************************************************************************ ** NOTICE ********************************* *********************** This document was converted from WordPerfect to ASCII Text format. Content from the original version of the document such as headers, footers, footnotes, endnotes, graphics, and page numbers will not show up in this text version. All text attributes such as bold, italic, underlining, etc. from the original document will not show up in this text version. Features of the original document layout such as columns, tables, line and letter spacing, pagination, and margins will not be preserved in the text version. If you need the complete document, download the WordPerfect version or Adobe Acrobat version, if available. ********************************* ******************************** October 8, 1999 STATEMENT OF COMMISSIONER MICHAEL K. POWELL, CONCURRING IN PART AND DISSENTING IN PART Re: Memorandum Opinion and Order, Applications of Ameritech Corp., Transferor, and SBC Communications Inc., Transferee, For Consent to Transfer Control of Corporations Holding Commission Licenses and Lines Pursuant to Section 214 and Section 310(d) of the Communications Act and Parts 5, 22, 24, 25, 63, 90, 95, and 101 of the Commission's Rules (CC Docket No. 98-141) Although I support approval of this merger between SBC and Ameritech, I respectfully disagree with the manner in which the draft weighs the transaction's potential harms and benefits. Specifically, I find fault with the underlying public interest standard, and its application in this proceeding sharpens my concerns with its pitfalls. The approach of rounding up "voluntary" conditions to compensate for largely unrelated potential harms is fraught with public policy problems. Even under this defective standard, however, I believe this Order, though carefully written, misses the mark. Although I concur in the conclusion that there are public harms that might well result from this combination that are not entirely offset by the applicants' asserted benefits, I am unsatisfied that any one of these harms bears the weight assigned to it in this Order. Thus, I believe fewer conditions, tailored to address the specifically identified harms, would have been the correct result. I. The Formulation of the Public Interest Standard as an Unconstrained Balancing Test is Both Substantively and Procedurally Flawed Consistent with my long-standing concerns regarding our license transfer process, I have fundamental difficulties with the public interest standard as developed and applied in this Order. Simply put, I am very uncomfortable with a standard that places harms on one side of a scale and then collects and places any hodgepodge of conditions no matter how ill-suited to remedying the identified infirmities on the other side of the scale. This balancing approach leads to a number of problems: First, the approach creates a great temptation to load up the benefits side of the scale with a big wish list of conditions that are non-germane to the merger's harmful effects. Second, the approach makes it easier for identified harms, even significant ones, to be visited upon the public in exchange for other benefits. Third, the conditions that are sought are more often surrogates for policies and rules of general, rather than merger-specific, applicability, but without the extensive deliberative process and the check of judicial review normally afforded a rulemaking. And fourth, the process of obtaining "voluntary" conditions inevitably involves bilateral negotiations with the parties that leave the integrity of the Commission's process vulnerable to criticism. I consider each of these in more detail below. A. The Problem of the Mountain and the Pebble To conceptualize the problems with the public interest standard when reviewing a license transfer (i.e., a merger), consider a simple balancing scale of the "see-saw" variety. On the left side of the scale are public interest harms and, on the right, public interest benefits. The balancing approach used in this Order simply requires that the benefits outweigh the harms. If the harms weigh but an ounce more than the proposed benefits, the standard (if faithfully applied as articulated) would require us to block the merger. This approach is troubling on one level, for if the government were neutral with respect to the asserted benefits, it still could be compelled to stop a merger based on essentially negligible harms. This balancing approach becomes even more disconcerting, when the harms we identify require extensive speculation and hypothesis about predicted behavior (as is the case with the "the precluded competitor doctrine") rather than detriments that are more concrete, or at least more time-tested. This is so, for the margin of error in our theories alone may encompass the putative harm, yet we might block the merger nonetheless if unimpressed with the proffered benefits. I believe this describes the case at bar. The more serious problem arises with the public interest "scale," however, when the Commission, rather than weighing the harms against the proffered benefits, attempts to tip the balance by adding weight to the benefits "platter" with conditions a mountain of goodies designed to leave us, on balance, fat and happy. The public interest standard, as the Commission applies it, does not require that the conditions cure or remedy the identified harms. The conditions need only outweigh the harms. And, the standard does not place any limit on how much heavier than the harms the conditions are. Thus, the Commission is free to compensate for a pebble of harm on one side of the public interest scale by throwing a mountain of purportedly beneficial conditions on the other side of the scale, as we have done here. In other words, when conditions are not calibrated to remedy harms, there is no constraint on how voluminous or unrelated they might be. The consequence of this approach is that the slightest harm opens up a quarry of "would-be- nice-to-haves" that can be piled on the scale. Moreover, the coercive effect of having the applicants over a barrel hoping to gain merger approval dramatically improves the chances that the companies will "agree" to abide by the conditions. Thus, the temptation and the enticement to stack the scale with precious gems is irresistible to competing companies