The views expressed in these remarks are my own, and do not necessarily reflect those of the Federal Trade Commission or any other individual Commissioner.
I want to thank you for giving me the opportunity to speak before you. The food marketing industry today represents a vital sector of American industry. The industry is changing rapidly, offering consumers greater choice and higher quality. Today we see the rise of wholesale clubs and certain mass merchandisers, which are beginning to give traditional full-service supermarkets, in some areas, a run for their money. 1 At the same time, private label products are increasingly being pushed by traditional supermarkets, a development which has generally benefitted consumers by providing them with quality products at lower prices.2 My general message to you is that a competitive food distribution industry is a healthy one, one that is good both for consumers and producers.
Of course, ensuring a competitive market is the FTC s business and, in your industry, one of our main focuses concerns mergers of supermarket chains. In highly concentrated markets, mergers may lead to significant reductions in competition. By reducing the number of players in the market, a merger can make it easier for the remaining firms to maintain explicit or tacit agreements to fix prices, to reduce the quality of services offered, or to otherwise increase joint profits at the consumers expense. It is also possible that a merger may enable a single firm to have such dominance that it can increase prices or reduce quality unilaterally. In the last year, the Commission has taken action in four supermarket mergers in different parts of the country. I want to first tell you about these actions and then discuss how the Commission goes about analyzing supermarket mergers to determine whether there is a likelihood that a merger may be anticompetitive. The Commission s antitrust jurisdiction, however, is not just limited to analyzing horizontal mergers. We are also tasked with enforcement of the Robinson-Patman Act, which bars certain types of non-cost justified price discrimination. I want to conclude by giving you a sense of my views about the proper role of government enforcement of the Robinson-Patman Act.
The Commission most recently entered into proposed consent agreements with Schnucks Markets, Inc. and Schwegmann Giant Supermarkets, Inc. concerning their acquisition of supermarkets currently owned by National Holdings, Inc. 3 The agreement settles Commission allegations that Schnucks purchase of National stores in St. Louis and Schwegmann s purchase of National stores in New Orleans could reduce competition substantially, thereby potentially increasing prices and reducing the quality of products and services available to consumers. The agreement requires Schnucks and Schwegmann, both of whom currently compete directly with National, to divest 24 stores in the St. Louis area and seven stores in New Orleans. One notable aspect of this settlement is the Commission s close coordination with the Missouri Attorney General s Office in investigating Schnucks acquisition of the National supermarkets. I think you will be seeing more coordinated investigations between States Attorneys General and the Commission in the future. Second, the Commission has recently accepted as final a consent agreement with Red Apple Companies, Inc., settling allegations that purchases between 1991 and 1993 of Sloan s supermarkets by Red Apple in Manhattan could substantially lessen competition in violation of the federal antitrust laws. 4 The consent agreement requires that Red Apple divest six Manhattan supermarkets. Finally, the Commission has also given final approval to a consent agreement involving Penn Traffic Company s proposed acquisition of 45 Acme grocery stores owned by American Stores that are located in northeastern Pennsylvania. 5 The agreement requires Penn Traffic to divest three stores in Towanda, Pittston and Mount Carmel, Pennsylvania.
These settlements were the result of intensive investigation by Commission staff into the proposed acquisitions to determine if the merger could cause competitive problems. In analyzing a supermarket merger, staff follows the joint FTC-Department of Justice Horizontal Merger Guidelines, issued in 1992. Those guidelines provide a helpful framework for understanding supermarket mergers. Under those guidelines, the first task is to determine the relevant product market and geographic market, then to look at what anticompetitive effects are probable, then to examine whether entry would prevent the likely anticompetitive effects and, finally, to see if the parties have any defenses. This all may sound complicated, but it is really just common sense. Defining relevant product and geographic markets sets the boundaries for competition. Then, we look at the level of concentration in the market, utilizing market share information. We look at whether the competitive conditions are such that there is a greater likelihood of a price increase or reduction in quality or output as a result of a merger. We then conclude by looking at whether new companies could enter easily to defeat an anticompetitive effect and, also, look at any asserted defenses, such as that the merger will increase efficiencies or is necessary because one of the entities is a failing firm. I want to give you some sense of how each step of this analysis generally applies to supermarket mergers.
