Beaton-Wells, C --- "Mergers Without Markets? Unilateral Effects Analysis in the United States and its Prospects in Australia" [2006] UMelbLRS 6

Last Updated: 29 September 2009

Mergers without markets? Unilateral effects
analysis in the United States and its prospects
in Australia

Caron Beaton-Wells*

In both the United States and Australia an initial but crucial step in the
analysis performed for the purposes of merger regulation has been definition
of the relevant market(s) in which competition may be affected. In recent
years, the federal agencies in the United States have adopted an approach
that excludes this step. Instead, their approach has been to examine the
potential unilateral effects of a merger by assessing “directly”, using empirical
techniques, the possibility of a supra-competitive price increase by the
merged firm. This has been seen as a preferable approach in many ways to
the traditional structural analysis to which market definition is integral. Is it
likely that Australian regulators and courts would adopt a similar approach?
This article addresses that question, concluding that there are good reasons
for regarding such a prospect as remote.

INTRODUCTION

Recent developments in antitrust enforcement in the United States reduce considerably, if not remove altogether, the requirement to define a relevant market for the purposes of assessing the competitive implications of a particular transaction.1 In the merger context, this development generally is associated with an analysis of the possible unilateral effects of the acquisition.2 Driven largely by the growth in differentiated products and services industries and advances in empirical techniques for predicting merger outcomes, unilateral effects analysis has been the primary focus of merger activity by the United States’ enforcement agencies since the early 1990s.3 Those agencies enforce the prohibition in s 7 of the Clayton Act 1914 on acquisitions “where in any line of commerce in any section of the country, the effect of such acquisition may be substantially to lessen competition, or to tend to create a monopoly”.

In Australia market definition is regarded as an essential preliminary step in applying those
prohibitions under Pt IV of the Trade Practices Act 1974 (Cth) (TPA) that necessitate an assessment of market power and anticompetitive effects.4 Such prohibitions include s 50 which, not unlike s 7 of the United States’ statute, is directed at acquisitions that have the effect or are likely to have the effect of substantially lessening competition in a market. This is the test that is applied by the Australian Competition and Consumer Commission (ACCC) in deciding whether to provide a clearance for a

* BA/LLB (Hons); LLM (Melb); PhD (Melb); Senior Lecturer, Melbourne Law School, University of Melbourne. The author is
grateful to Darryl Biggar, Frances Hanks and Philip Williams for their comments on an earlier draft of this manuscript. Any
errors or omissions are the responsibility of the author alone.
1 See generally Keyte J and Stoll N, “Markets? We don’t Need no Stinking Markets! The FTC and Market Definition” (2004) 49
(Fall) Antitrust Bulletin 593.
2 This analysis in turn tends to be associated primarily with horizontal mergers. Different issues tend to arise in connection with
vertical and conglomerate (diversifying) mergers. See, for example, United States Department of Justice, Non-Horizontal
Merger Guidelines, June 14, 1984.
3 Baker J, “Unilateral Competitive Effects Theories in Merger Analysis” (1997) 11 Antitrust 21 at 21. Until the revision of the Department of Justice and Federal Trade Commission Horizontal Merger Guidelines in 1992, the focus had been very much on coordinated effects (that is, the potential for a merger to engender collusion between the merged firm and remaining rival firms). One of the more significant changes resulting from the revision of the Guidelines in 1992 was the introduction of the concept that a merger may result in lessened competition through unilateral effects, even where the merger does not increase the likelihood of successful coordinated interaction.
4 See ss 45, 46, 47 and 50 of the Trade Practices Act 1974 (Cth).
proposed merger.5 It is also the test that is applicable on the far less common occasion that a merger is challenged in the Federal Court. While a different test applies to the determination of applications for authorisation by the ACCC and, on review, the Australian Competition Tribunal (Tribunal),6 an important aspect of that test involves consideration, as under s 50, of the anticompetitive detriment likely to arise from the proposed acquisition.7 At the same time, the task of defining a market has proven to be one of the most challenging and controversial aspects of Australian competition law.8 It not only consumes a significant proportion of the time and expense involved in examining whether the relevant tests apply, but is also often
determinative of the outcome such that the substantive issues of market power and effects on
competition are never reached. For these reasons there is obvious merit in examining the
developments in the United States and assessing the prospects for their adoption in Australia.9 To that end, this article:
1. outlines the approach taken to assessing the competition implications of a merger in the United States and Australia under the regulatory guidelines in each jurisdiction and, in particular, the role of unilateral effects in that approach;
2. examines briefly the economic theory underpinning unilateral effects analysis, as developed principally in the United States;
3. considers the ramifications for market definition of the adoption of unilateral effects analysis, with reference to recent United States’ experience; and
4. assesses the prospects of the United States’ approach to unilateral effects analysis taking hold in Australia.

MERGER REGULATION AND UNILATERAL EFFECTS

In the United States the approach involved in determining the antitrust implications of a merger is set out in guidelines published jointly by the Department of Justice and Federal Trade Commission (US Guidelines).10 The US Guidelines articulate a sequence of steps, commencing with market

5 The ACCC has a long-standing practice of entertaining applications for informal clearances. An informal clearance is essentially a non-statutory promise by the ACCC not to prosecute a proposed merger on the grounds that it does not raise substantial competition concerns. The 2005 amendments that the government proposed to the Act (see Trade Practices Legislation Amendment Bill 2005), in accordance with recommendations made by the Dawson Committee of Inquiry in its Review of the Competition Provisions of the Trade Practices Act, 2003, included the introduction of a parallel formal clearance procedure. However, the Bill was amended by the Senate, excising the schedule dealing with mergers and, as at the date of publication of this article, the fate of the Bill remains unclear.
6 Pursuant to the amendments proposed to be made by the Trade Practices Legislation Amendment Bill 2005 (see n 5 above), applications for merger authorisations will be determined by the Tribunal rather than the ACCC.
7 The test is whether the ACCC is satisfied that the acquisition “would result or be likely to result in such a benefit to the public that [it] should be allowed to take place” (see s 88(9)). While it does not expressly involve a weighing process between public benefit and anticompetitive detriment, the test, as applied in practice, is whether there is an overall or net public benefit once all the circumstances including any anticompetitive or other detriment has been taken into account.
8 The question of the relevant market has been described as “the most vigorously (and expensively) litigated and discussed question in our courts and tribunals”: Baxt R, “The Australian Concept of Market – How it Came to Be” in Richardson M and Williams P (eds), The Law and the Market (1995) 10 at 28.
9 To date the United States’ developments have not received much attention in published Australian competition law literature.
Cf the brief discussion in Robertson D, “The Regulatory Assessment of Mergers (and Things like Mergers)” (2000) 7 CCLJ 201
and more recently and in greater detail, Werden G and Hay D, “Bringing Australian Merger Control into the Twenty-First Century by Incorporating Unilateral Effects” (2005) 13 CCLJ 119 (an article published after the present article was submitted for publication). Werden and Hay argue that the Australian Guidelines should be amended to adopt the structure of the US Guidelines, differentiating between and then in turn giving more extensive treatment to each of unilateral effects and coordinated effects theories. The article does not expressly advocate the abandonment of the market definition-market share approach and does not deal with the arguments against such abandonment, as set out below. Rather, it appears to suggest that unilateral effects analysis should supplement the structural paradigm and, indeed, should be preferred in the case of mergers in differentiated products industries given the limitations of the latter in that context..
10 Department of Justice and Federal Trade Commission, Horizontal Merger Guidelines (April 2, 1992, as revised April 8, 1997) (US Guidelines).
definition,11 followed by an inquiry into market shares and levels of concentration that raise
presumptions regarding the legality or illegality (as the case may be) of the merger,12 followed by an assessment of anticompetitive effects. Such effects are divided into two broad categories, involving:
1. coordinated interaction between the merged firm and other rivals (coordinated effects);13 and
2. unilateral action on the part of the merged entity (unilateral effects).14

Having considered these two possible scenarios arising from the merger (bearing in mind that
they are not intended to be mutually exclusive), the US Guidelines deal next with the condition of entry in order to determine whether it would be timely, likely and sufficient to counteract any anticompetitive effects.15
This is followed by evaluation of whether there are efficiencies likely to be generated by the merger that may offset such effects16 and finally whether the merger would prevent
the exit of a so-called “failing firm”.17 The analytical framework articulated in the US Guidelines is applied, not only by the federal agencies in reviewing merger proposals, but also by the courts in determining the legality of a merger under s 7 of the Clayton Act.18

While conceptually consistent with the United States’ approach, the guidelines published by the ACCC in relation to mergers (Australian Guidelines) organise the analysis slightly differently.19 Under those guidelines, following market definition20 and calculation of market shares and concentration levels,21 an analysis of possible anticompetitive outcomes (depending on whether the specified market concentration thresholds are met) proceeds by examination of a series of factors in order as follows –
import competition,22 barriers to entry,23 and then other factors (including both structural and
behavioural factors).24 These factors reflect, to a large extent, the non-exhaustive list provided in s 50(3) of the TPA of factors that a court must consider in deciding whether or not an acquisition falls foul of the prohibition in that section.

While “unilateral effects” as such, are not singled out explicitly for attention,25 they underpin
almost all of the analysis provided for in the Australian Guidelines (market shares, barriers to entry, countervailing power and so on are all seen as factors that either will facilitate or detract, as the case may be, from an exercise of unilateral power on the part of the merged entity). To some extent this is a function of the fact that, prior to 1993, the test under s 50 of the TPA was concerned with whether the acquisition would create or strengthen substantially a position of dominance on the part of the merged firm. It is generally understood that that test was amended, lowering the standard from dominance to substantial lessening of competition, so as to catch not only acquisitions that would

11 US Guidelines, s 1 (1.0-1.322)
12 US Guidelines, s 1 (1.4-1.522).
13 US Guidelines, s 2.1.
14 US Guidelines, s 2.2
15 US Guidelines, s 3.
16 US Guidelines, s 4.
17 US Guidelines, s 5.
18 Rill J, “Practicing What They Preach: One Lawyer’s View of Econometric Models in Differentiated Products Mergers” (1997)
5 George Mason Law Review 393 at 393.
19 Australian Competition and Consumer Commission, Merger Guidelines, June 1999 (Australian Guidelines). See the
diagrammatic depiction of the analysis in Figure 1.
20 Australian Guidelines, [5.26], [5.34]-[5.86].
21 Australian Guidelines, [5.27]-[5.28], [5.87]-[5.103].
22 Australian Guidelines, [5.29], [5.104]-[5.114].
23 Australian Guidelines, [5.30], [5.115]-[5.128].
24 Australian Guidelines, [5.31], [5.129]-[5.179].
25 There is a brief reference to the distinction between unilateral market power and coordinated market power at [5.11] of the Australian Guidelines. There are additional references to unilateral effects in [5.88], [5.89] and [5.97]. For a detailed discussion of the inadequacies of the Guidelines in this regard, see Werden and Hay, n 9.
enable the merged entity to exercise unilateral market power, but also those that might facilitate coordinated market conduct.26 That said, coordinated effects are not given the prominence that they are given in the US Guidelines.27 In the Australian Guidelines, such effects are dealt with in a discrete section towards the end of the guidelines,28 under the heading “Other factors”, together with factors such as the availability of substitutes, the removal of a vigorous and effective competitor, vertical integration, dynamic characteristics of the market and so on.29

Under the Australian framework, efficiencies are only treated as relevant to the issue of
anticompetitive effects to the extent that they may create or enhance a competitive constraint on the merged firm or undermine the conditions for coordinated conduct.30 The “failing firm” scenario is dealt with in the context of considering whether the merger would involve the removal of a vigorous and effective competitor.31