and the Commission itself. B. "Poor Joshua!" The second difficulty I have with the Commission's merger standard is that in theory, it will allow a merger to go forward that it finds will harm the public, as long as the public gets something good in return. In the humorous extreme, one could analogize this to allowing a stranger to beat your dog as long as he commits to giving the dog a bone and some fun squeaky toys. No doubt, the "ol' boy" has been quite anxious to get a bone and add to his saliva-laden collection of playmates, but not at the expense of a beating. Of course, this analogy is perhaps less humorous if one assumes that the public interest is entitled to better treatment than your dog. Jests aside, the point is that when merger conditions are not designed to remedy harms, all the unrelated benefits in the world will not cure the loss to the public. I do not claim that in this Order we have allowed unconscionable harm in exchange for some goodies. I happen to believe the harms in general are overstated. Yet if one is convinced of the significance of a proposed merger's harms (as might be the case in future reviews), it is unsettling that the merger would proceed without significantly mitigating those harmful effects with remedial conditions. In this case, for example, nothing in the conditions we impose here reverses the fact that Ameritech will be lost as an independent benchmark for comparative analyses. If this loss is truly significant and harmful, one could argue that we have abdicated our duty to safeguard the public interest by approving the transfer and thereby allowing that harm to befall the public. C. Wither Thoughtful Deliberation? The exercise we have just endured in this proceeding has resulted in another development I believe is unwise and may wither our ability to deliberate on these issues thoughtfully. I think it a profound mistake to use license transfer proceedings as a way to advance policies of general applicability that are otherwise, and more appropriately, the subject of rulemakings. My reasons are three. First, no matter how much we try to include other parties, a merger review is primarily an intimate, bilateral dance between the government and the applicants. The nature of this dance is one of negotiation. If a merger is overwhelmingly anticompetive, there is nothing to discuss. If unquestionably positive, the same. But where there are some harms and the question is finding a set of conditions that will allow the merger to proceed, which is more often the case, the tango proceeds until there is a meeting of the minds between the government and its suitor. The parties inevitably go back and forth in an effort to find a compromise where the government gets a satisfactory list of conditions, but not so many that the applicants walk away from the deal. This is precisely what occurred with SBC and Ameritech. The Commission extracted these conditions during protracted negotiations with staff under the cloud that the merger would be rejected absent sufficient conditions. Further, the conditions were revised repeatedly based on staff input regarding what size and scope of commitments would, in staff's view, warrant approval of the merger. I am of the view that this inherently bilateral process does not and cannot take sufficient account of the issues that might be raised by third-parties. Moreover, I believe that the review process is not sufficiently broad in scope to consider and develop adequately the components of a policy or rule with far reaching implications. The understandable focus of such a process is to get through the merger, not the more comprehensive development of policy. Second, by importing parts of rulemakings and transforming them into merger conditions, we risk substantially confusing both the industry and state commissions with respect to rules previously adopted. For example, one of the conditions on the instant merger requires that the applicants provide certain unbundled network elements (UNEs) at a discount that may run well below forward-looking cost. Yet the Commission has stated repeatedly that pricing elements at forward-looking costs sends the correct, economically- efficient signals to entrants deciding whether to use the incumbent's facilities or build their own. This raises the question whether it can truly be considered a benefit to condition the merger with UNE prices that are, by definition, inconsistent with economic efficiency. The conditions here overlap significantly with many of our ongoing proceedings to implement the Telecommunications Act of 1996, including our advanced services proceedings, the UNE remand, and Bell Company applications to provide long distance service pursuant to section 271. In tackling these other proceedings, the Commission must consider more than the interests of SBC and Ameritech. Relying on conditions that overlap with more general proceedings will require us, as well as state regulators and courts interpreting our decisions, to distinguish carefully this conditioning exercise from our broader duties under the Act. Although I am reasonably confident that statements in the Order will make it easier for other decision makers to limit the conditions we adopt here to this merger as a legal matter, I doubt that such statements will prevent competitors' advocates from having to do the distinguishing. In this regard, I note that, in ex parte meetings and filings, several competitors have expressed a preference for approval without conditions, rather than risk the possibility that these conditions will be used to contravene prior interpretations of our rules, or prejudge decisions in future proceedings. They take this position, notwithstanding the conditions' purported benefits to entrants. Third, I personally am uncomfortable essentially promulgating rules without the deliberative process of notice and comment normally afforded in a comprehensive rulemaking. Moreover, I think it unacceptable to pursue matters as conditions where they are insulated from judicial review. In a classic rulemaking, parties have the right to petition for review in court. But when a merger is approved with conditions, the applicants are unlikely to pursue a challenge to terms that regulators will claim they acceded to "voluntarily" as the price for gaining favorable approval. D. Unquestioned Integrity Finally, the process we