Charles E. Biggio
Antitrust Law Committee
May 15, 1996
Good afternoon. It's a pleasure to be here this afternoon. I'm grateful for the opportunity to talk with you a bit about the Antitrust Division's merger program. I'm not going to focus on just one topic. I thought I'd give you an overview of what we've been up to. I'll touch a bit on our workload--which is tremendous, with no let up in sight--and then cover some of the substantive and procedural issues we've been facing, pointing as I go along to some of the key enforcement actions we've taken recently to illustrate how we handle these issues. At the end, I'll be happy to take your questions, so long as you permit me the usual right to duck any question about pending cases.
Driving our work, of course, is the current merger wave. We're swamped at the Division, and I know the FTC would say the same thing. Last year's merger activity was unprecedented. By one count, 8,956 mergers were announced last year, worth a total of $458 billion.
As you would expect, this increase in merger activity means more work for the Division:
There are a couple interesting things to note about this merger wave, at least for antitrust purposes. First, I think the mergers we're seeing now are different in an important way from the merger wave in the 1980's. We're not seeing so many "LBO" financial mergers. These financial deals virtually never posed antitrust problems. Instead, today we're seeing more strategic mergers between competitors or firms with complementary products or technologies. These types of deals are much more likely to have competitive problems, or at least ambiguous competitive consequences.
What brought this home for me was an article earlier this month in the Wall Street Journal on the boom in high-tech mergers. Instead of developing new products and technologies internally, many high-tech firms are choosing to buy new businesses. The reasons given in the article were: the acquirer's rich stock prices, consolidation in glutted product categories, the demand by corporations to use fewer vendors and more liberal telecommunications laws. "Above all," the article said, "hardware and software companies are racing to exploit explosive communications markets and can't afford the time it would take to build new businesses internally." As an antitrust lawyer, I went through these reasons saying, well, consolidation of glutted product categories, that could be bad for competition. On the other hand, the race to exploit new markets requiring faster movement than can be achieved internally, this sounds OK, but I also thought that there could well be some entry barriers somewhere leading firms to a buy vs. make decision. What I took away from this article is: The strategic mergers we're seeing today are hard to analyze. They take time and care to review.
Another interesting aspect of our current merger wave is the impact of deregulation. Everyone who thought the Telecom Bill wouldn't spur the regional operating companies to merge were surprised last month. In another area, deregulation of the electric utilities, both at the federal and state levels, coupled with technical advances in the ability to transmit power over longer distances, has spurred a merger wave in this industry. I believe there are at least ten utility mergers pending around the country. The banking industry is also consolidating in response to technological and regulatory changes. As I noted earlier, we review about 2,000 bank mergers a year. Deregulation is contributing to the growing number of deals we have to review, and it is contributing to the complexity of issues we have to grapple with.
My point is that, unlike the merger wave of the 1980s, we're seeing not only quantity, but also quality. Transactions today are squarely presenting a host of important and complicated issues: Will a merger retard innovation? Will the vertical aspects of, say, a media merger, raise competitive concerns? Will the merger give rise to a unilateral anticompetitive effect? I daresay that some mergers may well cause us to look at potential competition. All this makes our job interesting, but also a little bit harder.
Within the Division, we've taken steps to deal with our increased workload. For example, we have recently realigned our sections so that Litigation I Section, still under Chief Tony Nanni, will handle only criminal matters. Litigation II Section and the Merger Task Force will do only mergers. Bob Kramer remains the chief of Lit II, but Lit I and Lit II have swapped Assistant Chiefs: Willie Hudgins has moved over to Lit II and David Blotner has moved over to Lit I. Conrath and Horwitz remain the Chief and Assistant Chief of Merger Task Force. We believe this reorganization will allow us to manage our scarce merger resources more efficiently.
Another example of our efforts to handle our increased workload is a happy by-product of Anne Bingaman's efforts to improve cooperation between the Division and the state attorneys general. In the Kimberly-Clark/Scott Paper merger, for example, we worked hand-in-hand with interested states from the beginning of our investigation. Where appropriate, we try to work with state investigators, interviewing and even deposing witnesses together, sharing documents where the parties consent or where the law permits. In three cases, the Morton Plant-Mease hospital merger, the Waste Management/ Reuters merger and the Kimberly-Clark/Scott merger, we filed joint complaints, the first two with Florida, and that last with Texas. In addition to furthering the effort to achieve consistent antitrust enforcement, cooperating with the state AG's has also been a big help in extending our resources.
