In new research, Filippo Lancieri and Tommaso Valletti analyze the shortcomings of the current merger review system and defend stronger rebuttable structural presumptions as an important step forward.
Companies acquire or merge with each other for a variety of worthwhile reasons, ranging from managing uncertainty to gaining scale to accessing new resources. They also do it for less desirable reasons, including to gain market and political power, raise barriers to entry, or extract higher rents from consumers, workers, or suppliers. We must separate the good from the bad cases, and that is why most market economies require certain mergers to be vetted by antitrust authorities before they can be completed. In a new article, we analyze the shortcomings of the current merger review system and defend stronger rebuttable structural presumptions as an important step forward.
Defining the scope and rules of merger review is a complex task and more art than science. For example, countries differ significantly on grounds for blocking mergers—some focus mostly on prices, while others allow various public interest considerations to guide these decisions. Policymakers must also establish the thresholds that trigger mandatory merger notification, and these thresholds vary in size (e.g. turnover above a certain value) and in nature (e.g. market share of a certain percentage).
The decisions about which mergers to vet can have significant consequences. For example, by some estimates, almost 350,000 M&A transactions occurred in the United States over the past 20 years, ranging from small “acquihires” to very large and complex multi-billion-dollar mergers that spanned the globe. In the same period, the U.S. Justice Department and Federal Trade Commission received 31,500 notifications for transactions above legally established thresholds. About 970 transactions received a “closer look,” and roughly 300 were either abandoned or blocked. In other words, two sophisticated and well-resourced authorities became aware of only nine percent of U.S. M&A transactions over the past 20 years, scrutinized 0.3 percent, and blocked 0.1 percent. The figures are similar in Europe. Meanwhile, industrial concentration, mark-ups, and profits—proxies for market power— rose across both economies. Productivity growth stagnated.
Weak antitrust enforcement is not solely responsible for these developments, but it is certainly part of the story. Indeed, analyses usually find that mergers increase prices and decrease output. A recent literature review of post-merger studies that focused only on vetted mergers (that is, cases approved by antitrust authorities) found price increases in almost half of the analyses. A growing number of studies also show significant consumer harm resulting from mergers just below the mandatory notification thresholds. It is hard to look at these and other data and analyses—such as those showing the growing use of delay and obfuscation by antitrust defendants—and believe that merger review is working as intended.
Our article proposes using strong, rebuttable, structural presumptions in merger review to help reform this problematic system. These presumptions hold that all transactions above a given threshold are prima facie illegal but allow merging parties to rebut this conclusion with evidence of countervailing benefits.
These structural presumptions are grounded in solid economics. In the absence of efficiencies, mergers among rivals (and, in particular, horizontal mergers) typically lead to increased market power, higher margins above costs, and consumer harm. Recent work has shown that mergers can also lead to less investment in innovation, harm to workers, or increased lobbying expenditures that strengthen companies’ political power. A structural presumption acknowledges these facts, and raising the legal thresholds to approve mergers would help mitigate a lack of enforcement.
These rebuttable presumptions also change the practical dynamics of merger review. Antitrust authorities are severely underfunded (partially by design, as companies lobby to cut funding and weaken enforcement) and have much less information about a given market than do the merging firms in that market. Therefore, it makes sense to require those firms to prove that a merger will be beneficial. This reversal of the burden of proof would also dissuade firms from blocking or delaying review and would allow authorities to focus on what really matters—whether the parties presented solid evidence that the merger will increase welfare.
Finally, rebuttable structural presumptions protect pro-competitive mergers and do not punish efficient firms. They allow any firm to prove that a merger would benefit markets and the public. They also encourage large firms to grow organically, a process that can lead to new market entry and generally benefits consumers and suppliers.
We envision two approaches to implement these presumptions in practice: one incremental and the other a fundamental reform.