ECONOMIC THEORY OF UNILATERAL EFFECTS

In basic economic terms, an individual firm is understood to have unilateral market power if it can raise price above competitive levels without losing sales to rivals to such an extent as to make the price increase unprofitable.32 Broadly speaking, based on the traditional model of a dominant firm, economists recognise two situations in which unilateral market power might arise. The first concerns cost differences and the second, differentiated products.33

The first situation involves a dominant firm and a number of “fringe” competitors producing a
homogeneous product. In this model, the only difference between the dominant firm and the fringe firms is that the former has a substantial cost advantage over each of the latter. As a result the dominant firm is able to set its profit-maximising price significantly above its marginal cost because of the fringe firms’ cost disadvantage. That disadvantage prevents them from expanding their sales at the price determined by the dominant firm.34

In the second situation, it is differences between competitors’ products, as perceived by
customers, rather than differences in their costs that enable a firm to exercise unilateral market power. There is extensive theoretical economic literature on the relationship between product differentiation and market power.35 For present purposes, however, it is sufficient to note that a firm, the products of

26 See Corones S, Competition Law in Australia (3rd ed, Lawbook Co., 2004) pp 296-297. This amendment was effected from 21 January 1993 by the Trade Practices Legislation Amendment Act 1992 (Cth), consequent upon the recommendations of the Senate Committee on Legal and Constitutional Affairs, Mergers, Monopolies and Acquisitions – The Adequacy of Existing Legislation Controls, Canberra, 1991 (Cooney Committee).
27 In practice also, a brief review of the ACCC’s decision-making with respect to merger proposals (submitted for informal clearance purposes) in recent years suggests that its focus has been primarily on unilateral effects. See the public competition assessments issued by the ACCC in the period 2003-2005, available at http://www.accc.gov.au (viewed 4 October 2005). There have been cases in the past, however, in which ACCC opposition to a merger has been based on potential coordinated effects: see, eg, Wattyl/Courtaulds [1996] ATPR (Com) 50-232. To the extent that generalisations are possible, it is probably true to say that where the merger is in a differentiated products industry, the ACCC’s focus has been on unilateral effects; whereas where the industry in question involves homogeneous products, the focus has been on the potential for coordinated conduct as a result
of the merger: Corones, n 26, p 324.
28 Australian Guidelines, [5.167]-[5.170].
29 Australian Guidelines, [5.129]-[5.179].
30 Australian Guidelines, [5.171]-[5.174].
31 Australian Guidelines, [5.138]-[5.147].
32 Starek III R and Stockum S, “What Makes Mergers Anticompetitive?: ‘Unilateral Effects’ Analysis under the 1992 Merger Guidelines” (1995) (Spring) 63 Antitrust Law Journal 801 at 803.
33 See, eg, Carlton D and Perloff J, Modern Industrial Organization (2nd ed, 1994) pp 158-159.
34 Carlton and Perloff, n 33, pp159-166.
35 For a survey of this literature (albeit now somewhat out-of-date), see Eaton B and Lipsey R, “Product Differentiation” in Schmalensee R and Willig R (eds), Handbook of Industrial Organization (1989).
which are differentiated from those of rivals and attract strong customer loyalty as a consequence, may be able to raise its price above competitive levels without incurring a sufficient reduction in sales as to make the price increase unprofitable.

For reasons outlined below, economists generally agree that, while market share is a factor that affects the ability of a firm to exercise unilateral market power, there is no particular market share threshold over which a firm may be presumed to have such power. Even a relatively large share, of itself, cannot be considered definitive.36 Rather than market share, a test commonly applied in determining whether or not a firm is able to exercise unilateral market power is a test of the profitability of a small but significant non-transitory price increase (SSNIP) imposed by the firm acting unilaterally. A firm that is able to impose a SSNIP profitably is said to enjoy unilateral market power.
This is the same test as is employed for the purposes of market definition (being, in the United States, the profitability of a price increase for a hypothetical monopolist).37 In both instances the profitability of the SSNIP turns on the extent to which customers would reduce their purchases in response to the increase.38 A firm will impose a supra-competitive price increase unilaterally only if so few marginal customers (that is, customers who would be responsive to a SSNIP) would switch to rival firms such that the price increase will remain profitable.39

In merger policy, unilateral effects analysis is concerned with the potential for a merger to create a firm that possesses unilateral market power in the sense described above. In other words, it is concerned with whether the merged firm will be in a position to raise prices unilaterally (or reduce output or otherwise act anticompetitively). The US Guidelines identify criteria that will be considered in determining the potential for unilateral anticompetitive effects arising from a merger in two different settings. The first setting is one in which firms are distinguished primarily by differentiated products (s 2.21) and the second is one in which firms are distinguished primarily by their capacities (s 2.22). In recent years, agency attention has focused primarily on mergers in the first, differentiated products, setting.

Underpinning the distinction between these two settings are two benchmark models of
competitive behaviour discussed in modern industrial organisation theory – the differentiated product

36 Starek and Stockum, n 32 at 804, citing Pitofsky R, “New Definitions of Relevant Market and the Assault on Antitrust” (1990) 90 Colum L Rev 1805 at 1810. For a recent explanation by an Australian author as to why market share/concentration information, on its own, is insufficient and, even potentially misleading, as an indicator of market power, see Biggar D, Competition Policy Without Market Definition, Internal ACCC Working Paper, 30 August 2005 (copy on file with author).
37 As has been pointed out recently (and discussed further below), the SSNIP test plays a dual role, both as a test for the competitive effects of a merger and as a test for determining market boundaries: King S, The SSNIP Test in Merger Analysis: A Critical Appraisal, Internal ACCC Working Paper (2005) (copy on file with author). The hypothetical monopolist version of the SSNIP test is adopted in the US Guidelines, s 1, and has been incorporated in similar form in several other jurisdictions, including Australia. Starting with the smallest possible group of competing products, the US Guidelines ask whether a hypothetical monopolist over that group of products would profitably impose at least a SSNIP, generally deemed to be about 5% lasting for the foreseeable future. If a significant number of customers respond to a SSNIP by purchasing substitute products having a considerable degree of functional interchangeability for the monopolist’s products, then the SSNIP would not be profitable. Accordingly, the product market must be expanded to encompass those substitute products that constrain the monopolist’s pricing. The product market is expanded until the hypothetical monopolist could profitably impose a SSNIP.
Similarly, in defining the geographical market, the guidelines hypothesise a monopolist’s ability profitably to impose a SSNIP, again deemed to be about 5%, in the smallest possible geographic area of competition. If customers respond by buying the product from suppliers outside the smallest area, the geographic market boundary must be expanded. The hypothetical monopolist test is endorsed in the Australian Guidelines, [5.44] in conjunction with the price elevation test ([5.41]-[5.42]), as articulated by the then Trade Practices Tribunal in Re Queensland Co-Operative Milling Association Ltd and Defiance Holdings Ltd (1976) 25 FLR 169 at 189-190; [1976] ATPR 40-012.
38 Note that in Australia, the SSNIP test incorporates supply-side as well as demand-side responses to a price increase by the relevant firm. See Queensland Wire Industries Pty Ltd v Broken Hill Proprietary Co Ltd [1989] HCA 6; (1989) 167 CLR 177.
39 See further, Biggar, n 36, explaining at [11] that economically, the profitability of an increase in price at the margin depends on two things that have nothing to do with market definition: the elasticity of the residual demand curve facing the firm and the marginal cost of the firm.
Bertrand model and the homogeneous product Cournot model.40 A more traditional economic analysis based on the theory of the dominant firm, discussed above, would have focused on cost differences rather than on capacity differences. However, a dominant firm model could be seen as a sub-category of s 2.22 of the US Guidelines to the extent that the dominant firm’s capacity is distinguished from the capacities of fringe firms based on its cost advantages. In this model fringe firms “are effectively capacity-constrained because they cannot expand their output without incurring higher costs”.41

Differentiated products

In relation to the first setting, the US Guidelines recognise that, as products may vary in their degree of substitutability, competition between firms may be “localised” so that a firm competes more directly with those firms that sell relatively close substitutes. Based on this concept of localised competition, a merger of two firms that produced close substitutes prior to the merger may enable the merged entity to raise prices unilaterally given that:

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, capturing such sales loss through merger may make the price increase profitable even though it would not have been profitable premerger.42

With this potential scenario in mind, substantial unilateral price elevation in a market for
differentiated products is said under the guidelines to require that two conditions be satisfied: (1) there be a significant share of sales in the market accounted for by consumers who regard the products of the merging firms as their first and second choices, and (2) repositioning of the non-parties’ product lines to replace the localised competition lost through the merger be unlikely.

As to the first condition, economic theory stipulates that it is the perceived substitutability of the two firms’ products (ie, their “closeness”) that is the most important factor in determining the market power that will be generated by a merger in a differentiated product setting.43 Such closeness has a critical effect on the profitability of a post-merger price increase because “the more closely substitutable two products are (relative to substitutability with other products), the greater will be the degree to which substitution away from each of the products of the merging firms due to a price increase will be ‘internalized’ into the merged entity”,44 and such localised competition as existed between the firms pre-merger lost as a result.

Market share and concentration levels play a lesser role in this analysis than in other contexts
given that the requisite “closeness” of the merging firms’ products is not directly related to these indicia. As explained by two leading antitrust economists in the United States:

In sharp contrast to the situation with homogeneous goods industries, there is no reason why the shares in any delineated market in a differentiated products industry are indicative of the relative importance of each merging firm as a direct competitor of the other. Also, shares in a narrow market tend to be misleading in that they ignore what may be substantial competition from outside the market, and shares in a broad market tend to be misleading in that they ignore the fact that competition is localized.45

Despite this, in a manner that has attracted some criticism,46 the US Guidelines treat market shares and concentration as proxies for substitutability by specifying that:

40 See Fudenberg D and Tirole J, “Non-cooperative Game Theory for Industrial Organization: An Introduction and Overview” in Schmalensee R and Willig R (eds), Handbook of Industrial Organization (1989). For an explanation of the particular relevance of these models to unilateral effects theory, see Werden and Hay, n 9.
41 Starek and Stockum, n 32 at 814.
42 US Guidelines, s 2.21.
43 Starek and Stockum, n 32 at 816.
44 Starek and Stockum, n 32 at 816.
45Werden G and Rozanski G, “The Application of Section 7 to Differentiated Products Industries: The Market Delineation Dilemma” (1994) 8 (Summer) Antitrust 40 at 41.
46 See, eg, Hausman J and Leonard G, “Economic Analysis of Differentiated Products Mergers Using Real World Data” (1997) 5 (Spring) George Mason Law Review 321 at 337-339. Cf the defence of this aspect of the guidelines in Rill, n 18 at 396.

