Antitrust and the Regulatory Enterprise ANTITRUST AND THE REGULATORY ENTERPRISE Herbert Hovenkamp* I. Introduction ................................................................ 335 II. The Place of Antitrust in Regulated Markets ........... 341 III. Many Sources of Regulatory Policy but a Single Antitrust Concern ...................................................... 343 A. Federal Regulatory Immunity ............................. 344 B. "State Action" Immunity ..................................... 346 C. Noerr-Pennington Petitioning Immunity ........... 348 D. A Unified Rule for Antitrust Regulatory Im m unity ............................................................. 350 IV. Deregulation-Politics and Economics ...................... 354 A. Traditional Economics and Deregulation ........... 360 B. Shifts in Competition Theory .............................. 364 V. Antitrust under Deregulation: Telecom m unications .................................................. 366 V I. C onclusion .................................................................. 377 I. INTRODUCTION Since before the Middle Ages economic policy in western society has gone through cycles of greater and lesser confidence in markets. When confidence is low, the government intervenes with everything from price regulations and output limitations to licensing restrictions on new firms.1 Often economists are not very sensitive to this very long and cyclical history. They write as if today we * Ben V. & Dorothy Willie Professor of Law and History, University of Iowa. 1 See RICHARD H.K. VIETOR, CONTRIVED COMPETITION: REGULATION AND DEREGULATION IN AMERICA (1994) (arguing that prosperity and economic growth generally produce strong confidence in markets and a political bias against regulation, while recessions and depressions produce political momentum for regulation). COLUMBIA BUSINESS LAW REVIEW finally have the right answers.2 But just as certainly as the strongly pro-market period of classical political economy was followed by Progressivism in the 1920s and the New Deal in the 1930s, so too the strong free market ideologies of today will someday yield to renewed interest in regulation. This cyclical history inclines people to view antitrust and regulation as competing models for determining the appropriate scope of state intervention in the micro- economy.3 At the margin they certainly are competing, because we are never certain about where the boundary lies. However, a better way to view the two enterprises is as complementary products. We live in a world in which the great majority of markets clear at efficient, or something close to efficient, levels of output. Even now though, a few markets are exceptional. Because of serious imperfections- chiefly imbalances of information, free rider problems, and significant scale economies-some markets fail to clear at efficient levels, at least within the time frame that our government thinks is appropriate. At some level even the most hard core neoclassicists concede that some markets are efficient only if the state intervenes. For example, while extreme free marketers might rail at the excesses of regulation or antitrust, they tend to accept the system of intellectual property ("IP") rights as if it were handed from a mountaintop. In fact, however, the existing IP system is a very elaborate effort to correct a market failure, in this case free riding that occurs when innovations are too freely copied, and the corresponding decrease in the incentive to innovate. Anyone who does not believe that the IP laws are a form of regulation has not read the Patent, Lanham, or Copyright Acts and the maze of technical rules promulgated under them. To be sure, IP laws create property rights. But so do 2 See Judge Posner's lament that economic history has largely disappeared from the study of economics. RICHARD A. POSNER, OVERCOMING LAW 411 (1995). ' See Jim Chen, Regulatory Education and its Reform, 16 YALE J. ON REG. 145, 147-148 (1999) (book review); JAMES WILLARD HURST, LAW AND MARKETS IN UNITED STATES HISTORY 21 (1982). [Vol. 2004 No. 2:335] ANTITRUSTAND THE REGULATORY ENTERPRISE 337 state created exclusive franchises and filed tariffs. In fact, the detailed regulatory regimes that we call the IP laws are filled with very rough guesses about the optimal scope of protection-ranging from the duration of patents and copyrights to the scope of patent claims and fair use of copyrighted material. The range of government estimation that goes on in the IP system is certainly as great as in regulation of, say, retail electricity or telephone service. Further the IP regime is hardly immune from the legislative imperfections that public choice theory uncovers.4 A good example is the Copyright Term Extension Act, recently upheld by the United States Supreme Court.5 It is hard to come up with a serious argument that retroactive extensions of old copyrights, designed mainly to protect Disney's rights in Mickey Mouse, serve any reasonable purpose in furthering innovation. Further, the IP system invites and has produced many modifications of antitrust doctrine and invitations to create new doctrine. One example is the recent circuit split in cases involving settlement agreements to infringement ' Robert P. Merges, One Hundred Years of Solicitude: Intellectual Property Law, 1900-2000, 88 CAL. L. REV. 2187, 2223 (2000). As one writer described the extent of Congress' capture by entertainment and related interests: Copyright interest groups hold fund raisers for members of Congress, write campaign songs, invite members of Congress (and their staff) to private movie screenings or sold out concerts, and draft legislation they expect Congress to pass without any changes. In the 104th Congress, they are drafting the committee reports and haggling among themselves about what needs to be in the report. In my experience, some copyright lawyers and lobbyists actually resent members of Congress and staff interfering with what they view as their legislation and their committee report. With the 104th Congress we have, I believe, reached a point where legislative history must be ignored because not even the hands of congressional staff have touched committee reports. William F. Patry, Copyright and the Legislative Process: A Personal Perspective, 14 CARDOZO ARTS & ENT. L.J. 139, 141 (1996). 5 17 U.S.C. § 302(a) (2000); Eldred v. Ashcroft, 123 S. Ct. 769 (2003). COLUMBIA BUSINESS LAW REVIEW suits by pioneer drug manufacturers against the makers of generics .6 The subject of this essay is not the history of markets or regulation,7 but it begins with a fleeting look at a few decades of regulatory regimes and the accommodations that antitrust policy has made to them. Most economic justifications for regulation in the modern era have rested on various theories of "market failure," a term used to describe physical or technological conditions that prevent a market from performing efficiently. The classic market failure is the natural monopoly, or industry in which the cost of service declines as volume increases, all the way up to the market's saturation point. For example, a hard network such as the retail electric power grid can serve additional customers by adding capacity to the existing grid much more cheaply than by producing a second grid, or a nineteenth century railroad track between two towns can accommodate second, third and fourth trains much more cheaply than it would be for other railroad companies to lay their own tracks. In many of these markets the political, or regulatory, solution we have adopted is to permit a single firm to have a monopoly over a 6 Cf. Cardizem CD Antitrust Litig., 332 F.3d 896 (6th Cir. 2003); Valley Drug Co. v. Geneva Pharm., Inc., 344 F.3d 1294 (11th Cir. 2003). See Herbert Hovenkamp et al., Anticompetitive Settlements of Intellectual Property Disputes, 87 MINN. L. REV. 1719 (2003); Herbert Hovenkamp et al., Balancing Ease and Accuracy in Assessing Pharmaceutical Exclusion Payments, 88 MINN. L. REV. 712 (2004). 7 A small sampling of readings in the history of regulation, mainly in the United States, includes JOHN MCMILLAN, REINVENTING THE BAZAAR: THE NATURAL HISTORY OF MARKETS (2002); VIETOR supra note 1; THE REGULATED ECONOMY: A HISTORICAL APPROACH TO POLITICAL ECONOMY (Claudia Goldin & Gary D. Libecap eds.,1994); Herbert Hovenkamp, The Rise of Regulated Industry, in ENTERPRISE AND AMERICAN LAW, 1836-1937 105 (1991); Hebert Hovenkamp, Regulatory Conflict in the Gilded Age: Federalism and the Railroad Problem, 97 YALE L.J. 1017 (1988); Elizabeth Sanders, The Regulatory Surge of the 1970s in Historical Perspective, in PUBLIC REGULATION: NEW PERSPECTIVES ON INSTITUTIONS AND POLICIES (Elizabeth Bailey ed., 1987); THOMAS K. MCCRAW, PROPHETS OF REGULATION (1984); STEPHEN BREYER, REGULATION AND ITS REFORM (1982); HAROLD U. FAULKNER, THE DECLINE OF LAISSEZ FAIRE, 1897-1917 (1951). [Vol. 2004 No. 2:335] ANTITRUST AND THE REGULATORY ENTERPRISE certain portion of the market, but to regulate that firm's prices. To describe these regulatory decisions as political is apt because cost-of-service regulated monopoly is a policy choice. There may be alternative policies that work just as well, or alternative technologies that permit more traditional competitive solutions. For example, while it is very likely cheaper to have a single power grid serving a town, or a single line of tracks stretched between two towns, that does not necessarily entail that a single power company or a single railroad must own and operate those facilities. The grid or the tracks could be owned by the state, as the public highway system is, and then competing carriers could make use of the system subject to some traffic rules that prevent collisions or other conflicts. Alternatively, competing service providers could own and operate the network jointly, agreeing about their own traffic regulations to prevent conflict, but competing on price, quality and other terms of service. Realization of these possibilities accounts for much of the deregulation in the last few decades and also defines a role for antitrust policy. Disentangling networks from monopoly has been one of the great accomplishments of the deregulation movement in the last two decades. Another set of market failures thought to justify regulation is information asymmetries, which are a special subset of market failures. These include markets for complicated products or services that consumers frequently do not understand, markets for financial instruments and risk (such as insurance), or markets where buyers and sellers cannot efficiently meet each other to negotiate a price-such as for taxicabs that must be hailed from the street. The taxicab fare structure is complicated because taxi rides differ widely based on both distance and time. It would be very inefficient for a taxi to advertise its fares on the car doors, and even less efficient for prospective passengers and drivers to negotiate fares each time a cab is hailed. By contrast, limousine services, or cars that are called in advance, often enter long term contracts with COLUMBIA BUSINESS LAW REVIEW businesses to provide airport and other local transportation. They operate under traditional price competition. Over the years economists and policy makers have had widely different opinions about the extent to which various types of information failures are thought to justify regulation. At one extreme is the New Deal position that consumers are extremely vulnerable to any kind of complexity, particularly where products or services are differentiated from one seller to another. As a result, broad government regulation is needed of everything from product quality to advertising to terms of service and perhaps even price. This has always been the nominal justification for extensive government regulation of financial and insurance markets, including banking, transportation and professional services. At the other extreme is what might be termed the Chicago School position that information like any other product is itself subject to competition, which forces firms to produce the optimal amount and quality of it." Firms that fail to inform, or that misrepresent, will be disciplined because customers will not return to them. As a result, extensive regulation is not necessary and the common law of contract and fraud is sufficient to correct most problems. Finally, the gap between regulatory theory and regulatory practice is enormous. 