I'd like to switch gears a bit and touch on some of the substantive issues we've been dealing with recently.
Differentiated Products/Unilateral Effects
We've recently had a number of mergers that have afforded us the opportunity to focus on a merger's competitive effects in a differentiated product market. These types of mergers present the issues of whether the merged firm will be able to raise prices unilaterally, though collusion is also an issue we look at.
Oversimplifying greatly, a unilateral anticompetitive effect arises where a merger lets the combined firm raise the price of one or both of the formerly competing products, without regard to the constraints imposed by the remaining competitors. This can happen because the products of the two merging firms are sufficiently close substitutes for a significant number of consumers that, once that competition is eliminated, a large number of consumers would still buy one of the two merged products, and the loss of sales to the remaining more distant competitors would not be great enough to make the price increase unprofitable. The merged firm would reap the benefits of the higher price for the first product, without losing sales to another firm, since switching consumers in large part would switch to the firm's other (acquired) product. This is possible because the alternative products have differing sets of attributes that consumers don't perceive to be adequate substitutes. In effect, the competition that took place before the merger would be internalized in the merged firm, which would no longer suffer a loss of profits due to the competition, because it wouldn't care which of the now merged products a consumer chooses.
Against these "demand-side" considerations, we weigh certain supply side factors. Specifically, if it is likely that a remaining competitor will reposition its product so that it will become a sufficiently attractive new substitute for the merged firm's products, the likelihood of a unilateral effect may be reduced. A firm can reposition its product by changing the product's attributes or by changing the brand identity of the product. An example of the former case would be a software product adding features to make it more like a higher-end program. A firm can reposition a brand by targeting its advertising to a different group of consumers. You may have noticed McDonalds starting to reach out to older consumers in its recent ad campaign. In addition to product repositioning, cost savings realized by the merged firm may in certain circumstances create an incentive for the firm to lower its price, independent of competition. When such synergies are clearly proved to us, we'll take them into account when evaluating whether a unilateral price increase is likely. I should note that the 1992 Guidelines sets out a threshold of 35 percent combined market share for establishing presumption of a unilateral effect. While we take this threshold seriously, you should not take it as a guarantee that we will not pursue a unilateral effects case where the combined share is below 35 percent.
A recent example of the Division's efforts in this area is International Baking's purchase of Continental Baking from Ralston-Purina. Before the merger Continental was the #1 wholesale baker in America; Interstate was #3. Each make a wide range of cake and bread products. The markets of competitive concern were white pan bread in five cities.
Continental's primary white bread brand is Wonder; Interstate sells under a number of brands, depending on the region. Interstate's major brands are Butternut, Sunbeam, Mrs. Karl's and Weber's. White bread, believe it or not, has special attributes that a significant core group of consumers view to be special; notably its distinctly bland flavor and texture. I guess kids often refuse to eat any other kind of bread. Actually, while one might think that all white bread is alike, in fact, the parties and the other bakeries spend a lot of effort developing and nurturing brands that attempt to distinguish one white bread from another. It is this pronounced brand equity that differentiated these products.
In the relevant geographic areas, the Division found that Interstate and Continental were either the only two or two of three major premium brands of white bread. Private label also accounted for a significant chunk of sales; but there were no other brands of significance. Thus, the acquisition would have increased significantly the combined shares of the merging firms. Based on the evidence, we concluded that a price increase for Wonder would have resulted in Interstate's brand picking up most of the diverted sales, and vice versa. Private label, while a factor, would not have picked up sufficient sales to defeat a significant price increase. Because of the significance of brands, entry or repositioning by new bakeries or expansion by existing suppliers would not be an adequate competitive check. This was evidenced by a history of several failed attempts at entry by pretty substantial companies with strong brands. We also concluded that any cost savings that might be obtained in the merger would not offset the predicted price increases in the relevant markets.
