The following HTML version of this document does not include footnotes and certain formatting. Click here for a printer-friendly version of this document in Adobe Acrobat.Acrobat Reader can be downloaded for free from Adobe.
STATEMENTOfALBERT A. FOEROn behalf ofTHE AMERICAN ANTITRUST INSTITUTE
Re:S. 1687 and S.1854Federal Trade Commission Reauthorization Act of 1999The Hart-Scott-Rodino Antitrust Improvements Act of 1999
SUBCOMMITTEE ON CONSUMER AFFAIRS, FOREIGN COMMERCE, AND TOURISM OF THE COMMITTEE ON COMMERCE, SCIENCE AND TRANSPORTATION UNITED STATES SENATE
These are comments of the American Antitrust Institute concerning S. 1687, the Federal Trade Commission Reauthorization Act of 1999, and S. 1854, the Hart-Scott-Rodino Antitrust Improvements Act of 1999. The American Antitrust Institute is an independent, non-profit education, research, and advocacy organization that believes the antitrust laws should play a more expansive and effective role in the national economy. I am Albert A. Foer, President of the AAI.
The Federal Trade Commission Reauthorization Act of 1999
Early last year the American Antitrust Institute published a 40-page history of antitrust funding, titled "The Federal Antitrust Commitment: Providing Resources to Meet the Challenge." Copies were made available to the appropriation committees at that time and can be found at the AAI website. While some of the numbers require updating, the argument we offered remains valid, namely that the two federal antitrust agencies, i.e., the Federal Trade Commission and the Department of Justice Antitrust Division, are substantially underfunded and understaffed, in view of the law enforcement challenges they face. In our paper, we urged Congress to contemplate a three-year expansion of federal antitrust personnel on the order of 30-50%. At the time we wrote, the FTC's budget for f.y. 1999 was $119 million. The increase in S. 1687 to an authorization of $149 million in f.y. 2001 would be 25% and to $156 million in f.y. 2002 would be 31%. Given the continuation of the unprecedented challenges that are before us, we believe that the authorizations being proposed in S. 1687 would be at the extreme very low end of what is needed. The FTC's request to increase its FTEs from 979 to 1133 in f.y. 2001 seems well-justified. This would be permitted by a funding level of $165 million.
What are the challenges that require additional federal antitrust attention? First, we have a merger wave of unprecedented size and scope, which is rapidly restructuring the American and the world economy. The number of merger filings tripled from 1,529 in fiscal year 1991 to 4,642 in fiscal year 1999. We are seeing something like a 15% increase this year. The dollar value of merger filings increased over 11 fold during this period, an indication that mergers are becoming larger and more complicated. Merger analysis is labor intensive. Much depends on careful fact-gathering and analysis. Neither the FTC nor the Justice Department is staffed adequately to deal with what is arriving each and every day. The result is a rapid movement toward more and more concentration in market after market.
Second, the promise of deregulation has not been kept. That promise was that antitrust would replace direct economic regulation as the public's protection against the abuse of market power. It only partially happened, with the result that consumers have not yet received all of the benefits from competition to which they are entitled. More needs to be done in the formerly regulated sectors of our economy and special care must be given to the deregulation of electricity, where the rules of the new game are yet to be written. The FTC has played an extraordinarily important role in helping the States adjust to electricity deregulation, but it has depended very largely on one lone economist spending only half of his time on this subject.
Third, the economy has become more of a global marketplace, changing the modes and challenges of antitrust. Even as some markets become freer and more competitive, others are ruled by international cartels. While these are more often under the Antitrust Division's beat than the FTC's, both agencies have found that globalization increases their workload. For example, antitrust analysis now regularly requires understanding the international dynamics of an industry. Discovery becomes more time consuming and complicated. Cooperation with foreign antitrust agencies - which have proliferated since 1989-takes time.
Fourth, new technologies have created new antitrust issues, such as the potential of network effects creating persistent monopolies. The FTC already plays an important role in educating the public on the consumer and competition aspects of rapidly changing technology-driven markets, but merely keeping on top of developments, not to mention developing and carrying out prudent public policies, is especially time-consuming at this critical time in economic history.
The FTC has done a remarkable job of streamlining and improving its productivity. However, today, the agency can only be described as "strapped". The FTC's budget request does not contain new programs. It does request additional staff to carry out a number of tasks Congress has imposed, such as implementing new identity theft legislation, and it requests funding for certain internal technological improvements, to make it possible, for example, to file Hart-Scott-Rodino information on -line. Mainly, the FTC seeks funding for additional staffing made necessary by the demands of the merger wave and the rapid emergence of e-commerce. We urge that the authorization be increased to better recognize these needs.
