US: Economics

Virtually all antitrust cases brought under ‘rule-of-reason’ in the United States require the plaintiff to show that the defendant has market power or, in the case of mergers, that the merger is likely to create or enhance the merging firms’ market power. In principle, market power is the ability of a firm (or a group of firms) to raise and sustain price above competitive levels. Historically, assessments of market power have been linked closely with and dependent on the determination of the ‘relevant market’ – delineated by a set of products and a geographic area – where effects of the firm’s (or group of firms’) ability to raise prices is likely to be felt. In practice, determination of the relevant market has been considered a necessary predicate to a market power analysis in most antitrust cases.¹ Firms with a relatively high market share in the relevant market are thought to have market power.

In recent years, however, the antitrust community in the United States has questioned whether market definition should play such a central role in the determination of market power, and it has confronted the question of how much market share is enough to raise concerns. Specifically, the antitrust community has debated whether one needs to define a relevant market to establish market power or whether other types of evidence can provide more useful insight on the question of whether the behaviour or deal under consideration raises competitive concerns.² This debate has been fuelled, in part, by the 2010 Merger Guidelines, issued by the Federal Trade Commission (FTC) and the Department of Justice (DOJ) (collectively the ‘agencies’), which explains that, in the merger context, ‘[t]he Agencies’ analysis need not start with market definition’³ and that they will ‘consider any reasonably available and reliable evidence to address the central question of whether a merger may substantially lessen competition.’⁴ The antitrust community has also wrestled with whether market power requires a threshold market share or, in other words, whether firms with relatively low market share can possess market power.

This article examines these questions in the context of recent antitrust activity in the United States. We describe the economic issues and the recent developments in the assessment of market power, including increasing reliance on ‘direct effects evidence’ to evaluate market power. Based on our review of recent court and agency decisions, we observe that:

• reliance on market definition to infer market power remains a critical element of virtually all antitrust examinations;

• relatively low market shares (eg shares substantially under 50 per cent) can be found to be consistent with a firm having market power; and

• direct effects is more of a complement rather than a substitute for structural evidence, at least for litigated matters.

Defining markets

A relevant market consists of a set of products and geographies within which consumers view the products as reasonably interchangeable and the firms supplying those products face similar competitive conditions. At its core, consumers and their willingness to substitute among products discipline firms and prevent them from raising prices. As such, the determination of relevant market begins with an understanding of demand substitution or consumers preferences among the products (or services) at issue.

In practice, market definition involves an understanding of the products, their attributes, consumer usage habits and the geographic reach of the parties. Analytically, the degree of interchangeability is characterised by the cross-price elasticity of demand. Products and firms that exhibit relatively high cross-price elasticities of demand are considered to be substitutes. In the merger context, it is often feasible and instructive to use information about consumer preferences to perform a hypothetical monopolist test to determine whether a hypothetical monopolist over the products in the candidate relevant market could profitably impose at least a small but significant and non-transitory increase in price (ie, an ‘SSNIP’ of 5 to 10 per cent).⁵ One might also perform such a test in a unilateral conduct case, though it may not be appropriate if prices are already elevated because of the alleged conduct.⁶

A series of recent cases provide insights into practical approaches to product and geographic market definition. For example:

• In the DOJ’s challenge of American Express’s merchant restraints – which barred merchants from steering customers to other credit cards at the point of sale – the parties debated relevant product market.⁷ The DOJ asserted a market definition that included only general purpose credit and charge card network services (credit cards), and American Express asserted a relevant market that included both credit and debit cards. The analysis included a discussion of functional attributes of credit and debit cards, an analysis of consumer usage habits, and a hypothetical monopolist test to evaluate whether a hypothetical monopolist over all credit cards could profitably raise merchant fees by a SSNIP, without losing too much merchant volume to debit.

• In the FTC’s challenge of St Luke’s acquisition of Saltzer Medical Group – a hospital system’s acquisition of physician practice in the state of Idaho – the parties debated geographic market.⁸ The FTC asserted a narrower geographic area consisting of the city of Nampa, Idaho, and the merging parties argued a broader geographic area including the cities of Nampa, Boise and cities in between. The parties studied patients’ willingness to travel for primary care physician services, and they considered whether a commercial payer could successfully market a provider network without Nampa-based primary care physicians to Nampa residents.

As these examples demonstrate, there are seldom bright lines that distinguish products or geographies. Simply observing that there is some substitution or that there could be substitution in face of a dramatic price increase does not mean that those products belong in the same relevant product market or that the geographic market should be broadened. Rather, market definition requires that consumers be willing to substitute between products or firms in the response to a five or ten per cent price increase.