To determine product market in a retail setting, we generally look at a cluster of products and services. Starting by examination of the actual stores of the parties that are merging, we would go on to look at a narrow group of similar food retailers -- perhaps full-line, self-service supermarkets of a particular large size. We then try to determine what would happen if a hypothetical monopolist who controlled all supermarkets of that size imposed a small but significant and nontransitory price increase: would it be profitable or would a sufficient number of consumers switch to other outlets for food to make the price increase unprofitable? Now, as this analysis suggests, product markets need not be airtight -- a few consumers might be likely to switch but if enough remained to make the price increase profitable overall (in comparison to the lost profit from consumers to sources outside the hypothetical product market), a relevant product market would be established.6 In general, the Commission has found that full-line, self-service supermarkets with annual sales volumes of $2 million or more represent a relevant product market. Consumers view large supermarkets as providing a unique service, offering a broad array of products for one-stop shopping. One question that has come up is whether warehouse clubs or certain mass merchandisers should be included in the relevant product market. This is a fact-based issue that depends on the local conditions in a particular area and, most importantly, the perceptions and competitive actions of customers and competitors in a particular area. In most areas, we have found that warehouses and mass merchandisers do not offer anywhere near the variety of products and services that traditional supermarkets offer or the convenience that shoppers are looking for. On the other hand, there may perhaps be certain unique areas where mass merchandisers and warehouse clubs market penetration is so deep and well-established and their offerings so extensive that one might argue that they compete against traditional supermarkets.
The question of geographic market is largely answered by examining how far consumers will drive to go to a supermarket. Consumers tend to shop at supermarkets close to home. In an area where most people drive to a supermarket, the most appropriate geographic market may be the Metropolitan Statistical Area or smaller discrete areas contained therein. Of course, where supermarkets draw consumers from a very limited area, such as in Manhattan where people walk to supermarkets, a much narrower market may be more appropriate. It is also important to remember that the test for a relevant geographic market does not require that the boundaries be absolutely airtight. The test is whether a sufficient number of consumers would remain so that a price rise would be profitable over all -- that is, whether the additional profit from the price increase over the remaining customers exceeds the profit lost from the consumers on the fringe whose are likely to leave the geographic area. 7 Consequently, consumers could still be lost on the fringes of a geographic market, and yet the market could still be sufficiently established for antitrust purposes.
Once a product market and geographic market have been determined, the Commission looks at the level of market concentration, which generally depends on the number of players and their individual market shares. We use an index called the Herfindahl-Hirschman Index ( HHI ) of market concentration. The HHI is calculated by summing the squares of the individual market shares of all the participants. A market with an HHI below 1000 can be broadly categorized as unconcentrated, while a market with an HHI between 1000 and 1800 is considered moderately concentrated, and a market with an HHI above 1800 is viewed as highly concentrated. In the highly concentrated region, a merger that increases concentration by more than 100 points is viewed as presumptively anticompetitive, while a merger in the moderately concentrated region that increases concentration by more than 100 points or a merger in the highly concentrated region that increases concentration by between 50 and 100 points is viewed as raising significant competitive concerns.
However, we do not rest at merely calculating a quantitative concentration level, but look at the likelihood of an anticompetitive result as well, examining carefully whether conditions in the market are conducive to anticompetitive effects. There are two types of anticompetitive effects likely: one the result of unilateral effects and the other the result of coordinated interaction.
Unilateral effects are generally most likely where the resulting combination is either a near monopoly or a dominant firm. The concern there is that the resulting combination has such significant market power that it can successfully increase prices or reduce quality or output on its own. Unilateral effects are also possible in special situations where two supermarkets are the closest substitutes to residents in a differentiated subsection of a larger geographic market.
A second possibility is that, as a result of the merger, coordinated interaction among the remaining supermarkets in the geographic market is more likely. Coordinated interaction behavior includes tacit and express collusion, and may or may not be lawful in and of itself. Among the conditions that can indicate a susceptibility to collusion are high market concentration, difficult entry, relatively homogenous products and the ability to punish and detect cheaters to any collusive arrangement. It is sometimes difficult to understand the concern with collusion, particularly in an industry like supermarkets which offer many different types of goods and services. How, you might ask, could competing supermarkets ever coordinate? The answer is that coordination can be successful even if it is not complete or absolutely perfect. Coordination could concern one aspect of competition and still be successful in such a way that it harmed consumers -- for example, a tacit agreement to end the practice of doubling manufacturers coupons. As the 1992 Merger Guidelines note, the terms of coordination may be imperfect and incomplete -- inasmuch as they omit some market participants, omit some dimensions of competition, omit some customers, yield elevated prices short of monopoly levels, or lapse into episodic prices wars -- and still result in significant competitive harm. 8
In determining whether entry is likely to eliminate the risk of an anticompetitive outcome, we would look at how easy it would be for a new entrant to set up supermarkets in a particular area within two years or less. That is, what are the barriers to setting up new supermarkets? Developers and realtors who participate in locating sites for supermarkets can give us a good sense of how easy it would be to establish new stores. We would generally examine zoning restrictions and other site plan approval requirements as well as gather information about the difficulty of finding locations suitable for large food retailing spaces.
