A Survey of Evidence Leading to Second Requests at the Ftc

A SURVEY OF EVIDENCE LEADING TO SECOND REQUESTS AT THE FTC

DARREN S. TUCKER*

Antitrust practitioners and other observers follow closely the enforcement patterns and trends of the two agencies charged with antitrust merger enforcement in the United States, the Federal Trade Commission and the U.S. De- partment of Justice Antitrust Division (Agencies). To provide guidance to the antitrust bar and business community, and to increase transparency, the Agen- cies have issued a variety of materials describing their merger enforcement policies and practices, including the Merger Guidelines, the Merger Guide- lines Commentary,1 policy guides and best practice reports,2 closing statements, and speeches. In addition, over the last decade, the Agencies have published a number of studies describing the types of evidence—such as market concentration, entry barriers, and “hot” documents—that led the Agencies to pursue enforcement action against proposed mergers and acquisitions that were subject to a Request for Additional Information (Second Request).3

* Attorney Advisor to FTC Commissioner Joshua D. Wright; formerly Attorney Advisor to FTC Commissioner J. Thomas Rosch. The views expressed here are the author’s and are not purported to represent the views of the Commission or any Commissioner. The author thanks Malcolm Coate, Leemore Dafny, and Ken Heyer for helpful comments and Kelsey Buntjer and Monica Kumar for research assistance. 1 U.S. Dep’t of Justice & Fed. Trade Comm’n, Commentary on the 1992 Horizontal Merger Guidelines (2006) [hereinafter Commentary], available at http://www.ftc.gov/os/2006/03/Com- mentaryontheHorizontalMergerGuidelinesMarch2006.pdf. 2 Merger process initiatives, workshop materials, remedy guidelines, enforcement reports, and policy statements for the Antitrust Division are available at http://www.justice.gov/atr/ public/merger-enforcement.html, and for the FTC at http://www.ftc.gov/bc/mergers.shtm. 3 Generally, these studies indicated that the presence of hot documents or strong customer complaints was correlated with a decision to challenge a transaction (i.e., by seeking a remedy or an injunction), while low entry barriers were correlated with a decision not to challenge. The level of and change in concentration were less accurate predictors, except at very low or very high levels. See Fed. Trade Comm’n, Horizontal Merger Investigation Data, Fiscal Years 1996–2011 (Jan. 4, 2013) [hereinafter 2013 Merger Data Report], available at http://www.ftc. gov/os/2013/01/130104horizontalmergerreport.pdf; Fed. Trade Comm’n, Horizontal Merger In- vestigation Data, Fiscal Years 1996–2007 (Dec. 1, 2008), available at http://www.ftc.gov/os/ 2008/12/08120hsrmergerdata.pdf; Fed. Trade Comm’n, Horizontal Merger Investigation Data, Fiscal Years 1996–2005 (Jan. 25, 2007), available at http://www.ftc.gov/os/2007/01/P035603

591

78 Antitrust Law Journal No. 3 (2013). Copyright 2013 American Bar Association. Reproduced by permission. All rights reserved. This information or any portion thereof may not be copied or disseminated in any form or by any means or downloaded or stored in an electronic database or retrieval system without the express written consent of the American Bar Association.

Electronic copy available at: http://ssrn.com/abstract=2215678 592 ANTITRUST LAW JOURNAL [Vol. 78

These studies provide valuable insights regarding the relative importance of various categories of evidence considered by the Agencies during the course of a formal Second Request investigation. However, as practitioners are aware, whether a transaction will be subject to an enforcement action is not the only question the parties must consider when weighing the antitrust risks associated with a potential transaction. The length and intensity of the review process is also an important factor. In particular, transacting parties and their counsel often consider the risk that a transaction will be subject to a Second Request. Although the Agencies and the parties work collaboratively to expe- dite the Second Request process, the issuance of a Second Request significantly lengthens the Hart-Scott-Rodino (HSR) waiting period—by an average of five months4—and dramatically increases the burden on the merging parties and the reviewing agency. Moreover, a high percentage of transactions subject to a Second Request result in some form of enforcement.5 As a result, merging parties often devote significant resources to examining whether a proposed transaction will trigger a Second Request and to convincing the investigating agency not to issue a Second Request or to minimize its scope. This study examines the theories of harm and types of evidence associated with in-depth merger investigations at the FTC during a recent four-year period and, based on this examination, offers practical guidance regarding the factors that appear to be most relevant at the initial phase of an FTC merger investigation. The FTC’s assessment of defenses frequently asserted by merging parties and the average length of time to conduct investigations are also presented. To accomplish this, this study uses enforcement data that are un- available to practitioners outside the FTC: memoranda written by agency staff recommending the issuance of Second Requests.6 As explained below, while any study based on these memoranda has limitations, these materials provide a unique window into the FTC’s process for making enforcement decisions. horizmergerinvestigationdata1996-2005.pdf; Fed. Trade Comm’n, Horizontal Merger Investiga- tion Data, Fiscal Years 1996–2003 (Feb. 2, 2004 rev. Aug. 31, 2004), available at http://www. ftc.gov/os/2004/08/040831horizmergersdata96-03.pdf; Fed. Trade Comm’n & U.S. Dep’t of Jus- tice, Merger Challenges Data, Fiscal Years 1999–2003 (Dec. 18, 2003), available at http://www. ftc.gov/os/2003/12/mdp.pdf. 4 See Tables 7a, 7b, infra. 5 In 75 percent of the surveyed transactions (58 of 77 transactions), the FTC obtained some form of relief, including a fix-it-first remedy or a termination of the transaction, after issuing a Second Request. See also LAW BUSINESS RESEARCH, THE MERGER CONTROL REVIEW 468 (Ilene Knable Gotts ed., 3d ed. 2012) (“A high percentage of the transactions in which an agency issues a second request will result in some type of enforcement action (i.e., court challenge, consent decree, restructuring).”). 6 As described in more detail in Part II, the data used in this study do not permit rigorous testing for changes in agency review of proposed transactions because of the lack of analogous data on FTC-reviewed mergers that did not receive Second Requests. Despite this limitation, the comparisons provided below may be useful to parties in gauging whether the new Guidelines signal a change in FTC merger practices.

Electronic copy available at: http://ssrn.com/abstract=2215678 2013] SURVEY OF EVIDENCE LEADING TO FTC SECOND REQUESTS 593

By providing data on the FTC’s initial-phase merger enforcement decisions, this study may help practitioners focus their arguments on topics likely to be of interest to career staff at the FTC and avoid emphasizing topics that, although cited in the Merger Guidelines, receive little attention in FTC memoranda discussing whether to issue a Second Request. This study, when viewed in conjunction with prior studies, may also offer insight on how the importance of certain forms of evidence may vary at different stages of an investigation.7 A related question is whether and to what extent the FTC’s approach to the Second Request decision has changed in recent years. In August 2010, the Agencies released an updated version of the Merger Guidelines,8 a milestone that was greeted with a mix of praise and criticism. Many practitioners and academics commended the greater analytical flexibility, higher concentration screens, and inclusion of additional topics such as partial acquisitions and power buyers. Nevertheless, a number of observers raised concerns that the 2010 Guidelines downplayed the role of market definition, emphasized unreliable economic tools, permitted challenges based on exclusionary conduct concerns, and placed too little attention on non-price and dynamic effects.9 In addition, some questioned the Agencies’ claim10 that the revisions to the Guidelines were limited to codifying existing agency practice and wondered whether the 2010 Guidelines would result in a significant—and more aggressive—shift in merger enforcement at the Agencies. This study does not attempt to resolve the broader debate regarding whether the 2010 Guidelines have ushered in a new era of more intensive or less bounded merger enforcement. However, by comparing the theories of harm and sources of evidence cited by FTC staff when justifying the need for in- depth merger investigations before and after the new Guidelines became ef-

