Government Failure Versus Market Failure C€¦ · Government failure, then, arises when government...

57First Quarter 2007

b o o k e x c e r p t

CCliff Winston, a senior fellow at the

Brookings Institution in Washington, is

best known for his analyses of the airline

and trucking industries – more particularly, for his fi nding that transportation deregu-

lation has generated humungous effi ciency

gains. But a closer look at his work reveals a

broader theme: even where regulation can be

justifi ed in principle, the practical consequences

have ranged from less than optimal (think envi-

ronment) to wretched (think international

trade). Read this excerpt from his new book,

Government Failure Versus Market Failure,* for a

devastating critique of the government’s role in our market economy. — Peter Passell

Government Failure Versus

Market Failure

b y c l i f f w i n s t o n

*Published with permission from the Brookings Institution Press. All rights reserved.k-p

ho

to

s/a

la

my

58 The Milken Institute Review

Shortly after he took offi ce, President George W. Bush nominated Harvard profes-sor John Graham to head the Offi ce of In-formation and Regulatory Affairs within the White House’s Offi ce of Management and Budget. Graham was known to be a strong advocate of using cost-benefi t analysis to as-sess environmental, health and safety regula-tions. If, for example, the Environmental Pro-tection Agency proposed a regulation that saved 100 lives, but at a cost of $1 billion per life, Graham would oppose the regulation and encourage the EPA to craft an alternative rule that could save these lives at a (much lower) cost in line with conventional estimates of the statistical “value of life.” Or if the Nation-al Highway and Traffi c Safety Administration proposed a regulation that forced automakers to adopt a specifi c technology to reduce fuel consumption, but the resulting benefi ts were less than the costs of implementing the tech-nology, Graham would oppose the regulation on the grounds that its net benefi ts to society were negative.

To an economist, these positions are emi-nently reasonable. But many commentators and policymakers fl atly dismiss assessments based on cost-benefi t analysis, apparently willing to substitute good intentions – or their own political agendas – for analysis. And this inclination to refuse to acknowledge that government intervention can have costs as well as benefi ts raises a fundamental concern: is government policy truly enhancing the de-gree to which our economic system meets the material wants of Americans?

In theory, government seeks to improve the effi ciency of markets by correcting their

failure to equate prices with costs. The possi-ble causes of such market failure range from monopoly power, to a lack of information available to buyers, to the existence of costs or benefi ts that affect parties not directly in-volved in transactions.

On what basis is one to conclude that a policy to correct a market failure is success-ful? The fi rst consideration is whether gov-ernment has good reason to intervene in the fi rst place: Is there evidence of a serious mar-ket failure to correct? The second is wheth-er government is actually improving market performance: Is it reducing the “deadweight” loss – that is, the economic value lost to all parties – associated with market failure? Of course, intervention can be an expensive suc-cess by generating benefi ts that exceed costs, but nonetheless incurring excessively high costs to obtain the benefi ts.

Government failure, then, arises when government has created ineffi ciencies because it should not have intervened in the fi rst place, or when intervention could have gener-ated greater net benefi ts. Of course, the ideal of a truly effi cient market (before or after in-tervention) is rarely, if ever, observed. From a policy perspective, market failure should be a matter of concern when performance signifi cantly deviates from the reasonable benchmark. Similarly, government interven-tion should be questioned when economic welfare is actually reduced, or when valuable resources are plainly wasted in the pursuit of greater effi ciency.

Economic theory can suggest policies to correct market failures, but the effect of gov-ernment on economic welfare can only be

The aim of science is not to open the door to

infi nite wisdom, but to set a limit to infi nite error. – Bertolt Brecht


assessed with empirical evidence. For more than a century, the primary market-failure policies have included antitrust policy and economic regulation to curb market power, policies to address imperfect information and external costs and benefi ts, and public fi nanc-ing of socially desirable services that the pri-vate sector has insuffi cient incentive or ability to provide.

Initially, economists assessed these pol-icies on conceptual grounds, culminating in Milton Friedman’s classic attack in the 1960s questioning government’s role in al-most all areas of economic life. Charles Schul-tze (President Carter’s chief economic advis-er) was one of the fi rst to systematically raise doubts about the effi cacy of government pol-icies where the doubts were based on empir-ical evidence. Charles Wolf (the former dean of the RAND Graduate School) introduced the term “non-market” failure as an alterna-tive to market failure, and suggested that such failure may be of the same order of impor-tance as market failure.