9 We undoubtedly regulate many more markets than require regulation, and we regulate many things that need not be regulated, assuming that the goal of regulation is economic efficiency. The last point is important, however. The neoclassical economic critiques of regulation all rest on the premise that efficient allocation of resources is the only goal of regulatory policy, but policy making in a democratic society has always incorporated numerous and varied goals, some of which are inconsistent with each other. For example, supplying universal service in electricity or telecommunications may 8 George Stigler, The Economics of Information, 69 J. POL. ECON. 213 (1961). 9 On this point, see Steven P. Croley, Theories of Regulation: Incorporating the Administrative Process, 98 COLUM. L. REV. 1 (1998). [Vol. 2004 No. 2:335] ANTITRUST AND THE REGULATORY ENTERPRISE require serving some people at a price below short run marginal cost, and this may require a rate structure that forces other consumers to subsidize them, or we may simply subsidize services to the elderly or the poor. In sum, regulatory policy has to one degree or another incorporated ideas about wealth distribution along with its concerns for allocative efficiency. II. THE PLACE OF ANTITRUST IN REGULATED MARKETS One consequence of regulation is a reduced role for the antitrust laws. When the government makes rules about price or output, market forces no longer govern. To that extent antitrust is shoved aside. A corollary is that as an industry undergoes deregulation, or removal from the regulatory process, antitrust re-enters as the residual regulator. Since our fundamental criterion for determining antitrust immunity in regulated industries is the extent of unsupervised private discretionary conduct, the natural result of deregulation is an increased role for the antitrust laws. In general, the more extreme the deregulation-that is, the more that the market is opened to ordinary competitive forces-the greater the role for antitrust. Newly deregulated markets have thus been one of the few places in the last two decades where antitrust has been a growth industry. As previously indicated, regulation is one of those topics about which economists have disagreed over a very wide range. The differences vary with time, with ideology, and even with the type of training that economists have. More fundamentally, attitudes toward regulation differ with one's opinion about whether regulation should take noneconomic goals into account, such as wealth distribution or protection of diversity. Economic attitudes toward regulation have also cycled over time. From the Middle Ages to the age of the great classical political economists such as Adam Smith and John Stuart Mill, price regulation even of ordinary commodities was relatively common. In sharp contrast the classical period, which characterizes most of the nineteenth century, was dominated by concerns for liberty of contract and free markets, and a corresponding abhorrence of many forms of regulation. 10 Then, beginning during the Progressive Era and escalating during the New Deal, economic policy makers lost much of their confidence in open markets and at one time or another attempted to regulate virtually anything that could be regulated, creating giant bureaucratic agencies in the process. The high point of this regulation was in the 1970s, and since then we have experienced a free market revolution that may be starting to subside. Without speculating about where we are going next, I observe only that this ideological pendulum is likely to swing back and forth indefinitely. The solutions we achieve for regulatory problems are invariably political as well as technical. To the extent they are political, a democracy is unlikely to produce final answers. However, a critical premise for understanding the relationship between antitrust and regulation is that it is not antitrust's purpose to throw itself into this centuries-old policy battle. On the question of which markets should be regulated, and by what means, economists have the initial say and legislatures have the final say. It is not the role of federal judges or antitrust juries to decide whether price or entry regulation in the provision of electricity or taxi service is appropriate. Antitrust law takes a market's regulatory structure as given, warts and all, and tries to prevent injuries to competition that the regulatory process leaves untended. Competing with a regulatory regime or substituting its judgment for that of government officials is not antitrust's purpose. This is why we say that antitrust's role is "residual." It picks up only where regulation leaves off. '0 See HERBERT HOVENKAMP, ENTERPRISE AND AMERICAN LAW, 1836- 1937 (1990); WILLIAM J. NOvAK, THE PEOPLE'S WELFARE: LAW & REGULATION IN NINETEENTH-CENTURY AMERICA (1996). [Vol. 2004COLUMBIA BUSINESS LA W REVIEW No. 2:335] ANTITRUST AND THE REGULATORY ENTERPRISE 343 III. MANY SOURCES OF REGULATORY POLICY BUT A SINGLE ANTITRUST CONCERN In the United States regulation comes from three different levels. Federal, state, and local governments all regulate. Sometimes all three regulate the very same markets (such as telephone service, land use, and housing discrimination). Even though antitrust law is predominantly federal, it does not run roughshod over state and local regulation with a broad preemption sweep. Many of the constraints imposed by state price regulation of electricity, locally regulated zoning or taxicab fares, licensing requirements for sale of alcoholic beverages, or state regulation of the insurance industry would be antitrust violations if imposed by private parties.1 But antitrust leaves them alone. If a state or local regulation is valid, federal antitrust is the residual regulator as to that provision, just as it is to federal regulation. There has been an unhelpful tendency in antitrust to use differing doctrinal formulations depending on the level of sovereign imposing the regulation. To be sure, the issues are not precisely the same. When the relevant regulatory regime is federal, such as for sales of corporate securities, 2 much of telecommunications, or interstate energy regulation, the courts try to harmonize two potentially conflicting federal regulatory regimes. For example, we would not presume that Congress intended federal regulation of securities by the Securities Exchange Commission to conflict with federal antitrust regulation. While inconsistent regulatory requirements do arise, the results are chaotic and inefficient, and we would not presume that Congress intended them. '1 See, e.g., 1 PHILLIP E. AREEDA & HERBERT HOVENKAMP, ANTITRUST LAW 221-227 (2d ed. 2000). 12 See Herbert Hovenkamp, Antitrust Violations in Securities Markets, 28 J. CORP. L. 607 (2004). See, e.g., Friedman v. Salomon/Smith Barney, Inc., 313 F.3d 796 (2d Cir. 2002), cert. denied, 124 S. Ct. 152 (2003); MFS Sec. Corp. v. N.Y. Stock Exch., Inc., 277 F.3d 613, 621 (2d Cir.) (2002), cert. denied, 536 U.S. 924 (2002). State and local government regulation emanates from an inferior level of government. Under the Supremacy Clause of the Constitution federal antitrust could preempt conflicting state regulation, just as the federal law of labor relations or corporate securities preempts inconsistent state law. But that is not what has happened. A basic tenet of federalism is that states are entitled to regulate their own internal economies, and also to decide how much regulatory authority to cede to their municipalities, counties, or other units of local government. 13 As a result, federal courts addressing potential conflicts between state or local regulation and the antitrust laws ask a series of questions that are somewhat similar to the ones they ask when the regulation is federal. I say "somewhat similar" because largely by historical happenstance we have created different verbal formulations for dealing with different types of regulatory conflict. Antitrust would be tidier and could produce a more satisfactory accommodation with regulation if these rules were simplified and unified. The relation between federal antitrust and various regulatory regimes is covered by three sets of antitrust rules: the doctrine of federal regulatory immunity, the "state action" exemption, and the Noerr- Pennington immunity for petitions to any level of government. A. Federal Regulatory Immunity Federal regulatory immunity comes in two kinds, express and implied. When the antitrust immunity is express, the role of the court is mainly to ensure that the statute's requirements have been satisfied. For example, a labor strike would be an unlawful price-fixing agreement among employees but for an express antitrust immunity created by 13 See Parker v. Brown, 317 U.S. 341 (1943) (state regulation); Cmty. Communications Co. v. City of Boulder, 455 U.S. 40 (1982) (municipal regulation); 1 AREEDA & HOVENKAMP, supra note 11, 221. [Vol. 2004COLUMBIA BUSINESS LA W REVIEW No. 2:335] ANTITRUSTAND THE REGULATORY ENTERPRISE federal legislation. 4 A federal statute called the Health Care Quality Improvement Act creates an antitrust exemption permitting hospital boards to discipline or dismiss physicians for acting improperly, provided they comply with the statute's procedural requirements. 5 Implied federal immunity is a more difficult concept. It exists when the statute in question does not expressly confer an antitrust immunity, but immunity seems necessary if we are to avoid conflicts between regulatory and antitrust requirements. In general, courts try to ensure that application of the antitrust laws does not interfere with the agency's ability to do its proper job. In addition, they are more likely to find an immunity if the agency itself considered competitive concerns in making its decision. This is not always the case. For example, a government regulation may be concerned entirely with safety or the financial solvency of regulated firms, and agency supervision may largely ignore competitive concerns. By contrast, where the regulatory regime controls such things as pricing, entry by new firms, joint market behavior, mergers, or possible exclusionary practices, antitrust immunity may be essential if regulatory goals are not to be frustrated. The courts typically find implied immunity if they believe that an antitrust suit would interfere with the agency's operations, or if the agency has had a matter under study, even though it may not have reached a resolution. For example, in the Options Trading case recently decided by the Second Circuit, the court granted antitrust immunity to an agreement among the major stock exchanges forbidding a particular stock option from being listed on more than a single exchange. 6 The Securities Exchange Commission had been studying the problem of options listing for many years, approving them during some periods and disapproving them 14 15 U.S.C. § 17 (2000). See 1A AREEDA & HOVENKAMP, supra note 11, 255-57. 