Where market concentration data fall outside of the safe harbors [specified in the Guidelines], the merging firms have a combined market share of at least thirty five per cent, and data on prouct attributes and relative product appeal show that a significant share of consumers of one merging firm’s product regard the other as their second choice, then market share data may be relied upon to demonstrate that there is a significant share of sales in the market accounted for by consumers who would be adversely affected by the merger.47

It should also be noted that, in connection with the first condition for finding anticompetitive unilateral effects in a differentiated product setting, the US Guidelines appear to emphasise only the relative closeness of a buyer’s first and second choices. But the relative closeness of the buyer’s other choices must also be considered in analysing the potential for price increases.48 Accordingly, as an additional requirement, it must be shown not only that the merging firms produce close substitutes but also that other options available to the buyer are so different that the merged entity will not be constrained from acting anticompetitively.49 Thus, it has been said (somewhat controversially) that the merged firm must be shown to be likely to enjoy a “post-merger monopoly or dominant position, at least in a ‘localized competition space’”.50

As to the second requirement of a unilateral effects claim identified in the US Guidelines, it is
concerned with the ability and willingness of rival firms to reposition their product offerings so as to constrain the unilateral market power that the merged firm otherwise would possess. This involves consideration not only of the cost involved in such repositioning for the rival firm(s) in question but also the ease with which such repositioning could be achieved and its likely effectiveness given the search and switching behaviour of the relevant buyers.51

Capacity differences

The second setting in which unilateral market power might arise from a merger is described in the US Guidelines as a setting in which products are “relatively undifferentiated”.52 In this context, unlike in the first setting, buyers consider the products of non-merging firms as well as those of the merging parties to be relatively close substitutes. Based on the Cournot model, the potential unilateral effect contemplated in this section of the guidelines is an effect that arises in markets where “capacity distinguishes firms and shapes the nature of their competition”.53 In such markets, when the merging parties have a combined share of at least 35%, the guidelines stipulate that the merged firm “may find it profitable to raise price and reduce joint output below the sum of their premerger outputs because the lost markups on the foregone [sic] sales may be outweighed by the resulting price increase on the
merged base of sales”.54

The key issue in this setting is seen as the ability of the non-merging firms to respond to a price increase by the merged entity with “increases in their own outputs sufficient in the aggregate to make the unilateral action of the merged firm unprofitable”.55 Clearly, if products are close substitutes, and competitors can quickly and easily expand their productive capacities, anticompetitive behaviour will be difficult to accomplish or maintain. As previously indicated, whether or not such a response is

47 US Guidelines, s 2.211.
48 The guidelines later acknowledge as much in s 2.212, which recognises that if a buyer’s other options include “an equally competitive seller not formerly considered, then the merger is not likely to lead to a unilateral elevation of prices”.
49 Areeda P, Hovenkamp P and Solow J, Antitrust Law (2002), Vol 4, Part 2, Ch 9, [914f].
50 The requirement that the merged firm enjoy a monopoly or dominant position was articulated in the recent unilateral effects case, United States of America v Oracle Corporation 331 F Supp 2d 1098 at 1118 (ND Cal 2004), discussed below. Whether the court was correct in this approach has been debated in the United States: see, for example, the commentary on the case by various authors in “Roundtable Discussion: Unilateral Effects Analysis After Oracle” (2005) 19(2) Antitrust 8.
51 Starek and Stockum, n 32 at 819.
52 US Guidelines, s 2.22.
53 US Guidelines, s 2.22.
54 US Guidelines, s 2.22.
55 US Guidelines, s 2.22.
possible or feasible will depend in part on the cost advantages enjoyed by the merged firm. The guidelines indicate that it is the ability to expand output within two years without increasing costs that is relevant.56

Notwithstanding the 35% presumption in the guidelines, many argue that market share in this
setting, as in the first setting, should not be seen necessarily as a reliable indicator of unilateral market power:

Anti-competitive prices are only profitable when buyers have limited opportunities to substitute. If buyers have access to suppliers that are able to supply them with a relatively homogeneous product, the market share of such firms is of limited significance to the effect of potential substitution on a firm’s market power. Thus, the importance of market share in this type of market is predominantly its potential reflection of constraints on firms’ productive capacities.57

UNILATERAL EFFECTS ANALYSIS AND MARKET DEFINITION IN THE UNITED STATES

In antitrust law, market definition is seen as the first step in an essentially structural analysis of the competitive effects of a particular transaction, including a merger. In particular, it is necessary so as to facilitate the calculation of market shares and levels of concentration which, in the case of mergers, are relied upon to classify the acquisition as presumptively legal or illegal. If the latter, a range of other factors are then considered; however, traditionally, it has still been delineation of the relevant market and calculation of shares and concentration in that market on which the fate of the merger has hinged.58

In the United States this structural approach to merger regulation was established by a series of early Supreme Court cases, starting with United States v Columbia Steel Co [1948] USSC 97; 334 US 495 (1948) which introduced the term “relevant market” in 1948 and was the first horizontal merger case to focus intensely on market shares.59 The court dismissed the challenge to the merger, and congressional dissatisfaction with that result was significant in precipitating amendment of the merger law in 1950.60
Thereafter, the basic principles followed by the lower courts were established in the first two Supreme Court decisions in horizontal merger cases that arose for substantive determination under the 1950 amendments.61 In Brown Shoe Co v United States 370 US 294 (1962) (Brown Shoe), the Supreme Court held that “the proper definition of the market is a ‘necessary predicate’ to an examination of the competition that may be affected by the horizontal aspects of the merger” (at 335). The court established criteria for market delineation (discussed further below), and market shares were emphasised in its analysis of anticompetitive effects (at 325, 343). In United States v Philadelphia National Bank [1963] USSC 166; 374 US 321 (1963) the court held that a merger which produces a firm controlling an undue percentage share of the relevant market, and results in a significant increase in the concentration of firms in the market, is so inherently likely to lessen competition substantially that it must be enjoined in the absence of evidence clearly showing that the merger is not likely to have such anticompetitive effects (at 363).

56 US Guidelines, s 2.22.
57 Starek and Stockum, n 32 at 820.
58 Note (no author cited), “Analyzing Differentiated-Product Mergers: The Relevance of Structural Analysis” (1998) 111 Harv LR 2420 at 2423 (referred to hereafter as “Harvard Law Review Note”).
59 Harvard Law Review Note, n 58 at 508, 512-513.
60 Columbia Steel [1948] USSC 97; 334 US 495 (1948) was decided under the original Clayton Act enacted in 1914. The original s 7 prohibited the acquisition by one corporation of the stock of another if the effect “may be to substantially lessen competition between such corporations”. The Celler-Kefauver Act 1950 amended the relevant test under the Clayton Act so as to prohibit mergers the effect of which may be to “substantially lessen competition – in any line of commerce – in any section of the country”. See generally Werden G, “The History of Antitrust Market Delineation” (1992) 76 Marquette Law Review 123 at 129-130.
61Werden G, “Simulating the Effects of Differentiated Products Mergers: A Practical Alternative to Structural Merger Policy” (1997) 5 (Spring) George Mason Law Review 363 at 364-365.
The Supreme Court has not revisited these decisions and their basic tenets have been followed faithfully and consistently in the lower courts.62 It thus has been the US Guidelines, as promulgated by the federal agencies, that have been the most influential articulation and extension of the approach stipulated in Brown Shoe and Philadelphia National Bank.63 Notably, when the guidelines were first published by the Department of Justice in 1982, commentators singled out their approach to market definition as “their most important contribution”,64 their “noteworthy intellectual feat”65 and “their most innovative aspect”.66 Since then, many courts have followed the structural approach taken in the guidelines, including the approach advocated specifically in relation to market definition.67

There is some irony in the fact that, at about the same time as the structural paradigm was being endorsed by the Supreme Court, probably the most influential economic criticism of the market definition-market share approach was being formulated.68 This criticism was most strident in the context of differentiated industries. In 1950 Edward Chamberlin, the pre-eminent economic authority in the United States on product differentiation, dismissed the whole idea of structural analysis, observing:

“Industry” or “commodity” boundaries are a snare and a delusion – in the highest degree arbitrarily drawn, and, wherever drawn, establishing at once wholly false implications both as to competition of substitutes within their limits, which supposedly stops at their borders, and as to the possibility of ruling on the presence or absence of oligopolistic forces by the simple device of counting the number of producers included.69

Chamberlin’s objections have been emphasised repeatedly, as well as elaborated upon, in the economic literature since then.70 However, it was not until the mid-1990s that they began to be reflected in official merger regulation policy in the United States. From about this time it became evident that the Federal Trade Commission had:

begun to eschew the “traditional” structural method of proving harm to competition in favour of a more fashionable and flexible approach. Gone are the days when a precisely delineated market definition and rigid structural market analysis were invariably the starting point in [antitrust] cases. Indeed, the position often now articulated by the Federal Trade Commission is that a rigorous structural analysis, including market definition, is essentially unnecessary where it believes there is “direct” evidence of harm to competition.71

Notably, these observations are not confined to agency policy in the merger context. There were even earlier signs of a shift to a so-called “direct effects” approach in relation to non-merger

62Werden, n 61 at 365.
63 The first version of these guidelines was published by the Department of Justice in 1968 and they were revised in 1982 and
1984. The next, 1992, revision was published jointly by the Department and the Federal Trade Commission.
64 Baker D and Blumenthal W, “The 1982 Guidelines and Preexisting Law” (1983) 71 California Law Review 311 at 322.
65 Ordover J and Willig R, “The 1982 Department of Justice Merger Guidelines: An Economic Assessment” (1983) 71
California Law Review 535 at 539.
66 Baxter W, “Responding to the Reaction: The Draftsman’s View” (1983) 71 California Law Review 618 at 622; Sullivan L,
“The New Merger Guidelines: An Afterword” (1983) 71 California Law Review 632 at 638.
67Werden G, “Market delineation under the Merger Guidelines: A Tenth Anniversary Retrospective” (1993) 38 (Fall) Antitrust Bulletin 517 at 518-519 (see cases cited in n 8).
68 See, eg, Mason E, “The Current Status of the Monopoly Problem in the United States” (1949) 62 Harv LR 1265 at 1274.
69 Chamberlin E, “Product Heterogeneity and Public Policy” (1950) 40 American Economic Review (Papers and Proceedings) 85 at 86-87.
70 See, eg, Werden and Rozanski, n 45 at 40, observing that “[o]ne may object to nuance and tone in Chamberlin’s comment but his basic observations about the limitations of delineated markets and market shares are unassailable and remain at the heart of merger litigation involving differentiated products”.
71 Keyte and Stoll, n 1 at 593. See also the observation in Stoll N and Goldfein S, “Markets! We Don’t Need Markets!” (2003) 229 New York Law Journal 3, that: “Increasingly, the Federal Trade Commission and Department of Justice are placing more emphasis on the identification of close substitutes than upon the precise delineation of product market boundaries. Simply stated, if competitive harm is predicted to result from the elimination of a close rivalry of the merging firms’ products, it becomes unnecessary to delineate precisely the bounds of the relevant product market.”
transactions. However, unlike in the merger arena, such an approach has received judicial, including Supreme Court, endorsement in both anticompetitive agreement and monopolisation cases.72 The first Supreme Court articulation of a concept of “direct effects” occurred in Federal Trade Commission v Indiana Federation of Dentists [1986] USSC 113; 476 US 447 (1986), in which the court held that the defendant could be found in violation of the Sherman Act 1890 even though the Commission had failed to undertake “elaborate market analysis”.73 The case involved a group boycott that restricted the supply of dental information to insurance companies. The court stated that proof of market power is a “surrogate for detrimental effects” (at 460). Therefore, “proof of actual detrimental effects, such as a reduction of output” (at 460) could eliminate the need to examine market power, and hence to engage in a market definition exercise in order to determine whether the defendant possesses such power. The
Commission had shown that the horizontal restraint at issue had reduced output well below
competitive levels, so that (as was held) a demonstration of market power on the part of the defendant Association was unnecessary. United States’ courts have since applied the same reasoning in numerous other non-merger cases.74

