LECG, Loyola University
Horizontal Merger Guidelines Review Project
HMG Review Project - Comment, Project No. P092900 James Langenfeld [Director, LECG, and Adjunct Prof., Loyola University Chicago School of Law. Former Director for Antitrust, Bureau of Economics, FTC, and Honoree, Celebration of the Twentieth Anniversary of the 1982 Merger Guidelines.] As a contributor to the 1992 Horizontal Merger Guidelines ("HMGs") revision and the 1993 Statements of Antitrust Enforcement Policy in Health Care, I know the challenges the agencies face in potentially revising the HMGs and commend them for obtaining public input. This summary highlights my comments contained in the first attachment, and references the question numbers. The Agencies appear to welcome other comments on the Merger Guidelines, so I am also attaching two articles where I recommend revising the1984 Non-Horizontal Merger Guidelines because the Agencies indicate they will carefully analyze potentially important non-horizontal mergers, and the 1984 Guides do not describe Agency practice or incorporate the substantial learning and experience since 1984. 1. Usually one or two critical issues determine whether a merger should be investigated, but the Agencies should be cautious in eliminating steps if they plan on challenging a merger. 2. The Agencies and outside parties often use evidence or analyses of competitive effects mentioned, so the HMGs should describe them and when they are useful. 3. The HMGs should explain that functional substitutes may be a starting point for market definition, but there should be a next step that estimates what portion of the functional substitutes economically constrains the pricing of merging firms. There should be no radical changes in current HMGs market definition, because there is no consensus on how to improve it. 4. The "next best substitute" or "smallest market" principles help establish a systematic basis for defining markets when there is no robust evidence of the own price elasticity of demand. 6. & 7. Defining geographic markets and calculating market shares based on the locations of customers or of suppliers depends on the nature of competition for the products at issue, and the HMGs' should explain that. 9. The HHI thresholds do not reflect Agency practice in most instances, and anecdotal evidence suggests the post merger HHIs trigger level should be 2,500. 10. The HMGs should explain that there can be a reduction in competition in part of a broader market with substantially differentiated products. Diversion ratios and price/cost margins are useful in evaluating unilateral effects, and should be one of a set of economic analyses for determining overall competitive effects. It is much more reliable to evaluate direct competition between the merging firms with differentiated products than to infer diversion ratios from market shares. Merger simulation models should be discussed, including their limitations. The HMGs should have a more detailed discussion of repositioning, and the evidence that the agencies consider. 11. & 12. The HMGs' discussion of price discrimination should address in more detail the seller's ability to target customers and prevent arbitrage. Any discussion should be directly linked to buyer power and customers' ability to switch to substitutes, vertically integrate, or create new suppliers. 13. In most - but not all - markets, there is no way to accurately assign market shares to uncommitted or committed entrants, so entry analysis is often similar for both. 18. The HMGs should explain how remedies address competitive effects, and the evidence indicating the remedies will be effective. 19. The "Commentary on the Horizontal Merger Guidelines" has proven of limited usefulness in determining how the agencies will act. Including these types of examples in the HMGs will likely lead to attempts to characterize mergers as one of the examples, when there are seldom two mergers that have identical analyses. 20. The Agencies should learn from their successes and mistakes, so the HMGs should reflect retrospectives.