15 42 U.S.C.A. §§ 11,101-11,152. See 1A AREEDA & HOVENKAMP, supra note 11, 250. 16 In re Stock Exchs. Options Trading Antitrust Litig., 317 F.3d 134 (2d Cir. 2003). COLUMBIA BUSINESS LAW REVIEW during other periods after concluding that multiple listing led to excessive chaos in securities sales. 7 An immunity was proper even though the SEC had been indecisive. If a group of experts studying an issue for many years could not come to a clear understanding, then it would be imprudent to permit the issue to be decided by a jury trial in an antitrust case. Doing so would render agency regulation moot on that point. To be sure, reasonable people might believe that the issue should never have been turned over to an agency to begin with, but that complaint is not for antitrust to answer. B. "State Action" Immunity Regulation by state and local government is not only pervasive, but it is also probably more susceptible to political influences than federal regulation is. States and local governments regulate residential rents, liquor pricing, intrastate trucking rates, insurance, and taxi fares. They grant exclusive rights for municipal waste disposal, cable television, billboards, and ambulance service." Many of these regulations would not constitute antitrust violations even if there were no immunity, but some supposedly would. The concept of state action under federal antitrust law is very different from the Fourteenth Amendment rule that various Constitutional protections apply only to state action. The Fourteenth Amendment concept is broad and refers to all state and local government decision making, and even to 17 Id. at 146-148. 18 See, e.g., Fisher v. City of Berkeley, 475 U.S. 260 (1986) (upholding rent control ordinance); California Retail Liquor Dealers Ass'n. v. Midcal Aluminum, Inc., 445 U.S. 97 (1980) (liquor pricing); Liquor Corp. v. Duffy, 479 U.S. 335 (1987) (posting of liquor prices); S. Motor Carriers Rate Conference, Inc. v. United States, 471 U.S. 48 (1985) (intrastate trucking); Town of Hallie v. City of Eau Claire, 471 U.S. 34 (1985) (sewage treatment); Cmty. Communications Co. v. City of Boulder, 455 U.S. 40 (1982) (cable television); Campbell v. City of Chicago, 823 F.2d 1182 (7th Cir. 1987) (taxicab fares); City of Columbia v. Omni Outdoor Adver., 499 U.S. 365 (1991) (billboards); Ambulance Serv. of Reno v. Nevada Ambulance Servs., 819 F.2d 910 (9th Cir. 1987) (ambulance); FTC v. Ticor Title Ins. Co., 504 U.S. 621 (1992) (insurance). [Vol. 2004 No. 2:335] ANTITRUST AND THE REGULATORY ENTERPRISE public officials acting under color of state law when they engage in actions such as race discrimination. In contrast, antitrust state action refers only to the policies of the state acting as sovereign. Regulatory actions that are not properly authorized by the state are simply not immune. The antitrust state action doctrine can be reduced to two requirements, generally called authorization and supervision. Under the first requirement, the state must have "clearly articulated" and "affirmatively expressed" its wish to displace ordinary competitive processes with some form of regulation covering the challenged conduct. Second, if the conduct of private parties is at issue, and not merely that of the government itself, then the conduct must be "actively supervised" by some kind of state agency or official. For example, the state could not simply pass a statute authorizing building contractors to fix prices and then leave them free to do so entirely on their own.19 The state action immunity is a creature of federalism, not of federal regulatory policy. Its purpose is not to protect federal regulatory or competition goals, but to give appropriate recognition to state regulatory power. So when a court applies the state action doctrine it must try to avoid making substantive judgments about whether the state regulation at issue is a good idea. Suppose a state with a powerful potato growers' lobby wants to regulate the price of potatoes even though every economist in the country is willing to testify that using regulation to determine the price of potatoes is a terrible idea. If the state passes a statute that unambiguously calls for potato price regulation, and a potato agency created for this purpose actively administers the rate-making process, then the antitrust court has no choice but to find the regulation immune from the antitrust laws. Congress could of course intervene if it wants to, but 19 On the two requirements, see California Retail Liquor Dealers Ass'n., 445 U.S. 97 (1980). The authorization requirement was developed more fully in S. Motor Carriers Rate Conf., 471 U.S. 48, and the active supervision requirement in Ticor Title Ins. Co., 504 U.S. 621 and Patrick v. Burget, 486 U.S. 94 (1988). See 1 AREEDA & HOVENKAMP, supra note 11, T 224-225 (authorization); id. [ 226-227 (supervision). COLUMBIA BUSINESS LAW REVIEW the state action immunity doctrine presumes that it has chosen not to do so. Anticompetitive special interest regulations have often been immunized from antitrust challenge by the state action doctrine.2 ° While this is disconcerting as a matter of policy, the more important principle is that correcting flaws in political processes is not an antitrust task. We would not interpret the Sherman Act to give district court juries a wide-ranging mandate to ensure that government regulation produces only efficient, or competitive, results-certainly not without a very clear indication that Congress had this in mind. There is no evidence in the legislative history of the antitrust laws that Congress was concerned with anything other than privately created restraints. Further, juries of laypersons are particularly ill suited to determine the proper boundaries of regulatory intervention. At best, they could sort out the worst abuses, such as those that are a result of bribery or other corruption of the political process. But we already have plenty of laws that reach these practices. C. Noerr-Pennington Petitioning Immunity The Noerr-Pennington petitioning immunity is undergirded by the First Amendment guarantee that Congress will "make no law.., abridging.., the right of the people... to petition the Government."21 The basic antitrust immunity was defined in two Supreme Court decisions holding that citizens have a right to ask the government for regulation, even if the regulation that they want is anticompetitive. The Noerr case held that the railroad industry had a right to seek state legislation that would impose high costs and other obstacles on truckers.2 The Pennington case held that a labor union and mechanized coal 20 E.g., Omni Outdoor Adver., 499 U.S. 365 (immunizing city council's passage of regulatory provision that favored signs of politically well- connected business owner while excluding those of competitor). 21 U.S. CONST. amend. I. 22 R.R. Presidents Conference v. Noerr Motor Freight, Inc., 365 U.S. 127 (1961). [Vol. 2004 No. 2:335] ANTITRUSTAND THE REGULATORY ENTERPRISE 349 mines had a right to petition the government for a higher minimum wage for companies seeking to sell coal to the government. 3 This rule made it difficult for higher cost, unmechanized mines to compete. Subsequent decisions extended the petitioning immunity to judicial petitions.24 Noerr-Pennington immunity relates to government regulation in two general ways. First, regulation is often the result of interest group pressures, and Noerr-Pennington protects the right of these interest groups to request virtually any kind of regulation they wish. The Supreme Court has immunized requests from industry participants to permit them to cartelize their markets, and requests from one manufacturer to make its rival's product unlawful or to place costly regulatory burdens on it. 25 The second way that the Noerr-Pennington doctrine relates to regulation is through the litigation and administrative process. The doctrine permits people or firms to file lawsuits in a court or complaints before a regulatory agency, even though their motives are anticompetitive. The petitioning immunity is extremely broad, but it does come with one strong exception. There is no right to petition if the petition itself is a sham-that is, if the petition is not intended to obtain from the government a response favorable to the petitioner, but only to harass or suppress a rival. The classic example is the baseless lawsuit. For example, a dominant firm might file patent infringement suits against every new entrant into its market, even though it knows the suits are invalid. The dominant firm intends to win, not 23 United Mine Workers v. Pennington, 381 U.S. 657 (1965). 24 E.g., Profl Real Estate Investors, Inc. v. Columbia Pictures Indus., 508 U.S. 49 (1993). See 1 AREEDA & HOVENKAMP, supra note 11, 205. 2 In addition to the Noerr and Pennington cases cited above, see Omni Outdoor Advert., 499 U.S. 365, which found antitrust immunity when a politically well-connected firm that sold billboards used its political influence to get an ordinance passed that excluded its rival's billboards from the market. 26 A good, relatively recent example is Kottle v. Northwest Kidney Cntrs., 146 F.3d 1056 (9th Cir.1998), cert. denied, 525 U.S. 1140 (1999). COLUMBIA BUSINESS LA W REVIEW with a courtroom victory, but by imposing high litigation costs on a smaller rival.27 In addition to these three doctrines of regulatory immunity is a fourth approach that avoids immunity jargon altogether. In the Town of Concord case, Judge Breyer examined the threat to competition posed by a claimed antitrust violation and considered whether the regulatory process was sufficient in that case to make an anticompetitive outcome unlikely.28 He concluded that the threats of price manipulation brought by municipal utilities against a large power wholesaler were not very credible. In fact, the plaintiffs could have obtained power from alternative sources had they wanted to, so their claim that they were being squeezed by the defendant's high wholesale prices was relatively weak. Further, the regulatory process in place was highly unlikely to permit anticompetitive consequences of the type that the plaintiffs claimed. D. A Unified Rule for Antitrust Regulatory Immunity Concerns for federalism certainly compel us to allow ample room for state and local regulation. At the same time, the Supremacy Clause may force broader deference to federal than to state regulation. As a result, the scope of power to regulate and the proper domain of regulation will always be somewhat different for the states than for the federal government. But this hardly dictates that federal antitrust policy treat these forms of regulation differently in determining the reach of the antitrust laws into regulated industries. The antitrust immunity doctrines come into play only after the tribunal has decided, or at least assumed, that the regulation in question is valid under both federal and state law. Once we know that a regulation is valid, then insofar as antitrust 27 On "sham" petitioning, see California Motor Transp. Co. v. Trucking Unlimited, 404 U.S. 508 (1972); 1 AREEDA & HOVENKAMP, supra note 11, 1 204-208. 