This approach is seen as an abbreviated version of the full rule of reason analysis.75 It allows for plaintiffs to meet the burden of proving anticompetitive effects “directly” through evidence of actual adverse effects on price, output or other competitive dynamics rather than “indirectly” through a structural analysis of the defendant’s market power and an inference that the impugned transaction will injure competition. Such an approach, it has been said, has become a standard mechanism by which the Federal Trade Commission and private plaintiffs “can avoid the market definition quagmire of a standard rule of reason analysis”.76 In the merger context, it is associated with unilateral effects and has generated a movement in merger cases to define more narrowly the market in which the parties engage in competition. Indeed, some have even observed that in such cases where there is evidence of actual detrimental effects, market definition has been treated as “a mere formality, if not
completely disregarded”.77

The Department of Justice is said to have initiated a move away from market definition and
structural analysis in the merger context in United States v The Gillette Company 828 F Supp 78 (DDC 1993) (Gillette). In that case the government challenged a proposed merger between two pen manufacturers, arguing that the relevant market was for “high quality refillable fountain pens that have an established premium image among consumers” (at [4]). Although a market as such was pleaded, the government’s expert economist offered his opinion about the unilateral effects of the merger without reference to market definition. Specifically, Dr Rozanski testified that his analysis of competitive effects involved considering only the products that are “close substitutes” to the fountain

72 For a review of recent cases, see Keyte and Stoll, n 1 at 612-624.
73 Keyte and Stoll, n 1 at 613.
74 See Keyte and Stoll, n 1 at 613-624. But note that its interpretation has not been free of controversy. In a recent 7th Circuit court decision, for example, the Indiana Federation of Dentists approach was read narrowly, the court concluding that evidence of actual anticompetitive effects can suffice to meet an antitrust plaintiff’s burden solely in horizontal, not vertical cases. In vertical cases, plaintiffs must prove at least the “rough contours” of a relevant market and substantial market shares. The court also appeared to denigrate the value of actual effects evidence, characterising it as a “relaxed” evidentiary standard compared to market share evidence and implying that reliance on direct evidence in the “wrong” context could lead to false positives and hence over-deterrence. See Republic Tobacco Co v North Atlantic Trading Co [2004] USCA7 432; 381 F 3d 717 (7th Cir 2004), commented on in Gavil A, “On the Utility of ‘Direct Evidence of Anticompetitive Effects’” (2005) 19(2) (Spring) Antitrust 59.
75 See Leary T, Federal Trade Commissioner, A Structured Outline for the Analysis of Horizontal Agreements (2003 Spring Meeting of the Antitrust Section of the American Bar Association in Washington DC, 3-4 March 2004), available at http://www.ftc.gov/speeches/leary/chairshowcasetalk.pdf (viewed 13 September 2005): “Where the market effects of a particular restraint are likely to be ambiguous the plaintiff’s prima facie case typically defines a relevant market and demonstrates that the restraint affects a significant share of that market. It may not be necessary to define markets and calculate market shares, however, if there is more direct evidence of likely or actual effects on prices or output ... This is still a full rule of reason analysis. Such analysis involves proof of competitive market effects, either based on direct evidence of such effects or indirect inferences from market shares.”
76 Keyte and Stoll, n 1 at 612.
77 Keyte and Stoll, n 1 at 626.
pens sold by the merging parties.78 Ultimately, the court held that market definition was required, that the appropriate product market was broader than that alleged by the government and that it should be defined as the market for all premium writing instruments (at [8]). Nevertheless, the case is seen as significant for indicating a government view that “marketwide effects are not relevant when a unilateral effects analysis shows that the merger will eliminate some direct competition between the combining firms”.79

The subsequent case of Federal Trade Commission v Staples Inc 970 F Supp 1066 (DDC 1997) (Staples) is widely regarded as one of the first occasions on which the Commission was successful in a contested action in de-emphasising detailed market analysis and emphasising direct unilateral effects evidence.80 The proceeding concerned a proposed merger between Staples, an office supplies superstore chain and its rival, Office Depot. From an antitrust perspective, the focus was very much on the definition of the market in which these two firms competed. The Commission contended that the merged firm would operate in an extremely narrow and concentrated, “office-supply superstores” market (at [5]). Staples characterised this definition of the market as “contrived” and countered that the appropriate product market in which to assess the competitive consequences of the merger was the
market for the “sale of office supplies” (at [5]), of which a combined Staples-Office Depot accounted for only 5.5% of total sales in 1996.

In support of its position, the Commission presented data showing that the prices for office
supplies were higher in locations with fewer superstores. Thus, according to the pricing history of the two firms, it was the number of superstore firms that had the most significant effect on prices in the market. Prices were lowest in three-chain markets, higher in two-chain markets, and highest in markets with a superstore monopoly. The difference in prices between one-chain cities and three-chain cities was approximately 13% (at [5]), a significant difference in retailing where profits and profit margins are usually only a small percentage of sales volume.81 Describing this data as “compelling” (at [7]), the court held that the Commission had discharged its burden in proving the alleged superstore market and further, based on much the same evidence, that the proposed merger would lead to increased prices and hence would substantially lessen competition in that market (at [10]-[11]). The
availability and strength of the pricing evidence has been seen as making Staples unique.82
The court plainly had reservations about accepting the Commission’s market definition and went to some lengths to stress that its findings should not be taken as a precedent for outcomes in future cases.83 Nevertheless, in the end, the court’s reservations effectively were
trumped by the direct effects

78 Stoll and Goldfein, n 71.
79 Stoll and Goldfein, n 71.
80 Keyte and Stoll, n 1 at 627.
81 Baer W and Balto D, “New Myths and Old Realities: Recent Developments in Antitrust Enforcement” (1999) 2 Columbia Business Law Review 207 at 217.
82 Keyte and Stoll, n 1 at 629. Its significance was illustrated by a case that followed shortly thereafter, United States v Long Island Jewish Medical Centre 983 F Supp 121 (FDNY, 1997), in which the government was unable to persuade the court to adopt its unilateral effects argument, mounted in opposition to a hospital merger. The court was prepared to analyse the potential impact of the merger of two hospitals based on an assessment of “direct evidence”, said to show that the merged entity would be able to raise prices unilaterally. However, it concluded that the evidence that was presented to this end was “totally
speculative” (at 143), evidently being not of the same calibre as the type of substantive economic data provided by the government in Staples.
83 The court observed, for example, that: “it is difficult to overcome the first blush or initial gut reaction of many people to the definition of the relevant product market as the sale of consumable office supplies through office supply superstores. The products in question are undeniably the same no matter who sells them, and no one denies that many different types of retailers sell these products.” See Federal Trade Commission v Staples Inc 970 F Supp 1066 (DDC 1997) at [7]. Having reviewed the evidence and accepted the Commission’s proposed market, the court concluded by saying: “In addition, the Court is concerned with the broader ramifications of this case. The superstore or ‘category killer’ like office supply superstores are a fairly recent phenomenon and certainly not restricted to office supplies ... It remains to be seen if this case is sui generis ... For these reasons, the Court must emphasize that the ruling in this case is based strictly on the facts of this particular case, and should not be construed as this Court’s recognition of general superstore relevant product markets.” See Federal Trade Commission v Staples Inc 970 F Supp 1066 (DDC 1997) at [16]-[18].
evidence.84 As one commentator has noted, this evidence essentially “saved” what many regarded as an “unusually narrow” market definition propounded by the Commission.85 At the same time, it should be noted that the court did not refer to the theory of “unilateral effects” as such and that, while Hogan J plainly relied on the logic and evidence presented in support of direct effects, he also went to some lengths to apply other qualitative or “practical indicia” traditionally associated with the concept of a sub-market, discussed below.86 It has been suggested that Hogan J adopted this traditional framework “in a conscious effort to downplay novelty in order to avoid creating an issue for appeal”.87

More recently, the government attempted to challenge a merger on the basis of a similarly narrow market definition in United States v Oracle Inc 331 F Supp 2d 1098 (ND Cal 2004) (Oracle). In that case, it sought to enjoin the acquisition by the software firm, Oracle, of a large competitor, PeopleSoft Inc. The government alleged that these two firms operated in a narrow market for enterprise resource planning system software used in human resource and financial management applications in large complex enterprises (at 1101 - 1107). The acquisition, it was argued, would lead to the loss of the localised competition said to exist between the two firms in this market (at 1116).

For reasons that appeared to relate primarily to deficiencies in its evidence, the court rejected the market definition submitted by the government, concluding that there was a wider range of software solutions, customers and vendors that would be relevant to assessing the competitive effects of the acquisition than was recognised by the government’s case.88 However, unlike in Staples, the court discussed at length and appeared to accept in express terms the economic theory of unilateral effects. Walker J was even encouraging of the presentation of econometric evidence in the proof of such effects.89 However, he also urged caution in relying on exceptionally narrow markets, particularly in differentiated products industries (at 1118-1123). The potential contradiction in this approach, together with several other features of the case, has made it a talking point in antitrust circles in the United States.90

Interpretation of the approach taken in and the ramifications of each of these cases, as well as others in which unilateral effects analysis has been argued,91 tends to differ depending on the perspective from which the interpretation is being offered. As previously indicated, there are those who point to these actions as proof of a slow but steady abandonment of the structural approach to merger enforcement by federal agencies. Agency officials, however, are not prepared to acknowledge as much. Indeed, some have gone so far as to argue that the framework set out in the US Guidelines,

84 Keyte and Stoll, n 1 at 629.
85 Grimes W, “Antitrust and Systematic Bias Against Small Business: Kodak, Strategic Conduct and Leverage Theory” (2001) 52 Case Western Reserve Law Review 231 at 274.
86 The court specifically equated the government unilateral effects analysis with one of these indicia, “sensitivity to price changes”: see Federal Trade Commission v Staples Inc 970 F Supp 1066 at 1075 (DDC 1997).
87 Baker J, Econometric Analysis in FTC v Staples (Prepared Remarks of Jonathan Baker, Director, Bureau of Economics, Federal Trade Commission Before the American Bar Associations Antitrust Section Economics Committee, 18 July 1997 (revised 31 March 1998) available at http://www.ftc.gov/speeches/other/stspch.htm (viewed 16 August 2005).
88 For a review of the evidence in the case and the implications of the approach taken by the court to the evidence, see Beaton-Wells C, “Customer Testimony and Other Evidence in Australian Antitrust Assessments: Searching for the Oracle” (2005) 33 ABLR 448.
89 The judge observed, for example, that “modern econometric methods hold promise in analyzing differentiated products unilateral effects cases”. In particular, Walker J referred to merger simulation models that “may allow more precise estimations of likely competitive effects and eliminate the need to, or lessen the impact of, the arbitrariness inherent in defining the relevant market”. See United States v Oracle Inc 331 F Supp 2d 1098 at 1122 (ND Cal 2004). These models are discussed below.
90 See, for example, the commentary by various authors in “Interpreting Oracle: The Future of Unilateral Effects Analysis” (2005) 19(2) (Spring) Antitrust 8; The Oracle/PeopleSoft Decision: The Implications for Merger Analysis in High-Tech Industries (Paper presented as part of the Antitrust TeleSeminar Series organised by the Antitrust Law Section of the American Bar Association, 4 November 2004 (slides on file with author)).
91 See, for example, New York v Kraft General Foods Inc 926 F Supp 321 (SDNY 1995); Federal Trade Commission v Swedish Match 131 F Supp 2d 151 (DDC 2000).
including the preliminary requirement of market definition, continues to be applied.92 This is so
despite the fact that uncontested agency action taken against a growing number of proposed mergers has been based on very narrow product groupings (“refrigerated pickles”93 and “super-premium ice-cream”,94 among them) that traditionally would have been difficult to characterise as antitrust markets. In each of these cases, the Commission’s position has been that, by eliminating the head-to-head competition between the two firms in question, the proposed merger would lead to higher prices for the relevant products. This was sufficient to satisfy the agency that the merger would lessen competition substantially, thus making an “elaborate market analysis” unnecessary.95

As the patchy record of the government’s litigated actions attests, however, it is by no means clear that the United States courts are ready to embrace merger analysis without market definition, or even merger analysis based on the narrow markets involved in the application of unilateral effects theory.96

This has prompted observations that: While the movement away from the old “mechanical” approach using rigorously defined markets and market share analysis is “accelerating” it is “unclear how far the agencies will take it or whether a court would ever go all the way and forego the use of market definition, market share and concentration”.97

And, further:

it can be expected that courts will be more than a little reluctant to abandon Supreme Court precedent, even when it is more than a third of a century old. More recent Supreme Court precedent may have to pave the way.98

UNILATERAL EFFECTS ANALYSIS AND MARKET DEFINITION IN AUSTRALIA

Whether Australian regulators or courts are likely to embrace an approach to merger analysis based on unilateral effects that excludes the step of market definition is a question that raises the following issues:
• the status of market definition as a statutory requirement in Australia;
• the nature and purpose of that requirement in competition law analysis; and
• the extent to which the so-called “direct effects” of a merger are susceptible to “proof” in either the agency or adversarial context.