have followed in this review has exposed probably unavoidably the Commission's integrity to criticism. The effectiveness of a quasi- legislative and quasi-judicial institution like ours depends on its processes being above reproach. I believe sincerely that the Bureau and the Commission have conducted themselves ethically throughout the review. Nonetheless, the nature of pursuing negotiated conditions is that there will always be charges that the applicants obtained some advantage through the bilateral mechanism of give and take. We hear this refrain now ringing in our ears. The risk to our credibility is particularly acute where, as we have now done, we pursue conditions that do not go simply to the harms occasioned by the merger, but reach further into the rights and concerns of other parties. In sum, I would say this: I do not doubt that the unique process employed in this review was initiated with the best intentions and with the highest regard for due process for all those involved. Although this process may have come up short in some respects, I do believe it was conducted with integrity and a deep commitment to fairness. Nonetheless, I would not like to see it repeated. In particular, I do not subscribe to an essential assumption of this process, that is, the idea that a regulated entity can "voluntarily" offer and commit to broad-ranging legal obligations and penalties. There is never anything voluntary about the regulatory relationship. And, even if there were, I do not believe that the guiding structures of the regulatory process (either rulemaking or adjudication) should be supplanted by a unilateral offer from a license transfer applicant. II. The Order Overstates the Gravity of the Identified Harms That Could Result From This Merger and Thus Invites Too Many Conditions Even assuming that it would be prudent for our public interest merger review to continue relying on the faulty balancing approach employed in this Order, I would still feel obliged to point out the many deficiencies and inconsistencies in the analysis. In sum, upon careful reflection on the various potential harms that the Order asserts will result from the merger, it is clear that we have made too much of these harms and thereby imposed far too many compensatory conditions. A. Precluded Competitor Analysis As the first potential harm, the Order asserts that allowing SBC and Ameritech to merge will prevent them, to some appreciable extent, from entering each others' local markets. Under traditional horizontal merger analysis, however, there would be few grounds to disapprove the present combination. Although both applicants enjoy market power in their own regions, they do not compete with each other and thus their merger would not result in a significantly greater accretion of power or result in substantial anticompetitive effects. Nevertheless, there is a doctrine in antitrust jurisprudence known as the "actual potential competition" doctrine that conceivably could have some applicability. In short, the doctrine holds that even if merging parties are not present competitors, the government may nonetheless seek to block the merger, if it can demonstrate that the two parties intended to compete with each other, would have done so absent the combination, and the effect is a significant concentration of market power. Key to this analysis is that the government must show by a preponderance of the evidence that the parties would in fact have entered. Prevailing on this theory is rare, and it has not had sufficient viability for the Department of Justice (DOJ) to block previous Bell Operating Company (BOC) combinations. Beginning with Bell Atlantic/NYNEX, the Commission began employing a variant on the potential competitor doctrine that it called the "precluded competitor" doctrine, also known as "transitional markets analysis." In truth, it is substantially similar to the potential competition doctrine in antitrust. It does have a twist, however. It takes cognizance of the fact that potential competitors have been precluded by law from entering each other's markets. And, I suppose, because firms have been prohibited from entering certain markets, the suggestion is that one will find little evidence (as would be required in antitrust) of plans to compete against each other in those markets. Moreover, as the yoke of regulation is removed, the theory presumes that companies previously precluded will enter each others' markets. In essence, by suggesting that legal restrictions have dissuaded firms from making plans to enter new markets, the "precluded competitor" doctrine lightens the government's burden to show plans of actual competition. Perhaps recognizing the thinness of the precluded competitor argument, the Order seems to concede toward the end of its analysis that "[n]either firm was likely to enter most of the other's territory," and that "[w]ere the loss of each firm's entry into the other's territory the only public interest harm produced by this merger, the overall balance would be much closer." I agree that the potential competition analysis in this Order is thin. Assuming some margin of error for the various speculations regarding how this merger will reduce potential competition, I would submit that the risk of harms associated with any such reduction is essentially negligible. Below, I address the particular weaknesses I see with the doctrine's applicability to this case and more generally. 