I. Introduction
The United States and its major trading partners are increasingly interdependent in a shrinking, and sometimes perilous, world. Government policies in one country have a significant effect on enterprises in other countries and merger policies, in particular, have recently stimulated a great deal of commentary and lively controversy. I do not intend to revisit any controversies today but rather focus on an issue that I believe is even more fundamental, namely, the consistency of merger policy in the United States over time. When we in the U.S. talk about the desirability of long-term international convergence of merger policies, our trading partners are surely entitled to know whether our ideas about the appropriate path are subject to change without notice whenever a new administration takes office.
The conventional narrative - - fostered by our popular press and by some expert commentators - - is that merger policy in the U.S. has evolved like a pendulum swings. The pendulum is supposed to have swung dramatically from an aggressive enforcement policy in the 1960s and 1970s to a permissive policy in the Reagan and Bush years (1981-92), followed by a swing back to an aggressive policy in the Clinton years.(2) Other commentators believe that the pendulum began to swing back during the Bush years,(3) but the persistent image is one of reaction in the 1980s followed by a strong counter-reaction in the 1990s.(4)
The narrative has the same appeal as the plot line of a well constructed three-act play and, depending on the rhetoric, can be usefully employed by commentators of widely differing political views. For most, the antitrust of the 1960s and 1970s was fundamentally misguided; a small number still look upon it as a golden age. For some, the scaled-back government of the Reagan era laid the groundwork for almost two decades of unprecedented prosperity; for others, it led to speculative excess in a decade of greed. Current enforcement policies have been hailed by some as the restoration of needed balance and excoriated by others as the handiwork of government agencies out of control, once again.(5)
For an audience like this one, however, the conventional narrative may raise the possibility of a serious practical problem. How credible would a merger policy be if it really were so susceptible to change, depending on the outcome of Presidential elections? If the identity of enforcement personnel is so critical, how can there be any assurance of continuity in an environment where the average tenure of an Assistant Attorney General for Antitrust is about two years and the average Federal Trade Commissioner serves for just under five?(6)
The basic theme of this paper is that this potential problem is not a serious one because the conventional narrative is simply not true. The history of merger policy in the United States does not reveal either a steady state, or a wildly oscillating universe. What we observe is a rather dramatic change that was foreshadowed well before the Reagan election in 1980, followed by a period of gradual evolution at the margins since that time. Those who have an interest in continuity can be reassured, not only by the published commitments of incumbent enforcement officials in the U.S.(7) but also by the historical record. The word "essential" helps the title of this paper to convey a double meaning - - the title should be read to say, first, that the core principles of merger policy have remained stable for the last twenty years, regardless of political shifts, and, second, that this stability is important.
II. A Capsule History of U.S. Merger Policy
A. The Starting Point
The first chapter of the conventional narrative is accurate. Midway through the 20th Century, antitrust policy generally - - and merger policy, in particular - - was shaped by a mixed stew of economic misconceptions, social and political concerns, and a generous dose of nostalgia for a bygone era. The Celler-Kefauver Amendments to Section 7 of the Clayton Act(8) made sense to the extent they eliminated the artificial distinction between acquisitions of assets and acquisitions of stock but they were also based on misplaced concerns about increasing "aggregate concentration" in the U.S. economy, which was not only supposed to create inflationary pressures(9) but also seemed to portend the demise of family-sized businesses.(10)
The Supreme Court gave effect to a policy that appeared to favor the presentation of small-scale enterprises(11) even if consumers paid the price for some losses in efficiency. Moreover, mainstream economics decreed that there was a firm correlation between market concentration and elevated prices, and courts took judicial notice.(12) The combined effect was a merger policy that condemned relatively insignificant horizontal overlaps(13) and deemed a "trend toward concentration" as particularly pernicious.(14) Efficiencies were an aggravating rather than a mitigating factor because they could confer a "competitive advantage" that would fuel further concentration.(15) There was also a persistent belief that the preservation of a decentralized economy served important social and political goals.(16)
This set of policies began to crumble well before President Reagan appointed new agency heads in 1981. Opinions may differ on the important early landmarks but, in my view, the following events are worthy of mention.
- In 1968, Donald Turner, the head of the Antitrust Division under former President Johnson, published the first set of merger guidelines.(17) The horizontal and vertical market shares set forth as indicia of concern are extraordinarily low, by contemporary standards, but it is significant that the government acknowledged any limitations on its discretion when it seemingly could win any merger case it wanted.(18) Turner has never been given sufficient credit for what some at the time considered to be an act of considerable moral courage.
- In 1974, the U.S. Supreme Court rendered an opinion in the General Dynamics merger case,(19) which was a marked departure from earlier antitrust jurisprudence. The Court not only held for the defendants but also did so notwithstanding apparently high overlapping market shares. The opinion's analysis of specific industry facts, which indicated that historic market shares did not reflect future competitive significance, provided the precedent for the kind of detailed factual inquiry that is routinely followed today.
- 1974 was also the year when a long-simmering academic dispute over the economic significance of industrial concentration came to the attention of the larger antitrust community. The occasion was the publication of Industrial Concentration: The New Learning,(20) proceedings of a conference that featured an open debate between the proponents and the opponents of the received wisdom on oligopoly behavior. Although the Columbia University Center for Law and Economic Studies, which organized the conference, scrupulously allotted equal time to the opposing views, it is fair to say that publication of the proceedings seriously undercut the case for a merger policy based strictly on statistical measures of concentration.
- Beyond the intellectual debate, there was a growing recognition of U.S. vulnerability to international competition. Major U.S. industries, like automotive, were seriously affected by imports. Populist policies, which sacrificed efficiency in aid of other social and political goals, no longer appeared viable in the new competitive environment.
- In 1977, President Carter's newly-appointed Chairman of the Federal Trade Commission, made a highly publicized speech,(21) which created a bipartisan uproar and inadvertently provided a potent weapon to those who were wary of an antitrust based on nebulous social and political objectives. With rhetoric that bordered on self-parody, he advocated a "competition policy" that would consider things like "environmental harms . . . resource depletion, energy waste, environmental contamination, worker alienation, [and] the psychological and social consequences of producer-stimulated demands." These views not only alarmed people opposed to big government generally but also many mainstream antitrust supporters and those concerned with encroachment on the turf of other agencies and other constituencies.
- In 1977, the Supreme Court decided the Sylvania case,(22) with an opinion that borrowed liberally from a mode of economic analysis articulated by what later came to be known as the "Chicago School" or the "Law and Economics" movement. Sylvaniawas not a merger case, but the opinion cast a long shadow over merger jurisprudence in lower courts because of the implicit endorsement of economic views that also had an application to merger law. Most notable was the Court's implicit-assumption (later reinforced by other decisions) that the economic welfare of consumers, rather than other concerns, was the appropriate objective of antitrust. In addition, the Court was willing to entertain arguments on the efficiency justifications for business strategies.
- The year 1977 was a turning point for the Supreme Court in other antitrust cases, as well. The Court decided Brunswick, which articulated the antitrust injury requirement;(23) it decided the Fortner II case, which retreated dramatically from the extremist views on tying adopted in Fortner I;(24) and it decided Illinois Brick, which restricted private damage recoveries by indirect purchasers.(25) Any sensible observer at the time would have to conclude that antitrust policy had taken a fundamental turn.
- Reversing earlier trends, the antitrust agencies started to lose merger cases in the late 1970s. For example, from 1977 to 1983, when the Circuit Courts were finished with the FTC's late 1970s cases, the Commission won only eight of 22 merger cases.(26) In other substantive areas as well, the federal courts repeatedly expressed hostility to expansive notions of antitrust enforcement.
- In 1976 Congress passed the Hart-Scott-Rodino legislation(27) that, among other things, required pre-notification of mergers that satisfy certain threshold criteria. It may seem odd to cite this development as an event that foreshadowed a less restrictive merger policy, but I believe the statute ultimately had this entirely unanticipated effect. Agency staff members began to review a large number of transactions in a setting less adversarial than the preceding selective investigations. The agencies developed a familiarity with various industries and began to appreciate that most mergers did not have the consequences predicted by once-popular assumptions about oligopoly behavior.
The stage was thus set for the overtly expressed antitrust policy that followed the 1981 appointments of William Baxter and James Miller as heads of the antitrust agencies. These appointments unquestionably were significant events, but substantial change was inevitable in any event, in light of developments in mainstream economics and the trend of judicial decisions. The more important question, however, is the extent to which the reforms introduced during the Reagan era have survived into the present day.
B. The Lingering Legacy of the 1980s
The fundamental elements of the 1980s "revolution" - - if, indeed, it can be so characterized - - have survived intact. People tend to forget that, initially, the most controversial part of the 1980s enforcement agenda was the insistence on the primacy of consumer welfare concerns, and the consequent exclusion of other social and political factors. A shrinking but still prominent group of antitrust commentators believed this vision was unduly cramped. Anyone who wants to revisit the vehemence of this debate need only consult the recorded proceedings of the National Institute on Antitrust and Economics, presented in the early 1980s by the ABA's Section of Antitrust Law.(28) The supposed shortcomings of this economic approach were summarized as follows by a distinguished practitioner in the field of merger law:
I suggest to you that, like Halley's Comet, once in every 100 years, when the moon is full, we all take a binge on free market theories.
