James and Muris
Re: Initial Recommendations
Date: December 21, 2001
You asked us to review the current FTC/DOJ process for allocating antitrust investigations (the "clearance process"), and to suggest changes to improve it. In particular, you asked us to consider ways in which non-controversial allocation decisions could be made more quickly, thus leaving more time in the original 30-day stay period for substantive investigation and discussion with the parties. You also asked us to evaluate current procedures for dealing with contested allocations, and to suggest ways to improve the resolution of the relatively rare inter-agency disputes.
In carrying out this assignment, we have reviewed the current clearance process, examined the historical record, interviewed FTC and DOJ personnel involved in that process, and applied our considerable aggregate experience within both agencies in evaluating various alternatives to the current procedures used to determine which agency should have the responsibility to investigate particular matters. Our focus is not to assign blame; indeed, the limited problems we have identified with the current process have been around for many years and span many administrations. Instead, the following represent our best judgment of ways in which the current process could be made more efficient -- i.e., produce faster decisions with fewer disputes, and resolve any disputes quicker and with as little interagency conflict as possible.
I. The Current Situation
We should begin by saying that, just looking at the numbers, the current clearance process works very well in the vast majority of situations, notwithstanding a number of systemic and other impediments to effective operation. We were all impressed, in reviewing the actual statistics, at the fact that the vast majority of clearance requests are resolved quickly, efficiently and without dispute, even though there are significant differences in internal procedures at the agencies, different understandings of the process and the product divisions that drive many of the decisions, and even different terminology. It is the exception where this is not the case.
Unfortunately, in those exceptional cases, we believe the agencies too often focus on institutional prerogatives and interests, and undervalue the public interest in effective and efficient antitrust enforcement. In addition, these exceptions are highly visible inside the agencies, and frequently outside as well; the result has been unnecessarily adversarial relationships in what should be highly cooperative agencies, and in some cases significant delays in antitrust enforcement efforts and actions. As a result, these disputes have significance for both inter-agency relationships and public perceptions out of proportion to their rarity. It is for these reasons that you have asked us to recommend improvements in the current process.
The federal antitrust agencies have limited resources; every dollar saved in eliminating unnecessary administrative conflict is another dollar that can be used for substantive enforcement efforts. We understand that both agencies recognize that the clearance process can be improved, and wish to make whatever changes are appropriate to ensure that the clearance process is as efficient as possible in serving their mutual, and the public, interest in effective federal antitrust enforcement. What we are recommending here are a combination of the best practices of both agencies, and a number of refinements that we believe will make the process work better in the future than it has in the past.
II. The Goals of the Clearance Process
Consistent with the stated positions of both agencies, the objective of the clearance process ought to be assignment of a matter to the most appropriate agency as quickly as possible, thus allowing that agency to move forward in its investigation. This is especially important in HSR situations, where there are statutory time periods that must be met, but non-HSR situations also frequently require, and will always benefit from, prompt investigation. Thus, the two critical elements of the clearance process are (1) speed and (2) correct decisions. Our recommendations are primarily aimed at these objectives.
While both objectives are important, we believe they are stated in the correct order of priority. The clearance process is necessary because there are two federal antitrust enforcement agencies with largely overlapping jurisdiction. Because it would be wasteful and unfair to encourage or allow simultaneous or sequential investigations by both federal antitrust agencies of the same conduct by the same actors, it is necessary to have a system for allocating potential antitrust investigations between the two agencies. But because both federal agencies are fully competent to investigate almost all possible violations (the Department of Justice has exclusive jurisdiction over possible criminal violations, and there are a small number of other areas of exclusive jurisdiction), it is hard to imagine a clearance decision producing a seriously wrong result -- i.e., one that would materially impair effective antitrust enforcement.