28 Town of Concord v. Boston Edison Co., 915 F.2d 17 (1st Cir. 1990), cert. denied, 499 U.S. 931 (1991). [Vol. 2004 No. 2:335] ANTITRUST AND THE REGULATORY ENTERPRISE is concerned it should not matter very much whether the market interference in question emanates from the federal government or the states. Further, just as correcting defects in the federal political process is not antitrust's task, nor is antitrust the proper tool for oversight of state or local political processes. While the technical rules for these various immunity doctrines sound quite different from one another, all of them can be reduced to a single set of principles. The first is that antitrust is concerned about the private, discretionary exercise of market power, not with government decision making. So what we really want to know is whether a private actor is causing competitive harm without adequate government oversight of what it is doing. When the harm to the plaintiff is caused by the government itself, there is no antitrust claim. Nor is there an antitrust violation when the harm is caused by a private party, but only after that party's conduct has been reviewed and approved by a disinterested government agency, or the agency is active and has the power to review the conduct in question. The last clause is essential, for the immunity must exist both before and after the agency has acted. Otherwise people would simply go to antitrust courts and bypass the agency. The relevant question is whether a regulatory regime is in place that either has, or is likely to, review and control the challenged conduct. No immunity is granted if the regulatory agency lacks jurisdiction over the practice, has jurisdiction but is not authorized to take competition concerns into account, or simply rubber stamps what the private firms want. Importantly, however, this same set of queries defines the scope of the state action immunity. What we really want to know is whether the state authorized the challenged conduct by passing a statute that mandated it, contemplated that it would occur, or specifically permitted it. If the conduct is of the state or a government subdivision itself, that is the end of the inquiry. However, if the conduct is of a private party, then we also need to know whether a government agency or official "actively supervised" the conduct. Finally, the Noerr-Pennington inquiry immunizes any petition to the government where the ultimate harm is caused by the government action itself. If I petition the city council for an ordinance approving my waste disposal facility and disapproving yours, and the council passes this ordinance, then you have no antitrust claim against me. On the other hand, if I know that my petition to the government is frivolous, but I file it only because I know that you lack the resources to defend against it, then your harm is not caused by the government's decision but by my filing itself. In that case the injury to competition results from private conduct that has not been effectively supervised.2 9 In sum, a single rule for antitrust immunity, no matter what its source, requires two things. First, the regulatory regime must be lawful and have jurisdiction over the conduct that is the subject of the antitrust complaint. "Jurisdiction" means the authority to evaluate conduct, including its competitive consequences, and approve or disapprove it. Second, if challenged private conduct is discretionary-that is, if a private firm could have done something a different way that causes less competitive harm-that conduct must either have been reviewed and approved by the agency or must be under ongoing study, or the agency must have manifested its ability and will to evaluate the conduct if asked. This outcome is completely consistent with a proposal made by Timothy J. Muris, Chairman of the Federal Trade Commission ("FTC"), that would narrow the scope of the petitioning immunity in areas where government involvement in the outcome is minimal. 0 One area is tariff filings by regulated firms. A regulated firm is often required to file a tariff with a regulatory agency, which is basically a request for a certain rate or other practice. Officially, the 29 See 1 AREEDA & HOVENKAMP, supra note 11, IT 204-208. 30 Timothy J. Muris, Looking Forward: the Federal Trade Commission and the Future Development of U.S. Competition Policy, 2003 COLUM. Bus. L. REV. 358, 368-75. See also 1 AREEDA & HOVENKAMP, supra note 11, 210. [Vol. 2004COLUMBIA BUSINESS LA W REVIEW No. 2:335] ANTITRUSTAND THE REGULATORY ENTERPRISE 353 agency evaluates each tariff filing as part of its regulatory obligations. In reality, such filings are often merely rubber stamped by the agency or go into effect with no agency consideration at all. Another area is so-called "Orange Book" listings of pharmaceutical and related patents with the Food and Drug Administration ("FDA"). While patent extensions for new drug applications must be listed with the FDA in order to have patent protection, the FDA does not actually review the patents.31 In cases such as these the private party rather than the government is the actual decision maker, and these are precisely the types of cases where antitrust immunity is inappropriate." In its Trinko decision the Supreme Court took a first step in this direction.3 The Court held that the value of adding antitrust to a regulatory enterprise must be determined by asking two questions. First, how well is the regulatory enterprise itself doing its job of identifying and controlling competitive harms? The better it is doing, the less incremental value antitrust will provide.34 Second, how much confidence do we have that application of the antitrust laws will improve competition in the situation at hand? If antitrust will be particularly difficult to apply and prone to error, then it would likely do more harm than good in a well-regulated regime.3 The Court then had little 31 See Muris, supra note 30, at 372; In re Buspirone Patent Litig., 185 F. Supp. 2d 363 (S.D.N.Y. 2002); 1 AREEDA & HOVENKAMP, supra note 11, 210c (Supp. 2003). 32 My proposal is also consistent with Mr. Muris' proposal that in "state action" cases courts should look carefully to make sure that the state actually did intend to displace competition, and consequently preclude the application of the antitrust laws, with a particular regulatory scheme, and that the state is in fact actively supervising any private discretionary conduct. Muris, supra note 30, at 377-78. "3 Verizon Communications, Inc. v. Law Offices of Curtis Trinko, LLP, 124 S. Ct. 872 (2004). 3 Id. at 881-82. " Id. at 881 ("One factor of particular importance is the existence of a regulatory structure designed to deter and remedy anticompetitive harm. Where such a structure exists, the additional benefit to competition provided by antitrust enforcement will tend to be small, and it will be less plausible that the antitrust laws contemplate such additional scrutiny."). difficulty concluding that state and federal regulators had been doing an adequate job of supervising interconnection disputes between dominant and competitive telephone companies." Further, the antitrust "essential facility" doctrine that the plaintiffs were invoking would be particularly prone to error. As a result, antitrust was not likely to be worth its costs. In getting to this result the Court did not rely on any of the established immunity doctrines. 7 Significantly, local telecommunications are regulated by both federal and state regulatory agencies. In speaking of the effectiveness of regulation, the Court spoke of the two interchangeably, and there was no reason for it not to. Its concern was not with the source of the regulation for its own sake, but rather whether the regulation was doing an adequate job of addressing the competitive concerns at hand. IV. DEREGULATION-POLITICS AND ECONOMICS The 1960s and 1970s were the high point of federal regulation, whether measured by the number of industries that were regulated, the extent of the regulation, or our policy makers' overall confidence that government regulation allocated resources better than ordinary market forces. But beginning in the late 1970s and extending to the present day, the federal government has undergone comprehensive 36 Id. at 882 ("The regulatory response to the OSS failure complained of in respondent's suit provides a vivid example of how the regulatory regime operates. When several competitive LECs complained about deficiencies in Verizon's servicing of orders, the FCC and Public Service Commission ("PSC") responded. The FCC soon concluded that Verizon was in breach of its sharing duties under § 251(c), imposed a substantial fine, and set up sophisticated measurements to gauge remediation, with weekly reporting requirements and specific penalties for failure. The PSC found Verizon in violation of the PAP even earlier, and imposed additional financial penalties and measurements with daily reporting requirements. In short, the regime was an effective steward of the antitrust function."). 3' The Court read the Antitrust Saving Clause in the 1996 Act as precluding a federal regulatory immunity. See id. at 878. COLUMBIA BUSINESS LA W REVIEW [Vol. 2004 No. 2:335] ANTITRUST AND THE REGULATORY ENTERPRISE 355 deregulation to varying degrees in nearly every regulated market." Federal deregulation of specific markets began during the Carter Administration with statutes that deregulated various parts of the trucking and railroad industries. In the passenger airline market, the Civil Aeronautics Board ("CAB") once controlled both new entry of passenger airlines and comprehensively regulated ticket prices. But federal legislation abolished the CAB and opened the market to new entry and deregulated prices. The remaining forms of airline regulation deal mainly with its transportation aspects- routing, safety, pilot work schedules and the like, while leaving price determination and new entry largely to the firms themselves. In 1996 the Interstate Commerce Commission, which regulated mainly trucking and the railroads, was also abolished. 9 In telecommunications, competition now largely controls the long distance portion of the market, although there is still a great deal of regulation of local service. The long-run goal of the Telecommunications Act of 1996, of which more is said later, ° is a fully competitive system covering not only long distance but also local services. Much of deregulation has been a consequence of technological change. Ceteris paribus, both private choice and government policy rightfully prefer technologies that are more readily capable of being sold in competitive markets. The single biggest change has occurred in the "' See Alfred E. Kahn, Deregulation: Looking Backward and Looking Forward, 7 YALE J. ON REG. 325 (1990); Stephen Breyer, Antitrust, Deregulation and the Newly Liberated Marketplace, 75 CAL. L. REV. 1005 (1987); Thomas G. Krattenmaker, Implications of Deregulation for Antitrust Policy, 53 ANTITRUST L.J. 211 (1984). For a brief history, see Richard D. Cudahy, The Folklore of Deregulation (With Apologies to Thurman Arnold), 15 YALE J. ON REG. 427 (1998). On overall effects, see Richard A. Posner, The Effects of Deregulation on Competition: The Experience of the United States, 23 FORDHAM INT'L L.J. 7 (2000). 39 See 49 U.S.C. § 701 (2000) (terminating ICC and transferring its members and some jurisdiction to Surface Transportation Board). 40 See discussion infra Part V. telecommunications industry as a result of the gradual displacement of various portions of the hard-wired telephone network with microwave, satellite, and other wireless forms of communication. While the hard-wired network that continues to dominate local phone service was thought to be a natural monopoly, wireless telecommunication was not. Deregulation's effects showed up first in long distance, where firms such as MCI and Sprint entered competition with AT&T, whose statutory monopoly had once covered all services. Sophisticated computers and switching equipment in both the telephone industry and the electric power industry have permitted rapid transfer of services from one provider to another, facilitating the integration of long distance, local, and cellular telecommunications in one industry; and the competitive wholesaling of electric power in the other.41 But changes in technology tell only part of the story. Beginning in the 1960s the economic rationales for deregulation were developed in two distinctive bodies of literature. The first severely criticized regulatory agencies for being too expensive, for often being "captured" by the very firms they were supposed to be regulating, and for developing solutions that were highly imperfect and inconsistent with efficient allocation of resources. The second body of literature argued that the traditional economic conception of competition was too narrowly drawn, particularly when applied to industries thought to be natural monopolies. 