* * *
But, history tends to suggest also that we quickly are restored to some sense and to the mainstream traditions, rooted in the firm belief that our antitrust laws and similar legislation are not based only on simplistic notions about frictionless or static theories of markets, but also on a firm recognition that these laws incorporate three elements: (1) economic theory; (2) policy goals relating to control over the discretion and the power of the industrial sector of our economy by a few; and (3) the fear that excessive concentration tends to promote antidemocratic tendencies in the society as a whole.(29)
Some on the panel were equally indignant, if slightly less colorful.(30)
This objection has simply vanished from the mainstream debate over antitrust policy. In the last 20 years of close involvement with merger law - - including, most recently, two as a member of the Federal Trade Commission - - I am not aware of a single instance where non-economic factors have played any part in the ultimate decision of either federal agency.(31) The debate on all sides always begins with the economic premises that were formally introduced, amid vehement controversy, in the 1980s.
A focus on economics does not mean that merger cases are easy, with solutions that spill out of computers. Cases are highly fact-intensive. People with the same basic philosophy can and do come to different conclusions on individual cases. New theories of potential competitive harm have been advanced by economists who accept the primacy of consumer welfare analysis. The U.S. antitrust system has an intrinsically evolutionary quality; it is inevitable and desirable that doctrine will change as our understanding of economic phenomena improves.(32) Increasingly, we recognize that there are many things we do not know, or perhaps can never know, with certainty.(33)
Controversy continues, but the areas of fundamental agreement seem far more extensive than they were in the early 1980s. If our arguments were like a game of football, it could be said that we no longer battle over a 100 yard playing field but seem to play the game between the 45 yard lines. I suggest that there has been a gradual evolution over the last 20 years but no counter-revolution. This interpretation of history will be supported below by reference (i) to the agencies' own authoritative statements, (ii) the overall statistical record, and (iii) discussion of some individual cases.
Before turning to this objective evidence, however, it might be instructive to look at some prescient statements of the late William Baxter, President Reagan's first nominee to head the Antitrust Division and an outspoken advocate of a new approach to antitrust enforcement.
Speaking in 1985, he referred to "a continuing process of slowly improving understanding of how our economy works, about how enterprises in different kinds of markets interact with one another."(34) This understanding, he said,
". . . is a slowly growing body of knowledge of industrial organization . . . [that] is slowly and tortuously leading the legal profession away from the most inane outcomes it has produced over the last 30 or 35 years . . . in the direction of a fairly sensible antitrust policy, a policy based on whatever it is we know at any particular moment about the economics of industrial organization."(35)
In other words, one of the leading advocates and practitioners of 1980s economics expressly acknowledged at the time that it was not the end point of learning on the subject, and I do not know of any mainstream thinker who would make that claim today. One final Baxter quotation from the year 1986, for example, anticipates what has been perhaps the most lively debate in post-1980s economics, namely, theories about the anticompetitive potential of strategic behavior that falls outside traditional definitions of predation or exclusion:
"A great deal of theoretical work is going on in this area of cost-imposing activities, where you are imposing costs on your rivals. Now, I must say, not much that I regard as administrable as a matter of law has emerged from that yet, but I think there is reason to believe that it may well and that this may become a more active area of antitrust."(36)
Fifteen years later, work continues in this area - - and consideration of these effects is part of the ongoing evolution of merger law. I suggest we still have not found an "administrable" body of law, but developments in this or other areas depend a lot more on the expansion of knowledge than they do on the outcomes of elections.
III. The Persistence of Basic Themes in Merger Guidelines
Over time, there have been substantial differences in the tone and emphasis of public statements by different enforcement officials. It may be true that some officials in the 1980s tended to talk too much about the hypothetical cases they would not bring and some officials in the 1990s tended to talk too much about hypothetical cases that they would bring - - thus, alarming different groups of people. They sometimes seem to have been driven by the same bureaucratic imperative to distinguish themselves from their predecessors - - to highlight change, rather than continuity. And, of course, the media tend to exaggerate this impression; "more of the same" does not make a very compelling story.
However, anyone who looks at the definitive expressions of merger enforcement policy, contained in successive versions of the merger guidelines issued in 1982, 1984, and 1992 (including the 1997 efficiencies revisions), will note remarkable similarities. These guidelines, either issued or maintained during the administrations of the three Presidents who served from 1981 to 2001, not only outlined an analytical framework but also stated what the agencies intended to do.(37)
Some observers have argued that actual practice varied from stated intention and that there were always some "stealth" guidelines, best known to cognoscenti within the Beltway, which varied from the published versions. To some extent, this has been true on a case-by-case basis - - in fact, there have even been variations between different merger shops in a single agency at any given point in time. But, the available evidence does not support the inference that these variations were part of a systemic effort either to apply the published guidelines narrowly in the 1980s, to facilitate mergers, or broadly in the 1990s, to block them. A better explanation is that variations in application at any particular times, as well as variations in the language of successive guidelines over time, resulted from evolutionary changes in economic theory and from the gradual accumulation of industry expertise in the agencies. And, as will appear, the fact that a relatively insignificant number of mergers have been challenged throughout the two decades would indicate that the guidelines do provide accurate guidance to antitrust counselors.
In order to track the evolution of merger enforcement policies, it may be instructive to select and compare some parallel excerpts from successive versions of the guidelines.(38)
A. Underlying Themes
The central theme throughout has been that "mergers should not be permitted to create or enhance market power or to facilitate its exercise."(39) Moreover, William Baxter explained that both the 1982 and 1984 Guidelines were based on an "oligopoly theory," namely the proposition that it is easier for firms to coordinate their pricing behavior "if products tend to be homogeneous and the number of firms tend to be small. The criteria . . . represent a sort of prophylactic approach to conspiratorial collusion."(40) This sensitivity to statistical probabilities in merger enforcement continues to the present day.
B. The Common Search for Markets of Economic Significance
The 1982, 1984 and 1992 Guidelines incorporate subtle variations in the way they define markets. The 1984 Guidelines, for example, incorporated supply substitution factors directly into the market definition process, while the 1992 Guidelines focus "solely on demand substitution factors," and rely on supply substitution factors to identify "present and possible prospective participants in a market once that market has been defined."(41) In similar vein, the 1992 Guidelines incorporate refinements in the pricing assumptions for market definition methodology,(42) and introduce distinctions between so-called "committed" and "uncommitted entrants," in calculating market shares.
However, refinements of this kind (which only antitrust aficionados can appreciate and sometimes can be extremely difficult to apply) do not alter basic enforcement philosophy in one direction or another. The most significant thing is the ongoing reaffirmation of the basic principle that merger policy is informed by our best efforts to understand the economics of real markets, not politics or other social concerns. Those alternative avenues are as firmly avoided today as they were in the 1980s.
C. The Significance of Concentration
Statistical calculations of concentration have, if anything, become progressively less significant as we move from 1982, through 1984 and 1992, and into the present day. You would have thought that the movement would be precisely in the opposite direction if enforcers in the 1990s were trying to swing the pendulum back from an ostensibly laissez faire regime in the 1980s.
The basic tripartite split between markets that are "unconcentrated" (HHI under 1000), "moderately concentrated" (HHI between 1000 and 1800), and "highly concentrated" (HHI over 1800) has persisted in all three versions of the published Guidelines. But, it is important to consider the associated rhetoric. In the "highly concentrated" area, the 1982 Guidelines say that a challenge is "likely" if the delta is greater than 100, and that the agency "will resolve close questions in favor of challenging the merger."(43) This language was modified just two years later in order "to correct any misperception that the [1982] Guidelines are a set of rigid mathematical formulas that ignore market realities and rely solely on a static view of the marketplace."(44) In 1984, the DOJ stated that it "will not challenge mergers solely on the basis of concentration and market share data," but cautioned that in the highly concentrated range, other factors would trump statistics "only in extraordinary cases."(45) By 1992, the Guidelines refer to a presumption of illegality in this highly concentrated range, but go on to say that the presumption "may be overcome" by other factors, without the predecessor caveats about the height of the hurdle.(46)
The most likely explanation for this progressive shift of emphasis is obviously not ideology, but the accumulation of experience. For example, agency staffs had come to rely significantly on the opinions of customers, and had found that customer concerns often did not correlate with the HHI presumptions. Staff members had acquired this kind of expertise in a variety of industries and presumably had concluded, as expressed in the 1992 Guidelines, that "the larger universe of mergers . . . are either competitively beneficial or neutral."(47)
It is only fair to acknowledge the argument that the potential for troublesome "unilateral effects," articulated for the first time in the 1992 Guidelines,(48) is a backdoor way of focusing on small market segments (once called "submarkets"), with a consequent potential for inflating concentration statistics and creating new enforcement opportunities. Unilateral effects analysis has difficulties of its own that are still the subject of debate,(49) but I do not believe it would be accurate to claim that it was deliberately designed to catch more mergers, in a counter-reaction to merger policies of the 1980s.
In the first place, unilateral effects analysis was never described as a counter-reaction when the 1992 Guidelines were released. On the contrary, the analysis was expressly stated to be based on "a concept that was implicit in the "Leading Firm Proviso," contained in Section 3.12 of the 1984 Guidelines.(50) In the second place, unlike the facially similar "submarket" concept, unilateral effects analysis depends on sophisticated economic concepts applied to specific facts; it is not a theory that can be loosely applied in aid of a more populist antitrust.