7 For example, the Agencies have stated that the structural presumption may be given more weight in early stages of investigations. See Commentary, supra note 1, at 2 (“Evidence that the merged firm would have a relatively high share of sales . . . or that the market is relatively highly concentrated may be particularly significant to a decision by either of the Agencies to extend a pre-merger investigation pursuant to HSR by issuing a [Second Request].”). 8 U.S. Dep’t of Justice & Fed. Trade Comm’n, Horizontal Merger Guidelines (2010) [hereinafter 2010 Guidelines or 2010 Merger Guidelines], available at http://ftc.gov/os/2010/08/ 100819hmg.pdf. The Guidelines “outline the principal analytical techniques, practices, and the enforcement policy” of the Agencies with respect to horizontal mergers. Id. § 1. In addition, the federal courts and foreign competition authorities often look to the Guidelines when developing standards for evaluating horizontal mergers. 9 For example, the Antitrust Source published a collection of critical essays in its October 2010 issue, which is available at http://www.americanbar.org/content/dam/aba/publishing/ antitrust_source/Oct10_FullSource.authcheckdam.pdf. 10 See, e.g., Press Release, Fed. Trade Comm’n, Federal Trade Commission and U.S. Depart- ment of Justice Issue Revised Horizontal Merger Guidelines (Aug. 19, 2010), available at http:// www.ftc.gov/opa/2010/08/hmg.shtm (Christine Varney, Assistant Attorney General: “The revised guidelines better reflect the agencies’ actual practices.”). 594 ANTITRUST LAW JOURNAL [Vol. 78 fective, this study offers new insights into how the 2010 Guidelines may have affected the decision to issue Second Requests at the FTC.

I. BACKGROUND AND METHODOLOGY Under the HSR Act,11 parties to mergers and acquisitions exceeding certain thresholds must file premerger notification forms with the Agencies and observe a waiting period, which is typically thirty days, prior to consummating their transaction. Once the Agencies receive the HSR notification forms, the acquisition is “cleared” to one agency for review. After conducting a preliminary investigation of the notified transaction, the reviewing agency can allow the waiting period to expire or extend the waiting period by issuing a Request for Additional Information, also known as a “Second Request,” to each party. The issuance of Second Requests prevents the parties from closing their transaction until they produce additional documents and data relating to the product lines of concern and observe a second waiting period. Typically, during the Second Request process, the reviewing agency will also obtain similar materials from relevant third parties through the use of compulsory process. At the end of its review, the investigating agency can close its investigation and allow the deal to go forward, enter into a negotiated consent agreement with the parties, or challenge the transaction in court. In some cases, a decision to close an investigation may be based on the parties’ decision to aban- don their transaction or the parties’ unilateral efforts to remedy the agency’s substantive concerns. An agency’s decision whether to issue Second Requests is therefore a critical stage of the HSR process, and the Agencies do not take these decisions lightly. For example, at both Agencies, career staff must obtain approval from agency management before issuing Second Requests. At the FTC, when a transaction raises competitive concerns that cannot be resolved during the initial waiting period, staff prepares memoranda explaining the need for Second Requests. These memoranda summarize the evidence gathered to date, describe the competitive concerns with the transaction, and identify defenses raised by the parties. Based on these materials and deliberations involving staff and the Merger Screening Committee,12 the Director of the Bureau of Competition will decide whether to issue a Second Request. In recent years,

11 15 U.S.C. § 18a; see also 16 C.F.R. pts. 801–803 (implementing regulations). 12 The Merger Screening Committee consists of the Directors and Deputy Directors of the Bureaus of Competition and Economics and representatives from the Office of the Chairman, the Premerger Notification Office, the Office of Public Affairs, the Office of Policy Coordination, and the Office of International Affairs. Investigating staff also attends Committee meetings to discuss their recommendations and to answer questions. For a more detailed explanation of this process, see ABA SECTION OF ANTITRUST LAW, THE MERGER REVIEW PROCESS: A STEP-BY- STEP GUIDE TO FEDERAL MERGER REVIEW (4th ed. 2012). 2013] SURVEY OF EVIDENCE LEADING TO FTC SECOND REQUESTS 595 fewer than 5 percent of all transactions reported under the HSR Act triggered Second Requests.13 This study examines information extracted from non-public Merger Screen- ing committee memoranda prepared by staff in the FTC’s Bureau of Competi- tion and Bureau of Economics. All horizontal transactions that resulted in a Second Request at the FTC from August 2008 to August 2012—two years before and after the issuance of the 2010 Merger Guidelines—are included in the results. Transactions involving only vertical concerns were excluded,14 as were non-HSR-reportable transactions, even if such transactions were subject to a formal investigation. The survey includes a total of 77 investigations, 38 of which occurred prior to the 2010 Guidelines and 39 of which occurred subsequent to the 2010 Guidelines. Unless otherwise indicated, the information reported in Tables 1 through 11 is based on all 77 investigations in the survey. As with prior merger data studies, this study presents results by the parties’ industry where possible. The industry categories used here differ somewhat from prior reports, however. For example, grocery store and energy mergers are not treated as separate categories because of the small number of Second Requests issued in transactions involving those industries during the study period. Descriptions of each industry category and the number of transactions in this study falling within each category are in Table 1.

TABLE 1: INDUSTRY CATEGORIES

Category Description Number of Transactions Healthcare/ FDA-regulated pharmaceuticals and medical 21 total Pharma devices; over-the-counter medications 12 pre-, 9 post-2010 Guidelines Healthcare/ Hospitals; physician services; medical 12 total Providers laboratories; pharmacies 5 pre-, 7 post-2010 Guidelines Energy & Production, distribution, and retailing of energy 16 total Chemicals sources, including petroleum; chemicals 8 pre-, 8 post-2010 Guidelines Computer hardware, software, and components; 10 total Technology Internet-based services 3 pre-, 7 post-2010 Guidelines Grocery stores; food; instruments; equipment; 18 total Other funeral services; distributors (other than energy 10 pre-, 8 post-2010 Guidelines sector); miscellaneous

13 See Fed. Trade Comm’n & U.S. Dep’t of Justice, Hart-Scott-Rodino Annual Report for Fiscal Year 2011 at Figure 2, available at http://www.ftc.gov/os/2012/06/2011hsrreport.pdf (noting that the percentage of transactions reported under the HSR Act receiving Second Requests varied between 2.5 percent and 4.5 percent over the 2002 to 2011 fiscal year period). 14 For transactions raising both horizontal and vertical concerns, the survey included only portions of the Merger Screening memoranda related to the horizontal concerns. 596 ANTITRUST LAW JOURNAL [Vol. 78

Because this study is not based on a random sample of merger investigations, the results may not reflect FTC merger practice generally and cannot be used to infer a causal relationship between the presence of certain forms of evidence and the likelihood of a Second Request. Agency staff likely consid- ers a wider range of evidence and conducts a more thorough review of transactions that result in a Second Request than those that do not. In addition, agency staff prepares Second Request process memos only when they believe more in-depth investigation is warranted. Similar memos are not prepared when staff concludes that no further investigation is necessary and that the HSR waiting period should be allowed to expire. As a result, this study does not attempt to compare the evidence that led staff to seek Second Requests to the evidence that led staff to seek termination of the initial HSR waiting period. Another caveat is that the memoranda presented to the Merger Screening Committee reflect only the views of investigating staff, but others within the FTC decide whether to issue a Second Request. Even among agency staff, there can be disagreements between the lawyers and economists as to the inferences to be drawn from certain evidence.15 Agency management, whose views at this stage of the investigation are usually not memorialized in any detail, may determine to issue Second Requests based on different considerations or on a subset of the considerations identified in Merger Screening memoranda. Likewise, the survey results may not reflect the views of the Commission or any of the five Commissioners.16 The second part of this study, which examines the evidence and theories appearing in Merger Screening memoranda before and after the release of the 2010 Merger Guidelines, has two other potential biases. The first is that agency staff may have started to adjust their analytical tools before the 2010 Merger Guidelines were issued based on expectations of what those Guide- lines would contain.17 The second is the possibility of agency inertia. Because staff may have taken some time after publication to become familiar with the