Once the Commission has found a relevant product market and geographic market and concluded that anticompetitive effects are likely and that there are high barriers to entry, the Commission will also examine any asserted defenses: for example, that the merger generates substantial efficiencies that outweigh the likely anticompetitive effects or otherwise is necessary to prevent the exit of a failing firm.
I hope that this provides you with a sense of the analysis that the Commission employs when it is examining a supermarket merger. It may seem frustrating that much of this analysis is necessarily so fact-specific, depending as it does on the particulars of the market involved. I have, however, tried to give you a sense of some of the issues that recur in examining supermarket mergers.
I also want to give you a practical pointer when you are considering whether to acquire another supermarket chain. In at least two recent examples, chains have sought to sell supermarkets to drug store chains -- in effect taking the stores out of the market by selling them to someone who would transform them into drug stores -- in order to solve the competitive overlap problem in a proposed supermarket acquisition. One occurred at the same time that the supermarket was acquiring a supermarket competitor. The second involved an already- consummated supermarket acquisition that did not meet the threshold size to require that it be filed for Hart-Scott-Rodino review. The respondent there attempted to sell certain stores to a drug store chain after the Commission had begun an administrative action challenging the acquisition. The respondent also sought the buyer s commitment to never operate the purchased stores as supermarkets. Putting aside the question whether the Commission should be working with parties to fix competitive overlaps without requiring a consent agreement, I want to tell you that actions such as these are simply not wise. The Commission is concerned that such actions withdraw stores from the supermarket market for reasons that are not related to their optimal usage and hence deprive consumers of desired outlets for purchasing food. Remember that one of the anticompetitive effects antitrust regulators are concerned about is the reduction of output. Indeed, from an antitrust perspective, reduction of output is the classic effect of a monopolist or cartel bent on exacting monopoly rents. Consequently, you can see how troubling such actions are from our perspective. They obviously become even more disturbing to the extent that they leave the merged entity with a dominant or near- monopoly position in the marketplace or involve agreements between the seller and the buyer not to operate the stores as supermarkets. And the Commission has sufficient tools to deal with such situations and will use them when necessary to prevent anticompetitive outcomes.9
Now, let me be clear that I do not intend to suggest that there is something wrong about a supermarket chain selling stores to a drug store chain. The Commission has no concern whatsoever about supermarkets selling stores to drug chains as a general matter. Rather, the Commission is concerned when stores are sold to drug stores in an effort to fix a perceived competitive overlap problem in a transaction involving the acquisition of a competitor s supermarkets.
I hope this gives you a general sense of our approach to supermarket merger review under the antitrust laws. I want to turn now to another law that we enforce -- the Robinson-Patman Act.
I know that there have been, at various times, allegations of violations of the Robinson-Patman Act involving the food marketing industry. I do not want to comment on any particular allegation or investigation. I do, however, want to give you a sense of my initial feeling about government enforcement of the Robinson-Patman Act.
As a new Commissioner at the FTC, one of the things I have learned is that the fundamental principle of the antitrust laws is protecting competition. Protecting competition is the best way to maximize consumer welfare by ensuring the availability of a wide selection of high quality goods at low prices. By contrast, the Robinson-Patman Act has a different agenda -- to some extent, it seeks to protect competitors, not necessarily competition. 10 The Robinson-Patman Act generally prohibits sellers from selling like goods to functionally similar buyers at different prices. It represents Congress belief that large firms could dominate markets through predation and other forms of economic warfare directed against smaller firms. The problem is that sometimes, especially as interpreted in the past, the Robinson-Patman Act may conflict with the fundamental tenets of antitrust law. For example, a price cut that harms competitors may in fact benefit consumers and, consequently, aggressive Robinson-Patman Act enforcement may sometimes unwittingly protect less efficient competitors and, ultimately, increase costs to consumers. 11 Reconciling these two divergent regimes is no easy task, and I have no simple solutions.