41 On telecommunications, see PETER W. HUBER ET AL., FEDERAL TELECOMMUNICATIONS LAW ch. 1 (2d ed. 1999); Jerry A. Hausman & J. Gregory Sidak, A Consumer Welfare Approach to the Mandatory Unbundling of Telecommunications Networks, 109 YALE L.J. 417 (1997). On electric power, see Michael 0. Wise, Overview: Deregulation and Antitrust in the Electric Power Industry, 64 ANTITRUST L.J. 267 (1996); Lee A. Rau, Open Access in the Power Industry: Competition, Cooperation and Policy Dilemmas, 64 ANTITRUST L.J. 279 (1996); David S. Copeland, Requiring Transmission Access by Electric Utilities: the Shifting Roles of Regulation and Antitrust, 64 ANTITRUST L.J. 291 (1996); Herbert Hovenkamp, The Takings Clause and Improvident Regulatory Bargains, 108 YALE L.J. 801 (1997). COLUMBIA BUSINESS LA W REVIEW [Vol. 2004 No. 2:335] ANTITRUSTAND THE REGULATORY ENTERPRISE 357 A. REGULATION AND PUBLIC CHOICE One characteristic of regulatory thinking from the New Deal through the 1950s was an assumption that regulation served the public interest by protecting consumers from price gouging or other abuses at the hands of powerful firms that the market was unable to discipline. However, it often turned out that the principal beneficiaries of industry regulation were the regulated firms themselves, who were shielded from competition and guaranteed profit margins.42 Regulation also distorted the mix of capital and other resources and the incentives for cost reduction and innovation.43 The first significant political critique of the regulatory process from a government insider came in 1960, in James M. Landis' Report to then President Elect John F. Kennedy." A generation earlier Landis had presented his own, sharply contrasting conclusions in The Administrative Process, which praised the virtues of government regulation for protecting consumers, and for supervising business activities much more efficiently than the courts could supervise them.45 However, the 1960 Report expressed considerable disillusionment over the modern regulatory enterprise, and dismay at the extent of industry orientation shown by the various regulatory agencies. Landis' Report concluded that the Civil Aeronautics Board, the Interstate Commerce Commission, the Federal Power Commission, and the Federal Communications Commission had all taken sides with the regulated firms at the expense of the public 42 Interestingly, this critique was developed both by writers on the political left and by those on the political right. E.g., GABRIEL KOLKO, RAILROADS AND REGULATION: 1877-1916 (1965) (New Left); George Stigler, The Theory of Economic Regulation, 2 BELL J. ECON. & MGMT. SCI. 3 (1971) (Chicago School). 43 E.g., MICHAEL A. CREW & PAUL R. KLEINDORFER, THE ECONOMICS OF PUBLIC UTILITY REGULATION 120-24 (1986); Harvey Averch & Leland L. Johnson, Behavior of the Firm Under Regulatory Constraint, 52 AM. ECON. REV. 1052 (1962). " JAMES M. LANDIS, REPORT ON REGULATORY AGENCIES TO THE PRESIDENT-ELECT (1960). 45 JAMES M. LANDIS, THE ADMINISTRATIVE PROCESS (1938). COLUMBIA BUSINESS LAW REVIEW interest. Further, the cost of operating the regulatory machinery was significantly larger than anticipated.46 In addition, economists and political scientists such as James Buchanan, Gordon Tullock and Mancur Olson began to write about the theory of interest groups, explaining that small, well-organized, homogenous interest groups could reach legislative decision makers much more effectively than larger, heterogenous, and more poorly organized interest groups can. For example, taxicab owners are relatively few and have common interests about which they feel strongly. As a result they find it quite easy to organize a taxicab owners lobby. By contrast, taxicab passengers are extremely numerous and the cost of taxi rides is a relatively small part of each passenger's budget. Further, the costs of developing an effective lobbying organization could be very high in relation to the benefit that individual members could be expected to receive. Their very numerosity plus the small size of the perceived interest inclines each individual taxicab passenger to shirk from heavy political involvement in the taxicab pricing structure, preferring to rely on someone else. The result is that when the regulatory agency collects information and opinion upon which to base regulatory decisions, the taxicab owners speak with a single, well organized and well informed voice. By contrast, the taxicab passengers either fail to speak, or else their spokespersons are less well informed, idiosyncratic, or inconsistent with one another. It should then be no surprise that the legislation ends up favoring cab owners over passengers, even though the interests of cab passengers in the aggregate are far larger. Although there is much variation in both the nature and quality of public choice scholarship, one pervasive theme is that both regulatory legislation and government agency decision making often fail to find the public interest because 46 LANDIS, supra note 44; LANDIS, supra note 45; see MORTON J. HORWITZ, THE TRANSFORMATION OF AMERICAN LAW: THE CRISIS OF LEGAL ORTHODOXY, 1880-1960 (1992). (Vol. 2004 No. 2:335] ANTITRUST AND THE REGULATORY ENTERPRISE 359 their processes are captured by the very groups whom they seek to control." While public choice analysis is often used to support arguments against regulation, public choice is really an argument about why government sometimes makes decisions that favor one particular interest group rather than the interests of society as a whole. While the theory explains socially harmful decisions to regulate, it also explains socially harmful decisions not to regulate.8 Or it explains harmful deregulation just as much as harmful regulation. For example, public choice theory may explain why a small, well organized group of trucking firms can succeed in protecting interstate trucking from competition at the expense of consumers. But it also serves to explain why a small, well-organized group of cigarette or gun manufacturers can prevent Congress from enacting legislation restricting the distribution of smoking materials to minors or holding gun manufacturers accountable for the social cost of widespread access to weapons. Public choice theory can account for bad regulation or bad failures to regulate, but the theory itself tells us little about whether any particular instance of regulation or failure to regulate is socially harmful or beneficial. Another weakness of public choice theory is its lack of empirical robustness. In a world in which public choice fully explained regulatory choices we would expect regulation in markets that satisfied the Buchanan/Tullock requirements for well organized special interest groups. Whether freedom 47 JAMES BUCHANAN & GORDON TULLOCK, THE CALCULUS OF CONSENT (1962); MANCUR OLSON, THE LOGIC OF COLLECTIVE ACTION: PUBLIC GOODS AND THE THEORY OF GROUPS (2d ed. 1971). A good survey of the literature on public choice is DANIEL A. FARBER & PHILIP P. FRICKEY, LAW AND PUBLIC CHOICE: A CRITICAL INTRODUCTION (1991). For specific application to regulation, see George J. Stigler, The Theory of Economic Regulation, 2 BELL J. ECON. & MGT. SCI. 3 (1971); Richard A. Posner, Theories of Economic Regulation, 5 BELL J. ECON. & MGT. ScI. 335 (1974). ' For further development of this point, see Herbert Hovenkamp, Judicial Restraint and Constitutional Federalism: The Supreme Court's Lopez and Seminole Tribe Decisions, 96 COLUM. L. REV. 2213 (1996). from price competition or entry restrictions obtained would have little to do with whether regulation in a market is justified by traditional neoclassical factors. But in the world we live in we find a fairly (although by no means perfect) "neoclassical" correlation. That is, regulation of retail electricity, natural gas, and taxicab service is well nigh universal, even when the choice is left entirely up to the fifty states. By contrast, we do not find such regulation in the markets for office supplies, personal computers, or children's clothing.49 One is hard pressed to believe that retail electricity is price regulated because power companies have the most effective lobbyists or low organization costs, while the grocery industry is competitive because its lobbying is less effective and its sellers cannot organize effectively. Rather, the traditional neoclassical doctrine of natural monopoly probably accounts for the regulatory differences between these markets. Public choice theory can explain at least part of the movement toward deregulation. The relative cost of organizing various types of interest groups may have changed over the years. In particular, mass electronic communications may have made the organization of large diverse groups relatively cheaper and thus enabled them to speak with a stronger political voice. As a result, the balance of political power may have shifted away from investors in such businesses as monopoly utilities and toward their consumers. B. TRADITIONAL ECONOMICS AND DEREGULATION At least in the case of price regulation the traditional theory of natural monopoly seems to do a better job than public choice theory does of predicting which markets are regulated and which are not.5" By the same token, " See Herbert Hovenkamp, Legislation, Well-Being, and Public Choice, 57 U. CHI. L. REV. 63, 99 (1990). "0 Cf. Croley, supra note 9, at 55; Mark Kelman, On Democracy-Bashing: A Skeptical Look at the Theoretical and "Empirical" COLUMBIA BUSINESS LA W REVIEW [Vol. 2004 No. 2:335] ANTITRUST AND THE REGULATORY ENTERPRISE deregulation is much more readily explained by changes in economic theory that have broadened our conception of competition.51 While public choice theory explains how the political process can produce bad regulation, more traditional economics provides substantive reasons for regulation's numerous weaknesses and failures. One fundamental critique of regulation is that the process costs too much in proportion to any benefits it produces, and that bad results are magnified by chronically inadequate commitment of resources. Regulation is typically very expensive. Further, the sizeable budgets of the various regulatory agencies account for only a small percentage of the costs. The cost of regulatory compliance by firms required to file a tariff describing every proposed change in price or product, and perhaps dispute or even litigate the proposal with the agency itself or various private opponents is far larger. One consequence of high administration costs coupled with disappointing results is that legislatures chronically budget too little for regulatory decision making. As a result, the agencies do an even poorer job, often leaving large portions of the regulated market open to de facto private discretion. In extreme cases so-called "regulation" becomes nothing more than legalized private collusion.52 An equally damaging critique is that, while regulation seeks to approximate competitive behavior, the various formulas it applies distort incentives and reward inefficiency. 