D. Efficiencies
Here, it is clear that the Guidelines have become progressively more hospitable to efficiency defenses, a trend that is again contrary to what you would expect if the conventional narrative were true. William Baxter's 1982 Guidelines stated that "a claim of specific efficiencies as a mitigating factor" would be considered only in "extraordinary cases."(51) By 1984, the Department expressed concern about the "restrictive, somewhat misleading tone" of his statement.(52) Accordingly, the 1984 Guidelines acknowledge that efficiencies will be considered, but cautioned that they had to be demonstrated by "clear and convincing evidence."(53)
By 1992, the special evidentiary burden had been dropped, but the Guidelines introduced a sliding scale approach: "The expected net efficiencies must be greater the more significant are the competitive risks identified . . ."(54) This liberalization of the efficiencies defense was expressly commended by then Professor Robert Pitofsky,(55) who went on to chair the Federal Trade Commission in the late 1990s. The increased importance of efficiencies in agency analysis was underscored by the Guidelines revisions in 1997, which significantly expanded the discussion.
E. Treatment of Entry
Beyond the initial (and subsequently abandoned) complaints about neglect of non-economic factors in 1980s merger enforcement, the most vehement criticism has been that enforcers at that time tended to give undue weight to the potential for entry when evaluating mergers at high concentration levels.(56) To the extent that this criticism was directed at the agencies, it was exaggerated. The most noteworthy examples of entry trumping concentration were judicial decisions upholding mergers that had, in fact, been attacked by the Department of Justice.(57) Nevertheless, it is true that the 1992 Guidelines were overtly intended to shift the entry analysis from "whether effective entry could occur given the conditions of entry to whether it would occur given the alternative profit opportunities facing possible potential entrants."(58) This shift in emphasis is further articulated by the requirements that potential "entry would be timely, likely, and sufficient in its magnitude, character and scope to deter or counteract the competitive effects of concern," with further refinements that amplify on these terms.(59)
Facially, these changes tilt in a pro-enforcement direction in the 1990s - - in contrast to the 1990s changes in the weight given to concentration statistics and the standards for proving efficiencies, which facially point in the opposite direction. But, this change is hardly sufficient to support a broad claim that 1990s enforcement was a counter-reaction to perceived deficiencies of the 1980s. In addition, the jury is still out on whether the entry refinements in the 1992 Guidelines are really practical.(60)
F. Vertical Issues
A chronological comparison of authoritative policy statements is not possible in this area because the treatment of vertical issues in the 1984 Guidelines has never been revised.(61) At first glance, this inaction seems puzzling because there probably has been more intellectual speculation about vertical issues than any others. There is growing concern that vertical relationships may give rise to strategic behavior that can affect competition in ways not contemplated by the 1984 Guidelines.(62) (It is also true that the antitrust agencies have expressed more concern about vertical mergers in the middle 1990s, a subject that will be discussed in Section V below.)
In my view, the 1984 Guidelines have not been revised because the so-called "post-Chicago" learning in this area is still difficult to apply in practice, and there is no body of recent case law that the agencies are seeking either to reaffirm or to limit (as the 1992 Guidelines attempted when dealing with entry). In any event, the ongoing evolution of agency policies on vertical mergers cannot fairly be characterized as a reaction to the benign neglect of the area in the 1980s. It is better described as a manifestation of the expansion of economic understanding - - something that was expressly contemplated by the framers of 1980s antitrust policies.(63) If and when revised guidelines emerge on vertical merger issues, they will not shift the pendulum back in the direction of pre-1980s jurisprudence but will reflect entirely new learning.
G. A Summary Comment on the Evolution of Guidelines
It seems clear to me that successive iterations of merger guidelines, which are the most authoritative statements of enforcement intentions, do not support the simplistic myth of reactionary minimalist enforcement in the 1980s and a counter reaction throughout the 1990s. The guidelines reveal a more complex process that is not only responsive to current economic understanding but also to practical enforcement experience.
The practical side needs to be emphasized. The strong concentration presumptions in the 1982 Guidelines were soon seen to be impractical and began to be softened only two years later. The same can be said for the early expressions of hostility to efficiency claims. The trend in the treatment of entry arguments moved in the opposite direction. But, wholly apart from trends, there is always the risk that the most economically precise articulation of policy will set out standards of proof that are simply impossible to apply in the real world. In my view, some parts of the 1992 Merger Guidelines, now 10 years old, are subject to this criticism. It is not that the Guidelines go too far, or not far enough, in the direction of merger enforcement. The criticism is rather that they introduce concepts that have not been proven to be useful and are thus now largely ignored.
This is particularly true in the treatment of entry. The hypothetical "uncommitted entrants," who can enter and exit without incurring significant costs,(64) have proven as elusive as the Abominable Snowman; the category appears to be an empty box, like its theoretical godparent the "contestable market." Similarly, the extended "three step methodology" for evaluating the prospects of potential "committed" entrants(65) depends on the existence of data (and the human patience and capability to balance it)(66) that is simply not available.(67)
Others might disagree, but the important point is that disputes in the real world of merger enforcement take place "between the 45 yard lines"; they are not disputes over the fundamental direction of antitrust enforcement.
IV. The Statistical Record
One way to test whether the agencies are really doing what they say they will do is to look at the overall statistical record. In other words, if it is really true that mergers were much more aggressively prosecuted in the 1990s than they were in the 1980s, you would expect to see clear statistical evidence of it.(68) What does this evidence show?
Table I in the Appendix to this paper shows for years 1981 through 2000, the total number of transactions notified under the HSR law;(69) the number and percentage of transactions that were the subject of second requests, either by the FTC or the Department of Justice; and the number of transactions that were challenged.(70)
When analyzing these statistics, it is important to recognize that they generally are based on a fiscal year. Fiscal 1981, for example, began in October 1980, so it includes almost four months of the Carter administration and eight months of the Reagan administration. In addition, there is some time lag before any administration can put people in place to influence the direction of the agency - - particularly in the FTC, which is structured to be bipartisan and commissioners serve for fixed terms with staggered expiration dates. To make matters even more difficult, a precise year-to-year comparison of the ratio between enforcement actions and total filings is not meaningful because a transaction notified in one year may be challenged in a later year.
In order to compensate roughly for these problems, it makes sense to introduce a time lag of one fiscal year when comparing administrations. Thus, the "Reagan years" will be considered to begin in 1982, "Bush years" in 1990 and the "Clinton years" in 1994.
Some numbers in Table I would seem at first glance to support one version of the conventional narrative, and others do not. A very small percentage of notified mergers are challenged throughout the twenty year period, but the relative number of enforcement actions generally appears lowest in both antitrust agencies during the later Reagan years (1986-1989) and does increase markedly during the Bush years (1990-1993). Thereafter, however, the statistics show a decline in enforcement levels at the Federal Trade Commission during the Clinton years, and a continued increase at the Department of Justice. This pattern would make no sense if the conventional narrative were true. It is unlikely that an observer of any ideological persuasion would believe there were significant philosophical differences between the two agencies at that time.
Table II is another statistical comparison, which aggregates the relative frequency of enforcement actions by presidential terms. Some curious patterns, in addition to those noted above, now become evident.
It is even more obvious that the enforcement percentages in the Commission and the Justice Department diverge widely throughout. One possible answer is that there is a time lag in Commission responses because of the ongoing presence of Commissioners appointed in an earlier administration. This does not seem to explain the data here, however, because the Commission numbers swing just as sharply as the Justice numbers do - - in fact the difference between the Commission percentages in Reagan's second term and those in the Bush term is the widest of all. The two antitrust agencies just do not seem to track very well.
Moreover, if the conventional narrative is true, how can it account for the fact that the Commission enforcement percentages in bothReagan terms are virtually identical to the percentage in Clinton's second term?(71) Throughout the period attributed to the Clinton administration, all serving Commissioners but one were appointed by President Clinton.(72)
Other factors seem to be affecting the statistics. It is obvious that comparisons over time are only valid to the extent that the universe of merger filings is homogeneous. The relative number of enforcement actions is so small throughout that different industry patterns could have a significant effect on the results. In the late 1990s, for example, following partial de-regulation, the Justice Department reviewed a large number of mergers in radio broadcasting, and negotiated largely non-controversial consent decrees. Justice Department data(73)indicate, for example, that 15 "radio/TV' actions were brought in 1998 and 10 were brought in 1999. If these were eliminated from the Department's enforcement statistics, set out in Table I, the indicated percentages would drop from 1.1% to 0.8% in both 1998 and 1999 - - numbers that closely correspond to the Commission's ratios (0.7%).(74) I do not claim that these revised percentages are definitive because there could be special factors that tilt the other way, but this example shows how sensitive the numbers are to temporary anomalies.
The adjustment mentioned immediately above affects the numerators of the fraction used to calculate enforcement percentages; other adjustments could affect the denominator. Fiscal 2001 statistics have been eliminated from the tables altogether because early in 2001 the filing thresholds were raised and the number of filings was dramatically reduced. This reduction would tend to elevate the enforcement percentages in both agencies because the smaller transactions, which are no longer notifiable, were less likely to raise antitrust issues.(75)
There is an additional factor affecting the denominator which very likely reduced the relative number of merger challenges in both the late 1980s and the late 1990s. It is apparent from the statistics that there were significant merger "waves" in both periods. The ubiquity of the conventional narrative is revealed by the fact that the 1980s merger wave was widely attributed to supposedly lax merger enforcement,(76) while only fringe commentators(77) similarly attributed the 1990s wave to enforcement policies. In fact, it is likely that there really was a cause-and-effect relationship between the two phenomena but it works the other way around. Mergers are created by a convergence of general economic forces but the agencies only have the resources to challenge a relatively small number,(78) so the enforcement ratios decline when the waves hit.