1. "Precluded Competitor" Analysis May Have Outlived Its Usefulness Even if one were to accept the validity of this precluded competitor/transitional markets analysis, one must recognize that this framework must itself be transitional. In particular, this framework does not provide a basis for indefinitely presuming a company will enter another market, if that company remains free to deploy its resources elsewhere. Once the preclusion has been removed from the law, the clock is ticking on the doctrine and its presumption because there is no longer any legal barrier to firms making plans to enter that market. Thus, the further in time one gets away from the legal prohibition, the higher the government's burden to demonstrate that companies chose not to enter a market because of a legal restriction, rather because of unrelated factors (e.g., difficulty of entry, revenue potential). In this case, the application of the precluded competitor doctrine is, at best, strained. It has been nearly four years since the passage of the 1996 Act. For this entire period, SBC and Ameritech have not been precluded from entering each others' local markets, nor have other previously precluded competitors. Yet we see very little evidence of the BOCs, GTE or other incumbent local companies choosing to enter new local markets. There is evidence to suggest that the vast majority of local entry has come from new competitive local exchange carriers (CLECs) that have been spawned by the Act, as well as interexchange carriers, rather than from incumbents hoping to compete broadly for end users outside their regions. At the same time, BOCs have not been deterred from making plans to enter other markets from which they had been precluded before the Act, such as long distance. Given the striking absence of evidence that BOCs are poised or planning to displace smaller CLECs as the most active local entrants outside the BOCs' respective territories, I submit that the inference embodied in the precluded competitor doctrine is now too hard for us to make. At some point, the presumption that local incumbents will enter other incumbent's markets, but for preclusion, is undermined by the utter lack of evidence supporting that presumption. In this Order, the Commission seems to have reached that point. Of course, the Order enjoys the benefit of some actual evidence that SBC and Ameritech had planned to compete in some fashion in two cities, St. Louis and Chicago, and that they will no longer do so as a consequence of the merger. The Order goes on, however, to find that SBC and Ameritech are precluded competitors throughout each other's regions, not based on evidence but based on little more than a belief that the parties have both the incentive and ability to be competitors. Perhaps this is true, but it is important to point out that the Order does not fully explore factors that might affect the companies' incentives, or the true extent of their ability, with respect to entering new markets to provide wireline local exchange service. 2. The Order Undercounts the Number of Significant Potential Competitors The Order finds that SBC and Ameritech are precluded competitors with respect to the mass market, but not with respect to the large business markets. With respect to the former, it asserts that SBC and Ameritech are among only a small handful of firms with the ability to compete against the incumbent local monopolist. Apparently, these four include the three major interexchange carriers (i.e., MCI/WorldCom, AT&T and Sprint) and GTE. I believe limiting the list of significant competitors to these few firms gravely underrepresents the source of potential competition and is inconsistent with the statute's aspirations and our own local competition policy judgments. Perhaps the most glaring absence among the count are the other BOCs (i.e., Bell Atlantic, Bell South and U.S. West). The Order dedicates a great deal of ink to explaining the unique virtues of BOCs as competitors in the local mass market, including their size and their experience. This discussion is intended to distinguish SBC and Ameritech as singularly important potential competitors. Yet most of their strengths could just as easily be attributed to the other BOCs. Surely, based on the reasoning in the Order, the unique local experience of the other BOCs makes them at least as viable as the major long distance companies, if not more so. It is true the other BOCs may be hampered by the fact they are not able to link in-region and out-of-region markets through interLATA service prior to obtaining section 271 approval, but this also would be true of SBC and Ameritech themselves, were we not allowing them to merge. And this situation will change over time as BOCs obtain relief under section 271. Further, although I agree that the advantages of adjacency enhance SBC and Ameritech as competitors to each other as compared to other BOCs, I would note that the applicants are not (as the Order acknowledges) contiguous in most markets. Other BOCs currently serve areas adjacent to the applicants' respective territories. In addition, the territories of both BellSouth and US West each abut the territories of both SBC and Ameritech. I also believe it error to dismiss entirely the competitive significance of CLECs. Indeed, when one looks at a particular local market (say a certain city) one can often find CLECs providing significant offerings in the market. Indeed the Act itself and our policies envision that CLECs will be critical competitive entrants. We extol their virtues in our local competition reports and we regularly take actions to facilitate their entry and viability. In this Order, however, we have chosen to express a preference for grand competitors that can seemingly enter whole regions of the nation in one fell swoop and compete. Certainly, the IXCs have some capacity for entering in this manner, but on a market-by-market basis there are many markets in which CLECs are a more substantial competitive presence than the IXCs. Yet, they do not count in our analysis of most significant competitors. Were we to count the full panoply of other potential competitors, many of which are in reality as significant as those we dub "significant" here, I believe the force of the Order's assessment of harms would be substantially weakened. 3. The Order Fails to Explain Adequately Why Any Reduction in Potential Competition Here is Significant Even if one were to accept that this merger would reduce the number of potential competitors by one, it never explains why that is one too many. Surely, we must show why the specific number of firms that remain in the market is insufficient to nurture the benefits of competition. Even using the number of potential competitors the Order is willing to count, why is four remaining potential competitors insufficient? We are left to wonder. This points out the mushy, speculative nature of precluded competitor analysis. Because we never examine rigorously whether a particular competitor is likely to enter the market, it is difficult to make a compelling case that the merger unacceptably restricts competition by eliminating that competitor as an independent entity. A few years back, the Commission could have adopted an across-the-board policy of maximizing potential competition