3 The traditional approach to cost-of-service rate making, as in electric power, is a prime example. Under competition, prices are forced to marginal cost and each firm is forced to operate as efficiently as possible so that it will Practice of the Public Choice Movement, 74 VA. L. REV. 199 (1988); Daniel A. Farber & Philip P. Frickey, The Jurisprudence of Public Choice, 65 TEX. L. REV. 873, 895-900 (1987). "' See discussion infra text accompanying note 49 (discussing theory of contestable markets). 52 E.g., FTC v. Ticor Title Ins. Co., 504 U.S. 621 (1992). 53 See W. VISCUSI ET AL., ECONOMICS OF REGULATION AND ANTITRUST, chs. 10-12 (3d ed. 2000). not lose market share to rivals. Rate of return regulation distorts these incentives. Prices cannot be set at short-run marginal cost because that figure is too low to provide a firm with a reasonable return on its very large capital investment. As an alternative, the regulatory agency generally permits the regulated firm to recover a fair rate of return on its capital investment; and then permits it to pass through its variable costs, such as labor and fuel. But this formula distorts basic incentives. To the extent the firm receives its fair rate of return on both efficient and inefficient investments, it has no incentive to distinguish the two. The regulated firm tends to prefer capital intensive technology, which enlarges the base upon which its rate of return is computed. An ordinary competitive firm grows only to the point that incremental gains match incremental costs, but the classically price regulated firm never reaches that point because it is promised a fair rate of return on all its investment. Therefore, there is always a positive return to expansion.5 4 Finally, rate-of-return regulation undermines the incentive to innovate. Indeed, to the extent that cost savings reduce the rate base, the incentive to innovate can be negative. For example, if an older switch costs $10,000 while a newer replacement could be innovated that would do the same work but cost only $8000, the competitive firm develops and deploys the switch, and profits by pocketing the cost difference. By contrast, the regulated firm experiences only a decline in its rate base. Under the old technology it earns its fair rate of return on $10,000, while under the new technology it earns that same rate of return, but only on an $8000 base. Carefully designed regulatory policy partially controls these distortions, and close scrutiny by the regulatory agency prohibits at least some of the over investment. But the compensation is never complete, and careful agency oversight is expensive. As a result comprehensive price and service rate making offers, at best, a very rough approximation of the competitive price, deters innovation, ' See Averch & Johnson, supra note 43. [Vol, 2004COLUMBIA BUSINESS L-A4W REVIEW No. 2:335] ANTITRUSTAND THE REGULATORY ENTERPRISE perhaps significantly, and costs much to administer. These arguments are generally thought to be quite robust, and to counsel against regulation unless it is very clear that competition simply cannot work in a certain market. Avoiding excessive regulation consists of two things: First, the policy maker must avoid or abandon regulation in markets where it is unnecessary. For example, cost-of- service rate making was probably never warranted for industries such as interstate trucking and passenger air transportation to begin with.55 While these industries are subject to some unusual network complexities, they are capable of sustaining numerous competitive firms, and competition almost certainly provides a better outcome than regulation. Second, the domain of regulation must be no broader than necessary to cover the natural monopoly or other "market failure" that justifies the regulation. The New York City taxicab market provides one illustration. For taxis that have to be hailed from the street the transaction costs of price competition are probably very high. The driver and the passenger would have to negotiate a price, taking uncertainties such as traffic and highway construction into account; consequently, strangers to the city would be at a significant disadvantage. Taxi price regulation is hardly perfect, but it does deal with these problems. However, New York City has another class of automobile transportation, called limousines or non-medallion cars, which are not hailed from the street. They are called in advance, mainly by local businesses or frequent travelers. Often they work under relatively long-term contracts with firms that purchase local transportation services frequently, such as publishers and law firms. That portion of the market works quite well under price competition and should be left to function that way. " See BREYER supra note 7, at 197-284 (discussing trucking); MELVIN BRENNER ET AL., AIRLINE DEREGULATION (1985) (discussing airlines). C. SHIFTS IN COMPETITION THEORY Another economic argument against overbroad regulation is the result of a shift in economic thinking about the nature of competition. The neoclassical competition model developed during the late nineteenth and early twentieth centuries hypothesized competitive markets consisting of a large number of relatively small firms, no single one having power to reduce market-wide output and thus increase the price. 6 Within this model, competition is thought to be threatened by significant decreases in the number of firms in the market. As a result, significant economies of scale-the lower costs that attend higher output-are seen both as benefiting consumers by permitting lower production costs, and as injuring them by turning markets into oligopolies. The extreme case of scale economies is "natural monopoly," in which one firm occupying the entire market has lower costs than any possible combination of two or more firms. In that case no amount of competition is consistent with maximum productive efficiency. By definition any combination of two or more firms will have higher costs than the single firm has. Only price regulation was thought to approximate competitive behavior. But competition within regulated industries has turned out to be more robust than was formerly supposed. Even natural monopoly markets might perform competitively if entry and exit are not unduly costly. "Contestability" theory posits that competition for a market can be as effective as competition in a market: If the incumbent firm charges too much, a second firm will quickly enter, and there will be a brief price war until one is forced to exit. The ongoing threat of such entry forces the incumbent firm to charge competitive prices. Similarly, a publicly awarded "franchise monopoly," such as one for cable television, can be awarded 6 For a very brief summary, see HERBERT HOVENKAMP, FEDERAL ANTITRUST POLICY: THE LAW OF COMPETITION AND ITS PRACTICE ch. 1 (2d ed. 1999). On the history, see FRANK M. MACHOVEC, PERFECT COMPETITION AND THE TRANSFORMATION OF ECONOMICS (1995); George Stigler, Perfect Competition, Historically Contemplated, 65 J. POL. ECON. 1 (1957). [Vol. 2004COLUMBIA BUSINESS LA W REVIEW No. 2:335] ANTITRUST AND THE REGULATORY ENTERPRISE to the firm offering to charge the competitive price; every few years, the franchise would again be put up for such bidding. Even though the market has room for a single provider at any given time, price will be competitive if a competitive bidding process is used to select and, if necessary, replace the franchisee.57 While the theory of contestability is sensible, specific markets have posed numerous questions about practical application. The main problems are sunk costs and bottlenecks. A sunk cost is an investment that cannot be recovered by an exiting firm. For example, the first cable television company to enter the market must build a cable network. When its franchise expires the network has a significant useful life remaining. If a newcomer must build a new network, then the incumbent will have an overwhelming advantage in the second round of franchise bidding. The incumbent cannot be expected to transfer the network voluntarily to a new bidder. Legislation might require it to do so, but it would also have to determine the price. Alternatively, the network could be publicly owned and merely operated by the private firm-but that solution results in the very public ownership that our regulatory enterprise seeks to avoid. To date, there have not been practical solutions to this problem. As a result, franchise bidding in markets that contain costly, durable and nontransferable networks, such as cable television or local electric power grids, is a rarity." The voluminous literature includes WILLIAM J. BAUMOL, JOHN C. PANZAR & ROBERT D. WILLIG, CONTESTABLE MARKETS AND THE THEORY OF INDUSTRY STRUCTURE (1982); Joseph F. Brodley, Antitrust Policy Under Deregulation: Airline Mergers and the Theory of Contestable Markets, 61 B.U. L. REV. 823 (1981); Harold Demsetz, Why Regulate Utilities?, 11 J.L. & ECON. 55 (1968). For criticisms of the theory of contestability, see William G. Shepherd, "Contestability" vs. Competition, 74 AM. ECON. REV. 572 (1984). ' The classic piece on this problem is Oliver E. Williamson, Franchise Bidding for Natural Monopolies-In General and with Respect to CATV, 7 BELL J. OFECON. 73 (1976). 365 The other problem impeding contestability is bottlenecks. A good example is the airline industry. Sunk costs on any given route are low because, unlike a cable television network, an aircraft can be cheaply moved from one route to another. Even though the route from Omaha to Sheboygan has room for a single flight per day, if the incumbent carrier started charging monopoly prices a competing carrier would be able to swoop in and take its business at a lower price. Fearing this, the incumbent would have to price at the competitive level. But airplanes no longer land in cow pastures. They require airports and gate space and in most cases the latter is hard to come by. Today hub dominant carriers control a majority of the gate space at many United States airports. As long as that remains the case, contestability will not produce the promised results even in the passenger airline market, which was initially believed to be the darling market for the theory.59 While contestability theory has never produced the results it once promised, it has served to refocus antitrust thinking on the importance of competitive entry in disciplining monopoly. Even a firm that is alone in its market may not have significant power if new entry is quick and cheap. Such a market is not a good candidate for either price regulation or aggressive antitrust enforcement. V. ANTITRUST UNDER DEREGULATION: TELECOMMUNICATIONS The natural effect of deregulation is to enlarge the domain of antitrust by removing or narrowing antitrust immunities. For example, trucking and airline regulation has very largely placed those industries under antitrust control. Many of the collisions of antitrust and regulatory 5 See Elizabeth E. Bailey & John C. Panzar, The Contestability of Airline Markets During the Transition to Deregulation, 44 LAW & CONTEMP. PROBS. 