The relevant market: A concept still in search of a definition

Journal of Competition Law and Economics

Volumn 5, Issue 2, 2009, Pages 297-333

  ten Kate, Adriaan ^a   Niels, Gunnar ^a  

^a Oxera ^*  (United Kingdom)

Author keywords

[No Author keywords available]

Indexed keywords

EID: 66849101768     PISSN: 17446414     EISSN: 17446422     Source Type: Journal    
DOI: 10.1093/joclec/nhn029     Document Type: Article

References (45)

1. For an extensive description of the emergence of the relevant-market concept in antitrust analysis, see Gregory Werden, The History of Antitrust Market Delineation, 76 MARQ. L. REV. 123 (1992 ).
2. U.S. Department of Justice, Merger Guidelines (1982), reprinted in 4 Trade Reg. Rep. (CCH) 102 and U.S. Department of Justice and the Federal Trade Commission, Horizontal Merger Guidelines (1992). There was some guidance provided in the 1968 Guidelines, but that was not considered a proper definition of the relevant-market concept (it was phrased in terms of "grouping of sales" and "area of effective competition among the included sellers"). According to the Task Force on Productivity and Competition headed by George Stigler, this guidance was "so loose and unprofessional as to be positively embarrassing."
3. See the Report of the Task Force in 1 J. Reprints for Antitrust Law & Economics (1969), p. 827.
4. See Gregory Werden, The 1982 Merger Guidelines and the Ascent of the Hypothetical Monopolist Paradigm, 71 ANTITRUST L. J. 253 (2003).
5. Bureau of Competition, Merger Enforcement Guidelines, April 1991.
6. European Commission. Commission Notice on the definition of the relevant market for the purpose of Community competition law, 1997 O.J. (C 372), par. 5.
7. There is much recent literature on how the HMD is applied in practice, ranging from articles dealing with critical loss analysis to operationalize the HMD - see, most recently, Joseph Farrell and Carl Shapiro, Improving Critical Loss Analysis, ANTITRUST SOURCE (2008)
8. and Øystein Daljord, Lars Sørgard and Øyvind Thomassen, The SSNIP test and market definition with the aggregate diversion ratio: A reply to Katz and Shapiro, J. COMPETITION L. & ECON. 4: 263 (2008) - to articles analyzing how competition authorities have applied HMD in real-world cases - for example, Graeme Reynolds and Chris Walters, The use of customer surveys for market definition and the competitive assessment of horizontal mergers, J. COMPETITION L. & ECON. 4:411 (2008)
9. and Malcolm B. Coate and Jeffrey H. Fischer, A practical guide to the hypothetical monopolist test for market definition, J. COMPETITION L. & ECON., forthcoming. We do not discuss these here.
10. Note that the point on supply characteristics we are making in this article is different from another criticism that is sometimes levelled against the demand-side focus of the Merger Guidelines, which is that even under the Guidelines one sometimes delineates markets based on supply-side substitution, for entirely pragmatic reasons. For example, size 7 and size 9 shoes are not demand substitutes, but making this distinction does not add anything to the competition analysis given that the supply conditions (such as producer market shares) for both sizes will be virtually the same. In fact, note 14 in the 1992 Guidelines explicitly refers to this usage of supply-side substitution "as a matter of convenience" (although the purist would note that even in this case the Guidelines would still speak of markets in the plural rather than of one market).
11. U.S. Department of Justice, Merger Guidelines (1984), reprinted in 4 Trade Reg. Rep. (CCH) 103.
12. Morris A. Adelman, Economic Aspects of the Bethlehem Opinion, 45 VA. LAW REV. 684 (1959).
13. A similar idea that preceded the notion of hypothetical monopolization was that of hypothetical cartelization of the market, that is, which firms would need to be included in the cartel to make the price increase sustainable? For reasons that we do not expand upon here, the notion of hypothetical cartelization suffers from a number of conceptual problems and was soon abandoned in favour of hypothetical monopolization.
14. Already before the 1984 Guidelines were issued, Werden argued that the 1982 Guidelines should be interpreted in terms of profit maximization. Gregory J. Werden, Market Delineation and the Justice Department's Merger Guidelines, 1983 DUKE L. J. 514, 542 (1983).
15. Rather than citing examples of unawareness of the difference between the break-even and the profit-maximizing versions, we refer to Gregory Werden who has frequently drawn attention to the distinction. See, for example, Gregory J. Werden, Demand Elasticities in Antitrust Analysis, 66 ANTITRUST L. J. 363 (1998).
16. For that to happen, there must be multiple local optima for the monopolist. An example of such a situation is given in Werden (1983), supra note 10, at p. 543.
17. See Gregory J. Werden, Market Delineation under the Merger Guidelines: A Tenth Anniversary Retrospective, 38 ANTITRUST BULL. 517, 538 (1993).
18. U.S. Department of Justice & Federal Trade Commission, Horizontal Merger Guidelines (1992), reprinted in 4 Trade Reg. Rep. 104.
19. As we explain below, this assumption generally exaggerates the price discipline other products exercise over the products within the group.
20. A less important addition is in the words "at least." The authors probably realized that a "large and significant increase in price" would also be sufficient. Technically, they could have achieved the same goal by suppressing the words "small but." According to Werden, "The word 'small' must have been intended to indicate that a price increase can be significant even though it is small." See Werden, Market Delineation, supra note 10, at 544.