by preserving as many potential major incumbent LECs as possible. Similarly, Congress could have prohibited such combinations in the Act, but it did not. But given that the Commission has already allowed some consolidation among these companies, we must be able to draw a defensible line between permissible mergers among these companies and impermissible ones. This the Order has simply failed to do. B. Benchmarking The second public interest harm identified in the Order is that occasioned by the loss of one additional data point for comparing practices among "major" incumbent LECs. The theory is that the presence of more, rather than fewer, companies to compare increases the possibility that a maverick will confirm through its actions the viability of a course that regulators prefer over options advocated by other carriers. Additionally, the Order maintains that a number of regulatory devices depend on averaging the practices of companies and that the loss of Ameritech compromises the regulator's ability to do so. At the outset, it is difficult to surmount the discomfort one feels in asserting the importance of preserving our ability to regulate as a basis for blocking a merger. Nonetheless, in fairness, the Act does take a rather regulatory course to reach de- regulation, and the concern about maintaining the tools to do so is not completely far- fetched. I accept that something will be lost with Ameritech's departure. When I examine the public interest harm of this loss, however, I give it substantially less weight than does the Order, as I explain more fully below. 1. Benchmarking, While Useful, Has Not Been and Will Not Be as Essential to Regulation as the Order Suggests To begin, I must say that I am a bit dubious that benchmarking is as much a cornerstone of our regulatory edifice as the Order claims. Certainly, forms of comparative analysis have been and will continue to be used and the Order, as one would expect, has beaten the bushes to produce a few examples. But even given these examples, I believe the Order overstates both the importance of benchmarking to past regulatory efforts and the indispensability of benchmarking to future regulation. With respect to past efforts, for example, although the Commission has no doubt noted variations among carriers in the course of its proceedings, it is simply not true that this Commission has relied heavily or regularly on benchmarking per se (i.e., "average" or "best practices" benchmarking) to carry out its duties in a way that would be seriously threatened by this merger. With respect to future benchmarking, the Order exaggerates the impact of this merger by categorically rejecting the possibility of meaningful benchmarking between the major incumbent LECs and all other LECs. This rejection is based on general differences between these two groups, such as the size and scope of their networks. The Order fails, however, to explain why these general differences matter for purposes of comparing specific practices. For example, to the extent large and small incumbent LECs use the same switch technology, the fact that a small ILEC can perform a switch functionality should be probative of whether a large ILEC could perform the same function. But rather than examine the extent to which indisputable differences between large and small incumbents may or may not preclude benchmarking of specific practices, the Order pronounces, in essence, that large and small incumbents are just "too different" to provide meaningful benchmarks for each other. This sweeping assertion that large and small incumbents are different for all important regulatory purposes seems in tension with the statutory framework, which imposes many of the same requirements on all incumbents, regardless of size. I also find it telling that virtually all of the examples of past Commission benchmarking cited in the Order relate to efforts to impose requirements on all incumbent LECs, both large and small. If the Order were correct in suggesting that there are few, if any, meaningful comparisons between large and small incumbent LECs, we should be faulted for employing a "one size fits all" approach with respect to the "all incumbents" rules adopted on the basis of such past benchmarking. Further, I believe the Order waxes a bit too plaintive to the extent it suggests that regulators will be analytically powerless to compensate for a reduction in the number of benchmark firms in calculating averages. This seems to sell regulators' statistical abilities short. If regulators became concerned that taking a simple average of the industry would allow a large carrier to skew the results, for example, they could compensate for such concentration among benchmark firms through some sort of weighted average that would lessen the impact of the large firm's practices in calculating the average. Yes, it is useful when we can find support from one BOC for a practice that competitors or we regulators favor, but to which other BOCs object. But in the time I have served on and observed the Commission, I have not witnessed the consistent use of benchmarking at the level of near mathematical precision suggested. In any event, although I seriously question the claim that we systematically rely on BOC benchmarking to develop public policy to the degree implied in the Order, I will not dwell over-long on this claim, for I find shortcomings that are more significant in the benchmarking analysis. 