15 When either legal or economic staff determined a form of evidence or theory of harm to be relevant, it was included in the study results even if the other group of staff did not address the issue or concluded that it was inconsequential. For Tables 10 and 11, evidence is considered relevant regardless of the inferences drawn by agency staff. For example, a diversion analysis suggesting that the merger parties are not close competitors would be counted in Table 10. 16 Other than the Chairman, the Commission is not involved in the decision to issue Second Requests. However, the use of compulsory process, which is used to obtain relevant materials from third parties, must be authorized by the full Commission. Staff typically provides the Merger Screening memoranda, along with a brief cover memo from the Bureau of Competition, to the Commission to justify the request for compulsory process. 17 The Agencies released a draft version of the 2010 Guidelines for public comment several months before issuing the final version. See U.S. Dep’t of Justice & Fed. Trade Comm’n, Hori- zontal Merger Guidelines: For Public Comment (Apr. 20, 2010), available at http://www.ftc.gov/ os/2010/04/100420hmg.pdf. 2013] SURVEY OF EVIDENCE LEADING TO FTC SECOND REQUESTS 597 new Guidelines, the study results may understate any changes to current agency practice. Notwithstanding these limitations, the study results can be used to evaluate the relative importance of certain factors on agency decision making and, in some cases, how the relative importance of these factors has changed over time. The results also provide a basis for assessing the comparative weight of these factors in transactions across different industries.

II. STUDY FINDINGS—FACTORS ASSOCIATED WITH SECOND REQUESTS Table 2 identifies the theories of competitive harm that were identified by agency staff in support of their recommendation to issue a Second Request for each transaction in the survey. Information is provided for each transaction rather than for each market raising concerns because Merger Screening memoranda do not always associate a specific theory of harm with a particular market. For simplicity, unilateral effects, coordinated effects, potential competition, exclusionary conduct, and monopsony are treated as distinct and non- overlapping theories of harm.18 Table 2 also indicates where agency staff identified price discrimination or loss of non-price competition (e.g., product quality, variety, service, or innovation) as a means by which the merger might lead to unilateral or coordinated effects.19 The table excludes two transactions—one merger in the energy sector before the 2010 Guidelines and one hospital merger after the 2010 Guidelines—where the Merger Screening memoranda did not identify a candidate theory of harm; in both cases, staff had an unusually short period of time to conduct the initial-phase investigation. The survey results indicate that:

18 Each of these theories of harm is defined by reference to their descriptions in the Merger Guidelines. See 2010 Guidelines, supra note 8, § 6 (unilateral effects are the “elimination of competition between two [merging] firms”); § 7 (coordinated effects “involves conduct by multi- ple firms that is profitable for each of them only as a result of the accommodating reactions of the others”); § 5.1 (potential competitors are firms that do not “currently earn revenues in the relevant market”); §§ 1, 12 (monopsony is the “[e]nhancement of market power by buyers”); see also id. § 3 (price discrimination concerns arise when a merger’s effects “vary significantly for different customers purchasing the same or similar products”); §§ 1, 6.4 (non-price effects include “reduced product quality, reduced product variety, reduced service, or diminished innovation”). The 2010 Guidelines do not define exclusionary conduct, but this term usually refers to conduct prohibited under Section 2 of the Sherman Act, such anticompetitive tying or exclusive dealing. See id. § 1 (stating that the Agencies employ Section 2 of the Sherman Act when reviewing mergers). 19 For example, a transaction raising concerns that it would lead to coordination on non-price aspects of competition would be counted in both the coordinated effects column and the non- price column. 598 ANTITRUST LAW JOURNAL [Vol. 78

• In over 90 percent of the initial-phase investigations in the sample, staff raised unilateral effects concerns; • Staff raised coordinated effects concerns in slightly more than half of the cases in the sample; • Staff raised potential competition concerns in one quarter of the cases; • Exclusionary conduct and monopsony concerns were less common than other theories of harm; • Staff was concerned that anticompetitive effects might be effectuated through price discrimination in approximately one in six transactions; and • Staff raised non-price concerns in almost half of all transactions.

There were some significant variations by industry. Staff raised unilateral effects concerns in every memorandum in the sample involving the healthcare provider industry, and raised such concerns in over ninety percent of cases in every other industry except pharmaceuticals. Coordinated effects concerns were common in energy and chemicals markets but were infrequent in healthcare markets. A majority of transactions in the pharmaceuticals industry presented potential competition concerns, but only 11 percent of other transactions raised such concerns. Exclusionary conduct and monopsony concerns arose in some but not all industries. Price discrimination concerns were most common in energy and chemicals transactions, but rarely arose in transactions involving healthcare industries. Non-price concerns arose in the technology industry at nearly three times the rate as in the energy and chemicals industries.

TABLE 2: THEORY OF COMPETITIVE HARM (BY TRANSACTION) Unilateral Coordinated Potential Exclusionary Price Non-Price Monopsony Effects Effects Competition Conduct Discrimination Concerns All 91% 56% 25% 16% 7% 16% 47% Industries Healthcare/ 76% 33% 62% 19% 5% 5% 52% Pharma Healthcare/ 100% 27% 9% — — 9% 36% Providers Energy & 93% 87% 13% — — 33% 27% Chemicals Technology 90% 50% 20% 30% — 30% 80% Other 100% 78% 6% 28% 22% 11% 44%

Table 3 summarizes the frequency with which staff identified certain factors in support of their recommendations to issue Second Requests for the 77 investigations in the survey. “Market structure” refers to concerns relating to the number and relative size of producers, even if precise market shares are 2013] SURVEY OF EVIDENCE LEADING TO FTC SECOND REQUESTS 599 not available. A transaction raises “customer complaints” or “competitor complaints” where customers or competitors express strong concerns that the transaction would result in an anticompetitive effect. Documents are considered “hot” if they predict that a transaction will result in anticompetitive effects or a significant loss of competition.20 “Significant entry barriers” exist where staff determined that de novo entry would be unlikely to deter or counteract the anticompetitive effects of concern.

TABLE 3: SOURCES OF EVIDENCE (BY TRANSACTION) Market Customer Competitor Significant Hot Docs Structure Complaints Complaints Entry Barriers All 100% 47% 14% 29% 86% Industries Healthcare/ 100% 33% 10% 14% 95% Pharma Healthcare/ 100% 67% — 50% 83% Providers Energy & 100% 75% — 13% 88% Chemicals Technology 100% 30% 30% 60% 70% Other 100% 33% 33% 28% 83%

Market structure and significant entry barriers were the most frequently cited factors in Merger Screening memoranda leading to Second Requests. Customer complaints (cited in nearly half of the sample cases) and hot documents (cited in nearly a third) were the third- and fourth-most frequently cited factors. Competitor complaints were cited much less frequently and, in some industries, appeared to have played no role in staff recommendations to issue Second Requests. These results differed somewhat from prior merger data reports examining final enforcement decisions. For example, the FTC’s most recent report indicated that, in making final enforcement recommendations, agency staff referenced customer complaints at a 68 percent rate, hot documents at a 14 percent rate, and entry barriers at a 100 percent rate.21

20 Some examples of hot documents in the surveyed transactions include a “4(c)” document touting that an important benefit of a hospital merger was the “enhancement of collective bar- gaining power with third-party payors,” a Board document of a medical products company stating that the target company would be the only “possible significant threat” to the company absent the transaction, and an email among senior executives of a technology company stating that the firm to be acquired was “the only competitor they worry about.” In contrast, documents describing general industry conditions—such as a highly concentrated market or high entry barriers— were not considered to be hot documents. 21 In the most recent FTC merger data report, hot documents were present in 25 of 175 enforcement actions, customer complaints were present in 111 of 164 enforcement actions, and high entry barriers were present in 175 of 175 enforcement actions. See 2013 Merger Data Re- port, supra note 3, Tables 5.1, 5.2, 7.1, 7.2, 9.1, 9.2. 600 ANTITRUST LAW JOURNAL [Vol. 78

Once again, the results varied by industry. In the pharmaceutical industry, for example, agency staff appears to have relied less on customer complaints and hot documents and more on evidence of significant entry barriers, suggesting a greater reliance on structural considerations. In contrast, in the technology industry, agency staff cited to competitor complaints and hot documents more often than in cases involving other industries and were more willing to recommend the issuance of Second Requests when entry barriers were not significant than in other industries. Table 4 identifies the number of significant competitors pre- and post- merger for each market of concern in the 77 surveyed investigations. The results exclude markets raising potential competition concerns because, by definition, at least one of the merging firms in those markets was not a significant competitor. The results also exclude markets for which agency staff was unable to determine which firms qualified as significant competitors prior to the issuance of Second Requests. Merger Screening memoranda typically defined a significant competitor as a firm that offered products or services that satisfied the needs of most customers in the relevant market or that exceeded a market share threshold, which was often five percent. In a few markets, the number of significant competitors did not change as a result of the transaction under review because one of the merging parties was a market participant but did not appear to be a significant competitor or because the transaction involved the acquisition of a limited set of assets from a company that remained in the market.

