Horizontal Merger Guidelines Review Project #545095-00002

Submission Number:
Kent Bernard
Bernard Law Office
Initiative Name:
Horizontal Merger Guidelines Review Project

The core of the Guidelines and the approach taken in them has proven to be workable, and to provide fairly predictable results. While some areas of the Guidelines could use revision, the questions posed suggest that the Guidelines are going to be used in a more prescriptive way, which would be a mistake. Antitrust in the United States remains perhaps the ultimate common law discipline. Attempts to craft detailed guidelines to function almost as regulations are unwise. The numbers in this response correspond to the numbered items in the Questions for Public Comment. We have only responded to those questions where we believe that we have something to contribute to the debate. 1. This raises the fundamental question of the nature of the Guidelines. The proposed changes would be appropriate if the Guidelines were regulations, and the Commission or the Division were bound to follow the text. But since that is not what the Guidelines are for, there is no need for such changes and making them here would require a cascading series of changes in the material following. 2. Agencies should use evidence of what happens in the real world, rather than simply making inferences basic on models and structure. (a) Looking at what actually takes place after a consummated merger is a good approach. But the agencies should also look at what happens when a merger is blocked - does the anticipated competition take place, do the potentially competing products actually get developed and launched? Much current analysis is done on the basis of economic models and hypothetical conduct. It would be good to test those models against reality, and use the results. (b) Post-merger plans are irrelevant, post merger conduct is the key. The focus should not be on what someone puts down on paper, but what they actually implement. 5. The SSNIP test seems to work in many contexts, although it is clearly a blunt instrument Business people are not accustomed to dealing with questions about how they would behave given a hypothetical input price increase in a market in which all other things were held constant. The question of the percentage increase that should be used is more a political issue, than a legal one. What we say by the number is how much of a price increase should we be prepared to tolerate as a result of the merger. The number here, and the SSNIP test itself, cannot be the only tool used. For certain products, certain customers can and will move for a very small price difference, while others may be locked in and be vulnerable to price discrimination. Again, facts are the determinative factors here, rather than the theory of the number used. 6. If you are looking at the impact on customer, it makes sense to ask early in the game where those customers are located. But except for goods with high shipping costs, or where tariff or other barriers are present, the geographic market may not be contentious in most cases. 8. The HHI is a vast improvement over the old four firm concentration ratio approach, but the current Guidelines still would lead to somewhat absurd results if they were applied according to their terms. It would be useful to have a retrospective analysis of mergers cleared and challenged, and the HHI numbers and increases in each case. Many would find it somewhat unproductive for the agency or the division to challenge a merger of two 5% players in an industry simply because there was another company with a 45% share, which is what the Guidelines suggest should happen. The Guidelines should reflect actual practice - Let's get the data, and see where they take us. 12. Large buyers, or countervailing market power, is a legitimate issue. Where the product in question is a commodity, the presence of large buyers may well mean that a given market share does not reflect actual market power. The SSNIP test would only capture this if the agency asked the "right" buyer. In the writer's experience, a case where there was one very large buyer in fact resulted in the odd situation where the seller with the largest market share in fact had the least market power - it was willing to supply the most at the lowest margin. It is definitely worth exploring what happens to small buyers if there is a large buyer. Do sellers attempt to recoup margins by selling at higher prices to smaller buyers and, if so, is this an issue for merger law? Again, an investigation of the facts will tell us far more than arguments over theories of economics. 18. It would be helpful to have more clarity on how the agencies approach merger remedies, however item (c) suggests a predetermined conclusion that is out of place here. There are shortcomings to all approaches in merger cases (horizontal and otherwise). We are trading off ease of administration against precision of result, and are constantly making judgments at various levels as to whether allowing certain conduct is likely to be benign in the long term or will undo the reason for the remedy itself. For example, the Commission has consistently limited the use of out licensing with a royalty, on grounds that have ranged from odd economics to arguments about corporate behavior that have little or no support in fact. One can refute the economics (see attached paper), but when the explanation then becomes one of what companies "would" do, ungrounded in any case or real world fact pattern there is little one can do. The result may well be a tax on mergers that is unrelated to any legitimate competitive goal. Many of the questions put seem to assume a theoretical debate. Our suggestion is that there is a database of transactions that have been cleared, and blocked. Mining that data could tell us not only what HHI and increase tests are actually being applied and what the courts have done with them, but could give us insight into how the SSNIP test could be applied and what the outcomes have been on mergers challenged and cleared. We appreciate the opportunity to submit these comments, and are prepared to discuss them with you at the New York City session if you so desire. Kent Bernard