2. The Risks Asserted With Respect to Benchmarking Lack Empirical Support and a Principled Basis for Distinguishing Permissible Mergers From Impermissible Ones Overall, I would submit that the Order's treatment of comparative practices analysis propounds an impressive theory but fails to draw an adequate connection between that theory and reality. The benchmarking analysis is too general and speculative, offering little or no data to support the theory that this particular merger will significantly frustrate the Commission's efforts to promote competition. For example, in support of the conclusion that anticompetitive practices will metastasize from one applicant to the other post-merger, the Order cites examples that are not even clearly anticompetitive. Further, I am not persuaded that the Commission has explained adequately where it draws a line between a sufficient number of major LEC benchmarks and an insufficient number. As with respect to the precluded competition analysis, I find this aspect of the benchmarking section analytically unsatisfying. The Order makes a positive case for the value of comparative analyses, and generally explains that the loss of a data point marginally diminishes the value of such analyses. The Order fails, however, to specify the breaking point at which our ability to regulate is unacceptably compromised. The Order provides virtually no guidance regarding how many benchmarks regulators need to conduct effective comparative analyses. Indeed, the Commission approved prior major LEC mergers that caused the loss of a benchmark, but that infirmity did not prompt us to block the merger. Why was the reduction of major incumbent LEC benchmarks from seven to six acceptable, whereas the reduction from six to five is not? There may be an answer, but the Order provides none. This point is not trifling. It is always possible to identify an incremental loss to the public when one competitor (in this case, a potential competitor) is lost through consolidation. But where the government is prepared to accept a welfare loss by permitting some firms to merge, and not others, it must identify a defensible benchmark of its own to rationalize the choices that it makes and to signal clearly the limits of consolidation to parties that may be considering a merger. Sadly, all that one is left with after reading the Order's benchmarking analysis (and, indeed, its discrimination analysis) is the sense that, for some reason, the Bell Companies and perhaps GTE are on the "too large to merge" side of the dividing line between permissible and impermissible mergers. If this was supposed to be the moral of the benchmarking and discrimination stories in this Order, I would have preferred to relay that moral more directly, rather than through these theoretical constructs. 3. The Order Affords Insufficient Weight to Regulatory and Other Incentives and Costs that Will Discourage the Applicants from Colluding to Resist Local Competition The Order raises the fear that going from six major LECs to five will raise the possibility of coordinated behavior with respect to price and other aspects of the product, thereby threatening the variety of practices among which regulators can find procompetitive benchmarks. Undoubtedly, the fewer competitors there are, the easier it is to collude. There are a number of factors present in the telecommunications context, however, that cause me to question whether this merger will result in enough unity of interest for the collusion danger to be significant. In particular, even if one were to concede that the Order demonstrates that the applicants will have at least a marginally enhanced ability to collude, the Order fails to demonstrate persuasively that the applicants will have significantly greater incentives to collude. The Order fails in this task by presenting an oversimplified assessment of the likelihood that the applicants can and will coordinate their behavior to avoid opening local markets or otherwise produce public interest harms. As an initial matter, I would note that many of the usual characteristics of collusive markets are absent in the context of major incumbent LEC benchmarking. For one, major LECs generally are not direct competitors. Normally, one would expect to find collusion where competitors have more to gain by sharing "the pie" than fighting over it. But here each major incumbent LEC's local service profits are not depressed by competition from other major LECs. Indeed, in the context of major incumbent LEC benchmarking, it seems unlikely that each LEC will be able to be reasonably certain that tacitly "agreeing" with other LECs to engage in a particular adverse practice will prevent regulators from identifying and requiring another practice that would result in lower profits for the incumbent. Thus, there is little direct economic incentive for major incumbents to collude for purposes of frustrating benchmarking, as the competitive consequences of doing so are largely unknown. Given that firms' self-interest generally will lead them to be more responsive to direct economic incentives than to indirect and uncertain ones, the incentive for major LECs to collude to "hide" procompetitive practices from regulators will be much weaker, at best, than in the context of traditional market collusion. But the primary manner in which the Order oversimplifies the analysis of collusive risk is by affording little or no weight to factors that militate against coordinated behavior among major incumbent LECs. First, although the Order essentially concedes that incumbent LECs do not resist opening their local markets to the extent they could, the Order affords insufficient weight to factors that result in such restraint by incumbents. Chief among these factors is incumbent LECs' continued need to curry favor with federal and state regulators in order to reduce regulatory burdens and obtain clearance to enter new markets. The paradigmatic example of these "regulatory incentives" at the federal level is the process by which Bell Companies must open their local markets pursuant to section 271's "competitive checklist" prior to obtaining in-region long distance authority. I would submit that section 271 still provides significant incentives for Bell Companies seeking to provide data services and otherwise serve lucrative corporate clients in their many locations throughout the country. Moreover, other federal proceedings provide similar regulatory incentives. In the context of access charge reform, price cap LECs will be able to obtain additional pricing flexibility if they can satisfy certain triggers (e.g., collocation) that recognize the extent to which the incumbent has allowed new entrants to compete in the local market. In the context of unbundling, the Commission's recent decision regarding unbundled network elements (UNEs) may encourage incumbents to provision network elements more efficiently in exchange for removing an element from the required list of UNEs, as the decision whether to unbundle was based in part on the ease with which CLECs can obtain UNEs. Even the Commission's authority to forbear from unnecessary regulation under section 10 may be viewed as encouraging incumbent LECs not to discriminate because section 10 requires the Commission to consider the extent