The court concluded in this case, which plaintiff Students for Fair Admissions has appealed, that relying solely on a subset of the information that USNA considered in its holistic admissions reviews likely leads to overstated estimates of the impact of race and ethnicity on USNA’s admissions decisions. The ruling explores the bounds on the probative value of statistical evidence of discrimination. The remainder of this article draws lessons on what those bounds are from which econometric arguments did and did not appeal to the judge.

An oversimplified regression model of admissions cannot reliably estimate the contribution of race to admissions decisions

Multiple regression analysis is a statistical technique used to quantify the relationship between an outcome of interest and several explanatory variables. In the context of race-based discrimination litigation, the outcome of interest might be college admissions, loan or hiring decisions, or quality of medical care. By including race alongside sufficient other relevant factors, researchers can estimate the independent effect of race on the outcome of interest. In SFFA v. USNA, the key statistical question was whether the regression model provided by SFFA’s expert included enough other relevant factors to isolate the effects of race and ethnicity on the Naval Academy’s admissions decisions from the effects of variables such as socioeconomic background.

USNA acknowledged considering race and ethnicity as one of many factors in its holistic review of applicants’ files. SFFA’s expert attempted to quantify the impact of race and ethnicity on admissions decisions using a logit regression model that explains the probability of admission with several factors, including race and ethnicity. Secretariat Managing Director Dr. Stuart Gurrea, serving as a testifying expert for USNA, argued that this model oversimplified the complex admissions process and produced unreliable estimates of the impact of race and ethnicity. The court agreed with this assessment, finding that SFFA’s model failed to account for many other factors the Academy considered in its admissions process.

Difference-in-Differences (DiD) analysis has been a popular method in econometrics for estimating causal effects and is often employed in antitrust litigation. The essence of DiD lies in comparing the changes in outcome variables of interest (e.g., price) over time between a group that is exposed to the alleged anticompetitive conduct and a control group that is not (e.g., comparing different groups of consumers, different firms, or different geographic regions). It gets its name “difference-in-differences” because it essentially combines two types of variation—the first from a before-and-after analysis and the second from comparing an affected and an unaffected group.

The key advantage of DiD is its ability to control for time-invariant unobservable factors that may influence the outcome of interest. By differencing out the common time trends between the groups that are and are not affected by the anticompetitive conduct (i.e., “treatment” and “control” groups), DiD isolates the treatment effect by focusing on the differential changes in outcomes that occur after the introduction of the treatment. The DiD methodology has been implemented in antitrust analyses in various settings.² In merger analysis, for example, DiD has often been implemented to estimate retroactively the impact of past consolidations to inform future policy.³

Despite its strengths, DiD is not immune to potential biases. Choosing the right quantitative tool, such as DiD, in an antitrust setting involves careful consideration of various factors to ensure the validity of the causal inference. Under the Daubert Standard,⁴ it is important for an expert to demonstrate the adequacy of a chosen tool, such as regression, and the appropriateness of a chosen research design.⁵ Since biases in the canonical DiD may arise from the violation of distinct conditions, there is no single recipe solution, and experts need to carefully analyze the case in question.⁶

There may be situations where a simple pre– and post-treatment formulation is not enough to capture the dynamics. For example, a company’s pricing policy may go into effect in distinct regions at different times as opposed to being simultaneously launched. There might be a need to study the effect of successive acquisitions by the same company in different markets. A firm may choose to implement a new policy to distinct groups of stakeholders at different times. As in these examples, the resulting bias of the estimates obtained by applying the standard DiD will be particularly problematic when there is heterogeneity in the treatment effect over time. However, there have been a few methodological alternatives proposed in the literature,⁷ some of which have been used in litigation.⁸ One could, for example, use a matching algorithm in each period to pick the best control group (where only those units that are untreated in that period are candidates),⁹ and once the control groups are selected, proceed as usual.