An additional 30 years’ worth of empirical evidence suggests that the cost of government failure may actually be considerably greater than that of market failure. More specifi cal-ly, the evidence leads to the conclusion that some policy interventions have exacted costs where no serious market failure existed, while others produced far less benefi t than could have been expected.

Government failures appear to be ex-plained by (a) the self-correcting nature of markets in some cases, which makes interven-tion unnecessary, (b) the shortsightedness, infl exibility and confl icting agendas of regu-lators, and (c) political forces that allow in-terest groups to hijack policymaking in quest of economic advantage (what economists call “economic rent”).

My negative assessment is not intended to suggest that all market intervention is coun-terproductive, but to focus attention on how

30 years’ worth of empirical evidence suggests

that the cost of government failure may actually be

considerably greater than

that of market failure.

©g

ab

e p

alm

er

/c

or

bis


I

current policy could be improved. This is not a futile exercise; in the past few decades, gov-ernment has become somewhat less inclined to pursue ineffi cient policies and has initi-ated some benefi cial reforms. For example, policymakers are less likely to try to correct a perceived market imperfection by regulat-ing prices – think natural gas and petroleum. By the same token, in some cases policymak-ers have enhanced economic welfare by de-regulating in order to allow markets to work, or by using market-based incentives to cor-rect market failure with minimal intervention – for example, by encouraging the trading of rights to pollute in order to minimize the cost of meeting pollution targets. Further applica-tions of these approaches are likely to reveal that they could be used far more broadly to good effect.

Although researchers have identifi ed seri-ous fl aws in other market failure policies such as antitrust and patents, our empirical know-ledge is too limited to permit confi dent ad-vice about how policy could be signifi cantly improved. Thus, additional research is clearly

needed to help guide the formulation of ap-propriate policy in these areas.

Over the past few decades, the econom-ics profession has begun to understand which policies have been successful and which have not. Equally important, they have begun to analyze why policymakers often fail to pursue socially desirable change. The gap between what economists already know and what gov-ernment chooses to ignore should not dis-courage the profession’s efforts. By docu-menting government performance and indi-cating how it can be improved, I hope we can do more than set a “limit to infi nite error.”

• • • • •

In the textbook model of perfect competition, fi rms earn a “normal” market rate of return in the long run. Of course, some fi rms may have superior technologies and management, which enable them to earn above-normal re-turns for long periods. But when a fi rm at-tempts to raise prices above costs by abusing its market power, antitrust can improve con-sumer welfare by stopping them.

Does the United States, in fact, face considerable threats to competition?

ke

vin

ph

illip

s/g

etty

im

ag

es


A market’s technological characteristics may give rise to a “natural” monopoly, an un-usual situation in which costs are minimized when a single fi rm serves the entire market. Here, competition leads to a monopoly survi-vor, and government might arguably improve economic welfare by regulating the prices charged by the monopolist.

It’s worth emphasizing, though, that the theoretical justifi cation for antitrust laws and economic regulation does not imply that gov-ernment intervention is always needed to curb market power and improve economic effi ciency – or that it has done so in practice. Nor is intervention warranted simply because one can cite examples of fi rms that have col-luded or can identify markets that are prone to concentration because of economies of scale. Policies to curb market power can be justifi ed only by evidence that they can and do increase output, and thereby raise social welfare. However, there is precious little evi-dence that such policies have worked accord-ing to the manual.

antitrust policyU.S. antitrust policy is the responsibility of the Department of Justice and the Feder-al Trade Commission. Justice enforces Sec-tion 1 of the Sherman Antitrust Act of 1890, which prohibits contracts, combinations and conspiracies in restraint of trade, as well as Section 2 of the Sherman Act, which bars at-tempts to monopolize markets through anti-competitive means. And, along with the FTC, Justice enforces Section 7 of the Clayton Act of 1914 (amended in 1950), which prohibits mergers between fi rms that threaten to lessen competition.

Note that the Clayton Act also prohibits fi rms from engaging in “tying” arrangements (requiring consumers to buy two products in order to get just one of them) and bars com-

peting fi rms from having overlapping boards of directors. The FTC may also initiate cases under Section 5 of the Federal Trade Com-mission Act for “unfair methods of competi-tion,” thereby providing it with the ability to combat abuses that the Justice Department attacks under Sections 1 and 2 of the Sher-man Act.

