Antitrust: The Case for Repeal - Mises Institute The Case for Repeal_1_0.pdfAntitrust: The Case for...

ANTITRUST

THE CASE FOR REPEAL

DOMINICK T. ARMENTANO

REVISED 2ND EDITION

Ludwig von Mises Institute

Auburn, Alabama

All rights reserved. Written permission must be securedfrom the publisher to use or reproduce any part of this book,except for brief quotations in critical reviews or articles.

The publisher has attempted throughout this book to distin-guish proprietary trademarks from descriptive terms by fol-lowing the capitalization styles used by the manufacturers.

First edition, titled Antitrust Policy: The Case for Repeal, origi-nally published by the Cato Institute, 1000 MassachusettsAvenue, Washington, D.C. 2001.

Copyright © 1999 by the Ludwig von Mises InstituteReprinted in 2007 by the Ludwig von Mises Institute

Ludwig von Mises Institute, 518 West Magnolia Avenue,Auburn, Ala. 36832; www.mises.org

ISBN: 978-0-945466-25-3.

Contents

Preface vii

Introduction: An Antitrust Overview xi

1. The Antitrust Assault on Microsoft. l

2. The Case Against Antitrust Policy 13

3. Competition and Monopoly: Theory andEvidence 31

4. Barriers to Entry 51

5. Price Discrimination and Vertical Agreements 69

6. Horizontal Agreements: Mergers and PriceFixing 81

7. Antitrust Policy in a Free Society 99

Index 107

iii

People of the same trade seldom meet together,even for merriment and diversion, but the con-versation ends in a conspiracy against the public,or in some contrivance to raise prices. It is impos-sible, indeed, to prevent such meetings, by anylaw which either could be executed/ or would beconsistent with liberty and justice. But though thelaw cannot hinder people of the same trade fromsometimes assembling together, it ought to donothing to facilitate such assemblies; much lessrender them necessary.

-Adam SmithThe Wealth of Nations

v

Preface

The flurry of federal and state antitrust activity against firmssuch as Toys "R" Us, Staples, Intel, and Microsoft may signalthe beginning of an unfortunate new era in enforcement.Antitrust regulation, like a relentless Terminator, is back inbusiness and the economic havoc it threatens is consider-able.

My position on antitrust has never been ambiguous: Allof the antitrust laws and all of the enforcement agencyauthority should be summarily repealed. The antitrust appa-ratus cannot be reformed; it must be abolished.

It is said that much is risked in calling for repeal. Any callfor repeal is likely to galvanize those interests committed toa return to the old-style, traditional enforcement policies. Inaddition, the antitrust "establishment"-attorneys, consult-ants, antitrust agency bureaucrats-would probably step upits attack on those who intend, from its perspective, to fur-ther "weaken" antitrust policy. Abolitionists would againbe portrayed as pro-business and anti-consumer, devoid ofany concern for consumer welfare or economic fairness.The most serious danger, presumably, would be that a prin-cipled opposition to all antitrust could delay importantantitrust reforms or even reverse some of the slight admin-istrative reforms already achieved.

Similarly, any serious movement to repeal is said to runthe risk of alienating the support of those critics of tradi-tional policy most responsible for the modest antitrust reformsthat we have seen to date. The majority of important antitrust

vii

Antitrust: The Case for Repeal

critics do not support the repeal of antitrust laws; in theirview, there is an appropriate role for antitrust policy in afree-market economy, although one that is reduced inscope from the traditional understanding. They wouldargue that antitrust is still necessary for combating cartels,very large horizontal mergers, and bona fide predatorypractices.

I emphatically disagree. There certainly are risks in work-ing for repeal, but there are even greater risks in not push-ing the intellectual argument against antitrust to its logicalconclusion. I will argue that the case against antitrust reg-ulation-any antitrust regulation-is far stronger than evenits most important critics are willing to acknowledge. I willargue that the employment of antitrust, even against privatehorizontal agreements, cannot be justified by any respectablegeneral theory or empirical evidence. But even more practi-cally, I will argue that the very modest administrative reformsthat we have seen can only be temporary. They were, afterall, only administrative reforms, and we have already fallenback into the quagmire of more traditional enforcementpolicies. The greater risk would be to remain content withsome modest "reform" agenda while leaving the entireantitrust institutional structure of private litigation, agencyenforcement, and court review essentially in place. It wouldbe far better in an entirely practical sense to abolish all ofthese institutional arrangements and simply be done withthe greater risk.

Many of the arguments I develop and cases , discuss inthis book will be familiar to readers of my Antitrust andMonopoly.1 New readers who find these ideas stimulat-ing-or infuriating-may wish to pursue some of them ingreater depth elsewhere.2 I intend, with this revised edition

1Dominick T. Armentano, Antitrust and Monopoly: Anatomy of a Policy Failure,2nd ed. (Oakland, Calif.: Independent Institute, 1990).

2Robert H. Bork, The Antitrust Paradox: A Policy at War with Itself (New York:Basic Books, 1978);· Yale Brozen, Concentration, Mergers, and Public Policy (New

viii

Preface

of Antitrust: The Case for Repeal, to reach a wider audienceand to promote a greater public understanding of the caseagainst antitrust regulation. Such an understanding stillappears necessary.

York: Macmillan, 1982); Fred L. Smith, Jr., "Why Not Abolish Antitrust?" Regulation 7(january/February 1983): 23-28; Frank H. Easterbrook, "The Limits of Antitrust," TexasLaw Review 63 (August 1984): 1-40; Fred S. McChesney, "law's Honor Lost: The Plightof Antitrust," Antitrust Bulletin 31 (1986): 359-82; William Shughart II, The Organizationof Industry (Homewood, III.: Richard D. Irwin, 1990); and Fred S. McChesney andWilliam F. Shughart II, The Causes and Consequences of Antitrust (Chicago: Universityof Chicago Press, 1995).

ix

Introduction: An Antitrust Overview

Although it is difficult to summarize more than a century ofantitrust enforcement in one observation, it is undeniablytrue that the antitrust laws have often been employedagainst innovative business organizations that have expand-ed output and lowered prices. That is most obvious in pri-vate antitrust cases (90 percent of all antitrust litigation),but it is also evident in the classic government cases aswell. Since antitrust regulation (at least the Sherman Act)was allegedly designed to prohibit business activity harm-ful to consumers' interests, much of antitrust policy as prac-ticed, appears terribly misguided and might be termed a"paradox.'"

The alleged paradox can be explained in several ways. Oneapproach is to challenge the "public interest" origins of antitrustpolicy.2 If the laws were originally meant to protect less efficientbusiness organizations from competition rather than to pro-mote the interests of consumers, then there is no paradox.From that perspective, antitrust regulation is just another histor-ical example of protectionist rent-seeking legislation, the overalleffect of which is to lessen economic efficiency.3

1For examples of the view that antitrust laws were created to serve consumers,see Hans Thorelli, The Federal Antitrust Policy (Baltimore, Maryland: The JohnsHopkins Press, 1955); and Robert H. Bork, The Antitrust Paradox: A Policy at War withItself (New York: Basic Books, 1978).

2rhomas J. Dilorenzo, "The Origins of Antitrust: An Interest-Group Perspective,"International Review of Law and Economics 5 (1985): 73-90.

3See, for example, Bruce L. Benson, M.L. Greenhut, and Randall G. Holcombe,"Interest Groups and the Antitrust Paradox," Cato Journa/6 (Winter 1987): 801-18; or

xi


It can also be argued that there has traditionally existedserious theoretical confusion over the meaning of "compe-tition." That confusion may have misled the courts and theadministrators of antitrust law.4 For example, when a firmlowers its price, is that competition or an attempt to monopo-lize? When a firm gains market share, is that evidence ofefficiency or a threat to competition? When business merg-ers are restricted by law, is competition enhanced orrestrained? When a firm engages in expensive research andinnovation that competitors cannot easily duplicate, is thatmonopolization? Faulty theorizing on these issues couldexplain a public policy attack on economic efficiency in thename of preserving competition.

Economic Theory and Antitrust Policy

The theoretical foundations of antitrust policy devel-oped generally from neoclassical microeconomics andwere refined by scholars specializing in industrial organiza-tion. And although industrial organization (10) theoryremained deeply rooted in pure competitioW and puremonopoly models, 10 economists in the late 1940s and1950s increasingly focused their analyses on those indus-tries that lay between pure competition and absolutemonopoly. Their goal: to understand the relationshipsbetween market structure, business· behavior, and overalleconomic performance.

Early 10 economists generally came to accept a deter-ministic relationship between market structure and econom-ic performance. If markets were competitively structured(small firms, homogeneous products, and ease of entry),then the market process led .automatically to an allocation

William Baumol and Janusz Ordover, "Use of Antitrust to Subvert Competition,"Journal of Law and Economics 28 (May 1985): 247-65.

4See, for example, Thomas J. Dilorenzo and Jack C. High, "Antitrust andCompetition, Historically Considered," Economic Inquiry 26 (July 1988): 423-35.

xii

Introduction

of resources whereby price, marginal cost, and ~inimumaverage cost were all equal. Alternatively, high market con-centration, collusion among firms, economies of scale, orproduct differentiation could create barriers to entry andmarket power that would misallocate economic resources.Early empirical data on market concentration and firm profit-ability appeared to support the general 10 hypothesis thatcompetitively structured markets performed better thanconcentrated markets.