DiD also requires the treatment and control groups to have similar trends over time in the absence of the alleged anticompetitive conduct. In practice, this means that, absent a merger, and with everything else held constant, prices in markets where both merging parties are present (treatment group) and markets where at least one of them is not (control group) would have trended in a similar fashion. A violation of this assumption need not be the end of DiD analysis, but it does require one to adjust one’s specifications, as the regression will no longer produce consistent estimates merely by incorporating time-independent variables. If this violation of the parallel trends happens due to an observable factor, it is possible to extend the assumption by conditioning on variables that are observable pre-treatment.¹⁰

Most DiD literature imposes the requirement that potential outcomes of a unit are unaffected by the treatment assignment of other units – in other words, the variable of interest for that unit only depends on whether that unit and that unit only has been exposed to the anticompetitive conduct, which guarantees independence and essentially rules out any spillover effects. In our earlier example, customers can only be affected if the conduct has occurred in their market, but ought to be unaffected otherwise, all else held constant. However, it is possible that, if individuals are connected by a network, there may be spillover effects. A growing literature has already accounted for some extensions of the general framework that account for these network effects,¹¹ but there will likely be many more developments in this area, which may particularly impact how antitrust litigation views competition when platforms are involved.¹² For example, one might consider how changes in Gen AI policy that are applicable only to European markets start affecting the way companies conduct business in the United States, despite the absence of any such policy change in the United States.

In conclusion, DiD remains a valuable tool for estimating causal effects, offering a quasi-experimental approach to understanding and estimating the economic implications of alleged anticompetitive practices. Recent econometric developments have significantly enhanced the method’s applicability, addressing concerns related to control group selection, unobserved heterogeneity, and group trends. By incorporating appropriate adjustments to their DiD specifications, antitrust experts can improve the robustness of their estimates, ensuring that antitrust enforcement remains grounded in sound economic principles and evidence-based reasoning. As econometrics continues to evolve, it is paramount that practitioners stay up to date with state-of-the-art quantitative techniques, allowing DiD analysis to contribute to more accurate and reliable causal inference in antitrust cases.

---

However, more generally, MFNs can also apply to vertical agreements between suppliers and buyers, where, for example, a seller promises a buyer that the buyer will always be offered the lowest price offered by the seller.² While the exact details of these provisions differ by contract, parties, and setting, MFN clauses generally require that one party to the transaction not offer better contractual terms to any other party.³

With the rise of technical platforms, an entity that facilitates interaction/transactions between one or more groups of users (e.g., consumers and suppliers),⁴ MFN clauses have made their way into agreements between platforms and platform participants. These are known as platform MFN (PMFN) clauses. Generally, PMFN clauses are imposed by the platforms on the sellers/suppliers and prohibit sellers/suppliers from offering buyers/consumers products or services more favorably (e.g. lower price, better offering) on any other platform or distribution channel.⁵ PMFNs can vary based on the reach of the provision. A “narrow PMFN” prevents a seller/supplier from offering more favorable products or services using its own distribution channel, while a “wide PMFN” extends this prohibition to all other platforms, in addition to the seller’s/supplier’s own distribution channel.⁶

A key difference between classic MFNs and PMFNs is that the platform is not purchasing a service or good from a seller/supplier; rather, the platform is paid a cut from sales that occur on the platform. Thus, rather than a traditional MFN restricting the price at which a supplier can sell to a buyer’s competitors, a PMFN puts a floor on the price the participants on the seller-/supplier-side of the platform can charge to the consumers through competing distribution sources.⁷ Given the nuances that distinguish PMFNs from MFNs, economists have added to the general MFN research with literature on PMFNs investigating both the potential procompetitive and potential anticompetitive impacts of those PMFNs.⁸