Evaluating market power

There are two basic approaches to establishing market power: the structural approach and the direct effects approach. The structural approach assesses market power indirectly, by determining whether the conditions and circumstances of the relevant market lend themselves to the unlawful exercise of market power. As such, structural analyses focus on the structure of the relevant market, as measured by indicia such as market shares and concentration measures (specifically, the degree of concentration of market shares). In contrast, the direct effects approach considers direct evidence of actual or potential anti-competitive outcomes to determine whether the conduct or deal at issue has already had (or will have) anti-competitive effects. Direct effects analysis relies upon the observation, perhaps through an appropriate natural experiment, of the firm’s (or firms’) ability to raise prices or otherwise restrict output.

Structural analysis

Proper identification the relevant market is critical to the implementation of structural analyses because the share of sales in the relevant market accounted for by the party or parties in the antitrust case are the main inputs to the analysis of market power using this approach. In short, the existence or potential for the exercise of market power is inferred from the market shares of all relevant parties in the relevant market – with more concentrated markets considered to be more prone or vulnerable to anti-competitive behaviour and outcomes.

As an economic matter, inferring market power from market structure is an inexact science which often requires a detailed understanding of industry dynamics. While it is commonly understood that the higher is market concentration, the more likely it is that firms possess market power,⁹ there is considerable debate over how much market share is ‘enough’ to conclude a defendant has market power. For example, in United States v American Express, the court explained that ‘Amex’s market share [of 26.4 per cent] alone likely would not suffice to prove market power by a preponderance of the evidence were it not for the amplifying effect of cardholder insistence.’¹⁰ The court went on to explain that ‘cardholder insistence’ referred to American Express’s ‘highly insistent or loyal cardholders’ and that ‘[t]he ability of merchants to resist potential anti-competitive behavior by Amex, including significant price increases, by shifting customers to less expensive credit card networks or other forms
of payment is severely impeded by the segment of Amex’s cardholder base who insist on paying with their Amex cards and who would shop elsewhere or spend less if unable to use their cards of choice.’¹¹

There are many examples, beyond credit card networks, where firms have a set of ‘highly insistent or loyal’ customers. For instance, viewers can be loyal to and insistent upon having access to particular television networks. Indeed, some television networks – like regional sports networks or broadcast television networks – are sometimes described as ‘must-have’ or ‘essential’ to ability of multichannel video programming distributors to compete with each other, even though individually, each network may command a relatively low share of ratings points.¹² Similarly in health care, specific health-care providers (eg, a local academic medical centre, a children’s hospital, or cancer treatment facility) may be essential to a payer’s ability to construct a commercially desirable provider network, even though there are other providers in the relevant market.

Direct effects

‘Direct effects evidence’ refers to various kinds of evidence of actual effects that may be available in different cases. Examples include:

• information about an acquiring company’s post-merger plans to raise prices;

• evidence that competition between the merging parties has led to lower prices or other competitive benefits;

• changes in prices or output from a consummated merger; and

• the results of natural experiments.¹³

In spatial markets, for example, direct evidence of higher prices in areas where merging parties do not compete head-to-head can shed light directly on whether a proposed merger is likely to facilitate the exercise of market power – regardless of whether the estimated market shares of the merging companies suggest that such outcomes are likely to occur.

Direct effects evidence can serve two different functions. First, because it can provide information on consumer substitution in the face of an actual price increase, direct effects evidence can be a useful input into the identification of the boundaries of the relevant market. Second, direct effects evidence can identify circumstances in which market power has actually been exercised, regardless of whether structural analysis supports a conclusion of market power. When it is possible, the assessment of direct effects evidence can provide a useful complement to structural evidence, particularly in situations where lower market concentrations may mask opportunities for the exercise of market power due to special industry or market characteristics.

Proving market power in court

We review here some notable cases that have been litigated in the US courts over the past five years. These cases decided by the agencies and the courts not only affirm the continuing important of structural analyses in market power determinations, but also reveal an increasing reliance on direct-effect evidence to sharpen and inform such determinations.