It was a short step from microeconomic theory, regressionanalysis, and some engineering studies on optimum plant sizeto recommendations concerning appropriate public policy. Ifpoor market structure led to economic inefficiency, then gov-ernment antitrust regulation might correct such "market fail-ures." For example, antitrust regulation could reduce orrestrain industrial concentration (anti-merger policy), restrictpredatory practices, prohibit horizontal price and outputagreements (anti-collusion· rules), and discourage other agree-ments within and among firms (prohibitions against tyingagreements and resale price maintenance) that might restraintrade and competition. Barriers to entry that appeared to shel-ter so-called dominant firms (product differentiation, for ex-ample) could be attacked under the antitrust laws to makethe marketplace more efficient.

The structure-conduct-performance perspective becamethe primary intellectual justification for traditional antitrustpolicy in the 1950s and 1960s.5 Within that framework,several classic antitrust cases were brought to curb pricediscrimination,6 tying agreements/ increasing industrial

5See, for example, Phillip Are~da, Antitrust Analysis: Problems, Text Cases, 2nd ed.(Boston: little, Brown, 1974); or EM. Scherer, Industrial Market Structure and EconomicPerformance, 2nd ed. (Boston: Houghton Mifflin, 1980).

61n the Matter of the Borden Company, 381 FTC 130 (1958); Borden Company v.FTC, 381 E 2nd 175 (1967).

7Fortner Enterprises, Inc. v. United States Steel Corporation and United StatesHomes Credit Corporation, 394 U.S. 495 (1969).

xiii


concentration,8 and the "exclusionary" practices and highmarket share of United Shoe Machinery9 and InternationalBusiness Machines.10

Theory Revisionism and Policy Reform

Criticism of the structure-conduct-performance frame-work and of traditional antitrust regulation increasedsharply in the 1970s. The so-called "new learning" chal-lenged some of the theoretical assumptions of the older 10paradigm (economic uncertainty 'generally replaced per-fect information in the newer economic analyses, for exam-ple) and questioned many of its important empirical pre-dictions.11 New learning theorists such as Harold Demsetzand Yale Brozen argued that increasing market concentra-tion was not necessarily associated' with inefficiency ormonopoly profits and that increased concentration couldlead to an increase in market efficiency that benefited con-sumers.12 In addition, careful reexaminations of earlierantitrust cases demonstrated that much of the historicalenforcement effort had been entirely misplaced. By the early1980s, each part of the traditional justification for vigorousantitrust enforcement had come under severe criticism byeconomists and law professors. That intellectual criticismhelped pave the way for some modest changes in antitrustenforcement.

8Brown Shoe Company v. United States, 370 U.S. 294 (1962); FTC v. Procter &Gamble Company, 386 U.S. 568 (1967).

9United States v. United Shoe Machinery Corporation, 110 F. Supp. 295 (1953).

10United States v. International Business Machines Corporation, Docket no. 69,Civ. (ONE) Southern District of New York (1969).

11 For an early collection of critiques of antitrust policy, see Harvey Goldschmid,H. Michael Mann, and J. Fred Weston, eds., Industrial Concentration: The New Learning(Boston: Little, Brown, 1974).

12See, for example, Harold Demsetz, "Industry Structure, Market Rivalry, andPublic Policy," Journal of Law and Economics 16 (April 1973): 1-10; and Yale Brozen,"Concentration and Profits: Does Concentration Matter?" Antitrust Bulletin 19 (1974):381-99.

xiv

Introduction

The so-called antitrust revolution of the late 1970s andearly 1980s was evidenced by several important factors.First, there was a decided shift in the mix of antitrust casesinitiated by the Department of Justice and by the FederalTrade Commission (FTC). Fewer mergers were challenged(under revised merger guidelines) than previously a,ndmore price fixing cases were initiated. Second, there was amodest decline in both private and public antitrust activity.Finally, the courts, includtng the Supreme Court, becameincreasingly skeptical of traditional antitrust theories ofmonopoly power.

The last factor was probably the most significant. In deci-sions such as those in Sylvania,13 Brunswick,14 IllinoisBrick,lS Broadcast Music,16 Monsanto,17 Zenith Radio,18 andSharp19 the Supreme Court broadened the rule-of-reasonperspective in antitrust law. These decisions were basedprimarily on orthodox microeconomic analysis, and theywere by no means entirely consistent or complete. But theclear trend in court decisions during the period definitelyrepresented a shift away from the traditional analyses anddecisions of the 1950s, 1960s, and early 1970s.

The New Antitrust ActivismThe enforcement revolution was short-lived. New

administrators at the Department of Justice and at the FTCduring the Bush and Clinton administrations expanded

13Continental T. \1:, Inc. v. GTE Sylvania, Inc., 433 U.S. 36 (1977).

14Brunswick Corp. v. Pueblo Bowl-crMat, Inc., 429 U.S. 477 (1977).

lSlIIinois Brick Co. v. Illinois, 431 U.S. 720 (1977).

16Broadcast Music, Inc., v. CBS, Inc., 441 U.A. 1 (1979).

17Monsanto Co. v. Spray-Rite Service Corp., 465 U.S. 752 (1984).

18Matsushita Electric Indus. Co. v. Zenith Radio Corp., 1067 S. Ct. 1348 (1986).

19Business Electronics Corp. v. Sharp Electronics Corp. 108 S. Ct. 1115 (1988).

xv


antitrust enforcement.2o For example, Bush appointeesjames F. Rill (Justice) and janet Steiger (FTC) both made itclear that they favored a wider and more vigorous enforce-ment effort than did their Reagan administration predeces-sors. Investigations and enforcement efforts were alsoexpanded during the Clinton administration under AssistantAttorney General Anne K. Bingaman and her successor atjustice, joel Klein. Besides the sharp increase in corporatecriminal fines collected for alleged price-fixing, the Clintontrust-busters (including the FTC) dramatically expanded thenumber of merger investigations, initiated questionablecases addressing vertical integration issues, supported theinternationalization of antitrust enforcement, and filed highprofile cases against firms such as Staples, Intel, and, ofcourse, Microsoft. Antitrust regulation, despite decades ofintellectual criticism, was back in business.

20Janusz A. Ordover, "Bingaman's Antitrust Era," Regulation 20, no. 2 (1997):21-26.

xvi

1. The Antitrust Assault on Microsoft

The 1998 antitrust suit brought by the Department ofJustice and twenty state attorneys general against theMicrosoft Corporation1 captures everything that is stillwrong with antitrust policy and demonstrates why the lawsmust be repealed.

A brief historical review of Microsoft's antitrust difficul-ties is in order. The Federal Trade Commission startedinvestigating Microsoft's software licensing practices in1990 but closed its investigation in 1992 without filingcharges. (This was significant since the FTC is expresslycharged with policing so-called "unfair methods of compe-tition.") But in an unusual development, the Clinton admin-istration's Justice Department, under Assistant AttorneyGeneral Anne K. Bingaman, picked up the aborted FTCprobe of Microsoft and sharply expanded its scope.2

After an additional two-year study, the Justice Depart-ment concluded that Microsoft's "per processor" licensing-fee system discouraged PC manufacturers from installingcompetitive software and that Microsoft's standard two-year lease unfairly foreclosed software rivals from the mar-ket. To avoid long litigation, Microsoft signed a consentdecree with the Department in 1994 and agreed to end its

1United States v. Microsoft Corp. Civ. Action No. 98-1232 (1998).

2Under pressure from Microsoft's competitors, Senator Howard Metzenbaum(Democrat, Ohio) and Senator Orrin Hatch (Republican, Utah), both urged Ms.Bingaman to reexamine the Microsoft case. See Wall Street Journal, August 2, 1993,p. B8.


per processor licenses and shorten its standard two-yearlease period to one year. u.s. District Judge Stanley Sporkinrefused to certify the agreement because it did not providean "effective antitrust remedy" and was not in the publicinterest, but he was overruled by a Court of Appeals. Theconsent decree became fully effective in 1995.

With one set of alleged restrictive practices resolved, thefederal antitrust authorities immediately focused on a newset associated with so-called "Internet access." The newconcerns stemmed from Microsoft's decision to integrate(or tie) various software applications into' its increasinglypopular Windows operating system.

First, in an unprecedented move, the Justice Depart-ment threatened to delay the introduction of Windows 95because Microsoft bundled its own on-line Internet service(Microsoft Network) with Windows. Then Justice andMicrosoft disagreed bitterly over Microsoft's decision to tieits Internet browser, Explorer, to its operating system. Thegovernment claimed that the bundled browser violated the1995 consent decree; Microsoft claimed that the decreeexplicitly allowed "integration" of the browser as well asother applications. An appellate court ruled definitively inMicrosoft's favor in June of 19983 but, in the interim, theDepartment of Justice and twenty states filed an antitrustsuit against Microsoft.

The suit claimed that Microsoft had a monopoly in oper-ating systems for personal computers, that it attempted ille-gally to leverage its monopoly power in operating systemsto other products or services, that it engaged in restrictiveagreements with PC manufacturers and Internet serviceproviders, and that its monopolization injured competitorsand consumers. A trial began in October 1998.