The difference in the popular perception of the two merger waves seems particularly strange when they are compared in either absolute or relative terms. Appendix Table 1 shows that the number of filings at the peak in 2000 was almost twice the peak number in 1989. The differences in dollar value are even more striking, as displayed in Tables III-V. Even if the values are recast to reflect constant dollars, or expressed as a percentage of an expanding gross domestic product, the merger wave of the 1990s dwarfed the wave of the 1980s that attracted such a volley of critical comment.(79)
This comment should not be interpreted as a criticism of this aspect of Clinton-era merger policies. (I would be hard put to do so because I personally acquiesced in the approval of some of the largest mergers in history.) The point is that these statistics should be considered when we reflect on the most intemperate criticisms of merger policies in the 1980s.
Other factors may affect the statistics. An immediate decline in the relative incidence of merger challenges after the early 1980s may also reflect the fact that an increasingly specialized group of merger counselors, informed by merger guidelines, could more accurately predict what the agencies were likely to do. This should be an ongoing effect. Problematic mergers are shot down or modified in lawyers' offices before they are ever notified in the first place. My intuitive sense is that this is a genuine phenomenon, but have no way to measure how prevalent it is or whether it is related to the basic thesis of this paper. We could indulge in further speculations about whether counselors collectively have given more or less aggressive advice in various periods of time. For example, did counselors push the envelope further in the 1980s than in the 1990s because they believed the conventional narrative? This would suggest that the 1990s statistics understate the impact of merger enforcement. On the other hand, it may be that the 1990s numbers are inflated by a growing agency willingness to accept relatively painless decrees and facilitate mergers that would not even have been attempted before. If true, this would tilt the statistics the other way. Again, I doubt that these questions can be definitively resolved.(80)
There is another aspect of the statistics that is worthy of mention. Table I shows that the percentage of mergers subject to a second request declined throughout the 1990s but, during the same period of time, a relatively greater percentage of second request transactions was challenged. The most likely explanation here is that the agencies, with accumulated experience, have become more efficient in selecting transactions for further review. As indicated, budgetary pressures during the recent period of intense merger activity may also have played a part. Again, the data do not disclose the answer.
These comments do not exhaust all possible interpretations of the statistical data and further refinements obviously could be made to account for specific industry variations that affect the validity of comparisons over extended periods of time. I do not claim that the numbers presented here conclusively prove any particular hypothesis about merger policies in the last three presidential administrations.
The fact that the data are inconclusive is, however, precisely the point of the demonstration. The claim that merger enforcement in the 1990s differed dramatically from enforcement in the 1980s is not supported by statistical evidence.
V. Some Individual Cases
Antitrust cases generally, and merger cases in particular, are highly fact-specific, and it is therefore difficult to make comparisons over time. Moreover, except for litigated cases, the basic facts are not a matter of public record and no individual can claim familiarity with the full factual background of the myriad cases considered by the agencies over a period of two decades. Finally, even a study of agency arguments in litigated cases does not necessarily provide much guidance about shifts in enforcement philosophy. Internal deliberations may proceed in a relatively balanced way but, once the agencies make up their minds, they may litigate to win - - citing the most favorable precedents available. The disconnect between agency goals, expressed in guidelines and speeches, and the tone of agency briefs has been noted elsewhere.(81)
Some generalizations may, however, be warranted.
One useful piece of evidence is negative. I am not aware that anyone has pointed to a set of mergers that were passed by the agencies in the supposedly lax 1980s that would have been attacked in the supposedly more aggressive 1990s. On the contrary, some transactions that were regarded as highly controversial and subject to intense scrutiny in the 1980s - - like the then-substantial oil company mergers(82) or the General Motors-Toyota joint ventures(83) - - would likely be less troublesome in the 1990s. In fact, far larger oil company mergers were approved in the 1990s (albeit with some curative divestitures),(84) and the decree obligations inGM/Toyota were quietly terminated.(85)
For a second piece of negative evidence, I am not aware of any study that shows the HHI statistical presumptions were applied at lower levels in the 1990s than in the 1980s, and the fragmentary non-public information available to me suggests that there is no such trend. Similarly, beyond the personal impressions noted above,(86) I am not aware of any evidence that would show a systemic difference on the treatment of entry or efficiency claims.
It is also important to note the significant cases brought by the federal agencies in the 1980s, which established principles that the agencies relied upon routinely in the 1990s and continue to use today. The most important example, perhaps, is Robert Bork's opinion in Federal Trade Commission v. PPG Industries,(87) which created the framework of analysis that the Commission and the Courts have relied upon in many subsequent cases. Other illustrations of significant opinions in merger cases initiated by the Reagan antitrust agencies include two written by Richard Posner, Federal Trade Commission v. Elders Grain, Inc.,(88) and Hospital Corporation of America v. Federal Trade Commission.(89)
As already mentioned, there is evidence of a 1990s increase in the number of cases based on "vertical" theories. It is not true, however, that the agencies today are bringing vertical cases under the 1984 Guidelines that the agencies perversely refused to bring when their Guidelines were fresh. What has happened is that the agencies have negotiated consents based on theories developed since the Guidelines were drafted, and there still is insufficient learning to determine when the theories are well founded and when they are not.
Prior to the 1980s, vertical merger jurisprudence was based on a simplistic "foreclosure" theory, without further analysis. It was only necessary to prove that post-merger firms would have the ability to deprive rivals of access to some customers or suppliers, with not-insignificant market shares.(90) The fundamental insight provided by the so-called Chicago School was that vertical "foreclosure" statistics were meaningless in the absence of some demonstration of competitive impact at a horizontal level; foreclosure is harmless if it merely results in a realignment of customer/supplier relationships.(91) Theories about the ways in which vertical foreclosure could create adverse horizontal effects were limited twenty years ago,(92) and it is therefore not surprising that virtually no vertical merger cases were brought.
The antitrust agencies today still accept the fundamental Chicago-school insight that vertical mergers are problematic only to the extent that they are likely to have an effect on competition at a horizontal level.(93) What has happened, however, is that relatively recent theories about strategic behavior - - specifically foreseen by William Baxter fifteen years ago(94) - - have provided the basis for a number of consent settlements, although they have not yet been tested in judicial or administrative trials.
A significant number of vertical cases involve 1990s mergers in the defense industry, where companies historically have had complex horizontal and vertical relationships in numerous separate projects.(95) The principal concerns have not been with total foreclosure, two-level entry or other matters mentioned in the 1984 Guidelines, but rather with strategic use of information gained in one project to obtain competitive leverage in other projects. The party principally concerned has been the Department of Defense, the sole customer. The preferred remedy for the problem has been a decree or a letter of agreement erecting "firewalls" to avoid the spillover use of competitive information.(96)
Another area of concern has been healthcare. Early in the 1990s, there were a number of acquisitions by pharmaceutical companies of so-called pharmaceutical benefit managers (PBMs) that interface between the manufacturers and the large health plans which provide coverage for their members. Among other things, the PBMs influence the drugs that are on the approved "formulary," and hence available to the plan members. The mergers went unchallenged until Eli Lilly proposed acquisition of PCS Health Systems, the largest remaining PBM in the country.(97) The FTC challenged the transaction and negotiated an order that was intended to ensure that, post-acquisition, PCS would not unduly give preference to the drugs manufactured by its parent Lilly.(98)
Without indulging in a detailed commentary on the case, it is noteworthy that the sketchy complaint accompanying the order did not elaborate on the reasons for believing that potentially favorable treatment of Lilly by a single PBM would give rise to competitive harm. As the dissent pointed out,(99) the complaint did not appear to rely on post-Chicago theories of competitive harm, or explain how the decree would work to preserve competition.
One possible response is simply to acknowledge that the Eli Lilly decree, and the similar decree entered later but in response to an earlier PBM acquisition by Merck,(100) are outliers with no particular precedential significance. There is no indication whatever that these decrees stand for a general return to the days when foreclosure alone was enough. I personally believe that the cases may foreshadow some entirely new approaches that could be of future significance for the analysis of both horizontal and vertical mergers. I have argued that, for some products or services, competitive harm cannot be adequately captured by a traditional focus on price and output - - issues of "quality" or "variety" are also significant.(101) Some may have believed that it is particularly important to preserve the "variety" of drug therapies available to individual subscribers in benefit plans served by companies like PCS, even if the "foreclosure" is incomplete and temporary.
This variety issue was an overt concern in the later intense Commission investigation that resulted in abandonment of the proposed merger of Barnes & Noble, the largest book retailer in the U.S., with Ingram Book Group, the largest book wholesaler. The expressed concern of Commission staff, inter alia, was that Ingram offered qualitatively superior service to the retailing trade, including a greater diversity of product, and that if Barnes & Noble deprived competing retailers of access to Ingram, it would reduce both price and non-price competition at the retail level.(102) An adverse impact on variety might be a particularly significant competitive concern in the market for non-fungible products like books, just as it may be for pharmaceutical products.