A concern for antitrust litigators and regulators is the potential for PMFNs to reduce price and/or product competition. A PMFN imposed by one platform may restrict a seller’s/supplier’s ability to lower the price (or vary products) to buyers/consumers on competing platforms. Without being able to offer lower consumer prices (or variation in products), platforms may struggle to differentiate themselves from a dominant platform in such a way as to compete effectively and attract enough consumers to survive and become a successful platform. Reduction in competition could then lead to higher platform fees and consumer prices.⁹ A PMFN may further reduce competition by lowering incentives for potential entrants to join the marketplace at all,¹⁰ which may also result in more concentrated markets.¹¹

These antitrust concerns have piqued interest in PMFN policies in both litigation and regulation. A notable example is the regulation of PMFNs in the hotel booking space by France, Italy, and Sweden in April 2015, which led to Booking.com and Expedia (the two largest online travel agency platforms) to restrict “wide” price parity clauses within the E.U.¹² Later, France prohibited all price parity clauses in for French hotels in July 2015, and Germany prohibited all price parity clauses—wide and narrow—for Booking.com in December 2015.¹³ As a key example of PMFN litigation, in 2021, a class action was brought against Amazon and the “Big Five” book publishers accusing them of colluding to fix the price of ebooks at artificially high rates using MFN clauses.¹⁴ This case closely mirrored a 2011 case against Apple and the Big Five publishers, in which the Big Five settled and Apple lost at trial and was ordered to pay $450 million.¹⁵

Given past litigation and enforcement related to PMFNs and the general increased scrutiny in the Big Tech space, we expect that PMFNs will continue or increase in being an area of antitrust interest. Further, the litigation and enforcement, along with the existing literature, relating to the potential impacts of PMFNs highlight the importance of rigorous economic analysis and sound expert economic testimony to provide cases with clear conclusions on which side of the competitive scale the PMFN falls.

---

In Antitrust analysis in the United States, the Small but Significant Non-Transitory Increase in Price (“SSNIP”) test is often a key component of market definition analysis, whether performed quantitatively or qualitatively. In this analysis, an entity is hypothesized to be a monopolist with respect to a product or set of products that are under consideration to be a relevant market. The analysis asks whether this hypothetical monopolist could profitably impose and sustain a small, but significant non-transitory increase in price (often defined to be a 5% price increase). If the answer is “yes,” the set of products under consideration is a relevant market. If the answer is “no,” the set of products under consideration is expanded successively until the SSNIP can be profitably sustained. The SSNIP test captures economic considerations of substitutability and cross-elasticity of demand by considering potential substitution among competing products and may therefore be used to define both the relevant product and geographic dimensions of the relevant market.

While the SSNIP test has dominated antitrust analyses in the United States, other tests of market definition and market power have seen increasing use in the European Union. One example of an alternative to an SSNIP is the Small but Significant Non-Transitory Decrease in Quality (“SSNDQ”) test. This test is similar to the SSNIP test in that it considers a hypothetical monopolist of a set of products or services but, instead of imposing a hypothetical increase in prices of this product or service, it analyses possible substitution effects following a decrease in the quality of these products or services. Like the SSNIP test, the SSNDQ test can be used to draw the boundaries of a relevant antitrust market and understand the potential for market power.

Notably, in zero-price markets, the European Union has recommended the use of an SSNDQ as an alternative to an SSNIP. While pricing power has historically been a hallmark of market power and may be readily observed in traditional industries, with the rise of the digital economy, many of the world’s largest firms now operate in the zero-price economy. For example, social media companies like Meta, search engines like Google, and digital apps like Yelp are “free” to users, in that users trade their data to the company and its advertisers to use the product or service for $0. Because of this, defining a market using the traditional SSNIP test, or evaluating market power through a pricing analysis, imposes both empirical and even conceptual challenges. Alternatively, using an SSNDQ test allows the fact finder to maintain the zero-price nature of a basket of potentially competitive goods while still considering how consumers may substitute away from certain products given a change in a product’s value – where value contains elements of both pricing and quality.