3United States v. Microsoft Corp., 147 F. 3d 935 D.C. Cir. (1998).

2

The Antitrust Assault on Microsoft

Microsoft's MonopolyWhether Microsoft had a monopoly in operating sys-

tems depends, of course, on a precise definition of monop-oly. A perfect monopoly, presumably, would control all ofthe available supply of a product in some well-defined rele-vant market with strong legal barriers to entry. SinceMicrosoft was said to license 90 percent of the operatingsystem software sold in new personal computers and sincethere were no legal barriers to entry in software, Microsoftdid not have a perfect monopoly. There were other operat-ing systems for personal computers available (Mac as,Unix, OS/2, Linux) and consumers could turn to them if theMicrosoft system were unavailable; in addition, new sup-pliers could always enter the market. Yet, legal scholars cit-ing precedent would argue that any market share above 70percent (with or without legal barriers) can constitutemonopoly under antitrust law.4

As we will elaborate in the following pages, the market-share theory of monopoly is confusing and ultimately mis-leading. Much depends on how the relevant market for theproduct is defined. More importantly, a firm could producea superior product at low cost and consumers could estab-lish that firm as the dominant supplier; the law, presumably,was not meant to restrict such beneficial behavior.5 Indeedmonopoly, however defined, isn't illegal under theSherman Act; "monopolization" is. What the law reallyrequires (after a threshold market position has been estab-lished) is a showing that the defendant engaged in so-calledmonopolistic practices. The important questions are: Howdid the firm come to obtain its market share? Did the firmunfairly exclude competitors from the market? Did it unfair-ly restrain the competitive process?

4United States v. E./. duPont de Nemours & Co., 351 U.S. 377 (1956).SUnited States v. Grinnell Corp., 384 U.S. 563 (1966).

3


In our view, Microsoft's dominant market share in operatingsystems evolved legitimately from a free-market competitiveprocess. The PC software industry was legally open andcontained many talented players (Sun, Netscape, Novell,Oracle, Apple, IBM), some larger than Microsoft, somesmaller. The market process in this industry has alwaysbeen characterized by intense innovation, rapid growth,sharply falling prices, and bitter rivalry (and occasionalcooperation) between rivals. The industry exemplifiesAustrian economist Joseph Schumpeter's vision of compe-tition as a process of creative destruction.

Microsoft achieved its market position by aggressivelyinnovating and promoting an open, standardized operatingsystem platform that integrated various applications (filesharing, fax utilities, network support) that had been avail-able separately. Hundreds of PC manufacturers, thousandsof software applications developers, and eventually mil-lions of consumers came to appreciate the advantages ofthe Microsoft Windows approach. A standardized and inte-grated operating system was less expensive to produce anddistribute, easier to use, and ultimately more beneficial forconsumers. As a consequence, some early market leadersstumbled and fell by the wayside while Microsoft emergedout of the competitive process with a legitimately-earnedmarket share.

Network Effects and Path Dependence

Some critics hold that market dominance in software isenhanced unfairly by so-called network effects.6 Successfulfirms like Microsoft are said to have unfair advantages oversmaller firms because a larger number of product users-larger networks-leads to expanded consumer benefits

6For an extensive discussion of the issues, see John E. lopatka and William H.Page, "Microsoft, Monopolization, and Network Externalities: Some Uses and Abu-ses of Economic Theory in Antitrust Decision Making," Antitrust Bulletin 40 (Summer1995): 317-70.

4


which leads, in turn, to even larger networks and profits fordominant firms.

It can be admitted that network effects can createdemand-side advantages for larger firms and increasing ben-efits for consumers that use their systems. Even further,economies of scale can also generate cost-side advantagesfor market leaders, making it even more difficult for smallerfirms to be competitive. But there is nothing economicallyunfair or regrettable about these developments.

In the first place, increasing returns and low marginalcosts are no iron-clad guarantee of long-run success; busi-ness history is filled with "first mover" firms that experi-enced dramatic losses in market share because of changesin consumer tastes and technology. Second, low costs andincreasing advantages for a large pool of network users arethe economic benefits of the free competitive process;they are never to be regretted. The competitive process issupposed to generate low costs and increasing benefits forconsumers and is supposed to punish low value, high costrivals. Competition is supposed to reward firms that inno-vate first, that build integrated systems, and that expandbefore their rivals do. Thus, to make such firms primeantitrust targets is a screaming contradiction to the allegedintent of antitrust law and reveals, instead, its true protec-tionist purpose.

Another consideration is the notion of path dependencewhereby an increasing returns monopolist is said to be ableto lock in some inferior technology while locking out rivalswith superior innovations. Presumably this has occurredrepeatedly in business history (the QWERTY keyboard isoften cited) and it is alleged to be a serious inefficiencyassociated with monopoly.

Myths die hard in the antitrust area. With costs correctlytaken into account, there is simply no empirical support forthe notion that inferior technology can exclude superior

5


technology-a kind of Gresham's law in innovation? TheQWERTY keyboard myth has been effectively debunked ashave other alleged examples such as the BetajVHS videorecorder format controversy.8 The lack of empirical supportis not surprising since path-dependent theorists have theinnovation story backwards. Market share, after all, is th~direct result of consumers rewarding firms that have con-tinuously rewarded consumers with superior innovations.Again, the antitrust assault on market leaders is an attackon demonstrable efficiency and on revealed consumerpreferences.

Restrictive Practices

The trustbusters had a very different perspective. Theyheld that Microsoft engaged in certain restrictive practiceswith original equipment manufacturers and Internet con-tent providers that had the effect of foreclosing the marketto important Microsoft rivals. Take, for example, the issueof the Internet browser. Since Microsoft bundled its ownbrowser, Explorer, with Windows, and offered Explorer freeof charge to PC manufacturers, rival browser makers-suchas market leader Netscape Communications-argued thatthey were increasingly foreclosed from the browser market.

But the antitrust issue is whether Netscape and otherswere unfairly foreclosed. When Microsoft licensed its soft-ware, it did not generally restrict PC manufacturers frominstalling competitive software.9 Microsoft did not have

75.J• Liebowitz and S.E. Margolis, "Path Dependence, Lock-in, and History,"Journal of Law, Economics, and Organization 11 (1995): 205-26.

,8S.J• Liebowitz and S.E Margolis, "Fable of the Keys," Journal of Law andEconomics 33 (1990): 1-25.

9Microsoft did not restrict PC manufacturers from adding on "competitive" soft-ware beyond the start-up screen. Microsoft did restrict licensees from writing outMicrosoft code, a not uncommon· feature in the software market; many ofMicrosoft's rivals also integrate functions and impose similar restrictions on deletingcode.

6


explicit exclusive dealing agreements with PC manufacturers.Prominent computer makers such as Dell, Compaq, Gateway,and thousands of so-called resellers that package almostone half of all new PC systems, were free to install Netscape'sbrowser Navigator (or any other browser) if they so desired.Thus, Microsoft's product integration in and of itself did notcreate any physical foreclosure of rivals. 10

Microsoft's successful product integration may well havelowered Netscape's market share, but that is another matterentirely. If consumers preferred the integrated browser fromMicrosoft, they may have lowered their demand for alterna-tive browsers; Microsoft would do more business and itsrivals would do less. But, as we will argue in the followingpages, this sort of consumer choice does not restrain tradeor reduce competition. Indeed, the competitive process isenhanced when firms take business away from other firmsand overall trade is expanded when, say, a fully integratedbrowser works more effectively for consumers.

The antitrust authorities also held that Microsoft was ableto leverage its monopoly power in operating systems intothe browser market and harm consumers. This argument isunconvincing. First, if Microsoft's operating system wasalready leased at a price which maximized profit, there wasno additional leverage to exploit browser users. In addition,it made no economic sense to dilute the value of a superi-or product (operating system) with an alleged inferior add-on product (browser). Finally, Microsoft gave away itsbrowser for free, poor evidence, indeed, of any leverage orconsumer injury. Clearly, an operating system with a freebrowser is better for consumers than one without a browseror one with a browser at some additional cost.

lOpe users can download browsers, including Navigator, directly from the web.Netscape reportedly distributed over 100 million copies of its own browser in 1998.Wall Street Journal, November 6, 1998, p. A3.

7


As usual, the government has the economic logic back-ward. Tying or product integration is not necessarily an ele-ment of monopolization; indeed, it can be an importantcomponent of vigorous rivalry. Microsoft's decision to inte-grate the ,prowser into the operating system was intendedto be a more effective way of competing with other firmsthat already had included Web browsing technology intheir operating systems (Apple Computer) and with newerrivals, like Netscape, that established a dominant positionwith an improved independent browser. Thus, when theantitrust authorities and Microsoft's rivals complainedabout integration or predatory pricing, they were actuallycomplaining about the rigors of the competitive process,not about any monopolization.

The same sort of argument applies to Microsoft's agree-ments with Internet service providers which were said tobe restrictive of competitors. The fact remains that all busi-ness contracts are restrictive. All contractual agreementsforeclose options and exclude some alternatives. And con-tracts that last a year are more exclusionary than those thatlast a week. But this approach to restrictive practices can-not be the focus of antitrust analysis-unless we want pub-lic policy to micro-manage all business contracts. The focusof antitrust analysis, assuming we have the laws, ought tobe: do private agreements effectively restrict market outputand raise market prices? Clearly, the evidence in the PCindustry is that free-market contractual agreements haveled to massive increases in output and sharp reductions inprices to consumers. That, frankly, should be the end of thematter.