TABLE 4: CHANGE IN NUMBER OF SIGNIFICANT COMPETITORS (BY MARKET) No 2→13→24→35→46→57→68→7 change All 7 244458225 4 1 Industries Healthcare/ 4 111724132 1 — Pharma Healthcare/ — 3 5 1 1—1— Providers Energy & 3 48144111 Chemicals Technology — 3 5 7 4 1 1 — Other — 3 9 12 — 1 — —

Table 5 indicates the concentration levels for each market identified by staff in Merger Screening memoranda as requiring further review through a Second Request. The Appendix contains a breakdown of this information by industry. These tables reflect all of the transactions in the survey, but exclude 2013] SURVEY OF EVIDENCE LEADING TO FTC SECOND REQUESTS 601 markets raising potential competition concerns and markets for which FTC staff was unable to estimate market shares due to insufficient information. Concentration figures are based on the Herfindahl-Hirschman Index (HHI).22 At the initial phase of most investigations, market shares and concentration levels are preliminary and may be based on estimates provided by the parties, other market participants, or analyst reports. Not surprisingly, the vast majority of markets for which staff recommended a Second Request were highly concentrated post-merger. Of the 129 markets studied, 121, or 94 percent, exceeded the 1,500/100 concentration screen, and 113, or 88 percent, exceeded the 2,500/200 concentration screen described in Section 5.3 of the 2010 Guidelines.23 All of the markets that fell below these thresholds involved transactions with other markets exceeding at least one of these screens.

TABLE 5: CHANGE IN CONCENTRATION—ALL MARKETS

Change in HHI 100– 200– 600– 1,000– 1,500– 0–99 2,500+ TOTAL 199 599 999 1,499 2,499 0–1,499 1 1 1,500–2,499 1 5 2 1 9

Merger 2,500–3,999 2 16 14 10 2 44 - HHI 4,000–5,999 4 6 8 15 13 2 48 Post 6,000+ 2 3 1 10 11 27 TOTAL 7 0 30 27 27 25 13 129

Table 6 describes the FTC’s preliminary views regarding merging parties’ efficiency claims for each transaction in the survey. “Supportive” indicates that, in agency staff’s view, at least some portion of the asserted efficiencies appeared to be credible, merger specific, and cognizable, as required by the Merger Guidelines. “Skeptical” indicates that, in agency staff’s view, the asserted efficiency claims did not appear to satisfy the Merger Guidelines re- quirements or were so small, even if accepted, as to be immaterial to the

22 The HHI measure of concentration consists of the sum of the squares of the market shares of competitors in the relevant market. The number of markets in Table 5 differs from the number of markets in Table 4 because, for some markets, agency staff was able to determine market shares or the number of significant competitors but not both. 23 Under Section 5.3 of the 2010 Guidelines, supra note 8, mergers involving an increase in the HHI of more than 100 points in a market with a post-merger HHI of at least 1,500 may warrant additional scrutiny. Mergers resulting in an increase in the HHI of more than 200 in a market with a post-merger HHI above 2,500 “will be presumed to be likely to enhance market power.” 602 ANTITRUST LAW JOURNAL [Vol. 78 overall competitive analysis. An efficiencies claim is considered “Detailed” where the parties provided a white paper or party documents that described a transaction’s efficiencies in detail. An efficiencies claim is considered “Basic” where the parties provided only top-level or non-specific efficiency claims, such as a general description of expected efficiencies with little or no support- ing material. At the initial phase of most investigations, merging parties usually offer few details regarding expected efficiencies, and the agency’s primary sources of information regarding efficiencies are documents submitted with the HSR notification and public statements and filings by the merging parties. In most of the investigations where outside counsel submitted detailed efficiency claims, agency staff found those claims credible (although not enough to off- set their concerns about the transaction). Staff memoranda in cases receiving a Second Request rarely acknowledge the potential for a transaction to spur dynamic efficiencies (e.g., enhanced innovation), even in technology industry transactions. Given the limitations of this study, it is unclear whether agency staff is skeptical of this type of efficiency, parties are not advancing dynamic efficiency claims, or mergers with a potential for dynamic efficiencies do not raise concerns sufficient to trigger a Second Request (and therefore would not lead the staff to draft a Merger Screening memorandum recommending a Sec- ond Request).

TABLE 6: STAFF’S INITIAL EVALUATION OF EFFICIENCY CLAIMS Supportive Neutral Skeptical TOTAL No Efficiencies Asserted — — — 19 Basic Efficiencies Claim 13 35 5 53 Detailed Efficiencies Claim 3 1 1 5 TOTAL 16 36 6

Tables 7a and 7b indicate the length of time to complete merger investigations at the FTC once a Second Request has issued. These results are based on all of the transactions in the survey except for those the parties abandoned or those the consummation of which was delayed for at least six months for reasons beyond the FTC’s control, such as a temporary abandonment of the transaction by the parties.24 The time is measured from the issuance of the Second Request to the public announcement of the agency’s resolution of the

24 Table 7a excludes one transaction the valuation for which was not identified in staff memoranda or in press releases announcing the transaction. Investigations resolved through fix-it-first remedies are included in the overall results for Table 7b but are not broken out in a separate category because of the small number (2) of these cases. 2013] SURVEY OF EVIDENCE LEADING TO FTC SECOND REQUESTS 603 case, such as a proposed order for public comment or a decision to challenge the transaction in court. The results indicate that transaction size is positively correlated with investigation length; however, the difference in time to review the smallest (under $500 million) and largest (over $5 billion) transactions was only three weeks. Investigations resulting in litigation required the most time to complete, but otherwise there was no correlation between an investigation’s length and the result of the investigation.

TABLE 7A: INVESTIGATION LENGTH BY DEAL VALUE (IN DAYS) Mean Median All (64 transactions) 156 136 < $500m (24 transactions) 146 130 $500m −$5b (28 transactions) 160 134 >$5b (12 transactions) 168 148

TABLE 7B: INVESTIGATION LENGTH BY RESULT (IN DAYS) Mean Median All (65 transactions) 158 137 No Relief Sought (19 transactions) 146 139 Consent Agreement (38 transactions) 158 129 Litigation (6 transactions) 177 164

III. STUDY FINDINGS—SECOND REQUEST RECOMMENDATIONS BEFORE AND AFTER THE 2010 MERGER GUIDELINES

A. NEW AND EXPANDED THEORIES OF COMPETITIVE HARM IN THE 2010 GUIDELINES The 2010 Merger Guidelines describe several types of competitive harm that were not identified in the 1992 Guidelines. For example, Sections 5.1 and 5.3 of the 2010 Guidelines explain how the Agencies evaluate mergers involving potential competitors,25 a topic generally not addressed in the 1992 Guide- lines except with respect to “uncommitted entrants.”26 In addition, the 2010

25 See 2010 Guidelines, supra note 8, § 1 (“These Guidelines outline the principal analytical techniques, practices, and the enforcement policy of the Department of Justice and the Federal Trade Commission . . . with respect to mergers and acquisitions involving actual or potential competitors . . . under the federal antitrust laws.” (emphasis added)). 26 U.S. Dep’t of Justice & Fed. Trade Comm’n, Horizontal Merger Guidelines § 1.32 (1992, rev. 1997) [hereinafter 1992 Guidelines], available at http://www.justice.gov/atr/public/ guidelines/hmg.pdf. The DOJ’s 1984 Merger Guidelines discussed potential competition analysis in more detail. See U.S. Dep’t of Justice, Merger Guidelines § 4.1 (1984), available at http:// 604 ANTITRUST LAW JOURNAL [Vol. 78

Guidelines state for the first time that the Agencies consider whether horizontal mergers may lead to forms of exclusionary conduct ordinarily challenged under Section 2 of the Sherman Act.27 In addition, the 2010 Guidelines provide more explanation or emphasis to several forms of competitive harm described in prior versions of the Guidelines, including non-price effects, the exercise of monopsony power, and price discrimination. Table 8 describes the rates at which staff cited these competitive harms in Merger Screening memoranda before and after the 2010 Guidelines. As with Table 2, this table excludes two transactions for which staff did not identify a candidate theory of harm because of the brevity of the initial merger review.