Market definition as a statutory requirement

As stated at the outset, in Australian competition law, market definition has been seen as “the essential first step” in an assessment of the likely competitive effects of a merger (as well as other transactions subject to a competition test). This was the terminology employed in the landmark statement by the Tribunal in Re Queensland Co-Operative Milling Association Ltd (1976) 25 FLR 169 at 189; [1976] ATPR 40-012 (QCMA), explaining its approach to the concept of a market for the purposes of the

92 Baer and Balto, n 81 at 218. Cf Leary, n 75.
93 See Federal Trade Commission Press Release, “Federal Trade Commission Votes to Challenge Hicks Muse’s Proposed
Acquisition of Claussen Pickle Company”, 22 October 2002. The merger was subsequently abandoned (see Keyte and Stoll, n 1
at 629).
94 See Stoll and Goldfein, n 71.
95 Keyte and Stoll, n 1 at 629-630.
96 For a summary of recent cases in which courts have rejected so-called “localised effects markets”, see Rill, n 18 at 406-409.
97 Keyte and Stoll, n 1 at 630.
98Werden, n 61 at 385; see also Blumenthal W, Why Bother? On Market Definition under the Merger Guidelines (Statement
before the FTC/DOJ Merger Enforcement Workshop, 17 February 2004 (copy on file with author)), noting that: “[t]he case law seems to require market definition, and it seems to require a stepwise approach in which market definition precedes analysis of competitive effect. One can only imagine the blizzard of quotations from Supreme Court cases that would confront any agency effort to do otherwise.” Cf the view expressed in Burton T, “Unilateral Effects Analysis in Assessing Anti-Competitive Mergers: The Judicially Approved New Approach to Challenging Mergers” (1999) 43 St Louis University Law Journal 1481.
TPA. Five months later the Swanson Committee approved the Tribunal’s approach and recommended that a definition of the term be inserted in the Act.99 The definition subsequently inserted is found in s 4E which provides that, for the purposes of the Act, “market” means:

[A] market in Australia and, when used in relation to any goods or services, includes a market for those goods or services and other goods or services that are substitutable for or otherwise competitive with, the first mentioned goods or services.

The QCMA approach to market definition has been followed consistently by Australian courts.100 It has been emphasised nevertheless that the identification of the relevant market should not be regarded as an end in itself, but rather as a means to an end, that end being resolution of the substantive issue, that is the likely effects on competition of the conduct at hand. This is generally what is meant by reference to “market” as an “instrumental”101 concept and to market definition as an exercise that should be approached purposively.102

Taken to an extreme, the purposive approach might be seen to warrant abandoning market
definition as an anterior step altogether, and subsuming the matters with which it is concerned
(demand and supply substitutes) into the substantive competition analysis. This is an argument favoured by some Australian commentators who are critical generally of the traditional “structure- conduct-performance” paradigm that has underpinned much of the competition law analysis in Australia to date.103 It is not an argument that appears as yet to have made inroads into the traditional two-step approach applied by the ACCC, the Tribunal and the courts. It does raise the question, however, as to whether the establishment of a “market” is in fact a requirement made mandatory by the provisions of the TPA. As pointed out in a recent paper by Brewster and O’Bryan, that question can only be answered in the affirmative.104 According to Brewster and O’Bryan’s analysis, in Pt IV of
the TPA:

the expressions “market” and “competition” are each used in a specific way and as discrete concepts. Competition is a process that occurs within a market. Furthermore, it is frequently necessary to identify the market in which a corporation competes in order to establish liability under the TPA. For example, s 45(2) generally prohibits contracts that substantially lessen competition. However, s 45(3) specifies that an agreement will only be unlawful in so far as it lessens competition in a market in which a party to the contract (or a related body corporate) competes. It is therefore a necessary element of a contravention to identify such a market ... Section 47(13)(c) also limits examination of competitive effect to markets in which the supplier or acquirer compete. The separate identification of a market is also a necessary step in s 50 analysis. Under sub-section (1), an acquisition will be prohibited if it has the likely effect of substantially lessening competition in any market. However, under sub-section (3), the court must have regard to various matters that relate to the specific market in which the acquisition

is alleged to have an anti-competitive effect including:

(a) the actual and potential of import competition in the market;

(b) the height of barriers to entry to the market;

99 Trade Practices Act Review Committee, Report to the Minister for Business and Consumer Affairs (August 1976),
[4.19]-[4.21].
100 See, in particular, by the High Court in Queensland Wire Industries Pty Ltd v Broken Hill Proprietary Company Ltd [1989] HCA 6; (1989) 167 CLR 177, and in every High Court and Federal Court decision involving the issue of market since then.
101 Re John Dee (Export) Pty Ltd [1989] ATPR 40-938 at 50,219. See similarly the statements made, for example, in
Queensland Wire Industries Pty Ltd v Broken Hill Proprietary Company Ltd [1989] HCA 6; (1989) 167 CLR 177 at 187, 200; Arnotts Ltd v Trade Practices Commission [1990] FCA 473; (1990) 24 FCR 313 at 328.
102 For explanations of this approach, see, eg, Norman N and Williams P, “The Analysis of Market and Competition Under The Trade Practices Act: Towards the Resolution of Some Hitherto Unresolved Issues” (1983) 11 ABLR 396 at 406; Brunt M, “‘Market Definition’ Issues in Australian and New Zealand Trade Practices Litigation” (1990) 18 ABLR 86 at 104, 126-127.
103 See, eg, Robertson, n 9; Smith R and Round D, “A Strategic Behaviour Approach to Evaluating Competitive Conduct” (1998) 5(1) Agenda 25; Round D and Smith R, “The Strategic Approach to Merger Enforcement by the ACCC: A Comment” (1998) 26 ABLR 227; Smith R and Round D, “Competition Assessment and Strategic Behaviour” (1998) 19 European Competition Law Review 225 at 232-233; Round D and Smith R, “Strategic Behaviour and Taking Advantage of Market Power: How to Decide if the Competitive Process is Really Damaged?” (2001) 19(4) New Zealand Universities Law Review 427.
104 Brewster D and O’Bryan N, Market Definition – Drawing Imaginary lines? (Paper delivered at Trade Practices Conference hosted by Law Council of Australia, 2004 (copy on file with author)).

(c) the level of concentration in the market;

(d) the degree of countervailing power in the market; ...

(f) the extent to which substitutes are available in the market or are likely to be available in the

market;

(g) the dynamic characteristics of the market, including growth, innovation and product differentiation;

(h) the likelihood that the acquisition would result in the removal from the market of a vigorous and effective competitor;

(i) the nature and extent of vertical integration in the market.

The Court cannot undertake the task required by sub-section (3) unless the relevant market is first
identified.105 To this analysis, reference to s 50(6) should be added. That provision states that in s 50 “market” refers to a “substantial market for goods or services” in Australia, a State or Territory, or region of Australia. As pointed out in the Australian Guidelines, “only when the relevant market has been delineated, can its substantiality be determined”.106 Hence, s 50(6) provides additional support for the view that market definition is a mandatory step in the application of the substantive prohibition in s 50.

If market delineation is a statutory requirement, the next question is whether this necessarily
precludes an approach to merger enforcement based on unilateral effects analysis along the lines developing in the United States. Section 7 of the Clayton Act requires the appraisal of a merger in some “line of commerce” and some “section of the country”. This language has been equated with a requirement of market definition.107 As previously indicated, in Brown Shoe 370 US 294 (1962) (just as in QCMA (1976) 25 FLR 169; [1976] ATPR 40-012), definition of the relevant market(s) in terms of both product and geographic area was described as a “necessary predicate” to making a determination about anticompetitive effects.108 Since then, in broad terms and applying the language of s 7, the judicial approach generally has been to determine: (1) the “line of commerce” or product market in which to assess the transaction; (2) the “section of the country” or geographic market in
which to assess the transaction; and (3) the transaction’s probable effect on competition in those product and geographic markets.109 The US Guidelines reflect this approach.110

Unilateral effects theorists do not regard the terms of s 7 and the significance that has been
attached traditionally to market definition by the courts as impeding a direct assessment of the
competitive effects of a merger (as distinct from an indirect structural assessment). Instead they argue that even in employing the direct effects approach, market delineation remains an element of the conceptual framework. The only difference is that, rather than being the first step in or precondition to the analysis, it becomes a consequence or product of the analysis.111 This shift in role for market definition is reflected in comments made by Areeda, Hovenkamp and Solow, explaining the appropriate conclusion to be drawn from a finding of unilateral market power: In cases in which a merger facilitates a significant “unilateral” price increase for a grouping of sales that was not a distinctive-looking market prior to the merger, the appropriate conclusion is that the merger

105 Brewster and O’Bryan, n 104, pp 5-7.
106 Australian Guidelines, [5.3].
107 Areeda et al, n 49 at [913a].
108 Brown Shoe Co v United States 370 US 294 at 335 (1962).
109 United States of America v Oracle Corporation 331 F Supp 2d 1098 at 1110-1111 (ND Cal 2004). In Oracle Walker J
expressed concern that defining narrow markets, in reliance on a concept of localised competition, could result in markets
“defined so narrowly that one begins to question whether the market constitutes a ‘line of commerce’ as required by section 7”
(at 1120).
110 See US Guidelines, s 1.
111 Two alternative approaches to this approach have been suggested by Blumenthal: (1) “to retain market definition as an initial step in the analysis, but ... import much of the competitive effects analysis into that step. Market definition and competitive effects would be simultaneously determined”; (2) an iterative approach – “beginning with a quick ‘virtual’ market definition that identifies candidate problem markets, proceeds to competitive effects analysis, and returns to ‘confirmation of market definition’ as a late step.” See Blumenthal, n 98.
has facilitated the emergence of a new grouping of sales capable of being classified as a relevant
market. This formulation meets the statutory requirement that the “effect” of a merger is anticompetitive
in some “line of commerce” and in some “section of the country.” That is, § 7’s “effect” usage invites
consideration of the market’s structure after the merger rather than before, and if a new grouping of
sales can be said to constitute a relevant market after the merger, that is an appropriate grouping for
measuring the merger’s competitive impact.112
This explanation has certain logical appeal. It also does not appear necessarily to fall foul of the statutory regime in Australia. There is nothing in the TPA stipulating expressly that definition of a relevant market be undertaken prior to, rather than as a consequence of, an analysis of competitive effects.113 There may be a concern, however, that what is identified as a result of the effects analysis is in fact a sub-market, rather than a market.114 This is of concern given that it is identification of the latter and not the former that is mandated by the relevant Pt IV prohibitions.
The concept of a submarket was referred to in QCMA (1976) 25 FLR 169; [1976] ATPR 40-012 and other early Tribunal decisions as involving what is essentially a segment or sub-set of a market in which the competition between substitutes is especially intense.115 Notwithstanding its lack of statutory profile, the Tribunal suggested in these decisions that the concept may be helpful as an analytical tool, “in clarifying how competition [in the market] works”.116 By contrast, in the United States, the early treatment of submarkets gave them a much greater significance. In Brown Shoe 370 US 294 at 325 (1962) the Supreme Court explained that:

[t]he outer boundaries of a product market are determined by the reasonable interchangeability of use or the cross-elasticity of demand between the product itself and substitutes for it.