Ironically, if Microsoft had restricted its licensing ofWindows to a few select firms only, it would have beenaccused of monopolizing in restraint of trade. If Microsofthad charged exorbit~nt prices for its intellectual property, it

8


would have been accused of exploiting its monopolypower. Or if it had refused to integrate applications soft-ware packages, it would have been accused of repressinginnovation and shelving developments in order to enjoythe quiet life of a monopolist.

Instead, Microsoft engaged in' precisely the oppositebusiness behavior. It licensed its software to any and alllegitimate PC manufacturers (throughout the world) whilelicensing fees for its operating systems had averaged lessthan 3 percent of the cost of the personal computers in1996.11 It progressively integrated various applications soft-ware at minimal cost to the consumer. And all of this wasaccomplished without any government subsidy, legal barri-ers to entry, or regulation. Yet the critics, misled by market-share statistics and the anguished sobs of competitors, stillspied some evil monopolization. And in their regulatoryfrenzy, they threatened to smash one of America's mostsuccessful business organizations.

The Lorain Journal Case

Robert H. Bork, a supporter of the government antitrustsuit against Microsoft, has argued that Lorain Journal/ 12 anobscure 1951 antitrust case, can serve as an exact parallelwith the case against Microsoft.13 In Lorain, the town's onlynewspaper engaged in strict exclusive-dealing advertisingagreements with local merchants in order to prevent themfrom supporting a rival radio station. The government suedsuccessfully to end the exclusive dealing contracts.

The facts and argument in Lorain have nothing to do withthe Microsoft situation.14 Microsoft's general licensing

11 Wall Street Journal, December 2, 1998, p. B6.

12342 U.S. 143 (1951). The lower court decision is 92 F. Supp. 794 (Ohio 1950).

13Robert H. Bork, Letter to the Editor, Wall Street Journal, May 15, 1998.1400minick T. Armentan9, "Why Robert Bark is Wrong: Microsoft and the Lorain

Journal Case," On Point, Competitive Enterprise Institute, August 19, 1998.

9


agreements with PC manufacturers did not require that theyboycott the products of Microsoft's rivals. Manufacturerswere generally free to load competitive software and werefree to promote their own content on the Windows open-ing screen. In addition, consumers were free to add or elimi-nate any product from Windows and free to replace theentire opening screen (if they wished) with a few mouseclicks. Moreover, Microsoft was not the only operating sys-tem (newspaper) in town, nor did it face one lonely gov-ernment-licensed competitor (radio station). Finally,Microsoft could make strong efficiency arguments for inte-grating its browser and operating system,15 arguments thatcould not be made conclusively for the strict exclusive-dealing contracts in the newspaper case. In short, LorainJournal and the case against Microsoft have nothing of sub-stance in common.

Through the Looking Glass

The Microsoft case highlights the intellectual bankruptcyof antitrust policy. The industry was legally open; therewere numerous competitors of various sizes; technologicalchange was rapid and continuous; outputs expanded andprices had fallen dramatically; the leading software firmlicensed its operating system widely and at reasonableprices; and competitors constantly complained about therigors of the competitive process. Ironically, the govern-ment's trial case against Microsoft was heavily predicatedon explicit evidence of vigorous competition: internalmemos and e-mail correspondence that speak clearly toMicrosoft's intent to bury its rivals and emerge victorious inthe software and browser·wars.16 In professional sports, such

15Robert A. Levy, "Microsoft and the Browser Wars: Fit to be Tied," Cato InstitutePolicy Analysis, no. 296, February 19, 1998.

16rhe government commandeered over 3 million pages of internal Microsoft cor-respondence. Much of the actual trial was taken up with debate over the meaning

10


locker room bravado would clearly be seen as evidence ofcompetitive rivalry. Only in the Alice in Wonderland worldof antitrust regulation could competitive free speech andrivalrous performance in the marketplace be transformedmagically into some sinister monopolization scenario.

Economics aside, the government prosecution ofMicrosoft was also a travesty of common-sense justice.Microsoft had a property right to the software that itowned and innovated profitably; it had a property right towrite any new code that improved computer applications;it had a property right to insist that licensees not write outany part of its operating system program; it had a propertyright to determine the length of its software lease and whatprice to charge for its property; and it had a property rightto freely compete or cooperate with any rival. Yet, antitrustsought to emasculate these basic rights and impose selec-tive restrictions on Microsoft's freedom while leaving itsenvious rivals conspicuously unrestricted.17

Finally, the government's attempt at industrial planningin the computer industry was hopelessly naive; the techno-logical framework and consumer preferences change fartoo rapidly. Regulation here will create additional incen-tives to litigate outcomes rather than have them market-determined. It will also create strong disincentives for dom-inant firms to innovate and compete aggressively for mar-ket share. In short, antitrust will have achieved the oppositeof the results intended: it will have punished success,restrained efficient competition and hampered economicgrowth.

Microsoft is simply the latest in a long line of firms thathas been punished for its virtues, for the simple fact that its

and intent of executive e~mail. See, for example, Wall Street Journal, November 17,1998, p. 86.

17The Department of Justice had sought a preliminary injunction to require thatMicrosoft offer Netscape's browser with Windows or, alternately, sell its own browserseparately. Wall Street Journal, May 19, 1998, p. A3.

11


overall efficiency resulted in a substantial market share.Antitrust's dirty little secret is that the laws have beenemployed consistently to hamper successful businessorganizations and protect their less efficient rivals.18 Onewould be hard-pressed to discover a more immoral or irra-tional public policy toward business, or one more worthyof repeal.

18As an example, United Shoe Machinery Corporation had held its dominant mar-ket position for decades with superior innovation and competitive pricing. Nonetheless,a lower court determined that United's overall efficiency had illegally "excluded"rivals and eventually the Supreme Court divested the company. See United States v.United Shoe Machinery Corporation, 110 F. Supp. 295 (1953) and United States v.United Shoe Machinery Corporation, 391 U.S. 244 (1968).

12

2. The Case Against Antitrust Policy

The uptick in antitrust enforcement and the irrational attackon Microsoft should not distract us from the larger andlonger picture: the intellectual case against antitrust regula-tion has been building for decades.

The most important theoretical development has beenthe increasing professional disenchantment with the so-called barriers-to-entry doctrine.1 This doctrine held thatcertain economic obstacles prevented smaller firms fromcompeting with so-called dominant firms, that barriersenhanced the market power of these leading companies,and that they served to harm consumer welfare. Yet, mostof these alleged barriers have proven to be economies andefficiencies that leading firms have earned in the market-place. Efficiency and successful product differentiation cancertainly limit rivalry with firms unable to match or surpasssuch innovation; superior economic performance canmake it difficult for new firms to enter markets or for oldfirms to expand their market shares. But none of this isunfair or unfortunate from any consumer perspective, andnone of it can rationalize an antitrust attack on the firmswith the superior performance.

A reexamination of the antitrust case evidence alsotended to support administrative reforms in antitrust policy.

1For an excellent criticism of the traditional barriers-to-entry doctrine, see Robert H.Bork, The Antitrust Paradox: A Policy at War with Itself, (New York: Basic Books,1978), chap. 16. See also Harold Demsetz, "Barriers to Entry," American EconomicReview 72 (March 1982): 47-57.

13


By at least the mid 1970s it was becoming clear that muchof the antitrust case history did not confirm the resourcemisallocations suggested by orthodox monopoly theory.Indeed, economic analysis of the leading antitrust casestended to demonstrate that the indicted corporations hadincreased their outputs and lowered their prices and hadbehaved generally as competitive firms would be expectedto behave in open markets facing direct or potential riv-alry.2 The thrust of antitrust policy in these cases was, if any-thing, to restrain the competitive performance of the lead-ing firm and thus protect the existing market structure ofgenerally smaller, less efficient business organizations.

The IBM Case

These findings were perhaps best exemplified in u.s. v.IBM, 3 the disastrous government antitrust case against theInternational Business Machines Corporation (IBM) thatcontributed significantly to the movement away from tradi-tional antitrust policy. IBM was indicted by the Departmentof Justice in 1969 and charged with illegal monopolizationof the general-purpose digital-computer-systems market.The suit held that IBM had systematically engaged in cer-tain exclusionary business practices that tended to restraintrade and create a monopoly in violation of the ShermanAntitrust Act (1890). The case finally went to trial in 1975.After more than six years in court and a trial transcript ofmore than 104,000 pages, the case was abandoned by thegovernment in 1982.

It was clear from the start that this government antitrustcase and the many private antitrust cases against IBM4

2Dominick T. Armentano, Antitrust and Monopoly: Anatomy of a Policy Failure,2nd ed. (Oakland, Calif.: Independent Institute, 1990).

3United States v. International Business Machines Corporation, Docket no. 69,Civ. (ONE) Southern District of New York (1969).