TABLE 8: THEORIES OF COMPETITIVE HARM BEFORE AND AFTER THE 2010 GUIDELINES Pre-2010 Guidelines Post-2010 Guidelines Number Frequency Number Frequency Exclusionary Conduct 6 16% 6 16% Monopsony 3 8% 2 5% Price Discrimination 7 19% 5 13% Non-Price Effects 15 41% 20 53% — Product Variety 4 11% 1 3% — Innovation Market 1 3% 1 3%

1. Exclusionary Conduct As described in Table 8, exclusionary conduct concerns were cited in sixteen percent of horizontal merger investigations that led to Second Requests both before and after the implementation of the 2010 Guidelines; however, exclusionary conduct did not appear to be the principal theory of harm in any investigation. The most common exclusionary conduct concern cited by staff was that the merged firm might raise rivals’ costs through the use of bundling, tying, or exclusive dealing arrangements with its customers. In a smaller number of transactions, staff indicated that the merged firm might foreclose access www.justice.gov/atr/hmerger/11249.pdf; see generally Darren S. Tucker, Potential Competition Analysis Under the 2010 Merger Guidelines, 12 SEDONA CONF. J. 273, 278–82 (2011). 27 2010 Guidelines, supra note 8, § 1 (noting that Section 2 of the Sherman Act is a relevant antitrust statute for merger review and stating: “Enhanced market power may also make it more likely that the merged entity can profitably and effectively engage in exclusionary conduct.”); id. § 2.2.3 (stating that competitors may have relevant information “especially in cases where the Agencies are concerned that the merged entity may engage in exclusionary conduct”); id. §6 (“exclusionary unilateral effects also can arise”). For criticism of this addition to the Guidelines, see D. Bruce Hoffman & Daniel Francis, Including Exclusion in the 2010 Horizontal Merger Guidelines, ANTITRUST SOURCE, Oct. 2010, http://www.americanbar.org/content/dam/aba/publishing/antitrust_source/Oct10_Hoffman10_21f.authcheckdam.pdf. 2013] SURVEY OF EVIDENCE LEADING TO FTC SECOND REQUESTS 605 to a needed input or be able to prevent regulatory approval of a rival’s product as a result of the transaction. Competitors of the merging firms were the source of most exclusionary conduct complaints; staff was more likely to find these complaints credible when customers voiced similar concerns and the acquiring firm was already imposing vertical restraints on customers or distributors.

2. Non-Price Effects Although the 1992 Guidelines recognized that a merger could diminish non-price and price competition,28 as well as enhance product quality or innovation,29 those Guidelines provided little explanation as to how the Agencies applied these principles. The 2010 Guidelines provide a more extensive discussion of non-price effects30 and include a new Section (Section 6.4) that describes how the Agencies evaluate whether a merger is likely to diminish innovation competition or product variety in an anticompetitive manner. Approximately half of the surveyed transactions both before and after the 2010 Guidelines raised non-price concerns, such as loss of quality or service resulting from the transaction. Only in a handful of these cases, however, was loss of non-price competition the primary concern; loss of price competition was the primary concern in nearly all transactions and was at least a secon- dary concern in every transaction. Agency staff cited non-price concerns in support of a Second Request recommendation more frequently subsequent to the 2010 Guidelines. Specifically, staff raised concerns about a loss of product variety, one type of non-price effect, in five transactions during the survey period. In most of these, staff pointed to party documents indicating that the buyer planned to eliminate an existing or future product line of the target company with the expectation that customers would shift to higher-margin products offered by the buyer. Staff also raised concerns that the products to be eliminated offered features that customers valued and were not otherwise available in the market- place. Agency staff cited product variety concerns in Merger Screening memoranda less often after the 2010 Guidelines than before. Merger Screening memoranda often expressed a generalized concern for loss of innovation competition—another type of non-price competition—but only a small number of transactions in the survey raised innovation market

28 1992 Guidelines, supra note 26, § 0.1 n.6 (“Sellers with market power also may lessen competition on dimensions other than price, such as product quality, service, or innovation.”). 29 Id. § 4 (efficiencies include “improved quality, enhanced service, or new products” and “may result in benefits even when price is not immediately and directly affected”). 30 2010 Guidelines, supra note 8, §§ 1, 2.2.1, 4. 606 ANTITRUST LAW JOURNAL [Vol. 78 concerns akin to the Genzyme/Novazyme case.31 For these transactions, staff was concerned that the merging parties appeared to be two of a small number of firms engaged in research to treat a particular medical condition or to develop a new treatment option. Innovation market concerns did not arise in cases receiving Second Requests outside the pharmaceutical industry.

3. Monopsony Power and Price Discrimination Another topic that received greater emphasis in the 2010 Guidelines is monopsony power. The 1992 Guidelines’ discussion of monopsony power was limited to a single paragraph in the introduction. In contrast, the 2010 Guide- lines devote an entire section to this topic (Section 12), with significantly more explanation of the framework employed by the Agencies to evaluate buyer-side market power concerns. Likewise, although the 1992 Guidelines discussed price discrimination,32 the 2010 Guidelines devote an early section to this topic (Section 3), which may signal greater attention to this subject. FTC staff raised concerns regarding the merged company’s ability to exercise monopsony power in seven percent of the transactions studied. These concerns arose primarily in the “other” industry category. Monopsony concerns arose at roughly equal rates in Merger Screening memoranda resulting in Second Requests before and after the 2010 Guidelines. FTC staff was concerned about a merged firm’s ability to raise prices to targeted customers through price discrimination in approximately sixteen percent of investigations. These concerns arose most often in the chemicals and technology industries. The frequency with which staff raised price discrimination concerns appears to have declined slightly after the 2010 Guidelines.