Since the Lilly case was settled by a decree, and the Barnes & Noble transaction was abandoned, possible variety issues were never litigated. If, however, issues of this kind are litigated - - in either a horizontal or a vertical context - - they will be defined in ways that are consistent with recent economic learning, not on the basis of theories discredited twenty years ago.
Had it gone to trial, a case like AOL/Time Warner(103) could also have raised some interesting vertical issues relating to the potential for strategic behavior, as well as issues of non-price competition. The AOL/Time Warner transaction was a vertical combination at three distinct levels: (i) media entertainment of various kinds, in which one wing of Time Warner had arguably the broadest portfolio in the world; (ii) internet service, in which AOL was the dominant "narrowband" provider with the arguable potential to become the dominant "broadband" company; and (iii) cable services, for which another wing of Time Warner was the dominant - - and often exclusive - - provider in approximately 20% of the country.
Consistent with Chicago-school learning, the Commission focused on the possible horizontal competitive effects of each of these three levels as a result of the vertical combination. However, we considered not only the possible impact of total foreclosure, but also the impact of strategies short of foreclosure suggested by post-Chicago theories. In this connection, we considered whether traditional "market share" statistics could adequately capture the potential for strategic deployment of Time Warner's entertainment portfolio.(104)
The interesting thing to me, as a participant in the decision, was the sophistication of the economic arguments advanced in support of, and in opposition to, the merger. This was not a dispute between adherents of "pre-Chicago," "Chicago" and "Post-Chicago" theories; it was a dispute between people who largely shared the same economic philosophy but had very different opinions about what was likely to happen in fast-moving industries with little history to draw on for guidance.
In my view, these are the defining characteristics of 1990s cases. They are difficult and fact-intensive. Because of the prevailing practice of settlement by decree - - which has been the subject of much commentary, important in its own right(105) - - there is very little recent case law to provide guidance. The agency decision-makers are, to a great extent, on their own. In these circumstances, it is significant that there is so much unanimity over outcomes and, when there are differences, they are not primarily driven by ideology. I have acknowledged before that there may be some individual differences in toleration for risks of over-enforcement (Type I) or under-enforcement (Type II), but these affect only a handful of cases at the margin.(106)
Conclusion
For the reasons set out in this paper, I suggest that the conventional narrative is wrong. Merger policies in the 1990s do not represent a counter-reaction to, or a repudiation of, policies in the 1980s. The fundamental policy shift of the 1980s - - namely, the focus on the economic welfare of consumers - - remains firmly in place. There still is an effort to assess effects in real markets, based on the best tools available. There have been evolutionary changes, but they are forward-looking and change was anticipated by those in charge of the antitrust agencies twenty years ago.
A number of reasons have been suggested for disconnect between the popular perceptions and reality. Policymakers, critics, commentators and journalists of all kinds are only heard in a noisy world when they say something different. A speech or a news article that emphasizes continuity will attract a lot less attention than one that emphasizes disagreement or change. (I recognize that this paper, which emphasizes disagreement with conventional wisdom, is subject to the same caveat. It is "different" to say that things are not different.)
An additional factor, worthy of separate study, is the longevity of career staff in both antitrust agencies. Career staff have greater industry-specific expertise than politically appointed policymakers and when cases are fact-intensive, as they are, staff's views inevitably carry great weight. As administrations change, policymakers come and go (albeit with time lags in the Commission), but career staff stay on to provide continuity. I do not know about the Antitrust Division of the Department of Justice, but the senior career staff involved in merger work at the Federal Trade Commission have served for an average in excess of eighteen years. In other words, they have served through most of the 1980s, they have served through the 1990s, and they are serving today in a new century.
The relative stability that we have seen over the last twenty years, in a period of vigorous policy debate, is a good omen for the future. If the evidence arrayed in this paper is at all persuasive, it should reassure our trading partners about the essential stability of merger policy in the United States.
Table 1 - Number of Reported Transactions,
Second Requests, and Enforcement Actions
| Second Requests | Enforcement Actions |
|||||||||
| Year | Filings | Number | % ofFilings | Number | % of Filings | % of Second Requests | FTC: Number | FTC: % of Filings |
DOJ: Number |
DOJ: % of Filings |
| 1981 | 762 | 69 | 9.1 | 17 | 2.2 | 24.6 | 13 | 1.7 | 4 | 0.5 |
| 1982 | 713 | 50 | 7.0 | 26 | 3.6 | 46.0 | 12 | 1.7 | 14 | 2.0 |
| 1983 | 903 | 48 | 5.3 | 12 | 1.3 | 25.0 | 5 | 0.6 | 7 | 0.8 |
| 1984 | 1119 | 77 | 6.9 | 23 | 2.1 | 22.1 | 14 | 1.3 | 9 | 0.8 |
| 1985 | 1450 | 84 | 5.8 | 14 | 1.0 | 15.5 | 9 | 0.6 | 5 | 0.3 |
| 1986 | 1660 | 71 | 4.3 | 13 | 0.8 | 18.3 | 7 | 0.4 | 6 | 0.4 |
| 1987 | 2170 | 58 | 2.7 | 20 | 0.9 | 34.5 | 12 | 0.6 | 8 | 0.4 |
| 1988 | 2391 | 68 | 2.8 | 34 | 1.4 | 50.0 | 23 | 1.0 | 11 | 0.5 |
| 1989 | 2535 | 64 | 2.5 | 28 | 1.1 | 43.8 | 19 | 0.8 | 9 | 0.4 |
| 1990 | 1955 | 89 | 4.6 | 52 | 2.7 | 58.4 | 35 | 1.8 | 17 | 0.9 |
| 1991 | 1376 | 64 | 4.7 | 42 | 3.1 | 65.6 | 29 | 2.1 | 13 | 0.9 |
| 1992 | 1451 | 44 | 3.0 | 22 | 1.5 | 50.0 | 14 | 1.0 | 8 | 0.6 |
| 1993 | 1745 | 71 | 4.1 | 32 | 1.8 | 45.1 | 21 | 1.2 | 11 | 0.6 |
| 1994 | 2128 | 73 | 3.4 | 50 | 2.4 | 68.5 | 28 | 1.3 | 22 | 1.0 |
| 1995 | 2612 | 101 | 3.9 | 561 | 2.3 | 60.4 | 43 | 1.6 | 18 | 0.7 |
| 1996 | 2864 | 99 | 3.5 | 57 | 2.0 | 57.6 | 27 | 0.9 | 30 | 1.0 |
| 1997 | 3438 | 122 | 3.5 | 57 | 1.7 | 46.7 | 26 | 0.8 | 31 | 0.9 |
| 1998 | 4575 | 125 | 2.7 | 85 | 1.9 | 68.0 | 34 | 0.7 | 51 | 1.1 |
| 1999 | 4340 | 111 | 2.6 | 76 | 1.8 | 68.5 | 30 | 0.7 | 46 | 1.1 |
| 2000 | 4749 | 98 | 2.1 | 80 | 1.7 | 81.6 | 32 | 0.7 | 48 | 1.0 |
Notes:
- Data for 1982 through 1985 is calendar year data. Data for other years is fiscal year data. There are two actions that are counted twice. Each of the FTC and DOJ "challenged" one transaction in the last quarter of calendar year 1985. The number of HSR filings (in which a second request could be issued) for the months Oct., Nov., and Dec. 1985 could not be identified; an adjustment was made to ensure that filings and challenges in that period were counted comparably.
- Challenge data is drawn from: (i) FTC Mission Accomplishments, for the fiscal years 1986 through 2000, and counts as challenges the sum of PI's authorized, administrative complaints issued, Part II consents accepted, and transactions withdrawn after issuance of a second request (in all cases, duplicate cases were eliminated); (ii) DOJ Workload Statistics for the years 1988 through 2000, counting as challenges the sum of merger actions filed in the district court, and transactions abandoned or restructured prior to filing a complaint as a result of an announced challenge; and (iii) HSR Annual Reports filed for the Calendar Years 1982-1985, and for the Fiscal Year 1986 and 1987 (one report), which in the text and footnotes describe cases brought and transactions abandoned. Duplicate cases eliminated. In all cases, data was collected as of date of first instance of challenge.
- Number of Transactions counts the number of transactions in which the agencies were authorized to issue a second request. This number is slightly lower than number of transactions filed for, accounting for mistaken filings, secondary filings and filings in which a party files to cross multiple notification thresholds in the same year. Data is drawn from the HSR Annual Reports for the Years 1999, 1992 and 1985, at Appendix A and C.
Table 2 - Merger Enforcement, Challenges as a % of HSR Filings
| 1982-1985 (Reagan I) |
1986-1989 (Reagan II) |
1990-1993 (Bush) |
1994-1997 (Clinton I) |
1998-2000 (Clinton II) |
|
| FTC | 1.0% | 0.7% | 1.5% | 1.1% | 0.7% |
| DOJ | 0.8% | 0.4% | 0.8% | 0.9% | 1.1% |
| TOTAL | 1.8% | 1.1% | 2.3% | 2.0% | 1.8% |
Notes: Data is drawn from the same sources as Table 1.