125 (1981); David R. Graham et al., Efficiency and Competition in the Airline Industry, 14 BELL J. OF ECON. 118 (1983). COLUMBIA BUSINESS LA W REVIEW [Vol. 2004 No. 2:335] ANTITRUST AND THE REGULATORY ENTERPRISE 367 law have faded as control by regulatory agencies has been narrowed or eliminated. At the same time, antitrust sometimes has a different look and feel in formerly regulated industries and the transition has not always been smooth. Antitrust has moved in most easily in markets that never should have been regulated to begin with, such as trucking, or where technological change has permitted competition among multiple firms, such as long distance communications." The road has been much bumpier in deregulated markets that continue to manifest many of the characteristics of natural monopoly, or where technologically complex networks make the transition to competition unusually difficult. Perhaps the bumpiest of all has been deregulation of local telephone and related services under the 1996 Telecommunications Act. One of the most elaborate networks in United States markets is the telecommunications system, which links virtually every person in the United States as well as the world. While the telecommunications network links millions of users, much of it is also competitively supplied by hundreds of sellers of basic local and long distance service, high speed data transmission and internet service, wireless service, and of course hundreds of different types of attachments, ranging from simple telephones to fax machines to high speed computers. This level of provider competition is a relatively recent development. As passed in 1934, the Federal Communications Act contemplated that a single firm, AT&T, would own and operate the entire telecommunications system. For example, someone buying phone service in the 1950s in the United States did not have a choice of local or long distance carriers. In most areas AT&T was the only supplier.61 60 See 1A AREEDA & HOVENKAMP, supra note 11, 251i. On the antitrust impact of deregulation in various industries, see id. 241. 61 See HUBERETAL., supra note 41, § 1.3. The monopoly extended even to instruments, and AT&T was aggressive about protecting it. For example, the "Hush- A-Phone" was a purely mechanical gadget shaped like a tiny inverted megaphone that clipped over the mouthpiece of a telephone. It enabled a caller to whisper into the phone and be heard by the person on the other end of the line, but not by eavesdroppers in the same room as the caller. AT&T complained and got the FCC to agree that the Hush-A-Phone was an unauthorized foreign attachment. However, the Commission was reversed by the court of appeals.62 That small victory turned out to be the camel's nose under the tent. A few years later Tom Carter, who had invented the Hush-A-Phone, won an even bigger victory with a decision permitting various communications devices to be connected to the AT&T system, provided that connection was through a protective coupler that protected the network from harm. The Carterphone was another purely acoustic device that was designed to channel a radio signal into the telephone system so it could be picked up by someone on the telephone network.63 Then MCI entered the telecommunications market with a set of booster antennas initially designed to do no more than enable truckers' CB radios to communicate with each other over longer distances. From that point, and in the face of sustained opposition from both AT&T and initially the FCC, MCI expanded its service to encompass wireless alternatives to long distance, which it could only do by interconnecting its equipment with the AT&T system.64 62 Hush-A-Phone Corp. v. United States, 238 F.2d 266, 268 (D.C. Cir. 1956). 61 Use of the Carterphone Device in Message Toll Telephone Services, 13 F.C.C. 2d 420 (1968), recon, denied, 14 F.C.C. 2d 571 (1968). While the device was purely acoustic, the language of the decision effectively paved the way for electronic interconnection by permitting attachment on the customer's premises of any device "so long as the interconnection does not adversely affect the telephone company's operations or the telephone system's utility for others." On the history of the gradual demise of the AT&T monopoly, see GERALD W. BROCK, THE TELECOMMUNICATIONS INDUSTRY: THE DYNAMICS OF MARKET STRUCTURE ch. 9 (1981); WALTER G. BOLTER ET AL., TELECOMMUNICATIONS POLICY FOR THE 1980'S: THE TRANSITION TO COLUMBIA BUSINESS LA W REVIEW [Vol. 2004 No. 2:335] ANTITRUST AND THE REGUL4TORY ENTERPRISE 369 The AT&T breakup, which occurred in 1982, was intended to assign the competitive and natural monopoly portions of the telecommunications network to distinct firms. 5 This division, plus the need for local and long distance companies to interconnect, led to many disputes. Prior to 1996, the late Judge Harold Greene, who had presided over the breakup, decided these disputes. However, later that year Congress passed the 1996 Telecommunications Act, which imposed very broad interconnection obligations on the Bells and placed the process under the joint regulation of the FCC and state regulatory agencies. Under that statute the ILECs, or incumbent local exchange carriers (the Bells), are obligated not only to interconnect with CLECs, or competitive local exchange carriers, but also to lease to them almost any input capable of being leased. In many although certainly not all cases the CLECs have made only a minimal investment in facilities themselves. The 1996 Act was intended to facilitate the growth of competition, but the type of competition it produces is odd. Imagine that a town has only one seller of bananas, which is the local Kroger grocery store. Seeking to promote banana competition, the town passes a banana competition ordinance requiring Kroger to sell bananas at a steeply discounted wholesale price to individual entrepreneurs who push banana carts around the store, perhaps underselling Kroger itself by a few cents. In this case Kroger supplies the store facility, storage, heat, light, and even the bananas themselves, with the small sellers supplying little more than their labor. The banana competition ordinance simply confuses competition with large numbers of retailers. True banana competition would require individual stores with their own COMPETITION (1984); PHILIP L. CANTELON, THE HISTORY OF MCI: THE EARLY YEARS (1993). 65 See United States v. AT&T, 552 F. Supp. 131, 143 (D.D.C. 1982), affd sub nom. Maryland v. United States, 460 U.S. 1001 (1983) (giving the Bell Operating Companies control over the local hard-wired monopoly but denied them the right to engage in any "non-monopoly business"). facilities, purchasing bananas on the market and retailing them to consumers. Nevertheless, this is what the 1996 Telecommunications Act does. Small CLECs can lease most of their inputs from the Bells and even locate some of their equipment on Bell property. They are entitled to purchase the equipment and services they need at regulated wholesale prices, and then resell the services in competition with the Bells. In the long run, such sharing requirements may promote competition by encouraging small resellers to invest in their own facilities and eventually emerge as full fledged competitors. The small banana retailers might save their money and eventually trade in their carts for their own stores. But more effective competition would likely have resulted if the Act had limited the number of CLECs and placed more stringent requirements on them-for example, requiring them to provide inputs that are reasonably capable of being produced under competitive conditions. Under such a policy, which is more similar to what the FCC does in cellular phone markets,66 there would be fewer CLECs but their quality would be higher. Many of the CLECs who are in business today could never survive in an ordinary competitive market of the kind that the 1996 Act contemplates as its goal. Indeed, one of the ironies of the 1996 Telecommunications Act is that it may actually have retarded the development of a competitive infrastructure for local telephony because it permits CLECs simply to lease everything they want at steeply discounted prices, rather than forcing them to build for themselves where they could reasonably do so. 67 The 1996 Telecommunications Act contemplates that compulsory interconnection agreements will give rise to disputes, and there have been many of them. Disputes are to 66 See HUBER ET AL., supra note 41, at § 10.4.2. 67 See J. Gregory Sidak, The Failure of Good Intentions: The Worldcom Fraud and the Collapse of American Telecommunications After Deregulation, 20 YALE J. ON REG. 207 (2003); Jerry A. Hausman & J. Gregory Sidak, A Consumer-Welfare Approach to the Mandatory Unbundling of Telecommunications Networks, 109 YALE L.J. 417 (1999). [Vol. 2004COLUMBIA BUSINESS LA W REVIEW No. 2:335] ANTITRUSTAND THE REGULATORY ENTERPRISE be resolved by the FCC, or in some cases state regulatory agencies. In the Trinko case such a dispute ended up in an FCC-administered consent decree in which Bell Atlantic (now Verizon) paid a penalty, mainly for delays in filling orders pursuant to an interconnection agreement.68 The plaintiffs in Trinko were consumers, claiming that the breach of the interconnection agreement was also an antitrust violation.69 Trinko illustrates why some kind of immunity doctrine can be helpful, even in a partially deregulated industry. Here, a government agency was intended by Congress to resolve interconnection disputes and it was actually doing that job far more expeditiously than any court could do it through jury trials. 0 Indeed, the Trinko antitrust complaint was filed early in 2000, and three and a half years later was still tied up in litigation over a motion to dismiss. Further, the antitrust disputes raised issues under the essential facility theory and the monopoly leveraging theory, two highly controversial antitrust doctrines that stretch our ' Verizon Communications, Inc. v. Law Offices of Curtis Trinko, LLP, 124 S. Ct. 872 (2004). See also Covad Communications Co. v. Bellsouth Corp., 299 F.3d 1272 (11th Cir. 2002), vacated by 124 S. Ct. 1143 (2004); Goldwasser v Ameritech Corp., 222 F.3d 390, 402 (7th Cir. 2000). On the earlier regulatory history, see Memorandum Opinion and Order, Application of Verizon New England, 16 F.C.C.R. 8988, 9034-35 (2001) (summarizing events); Order, Bell Atlantic-New York Authorization, 15 F.C.C.R. 5413 (2000) (consent decree). The New York PSC also addressed the problem. Order Addressing OSS Issues, MCI WorldCom, Inc. v. Bell Atlantic-New York, 2000 WL 1531916 (NY PSC Nos. 00-C-0008 et al.); Order Directing Market Adjustments and Amending Performance Assurance Plan, MCI WorldCom, Inc. v. Bell Atlantic-New York, 2000 WL 517633 (NY PSC Nos. 00-C-0008 et al.); Order Directing Improvements to Wholesale Service Performance, MCI WorldCom, Inc. v. Bell Atlantic-New York, 2000 WL 363378 (NY PSC Nos. 00-C-0008 et al.). 69 Law Offices of Curtis v. Trinko, L.L.P. v. Bell, 305 F.3d 89, 95 (N.Y. 2002). 70 The interconnection dispute arose in December 1999, and was resolved by an FCC approved consent decree in March 2000, about three months later. Order, Bell Atlantic-New York Authorization, 15 F.C.C.R. 5413, 5416 (2000). 372 COLUMBIA BUSINESS LAW REVIEW [Vol. 2004 definition of monopolistic conduct to the breaking point.71 The monopoly leveraging claim in this case accused Verizon of attempting to leverage a monopoly position in the wholesale market for telecommunications services into a dominant retail position. The problem with that claim was that any monopolist in the wholesaling of a product necessarily has a monopoly in the retailing as well, unless it decides to share distribution with others. That is, the rule would require compulsory wholesaling by upstream monopolists. For example, if Alcoa has a monopoly in aluminum ingot, it would be required by the Sherman Act to wholesale at least some ingot to independent fabricators, lest it create a second monopoly in fabrication as well.72 Such a rule would turn every upstream monopolist into a public utility, with the price and volume of its sales to independents regulated by a federal district court. The essential facility doctrine, about which much has been written, would require a firm to share essential inputs at terms imposed by the court. The important point is that these two doctrines would have effectively given antitrust courts full and duplicate control over interconnection disputes that are already being well supervised by regulators. The disputes are often extremely technical, and certainly not well suited for administration via jury trials. In Trinko the Supreme Court barred all leveraging claims and placed severe limitations on further uses of either the essential facility doctrine or the general section 2 law of unilateral refusal to deal. However, the Court also held that an Antitrust Saving cause in the 1996 Telecommunications Act barred a finding of implied immunity.7 3 71 See 3 AREEDA & HOVENKAMP, supra note 11, 652 (discussing leveraging); 3A AREEDA & HOVENKAMP, supra note 11, 770-774 (discussing the essential facility doctrine). 