to which competitors have been allowed to enter the market. There are no doubt many other federal and state regulatory incentives that will counterbalance incumbents' incentives to behave anticompetitively. It is sadly ironic that the Order laments the loss of Ameritech as an independent benchmark because of its venerable history of trying to open its local markets, even before passage of the 1996 Act. I similarly applaud Ameritech for such efforts, but I recognize that it did not break new ground out of some moral sense of what "good LECs" do. Instead, Ameritech was often merely aggressively pursuing its self-interest by attempting to curry regulatory favor and thereby win entry to new lucrative markets, such as long distance. Rather than lament the loss of Ameritech as an independent entity, perhaps we should acknowledge that we may have, in some sense, contributed to Ameritech's decision to merge, to the extent that our policies have made entry into other markets seem futile. Query, for example, whether Ameritech would have merged if it had had more confidence in our willingness to allow it to earn entry into the long distance market pursuant to section 271? Furthermore, there may be plenty of market-based incentives for a major LEC to break ranks with its putative co-conspirators. The case for collusion would be greater if these companies were competing to provide one product to the same set of customers. In fact, however, the communications market is much more multi-dimensional and dynamic. Companies are positioning themselves to compete in long distance, wireless, data and other services. In many of these areas, there are numerous competitors, thereby restricting the ability to collude. A carrier hoping to offer attractive packages of these services may be much less willing to agree to a truce on any one piece of the bundle. In addition, the Order affords insufficient weight to the cost and difficulty the applicants will face in coordinating behavior across the merged entity for all purposes, including discrimination. Intuitively, it is generally more difficult and costly to standardize activities across a larger holding company than across a smaller company. Thus, it is na‹ve to think that the applicants will engage in coordinated discriminatory behavior across the vast service region of the merged entity without seriously considering the costs and difficulties of actually coordinating that behavior. These costs and difficulties suggest that some efforts at coordinated discrimination will not be successful and others will not be attempted at all. 4. Variations in Practices Will Continue To conclude my critique of the Order's benchmarking analysis, I should add that I seriously question whether it is a given that a vastly larger company will necessarily lead to a more uniform set of policies, rigorously enforced from a central authority. Certainly, some of that will occur. However, the large size of the new company may also lead to a greater need to decentralize authority among the local operating companies, simply to promote efficient management and to allow the operating companies more flexibility to respond to varied and changing regulatory, economic and technical conditions across the merged entity's cast service area. Moreover, the combined company may have other reasons for permitting flexibility among its operating companies, such as encouraging innovation within the company. C. Discrimination The third public interest harm asserted to justify blocking the merger is that the merged entity will have an increased incentive and ability to discriminate against competitors. Like the preceding harms, this basis is not nearly as weighty as suggested by the Order. The Order's basic argument is that the merger increases the incentive and ability to discriminate in two ways: (1) by increasing the possibility of discriminating on both the originating and terminating end of a long distance call; and (2) by increasing the ability to reap the benefits in one region by raising the costs of a national competitor in another (so- called "spillover effects"). 1. Origination and Termination The possibility of discriminating in the provision of long distance is simple enough to understand. A single company that has a local exchange on both ends of a long distance call that it services will have an incentive to discriminate against other carriers on the termination side of the call. By doing so, its long distance service will be superior, prompting callers in the originating region to take service from it, rather than competitors. This concern with favoring one's own affiliate was the underpinning of the MFJ, as the Order notes. Judge Greene insisted that the "long lines" company (AT&T) be structurally separated from its local exchange facilities (the BOCs) for just this reason. What the Order fails to grasp, however, is that the 1996 Act made a fundamental change in the approach to this problem. Congress, expressly permitted companies to vertically integrate their local and long distance operations. Congress remained concerned, however, about the BOCs' bottleneck control over the local exchange. Yet rather than continue to bar BOCs from providing long distance, it established a comprehensive "checklist" that BOCs must satisfy under section 271 before they are permitted to enter the long distance market. The Order's concern with discrimination here results as much, or more, from the eventual vertical integration of local and long distance service by the applicants, than from the combination of local exchange regions. There is no question that the substantial local market share enjoyed by the BOCs magnifies the effect of this form of discrimination, but I believe Congress selected the safeguards for mitigating the harm. My concern is that the ability to offer both local and long distance service is expressly contemplated by the Act (though not by the MFJ). Congress addressed the dangers associated with local bottleneck control through a number of comprehensive statutory provisions, and through regulations that it charged the FCC to promulgate. Yet with a waive of a hand, we dismiss the adequacy of such safeguards, such as the separate affiliate requirement of section 272 that expressly prohibits favoring one's own affiliate over another. This, to my mind, contravenes the Congressional scheme. Moreover, it