21. One practical area where the difference is important is when the HMD is operationalized through critical loss analysis in real-world cases. There it can lead to approaches that are inconsistent with the HMD, as also noted by Joseph Farrell and Carl Shapiro, Improving Critical Loss Analysis, ANTITRUST SOURCE, Feb. 2008, at p. 3, http://www.abanet.org/antitrust/at-source/08/02/02-08.html.
22. Gregory J. Werden, MARKET DELINEATION ALGORITHMS BASED ON THE HYPOTHETICAL MONOPOLIST PARADIGM, U.S. Department of Justice, Antitrust Division, EAG Discussion Paper 02-8, August 2002.
23. In principle, it is possible that insiders produce some outside products at the same time, which would raise specific problems for the profit maximization by the hypothetical monopolist who is supposed to keep the prices of those outside products constant. We abstract from this possibility here.
24. See Werden, Retrospective, supra note 13, at p. 543.
25. The power set of a set is the set of all its subsets. Thus, groups of products are elements of the power set.
26. A partition of a set is a collectively exhaustive and mutually exclusive division of that set into subsets.
27. An example of this partitioning of the universe in the geographical dimension can be found in the investigation by the Mexican Federal Competition Commission of the distribution of liquid petroleum gas, in which the national territory was split up into thirty-seven regions considered to be the relevant markets. Comisión Federal de Competencia, Declaración de inexistencia de condiciones de competencia efectiva en el mercado de distribución del gas licuado de petróleo, Gaceta de Competencia Económica, Año 6, núm. 16, mayo-agosto 2003, pp. 275-342.
28. This point was also made in Gunnar Niels, The SSNIP Test: Some Common Misperceptions, 3:4 COMP. L. J. 267 (2004).
29. A minor addition is that our version of the HMD not only assumes that the terms of sale of all outside products are held constant (as in the original version) but also that the terms of sale of inside products outside the area are kept at the pre-merger level, an assumption that is absent from the official version of the HMD.
30. See Robert Pitofsky, New Definitions of Relevant Market and the Assault on Antitrust, 90 COLUM. L. REV. 1805 (1991).
31. It is important to note that these three factors are interrelated. In the theory of consumer demand, demand reduction is expressed in terms of own-elasticities and substitution between products in terms of cross-elasticities. In a Marshallian demand system, the own-elasticity of a product (taken as a positive number) equals one plus the relative-share-weighted sum of its cross elasticities.
32. This is the so-called adding-up restriction derived from the budget constraint. See, e.g., ANGUS DEATON & JOHN MUELLBAUER, ECONOMICS AND CONSUMER BEHAVIOR (Cambridge University Press, 1980).
33. Par. 1.0 of the Guidelines.
34. The failure of earlier Guidelines to address the issue of price reactions of outside products rendered their market definitions indeterminate. This holds for both the break-even and the profit-maximizing versions of the HMD.
35. See Gregory J. Werden, MARKET DELINEATION ALGORITHMS BASED ON THE HYPOTHETICAL MONOPOLIST PARADIGM, U.S. Department of Justice, Antitrust Division, EAG Discussion Paper 02-8, August 2002.
36. Additionally, if sufficient data were available to approximate the level of normal profits, one might as well use this data to determine market power directly (that is, market power defined as the ability to sustain profits above the competitive level for a significant period), and skip the intermediate stage of market delineation altogether. See Neils, supra note 25.
37. This is so unless outsiders were firms with infinite supply elasticities; keeping their prices fixed would be at odds with profit maximization and they might even incur losses.
38. Here we assumed that the prices of outside products are also at marginal cost and are kept at that level post-merger. If one would have assumed that the pre-merger prices of all the outside products are at their Bertrand equilibrium level, the results would have been even more dramatic. In fact, the price increase after the (non)merger would already be significant for values of d below twenty.
39. To get an idea of how substitutable the products are at d = 10, consider the case in which all prices are at forty and all quantities at sixty. From that position, an increase in the price for one of the products by one unit - that is, by 2.5 percent - would induce a decline in demand for that product of twenty-one units, of which one unit is demand reduction and twenty units - that is, one-third of the demand before the increase - leak away to the substitutes. In such circumstances, every brand of every product would constitute a relevant market by itself, and a merger between two firms producing different brands of the same product would not even be considered to have horizontal overlaps.
40. Due to the symmetry assumption, it is not clear which of the two products should join the original product in its relevant market. However, a small tremble in the parameters would be sufficient to break the tie.
41. The two-to-one relation between break-even and optimal prices was mentioned in Werden, Market Delineation, supra note 10, at 543, and more recently in Farrell & Shapiro, supra note 17, at 17.
42. Note that we are not advocating that the definition of the relevant market for antitrust purposes should only depend on demand substitution factors; see further below.
43. The prices of complements would go down.
44. See DEATON & MUELLBAUER, supra note 27; at ch. 1.2.
45. Occasionally, the second or third next best substitute may be sufficient to reach the cut-off level.