B. THE ROLE OF MARKET DEFINITION Market definition was not only a necessary, but also the first, step under the 1992 Guidelines;33 although, subsequent guidance from the Agencies sug- gested greater analytical flexibility. For example, the 2006 Merger Guidelines Commentary states that the Agencies need “not settle on a relevant market definition before proceeding to address other issues” and that, where other

31 See Statement of Chairman Timothy J. Muris in the Matter of Genzyme Corporation/ Novazyme Pharmaceuticals, Inc., FTC File No. 021 0026 (Jan. 13, 2004), available at http:// www.ftc.gov/os/2004/01/murisgenzymestmt.pdf. 32 1992 Guidelines, supra note 26, §§ 1.12, 1.22, 1.32, 1.42. 33 Id. § 0.2. 2013] SURVEY OF EVIDENCE LEADING TO FTC SECOND REQUESTS 607 considerations indicated that anticompetitive effects were unlikely, market definition was an unnecessary exercise.34 The 2010 Guidelines continue the trend of placing less weight on market definition and concentration, at least in the analysis of unilateral effects for differentiated products. Section 6.1 explains that “[d]iagnosing unilateral price effects based on the value of diverted sales need not rely on market definition or the calculation of market shares and concentration.”35 In contrast, concentration remains an important consideration in the analysis of coordinated effects.36 The data presented in this study are consistent with the position that the 2010 Guidelines have had little, if any, effect on the role of market definition in FTC initial-phase merger investigations. In the two years leading up to the 2010 Guidelines, FTC staff defined markets in every merger investigation resulting in a Second Request. In the two years following the new Guidelines, staff defined markets in all but one of these investigations. In that case, agency staff rejected several product markets proposed by market participants but did not offer a view on the correct product market. The Merger Screening Committee memoranda indicated that a “threshold question” for the Second Request investigation would be determining the correct relevant market. In addition, nearly all Merger Screening memoranda prepared by FTC legal staff, both before and after the 2010 Guidelines, contain sections titled “Prod- uct Market(s)” and “Geographic Market(s).”37 These results are consistent with the FTC’s continued practice of defining markets in cases in which it seeks relief. The two Commission merger decisions handed down since the 2010 Merger Guidelines were implemented (ProMedica and Polypore) defined relevant markets at the outset of the analy-

34 Commentary, supra note 1, at 2, 5 (“[T]he Agencies do not apply the Guidelines as a linear, step-by-step progression that invariably starts with market definition and ends with efficiencies or failing assets.”). 35 See also 2010 Guidelines, supra note 8, § 4 (“Some of the analytical tools used by the Agencies to assess competitive effects do not rely on market definition, although evaluation of competitive alternatives available to customers is always necessary at some point in the analysis.”); id. § 5 (stating that the Agencies will “normally” define markets and determine concentration levels); id. § 6.1 (“These merger simulation methods need not rely on market definition.”). The 2010 Guidelines could also be read to suggest that the Agencies may not always define markets in merger enforcement actions. See id. § 4 (“In any merger enforcement action, the Agencies will normally identify one or more relevant markets in which the merger may substan- tially lessen competition.” (emphasis added)). But see id. § 1 n.2 (“These Guidelines are not intended to describe how the Agencies will conduct the litigation of cases they decide to bring.”). 36 Id. § 7.1 (a coordinated effects concern requires that “the merger would significantly increase concentration and lead to a moderately or highly concentrated market”). 37 FTC legal staff usually prepares Merger Screening memoranda based on a template promul- gated by the Bureau of Competition. That template, which was revised in 2011, includes a section to discuss the relevant market. 608 ANTITRUST LAW JOURNAL [Vol. 78 sis.38 Likewise, in every Section 7 complaint issued by the Commission since the new Guidelines, the FTC has alleged one or more relevant markets.39 This is true for both litigated and non-litigated cases. Thus, to date, the FTC does not appear to have availed itself of the greater flexibility afforded by the 2010 Guidelines regarding the necessity of defining markets.

TABLE 9: MARKET DEFINITION BEFORE AND AFTER THE 2010 GUIDELINES Pre-2010 Guidelines Post-2010 Guidelines Number Frequency Number Frequency Relevant Market Defined 38 100% 38 97% Determination of Shares or 36 95% 34 87% Significant Competitors

C. ECONOMIC TOOLS IN THE 2010 GUIDELINES The 2010 Guidelines identify several economic tools, some of which appear for the first time in those Guidelines, to help define markets or assess a transaction’s competitive effects. The addition that has garnered the most attention is the upward pricing pressure (UPP) concept,40 which the Agencies use to assess the likelihood of unilateral effects in differentiated products mergers. Although the 1992 Guidelines pointed to the use of diversion ratios and margins to assess the likelihood of unilateral effects,41 the 2010 Guide- lines make explicit that these two figures, when multiplied together, provide a measure of the value of diverted sales. As the value of diverted sales increases, the greater the potential for a merger-induced price increase (all else being equal). The 2010 Guidelines also codify two well-known economic tools: critical loss analysis and merger simulation. Section 4.1.3 provides an overview of critical loss analysis and explains that “[w]hen the necessary data are available,” the Agencies may use critical loss analysis to corroborate the results from the hypothetical monopolist test. Section 6.1 states that the Agencies

38 ProMedica Health Sys., Inc., FTC Docket No. 9346, 2012 FTC LEXIS 58, at *39–73 (Mar. 22, 2012) (Commission Opinion); Polypore Int’l, Inc., Docket No. 9327, 2010 FTC LEXIS 97, at *29–51 (Dec. 13, 2010) (Commission Opinion), aff’d, Polypore Int’l, Inc. v. FTC, 686 F.3d 1208 (11th Cir. 2012). But see Polypore Int’l, Inc., FTC Docket No. 9327, 2010 FTC LEXIS 96 (Dec. 13, 2010) (Rosch, Concurring) (criticizing reliance on precise upfront market definition). 39 See FTC Competition Enforcement Database, Merger Enforcement Actions, FED. TRADE COMM’N, http://www.ftc.gov/bc/caselist/merger/index.shtml. 40 2010 Guidelines, supra note 8, § 6.1. 41 1992 Guidelines, supra note 26, § 2.21 (“Some of the sales loss due to the price rise merely will be diverted to the product of the merger partner and, depending on relative margins, captur- ing such sales loss through merger may make the price increase profitable even though it would not have been profitable premerger.”). 2013] SURVEY OF EVIDENCE LEADING TO FTC SECOND REQUESTS 609 may use merger simulation studies when the data are available, but offers the caveat that they do not treat this evidence “as conclusive in itself.” Price-cost margins take on greater significance in the 2010 Guidelines. UPP calculations and critical loss analysis both depend on margins.42 In addition, the 2010 Guidelines state for the first time that margins can be relevant by themselves. Section 2.2.1 explains that “if a firm sets price well above incre- mental cost, that normally indicates either that the firm believes its customers are not highly sensitive to price . . . or that the firm and its rivals are engaged in coordinated interaction.”43 The study results indicate that FTC staff usually does not rely on any of these forms of economic evidence in recommending Second Requests and that the 2010 Guidelines do not appear to have affected their usage at this stage of investigations. In no initial-phase investigation that led to a Second Request over the four years of this survey did staff calculate the predicted upward pricing pressure of a transaction or conduct a merger simulation, including one based on a willingness-to-pay model. Likewise, critical loss analysis appears to have been used sparingly. FTC staff performed a critical loss analysis in only a single initial-phase investigation leading to a Second Request during the four-year survey period. In that case, critical loss analysis was informative but not determinative as to the relevant market, and staff considered a variety of other evidence bearing on market definition. For mergers that received Second Requests, agency staff appeared to rely somewhat more often on margins, standing alone, in initial-phase merger investigations. In two transactions, staff pointed to the possibility of ongoing coordination based on high or increasing margins in industries with excess capacity. In another transaction, staff relied on one of the merging firm’s high margins as evidence that it was exercising market power and faced limited competition. In a fourth transaction, staff relied in part on margins for different groups of customers to conclude that price discrimination was unlikely. Two of these transactions were reviewed before and two were reviewed after the issuance of the 2010 Guidelines.