Table 3 - Number of Mergers, Divestitures and Disclosed Value (1968-2000)
| Year | Net merger and acquisitions announcements | Number of transactions with purchase price disclosed |
Total Divestitures |
Divestitures as % of Total |
Total dollar value paid ($ billions) |
Constant dollar value * ($ billions) |
| 1968 | 4462 | 1514 | 557 | 12.5 | 43.60 | 119.1 |
| 1969 | 6107 | 2300 | 801 | 13.1 | 23.70 | 62.4 |
| 1970 | 5152 | 1671 | 1401 | 27.2 | 16.40 | 41.7 |
| 1971 | 4608 | 1707 | 1920 | 41.7 | 12.60 | 31.1 |
| 1972 | 4801 | 1930 | 1770 | 36.9 | 16.70 | 40.0 |
| 1973 | 4040 | 1574 | 1557 | 38.5 | 16.70 | 36.6 |
| 1974 | 2861 | 995 | 1331 | 46.5 | 12.50 | 23.8 |
| 1975 | 2297 | 848 | 1236 | 53.8 | 11.80 | 20.3 |
| 1976 | 2276 | 998 | 1204 | 52.9 | 20.00 | 32.9 |
| 1977 | 2224 | 1032 | 1002 | 45.1 | 21.90 | 33.8 |
| 1978 | 2106 | 1071 | 820 | 38.9 | 34.20 | 49.0 |
| 1979 | 2128 | 1047 | 752 | 35.3 | 43.50 | 56.1 |
| 1980 | 1889 | 890 | 666 | 35.3 | 44.30 | 50.3 |
| 1981 | 2395 | 1126 | 830 | 34.7 | 82.60 | 86.0 |
| 1982 | 2346 | 930 | 875 | 37.3 | 53.80 | 53.8 |
| 1983 | 2533 | 1077 | 932 | 36.8 | 73.10 | 71.9 |
| 1984 | 2543 | 1084 | 900 | 35.4 | 122.20 | 117.8 |
| 1985 | 3001 | 1320 | 1218 | 40.6 | 179.80 | 171.7 |
| 1986 | 3336 | 1468 | 1259 | 37.7 | 173.10 | 167.7 |
| 1987 | 2032 | 972 | 807 | 39.7 | 163.70 | 155.3 |
| 1988 | 2258 | 1149 | 894 | 39.6 | 246.90 | 228.6 |
| 1989 | 2366 | 1092 | 1055 | 44.6 | 221.10 | 194.6 |
| 1990 | 2074 | 856 | 940 | 45.3 | 108.20 | 90.8 |
| 1991 | 1877 | 722 | 849 | 45.2 | 71.20 | 58.5 |
| 1992 | 2574 | 950 | 1026 | 39.9 | 96.70 | 78.5 |
| 1993 | 2663 | 1081 | 1134 | 42.6 | 176.40 | 141.5 |
| 1994 | 2997 | 1348 | 1134 | 37.8 | 226.70 | 180.6 |
| 1995 | 3510 | 1735 | 1199 | 34.2 | 356.00 | 278.3 |
| 1996 | 5848 | 2658 | 1702 | 29.1 | 495.00 | 377.0 |
| 1997 | 7800 | 3013 | 2108 | 27.0 | 657.10 | 498.6 |
| 1998 | 7809 | 3091 | 1987 | 25.4 | 1191.90 | 911.9 |
| 1999 | 9278 | 3384 | 2353 | 25.4 | 1425.90 | 1072.1 |
| 2000 | 9566 | 3757 | 2501 | 26.1 | 1325.70 | 960.7 |
*Constant dollar value is the annual dollar value divided by the seasonally adjusted Producer Price Index, by Stage of Processing, Total Finished Goods (1982=100), Table B-65, p. 349, Economic Report of the President, January 2001.
SOURCE: Mergerstat® Review, 2001, pp. 2 and 9. The Mergerstat® Review Research Department tracks publicly announced formal transfers of ownership of at least 10 percent of a company's equity where the purchase price is at least $1,000,000, and where at least one of the parties is a U.S. entity. These transactions are recorded as they are announced, not as they are completed. Open market stock purchases are not recorded. For sellers in the database with competing bids, only the highest offer is included in the calculation. Cancelled transactions are deducted from total announcements in the period in which the cancellation occurred, resulting in net merger-acquisition announcements for that period. The statistics reflect completed or pending transactions as of the end of the applicable period.
Table 4 - Merger and Acquisition Activity ( 1968-2000 )
Table 5 - Merger and Acquisition Dollar Value as a Percentage of GDP (1968-2000)
Endnotes:
1. Commissioner, Federal Trade Commission. I acknowledge the assistance of my advisor, Thomas J. Klotz, but take sole responsibility for the views expressed, which are not necessarily shared by any other commissioner.
2. See, e.g., John M. Broder, F.T.C. Rejects Deal To Join Two Giants of Office Supplies, New York Times, Apr. 5, 1997, at § 1, p. 1 (FTC challenge to merger of Staples and Office Depot "reflects a growing aggressiveness in antitrust enforcement by the F.T.C., which was relatively passive during the Bush and Reagan Administrations"); Robert D. Hershey, Jr., An Energetic Trustbuster, but Not a Boat Rocker, N.Y. Times, July 25, 1994, at § D, p. 1 ("Anne K. Bingaman, in her first year as the top Justice Department trustbuster, has transformed a dormant antitrust operation into a fountain of activity."); Al Kamen, In the Loop, Wash. Post, Aug. 4, 1993, at A15 (announcing appointment of Robert Litan to be deputy to Anne Bingaman, Assistant Attorney General for Antitrust; calling Litan's appointment part of effort "to reinvigorate the long-moribund antitrust operation after the pro-trust years"); James Vicini, Antitrust Enforcement Tougher Under Clinton, Chicago Sun-Times, Nov. 6, 1994, at 2 (Justice Department's Antitrust Division "had dwindled in importance under the Reagan and Bush Administrations"); Sharon Walsh, Antitrust Head Bingaman to Leave Post By Nov. 15, Wash. Post, Aug. 2, 1996, at A19 (Anne Bingaman is "credited with reviving the once-moribund antitrust division of the Justice Department");Administration's `Hot Jobs,' Part II, Wash. Post, Nov. 11, 1996, at A27 ("Under Anne K. Bingaman, the once moribund Antitrust Division became a hive of activity in bringing major cases against corporate giants.").
3. See, e.g., Robert Pitofsky, Antitrust Policy in a Clinton Administration, 62 Antitrust L.J. 217, 218 (1993) (Bush appointees to DOJ and FTC "abandoned the orthodoxy" of Reagan officials who "viewed government regulation as a source of much mischief"); David Francis,Big Business Gets Bigger, Record Pace, Christian Science Monitor, June 4, 1997, at 8 ("President Reagan kept antitrust policies relatively dormant. But Presidents Bush and Clinton reactivated the two antitrust agencies . . . .").
4. See, e.g., Eleanor M. Fox, Can We Control Merger Control? - An Experiment, in Policy Directions for Global Merger Policy 79, 84 (Global Forum for Competition and Trade Policy: 1999)("During the Reagan Administration . . . U.S. federal merger enforcement ground to a halt."); Walter Adams & James W. Brock, "Reaganomics and the Transmogrification of the Merger Policy," 33 Antitrust Bull. 309, 309-10 (1988). ("The Reagan Administration's most conspicuous antitrust achievement was its emasculation of the nation's merger policy.")
5. For recent examples, compare the rhetoric of former FTC Chairman Robert Pitofsky and the editorials of the Wall Street Journal. SeeRobert Pitofsky, An Antitrust Progress Report for the FTC: Past, Present and Future, Remarks before Antitrust and Business magazines, 1996 Conference of Business Development Assocs. Inc., Washington, D.C. (March 4, 1996) at 2 ("The Commission of the 1990s has tried to strike a middle ground between what many people believe was an excessively active enforcement in the 1960s and the minimalist enforcement of the 1980s."), available at </speeches/pitofsky/speech4.htm>; Robert Pitofsky, FTC Staff Report on Competition Policy: Six Months After, Prepared Remarks before the ABA Section of Antitrust Law Conference on the Changing Nature of Competition - Legal and Policy Implications, Washington, D.C. (Nov. 7, 1996), available at </speeches/pitofsky/rpaba11.htm>; Jeffrey H. Birnbaum, Washington's Most Dangerous Bureaucrats, 136 Fortune 118, 124 (Sept. 29, 1997)("Charles Rule, a former head of the Justice Department's antitrust division under Reagan, considers Pitofsky's vigorous antitrust actions 'a swing back to what was going on in the 1960s and 1970s.'")
6. These averages cover the last 30 years.
7. See Timothy J. Muris, Antitrust Enforcement at the Federal Trade Commission: In a Word-Continuity, Prepared Remarks before ABA Antitrust Section Annual Meeting, Chicago, Ill. (Aug. 7, 2001), available at </speeches/muris/murisaba.htm>; Charles A. James, Be Careful What You Wish For: Some Thoughts on the Merger Review Process, Address Before ABA Antitrust Section, Chicago, Ill. (Aug. 7, 2001), available at <http://www.usdoj.gov/atr/public/speeches/8764.htm> at 3.
8. Act of Dec. 29, 1950, Pub. L. No. 81-899, 64 Stat. 1225 (codified as amended at 15 U.S.C. § 18 (1994)).
9. See Willard F. Mueller, Industrial Concentration: An Important Inflationary Force, in Industrial Concentration: The New Learning 280, 281 (Harvey J. Goldschmid et al., eds. 1974).