FTC, et al v Sysco Corporation, et al

In 2015, the FTC moved to block the merger of Sysco Corporation and US Foods, Inc, the first and second-largest broadline food service distributors.¹⁴ The key point of contention between the FTC and the parties was the definition of the relevant market, which the FTC defined as broadline food service distribution sold to national customers and the parties defined as the entire food service distribution industry. The court agreed with the FTC, crediting its SSNIP test showing that the relevant market was restricted to national customers. According to the FTC:

• over 70 per cent of the time when Sysco or US Foods lost a bid opportunity for a national customer, it was to another broadline distributor; and

• only 50 per cent of national broadline customers would have to remain within the market to make a price increase by a hypothetical monopolist profitable.¹⁵

In this relevant market, the FTC calculated post-merger market share for the merged entity ranging from 71 per cent to 59 per cent.¹⁶ Based on this analysis, the court concluded that the level of market concentration created by this merger was presumed to be likely to enhance market power.’¹⁷ This outcome illustrates the continuing and prominent role of structural analysis in US merger review, at both the agencies and the courts.

United States of America et al v American Express Company et al

This case, which the lower court decided in 2015, illustrates the successful use of a combination of structural and direct effects evidence to prove the plaintiff’s claim that the defendant has market power. As explained above, the parties disputed the boundaries of the relevant market, with the DOJ asserting a relevant market of credit cards only and American Express asserting a market of both credit and debit cards. In the end, the court ruled in favour of the United States, identifying a relevant market of credit cards only in which American Express was the second largest card network, accounting for 26.4 per cent of charge volume. While this market share was considered significant by the court, it was direct effects evidence – namely, the evidence of cardholder insistence that prevented merchants from walking away from American Express – that underlies the court’s ruling that American Express has market power. The court wrote:

American Express’s highly insistent or loyal cardholder base is critical to the court’s finding of market power in this case. The ability of merchants to resist potential anti-competitive behavior by Amex, including significant price increases, by shifting customers to less expensive credit card networks or other forms of payment is severely impeded by the segment of Amex’s cardholder base who insist on paying with their Amex cards and who would shop elsewhere or spend less if unable to use their cards of choice.¹⁸

FTC, et al v St Luke’s Health System

In 2014, the FTC challenged a proposed merger between St Luke’s Health System and Saltzer Medical Group in which the geography of the relevant market was a key point of contention between the parties.¹⁹ The FTC asserted a relevant market for primary care physician services sold to commercially insured patients in Nampa, Idaho, and the merging parties asserted a broader geographic market that included Boise, Idaho. In defining the boundaries of the relevant market, the court was recognised that insurers could not successfully offer a network of primary care physician services to residents of Nampa, Idaho with all services located in another city. Accordingly, the court concluded that ‘Nampa is therefore the relevant geographic market.’²⁰ Within this relevant market, the merged entity would have accounted for almost 80 per cent of PCP services in Nampa. As a result, the court held, ‘[t]his substantial market share will give St Luke’s a dominant bargaining position over health plans in the Nampa market services.’²¹ In the end, St Luke was ordered to fully divest itself of Saltzer’s assets and unwind the acquisition.

United States of America v H&R Block, Inc et al

In 2011, the DOJ sought to enjoin the merger of H&R Block and 2SS Holdings (Tax Act), two providers of digital do-it-yourself (DDIY) tax preparation products. In defining the relevant market, the DOJ sought to limit the market to DDIY products, while the merging parties sought to include not only DDIY products, but also assisted tax preparation services. Based on the apparent absence of price competition between DDIY products and assisted tax preparation services, the court defined the relevant market to include only DDIY products. Within this narrower market definition, H&R Block had 15.6 per cent market share and Tax Act had a 12.8 per cent market share, resulting in a post-acquisition market share of 28.4 per cent. The court wrote, ‘[T]he totality of the evidence confirms that anti-competitive effects are a likely result of the merger, which would give H&R Block and Intuit control over 90 per cent of the market for digital do-it-yourself tax preparation products.’²² In effect, the elimination of one of the top three firms could facilitate tacit coordination between H&R Block and Intuit with obvious anti-competitive effects.²³ The merger was enjoined.

Michael C Malaney, et al, v UAL Corp, United Airlines, Inc, and Continental Airlines, Inc

This proposed merger between UAL Corp, United Airlines, Inc, and Continental Airlines, Inc was challenged, in 2010, by various individual plaintiffs seeking a preliminary injunction barring the merger.²⁴ The court rejected the plaintiffs’ claims owing to the failure to prove the existence of a viable relevant geographic and product market. In addition, the court held that the plaintiffs failed to provide any direct effects evidence to support their claim. The court wrote, ‘Plaintiffs have not demonstrated in any way that they will suffer any specific harm, including ‘increased airfares, decreased service, poorer service, and a constraint on the ability of other network carriers to compete.’²⁵ This decision, which was handed down shortly after the issuance of the 2010 Merger Guidelines, affirmed the continuing importance of market definition in merger analysis.