42 Margins are also used in the hypothetical monopolist test, which has been part of the Merger Guidelines since 1982, and can help validate merger simulation results. 43 See also 2010 Guidelines, supra note 8, § 4.1.3 (“Unless the firms are engaging in coordinated interaction (see Section 7), high pre-merger margins normally indicate that each firm’s product individually faces demand that is not highly sensitive to price.”). The 2010 Guidelines note, however, that “high margins are not in themselves of antitrust concern.” Id. § 4.1.3 n.6; see also id. § 2.2.1 n.3 (“High margins can be consistent with incumbent firms earning competitive returns.”). 610 ANTITRUST LAW JOURNAL [Vol. 78

The FTC staff’s infrequent reliance on upward pricing pressure, merger simulation, critical loss analysis, and margins in the surveyed initial-phase investigations should not be surprising, given that the information needed for these calculations is usually not available early in an investigation.44 These methodologies are likely to have greater significance at later stages of investigations or in litigation. For example, the FTC relied on merger simulation in its ProMedica and Western/Giant merger challenges,45 high margins to show ongoing coordination in Swedish Match,46 and critical loss analysis to define the relevant market in CCC/Mitchell and Arch Coal.47 The Antitrust Division has also offered these types of evidence in its merger challenges.48 In contrast, FTC staff appears to rely on diversion ratios with some frequency in initial-phase investigations that led to Second Requests, particularly in hospital mergers. Over the survey period, staff calculated diversion ratios in fifteen percent of transactions raising unilateral effects concerns;49 there was virtually no change in this rate before and after the 2010 Guidelines. FTC economists are often able to calculate preliminary diversion ratios in early stages of hospital merger investigations because of the availability of dis- charge data from state health agencies and analytic firms. Calculation of and

44 Cf. Elizabeth M. Bailey et al., Merger Screens: Market Share-Based Approaches Versus “Upward Pricing Pressure,” ANTITRUST SOURCE, Feb. 2010, at 1, http://www.americanbar.org/ content/dam/aba/publishing/antitrust_source/Feb10_Leonard2_25f.authcheckdam.pdf (“[A] lack of reliable estimates of the three key inputs would limit the extent to which UPP can be a useful screen in the first thirty days.”). 45 ProMedica Health Sys., Inc., FTC Docket No. 9346, 2012 FTC LEXIS 58, at *140–46 (Mar. 28, 2012) (Commission Opinion) (simulation study that predicted higher prices from the transaction, while “not central to our reasoning . . . further confirms our conclusions”); FTC v. Foster, 2007-1 Trade Cas. (CCH) ¶ 75,725 at 107,994 (D.N.M. 2007) (rejecting the FTC’s merger simulation because it “did not consider important and necessary issues”). But see ProMedica Health Sys., Inc., FTC Docket No. 9346, 2012 FTC LEXIS 60, at *4–13 (Mar. 28, 2012) (Rosch, Comm’r, concurring) (criticizing the use of merger simulation). In the Rockford hospital case, the defendants asserted that FTC’s failure to perform a merger simulation under- mined its prima facie case, an argument the district court rejected. See FTC v. OSF Healthcare Sys., 852 F. Supp. 2d 1069, 1086 (N.D. Ill. 2012). 46 FTC v. Swedish Match N. Am., Inc., 131 F. Supp. 2d 151, 168 (D.D.C. 2000) (stating that “wide margins” in the presence of “falling demand and excess capacity” indicated “anticompetitive behavior already exhibited within the market”). In Swedish Match, the FTC also advanced the position, consistent with the UPP concept, that high margins and diversion ratios suggest a likelihood of unilateral effects. See id. at 169 (“High margins and high diversion ratios support large price increases, a tenet endorsed by most economists.”). 47 FTC v. CCC Holdings Inc., 605 F. Supp. 2d 26, 40 (D.D.C. 2009); FTC v. Arch Coal, Inc., 329 F. Supp. 2d 109, 122 (D.D.C. 2004). 48 See, e.g., United States v. H&R Block, Inc., 833 F. Supp. 2d 36, 63–64, 86–88 (D.D.C. 2011) (finding the DOJ’s simulation study to have “some probative value” and its critical loss analysis to be helpful in validating the relevant market); United States v. Oracle Corp., 331 F. Supp. 2d 1098, 1170 (N.D. Cal. 2004) (concluding that the DOJ’s simulation study was “unreliable”). 49 These results exclude diversion ratios derived from market shares. 2013] SURVEY OF EVIDENCE LEADING TO FTC SECOND REQUESTS 611 reliance on diversion ratios is common in more advanced stages of FTC investigations and litigation involving unilateral effects concerns.50

TABLE 10: USE OF ECONOMIC EVIDENCE BEFORE AND AFTER THE 2010 GUIDELINES

Pre-2010 Guidelines Post-2010 Guidelines Number Frequency Number Frequency Critical Loss Analysis 1 3% — — Margins as Independent Consideration 2 5% 2 5% Upward Pricing Pressure — — — — Merger Simulation — — — — Diversion Ratios 5 15%* 5 14%* * Based on the 68 investigations in the study involving unilateral effects concerns.

D. OTHER FORMS OF EVIDENCE FIRST DESCRIBED IN THE 2010 GUIDELINES The 2010 Guidelines identify several other forms of evidence that, although known to be considered by the Agencies,51 were not mentioned in the 1992 Guidelines. The 2010 Guidelines, for example, state that the Agencies consider natural experiments and the existence of buyer power when assessing the likely competitive effects of a transaction.52 In addition, the 2010 Guide- lines identify—apparently, for the first time—the financial terms of a transaction as a relevant consideration in merger analysis at the Agencies.

TABLE 11: USE OF OTHER TOPICS BEFORE AND AFTER THE 2010 GUIDELINES Pre-2010 Guidelines Post-2010 Guidelines Number Frequency Number Frequency Credible Buyer Power 2 5% 9 23% Natural Experiments 5 13% 5 13% Financial Terms of Deal 5 13% 2 5%

50 See, e.g., FTC v. Whole Foods Mkt., 548 F.3d 1028, 1044 (D.C. Cir. 2008) (Tatel, J.); CCC Holdings, 605 F. Supp. 2d at 70–71; ProMedica, 2012 FTC LEXIS at *120–23. 51 For example, the Merger Guidelines Commentary states that the Agencies consider actual effects from consummated mergers, natural experiments, and buyer power when assessing the likely competitive effects of a transaction. See Commentary, supra note 1, at 2, 10, 17–18. 52 In addition, evidence of actual anticompetitive effects from a consummated merger “is given substantial weight.” 2010 Guidelines, supra note 8, § 2.1.1. 612 ANTITRUST LAW JOURNAL [Vol. 78

1. Natural Experiments Natural experiments are historical events such as recent mergers, entry, ex- pansion, or exit in the relevant market or similar markets that may be informative as to the competitive effects of the transaction under review.53 According to the survey results, FTC staff relied on natural experiments in about the same share of cases that received Second Requests, thirteen percent, both before and after the 2010 Guidelines. In these investigations, staff had evidence regarding changes in price or output or how customers responded to a prior transaction, entry, or exit in the relevant market. These market events were often informative not only as to the overall competitive effects of a transaction but also to other issues, such as market definition or the existence of buyer power. Staff’s memoranda usually noted, however, that the Second Request investigation would need to identify and account for any confounding factors, such as changing input prices, before placing significant weight on these market events.

2. Power Buyers The federal courts have recognized a “power buyer” defense for many years,54 but, until recently, the Merger Guidelines did not expressly mention power buyers. The 2010 Guidelines recognize this defense for the first time, but describe several limitations to its application.55 Since the issuance of the 2010 Guidelines, FTC staff has indicated with greater frequency in Merger Screening memoranda that power buyer claims may be valid and worthy of further investigation. This change was most nota- ble in the technology and “other” industry categories. In these memoranda, FTC staff was more likely to credit power buyer claims when customers had disciplined suppliers in the past, could credibly sponsor entry or self-supply, could switch suppliers with ease, or could use middlemen or group purchasing organizations to prevent price discrimination. On the other hand, both before and after the issuance of the 2010 Guidelines, FTC staff appeared to have consistently rejected arguments—at least in Merger Screening memoranda— that the presence of large customers, standing alone, could constrain a post- merger price increase.