The court went on to state (at 325) that:

within this broad market, well-defined submarkets may exist which, in themselves, constitute product markets for antitrust purposes. The boundaries of such a submarket may be determined by examining such practical indicia as industry or public recognition of the submarket as a separate economic entity, the product’s peculiar characteristics and uses, unique production facilities, distinct consumers, distinct prices, sensitivity to price changes, and specialized vendors.

The Brown Shoe distinction between markets and submarkets has attracted substantial criticism in the United States117 and is not adopted in the US Guidelines. It is not surprising then that supporters of direct unilateral effects analysis have been quick to point out that the “new grouping of sales”118 identified as a result of such analysis is not to be equated with a submarket. Areeda et al have made it clear, for example, that in the context of unilateral effects, there should not be:

a return to the language of “submarkets” – for example, one might say that while the relevant market in the hypothetical market (sic) of firms A, B, C ... J consists of the sales of these ten firms, a relevant submarket exists for the sales of merging firms B and C. Historically the term “submarket” has been used to identify artificially narrow groupings of sales on the basis of non-economic criteria having little to do with the ability to raise price above cost. After the merger occurs, postmerger firm BC can raise its

112 Areeda et al, n 49. See a similar argument made in Baker J, “Product Differentiation Through Space and Time: Some Antitrust Policy Issues” (1997) 42 (Spring) Antitrust Bulletin 177, in which the then Director, Bureau of Economics, Federal Trade Commission, advocated a notion of “res ipsa loquitur market definition”.
113 Note in this regard the observation by Robertson that: “Even if the Australian competition legislation does require that there be a substantial lessening of competition ‘in a market’, the real question for courts and regulators is what role market definitions play; do they come first with the competition analysis following after; or do they also come last, with the competition analysis forming the basis for the conclusion as to the scope of the relevant market in judicial or regulatory decision.” See Robertson, n 9 at 218.
114 This concern has certainly been expressed in the United States where the sub-market concept has fallen into disfavour: see, eg, Baker, n 112.
115 Re Queensland Co-Operative Milling Association Ltd and Defiance Holdings Ltd (1976) 25 FLR 169 at 190-191; [1976] ATPR 40-012; Re Tooth & Co Ltd; Re Tooheys Ltd (1979) 39 FLR 1 at 39; [1979] ATPR 40-113.
116 Re Tooth & Co Ltd; Re Tooheys Ltd (1979) 39 FLR 1 at 39; [1979] ATPR 40-113.
117 A detailed review of the criticisms that have been made in case law and commentary is set out in Werden, n 60.
118 Areeda et al, n 49 at [913a].
price significantly above cost, thus meeting the usual criteria for market definition and not requiring the
confusing “submarket” label. A somewhat better but by no means perfect usage is to say that the merger facilitates the appearance of a new, narrower grouping of sales, or “market,” in which firm BC occupies a sufficiently strong position so as to enable its price increase. In fact, the merger does not create such a market because a cartel of firms B and C would also have been able to increase price profitably, indicating that B and C were already a relevant market. However, before their union B and C felt one another’s competition, as well as that of other firms, more significantly than after the merger.119
Other United States commentators have expressed the view that the sensitivity about submarkets is exaggerated.120 The basic point being made in Brown Shoe, they argue, is that there are varying degrees of substitution among products. That concept – varying degrees of substitutability – is at the core of the unilateral effects doctrine which holds that a merged firm may be able to raise price (or reduce output) unilaterally if the products of the merging companies are the first and second choices for customers representing a substantial part of the market, and other firms are unlikely to be able to reposition their products to become closer substitutes for those of the merged firm.121 This is not the Brown Shoe submarket analysis, but rather recognition that within a relevant market, it is possible for a merger to eliminate localised competition between particularly close substitutes.122 In addition, it has
been argued that the “practical indicia” identified by the court in Brown Shoe remain useful in defining markets given that most of the indicia of submarkets identified in that case are related to substitutability in supply or demand. As explained by two former Federal Trade Commission personnel:

The agencies use Brown Shoe criteria and measure unilateral effects to determine what cases to bring, because merger analysis requires the agencies to examine the relationship between products made by merging firms and to determine whether the cross elasticities between them are so much more significant than the relationships between products made by other firms that the merger will confer market power on the new firm. In a somewhat cruder sense, that is what Brown Shoe submarkets are all about. The Brown Shoe-Court clearly was looking for a way to suggest that some products within a broad market may be closer substitutes for one another than other products in the market. The use of unilateral effects analysis is a more focused and disciplined effort to measure those relationships. When there are unique relationships among products made by the merging firms, as evidenced by how the firms behave in the marketplace and by quantitative analysis of past pricing behavior, the merger poses competitive problems. In these situations, the issue of the precise boundaries of the market becomes, and should become, secondary.123

The distinction between markets and submarkets aside, there are possibly more fundamental
difficulties associated with an approach to merger analysis that renders market definition something of an afterthought once the so-called “direct” effects of the merger have been identified. Four such issues, concerned with the nature and purpose of market definition, are discussed below. At least the first two have been raised and are as applicable in the United States context as they are in the Australian context.

The nature and purpose of market definition

First, the direct effects approach (and the empirical models employed in its service, discussed below) is very much focused on the loss of competition between the two merging firms, that is, on the loss of constraint that was imposed pre-merger by the acquired firm on the acquirer and vice versa. It is, in this sense, a very narrow inquiry. However, a comprehensive analysis of potential anticompetitive effects cannot be restricted in this way. All possible sources of constraint on the pricing discretion of

119 Areeda et al, n 49 at [913a].
120 Baer and Balto, n 81, n 21.
121 As reflected in the US Guidelines, s 2.21.
122 Baer and Balto, n 81, n 21.
123 Baer and Balto, n 81 at 218-219.
the merged firm must be taken into account, including constraints that may be imposed by the
behaviour of other rival firms and by buyers, as well as ultimately by the prospects of new entry.124

Second, and related to the first point, using empirical models to predict post-merger price
increases means that the direct effects inquiry is a relatively static one. Of its nature, it may be seen as an inquiry into the effects of a merger at a particular point in time and thus it fails to recognise competition as a “process rather than a situation”.125 For this additional reason, direct effects analysis (based on quantitative predictions of price effects) should not be held up as providing the complete answer to the question of whether the merger will facilitate an exercise of unilateral market power on the part of the merged firm. This answer can only be found in an analysis that takes full account of market-place dynamics and, in particular, the likely responses of existing or potential competitors to an attempt by the merger firm to raise prices or reduce output. Such responses may not be immediate but may nevertheless occur within sufficient time to allay concerns about the competitive effects of the merger in the long run.

Reservations about the narrow and static nature of a direct effects approach should be seen as catered for by the US Guidelines which, in assessing the extent to which the merger might facilitate or augment the power of the merged firm, require examination of repositioning by existing rivals as well as the prospects of new entry. However, there are those who argue that these elements of the analysis risk being overlooked or undermined by the direct effects approach. One observer of developments in the United States has argued, for example, that:

Although agency economists continue to acknowledge the Guidelines principles of market definition, competitive repositioning, and entry, these concepts seem to have been submerged in the econometric analysis shuffle. Instead, the economists appear to be employing an approach which undertakes to identify effects first, then define a relevant market and assess competitive responses later. This relegation of competitive dynamics analysis to the latter stages of merger review, after a price increase has already been predicted, represents a significant shift in the nature of merger enforcement at the federal agencies.

The Merger Guidelines reflect the recognition that federal antitrust enforcement agencies should take full measure of the “competitive story” surrounding a proposed transaction before concluding that it is likely to result in harm. The approach identified by the models’ advocates [referring to econometric models that predict price effects, discussed below], however, would allow the agencies to forego [sic] this analysis up-front on the basis that the models offer a more precise measure of post-merger pricing incentives ... [A]doption of this approach is tantamount to the creation of a presumption that all differentiated products mergers will result in competitive harm, absent some countervailing market force. Such a shift in the analytical paradigm, though theoretically subtle, is practically dangerous, particularly when the reliability of the models is hampered by limiting assumptions and/or data availability.126

In Australia, the very process of market definition is seen as playing a significant role in ensuring that all of the relevant sources of effective constraint operating on the firm in question are identified. That is consistent with its role in the discipline of economics. As one member of the ACCC has

124 The narrowness of the inquiry is, to a significant extent, a function of the empirical techniques, discussed below, that are invoked in aid of a “direct effects” analysis. See the comments to this effect made in King, n 37, p 5. One of the leading United States proponents of merger simulation, Werden, for example, has conceded that this empirical technique “does not, as it is generally practiced, allow the investigation of prospects for entry or product repositioning; they are assumed away. And it’s perfectly possible that firms in the real world do not play the same strategies that they do in the model, so it’s not necessarily the case that merger simulation tells you what’s actually going to happen after a merger.” See Werden’s views expressed in Whither Merger Simulation, available at http://www.antitrustsource.com, May 2004 (viewed 20 September 2005).
125 Re Queensland Co-Operative Milling Association Ltd and Defiance Holdings Ltd (1976) 25 FLR 169 at 190; [1976] ATPR 40-012. Again, a concession to this effect has been made by Werden (referred to in the preceding note) in relation to the use of merger simulations, pointing out that this technique “cannot predict the long-run evolution of an industry. It cannot say much about entry or product repositioning; it cannot say much about changes in marketing strategy. It indicates only relatively short-term effects: how prices will be adjusted by the merging firms after the merger, and how the non-merging firms will respond to those price changes.” See Werden’s views expressed in Whither Merger Simulation, available at http:// www.antitrustsource.com, May 2004 (viewed 20 September 2005).
126 Rill, n 18 at 401-402.
described it recently:

From the perspective of merger analysis, an economic approach means that defining the market involves the process of laying out the field of inquiry – what participants, both from the demand-side and the upply-side are likely to influence whether or not the acquisition in question will lead to a substantial essening of competition? Market definition is not conclusive. It is not an attempt to determine any potential lessening of competition directly. Rather, market definition lays out the boundaries ofanalysis.127

Market definition also ensures that market-place dynamics are taken into account by requiring that otential constraints, on both the demand and supply sides, be factored into the analysis.128 Thus, as Maureen Brunt has put it, the market concept serves “to define what is relevant and why. It supplies a check-list of considerations that bear on market constraints of business behaviour.”129 Approached in this way, market definition sets the scene for an assessment of whether and the extent to which the conduct in question (in this context, the merger) is likely to remove or weaken the constraints that represent competition at work. As such, the process of defining the market does far more than provide a basis for calculating market shares and concentration levels.130 Rather, it provides “the basis for a complete competition analysis”,131 and in this sense, is truly purposive. There is conceptual order and
clarity in this approach, the value of which ought not be underestimated. Some or all of this clarity may be lost in the re-ordering of the analysis advocated by United States theorists such as Areeda et al.