72 United States v. Aluminum Co., 148 F.2d 416 (2d Cir. 1945). 73 Trinko, 124 S. Ct. at 878 ("Section 601(b)(1) of the 1996 Act is an antitrust-specific saving clause providing that 'nothing in this Act or the amendments made by this Act shall be construed to modify, impair, or No. 2:335] ANTITRUSTAND THE REGULATORY ENTERPRISE That conclusion seems far less clear than the Court presumed. Historically, the Supreme Court has disliked such clauses and several members have declared that they are unhelpful 4 for good reason.75 A saving clause is meaningful only if it can be read against some status quo ante that must be saved, but the 1996 Telecommunications clause never does that. One might of course take the position that the saving clause creates for telecommunications the same kind of antitrust coverage that applies to ordinary commodities such as cement or lumber, with no consideration whatsoever for the regulatory environment of telecommunications. However, that construction is grossly inconsistent with the clause's mandate that the 1996 Act not be construed to modify the applicability of antitrust. Immediately prior to passage of the 1996 Telecommunications Act, interconnection agreements were overseen mainly under the supersede the applicability of any of the antitrust laws.' 110 Stat. 143, 47 U.S.C. § 152, note. This bars a finding of implied immunity."). " Rush Prudential HMO, Inc. v. Moran, 536 U.S. 355, 3B4 (2002); Egelhoff v. Egelhoff ex rel. Breiner, 532 U.S. 141, 153 (2001); Atherton v. F.D.I.C., 519 U.S. 213, 231 (1997); American Dredging Co. v. Miller, 510 U.S. 443, 459 (1994). 7' The legislative history of the saving clause is unilluminating. The Conference Committee observed this: in the future, the conferees anticipate that cable companies will be providing local telephone service and the BOC's will be providing cable service. Mergers between these kinds of companies should not be allowed to go through without a thorough antitrust review under the normal Hart-Scott- Rodino process .... By returning review of mergers in a competitive industry to the DOJ, this repeal would be consistent with one of the underlying themes of the bill-to get both agencies back to their proper roles and to end government by consent decree .... The repeal would not affect the [Federal Communications] Commission's ability to conduct any review of a merger for Communications Act purposes, e.g. transfer of licenses. Rather, it would simply end the Commission's ability to confer antitrust immunity. H. Rep. No. 104-458, at 201 (1996). AT&T consent decree78 administered by Judge Greene and were not generally addressed separately under the antitrust laws.77 However, it seems clear that this regime is not the status quo that Congress had in mind when it commanded that nothing in the 1996 Act should be construed to modify the applicability of the antitrust laws. Indeed, one of the purposes of the 1996 Telecommunications Act was to dissolve the AT&T consent decree. The regime administered under that decree no longer exists.7" Going even further back, prior to the entry of the AT&T breakup decree, the telephone company was a unitary price- regulated firm governed by the implied immunity rules that applied at that time to public utilities. First, under the doctrine of "primary jurisdiction," the courts sometimes dismissed antitrust claims or transferred them to the FCC in deference to its expertise. 9 Second, under these pre-consent decree rules, AT&T was not entitled to a blanket antitrust immunity, but rather immunity was given on a case-by-case basis to conduct that was compelled or supervised by the FCC or relevant state agency. 80 The leading decision was 76 United States v. AT&T, 552 F. Supp. 131 (D.D.C. 1982), affd sub nom. Maryland v. United States, 460 U.S. 1001 (1983). 77 E.g., United States v. Western Elec. Co., 767 F. Supp. 308 (D.D.C. 1991) (information services); United States v. Western Elec. Co., 890 F. Supp. 1 (D.D.C. 1995) (wireless); United States v. Western Elec. Co., 1987- 1 Trade Cas. (CCH) 67,452 (D.D.C. Jan. 28, 1987) (cellular). Numerous other decisions are cited in HUBER ET AL., supra note 41, § 4.3.2. 78 See H. Rep. No. 104-458, at 201 (1996) (noting that one purpose of 1996 Act was to "end government by consent decree"). 7' E.g., Carter v. AT&T Co., 365 F.2d 486 (5th Cir. 1966), cert. denied, 385 U.S. 1008 (1967) (finding that district court properly invoked primary jurisdiction doctrine to refer factual matters underlying antitrust dispute to FCC). "' See, e.g., MCI Communications Corp. v. AT&T, 708 F.2d 1081 (7th Cir. 1981), cert. denied, 464 U.S. 891 (1983) (holding that there was no immunity to the extent that AT&T's conduct was not supervised by a regulatory agency); Mid-Texas Communications Sys. v. AT&T, 615 F.2d 1372 (5th Cir. 1980), cert. denied, 449 U.S. 912 (1980) (finding that regulatory provisions did not confer a blanket immunity from antitrust laws; but actions based on good faith compliance with regulatory standard could individually be held immune); Phonetele v. AT&T, 664 F.2d 716, COLUMBIA BUSINESS LA W REVIEW [Vol. 2004 No. 2:335] ANTITRUSTAND THE REGULATORY ENTERPRISE MCI, handed down shortly before the AT&T consent decree. The Seventh Circuit held that mere "pervasiveness" of a regulatory scheme was insufficient to create an immunity.8 Rather, determination of immunity "requires an evaluation of the specific regulatory scheme involved and the administrative authority exercised pursuant to that scheme."82 The question in this case was "whether the activities that are the subject of MCI's complaint were required or approved by the Federal Communications Commission, pursuant to its statutory authority, in a way that is incompatible with antitrust enforcement .... ,"s With respect to interconnection agreements under the then existing scheme, the court concluded: Although the FCC has authority to compel interconnection ... the initial decision whether to interconnect rests with the utility, and the record shows that the FCC did not control or approve of AT&T's actions here. Nor has the FCC supervised AT&T's interconnection practices so closely that the FCC's approval could be inferred. 4 In sum, a Saving Clause standard that compels that antitrust rules not be modified would seem to preserve precisely the kind of immunity called for in this case- namely, immunity given on a case-by-case basis in situations where the agency has jurisdiction over the disputed conduct and is actually exercising its jurisdiction by resolving such disputes consistent with its statutory mandate. This approach seems quite consistent with recent Supreme Court holdings to the effect that saving clauses should not be read 731-735 (9th Cir. 1981), cert. denied, 459 U.S. 1145 (1983) (finding no general antitrust immunity, but that defendants could attempt to show specific immunity for the challenged acts); Sound v. AT&T, 631 F.2d 1324 (8th Cir. 1980) (refusing to find immunity for AT&T's unsupervised, discretionary conduct). 81 MCI, 708 F.2d at 1102. 82 id. 83 id. 8 Id. at 1103. so as to produce unnecessary conflict between two different legal regimes.85 The alternative interpretation would seem to be that the Saving Clause preserves some imagined historical regime in which the telephone companies' status as regulated industries was simply not relevant at all. Problematically, there has never been a time when the antitrust laws applied to the telephone system in the same way they apply to ordinary commodities. The Mann-Elkins Act of 1910, an amendment to the Interstate Commerce Act, first placed the telecommunications industry under significant federal regulation. 6 The Communications Act of 1934 then placed the industry under the control of the Federal Communications Commission.87 It is worth noting that the Telecommunications Act Saving Clause has not been interpreted so as to abrogate one other important antitrust immunity. The "filed tariff' doctrine immunizing from antitrust challenge rates that have been filed with a regulatory agency is alive and well under the 1996 Telecommunications Act. Judge Wood herself so held in Goldwasser.8 Significantly, the filed tariff doctrine is an implied immunity doctrine. It exists not because a specific statute mandates it, but because antitrust challenges to filed tariffs are thought to create needless conflict with the operations of the agencies with whom such tariffs are filed. The policy rationale for the filed tariff doctrine is far weaker for the general doctrine of implied regulatory immunity. The filed tariff doctrine immunizes antitrust claims even if the agency is somnolent and merely rubber stamps tariffs submitted to it by regulated firms. It thus often applies whether or not there is any real conflict between agency ability to function and application of the 8" See Geier v. American Honda Motor Co., 529 U.S. 861 (2000) (holding that clause saving state law tort claims did not operate so as to preserve such claims where there was an actual conflict between federal and state law). 86 Mann-Elkins Act, ch. 309, §7, 36 Stat. 539 (1910). 87 See 47 U.S.C. § 151 et. seq.; 77 CONG. REC. 12,452 (1934). ' Goldwasser v Ameritech Corp., 222 F.3d 390, 402 (7th Cir. 2000). [Vol. 2004COLUMBIA BUSINESS LA W REVIEW No. 2:335] ANTITRUSTAND THE REGULATORY ENTERPRISE antitrust laws. Indeed, the Supreme Court's most recent decision finding an antitrust filed tariff immunity conceded that the immunity was probably improvident and rested entirely on stare decisis.8 9 Considered in this light, the most sensible reading of the 1996 Act's Saving Clause is that it preserves intact the system of regulatory rules that existed prior to the consent decree and that continues to govern most regulated industries. Under these rules there is no blanket immunity from the antitrust laws. Further, behavior that is never disclosed to the agency, perhaps because it is surreptitious, is not immune. However, immunity is given for conduct that is within the jurisdiction of a regulatory agency and where the agency has acted or is actively considering whether to act. VI. CONCLUSION The most important principle for antitrust in the regulated industries is first, do no harm. While judges might blanch at the atrocities that governments have committed in the name of regulation, antitrust is not the appropriate vehicle to provide a cure. Rather, it has the much humbler task of preserving competitive incentives that are consistent with the regulatory regime that has been created, whatever the regime's internal merits. 9 See Square D. Co. v. Niagara Frontier Tariff Bureau, 476 U.S. 409, 420 (1986) (conceding that doctrine may be "unwise as a matter of policy," but Congress had ample opportunity to overturn it and had not done so).