4Many companies, including Greyhound, Telex, Cal Comp., and Memorex, suedIBM under the antitrust laws. Most of these cases were resolved in IBM's favor. See

14

The Case Against Antitrust Policy

were all fundamentally misguided. They were, in brief,attacks on entrepreneurial success and efficiency. Clearly,IBM had not restricted production to raise prices and prof-its; nor had it repressed invention and innovation. On thecontrary, IBM had achieved its considerable success andmarket share by taking unprecedented research-and-devel-opment risks, innovating superior products, and developingan unsurpassed, long-term corporate commitment to cus-tomer-support services.5 Most of the alleged unfair prac-tices, such as educational discounts and bundled hardwareand software, were only "exclusionary" of less efficient sell-ers-some larger than IBM, some smaller-that could notmatch IBM's overall market performance.

In addition, and contrary to the assertions of the gov-ernment and the private plaintiffs, IBM's considerable busi-ness success had not hurt the overall growth of non-IBMcompanies and the data-processing industry generally. IBMhad grown rapidly, but the industry had grown far morerapidly; IBM's share of domestic electronic data-processingrevenues declined from 78 percent in 1952 to 33 percentin 1972, hardly persuasive evidence of any monopoliza-tion.6 Assistant Attorney General William Baxter under-stood the true state of affairs when, in 1982, his office with-drew its absurd legal action, terming it "without merit."7

The collapse of the concentration doctrine also stronglyinfluenced a new direction in antitrust policy.8 Early empirical

Franklin M. Fisher, James W. McKie, and Richard B. Mancke, IBM and the DataProcessing Industry: An Economic History (New York: Praeger Publishers, 1983),pp.448-49.,

5Franklin M. Fisher, John J. McGowan, and John E. Greenwood, Folded, Spindled,and Mutilated: Economic Analyses and U.s. v. IBM (Cambridge, Mass.: MIT Press,1983).

6Ibid" p. 111.

7Wall Street Journal, January 11, 1982, p. 3.

8Harold Demsetz, The Market Concentration Doctrine, American EnterpriseInstitute-Hoover Institution Policy Studies (August 1973); idem, "Industry Structure,Market Rivalry, and Public Policy," Journal ofLaw and Economics 16 (April 1973): 1-9.

15


work in industrial organization had appeared to discover aslight positive correlation between market concentration (thepercentage of the market sales or assets controlled by asmall group of firms, usually 'four) and the average profitsearned by firms in such markets. Most of these studiesassumed that the so-called barriers to entry mentionedabove limited competition in the concentrated industriesand allowed firms monopoly profits.9

later research argued, however, that the higher profitsin the concentrated markets were more logically explainedby the fact that the leading firms had lower costs and thatthese efficient firms had grown more quickly than the lessefficient firms. In addition, over the long run, profit ratestended to decline in the high-concentration markets and toincrease in the low-concentration markets, indicating thatthe competitive-market process of resource reallocationwas alive and well. In short, evidence from the so-callednew learning undercut much of the rationale for the tradi-tional antitrust regulation of market concentration andhigh-market share.10 A new direction in antitrust policy wasinevitable and emerged in the 1980s.

But not all traditional antitrust policies were aban-doned. Antitrust was still very much concerned with price-fixing and market-division agreements between competi-tors (horizontal agreements), and neither the antitrustauthorities nor the courts relaxed their position that sucharrangements were normally illegal per see In addition, certain

9See, for instance, Joseph S. Bain, "Relation of profit Rates to Industry Concentra-tion: American Manufacturing, 1936-1940," Quarterly Journal of Economics 65(August 1951): 293; and H. Michael Mann, "Seller Concentration, Barriers to Entry,and Rates of Return in Thirty Industries: 1950-1960," Review of Economics andStatistics 48 (August 1966): 296-307.

10Ya1e Brozen, "Concentration and Profits: Does Concentration Matter?" AntitrustBulletin 19 (1974): 381-99; John R. Carter, "Collusion, Efficiency, and Antitrust,"Journal of Law and Economics 21, no. 2 (October 1978): 434-44. An excellent dis-cussion of the concentration and profit controversy appears in Harvey Goldschmid,H. Michael Mann, and J. Fred Weston, eds., Industrial Concentration: The NewLearning (Boston: Little, Brown, 1974).

16


interfirm cooperative joint ventures were still subject to reg-ulation by the appropriate antitrust authorities. Resale pricemaintenance and so-called predatory practices remainedillegal under the antitrust laws. The Department of Justiceand the Federal Traqe Commission still regulated horizon-tal mergers through revised merger guidelines. Andalthough the merger attitudes and guidelines were some-what more relaxed than they had been in previous years,the antitrust authorities continued to intervene in certainbeer, office supply, and telecommunications industry con-solidations. In short, although the focus of antitrust enforce-ment changed somewhat in the 1980s and early 1990s, theantitrust authorities still remained active in the areas of pricefixing, mergers, and restrictive practices, where it was allegedthat firms were able to harm social welfare.

There has been some progress made in moving awayfrom the gross irrationalities of old-style traditional enforce-ment. And some critics of traditional enforcement might betempted to be content with these modest administrativechanges, or even more tempted to call for additional reform,such as the general adoption of a rule of reason withrespect to certain "restrictive" practices, such as tying agree-ments or resale price-maintenance contracts. But the resur-gence of antitrust enforcement in the 19905 indicates thatthis reform approach is naive. Thus, it will be argued belowthat even additional reforms will not be sufficient and thatthe case against antitrust regulation is strong enough to jus-tify the complete repeal of all the laws.

'The Case for RepealThe case for the repeal of the antitrust laws can be

summarized as foHows:

First, the laws misconstrue the fundamental nature of bothcompetition and monopoly. Competition is an open marketprocess of discovery and adjustment, under conditions of

17

Antitrust: The Case for .Repeal

uncertainty, that can include interfirm rivalry as well asinterfirm cooperation. Within this competitive process, afirm's market share is not its market power, but a reflectionof its overall efficiency. Monopoly power, on the otherhand, is always associated with legal, third-party restraintson either business rivalry or cooperation, not with strictlyfree-market activity.

Second, the history of antitrust regulation reveals thatthe laws have often served to shelter high-cost, inefficientfirms from the lower prices and innovations of competitors.This protectionism is most obvious in private antitrust cases(in which one firm sues another) which constitute morethan 90 percent of all antitrust litigation.

Third, some of the antitrust laws, such as section 2 ofthe Clayton Act (1914) and the Robinson-Patman Act(1936), explicitly intend to restrict price rivalry in the nameof preserving competition. Government antitrust suitsagainst firms that price discriminate almost always result inthe defendant firm raising some of its prices to comply withthe law.

Fourth, section 7 of the Clayton Act, which restrictsmergers that may tend to lessen competition, is itselfdestructive of the competitive process. Restricting mergersand takeovers may i"nhibit the flow of production into thehands of more efficient managers. The anti-competitiveeffect of section 7 is especially evident in vertical integra-tion antitrust cases and in cases in which poorly perform-ing domestic firms may require merger or other forms ofcooperation in order to compete more successfully withforeign firms. Even with somewhat relaxed attitudes towardsome mergers and with revised merger guidelines, the FTCand the Antitrust Division of the Justice Department havecontinued to regulate, delay, and oppose many importantbusiness consolidations.

Fifth, the antitrust laws are a form of government regu-lation, and, like all government regulation, they tend to

18


make the economy less efficient. In the name of preservingcompetition, the efficient competitive process has itselfbeen impeded by antitrust intervention. Firms that intendto lower their prices may be restricted from doing so byantitrust law. Even more important and pernicious, firmsthat would innovate some new process or product mustconsider whether the innovation will give them an "unfair"competitive advantage or be termed "predatory" by theantitrust regulators or some competitor.

Sixth, the enforcement of the antitrust laws is predicatedon the mistaken assumption that regulators and the courtscan have access to information concerning social benefits,social costs, and efficiency that is simply unavailable in theabsence of a spontaneous market process. Antitrust regu-lation is often a subtle form of industrial planning and isfully subject to the "pretense-of-knowledge" criticism fre-quently advanced against government planning.

Seventh, the antitrust laws have been enforced arbitrar-ily, violate traditional notions of due process of law, andalways interfere with the rights of property owners or theirtrustees to make, or not make, voluntary agreements. AsAdam Smith observed more than two hundred years ago,a law that interferes with private and voluntary agreementscannot be "consistent with liberty and justice."ll

Finally, the modest progress made to date in antitrustreform has been only administrative. Administrativechanges and reforms are helpful and should not be under-estimated. But they should not be overestimated, either.The antitrust statutes-even the blatantly anticonsumerRobinson-Patman Act-remain firmly in place, and much ofthe current enforcement effort is still traditional in natureand, therefore, thoroughly misguided.

11 Adam Smith, An Inquiry Into The Nature and Causes of The Wealth of Nations(New York: Modern Library, [1776J 1937), p. 128.

19

Antitrust: The. Case for Repeal

Regulatory changes in the air-carrier industry illustratethe wisdom of total repeal as opposed to reform. By themid-1970s it had become clear that government regula-tion of this industry, under the Civil Aeronautics Act of1938, had worked to restrict entry, encourage wastefulpractices, and raise costs and prices generally to air-trans-portation consumers.12 Theoretical criticism of airline regu-lation by economists accelerated. The empirical evidencethat air-carrier regulation was inefficient and that a freemarket would work more efficiently became overwhelm-ing. Theory and evidence were then cogently crafted into asolid political case for massive deregulation of the air-carrierindustry.