What does the bread case, as well as the Kimberly-Clark/Scott case, which also involved differentiated tissue markets, suggest for arguing cases before the Division where the relevant product markets are differentiated? One way to think of this problem is in terms of what arguments defense counsel could make in a differentiated product market that might not necessarily be persuasive to us:
Geographic markets: A broad geographic market based on how far a product is shipped from a plant will not be persuasive where local competition is determined by the presence of competing brands in a the local area. After all, being able to ship a product to a grocery store won't do you any good if you can't sell the product because nobody recognizes your brand. In the bread case, it was true that bread could be physically shipped comparatively long distances. However, when you looked at pricing and where firms were actually selling, the fact that bread could be shipped over such distances didn't make much difference. It was where the brand name was recognized that counted. We concluded that the geographic markets were fairly local (e.g., Chicago, Peoria, San Diego and Los Angeles).
Production capacity: While productive capacity (or excess capacity) may in certain circumstances be a competitive constraint, the ability to produce will not given as much weight as the ability to sell. Again, in bread, there was plenty of capacity; but this fact didn't make much difference in terms of a firm's ability to sell the product. Again, it was the brand that mattered and that established a firm's market share.
Product market: Where products are differentiated, arguing that the market includes a wide range of potential substitutes may not be particularly persuasive. We'll be interested in examining the extent to which the products of the merging firms are sufficiently close (and isolated) substitutes that other products do not provide an effective constraint on price. Similarly, to the extent that the merging firms are not close substitutes, that will be an important factor in evaluating whether the merger is unlikely to result in a unilateral effect.
Entry/Repositioning: Even though it may be easy to begin making a product, if brand equity is important, simply building a facility will not translate into sufficient sales to defeat a unilateral effect. You'll also have to explain how the new entrant will attract consumers in sufficient numbers to replace the competition lost by merging close substitutes. Also, since the anticompetitive effect is tied to the significance of the individual brands of the merging parties, unless the new product is viewed by consumers to be a comparable substitute for one of the merging brands, entry or repositioning--even if physically easy--will not likely resolve our concerns.
We're really not doing anything that new here. The economic theory underlying our differentiated products analysis is well known and we are employing Part 2 of the 1992 Horizontal Merger Guidelines. What is noteworthy is the improvement in our ability to collect and analyze data, which allows us to predict with greater confidence when a transaction is likely to give the merged firm the ability to raise price by itself, without regard to the actions of the remaining competitors. You should be prepared for interrogatories and document requests in our second requests that seek data relevant to this inquiry. For example, for consumer products, detailed point-of-sale data is usually collected by third parties like IRI and purchased and carefully studied by companies. This type of data is highly probative, and you can expect that we'll be asking for it in our second requests and from third parties through CIDs, along with market studies, focus group analyses, consumer switching studies, and so on.
If you want to learn more about how the Division is likely to analyze a differentiated products merger, I recommend Carl Shapiro's article in the Spring 1996 issue of the ABA's Antitrust Magazine.(1) That article provides a more systematic discussion of how we look at these types of mergers.
I think we're seeing more deals that tee-up vertical issues. The FTC's enforcement action in the pharmaceutical benefit management merger (Eli Lilly/McKesson) is an example. The Division has also obtained relief in several vertical mergers. Specifically, the Sprint/France Telecom/Deutsche Telekom and MCI/British Telecom deals, as well as the TCI/Liberty Media transaction, involved vertical issues. We also looked carefully at the vertical issues in the Disney/ABC transaction, where we ultimately concluded that there was not likely to be a competitive problem. With the changing regulatory landscape in telecom, television, and other industries like electric utilities, I think we can expect more mergers that squarely present interesting vertical issues.
Generally, I believe we won't find competitive problems in the vast majority of vertical mergers (leaving aside, of course any horizontal issues that might also be presented by a merger).
If a vertical merger is problematic, it's problematic because of a probable downstream price or output effect. That effect does not arise simply because input prices are raised to non-integrated rivals--these increased input prices must translate into higher market prices or lower output downstream.
There are three basic theories under which consumers can be harmed by vertical mergers. First, there's foreclosure (i.e., raising rivals costs). Second, there is the increased chances of coordinated behavior (i.e., a horizontal ramification of vertical merger). Third, the avoidance of rate regulation.
Generally, for foreclosure to result in an anticompetitive effect, we believe both the upstream and downstream markets should be susceptible to the exercise of market power. Whether market power could be exercised in the upstream or downstream markets would be determined under the 1992 Horizontal Merger Guidelines.