completely diverges from the conclusion reached in prior BOC mergers that such safeguards are adequate to protect against this form of discrimination. It does so, with only the meekest explanation that this merger simply is bigger and that there are fewer major LECs left. The Order fails, however, to explain adequately why the size or number of major LECs should prevent the safeguards from remaining effective. It is no wonder that the Order returns to the analysis of yesteryear and the MFJ to justify its conclusions. Additionally, the Order disregards the countervailing incentive to curtail discrimination in order to obtain (and maintain) section 271 authority to enter long distance. This incentive is the cornerstone of the congressional regime. BOCs would be inclined to discriminate much more than they do when providing access to their market if they did not hope to satisfy regulators that they had in fact opened up their network on a nondiscriminatory basis to competitors. To gain that approval, a BOC will have to accommodate competitive carriers to a much greater degree than it might otherwise. Ironically, in rejecting an empirical study on the record that firms have no greater incentive to discriminate against downstream rivals as evidenced by prior mergers, the Order cites the desire to "accommodate competitive LECs in order to enter the long distance market" to explain positive evidence. This, of course, is exactly what 1996 envisioned and lies at the core of its interconnection and non-discrimination provisions. 2. Spillover Effect The Order also cites the increased benefits post-merger of raising a national rival's costs. I offer three responses to this argument: First, such local discrimination would violate a bevy of statutory provisions and regulations. Of course, the Order dismisses these safeguards as inadequate, but I question whether it is the safeguards that are inadequate, or our own ability to promulgate and enforce them. I note a disturbing trend whereby we use "voluntary" merger conditions as a vehicle for remedying the ineffectiveness of our rules and our enforcement. Second, as discussed above, we do not take full account of the incentive to satisfy the requirements necessary to gain approval to enter the long distance market. Unless we are prepared to say that Congress' design is irretrievably flawed, I believe we must consider this motivation as a partial check on the incentive to discriminate in the local exchange. Third, this argument I believe places entirely too much emphasis on national carriers. In the words of the Order "the local exchange market is just that, a local market." Well said. In fact, while there may be a handful of national local carriers, there are substantially more localized CLECs that would not necessarily be affected by the spillover theory of discrimination. 3. The Order Fails to Make a Compelling Case That the Merger Will Result in Enhanced Discrimination Against Enhanced Services Providers I also want to say a word about the special emphasis placed on the incentive to discriminate against advanced services. It certainly is unsurprising to find greater interconnection problems with new services that have nontraditional requirements. It is also unremarkable to find such requests take longer and are more challenging to accommodate. The novelty of such services alone, however, does not explain why a BOC has either a greater incentive or ability to discriminate against such providers, and I do not find it clearly explained in the Order. Logically, the same spillover effects cited with respect to local exchange services might be found with advanced services, but it is not evident that that there are heightened dangers resulting from the merger to them that justify a special set of conditions. I sympathize with the desire to facilitate the deployment of advanced services on favorable terms, but I believe the conditions adopted in this Order to advance that goal are not squarely supported by the analysis of the harms. 4. The Order Presents Little or No Evidence of Discrimination Resulting from Prior LEC Mergers Finally, I have some real concerns about the Commission's rejection of empirical evidence suggesting that its theory is flawed. Carlton and Sider argue that if the hypothesis is correct that the merged firm will have a greater incentive to discriminate against downstream rivals, such behavior should have appeared as a consequence of the Bell Atlantic/NYNEX and SBC/Pactel mergers. I find the rejection of their evidence unsatisfying. Theory and speculation are unavoidable in merger review. Normally, however, we expect the government to substantiate its claims with evidence, when available. We have twice approved BOC mergers where the theories espoused here would be applicable. In at least one case, we found, contrary to this Order, that the regulatory safeguards are adequate protections against downstream discrimination. And, in both cases we allowed the merger to go through. Yet we do not now offer any evidence from those mergers to substantiate our theory. Has Bell Atlantic increased its discriminatory actions in order to get the spillover effects in the former NYNEX region? Perhaps to the contrary. Bell Atlantic, while not perfect, is widely regarded as having done more to open its local market than most other BOCs, a far cry from what the theory might predict. Out of its desire to obtain long distance approval, Bell Atlantic has worked extensively to satisfy the regulatory requirements. Indeed, it has before the Commission now an application to enter long distance in New York that is considered substantial, though we do not yet know if it is completely satisfactory. III. Conclusion In closing, I wish to reiterate that, although I think that the potential harms from this merger are plausible but overstated, and that we should have limited the conditions to those tailored to remedy the identified harms, it is my firm belief that the unique process employed in this review was initiated with the best intentions and with the highest regard for due process and for the public interest generally. I also wish to thank the Commission's staff and my colleagues for the enormous amount of time and energy they have devoted to this proceeding. I just wish this had been time better spent, and I hope we can avoid these pitfalls in future merger reviews.