Of course, many types of price discrimination will make the market more inefficient and thus hurt consumers. For example, when an inefficient buyer with large market share (a so-called power buyer ) is able to exact non-cost justified discounts, market efficiency may suffer and consumers may ultimately pay more. Consequently, enforcement of the Robinson-Patman Act is not necessarily inconsistent with the antitrust laws. Indeed, the Supreme Court has recently affirmed that, for those Robinson- Patman Act allegations that are most likely to involve price cutting that is beneficial to consumers -- claims by a seller that it is injured by a rival seller s unfairly low prices, a law violation is found only when there is harm to competition, not just harm to competitors. 12
Professor Frederick Scherer has suggested that one way to reconcile the two regimes is to enforce the laws against price discrimination based on competition, efficiency, and plain common sense. 13 I think his proposal has a great deal going for it. It suggests to me that government law enforcers, as a matter of prosecutorial discretion, should examine possible Robinson-Patman Act investigations with careful attention to whether the allegations involve harm to the competitiveness or efficiency of a market and, likewise, whether relief can be obtained consistent with the principles of ensuring competition and market efficiency. As I noted before, instances where power buyers are able to obtain non-cost justified discounts and render the market more inefficient would seem to be a prime candidate for enforcement actions. By contrast, we should be wary of bringing an action that is likely to hurt consumers in any way -- by raising prices or reducing quality or output.
Now, Robinson-Patman Act investigations -- like supermarket merger reviews -- are necessarily fact-specific, so there is no magical formula for determining the best cases beforehand. Nevertheless, I do believe that what I have suggested offers some general guideposts for how I plan to look at such cases.
I hope I have given you a sense of my perspective on Robinson-Patman Act enforcement. We are committed to using our resources in a fashion that will prevent actions that are likely to result in the greatest harm to consumers. In this way, taxpayers can be assured that the Commission is giving them the best bang for the tax dollar buck. I also hope I have given you a better appreciation of the Commission s method of analyzing supermarket mergers. I always believe that business can handle adversity, but what it needs is some certainty. Although merger review is a necessarily fact-specific endeavor, understanding the way in which the Commission analyzes mergers should give you a better sense of when the Commission will challenge a particular supermarket merger. Thank you for your time today.
1 David Mills, Private Labels and Countervailing Power In Retail Distribution (Univ. Of Virginia 1992) (unpublished theoretical work); Philip Fitzell, Private Label Marketing in the 1990s (1992). See also More Shoppers Bypass Big-Name Brands and Steer Carts to Private-Label Products, Wall Street Journal at B-1 (Oct. 20, 1992).
2 Food Markets, Inc., R-7879 (Aug.4,1994).
3 Schnucks Markets, Inc./Schwegmann Giant Supermarkets, Inc., FTC File Nos. 941-0131, 941-0130 (March 8, 1995) (proposed agreement accepted for public comment) (Comm r Azcuenaga concurring separately).
4 Red Apple Companies, Inc., D-9266 (Feb. 28, 1995).
5 Penn Traffic Company, C-3577 (May 15, 1995).
6 The Coca Cola Bottling Company of the Southwest, No. 9215, slip op. at 8-9 (FTC, Aug. 31, 1994), petition for review filed (5th Cir. Nov. 23, 1994).
7 Id. at 49-50.
8 Merger Guidelines, 2.11.
9 Although antitrust theorists might debate whether Section 7 of the Clayton Act applies to a situation where there is no actual competitive overlap at the time of the supermarket acquisition, the Commission also enforces Section 5 of the Federal Trade Commission Act, which gives us authority to fill gaps in the coverage of the other antitrust statutes. It would seem that, even if Section 7 did not apply to such a situation, Section 5 would definitely apply. The Commission also has authority under the FTC Act to seek injunctive relief in federal court to preserve assets pending completion of administrative litigation, in the event that a respondent attempts to sell stores after the Commission has brought an administrative complaint.
10 [T]he legal focus of the competitive injury inquiry [in Robinson-Patman Act jurisprudence] is on the competitor, not the consumers . . . [W]hether or not output is restricted or prices are raised simply is not dispositive. Alan s of Atlanta, Inc. v. Minolta Corp., 903 F.2d 1414, 1418, n.6 (11th Cir. 1990).
11 As Judge (now Justice) Breyer remarked in a Robinson-Patman Act decision, a legal precedent or rule of law that prevents a firm from unilaterally cutting its prices risks interference with one of the Sherman Act s most basic objectives: the low price levels that one would find in well-functioning competitive markets. Barry Wright Corp. v. ITT Grinnell Corp., 724 F.2d 227 (1st Cir. 1983).
12 Brooke Group, Ltd. v. Brown & Williamson Tobacco Corp., 113 S. Ct. 2578 (1993). 13 F.M. Scherer, Industrial Market Structure and Economic Performance 582 (2d Ed. 1980).