Similarly, for a vertical merger to increase the chances of anticompetitive coordination, the upstream and the downstream markets must be susceptible to the exercise of market power. Other conditions are also important. For example, there must be substantial transactions between the integrated firm and one or more of its customers or suppliers. These relationships provide the conduit for the exchange of information that might result in coordinated activities. The information must be reliable and generally not otherwise available. Often, the suppliers and customers of the integrated firm will provide the integrated firm with misinformation. If the information is not reliable, it's not going to be a good mechanism to facilitate collusion. Also, if the information in question is already known to the market participants, the merger won't increase the likelihood of coordination. Finally, the coordination must result in the firms having the ability and incentive to raise prices in the downstream markets.
A vertical merger may also provide a regulated firm with the ability to shift costs to the regulated market. In doing so, the regulated firm could evade price regulation in that market. If a vertical merger increases the likelihood of this happening, I think it's uncontroversial that this would be a problem.
An example of our enforcement efforts is the Sprint/France Telecom/Deutsche Telekom transaction. The transaction involved a joint venture between Sprint and FT and DT (together with a large purchase of stock in Sprint by FT and DT) to provide global telecommunications services. We concluded that the vertical relationship between FT and DT on the one hand and Sprint on the other could reduce competition by placing other U.S. firms at a competitive disadvantage by creating incentives for the parties to discriminate against competitors in the terms and conditions of access to France's and Germany's monopoly networks and services.
The decree obtained by the Division would permit the deal to take place subject to certain restrictions on the parties' ability to discriminate against other firms seeking access to the French and German markets. Specifically, the decree prevents the parties from providing certain services until competitors have the opportunity to provide similar services in France and Germany. It also prevents them from giving Sprint more favorable access to FT's and DT's networks. In addition, restrictions were placed on the ability of Sprint to obtain proprietary, competitively sensitive pricing data that FT or DT might obtain in dealing with Sprint's competitors. The decree thus prevents the foreclosure of U.S. competitors from the German and French markets and reduces the chances of anticompetitive coordination by Sprint and its competitors.
As I said, we don't expect that there will be numerous challenges to vertical mergers. I would note, however, that the telecom and media deals are presenting vertical issues squarely, and the Division will look at these issues carefully. If you have a vertical case or would like to get a more detailed description of how the Division is likely to analyze a vertical merger, I suggest you read Steve Sunshine's speech at the 1995 ABA Spring Meeting.(2)
For my last substantive topic, I'd like to mention our old friend collusion. With all the talk about innovation markets, vertical issues, differentiated products, you may have thought we've given up on the risk of collusion as an anticompetitive effect. We haven't. One of our most recent cases involved a risk of collusion theory. Georgia-Pacific, one of the largest paper and wood products companies agreed to buy the gypsum wallboard business of Domtar, which is a Canadian company. Georgia-Pacific and Domtar were about the same size and were the third and fourth largest producers of gypsum board in the United States. The number one and two firms (US Gypsum and National Gypsum were each significantly larger than Georgia-Pacific and Domtar pre-merger. After the merger, Georgia-Pacific would be about as big as US Gypsum and National Gypsum. The remaining fringe firms are small, each operating only one or two plants.
Gypsum is a homogeneous product. No one said that brand name mattered, though service and reliability did have some value. Gypsum is a heavy product and expensive to ship long distances. As a result, the geographic markets tend to be regional. Following our investigation, we concluded that the merger was likely to increase the risk of collusion in the northeastern United States, where Georgia-Pacific and Domtar each have two plants. In the Northeast, US Gypsum, National Gypsum and Georgia-Pacific would have about 90 percent of the market. As I noted above, gypsum is a homogenous product sold mainly on price. Capacity, production and pricing data are widely available and price changes are normally announced well in advance of implementation. In addition, at least once every generation this century, civil or criminal actions have exposed successful price fixing arrangements among the dominant gypsum board manufacturers. Finally, the small fringe firms would not have the incentive or the ability to undercut a cartel, so these firms effectively could be ignored by the three big players. Based on these and other factors, we thought there would be a problem in the Northeast.
To meet our concerns, Georgia-Pacific entered into a consent decree with us requiring it to divest two of its wallboard plants in the Northeast to one or two independent firms. This will put significant capacity in the hands of a new firm that would be unlikely to participate in any express or tacit coordination by the three major producers.