The third issue raised by the possibility of defining the relevant market at only the post-merger
stage of the analysis is that it would be inconsistent with the well-established requirement in
Australian merger law that the merger’s effects be assessed by comparing competition in the market with the merger and without the merger.132 This approach allows for the appraisal of competitive effects to be evaluative, that is for the appraisal to determine whether an anticipated lessening of competition is likely in fact to be substantial, as is required by the terms of s 50. It is difficult to understand how the comparison critical to such an evaluation might be undertaken if there is, in effect, no assessment made of the state of competition in the market without the merger.

Finally, there is an argument that market definition is superfluous given that it is based on the same test (the SSNIP test) as the test employed to ascertain competitive effects. It is true that if, in defining the market, it is evident that the merging firms would find it profitable to impose a SSNIP, this points to such firms constituting a market unto themselves as well as to the merger being anticompetitive. However, the alternative scenario – in which the merging firms would not be able to impose a SSNIP profitably – must also be considered. In such a scenario the market would be defined
more broadly to include the competing firms to which buyers would turn in response to the SSNIP (thereby rendering it unprofitable). The SSNIP test thus plays an important part in ensuring that market boundaries are suitably drawn for the purposes of the competition analysis to follow – an analysis in which the profitability of a price increase by the merged entity is but one of a number of considerations relevant to assessing potential merger outcomes.133

127 King, n 37, p 3.
128 See Queensland Wire Industries Pty Ltd v Broken Hill Proprietary Company Ltd [1989] HCA 6; (1989) 167 CLR 177 at 196, 199, 200, 210.
129 Brunt, n 102 at 112.
130 This is not to suggest that such information is of no value whatsoever. At the very least it plays an important practical role in assisting regulatory authorities to distinguish between potentially problematic mergers and those that are unlikely to raise competition concerns (where the merger in question is between two firms with low market shares, for example). It also provides valuable guidance for the business community in this respect.
131 King, n 37, p 7.
132 This test has its origin in the reasons of the Full Court of the Federal Court in Outboard Marine Australia Pty Ltd v Hecar Investments (No 6) Pty Ltd [1982] FCA 265; (1982) 66 FLR 120. See, more recently, Stirling Harbour Services Pty Ltd v Bunbury Port Authority [2000] FCA 1381; [2000] ATPR 41-783 at 41,267 [12].
133 See King, n 37, pp 21-28.
To make this point, King has argued that a distinction ought to be drawn between an “informal” and a “formal” SSNIP test.134 The former, he argues, is the version of the test employed in Australia for the purposes of market definition. Rather than testing literally for the prospects of a post-acquisition price increase, it acts more as a “thought experiment”135 to prompt the analyst to consider all possible sources of potential constraint on the merged entity. King points out also that the test in this form is not the only market definition tool available.136 Others include a range of factors that are listed in the Australian Guidelines such as the end-uses of possible substitute products, the past behaviour of buyers, pricing correlations, switching costs, and so on.137

The so-called “formal” SSNIP test, by contrast, is directed specifically at the issue of competitive effects. Employing empirical techniques (discussed below), it asks and attempts to answer directly the question whether the merger will lead to a price increase imposed unilaterally by the merged firm. However, in this context also, it should be seen as only one of several relevant tests given that it highlights only one kind of possible anticompetitive outcome. A merger may still result in a substantial lessening of competition (for example, through reduction in quality, service or range) even if the merging parties fail to pass the formal SSNIP test.138

Proof of direct effects

To a significant degree, the shift in the United States away from the structural paradigm in analysing merger cases has been prompted by the development of new empirical techniques aimed at predicting unilateral price effects.139 These techniques, combining quantitative evidence with economic models of competition and firm behaviour, are used to determine whether, and to what extent, a merger will allow a firm to increase prices unilaterally after the merger. Such empirically derived predictions of post-merger prices are referred to as “direct” in the sense that they do not require definition of a market and assignment market shares.140 They have become an increasingly popular tool in the armoury of the United States’ agencies responsible for reviewing merger proposals.141

The argument that is made in favour of forgoing the market definition process and relying upon empirical predictions has been described in the United States as “simple and powerful”.142 As previously stated, market definition is seen as merely a means to an end. The end is to prevent firms from acquiring or increasing market power, that is, in basic terms, the power to raise prices to supra-competitive levels. Thus, so it is argued, “defining the market is unnecessary if econometric techniques can produce reliable and accurate predictions of price effects”.143 A similar view has been expressed recently by an Australian commentator, observing that:

The historic importance given to market definition in competition policy is overdone. If market definition were given the prominence proportional to its economic role, it would be relegated to not much more than a footnote. Instead, the focus would be placed on techniques for directly measuring the ability of firms to profitably raise prices. It is possible to carry out the task of competition policy

134 King, n 37, pp 28-32.
135 King, n 37, p 29.
136 King, n 37, p 30.
137 Australian Guidelines, [5.59], [5.62].
138 King, n 37, p 29.
139 See, eg, Overstreet T, Keyte J and Gale J, “Understanding Econometric Analysis of the Price Effects of Mergers Involving
Differentiated Products” (1996) 10 (Summer) Antitrust 30; Werden G and Froeb L, “The Effects of Mergers in Differentiated
Products Industries: Logit Demand and Merger Policy” (1994) 10 Journal of Law, Economics and Organisation 407; Werden,
n 61; Baker J, “Contemporary Merger Analysis” (1997) 5 George Mason Law Review 347; Werden G and Froeb L, “The
Antitrust Logit Model for Predicting Unilateral Competitive Effects” (2002) 70 Antitrust Law Journal 252; Epstein R and
Rubinfield D, “Merger Simulation: A Simplified Approach With New Applications” (2002) 69 Antitrust Law Journal 883.
140 See Hausman and Leonard, n 46 at 337-338; Werden and Rozanski, n 45 at 40, 41.
141 See, eg, Shapiro C, “Mergers with Differentiated Products” (1996) 10 (Spring) Antitrust 23 at 23.
142 Harvard Law Review Note, n 58 at 2426.
143 Harvard Law Review Note, n 58 at 2426.

enforcement without any reference to market definition – and the results are likely to be more

consistent, predictable and more soundly based in economic realities ...

The key point here is to note that there are a variety of approaches to directly estimating market power and/or the competitive impact of a merger that competition authorities can and do, already, use. The business of competition policy, it seems, can be carried out in a quite satisfactory manner, even without bothering with market definition.144

The question in choosing between structural analysis and empirically based predictions is seen therefore as a simple question of which method better predicts such effects.145 Economists generally consider empirical analysis to be superior in this regard.146 With this method, the likelihood that a firm will be able to raise prices unilaterally as a result of the merger is said to depend less on the merging firms’ market shares or the concentration of an arbitrarily defined market than on whether consumers view the merging products as close substitutes and whether there are other products that are highly substitutable with the merging brands (matters on which market shares and concentration are said to be uninformative, if not potentially misleading).147

The market definition-market share analysis is seen as particularly problematic in differentiated products settings.148 One of the main reasons for this is that, in such settings, there is no clear break in the chain of substitutes. Rather, products are offered over a broad spectrum of price and quality, making decisions about which products should be included and which excluded from the market an uncertain and unavoidably arbitrary exercise.149 Economists harbour the concern (not entirely unfounded, as the recent decisions in Gillette 828 F Supp 78 (DDC 1993) and Oracle 331 F Supp 2d 1098 (ND Cal 2004), referred to above, attest) that, in such circumstances, courts tend to define markets broadly, resulting in the appearance of a low market share for the merged firm and hence the danger that the potential for the merger to create or strength unilateral market power will be underestimated.150

In addition, in differentiated product industries, firms compete on a range of dimensions other
than price. As a result of this, products competing against each other in such a setting may have widely divergent prices. The possibility that products bearing very different prices are in the same market complicates the market definition exercise considerably.151 A related concern associated with market definition in differentiated product settings is that differences between products will be very much a matter of subjective customer perception.152 There are problems in quantifying such differences and the evidence of customer perceptions is notoriously difficult to compile and present, particularly in the context of litigation.153

Albeit attractive in their simplicity and not without merit, these arguments not only undervalue
the conceptual importance of market definition as a preliminary step in assessing competitive effects (as previously discussed), they also tend to overstate the practical utility of empirical models. From a practical perspective, there is little doubt that quantitative techniques have their limitations as a source of evidence to prove the direct effects of a merger, particularly in a litigated setting. The most significant of such limitations relate to data availability, choice of methodology and reliance on a series of restrictive assumptions. Thus, in comparing a market definition-led approach with an

144 Biggar, n 36 at [8]; [69].
145 Biggar, n 36 at [62]-[64].
146 See, eg, Shapiro, n 141; Werden and Rozanski, n 45; Hausman and Leonard, n 46.
147 Overstreet et al, n 139 at 31; Shapiro, n 141 at 28.
148 See generally Keyte J, “Market Definition and Differentiated Products: The Need for a Workable Standard” (1995) 63 Antitrust Law Journal 697; Werden and Hay, n 9.
149Werden and Rozanski, n 45 at 40.
150Werden, n 61 at 368-369; Church J and Ware R, Industrial Organization: A Strategic Approach (2000) pp 604-605.
151 Starek and Stockum, n 32 at 806-808.
152Werden and Rozanski, n 45 at 42.
153 See the discussion of consumer evidence in Beaton-Wells C, Proof of Antitrust Markets in Australia (Federation Press, 2003) Ch 5.
econometric analysis-led approach, it has been pointed out that:

Just as market definition often determines the merger’s legality under the structural approach, choices of methodology, underlying assumptions, and data drive the outcome in econometric analyses. The reliability of a specific set of predictions stands or falls based on whether the assumptions underlying the econometric model used are consistent with the circumstances of the merger and whether the data used are reliable.154

In basic terms, empirical analysis, as utilised in differentiated product merger cases, involves the performance of a merger simulation.155 This involves estimating own-price and cross-price elasticities and predicting price effects from such estimates.156 A firm’s ability unilaterally to raise prices depends on the elasticities of demand and supply it faces. Given that the data necessary to construct a full system of demand and supply equations generally are not available,157 economists have developed models that can estimate elasticities and predict unilateral market power with a reduced data set.158 These models make many simplifying assumptions. Hence, the reliability of a particular model’s predictions is very much dependent on whether the assumptions underlying that model are consistent with reality.159

Once the relevant elasticities are estimated, post-merger prices and output can be predicted using a simulation model.160 Merger simulation is, in simple terms, “a procedure through which the estimated demand parameters are combined with pre-merger prices and outputs, and processed systematically through a conventional economic model of short profit maximisation”.161 The reliability of predictions based on any given simulation model again depends on the validity of the underlying assumptions as applied to the case in question. Such assumptions relate to (1) competitive
interaction for the industry and whether it follows Bertrand competition (ie non-cooperative
price-setting); (2) marginal cost and whether it varies in the relevant range; and, most importantly, (3) the shape of the demand curves for the products of interest.162 In relation to the first of these assumptions, there is some debate over whether firms can be assumed to behave consistently with the Bertrand model of competition.163 The second assumption generally has not proven to be problematic.