10. S. Rep. No. 81-1775, (1950), reprinted in 1950 U.S. Code Cong. Serv. 4293, 4295 ("The extent to which the American economy has become concentrated and centralized in the hands of a few giant corporations was strikingly revealed . . . in figures presented by Dr. Wilfred I. King[.] . . . [H]is figures show that in 1946 . . . one-tenth of one percent of the total number of all American corporations -- the giant firms with assets of $100,000,000 and over -- owned 49 percent of the assets of all American corporations[.] . . . At the other end of the scale, 45 percent of the number of corporations -- small firms with assets of $50,000 or less -- owned less than one percent of the assets . . . . The enactment of the bill will limit further growth of monopoly and thereby aid in preserving small business as an important competitive factor in the American economy.").
11. See Brown Shoe Co. v. United States, 370 U.S. 294 (1962)(citing legislative history of Celler-Kefauver Amendments, expressing antitrust concern about permitting merger giving firm five percent market share, and further expressing concerns that large national chains with such small shares can adversely affect competition).
12. See, e.g., Stanley Works v. FTC, 469 F.2d 498, 504-05 (2d Cir. 1972)("Although the teachings of scholars cannot be dispositive, it would be foolhardy, in an area as complicated as this, to wholly disregard the lessons they have learned in a lifetime of study," and citing Carl Kaysen and Donald F. Turner, Antitrust Policy: An Economic and Legal Analysis (1959) and Joe Staten Bain, Industrial Organization (2d ed. 1968)).
13. E.g., United States v. Von's Grocery Co., 384 U.S. 270 (1966)(merger between third and sixth largest retail grocery chains in Los Angeles area with combined market share of approximately seven percent illegal); id. (condemning acquisition of leading firm in cabinet hardware sales with 22-24% market share by tenth largest seller with 1% share).
14. See, e.g., United States v. Pabst Brewing Co., 385 U.S. 546 (1966).
15. E.g., Foremost Dairies, 60 F.T.C. 944, 1059, 1080-81 (1962), modified 67 F.T.C. 282 (1965). Wesley J. Liebeler, Bureau of Competition: Antitrust Enforcement Activities, in The Federal Trade Commission Since 1970: Economic Regulation and Bureaucratic Behavior 65, 94 (Kenneth W. Clarkson & Timothy J. Muris eds. 1981).
16. Robert Pitofsky, The Political Content of Antitrust, 127 U. Pa. L. Rev. 1051 (1979).
17. United States Dep't of Justice, Merger Guidelines (1968), reprinted in 4 Trade Reg. Rep (CCH), ¶ 13,101.
18. United States v. Von's Grocery Co., 384 U.S. 270, 301 (1966)(Stewart, J., dissenting) ("The sole consistency that I can find is that in litigation under [Section] 7, the Government always wins.").
19. United States v. General Dynamics Corp., 415 U.S. 486 (1974).
20. Industrial Concentration: The New Learning, supra. note 9.
21. Michael Pertschuk, Remarks before the Annual Meeting of the Section of Antitrust and Economic Regulation, Association of American Law Schools, Atlanta, Ga. (Dec. 27, 1977). See also Michael Pertschuk, New Directions for the FTC, Prepared Remarks before Eleventh New England Antitrust Conference, Boston, Mass. (Nov. 18, 1977).
22. Continental T.V., Inc. v. GTE Sylvania, Inc., 433 U.S. 36 (1977).
23. See Brunswick Corp. v. Pueblo Bowl-O-Mat Inc., 429 U.S. 477 (1977)("The antitrust laws . . . were enacted for the protection of competition, not competitors.").
24. See United States Steel Corp. v. Fortner Enters., 429 U.S. 720 (1977)(Fortner II).
25. Illinois Brick Co. v. Illinois, 431 U.S. 720 (1977).
26. Bureau of Consumer Protection, Bureau of Competition & Bureau of Economics, Federal Trade Commission Law Enforcement in the 1980's: A Progress Report on the First Three Years of the Reagan Administration Leadership, October 1981 to October 1984, at 41 (Federal Trade Commission Oct. 1984); see also M. Howard Morse, Vertical Mergers: Recent Learning, 53 Bus. Law. 1217, 1223 (1998).
27. Hart-Scott-Rodino Antitrust Improvements Act of 1976, Pub. L. No. 94-435, tit. II, § 201, 90 Stat. 1390 (codified as amended at 15 U.S.C. § 18a (1994)).
28. National Institute on Antitrust and Economics, 52 Antitrust L.J., 515 (1983).
29. Stephen M. Axinn, A Lawyer's Response, 52 Antitrust L.J. 643 (1983).
30. See, e.g., Gordon B. Spivack, The Chicago School Approach to Single Firm Exercises of Monopoly Power: A Response, 52 Antitrust L.J. 651, 653 (1983). See also Kenneth M. Davidson, The Competitive Significance of Segmented Markets, 71 Cal. L. Rev. 445 (1983); Louis B. Schwartz, The New Merger Guidelines: Guide to Governmental Discretion and Private Counseling or Propaganda for Revision of the Antitrust Laws, 71 Cal. L. Rev. 575, 577 (1983) ("a foreseeable and less acceptable consequence of the Guidelines: the facilitation of the sort of 'legal smuggling' that results in the conversion of transitory policy into permanent legal constraints favoring antitrust defendants").
31. The same cannot be said of all state agencies.
32. William E. Kovacic & Carl Shapiro, Antitrust Policy: A Century of Economic and Legal Thinking, 14 J. Econ. Perspectives 43 (2000)(describing the influence of economic analysis on the development of legal antitrust doctrines in United States).
33. Federal Trade Commission Bureau of Economics, Empirical Industrial Organization Roundtable (Sept. 11, 2001), transcript available at </be/empiricalioroundtabletranscript.pdf;> Thomas B. Leary, Antitrust Economics: Three Cheers and Two Challenges, Remarks Before Charles River Associates Conference, Washington, D.C. (Nov. 15, 2000) and Remarks Before Western Economic Association, San Francisco, Cal. (July 7, 2001), available at </speeches/leary/learythreecheers.htm#N_1_>.
34. William F. Baxter, Antitrust: A Policy in Search of Itself, 54 Antitrust L.J. 15, 16 (1985)(emphasis added).
35. Id.
36. William F. Baxter, Counseling Your Client on Monopolization, Mergers and Joint Ventures Panel Discussion, 55 Antitrust L.J. 321, 327 (1986).
37. The power of the vision in these guidelines is demonstrated by the remarkable extent to which they have provided an analytic starting point for other countries designing their own guidelines. These other countries voluntarily selected these approaches in the marketplace of ideas.
38. These comparisons are selective because I want to highlight certain issues that have been the subject of controversy and because people would otherwise stop reading this paper. Someone more academically oriented may be inclined to do a comprehensive job.
39. United States Dep't of Justice, Merger Guidelines (June 14, 1982), reprinted in 4 Trade Reg. Rep (CCH) ¶ 13,102 (hereinafter "1982 Guidelines") at § 1; United States Dep't of Justice, Merger Guidelines (June 14, 1984), reprinted in 4 Trade Reg. Rep. (CCH) ¶ 13,103 (hereinafter "1984 Guidelines") at § 1; United States Dep't of Justice and Federal Trade Commission, Horizontal Merger Guidelines (Apr. 2, 1992), reprinted in 4 Trade Reg. Rep. (CCH) ¶ 13,104 (hereinafter "1992 Guidelines") at § 0.1.
40. William F. Baxter, Panel Discussion, 54 Antitrust L.J. 31, 33 (1985).
41. ABA Section of Antitrust Law, Commentary on 1992 Horizontal Merger Guidelines, in Antitrust Law Developments 1353 (3d ed. 1992).
42. Id. at 1354.
43. 1982 Guidelines at § III. A. 1(c).
44. See United States Dep't of Justice, Statement to Accompany Release of 1984 Merger Guidelines (June 14, 1984), reprinted in 4 Trade Reg. Rep. (CCH) ¶ 13,103.
45. Id. at § 2.
46. 1992 Guidelines at § 1.52.
47. 1992 Guidelines at § 0.1.
48. 1992 Guidelines at § 2.2.
49. See e.g., Jerry A. Hausman, Empirical Industrial Organization Roundtable, supra note 33, at 36 ("One thing with these merger simulations . . . [is] that they always show prices going up . . . [so that] in a sense they give an upper bound . . .").
50. ABA Section of Antitrust Law, supra note 41 at 1353, n.6.
51. 1982 Guidelines at § V. A.
52. United States Dep't of Justice, supra note 44.
53. 1984 Guidelines at § 3.5.
54. 1992 Guidelines at § 4.
55. Robert Pitofsky, Overview [of Merger Analysis in the 90's: The Guidelines and Beyond], 61 Antitrust L.J. 147, 148-49 (1992).
56. Id. at 148 ("In my view, the worst aspect of merger enforcement in the 1980s was the way the enforcement agencies and the courts were willing to buy stories about entry -- "Oh, yes, we'll become a monopoly and we might raise our price, but someone (never mind exactly who) will go into business, start shipping into our market, and defeat the price increase." I think that has been the preeminent weakness of U.S. merger enforcement in recent years. The ability to tell a story about possible entry based on "soft" evidence should not deter government action and should not influence the courts.").
57. See United States v. Waste Mgmt., Inc., 743 F.2d 976 (2d Cir. 1984); United States v. Baker Hughes Inc., 908 F.2d 981 (D.C. Cir. 1990); United States v. Syufy Enters., 903 F.2d 659 (9th Cir. 1990). See also ABA Section of Antitrust Law, Commentary, supra note 41 at 1361-62.