Finally, I'd like to say a few things about the remedies we will consider. We try to be flexible. Our goal is to resolve the competitive problems while permitting the parties to achieve the efficiency enhancing aspects of their deal.
Our bread case is a noteworthy effort of our flexibility. Recognizing that the brands were the key assets, our decree with the parties requires that only brands must be divested--together with such additional hard assets (plant, distribution system, etc.) as the purchaser may require to be competitive at the previous market share level associated with the divested brand. We concluded that there was sufficient bread baking capacity in the market that the parties need not be required to make capacity available if the purchaser could find it elsewhere. This flexible approach ensures that the key competitive assets (the brands) will be divested, while affording the defendants the ability to find a buyer that may not require manufacturing or distribution assets to be competitive, thereby reducing the burden on defendants. In contrast, in Kimberly-Clark/Scott, we concluded that it was important that the parties divest tissue machines along with the relevant brand names. In that case, we required the parties to offer a menu of four plants from which a purchaser or purchasers could select two. These two cases are reflective of what can be done in a differentiated product market. In some cases, the brands (and not necessarily productive capacity) are the key competitive asset; in others, plant and equipment may be important parts of the divestiture package. These cases show our efforts to remedy anticompetitive effects where they exist while not requiring that assets unnecessary to the buyer's competitive viability be divested.
Another example of our flexibility is our arrangement resolving our concerns in the Disney/ABC merger. There, we signed up a consent order with Disney that would require it to sell one of the two overlapping television stations in the Los Angeles market. We agreed, however, not to file the decree in light of Disney's voluntary commitment to sell one of the stations and in light of the expectation that the FCC would enter an enforceable order requiring the sale. However, if circumstances arise that could permit Disney to keep both stations, we would be able to file the consent decree obligating Disney to make the sale.
I'd like to close with two points where we aren't going to be flexible. First, except in extraordinary cases, we will insist that any divestitures be completed within a maximum of six months from the date a consent decree is filed. Not nine months, and not six months from the day the order is entered after a two month or more Tunney Act process. We'll insist on six months from the date we file the decree--or sooner where circumstances require. If the divestiture is not completed, a trustee will be appointed to carry out the sale. This has been our policy. I don't want to put words in the mouths of my friends at the FTC, but George Cary was quite emphatic at the ABA Spring Meeting that the FTC will also insist on a six-month divestiture period.
Second, if a fix-it-first isn't consummated before closing the main deal, we will insist on an order. Also, we will also insist on an order if the remedy crafted by us and the parties requires any ongoing relationship with the party brought forward to fix a problem. I have two examples of this. First, Waste Management acquired a solid waste transfer station used as an intermediate point in the transportation of solid waste out of southern Florida. A competitor of Waste Management had a contract right to use some of the capacity of the station to transfer waste it contracted to remove from south Florida to its landfill near central Florida. Without access to the transfer station, the competitor would be out of the market. Even though the competitor had a contract that would be enforceable against Waste Management preserving the competitor's ability to use the transfer station for five years, we insisted that Waste Management's contract obligation be embodied in a consent decree to ensure that it would not take steps to disadvantage the competitor.
Another example is our still-pending lawsuit in the clay matter. The parties anticipating a problem crafted a joint venture with a third firm that in our view amounted to nothing more than a supply agreement. The parties' view was that the agreement between Engelhard and ITC, the third party, was adequate and that no decree was required. Because the agreement between Engelhard and ITC was executory, could be changed at any time by mutual agreement, or could simply be ignored, we could not accept this arrangement as a proper fix. Indeed, we concluded that the parties' fix, even if embodied a consent decree, was inadequate. In any event, this case reflects our policy of not accepting fix-it-first arrangements that involve ongoing performance. We require clean remedies, such as divestiture, that don't require ongoing "promises to compete" in order for the arrangement to preserve competition.
That's all I've got. It has been a pleasure, and I'll be happy to take questions. Thank you.
1. Shapiro, "Mergers with Differentiated Products," Antitrust Magazine, Spring 1996.
2. "Vertical Merger Enforcement Policy," Remarks of Steven C. Sunshine, Deputy Assistant Attorney General, Antitrust Division, U.S. Dept. of Justice, before the ABA Section of Antitrust Law Spring Meeting, April 5, 1995.
Merger Enforcement At The Antitrust Division
Wednesday, May 15, 1996
Updated June 25, 2015