Of course, there are circumstances when one agency or the other may have more experience with a particular product or industry, or for some reason more expertise with a particular kind of investigation, and thus could arguably be more efficient in conducting a specific investigation. In these circumstances, it would be preferable to have the more experienced agency conduct the investigation. There is also value in generally avoiding different agencies having responsibility for similar investigations in the same industry or involving the same parties; it would be fundamentally unfair to have meaningfully different competitive rules apply to different firms in the same industry. But given the high degree of similarity in the substantive standards applied today by the two federal agencies, and the fact that they are both constrained by the same jurisprudence from a single federal judiciary, there are unlikely in most situations to be more than marginal substantive differences that would result from the identity of the investigating agency. Thus, the main value served here is to minimize the instances where the same or similarly situated private parties have to deal with the complications of dealing with different procedures and slightly different approaches whenever possible. For all these reasons, a purely random allocation system would not be optimal, but we should recognize that in the vast majority of situations the allocation of matters between agencies is not likely to have critical substantive effects.
On the other hand, the speed of the clearance process is always important. In an HSR merger matter, with a 30 day statutory stay running from the time of the filing, every day spent in the clearance process is a day lost to the investigation. At the end of the original 30-day stay, the only agency options (in some cases, the parties could withdraw and refile the HSR notification) are to issue a Second Request, thus extending the stay, or allow the transaction to close, thus eliminating the potential for preliminary relief. It will surprise no one that, where the agency is uncertain as to the competitive effects of a transaction, it will likely opt for issuing a Second Request. Since this imposes significant burdens on the merger parties, they will in most cases have an interest in seeing as much substantive work done by the agency during that first 30-day stay as possible. This work cannot be done before a clearance decision is made. Thus, every day lost in the clearance process is a day lost to substantive investigation, and increases the potential for additional (and perhaps unnecessary) burdens on the parties to the transaction.
Speed is also important in non-HSR situations. If the matter involves a merger that was not required to file an HSR notification, the public interest is best served by being able to investigate either prior to its closing or, if already closed, prior to significant scrambling of assets. If the matter involves potentially anticompetitive conduct, the public interest is best served by being able to take any appropriate action before it harms consumers, or to stop any ongoing harm as soon as possible. Thus, the speed of the clearance process is always going to be important, and our recommendations recognize speedy allocations as the most important goal of the clearance process.
Finally, it is also desirable to reduce to the maximum extent possible any inter-agency friction that the clearance process may produce. As our examination of the clearance process shows, the current clearance process works well in the vast majority of situations today. But there are occasional disputes, and those disputes have in the past caused considerable friction and hampered inter-agency cooperation. Antitrust enforcement is too important to our national economy to waste limited resources on inter-agency disputes. So long as we have two federal antitrust agencies with largely overlapping jurisdiction, it is important for them to work cooperatively to use their combined resources for the very important work they are assigned. Thus, we have sought to come up with procedures that will reduce potential friction points throughout the process, and allow decisions to be made quickly and effectively, even when there are good faith disagreements between the agencies.
III. Preliminary Recommendations
At this point in our work, we see several different ways in which the current process could potentially be improved. Our preliminary suggestions fall into four general categories: Standardization, Criteria, Communications, and Dispute Resolution.
The clearance process should be seen as a joint effort, not as an inter-agency process. Today, each agency has its own internal procedures, and to some extent its own vocabulary, and these small differences impair efficient cooperation. There is a spectrum of options available to decrease or eliminate this problem, ranging from standardization of computer systems and programs, lexicon, procedures to the creation of a joint clearance office. The goal should be the most efficient clearance process possible, and to this end, strict and binding time periods should be adopted, recognizing that speed of resolution trumps all other values.
Today, the agencies set internal deadlines in business days, not calendar days, and these guidelines are not absolute. This is understandable; in general, weekends are not intended to be workdays, and daily circumstances change. But the statutory deadlines of HSR are set in calendar days, not business days, and they are deadlines, not guidelines. We believe the agencies must conform their internal procedures to the real world. While we recognize the complications this recommendation presents, we believe that internal deadlines should be set in calendar days, not business days, to conform to the reality of the statutory deadlines. In addition, we believe that aggressive and binding clearance deadlines must be set in order to produce an effective and rapid clearance process. While we appreciate the rigidity of these suggestions, and recognize that there may need to be some mechanism for rare exceptions (perhaps only invoked by the agency head or some other senior official so that it is reserved for truly exceptional circumstances), we believe that aggressive deadlines with real consequences are an important management tool in any effective clearance program.