53 Id. § 2.1.2. The Agencies also consider evidence based on variations among similar markets. See id. 54 See, e.g., FTC v. Tenet Health Care Corp., 186 F.3d 1045, 1054 (8th Cir. 1999); United States v. Baker Hughes Inc., 908 F.2d 981, 986 (D.C. Cir. 1990). 55 2010 Guidelines, supra note 8, § 8. 2013] SURVEY OF EVIDENCE LEADING TO FTC SECOND REQUESTS 613

3. Financial Terms of a Transaction The 2010 Guidelines state that the financial terms of a transaction may be a relevant consideration in the Agencies’ competitive effects analysis. Sec- tion 2.2.1 explains that “a purchase price in excess of the acquired firm’s stand-alone market value may indicate that the acquiring firm is paying a premium because it expects to be able to reduce competition or to achieve efficiencies.”56 The survey results indicate a number of interesting findings with respect to the FTC’s consideration of the financial terms of transactions. First, investigating staff cited this source of evidence in significantly fewer Merger Screening memoranda following the release of the 2010 Guidelines. Second, the most frequent inference FTC staff drew from this type of evidence was that the transaction was unlikely to be anticompetitive. For example, in one acquisition, a low purchase price of the target helped corroborate the parties’ General Dynamics defense.57 Only in two investigations in the survey period did financial terms appear to raise concerns.58 Third, FTC staff has found a wider range of financial terms to be informative than those identified in the 2010 Guidelines, including the extent of revenue synergies, changes in debt levels, valuation of comparable companies, and the degree to which the transaction satisfied the buyer’s own acquisition criteria.

IV. CONCLUSION The survey results, when viewed in conjunction with prior merger data reports and other evidence, support the notion that the types of evidence on which the FTC relies varies at different stages of an investigation. For example, as an investigation draws closer to a final enforcement decision, customer complaints and high entry barriers increase in significance, while hot documents decrease in significance. Economic tools with substantial data require- ments, such as critical loss analysis and merger simulation, appear to be used primarily in later stages of investigations, while diversion ratios and margins (standing alone) appear to be used with some frequency throughout investigations.

56 The proper inferences to be drawn from the financial terms of a transaction can be counter- intuitive. See Darren S. Tucker & Kevin L. Yingling, Keeping the Engagement Ring: Apportion- ing Antitrust Risk with Reverse Breakup Fees, ANTITRUST, Summer 2008, at 70 (finding that higher reverse breakup fees are correlated with lower antitrust risk). 57 United States v. Gen. Dynamics Corp., 415 U.S. 486, 496–510 (1974) (permitting respon- dents to rebut an initial presumption of illegality with evidence that the competitive significance of one of the merging parties would decline absent the merger). 58 In one other transaction, legal and economic staff both found significant that the buyer paid a substantial premium for the target company but drew differing conclusions from that fact. 614 ANTITRUST LAW JOURNAL [Vol. 78

The survey results are also generally consistent with the Agencies’ position that the 2010 Merger Guidelines reflected existing practice and would have little effect on FTC merger investigations. Unilateral effects, coordinated effects, and potential competition remain the workhorses of horizontal merger analysis, and FTC staff continues to define markets in all cases where staff makes a Second Request recommendation. The results are also in accord with the Agencies’ claim that prior to the issuance of the 2010 Guidelines, agency staff was already employing the theories of harm and sources of evidence appearing for the first time in those Guidelines. Thus, the survey results are consistent with the evolutionary approach of the Guidelines and do not reveal any sharp departures from prior FTC practice. One unexpected finding is that subsequent to the 2010 Guidelines, FTC staff has referred less frequently to the financial terms of a transaction and to product variety concerns, two of the new topics in those Guidelines. This could be due to stricter standards in the 2010 Guidelines for these topics, better counseling to avoid the creation of relevant documents, or a change in the types of transactions reported to the FTC. In contrast, FTC staff has found power buyer claims—another topic appearing for the first time in the 2010 Guidelines—more likely to be credible and worthy of further investigation since the 2010 Guidelines were issued. This could indicate that FTC staff had been applying a stricter standard for power buyer claims prior to the 2010 Guidelines, that the inclusion of this topic in the 2010 Guidelines has en- couraged outside counsel to develop these claims, or that the ever-changing merger environment resulted in more cases where this issue arose. Another noteworthy finding is that FTC staff now raises non-price concerns, such as loss of product or service quality, in the majority of investigations leading to FTC Second Requests. These concerns arise particularly often in technology sector transactions. Although a greater focus by FTC staff on non-price concerns may present additional challenges for merging parties, it should be noted that even in investigations where staff cited non-price concerns, price concerns were usually still the focus. Finally, the survey results do not lend support for certain critiques of the 2010 Guidelines. For example, concerns that UPP might supplant market definition and concentration in unilateral effects analysis are not borne out by the study results, at least at the FTC. Likewise, the recognition of exclusionary conduct as a relevant form of competitive concern in the 2010 Guidelines has not resulted in a greater number of FTC Second Requests (in absolute or relative terms) prompted in part by this concern. Finally, the overall drop in the rate of Second Requests at the FTC since the 2010 Guidelines does not appear 2013] SURVEY OF EVIDENCE LEADING TO FTC SECOND REQUESTS 615 consistent with concerns that the greater analytical flexibility described in the 2010 Guidelines would lead to more aggressive merger enforcement.59

59 As with the other findings noted above, this result might reflect a change in the distribution of the types of mergers occurring, rather than a change in enforcement philosophy. The overall rate of FTC Second Requests was 2.1 percent in the two years prior to the 2010 Guidelines and 1.5 percent in the two years after. The rate of FTC Second Requests as a percentage of transactions cleared to the agency declined from 16 percent to 14 percent. The underlying data for these calculations are on file with the author. 616 ANTITRUST LAW JOURNAL [Vol. 78

APPENDIX Tables 5a through 5e indicate the concentration levels for each market identified by staff in Merger Screening memoranda as requiring further review through a Second Request. Results are provided by industry category. These tables reflect all of the transactions in the survey but exclude markets raising potential competition concerns and markets for which agency staff was unable to estimate market shares due to insufficient information. Table 5 is the aggre- gate of Tables 5a through 5e.

TABLE 5A: CHANGE IN CONCENTRATION—PHARMACEUTICAL MARKETS

Change in HHI 100– 200– 600– 1,000– 1,500– 0–99 2,500+ TOTAL 199 599 999 1,499 2,499 0–1,499 0 1,500–2,499 3 3

Merger 2,500–3,999 11 3 3 17 - HHI 4,000–5,999 3 4 5 8 5 1 26 Post 6,000+ 1 2 5 7 15 TOTAL 3 0 19 10 11 10 8 61

TABLE 5B: CHANGE IN CONCENTRATION—HEALTHCARE PROVIDER MARKETS

Change in HHI 100– 200– 600– 1,000– 1,500– 0–99 2,500+ TOTAL 199 599 999 1,499 2,499 0–1,499 1 1 1,500–2,499 1 1 2

Merger 2,500–3,999 1 3 1 1 6 - HHI 4,000–5,999 1 1 3 1 6 Post 6,000+ 1 1 2 4 TOTAL 0 0 4 5 2 5 3 19 2013] SURVEY OF EVIDENCE LEADING TO FTC SECOND REQUESTS 617

TABLE 5C: CHANGE IN CONCENTRATION—ENERGY AND CHEMICAL MARKETS

Change in HHI 100– 200– 600– 1,000– 1,500– 0–99 2,500+ TOTAL 199 599 999 1,499 2,499 0–1,499 0 1,500–2,499 1 1 1 3

Merger 2,500–3,999 2 2 3 1 8 - HHI 4,000–5,999 1 1 3 5 Post 6,000+ 1 1 TOTAL 1 0 3 4 5 4 0 17

TABLE 5D: CHANGE IN CONCENTRATION—TECHNOLOGY MARKETS

Change in HHI 100– 200– 600– 1,000– 1,500– 0–99 2,500+ TOTAL 199 599 999 1,499 2,499 0–1,499 0 1,500–2,499 0

Merger 2,500–3,999 1 1 2 2 6 - HHI 4,000–5,999 1 1 1 4 7 Post 6,000+ 1 1 1 3 TOTAL 2 0 2 3 7 1 1 16

TABLE 5E: CHANGE IN CONCENTRATION—OTHER MARKETS

Change in HHI 100– 200– 600– 1,000– 1,500– 0–99 2,500+ TOTAL 199 599 999 1,499 2,499 0–1,499 0 1,500–2,499 1 1

Merger 2,500–3,999 1 1 4 1 7 - HHI 4,000–5,999 1 1 2 4 Post 6,000+ 3 1 4 TOTAL 1 0 2 5 2 5 1 16