The Hart-Scott-Rodino Antitrust Improvements Act of 1999
We were also asked to comment on S. 1854, the Hart-Scott-Rodino Antitrust Improvements Act of 1999. We direct our comments to the following four questions:
(1) What reforms are appropriate with regard to "second requests"? (2) Is it appropriate to increase the thresholds for PMN reporting and, if yes, are the thresholds proposed in S. 1854 appropriate? (3) Is it appropriate to increase the fees payable for larger transactions and if yes, are the fees proposed in S. 1854 appropriate? And (4) What effect will a reformed fee structure have on antitrust enforcement?
(1) What reforms are appropriate with regard to second requests?
The Hart-Scott-Rodino Antitrust Improvements Act of 1976 brought a revolution to the antitrust laws with respect to merger enforcement. Before this landmark legislation, mergers and similar transactions were effectively, though not formally, out of the reach of the antitrust laws. It was nearly impossible to obtain an injunction that would stop a merger before it was consummated. Rather, the merger had to be attacked after the fact. This assured that litigation would be a prolonged process and that in most cases there would be no effective remedy, because, as the saying went, it's too difficult to "unscramble the eggs" once they have been cracked, stirred together and fried. H-S-R changed all this, very dramatically. Today, mergers are challenged before they are consummated. This has brought a much higher degree of certainty to the process, speeded it up, and made it possible to obtain effective remedies in those relatively few situations where the transaction would violate the antitrust laws.
The antitrust agencies have worked diligently to make the premerger notification program effective and they should be congratulated on their success. They have created an administrative system that identifies those mergers that are likely to be problematic and quickly allows the vast majority to move forward. With respect to those that may be problematic, they seek more information through a "second request". About 60% of second requests lead to successful law enforcement actions, and the second request process generally takes less than four months. These are impressive facts, given the complexity of today's mergers. It should also be mentioned that a 60% success figure is probably about as high as good public policy would warrant: if the agencies only sought more information on those situations which were more or less automatic "winners", they would not be casting their net widely enough.
S. 1854 would modify provisions of the H-S-R act with respect to second requests. Second requests would be limited to information that (a) is not unreasonably cumulative or duplicative, and (b) does not impose a cost or burden on the parties that substantially outweighs the benefits to the agencies in conducting their antitrust review. Parties would have the right to petition a U.S. magistrate to review whether a second request meets those standards. In addition, parties would have the right to petition a U.S. magistrate for a determination of substantial compliance with a second request. The waiting period after substantial compliance would be extended from 20 days to 30 days.
We would be pleased to see the waiting period extended from 20 to 30 days. This would facilitate a more careful evaluation of the materials received and would relieve some of the stress that the current timeframe imposes on everyone.
With respect to the other proposals, it is not clear to us that a significant problem exists, and we think the solution proposed is unworkable. Although you may hear horror stories about the amount of data being demanded by the agencies, these caricature the reality. First, the volume of second requests is not very large. In 1998, it covered 125 transactions, only 2.7% of transactions.
Second, a few very large, complicated, and hard-fought cases may color perceptions of what is involved. Over 85% of the FTC's second request transactions are resolved before there is substantial compliance with the request. Over 70% of the FTC's second requests result in productions of 50 document boxes or less and over 60% result in productions of under 20 boxes. The average time to complete a second request investigation is three-to-four months, and many are completed in much shorter time.
The agencies have a generally strong record of working with parties to narrow requests and accommodate to what is feasible. Given their workload and the deadlines that H-S-R imposes on them, the agencies have a strong incentive to reduce the size of document requests. The Antitrust Division has an internal appeal procedure for disputes, leading to the Deputy Assistant Attorney General. In 4 years, we understand, this process has been utilized only three times, and in two of the three situations, the Deputy sided with the merging parties. The FTC has an informal process that involves the Director or Deputy Director of the Bureau of Competition whenever there is a second request conflict between the parties and the investigation staff.
The reality seems to be that there has been only a handful of investigations requiring submission of large numbers of boxes of documents. Usually, these were multibillion dollar, multi-market transactions that transform entire industries. These are precisely the type of transactions where a detailed and probing examination is appropriate and where interconnections may not be apparent until data is obtained. When an agency issues a second request, it does not have enough information to evaluate the potential benefits of the information requested. Its information in hand is extremely limited. Unlike the parties to a transaction, the agency has no company officials engaged in the business in question to whom it can turn for information about the industry and its products, competitors and customers.
You may hear complaints that the agencies "game" the premerger rules, using second request demands to delay a transaction. The other side, too, plays games, of course, sometimes dumping tons of documents on an overworked agency staff in the hope of wearing them out and causing them to overlook a needle buried in the haystack.
Introduction of a formal right of appeal in the midst of a second request places a professionally conducted investigation in an untenable chicken-and-egg conundrum: the enforcement officials cannot necessarily determine the focus of the investigation until they have examined the data; if they are precluded from the data on the basis that they haven't identified their theories clearly enough, they will never obtain the data needed for a professional evaluation of the various possible theories.