We believe that the agencies should agree to a jointly staffed Clearance Office, and commit to the goal of resolving the vast majority of clearances, thus leaving the allocated agency staff free to contact third parties, by the end of the 5th calendar day following the initiation of the clearance process (defined as the time that the HSR filing is deemed to be officially made or the time that a specific clearance request is made by either agency). Internal agency procedures should be adjusted to enable this goal to be met, as outlined below.
Joint Clearance Office. The most obvious way to deal with the internal differences in procedure and vocabulary is to create a joint clearance office. The current system to some degree institutionalizes competition, not cooperation; even operating in good faith, there is some significant incentive (from peer pressure if nothing else) to advance the views and preferences of the agency one represents. We believe that this is generally overcome by the desire of all involved to do the right thing, but when it is not clear what the "right thing" is in the current environment, institutional interests can become more powerful. This process should not be competitive; how matters are allocated, from a "good government" perspective, should have little to do with what individual lawyers or even agencies would prefer. Given our view that it makes little difference in most cases which agency handles any particular matter, we believe that the most efficient way to manage the clearance process is to remove it as completely as possible from internal agency pressures.
Conceptually, this is fairly easy. A Joint Clearance Office could be established and staffed for the sole purpose of managing the clearance process. Ideally, the costs would be borne equally by both agencies, and evaluation and compensation would be based on success in efficient management of the process. At least in time, those involved would come to see their task as serving the public interest in efficient antitrust enforcement, and not in protecting or advancing the interests of a particular enforcement agency. But since we are dealing with the Federal Government, implementing this simple concept would be more difficult. The intricacies of joint funding are beyond our expertise, and the complications of joint staffing may be beyond anyone's expertise. But it is obvious from looking around that this can be done -- with temporary details and agency reimbursements, if in no other way. We defer to the experts on how this can best be accomplished.
Even if for some reason this cannot be fully accomplished in the near term, the basic goal could be met with the creation in one agency (probably, given its structure and general responsibility for the HSR process, the FTC) of a Clearance Office that was jointly staffed and charged with operating as an independent clearance management body. This is less desirable because it seems potentially less permanent, but with the proper agreements and public statements probably would accomplish the purpose for the time being while a more permanent solution was worked out. The advantages of this approach are so obvious that we find it hard to believe that it could not be accomplished with the cooperation of both agencies (and the Congress, if necessary).
Given the bureaucratic complexities, we have also considered other steps that each agency could undertake to accomplish a more efficient clearance process. The following recommendations assume continued separate staffing of this function at each agency:
1. Dedicated Staff. Assuming that the clearance functions remain in each agency, we believe that each agency should dedicate permanent staff to carry out this function. The effective operation of the clearance process is, properly viewed, a high priority of both agencies; it is their responsibility to efficiently carry out this responsibility in a way that does not unnecessarily increase burdens on third parties. It is not an administrative detail or something that can or should be done by lawyers or staff with other responsibilities, nor in our view should it be a task that is rotated on a regular basis among those with other primary responsibilities. Each agency should dedicate the necessary lawyer and administrative resources to carry out this responsibility efficiently and effectively.
2. Common Lexicon, Forms and Procedures. Today, both agencies use slightly different language, forms and procedures in carrying out the clearance process. We believe that these should be standardized. This will require either or both agencies to modify their current procedures and lexicon to some extent, but the benefits to standardization outweigh the minor cost of such changes. The lack of standardization contributes to unnecessary confusion and increases the potential for conflict. In addition, both agencies should use a common database and electronic communications process, to which both should have equivalent access at all times.
3. Commodity List. Today, there is no common agreed list of areas of primary responsibility. Indeed, the two agencies have different lists that use different terms to describe the same products or product areas, and the two lists do not conform, even taking account of the different terminology. There should be a common list, using a common lexicon, and we have attached a recommended broad commodity split to this Report. In coming to this recommendation, we have taken into account the historical focus of each agency, the importance of grouping like products and industries where possible, and the need to balance the likely workload between the agencies. Our recommendations, if accepted, would require some temporary adjustments between the agencies, and could prompt some personnel changes, but we believe these short-run costs would be far outweighed in the long run by a more rational allocation of commodities and industries between the two agencies.