Indeed, there is some evidence that a greater move towards quality considerations in market definition may be on the horizon. For example, in the DOJ’s recent case against Google related to its monopolization of general search services, Judge Amit Metha, in his opinion, cited evidence of Google’s ability to decrease the quality of its search engine as evidence of its monopoly power in the market for general search services. However, the potential import of an SSNDQ test to the United States would likely bring both additional opportunities and additional considerations for market definition analyses going forward.

In February 2023, Novant agreed to purchase Lake Norman Regional Medical Center (LNR) and Davis Regional Psychiatric Hospital (Davis) from CHS. Novant is one of the largest health systems in NC, operating multiple facilities across the state. On the other hand, CHS is a national for-profit health system, and the LNR and Davis facilities represent CHS’s most important assets in NC. CHS wanted to sell LNR because the facility needed substantial capital investments, which CHS was not willing to make. Given its poor performance and investment needs, The Davis facility was a former care hospital that, due to its poor performance and investment needs, was repurposed as a psychiatric facility. The transaction between Novant and CHS aimed to improve Novant’s competitive edge relative to Atrium Health (Atrium), NC’s largest health system.

The FTC decided to challenge the transaction in January 2024, initiating an administrative procedure and subsequently filing a complaint to block the deal in the US District Court for the Western District of NC. The FTC argued that LNR and Novant’s nearby hospital are head-to-head competitors and the main hospital options in the Eastern Lake Norman area, the relevant geographic market for the transaction. As such, LNR exerts competitive pressure on Novant, limiting Novant’s ability to increase prices.

In June 2024, the district court ruled in favor of CHS and Novant by rejecting FTC’s preliminary injunction request. However, the FTC appealed at the US Court of Appeals for the 4th Circuit, which, in a divided decision, reverted the district court’s decision and granted the request to enjoin the CHS-Novant deal. Following this setback, the parties abandoned the deal. The appeals court’s ruling does not explain why the district court decision had to be reverted, with the dissenting judge explicitly agreeing with the district court that the injunction is not in the best public interest.

Three features make this case of particular interest for antitrust economic analysis. First, one of the parties, CHS, had decided to exit the market and stopped actively competing. Second, the transaction seemed to have meaningful potential pro-competitive effects. Third, the buyer, Novant, committed not to increase prices for three years after the transaction.

Whether you are an economist, attorney, antitrust enthusiast, or just curious about Secretariat, we are glad you found us. This publication showcases insights from leading economists about recent developments in law and economics that may significantly impact the field of antitrust. This issue explores recent topics in the economics of antitrust analysis, with implications for merger analysis, market definition, and anticompetitive conduct.

In the first article of this issue, Dr. Pablo Varas covers important economic lessons from the abandoned deal between Novant Health and Community Hospital Systems in North Carolina. This article highlights aspects of the deal that differed from more typical healthcare acquisitions and explores how they may impact proposed mergers going forward.

The first article discusses the recently granted permanent injunction blocking the proposed merger of Penguin Random House (PRH) and Simon & Schuster (SS). The second article discusses a systematic approach to managing and using big data in litigation.

In the first article, Jason Albert discusses the Department of Justice’s (DOJ) monopsony allegation and market definition in the PRH and SS merger case. Dr. Albert notes that DOJ focused on authors as labor, argued that the merger would result in increased monopsony power, and is likely to focus on labor markets in future monopsony cases.

The first article discusses the Department of Justice’s (“DOJ”) first guilty plea in a criminal no-poach antitrust case. The second article also addresses no-poach antitrust issues, with a discussion of recent decisions regarding the McDonald’s and Burger King class action cases. The third article pertains to sponsored search auctions (“SSA”) and analyzes the importance of seller and bidder information in online advertising auctions.

In the first article, Erica Greulich discusses the recent case in which VDA OC LLC (“VDA”) plead guilty to claims that it violated antitrust law by participating in an alleged scheme to limit the wages of nurses. Dr. Greulich discusses how the guilty plea, along with other efforts by the DOJ, may further increase the likelihood that the DOJ will pursue more labor-side antitrust investigations.