It is important to note that the argument was not thatthe Civil Aeronautics Board (CAB) should do less in theway of regulation or that it should do something else. Theargument was that the CAB itself-the entire regulatorystructure-should be abolished and an open-market processbe allowed to operate in its place. The necessary and suffi-cient reform here was the total repeal of the economic reg-ulatory structure, which occurred when Congress terminatedthe CAB on January 1, 1985.13

Air-carrier deregulation would not have worked as wellhad the existing regulatory structure been maintained.Deregulation often requires a painful reallocation ofresources that is sure to hurt special interests, and in the air-carrier industry this process was especially painful. Strongsentiment quickly developed for reregulation, lest Americalose its"national transportation system." But in the absence

12George Douglas and James Miller, Economic Regulation of Domestic AirTransport (Washington, D.C.: Brookings Institution, 1974). For an account of theresults of airline deregulation, see John E. Robson, "Airline Deregulation: TwentyYears of Success and Counting," Regulation 21 (1998): 17-22.

13Some of the CAB's regulatory powers, including the authority to regulate airlinecomputer-reservations systems, were shifted to the Department of Transportation(DOT). See Regulation 9 (January/February 1985): 8. For the antitrust implications ofDOT regulation of computer-reservations systems, see Antitrust and Trade RegulationReporter 48, no. 1207 (March 21, 1985): 505.

20


of any continuing regulatory structure, the general laissez-faire momentum that had been set in motion could not bereversed. As in all such cases, the results of air-carrier de(eg-ulation have been enormously beneficial to consumers.

There is an important lesson for critics of traditionalantitrust policy here. The administrative changes in antitrusthave been transitory. Since the entire regulatory antitruststructure still exists-the laws, the courts, the agencies-thestructure has been activated and employed more strictly bydifferent administrators holding different theories. If thecase against antitrust regulation is overwhelming, the entireantitrust framework must be abolished.

Theories of Antitrust Policy

It will not be easy to repeal the antitrust system. Antitrustregulation is a firmly entrenched institution in America andhas been since 1890. This section explores some of the rea-sons for the persistent faith in antitrust regulation-despiteits record-and speculates on the more subtle meaning ofantitrust.

Antitrust as Public Interest

The primary reason for the widespread support forantitrust enforcement is a belief that the laws still serve,however imperfectly, to protect the economy (consumers)from the economic abuses commonly associated with pri-vate monopoly and private monopoly power. This per-spective can be termed the "public interest" theory ofantitrust policy.

The notion of competition is enormously popular inAmerican society. We expect and enjoy competition insports and in business. In business, competition is said tokeep organizations alert and efficient. Business competitiongives consumers quality products at low prices, providesbuyers with alternative suppliers, forces poorly managed

21


firms out of the market, and limits and restricts so-calledeconomic power.

Monopoly appears antithetical to all of this. Businessmonopoly is said to deaden initiative and efficiency, restrictproduction, raise prices, exclude competitors from the mar-ket, and misallocate economic resources. It can be eco-nomically and even politically dangerous. It is a short stepfrom these impressions to supporting a law that encouragescompetition and prohibits business monopoly-that is, anantitrust 'law.

Academic economists have crafted these impressionsconcerning competition and monopoly into an elaboratetheoretical paradigm that serves to legitimize someantitrust regulation. Put briefly, this theory holds that freemarkets can occasionally fail to work in the best interests ofsociety generally. This market failure can occur wheneverprivate business organizations gain monopoly power, thepower to restrict production and raise market price. Suchfirms can produce less and charge more, and they general-ly have higher costs than comparably competitive businessorganizations. A law that prohibits free-market monopo-lization would appear to promote increased outputs, lowercosts, and lower prices for consumers. Antitrust law, there-fore, exists to protect the public interest from the power offree-market monopoly.

There are at least two ways to analyze this public-inter-est perspective on antitrust policy. One way is to challengethe theoretical models of competition and monopoly uponwhich it is so heavily dependent. If the models are funda-mentally deficient, then the scientific case for antitrust isweakened substantially. The other way to challenge thepublic interest perspective is to study the actual conductand performance of business organizations that have beenconvicted under the antitrust laws. If such firms were foundnot to be restricting production and raising prices-if,indeed, they have been increasing outputs and lowering

22


prices-then the public-interest theory of antitrust regula-tion would be all but demolished.

Antitrust as RegulationAn entirely different perspective on antitrust policy is to

see it as an example of special-interest regulation. Govern-ment regulation in America has often been associated withspecial-interest groups, usually business groups, that haveattempted to use legislation to gain and hold economicadvantages (or rents) not obtainable in a free market.14

These advantages are often secured by legal barriers toentry and competition that serve to restrict production andincrease prices. Import quotas in the textile industry, forexample, have had the effect of protecting domestic textilecompanies from foreign competition while inflicting mas-sive economic losses on consumers.15

Antitrust, despite disclaimers, is government regulation.Whether antitrust was originally intended to promote andprotect special business interests can never be known withabsolute certainty, although there is some evidence this mayhave been the case.16 But, as will be demonstrated below,there is adequate evidence that antitrust has often beenemployed as special-interest legislation. In practice, antitrust

14George J. Stigler, "The Theory of Economic Regulation," BellJournal of Economicsand Management Science 2 (Spring 1971): 3-21; Sam Peltzman, "Toward a MoreGeneral Theory of Regulation," Journal of Law and Economics 19 (August 1976):211-40. For a review of the rent-seeking literature, see Robert D. Tollison, "RentSeeking: A Survey," Kyklos 35 (1982): 575-602.

15See "Economic Effects of Significant U.S. Import Restraints," Publication 2935(Washington, D.C.: International Trade Commission, December 1995). The eco-nomic losses were estimated at roughly $1 0 billion annually.

16Thomas J. Dilorenzo has shown that outputs in the "trust" industries-far frombeing restricted-expanded rapidly in the decade prior to the Sherman Act of 1890.He has also argued that Sen. John Sherman's motives in sponsoring the act may havebeen ambiguous. See Thomas J. Dilorenzo, "The Origins of Antitrust," InternationalReview of Law and Economics 5 (1985): 73-90. See also Thomas W. Hazlett, "Thelegislative History of the Sherman Act Reexamined," Economic Inquiry 30 (1992):263-76.

23


has been protective of existing market structures-much liketariff and quota protection-and has served to keep costsand prices higher to final consumers. Antitrust defendantshave lost cases because their efficient performance-lowprices and successful innovations-has been ruled "exclu-sionary" of less efficient competitors. In private cases,especially, antitrust has often been employed as a club byplaintiff firms anxious to restrain the price and innovationalrivalry emanating from efficient defendant corporations.17

And since private cases constitute the vast majority of allantitrust litigation, they reveal the fundamental nature ofantitrust policy. In short, antitrust, like almost all gov-ernment regulation, has often served to benefit some at thegeneral expense, a result fully anticipated by much of thepublic choice Iiterature.18

If this perspective on antitrust regulation is correct,antitrust law will actually be harder, not easier, to repeal-oreven to additionally reform. The social costs of such special-interest legislation such as antitrust are normally spread verythinly over society as a whole; consider for example, the percapita costs of nonsense cases such as the thirteen year gov-ernment war on IBM or the irrational assault on Microsoft.Yet, the benefits of antitrust regulation are frequently con-centrated on very special interests-the antitrust establish-ment-and those benefits can be substantial. Antitrust attor-neys, private plaintiffs, consultants, and the antitrustbureaucracy itself have much to gain from a continuation ofantitrust regulations and much to lose from any repeal of orreduction in antitrust enforcement. Consequently, thebeneficiary' groups have every incentive to strenuously resistreform and repeal and to denounce all antitrust critics in the

17William J. Baumol and Janusz A. Ordover, "Use of Antitrust to SubvertCompetition," Journal of Law and Economics 28 (May 1985): 247-65.

18See, for instance, Robert D. Tollison, "Public Choice and Antitrust," Cato Journal4, no. 3 (Winter 1985): 905-16. See also William F. Shughart II, Antitrust Policy andInterest Group Politics (New York: Quorum Books, 1990).

24


most strident tones. Ordinary citizens and consumers, onthe other hand, have little incentive to rally against theantitrust juggernaut, little incentive even to educate them-selves as to the antitrust facts of life. This cost-benefit calcu-lus makes any attempt to repeal the antitrust laws difficult,unless that calculus can be changed.

Antitrust as Industrial PolicyA third perspective on antitrust is to see it as one of

America's oldest industrial policies. Industrial policy is gov-ernment industrial planning, and much of antitrust policy isa kind of government planning. For example, the JusticeDepartment and the FTC publish detailed merger guide-lines that proscribe legally permissible business consolida-tions. Indeed, they often intervene in mergers, even whilepermitting them, requiring that firms sell certain assets orcompanies. As an example, the merger of Texaco andGetty Oil was FTC-approved pending the sale of 600 serv-ice stations, certain pipelines, and several refineries; theGulf-Chevron merger was FTC approved after an agree-ment was reached to sell 4,000 Gulf stations and a majoroil refinery.19

Further examples of antitrust industrial policy includethe FTC's authority to review the costs and benefits of jointbusiness ventures and to grant or deny approval of inter-firm cooperative agreements. The antitrust authorities canmove against firms that fix resale prices, charge low (preda-tory) prices, charge high (monopoly) prices, and chargeprices that are the same (collusion). And the FTC candecide to oppose the 1997 merger of Staples and OfficeDepot based upon some arbitrarily narrow definition of therelevant market (see chapter 6).