As you will see from the attached, we have tried to create logical groupings, and to allocate those groupings between the agencies based on the criteria noted above. The suggested commodity list is a macro document; it would benefit greatly from continued refinement into finer categories, and there are some areas where we have not yet arrived at a recommendation. We believe that the best way to complete this important task is by this group working together with agency staff, and we are willing to move forward on this immediately if you agree with this approach. This recommended allocation will not eliminate all close calls or disputes, but it should reduce their frequency. Taken together with the other steps suggested below for resolving disputes, we believe that adoption of this commodity split would improve the efficient functioning of the clearance process and hopefully of the federal enforcement effort as a whole.
4. Procedures and Time Periods. We recommend the following, which are designed to allow matters to be cleared to one agency by the end of the 5th calendar day after the clearance process is initiated:
6. Clearance Manual. All of the above features should be incorporated into a Joint Clearance Manual setting forth the common procedures and forms to be used, the time periods relevant to each phase of the process, and the agreed divisions of commodity responsibility. The Clearance Manual should be a public document, and any subsequent revisions should be made available to the public. This process is not well understood by the public, including many in the antitrust bar, and transparency here would eliminate confusion and misunderstanding.
Notwithstanding the rationalized commodity list suggested here, there will continue to be matters that do not fit neatly within assigned categories. Today, the principal decisional criterion for disputed allocations is prior agency experience. We believe that experience should remain the touchstone for clearance resolution. Experience offers the advantages of consistent governmental enforcement, increased predictability, and some efficiency gains. Despite these advantages, criticisms can be leveled at this criterion, perhaps principally the belief that use of experience creates an incentive for each agency to fight very hard over a particular dispute out of concern that loss of the investigation will lead to loss of the class of matters for that agency. On balance, however, we do not see a better solution.
Accepting the use of prior experience as the basic criteria for resolving disputes, there are still issues of how to evaluate particular kinds of experience. Given a rationalized commodity list and effective dispute resolution mechanism, we see little need for the very arcane rankings that have developed over the years, and would suggest that the more recent experience (within at least the last five years) and the more detailed experience (actual investigations trumping all other kinds of experience) be the deciding factors. Where neither agency has recent or significant experience, the matter should be allocated on the basis of any other factors likely to lead to a more effective and efficient evaluation of the conduct at issue, with the dispute resolution system suggested here employed promptly if no mutual agreement can be reached.
The most complex issues are likely to arise where there are combinations of multi-product companies or multiple possible competitive issues presented by the specific conduct to be evaluated. We believe that horizontal effects should prevail over vertical or conglomerate effects, and that the product areas or markets in which competitive issues are to be evaluated that involve the largest dollar volume of commerce should prevail over smaller products or markets. There may well be other criteria that would be useful to adopt, and we would be glad to continue to work with staff from both agencies to identify any additional relevant criteria to be applied in those instances where the new commodity allocation leaves the agency allocation decision uncertain.
The clearance process needs to be viewed as an effort to efficiently allocate the federal government's antitrust responsibilities, not an adversarial process for resolving inter-agency disputes. For the process to work optimally, the two agencies need to have a common process that fosters cooperation and timely decisions.
This premise underlies our recommendations for reliance on joint or at a minimum dedicated clearance staff. The job descriptions and the performance appraisals of those responsible for this function should emphasize interagency coordination and prompt dispute resolution. In addition, if this function is not provided by a joint office, there should be regular -- we think daily is not too frequent -- personal communications between the staff at both agencies dedicated to this function, if only to confirm where outstanding matters stand. If there is no joint office, there should be regular -- at least weekly -- face to face meetings by dedicated agency staff to review process and ongoing matters; good personal relationships will reduce the chances for friction. There should also be regular -- probably weekly -- reports, produced jointly by the dedicated agency staff, to the agency heads and other appropriate people of the relevant statistics: how many requests submitted? how long they took to be processed? how many disputed claims? how long they took to resolve? what was the basis for the resolution? These can be easily produced if there are the appropriate common software and database programs. There should also be regular face-to-face meetings of the senior officials responsible for this process, to review results and procedures and see where changes or improvements may be necessary. Regular communications will reduce the opportunities for disagreement and friction.
Finally, we suggest regular reviews -- at least for the next year every quarter -- jointly by the agency heads, with all the relevant people at both agencies, to review the operational statistics and hear from the participants about how the process is working and how it could be improved. The evaluations of those personnel involved in the clearance process should depend on how successfully they reduce delay and inter-agency friction.
