market takeovers become dangerous even if the market concentration in no single market is affected. This would pull the mask off those firms that really hold economic power.
Author’s Disclosures: The author reports no conflicts of interest. You can read our disclosure policy here.
Articles represent the opinions of their writers, not necessarily those of the University of Chicago, the Booth School of Business, or its faculty.
Subscribe here for ProMarket’s weekly newsletter, Special Interest, to stay up to date on ProMarket’s coverage of the political economy and other content from the Stigler Center.