It is also important to understand that a change in the rules will affect the strategic dynamics of merger enforcement. Many cases are efficiently settled because the merging parties know that they have documents that will reveal the anticompetitive nature of their transaction and that the data will have to be provided to the investigators. By increasing the probability that they will not have to provide data (maybe they can persuade the magistrate that it is not necessary), we would be reducing the probability of settlement in these cases.
As we see it, the intervention of a review by a magistrate sounds better on paper than it is likely to be in practice. Indeed, we expect it would slow down the merger review process and make it more expensive for all concerned, with the parties spending a lot of time sparring in and out of court. Magistrates tend to be busy and there is no guarantee that their intervention will be swift. In the worst scenario, if the magistrate happens to be unsympathetic to the antitrust laws or unfamiliar with how an antitrust case is put together, he or she could represent an open invitation for challenges that would have the effect of making it more difficult to obtain important evidence and properly evaluate the case. Moreover, a magistrate is used to participating in civil cases where the plaintiff has had more opportunity to develop the facts of the case; here, the government has typically had to rely on the initial filings and a few telephone calls. Its second request must necessarily be broad, often broader than what would be acceptable to a magistrate who is used to reviewing discovery in private civil litigation. This, it must be stressed, is a pre-discovery situation.
We are reluctant to see a generally satisfactory administrative process be revised by legislation, which can be inflexible and difficult to adjust at a later time. The agencies have demonstrated a willingness to cooperate with the private bar in streamlining the process. We believe there is a better way to deal with whatever second request problem exists: the enforcement agencies should be encouraged by this Committee to build on their current internal processes, perhaps creating a more formally defined route for appeal within the structure. Decisions in the relatively few situations that will occur can best be made by people who understand how an antitrust case is put together and can decide quickly. Importantly, the top officials of the agency (perhaps, in the case of the FTC, the General Counsel or a designated Commissioner) would have an appropriate background for understanding the dynamics of an investigation and an incentive not to waste resources. More to the point, however, is that we are talking about an investigation of a law violation, not a judicial determination of responsibility. It is premature at this stage to insert the judiciary into the process.
Some may argue that the enforcement agency does not need so much data, because they are only preparing for a motion for a preliminary injunction and more complete data can be obtained during discovery for trial. In the context of modern merger practice, this is misleading. It is almost always the case today that the decision of the Federal District Court in a preliminary injunction determines whether or not the transaction will go forward; in effect, the preliminary injunction hearing is the trial and the parties need all their ammunition for that one hearing. The public gets one bite at the apple and it should be a full bite.
(2) Is it appropriate to increase the thresholds for PMN reporting, and, if yes, are the thresholds proposed in S. 1854 appropriate?
The thresholds for reporting a planned merger were created in the original Hart-Scott-Rodino law, dating back to 1976. They have not been modified to reflect inflation. If the initial thresholds were in some sense "correct," then it is not unreasonable to modify them upwards one time and thereafter to index them for future inflation. On the other hand, the original thresholds were arbitrary then and remain arbitrary now.
The important question is how many anticompetitive mergers will escape federal attention if the threshold is lifted? How do the negative effects of this compare to the positive effect of freeing a substantial number of transactions from the reporting requirement?
The fact that a merger is relatively small does not necessarily mean that it will not have an adverse impact on competition, if the market itself is small. This is frequently the case in industries that serve a primarily local market, such as radio broadcasting or certain types of medical care. Moreover, there are many situations where the dollar size of the transaction is unrelated to its importance. For example, in the high tech industries, it is not unusual for a firm to have little in the way of revenue, but to own intellectual property that can create a new market or bring a big change to an existing market. To exempt small transactions from reporting means that the parties can go forward and consummate their deal without antitrust interference. If the government later determines that the transaction was anticompetitive, the only remedy is to try to "unscramble the eggs." The cost to the federal government of pursuing relatively small mergers after the fact is likely to prove to be too high, with too small a return. Therefore, the practical effect of raising the limit for reporting will also be to provide a de facto safe harbor from federal intervention for mergers that fall under the reporting threshold.
How many transactions will be affected? S. 1854 would raise the filing threshold from $15 million to $35 million. The impact of S. 1854 would appear to be, roughly, the elimination of 37% of the currently reported transactions. In absolute numbers, this would be approximately 1,700 transactions.
In fy 1998, there were a total of 17 second requests by the two agencies in the <$35 million range This is only 1% of the reported transactions under $35 million, and .4% of all transactions reported. Over 60% of the FTC's second requests in the 1999 fiscal year resulted in successful enforcement actions . It is likely that 7-8 mergers that would have been stopped or modified would fall through the net if the threshold is lifted to $35 million.