This process is, like Second Requests, one of the ways in which perceptions of the agencies are formed in the minds of those who interact with them, and thus it has importance beyond its intrinsic significance. Agencies that are perceived as efficient and sincerely interested in minimizing unnecessary burdens on third parties will find more support for additional resource requests and substantive efforts. Thus, the success of the clearance process, and as importantly the perceived success of the process, should be given a very high priority by both agency heads.
D. Dispute Resolution.
There will be a small number of matters that are not resolved through the normal agency process. As described above, these should be taken to the agency head as soon as it becomes obvious that the normal resolution process will not work. If the schedule proposed above is adopted, these will arrive on the agency heads' desks no later than the end of the 6th calendar day after the clearance process has been initiated.
The agency heads should not themselves seek to resolve the clearance dispute, but should decide only whether the matter is important enough to be submitted to the dispute resolution process. We believe that good personal relationships between the agency heads are too important to the overall federal enforcement effort to put them into positions as adversaries, and thus their role should be limited to making an agency decision that the resolution of this dispute is important enough to justify the invocation of the dispute resolution arbitration process. While we are reluctant to put absolute time deadlines on agency heads given the other important demands on their time, we suggest that decision should be made within 48 hours of submission to them. In order to ensure this can be done, whenever agency heads are unavailable they should designate another official to make this decision. A decision to not invoke the dispute resolution process by either agency head will result in clearance to the other agency.
Where both agency heads decide it is necessary, we suggest that the issue of which agency should handle the matter in question be submitted to a neutral third party for a recommended resolution, and that both agency heads commit to accepting the recommendation of that neutral third party. To avoid actual or perceived conflicts, the neutrals should be academics who are not directly engaged in representing parties before the agencies. If a panel of 3-5 such individuals were pre-selected, it is likely that at least one would be immediately available, especially given the very abbreviated process we envision, as described below. They need not be antitrust experts. We would suggest selecting the neutral for a particular matter by lot, repeating the process if for some reason that person is unavailable. When both agency heads determine that a matter should be submitted to the dispute resolution process, the neutral should be selected and the process completed within no more than 48 hours. The panel members should agree that any information they receive in this role, including the fact of the need for an arbitration, the identities of the parties and the outcome, will remain confidential to them personally, and will not be disclosed to any other person. The panel members should already be familiar with the Clearance Manual described above, and thus be prepared to hear and immediately tender a recommendation on any such disputes.
The process itself should be very lightweight and speedy. The neutral should be provided (in addition to the Clearance Manual previously provided) only with the prior written claims of both agencies produced during the clearance process, and no more than one hour should be set aside for debate and discussion, with each agency designating a single person to participate in that discussion on behalf of the agency. The neutral should tender a recommendation at the end of that discussion and that recommendation should be accepted by both agencies.
The sole basis for the neutral's recommendation should be his or her judgement as to which agency is more likely to be able to most efficiently carry out, in this particular matter, the public interest in effective antitrust investigations. Thus, the neutral should take into account the prior experience and expertise of each agency with respect to the particular industry, issues, and the parties involved, as well as any other factors deemed relevant to the public interest in efficient and effective antitrust investigations.
IV. Enforcement Mechanisms
One would hope that the good faith of the participants in this process would ensure its efficient operation. Still, even those of good faith can sometimes lose perspective or an appropriate sense of balance, and bureaucratic forces are not to be underestimated. Therefore, we recommend that the Final Report of this Group be published, along with the specific procedures adopted as a result of it (including the Clearance Manual), and that the agencies periodically jointly publish the results of the clearance efforts, including relevant statistics about how many clearance decisions were made at the various stages of the process.
We also recommend that the agreed commodity list, showing areas of primary responsibility where a particular agency will always have the investigatory lead absent very unusual circumstances, be published, so that private parties can know which agency is likely to review antitrust issues involving their particular industries.
Finally, we recommend that this or a similar group be invited to review the clearance process again following one year of experience with any recommendations adopted by the agencies, so as to permit suggestions for fine tuning or further refinements as dictated by practical experience.
Attachment -- Recommended Commodity Allocation
Recommended Commodity Allocations