190il and Gas Journal, January 23, 1984, p. 48; Oil and Gas Journal, February 6,1984, p. 84; Oil and Gas Journal, July 16, 1984, p. 43; Wall Street Journal, April 24,1984, p. 4. The Hart-Scott-Rodino Antitrust Improvement Act of 1976 requires pre-notification of certain size mergers.

25


This point about industrial planning and policy is empha-sized not to quibble over labels but to point out thatantitrust, like other government-planning policies, is subjectto criticism on the grounds that it always assumes the exis-tence of the information it requires for intelligent decisionsconcerning social efficiency. As will be argued later, thecost-benefit information that would be required for intelli-gent choices concerning mergers and divestitures is pro-duced and discovered only through a working out of theopen-market process and is knowable only to the particu-lar individuals involved in that process. Antitrust authoritiesand courts continually presume the existence of such infor-mation when they prohibit a merger, deny a joint venture,break up a company, or rule that certain prices are preda-tory. Yet, if antitrust regulators and courts cannot obtainaccurate information concerning future social costs andbenefits, no rule of reason in antitrust is really possible. Thus,the case against any antitrust regulation is all the stronger.

The AT&T Case

Those who argue that antitrust is not government-indus-trial planning will have difficulty explaining the historic deci-sion to break up the American Telephone and TelegraphCompany (AT&T) arguably the most significant employmentof antitrust regulation in the history of antitrust enforcement.This historic consent decree, among other things, divestedthe 22 operating telephone companies from AT&T andended a portion of a 1956 consent decree that had pre-vented AT&T from competing in nonregulated markets, suchas data processing.20 Ending the 1956 consent decree-alegal restriction on market entry and competition-wasentirely consistent with permitting a spontaneous marketprocess to exist in telecommunications and data processing.

20United States v. AT&T, 524 F. Supp. 1336 (1981); United States v. AT&T, 552 F.Supp. 131 (1982).

26


Divesting the operating 'companies and reorganizing theminto seven regional companies was, however, an unprece-dented experiment in antitrust industrial planning.

A number of economic arguments were employed tojustify the divestiture of the operating telephone companies.The first was that AT&T's ownership of the operating com-panies served as a bottleneck to potential long-distancecompetitors. AT&T's ownership of the operating companies,so the argument went, placed it in a position to deny anycompetitor fair access to the bulk of the business and resi-dential telephone market. The second argument was thatdivestiture would reduce the potential threat of cross-subsidization of revenues from regulated markets to unreg-ulated markets and end the necessity of restricting AT&Tfrom entering unregulated markets. Finally, the divestiturewould serve to weaken the grip of AT&T's Western ElectricCompany on the telephone equipment market (since theoperating companies had ordered the bulk of their equip-ment from Western), leading to additional innovation andlower equipment prices.

These arguments are not entirely implausible, and theAT&T divestiture may well have led to the results anticipated.But how did its supporters know that the assumed, futurebenefits of divestiture would exceed its costs? For example,even Robert W. Crandall and Bruce M. Owen, in theirexcellent discussion of the divestiture issues, concede thatthe absence of any direct evidence of AT&T's pre-divesti-ture vertical-integration joint economies made it "very diffi-cult to prove that the divestiture is necessarily welfareenhancing."21

Indeed, most consumer difficulties in telecommunica-tions did not relate directly to vertical integration and divesti-ture at all; government regulation, not vertical integration

21 Robert W. Crandall and Bruce M. Owen, "The Marketplace: EconomicImplications of Divestiture," in Disconnecting Bell: The Impact of the AT&TDivestiture, Harry M. Shooshan ed. (New York: Pergamon Press, 1984), p. 57.

27


per se, had been the primary obstacle to a truly open-mar-ket competitive process in telecommunications.22 TheFederal Communications Commission has long restrictedentry into long distance telecommunications and had reg-ulated the rates of the monopoly supplier, AT&T. Entry intolocal telephone markets had been legally restricted by stategovernments, and phone service and rates had been regulat-ed by public utility authorities; the dominant supplier was,again, AT&T. This regulation was not, of course, accidental.AT&T had a long history prior to divestiture of advocating gov-ernment regulation and monopoly in telecommunications,and of opposing attempts to increase competition bydecreasing government regulation.

Most of the alleged difficulties associated with AT&T'svertical integration-and most of the alleged benefits associ-ated with divestiture-were difficulties that would have beenovercome in time by complete deregulation. Cross-subsi-dization, for instance, becomes a serious issue only in a reg-ulated setting where a firm might choose, say, to financeprice-cutting wars in unregulated markets out oJrevenues orprofits earned in regulated markets. Ending the regulationends the possibility of "unfair" cross-subsidization. In addi-tion, Western Electric's near capture of the operating-comp-any market for telephone equipment is controversial onlybecause the operating companies can pass along, underregulation, all of the inflated equipment costs to the final con-sumer of phone services. In an openly competitive market,consistent noncompetitive purchases of materials by vertic-ally integrated companies would normally result in a severeloss of market share for those companies-a strong incentiveto change the practice. Again, it was regulation, not verticalintegration, that was the ultimate source of the difficulty.

22Roger G. Noll and Bruce M. Owen, "The Anticompetitive Uses of Regulation:United States v. AT&T," in John E. Kwoka, Jr. and lawrence J. White, eds., The AntitrustRevolution (Boston: Scott, Foresman, 1990) pp. 290-337.

28


Even the so-called bottleneck and access issues are for-ever clouded by the' fact that, under divestiture, no open-mar-

. ket access value exists for the rival long-distance companies.The current access fees are not market determined but areset under the authority of the FCC. In the absence of truemarket values, even supporters of divestiture cannot besure that the existence of rival wire-line suppliers actuallyimproved overall resource efficiency and advanced the elu-sive public interest.23

ConclusionsVery little academic or public credence is given to

antitrust policy as special-interest regulation or as govern-ment-industrial planning. Government regulation and plan-ning have been sharply criticized by economists and, byand large, have been professionally discredited.24 Whatsupport now remains for antitrust policy would appear todepend upon the public-interest perspective; that is, thebelief that some antitrust regulation is necessary to preventmarket failure.

In the following chapters, the public-interest theory ofantitrust will be critically examined to determine whether thestandard theories of competition and monopoly employedto explain market failure actually make sense and whetherthe classic antitrust cases contain evidence that free-marketmonopoly can exist and misallocate resources. If antitrusttheory and history are internally consistent, then someantitrust policy may be appropriate. If, however, they are

23There was early evidence that overall "consumer interests" were not advancedby the divestiture. See Paul w. MacAvoy and Kenneth Robinson, "losing by JudicialPolicymaking: The First Year of the AT&T Divestiture," Yale Journal on Regulation 2(1985): 225-62.

24See, for example, Robert W. Poole, Jr., ed., Instead of Regulation (lexington,Mass.: lexington Books, 1983). An excellent critical analysis of the entire govern-ment-planning paradigm by many authors can be found in Cato Journal 4, no. 2 (Fall1984). For a definitive book-length criticism of government planning see Don lavoie,National Economic Planning: What is Left? (Cambridge, Mass.: Ballinger, 1985).

29


inconsistent, then the public-interest perspective and the pol-icy it supports deserve to be rejected, not simply reformed.Without a scientific public interest justification, there is norationale for any antitrust regulation in a market economy.

30

3. Competition and Monopoly:Theory and Evidence

Much of the support for antitrust policy depends upon thecorrectness of the standard theories of competition andmonopoly. These can be briefly summarized as follows.

The Theories

Some economists define competition as a state of affairsin which rival sellers of some homogeneous product are sosmall-relative to the total market supply-that they individ-ually have no control over the market price of the product.1

These atomistic sellers take the market price as given andthen attempt to generate an output that maximizes theirown profit. The final outcome (equilibrium) of such a mar-ket organization of firms is that consumers obtain theproduct at the lowest possible cost and price. Such marketsare said to be "purely" competitive ("perfectly" competitiveif there is perfect information), and resources are said to beallocated efficiently.

Free-market monopoly involves some voluntary restric-tion of market output relative to the output forthcomingunder competitive conditions. Economists usually assumethat monopoly means that there is only one supplier of a

1The standard theoretical analysis of competition, monopoly, and resource mis-allocation can be found in any microeconomics text and in most texts on antitrustpolicy. See, for instance, William F. Shughart II, The Organization of Industry, 2nd ed.(Houston, Texas: Dame Publications, 1997).

31


product with no reasonable substitutes or that severalmajor suppliers of a product collude to restrict production.The economic effect of such monopolization is that marketoutputs are restricted-the monopoly restrains trade-andprices are increased to consumers. Such restrictions of pro-duction are also said to misallocate resources and reducesocial welfare.

The expression "misallocation of resources" is a power-ful one in economics. It signifies that scarce economicresources are not being put to their greatest economicadvantage. The implication is that some alternative alloca-tion of these resources could improve overall economicperformance.