58. 1992 Guidelines at § 3.0; see also ABA Section of Antitrust Law, Commentary, supra note 41 at 1362-64.
59. 1992 Guidelines at § 3.0.
60. See discussion infra § II(G).
61. 1984 Guidelines at § 4.2.
62. See, e.g., Michael H. Riordan & Steven C. Salop, Evaluating Vertical Mergers: A Post-Chicago Approach, 63 Antitrust L.J. 513 (1995); David Reiffen & Michael Vita, Comment: Is There New Thinking on Vertical Mergers?, 63 Antitrust L.J. 917 (1995); Michael H. Riordan & Steven C. Salop, Evaluating Vertical Mergers: Reply to Reiffen and Vita Comment, 63 Antitrust L.J. 943 (1995).
63. See, e.g., Baxter, supra note 34, at 16.
64. 1992 Guidelines at § 1.32.
65. Id. at § 3.0.
66. Even those most sympathetic to economic analysis have long noted problems of this kind. See, e.g., Ernest Gellhorn, The Practical Uses of Economic Analysis: Hope vs. Reality, 56 Antitrust L.J. 933, 945 (1987)("Even more important are the practical limits of economic information and thus of economic analysis."). See also Thomas B. Leary, supra note 33.
67. My personal preference would be to apply a test, based on (1) the history of entry or non-entry in the market(s) involved and (2) whether there are special circumstances that would affect the probative and predictive value of this history.
68. See also Malcolm B. Coate, Merger Enforcement at the Federal Trade Commission in Three Presidential Administrations, 45 Antitrust Bull. 323 (2000). Coate reaches similar conclusions using different data and methodology.
69. Because of differences in the manner that historical data was reported, the data in Table I are provided by calendar years for 1981 to 1985 and by fiscal year (beginning October 1) for 1986 to 2000. Thus, data for the last three months of calendar year 1985 are also included in fiscal year 1986.
70. Enforcement actions or "challenges" include transactions that were abandoned after a second request was issued.
71. As explained in the notes to the tables, the year 2001 was excluded because of a major change in the filing criteria.
72. The only non-Clinton appointee who served in 1998-2000 was Mary Azcuenaga who resigned as of June 1998.
73. See HSR Annual Reports to Congress, reprinted in Stephen M. Axinn, et al., 2 Acquisitions Under the Hart Scott Rodino Antitrust Improvements Act (Revised)(1988, updated 2000).
74. Changes in the regulatory environment for other industries could also distort statistical comparisons. For example, the unsuccessful objections of the DOJ to airline mergers in the 1980s would not be included as an enforcement action; a direct challenge in the 1990s (e.g., Continental/Northwest) would be included. Some deregulatory deals involving vertical issues (i.e., local monopolies combining with competitive upstream providers) resulted in enforcement activity in the 1990s and would very likely also have been challenged in the 1980s, were they possible.
75. A less comprehensive tweaking of the rules in 1996 was then forecasted to reduce filings by 7-10%, which could have a smaller upward impact on enforcement percentages.
76. Thomas G. Krattenmaker and Robert Pitofsky, Antitrust Merger Policy and the Reagan Administration, 33 Antitrust Bull. 211, 228 (1988)("One reason for our rather harsh judgment is the simple fact that, in the midst of an unprecedented wave of mergers, this [Reagan] Administration has challenged extremely few of them.").
77. Ralph Nader, Burden of Proof, CNN (Aug. 9, 2000)("Both parties are terrible on antitrust. . . . They've allowed mergers under Clinton of the giant HMOs, the giant hospital chains, the giant telecommunication companies. Their one bright light is . . . Microsoft. And I believe antitrust laws police anti-competitive practices and allow competition, especially small business competition . . .").
78. The Commission employed 403 FTE for its competition mission in fiscal year 1989 and 472 FTE in fiscal year 2001 during the merger wave. The Antitrust Division of the Department of Justice employed 508 FTE in fiscal year 1989 and 771 FTE in fiscal year 2001.
79. The criticism of the antitrust agencies in the late 1980s may have reflected some spillover concerns about "hostile takeovers" and "leveraged buyouts," that were prominently publicized during this period of time. These transactions, however controversial, usually did not raise antitrust issues, by any standard. Note that if they were particularly numerous, it would tend to have a downward effect on enforcement percentages in the late 1980s, but I do not know what the facts are.
80. As a former counselor, my impressions from a personal non-random, non-scientific sample is that the agencies were relatively more tolerant of entry stories in the 1980s and relatively more tolerant of business justifications at high HHI levels in the 1990s - - consistent with the overall thrust of their guidelines at these times.
81. See, e.g., Christopher J. MacAvoy, Excerpts of the Transcript From the Roundtable Conference (Jan. 11, 2000), in ABA Section of Antitrust Law, Perspectives on Fundamental Antitrust Theory 387, 405 (2001)("what the agencies actually do is a very nuanced and rich analysis. But when you sit down . . . [and] read the agency's brief that they submit to the court, you don't see nuance, you see . . .Philadelphia National Bank"). But see 1984 Guidelines at § 1 ("The Guidelines are designed primarily to indicate when the Department is likely to challenge mergers, not how it will conduct the litigation of cases that it decides to bring.").
82. Texaco Inc., 104 F.T.C. 241 (1984)(Texaco's acquisition of Getty Oil Co., combining sixth and tenth largest producers of crude oil);Chevron Corp., 104 F.T.C. 597 (1984)(Chevron's acquisition of Gulf Corp., combining seventh and eighth largest producers of crude oil).
83. General Motors Corp., 103 F.T.C. 374 (1984).
84. See, e.g., Exxon Corp., FTC Docket No. C-3907 (Jan. 26, 2001), available at </os/2001/01/exxondo.pdf>; BP Amoco p.l.c., FTC Docket No. C-3938 (Aug. 25, 2000), available at </os/2000/08/bparco.do.pdf>; Chevron Corp., FTC Docket No. C-4023 (Sept. 7, 2001),available at </os/2001/09/chevtexdo.htm>.
85. General Motors Corp., 116 F.T.C. 1276 (1993)(set aside order).
86. See supra note 80.
87. 798 F.2d 1500 (D.C. Cir. 1986).
88. 868 F.2d 901 (7th Cir. 1989).
89. 807 F.2d 1381 (7th Cir. 1986).
90. The 1968 Merger Guidelines suggested that concerns would arise if a supplying firm with a 10% market share merged with a purchasing firm that had a 6% market share. The 1962 Brown Shoe v. United States decision, 370 U.S. 294 (1962), condemned foreclosure of an even smaller share.
91. See Robert H. Bork, The Antitrust Paradox 224-45 (1978)(foreclosure effects through realignment of customer/supplier relationships could be mitigated "by throwing an industry social mixer" where displaced suppliers and displaced customers could meet). See also Fruehauf Corp. v. FTC, 603 F.2d 345, 352 (2d Cir. 1979)(rejecting reliance on foreclosure alone).
92. The theories articulated in the 1984 Guidelines were (i) increased barriers to entry, (ii) facilitation of collusion, and (iii) regulatory evasion. 1984 Guidelines at § 4.2.
93. See Morse, supra note 26, at 1227.
94. See supra discussion accompanying note 36.
95. See, e.g., Martin Marietta Corp., 117 F.T.C. 1039 (1994); Alliant Techsystems Inc., 119 F.T.C. 440 (1995); Hughes Danbury Optical Systems, Inc., 121 F.T.C. 495 (1996).
96. See, e.g., id.
97. See Morse, supra note 26, at 1239.
98. Eli Lilly and Co., 120 F.T.C. 243 (1995).
99. Id. at 258 (Azcuenaga, Comm'r, dissenting).
100. Merck & Co., FTC Docket No. C-3853, (Feb. 24, 1999)(decision and order), available at </os/1999/9902/9510097merck.do.htm>.
101. See Thomas B. Leary, The Significance of Variety in Antitrust, 68 Antitrust L.J. 1007 (2001).
102. See Richard G. Parker, Global Merger Enforcement, Prepared Remarks before the International Bar Ass'n, Barcelona, Spain (Sept. 28, 1999), available at </speeches/other/barcelona.htm>.
103. America Online, Inc., FTC Docket No. C-3989 (Apr. 18, 2001), available at </os/2001/04/aoltwdo.pdf>.
104. See also Time Warner Inc., 123 F.T.C. 171 (1997)(addressing similar variety issues in the entertainment business).
105. See, e.g., Michael L. Weiner, Antitrust and the Rise of the Regulatory Consent Decree, 10 Antitrust 4 (Fall 1995); Harry First, Is Antitrust "Law"?, 10 Antitrust 9 (Fall 1995); A. Douglas Melamed, Antitrust: The New Regulation, 10 Antitrust 13 (Fall 1995); Michael Malina, Some Thoughts on the Source of Antitrust Law in the Nineties, 63 Antitrust L.J. 853 (1995); Richard M. Steuer, Counseling Without Case Law, 63 Antitrust L.J. 823 (1995).
106. Thomas B. Leary, Three Hard Cases and Controversies: The FTC Looks at Baby Foods, Colas and Cakes, Prepared Remarks Before the Association of the Bar of the City of New York's Milton Handler Annual Antitrust Review, New York, NY (Dec. 4, 2001),available at </speeches/leary/learythree.htm>.