154 Harvard Law Review Note, n 58 at 2431.
155 There are other ways of quantitatively predicting the unilateral effects of a merger; however, merger simulations are the main way and have certainly attracted the most attention and debate in merger enforcement circles and the economic literature in recent years.
156 Own price elasticity is a measure of the responsiveness of quantity demanded of a product resulting from a change in its own price. Cross price elasticity of demand is a measure of the responsiveness of quantity demanded of one product in response to an increase in the price of another product.
157 See Rill, n 18 at 404, noting further that: “It is true that the advent of retail scanners has increased the opportunities to attempt demand modelling. However, even that data can be limited by the conditions surrounding its collection. Moreover, while economists have suggested that additional assumptions can be made that circumvent the need for perfect information, these assumptions necessarily weaken the models’ results.”
158 Harvard Law Review Note, n 58 at 2425. Leading models include the HLZ “Almost Ideal” Demand System; the Antitrust Logit Model and the Residual Demand Elasticity Model: Rill, n 18 at n 34.
159 Rill, n 18, noting in n 35 that the HLZ Model makes the fewest assumptions but requires quite detailed data; the Antitrust Logit Model makes assumptions about the firms’ marginal costs, non-price strategies of competition, and the shape of the demand curve; while the Residual Demand Elasticity Model holds constant market-wide demand and supply and also requires identification of firm-specific variation in cost and other factors that affect demand.
160 Harvard Law Review Note, n 58 at 2425.
161Werden, n 61 at 363.
162Werden, n 61 at 374-378 for a summary explanation of these assumptions.
163 See Muris T, “Economics and Antitrust” (1997) 5 George Mason Law Review 303 at 311; see also the views expressed in Whither Merger Simulation, available at http://www.antitrustsource.com, May 2004 (viewed 20 September 2005). Cf Werden, n 61 at 374-375, explaining the Bertrand model of competition and indicating that, in his view, this model is a “very reasonable assumption in most differentiated products industries”.
However, the third is the most sensitive – even slight miscalculations about the shape of the demand curve for the relevant products can yield erroneous conclusions about post-merger price increases.164

Given the present state of empirical discipline, there is often substantial disagreement over many of the aspects of a simulation study.165 The choice of methodology can be difficult, and reasonable economists inevitably disagree on the appropriateness of a particular model in any given situation.166 At least at present, no single model is seen as technically superior in all cases; “which model is appropriate in a given case depends on industry characteristics, the type of data available, how firms in the market compete, and other considerations”.167

In light of these difficulties it is hardly surprising that, despite the stridency and high profile of its advocates, empirical analysis is yet to become a major feature of merger litigation in the United States. Indeed, the use of such analysis by courts is seen as raising greater problems in merger cases than in other areas of law because “the econometric analysis is less well-established in the merger context, the methodological issues involved are generally more complex, and there is greater disagreement over the proper resolution of methodological difficulties”.168 The result is a battle between the experts which is almost impossible for a judge to resolve.

Even in the United States agency review context in which empirical analysis has been cultivated heavily by agency economists in recent years, there is growing recognition of its limitations and pitfalls. It has been argued recently by senior agency officials, for example, that the modelling choices made in merger simulations for internal agency purposes should be subject to the same rigorous standards applicable to expert opinions proffered as evidence in court proceedings.169 Furthermore, there is growing acknowledgment that the more “traditional” sources of evidence, such as industry

164 As one Federal Trade Commission official, Baker, has noted: “price simulation models are generally sensitive to assumptions about the way the elasticity of demand varies as output changes, a matter about which it is often difficult to gather information or engage in informed speculation”: Baker, n 112. It is true that, in an attempt to overcome this, some models attempt to measure price elasticity more directly (see Overstreet et al, n 139, reviewing such models). However, it has also been observed that “regular employment of [such] models may tend to artificially elevate the level of concern in unilateral effects cases. At best they may result in pitched theoretical battles between the parties’ economists over the exact level of price increase to be expected, with correspondingly less focus placed on competitive responses.” See Rill, n 18 at 405-406.
165 Harvard Law Review Note, n 58 at 2431.
166 Harvard Law Review Note, n 58 at 2431-2.
167 Harvard Law Review Note, n 58 at 2432, noting that: “The DOJ, for example, has used the Logit Model, the RDE Model and the Diversion Ratio on various occasions. Economist Carl Shapiro promotes the Diversion Ratio as a tool for evaluating unilateral effects, but economists Jerry Hausman and Gregory Leonard criticize any reliance on the Diversion Ratio. In evaluating the Logit Model, Hausman and Leonard comment that ‘econometric tests of [the model’s] assumptions have a high rate of rejection.’ They advocate their own model (the HLZ model), but economists Gregory Werden and Luke Froeb outline problems with it, and the frequent unavailability of requisite data limits the usefulness of this model” (footnotes omitted).
168 Harvard Law Review Note, n 58 at 2435.
169 See Werden G, Froeb L and Scheffmann D, “A Daubert Discipline for Merger Simulation” (2004) 18 (Summer) Antitrust 89. Werden and Froeb argue that merger simulation should be disciplined, whether in the courtroom or in the agency context, by Federal Rule of Evidence 702 and the principles established in Daubert v Merrell Dow Pharmaceuticals Inc [1993] USSC 99; 509 US 579 (1993) which limit the admissibility of expert testimony. In particular, they argue that the analyst performing the simulation must be an expert in the relevant field of economics; that the simulation employ sounds methods from that field and, most importantly, that those methods are applied reliably to the facts of the industry at hand. These authors argue further that every assumption made in connection with a merger simulation must be justifiable and, absent sufficient justification, a sensitivity analysis be performed so as to show the extent to which the predictions depend upon the assumptions. See further the views expressed by these same commentators in Whither Merger Simulation, available at http://www.antitrustsource.com, May 2004 (viewed 20 September
2005). Scheffmann, a former Director, Bureau of Economics, Federal Trade Commission, has been one of the most outspoken critics of merger simulation analyses and during his tenure at the Commission was responsible for resurrecting the focus on the coordinated effects of mergers.
witnesses and documents, remain valid and useful, and that quantitative analyses should be used to supplement rather than replace such sources, in determining whether or not a transaction violates s 7 of the Clayton Act.170

In Australia, qualitative evidence from industry sources is undoubtedly the most influential and
highly valued source of evidence on a range of issues arising under Pt IV of the TPA, including market definition and anticompetitive effects.171 By contrast, the employment of quantitative analysis has not been a common feature of Pt IV proceedings in Australia to date.172 Cases in which evidence of this nature has been adduced have been few and far between and in those instances in which it has been presented, it has been ineffective in the sense that it has had limited impact on the court’s findings.173 Judicial reservations towards this category of evidence were illustrated recently in the merger context in Australian Gas Light Co v Australian Competition and Consumer Commission (No 3) [2003] FCA 1525; (2003) 137 FCR 317; [2003] ATPR 41-966, the first substantive merger to come before an Australian court in over a decade.174 Detailed empirical evidence based on a residual demand analysis was presented by the ACCC in an attempt to show that the three large Victorian generators had the ability to increase market prices through their unilateral bidding in the national electricity market. This
evidence was examined closely by French J but ultimately failed to persuade him that the proposed acquisition met the test in s 50 of the TPA. For some, this result has only reinforced the impression that the substantial investment involved in compiling empirical evidence may not be worthwhile in Pt IV proceedings.175

Notably, even in the authorisation context in which economic modelling has become increasingly popular in recent years, recent comments by the Tribunal indicate that quantitative estimates of claimed public benefits will be subject to close scrutiny and that there may be circumstances in which, owing to difficulties with data and methodology, the better judgment is simply not to rely on such estimates.176

CONCLUSION

In 1993 one of the United States’ leading antitrust officials made the observation that “the [US] Guidelines’ approach may be subject to further refinement, but it appears that it will survive well into

170 Barnett T (Deputy Assistant Attorney-General, Antitrust Division, United States’ Department of Justice), “Substantial Lessening of Competition – The Section 7 Standard” (2005) 2 Columbia Business Law Review 293 at 304-311. See also the commentary in Bates J, “Customer Testimony of Anticompetitive Effects in Merger Litigation” (2005) 2 Columbia Business Law Review 279.
171 See Beaton-Wells, n 153, Ch 4 (for a general review of this category of evidence in market definition cases); pp 329-331 (for conclusions, specifically that “courts have valued both highly and consistently, evidence that is derived from the industry relevant to the case at hand concerning, essentially, the competitive dynamics in that industry (‘industry evidence’). Where it has been necessary to resolve inconsistencies or ambiguities in the evidence presented on the issue of market definition, industry evidence generally has been preferred over every other category of evidence. In this sense it can be characterised as the most effective type of evidence adduced on this issue”).
172 Beaton-Wells, n 153, Ch 6 (for a review of the few cases in which it has been presented).
173 In Trade Practices Commission v Australia Meat Holdings Pty Ltd [1988] FCA 244; [1988] ATPR 40-876 the court declined to place any weight on evidence of price correlations and gave specific reasons for doing so. In News Ltd v Australian Rugby Football League Ltd [1996] FCA 1256; (1996) 58 FCR 447; [1996] ATPR 41-466 the court either overlooked or chose to ignore the quantitative evidence (aimed at showing limited substitutability for rugby league). In Australian Rugby Union Ltd v Hospitality Group Pty Ltd [2000] FCA 823; [2000] ATPR 41-768 evidence of cross-elasticities was found not to be statistically significant and in Australian Competition and Consumer Commission v Universal Music (2001) 115 FCR 442; [2002] ATPR 41-855, qualitative evidence was preferred to the quantitative evidence presented on the issue of market power.
174 The Commission presented detailed quantitative evidence to support a decision that it had made to refuse informal clearance of AGL’s proposed acquisition of a 35% stake in the Loy Yang Power Station and Coal Mine. The analysis was made possible by the availability of data concerning electricity trading in the national electricity market.
175 See Merrett A, “Quantitative Analysis Again up in Lights” (2005) 13 TPLJ 90 at 96. Note also the observations made by Allsop J about the limitations of economic modelling, specifically the simplicity of the assumptions on which they are invariably based and hence their inaccuracy in representing reality, in Australian Competition and Consumer Commission v Baxter Healthcare Pty Ltd (with corrigendum) [2005] FCA 581; [2005] ATPR 42-066 at 43-039 [512]- [513].
176 See Re Qantas Airways Ltd [2005] ATPR 42,065 at 42,878 [206]-[208].
the next century”.177 In hindsight this has proven to be an insightful prediction. For while there have been bold assertions of an abandonment of the traditional structural approach to merger analysis by the federal agencies, serious limitations on such a development have also been recognised.

The first limitation arises from the terms of s 7 of the Clayton Act and its at least implicit
requirement of market definition, together with the resistance that has been shown to any possible resurrection of the submarket concept. The second limitation stems from the fact that definition of a relevant market and analysis of the structural elements of that market, including but by no means limited to market shares and levels of concentration, remain a theoretically valid and practically feasible method of predicting competitive effects. This is clearly the case for mergers in which coordinated effects may be of concern given that such effects are not susceptible to quantification in the same way as unilateral effects.178 However, even in the case of unilateral effects, there is growing acknowledgment of the weaknesses of so-called direct, empirically based, techniques for predicting merger outcomes (including by members of the United States’ agencies who have been responsible in large part for pioneering these techniques).

For essentially the same reasons, it is unlikely that Australian competition lawyers will be forced to confront “mergers without markets” for at least the foreseeable future. Indeed, this conclusion may be reached with even greater confidence in Australia than in the United States having regard to the clear statutory profile of the market concept under the TPA and the apparent lack of support (amongst the judiciary, if not more generally) for the introduction of econometric analysis as a substitute for qualitative sources of evidence in establishing competitive effects. Thus, barring significantamendment of the TPA and substantial development in quantitative techniques, it is predicted that market definition and the structural paradigm to which it is integral will remain key features of the Australian competition law landscape well into the present century.

177 Werden, n 67 at 555.
178 See Shapiro, n 141 at 23; also Harvard Law Review Note, n 58 at 2427-2431 regarding the potential for coordinated interaction in differentiated product industries.