Monopoly is said tomisallocate resources in two funda-mental ways. The first is termed "allocative inefficiency." Itimplies that the price consumers pay for a product undermonopoly-the monopoly price-exceeds the marginal costof producing that product. Consumers indicate their will-ingness to have suppliers produce more of some productby paying a price that exceeds the marginal cost of pro-ducing it. Firms with monopoly power, however, can maxi-mize their profits by restricting their production and keep-ing their prices up. Suppliers with monopoly power are saidto have no incentive to expand production to the pointwhere market price and marginal cost are equal. The con-sequence of such supply decisions is that resources are atleast somewhat misallocated and social welfare is reduced.

Monopolists are also said to be likely to expendresources to obtain monopoly positions and then expendadditional resources to retain them. Further, in the absenceof direct seller rivalry, monopoly suppliers can afford to beless efficient than competitive firms with respect to theirown use of resources. All of these extra expenses and inef-ficiencies can increase the cost function under monopolyrelative to competition and contribute to what is termed"technical inefficiency." In short, firms with monopoly

32

Competition and Monopoly: Theory and Evidence

power can produce less, charge more, and misallocate eco-nomic resources. Society would be clearly better off underconditions of competition, and the rationale for antitrustenforcement against monopoly is said to be obvious.

The Problem with Competition TheoryAlthough the standard theories of competition and

monopoly seem reasonable and would appear to rationalizesome antitrust enforcement, they pose some very serious dif-ficulties. Resource allocation under atomistic competitionmight well be efficient if perfect information existed or iftastes and preferences never changed, but it is difficult tounderstand the relevance of such a theory in a real worldof differentiated preferences, economic uncertainty, anddynamic change. The economic problem to be solved bycompetition is emphatically not one of how resourceswould be allocated if information were perfect and con-sumer tastes constant; with everything known and constant,the solution to a resource-allocation problem would be triv-ial. Rather, the economic problem lies in understanding howthe competitive market process of discovery and adjust-ment works to coordinate anticipated demand with supplyin a world of imperfect information. To assume away diver-gent expectations and change, therefore, is to assume awayall the real problems associated with competition and theresource-allocation process. Thus, although the standardefficiency criteria may be technically correct for a staticworld, they are irrelevant to actual market situations.

Market uncertainty and change may require differenti-ated products. They may also require some interfirm coor-dination, instead of independent rivalry, and even someprice cooperation. They may require some product andservice advertising, although none is required in the atom-istic equilibrium. These variables do not indicate that com-petition does not exist or that the competitive process is

33


defective or inefficient. They mean, simply, that the com-petitive process is in a necessary state of disequilibrium.The market process may, in the abstract, tend toward sometheoretical equilibrium, but it never reaches one.

Much of traditional antitrust enforcement has beenbased on erroneous notions of efficiency under static equi-librium conditions. Outputs falling short of the purely com-petitive-theoretical-output were said to have been"restricted." Market advertising, product differentiation, andinnovation were often said to be elements of monopolypower-not elements of a competitive pro'cess-that couldmisallocate resources and lower social efficiency. Any con-trol over market price was termed "monopoly power," andinterfirm cooperative agreements were regarded by econo-mists and the antitrust authorities with great suspicion. Yet,if the purely competitive equilibrium is not an appropriatewelfare benchmark, none of these traditional conclusionsmakes any sense.

An alternative perspective on competition is to see it asan entrepreneurial process of discovery and adjustmentunder conditions of uncertainty.2 A competitive processimplies that business organizations of various sizes contin-ually strive to discover which products and services con-sumers desire, and at what prices, and continually strive tosupply those products and services at a profit to them-selves and at the lowest cost.

This process of discovery and adjustment may encom-pass explicitly rivalrous behavior in the usual sense-directprice and nonprice competition-and it may also include var-ious degrees of interfirm cooperation, such as joint ventures

2F.A. Hayek, "The Meaning of Competition," in Individualism and EconomicOrder (Chicago: Henry Regnery, 1972), pp. 92-106. On the historical developmentof the distinction between the competitive process and the competitive equilibrium,see Paul J. McNulty, "Economic Theory and the Meaning of Competition," QuarterlyJournal of Economics 82 (November 1968): 639-56. ludwig von Mises termed thecompetitive process "catallactic competition." Ludwig von Mises, Human Action: ATreatise on Economics (New Haven, Conn.: Yale University Press, 1963), pp. 274-94.

34


and mergers. Interfirm cooperation and rivalry are notopposing paradigms from a market-process perspective.There is no a priori way, for example, to define the optimalsize of a cooperative business unit or, alternatively, the opti-mal number of rival firms for efficient market coordination.Even price agreements between firms may serve to reducerisk and uncertainty-during a recession, for example-andlead to an increase in market efficiency. (See chapter 6.)Cooperation and rivalry are voluntary alternative institu-tional arrangements by which entrepreneurs, under condi-tions of uncertainty, strive to discover


The standard textbook treatment often assumes amonopoly output restriction and then proceeds to com-pare that restricted output, unfavorably, with an atomisticequilibrium output level.3 But both the assumption and thecomparison are entirely misleading, for the atomistic equi-librium output level is neither possible nor relevant andcannot serve as the welfare benchmark for any compari-son. Moreover, it is difficult to understand how any outputlevel that is inefficient or generates substantial profits canbe sustained in an open market in the face of strong incen-tives to expand production.

Free-market monopoly power created through mergeror collusion is presumably the primary concern of theantitrust authorities. But if the economic effect of monop-olization is to raise prices above costs-marginal and aver-age-strong economic incentives then exist to expand cur-rent production and to encourage output by new firms. Ifproduction increases, prices will fall and the market willtend, other things being equal, toward a situation in whichprices and costs are equal.

What happens if a free-market monopolist attempts tosubvert this competitive process and discourage rivalrousentry by lowering prices? The reduced prices would induceadditional sales, and the market situation would then tendtoward the traditional competitive equilibrium. What hap-pens if a monopolist discriminates in price? Indeed, theremight be strong economic incentives to do so, but amonopolist that price discriminates will end up selling addi-tional output at some lower price, and, again, the market willtend toward the traditional competitive output. Certainly amonopolist that is inefficient cannot deter market entry;

3See, for instance, Edwin Mansfield, Microeconomics: Theory and Applications,5th ed. (New York: W.W. Norton, 1985), p. 294. The entire notion of a free-marketmonopoly price and output may be untenable. See Murray N. Rothbard, Man,Economy, and State (Princeton, N.J.: D. Van Nostrand, 1962), Vol. 2, pp. 586-615.Also see the Appendix in this chapter for an explanation of Rothbard's monopolytheory.

36


inefficiency will act as an invitation to entry and additionaloutput. On the other hand, a monopolist that is clearlymore efficient than potential rivals can deter entry, but itwould be the efficiency of the monopolist that would keepcompetitors out. Resources are not misallocated and thecompetitive process is not subverted when high-cost firmsare restrained from entering markets by the superior prod-uct or efficiency of existing suppliers.

Firms may intend to restrict market output through col-lusion and cartel agreements, but the realization would beeven more tenuous than that possible through a one-firmmonopoly. Not only would a cartel of suppliers encounterthe same incentives to expand production reviewed above,it would also face such difficulties as coordinating andpolicing its own supply-restriction schemes.4 Interfirmagreements to restrict rivalry could exist in a free market, asthey did occasionally under common law prior to theSherman Act, and they might even be able to stabilize tem-porarily some price fluctuations, but there is little reliableevidence that free-market collusion can allow conspiringfirms to capture monopoly profits.5 Moreover, interfirmcooperation may well have significant benefits that couldoverwhelm any possible negative output restriction. (Seediscussion in chapter 6.)

Likewise, the usual textbook discussions of inefficiencyunder monopoly are unconvincing. The standard argumentof allocative inefficiency is, in fact, contrived and mislead-ing. With new entry and output blocked by definition, a

4The difficulties of effective collusion are reviewed in Dominick T. Armentano,Antitrust and Monopoly: Anatomy of a Policy Failure, 2nd ed. (Oakland, Calif.:Independent Institute, 1990), pp. 133-37. See also George j. Stigler, "A Theory ofOligopoly," Journal of Political Economy 72, no. 11 (February 1964): 44-61.

5A negative relationship between collusion and profitability is found by PeterAsch and joseph j. Seneca in "Is Collusion Profitable?" Review of Economics andStatistics 58 (February 1976): 1-12. See also Howard Marvel, Jeffrey Netter, andAnthony Robinson, "Price Fixing and Civil Damages: An Economic Analysis," StanfordLaw Review 40 (1988): 561-78.

37


monopolist is said to misallocate economic resources rela-tive to their allocation under conditions of pure competi-tion. But this "misallocation" occurs only because the com-petitive process is assumed to be ended in atomistic com-petition (price, marginal cost, and minimum average costare all assumed to be equal) and because no competitivemarket process is allowed to begin under monopoly. If, o'nthe other hand, a competitive process always operatesunder free-market monopoly, and if it is assumed that nofinal atomistic equilibrium condition can ever exist, thenresource misallocation under free-market monopoly, assome unique social problem, simply disappears. Allocativeinefficiency would tend to disappear from the free-marketmono

Date post:	27-Jan-2021
Category:	Documents
Upload:	others
View:	3 times
Download:	0 times

Antitrust: The Case for Repeal - Mises Institute The Case for Repeal_1_0.pdfAntitrust: The Case for...

Documents