The incremental path would be to adjust merger guidelines. The revised 2023 U.S. Merger Guidelines lowered the burden of proof and tightened market-share thresholds, creating rebuttable presumptions for transactions that surpassed these thresholds. While this is an important step in the right direction—and one that should be followed by more authorities around the world—it also has downsides. First, these regulations can be changed by a new presidential administration. Second, and more importantly, these new guidelines still reflect a narrow focus that requires regulators to model the individual impacts of each merger, something that is usually done by defining a “relevant” antitrust market and then calculating transaction-specific market shares.
Defining “relevant markets,” however, is a complex exercise. For antitrust defendants, the path to approve a merger remains clear: hire many economic consultants to defend a broad relevant market definition, so that the structural presumptions are never triggered. Authorities, then, need to allocate their own scarce internal resources to produce studies or other analyses defending a narrow market definition that triggers the presumption. This game, however, is played on an unequal playing field—the merging parties have more resources to produce studies and control the information necessary to understand the market. Ultimately, authorities have less capacity to focus on what really matters: analyzing if the transaction will harm welfare because efficiencies do not outweigh increases in market power. This is a system that can continue to miss the forest for the trees.
Our preferred alternative is substantial reform that implements these structural presumptions at a less granular level and before a merger is reviewed. It would require all “large” firms to demonstrate that their M&A transactions will result in significant synergies shared with consumers (or consumers and suppliers). The presumption would apply to firms, not transactions.
Every metric that triggers this structural presumption will have advantages and disadvantages, just as the current merger notification triggers involve trade-offs and some arbitrariness (there is no scientific justification for why the size-of-transaction threshold in the U.S. is $119.5 million rather than, say, $75 million, or the current HHI levels in the guidelines). Policymakers can rely on a combination of firm-level and industry-level characteristics to establish this rebuttable presumption.
At the firm level, the presumption could be triggered by a combination of certain amounts of market capitalization, company turnover, and market share in broadly defined markets. This approach would be similar to Europe’s in the Digital Markets Act and the Digital Services Act. In practice, it would mean, for example, that the presumption would apply to all firms that had at least $X billion in annual turnover or more than $X billion in market capitalization or fair market value or that supplied at least X percent of goods or services to the total population of a given jurisdiction. Firms would notify the authorities that they met the criteria, and this would lead to a process that lasted for a certain number of years and could be renewed.
Thresholds would also be set at the industry level when indicators signaled the presence of very concentrated market structures. This would require antitrust authorities to collect and process industry-wide information on how different markets were structured, likely by using economic research units focused on tracking the performance of different sectors of the economy. These units would employ economists, statisticians, and other industry specialists and supply independent information to merger case managers. In practice, authorities would be defining aggregate levels of concentration according to industry metrics such as NAICS codes. The presumption would then apply to all transactions by firms holding more than X% of that industrial market. Again, firms would be notified in advance. These rougher codes are possibly less accurate than relevant antitrust markets, but they are also less costly to define, more consistent across different cases, and less subject to gaming.
Companies caught by either the firm or industry-level thresholds would be subject to mandatory notification of all their transactions and a prima facie presumption that these transactions would have anti-competitive effects.
This presumption could be rebutted by parties showing offsetting efficiencies. Competition policies generally do not account for or treat efficiencies in mergers well. That is probably because decisions about mergers are based more on relevant market definitions than on efficiencies, so less attention is devoted to the latter. Our understanding of efficiencies will improve as they grow in importance. In general, efficiencies must be merger-specific, verifiable and should not arise from anticompetitive reductions in output. Companies should show that they are unable to make the product or service in a reasonable way, that there are no alternatives less restrictive than the merger, and that the merger will create clear net-positive effects for consumers and, potentially, suppliers.
Finally, it is worth noting that this system would only apply to transactions above the thresholds established to trigger the presumption. All transactions below the thresholds would continue to be subject to the current antitrust system.
Merger enforcement always requires compromises between precision and ease of administration. The current approach, which purports to be precise, has produced systematic underenforcement. Adopting rebuttable structural presumptions would be an important step toward fixing the current system.
This article appeared first on Columbia Law School Blue Sky Blog under the title “Why Rebuttable Structural Presumptions Improve Merger Review.”
Articles represent the opinions of their writers, not necessarily those of the University of Chicago, the Booth School of Business, or its faculty.