Changes in government policies and procedures often have unanticipated consequences as parties "re-game" the system in light of the new rules. If we raise the threshold to permit 7-8 anticompetitive mergers to pass through the antitrust net, we are also sending a signal to others in the small-size category that they need no longer worry about antitrust. Their legal advisors will be able to counsel them that horizontal mergers previously unthinkable can now be undertaken with minimal concern of antitrust interference. A probable effect of the change in threshold, therefore, would be the triggering of more anticompetitive mergers in local and regional markets.
There is no commonly accepted methodology for converting these considerations into a consumer welfare loss. There is also no way to predict whether these mergers, most probably local or regional in effect, would be challenged (most likely, after the fact) by state antitrust authorities.
To the extent that elimination of 37% of the premerger filings frees up government staff, this should not be viewed as a savings; rather it would release resources that could be shifted from relatively quick reviews of small transactions to work on larger transactions, of which there is a plentiful supply. During the current merger wave, which has utterly stretched the resources of the two antitrust agencies, this reallocation might slightly relieve a little of the need for more personnel, but it should not result in noticeable budgetary savings.
We draw the following conclusions: (1) Raising the threshold to $35 million will allow approximately 37% of currently reported mergers to disappear from the antitrust radar screen. (2) At current merger rates, this will mean that something on the order of 17 transactions that would have merited further investigation will be left unattended and something on the order of 7-8 transactions that would have triggered successful enforcement actions will be allowed through the net. (3) These numbers need to be adjusted, on the one hand, for possible actions by State law enforcement agencies, and, on the other hand, for the de facto permission that lack of enforcement will give companies to engage in a larger but unpredictable number of anticompetitive mergers having a primarily local or regional impact. (4) Merging companies will clearly save over $100 million in filing fees and attorney fees, but consumers will pay many millions of dollars a year in additional monopoly rents. Depending on the magnitude of price increases that are the result of anticompetitive mergers and assumptions made with respect to point (3) above, the loss to consumers could be either less than or in excess of the savings for companies.
Because of the uncertainties in the tradeoff, we would recommend an alternative way of attacking the issue. Instead of giving a de facto waiver to all of these mergers, substantially reduce their filing fee and direct the enforcement agencies to reduce the initial reporting burden on small merging companies.
(3) Is it appropriate to increase the fees payable for larger transactions and, if yes, are the fees proposed in S. 1854 appropriate?
The fee paid for filing a premerger notice is a user fee and should in some rough way be apportioned to the costs imposed on the government. In a general way, it is likely that the larger transactions require more of the government's time and expenses, because larger transactions tend to be more complicated, involve more product lines and more geographic markets, thus more data and analysis. It is therefore appropriate to increase the fees for larger transactions.
S. 1854 increases the fee from $45,000 to $100,000 for transactions >$100 million. In 1998, this would have involved 1,326 transactions. We endorse this and further suggest that there should also be a megamerger category for the 193 (4.2%) of transactions above $1 billion. These are the most complicated and most costly to investigate, on average, and we urge that the user fee for these be considerably higher, e.g., $500,000.
(4) What effect will a reformed fee structure have on antitrust enforcement?
The federal antitrust effort has become dependent on H-S-R filing fees. By law, fees collected by the agencies in conjunction with the receipt of premerger notifications are for the exclusive use of the two antitrust enforcement agencies. The funds are split evenly between the Antitrust Division and the FTC, although the FTC uses roughly half of its funding for consumer protection.
In recent federal budgets, Congress has tied the agencies' funding to the premerger notification income. Again in 2000, 100% of the antitrust budget will come from filing fees. We are not persuaded that it is in the public's long-term interest for antitrust enforcement to be tied to the level of merger activity in this way. For the moment, it provides some stability in funding, but of course this can change if the merger wave slows down - or if Congress changes the formulation for income. When an agency "earns" its funding as a result of a certain type of activity, it is subject to a skewing of its activities in favor of that activity. Moreover, in the case of antitrust, the merger wave, which drives mandatory deadlines under the HSR law, has forced the agencies to focus on mergers, to the exclusion of many other types of situations that might better deserve their focus. Roughly speaking, the two antitrust agencies now must spend 75% of their resources on mergers, up from about 33% not so many years ago. This implies that other types of antitrust concerns are probably getting less attention than they should.
However, starting from the current situation, it is important that any modifications to the PMN filing fees should preserve the current anticipated level of funding. We have seen that the fees which will be lost at the <$35 million end of the spectrum would have amounted to $76.5 million in 1998. The new fees that would be generated at the >$100 million end would be $55,000 (the new fee of $100,000 minus the old fee of $45,000) times 1,326 transactions, or $72.9 million. This amounts to an appropriations reduction of approximately $3.6 million, and is therefore not revenue neutral.
Thank you for this opportunity to present the views of the American Antitrust Institute.