Conclusions

Based on our review of agency practice and the string of recent US court decisions, we find that:

•   reliance on market definition to infer market power remains a critical element of virtually all antitrust examinations;

•   relatively low market shares can be found to be consistent with a firm having market power; and

•   direct effects is more of a complement rather than a substitute for structural evidence, at least for litigated matters.

Notes

1. See United States v EI du Pont de Nemours & Co, 353 US 586, 593 (1957).
2. See, for example, Carl Shapiro and Joseph Farrell, ‘Antitrust Evaluation of Horizontal Mergers: An Economic Alternative to Market Definition,’ the B E Journal of Theoretical Economics, Vol. 10, Issue 1, 2010, available at: http://faculty.haas.berkeley.edu/shapiro/alternative.pdf.
3. US DOJ and FTC, ‘Horizontal Merger Guidelines’ (hereafter Merger Guidelines), available at www.justice.gov/atr/horizontal-merger-guidelines-08192010, at section 4.
4. Merger Guidelines, at section 2.
5. Merger Guidelines, at p. 9.
6. Thomas Krattenmaker, Robert Lande and Stephen Salop, ‘Monopoly Power and Market Power in Antitrust Law,’ available at: www.justice.gov/atr/public/hearings/single_firm/docs/222144.htm, at pp. 9–20.
7. United States of America et al v American Express Company et al 2015 US Dist LEXIS 20114 (EDNY, 19 February 2015) (American Express Decision).
8. Federal Trade Commission, et al v St Luke’s Health System, 2014 US Dist (D Idaho, 24 January 2014) (St Luke’s–Saltzer Decision).
9. The Merger Guidelines, for example, use market shares to classify relevant markets into ‘Unconcentrated Markets’ (which ‘ordinarily require no further analysis’), ‘Moderately Concentrated Markets’ (which ‘potentially raise significant competitive concerns and often warrant scrutiny’), and ‘Highly Concentrated Markets’ (which, under some circumstances, ‘will be presumed to be likely to enhance market power’). Merger Guidelines, at section 5.3.
10. American Express Decision, at p. 71.
11. In supporting its decision the Court noted a similar finding of market power in: United States v Visa USA, Inc 344 F.3d at 239-40, where Visa had a similar market share; and Toys ‘R’ Us, Inc v FTC, 221 F.3d 928, 937 (7th Cir. 2000), where the Toys ‘R’ Us accounted for 20 per cent share of national wholesale market. See American Express Decision, at p. 71.
12. According to the US Federal Communications Commission, for example, ‘[i]f the programming offered by a competitive MVPD [multichannel video programming distributor] lacks “must have” programming that is offered by the incumbent cable operator, subscribers will be less likely to switch to the competing MVPD.’ Report and Order and Notice of Proposed Rule Making, FCC 07-169, at paragraph 41.
13. Merger Guidelines, at section 2.1.
14. ‘[A] broadline foodservice distributor sells and delivers a “broad” array of food and related products to just about anywhere food is consumed outside the home.’ Federal Trade Commission, et al v Sysco Corporation, et al, 2015 US Dist LEXIS 83482 (DDC, 23 June 2015) (Sysco–US Foods Decision), at p. 4.
15. Sysco–US Foods Decision, at p. 26.
16. The FTC’s expert calculated market shares using two different data sets, including six variations of sensitivities on the second data set. Of these, the ‘[m]ost convincing to the court was …[the] final method of calculating shares…which assumed that all 16 of the top broadliners had the same national-local sales ratio as Defendants did…This variation almost certainly underestimated Defendants’ market shares.’ Sysco–US Foods Decision, at pp. 33–35.
17. Sysco–US Foods Decision, at pp. 32–33.
18. American Express Decision, at pp. 5, 36–39.
19. St Luke’s–Saltzer Decision, at p. 8.
20. St Luke’s–Saltzer Decision, at p. 8.
21. St Luke’s–Saltzer Decision, at p. 12.
22. United States of America v H&R Block, Inc et al, 833 F. Supp. 2d 36 (DDC, 10 November 2011) (H&R Block-Tax Act Decision), at p. 45.
23. H&R Block-Tax Act Decision, at pp. 33–37.
24. Michael C Malaney, et al, v UAL Corp, United Airlines, Inc, and Continental Airlines, Inc, 2010 US Dist LEXIS 106049 (ND Cal, 27 September 2010) (United-Continental Decision), at p. 10–11.
25. United–Continental Decision, at p. 15.