Does the United States, in fact, face con-siderable threats to competition? It is diffi -cult to provide a single measure of how com-petitive U.S. industries are, but the available evidence does not suggest that the nation’s economy is suffering from any serious un-derlying problem in this regard. For example, the late Peter Pashigian of the University of Chicago followed common practice in defi n-ing an imperfectly competitive market as one in which four fi rms had a combined market share above 70 percent. He found that in 1992 only 46 out of 398 manufacturing industries met that criterion. And preliminary evidence from 2002 actually shows a decline in concen-tration since then. Manufacturing, it should be noted, accounts for just 16 percent of GDP. However, concentration data are not readily available for other sectors.

In addition, the theme that domestic mar-kets are highly competitive is consistent with the common fi nding that the U.S. economy has experienced only small deadweight losses from non-competitive pricing. UCLA econ-omist Arnold Harberger’s famous 1954 esti-mate of the deadweight loss due to monop-oly at a minuscule 0.1 percent of GDP may be low. But more recent estimates summa-rized by Paul Ferguson of Lancaster Univer-sity in 1988 still suggest a fi gure of just 1 per-cent of GDP.

The relevant issue, therefore, is whether the apparent absence of serious anticompet-itive problems in the United States is really the result of more than a century of antitrust


policy directed at monopolization, collusion and mergers.

MonopolizationIf the antitrust policy made a difference, one would expect consumers to benefi t when the government prevailed in a monopolization case and the courts provided competitive relief such as divestiture. To test the propo-sition, my colleague Robert Crandall and I collected evidence on landmark cases where relief was indeed provided, including Stan-dard Oil (1911), American Tobacco (1911), Alcoa (1945), Paramount (1948) and United Shoe Machinery (1954). In each case we found that the court’s relief failed to increase competition or reduce consumer prices. Moreover, recent antitrust enforcement of monopolization, including cases against IBM, Safeway, A&P and Blue-Chip Stamps, also failed the test.

A possible exception to these fi ndings is the 1984 breakup of AT&T, which followed a 1974 monopolization case. But the key provi-sion of the decree that gave rise to the growth in long-distance telephone competition and lower rates – namely, Bell companies were re-quired to modify their network facilities to provide equal access to all long-distance car-riers – could have been imposed by the Fed-eral Communications Commission without the intervention of the Justice Department. Thus, antitrust policy was not necessary to re-strain a monopolist; rather it was necessary to change the anticompetitive policies of anoth-er federal regulatory agency.

Indeed, if the FCC had acted appropriate-ly, the large costs of breaking up AT&T could have been avoided. These costs include re-structuring AT&T to consummate the break-up, legal enforcement costs and the loss of effi ciencies in the vertical integration of local and long-distance service. In subsequent

decades, the telecommunications industry has evolved into a competitive free-for-all among three vertically integrated technolo-gies: fi xed wire-telephone companies (exclud-ing AT&T!), cable television companies and wireless carriers.

Why has antitrust been such a bust? One reason is that the protracted length of mo-nopolization cases (some of the ones noted earlier took more than a decade to resolve) meant that antitrust actions are likely to lag far behind market developments, and thus be less effective than market incentives in spur-ring competitive change. Another is that, not-withstanding a court’s intentions, the relief obtained by the government may simply have a negligible practical impact on consumers.

Take, for example, the recent Microsoft case. The merits of the case are not yet clear, but its limitations are consistent with the problems of other monopolization cases. First, it has turned out to be a lengthy case. The federal suit was fi led in May 1998 and some private actions are still unsettled. Mi-crosoft also faces an antitrust challenge in Europe. By the time the case is fully resolved, the information technology market is likely to have changed substantially, and the pre-sumed need to curb Microsoft’s market dom-inance may be gone. Challenges to a modest, court-directed remedy were rejected in 2004. But the remedy has left both sides question-ing the point of the prosecution.

CollusionEconomists have yet to fi nd that antitrust prosecution of collusion in restraint of trade has led to signifi cantly lower consumer pric-es. Michael Sproul of California State Uni-versity (Northridge) analyzed a sample of 25 price-fi xing cases between 1973 and 1984. He argued that if cartels had raised prices above competitive levels, prosecution should have


lowered them. Controlling for other infl u-ences, however, he found that prices had risen an average of 7 percent four years after an in-dictment. Sproul also found that prices rose, on average, even if one used a starting point during the investigation but before the date of the indictment.

Retrospective assessments of specifi c price-fi xing prosecutions have also found that consumers did not benefi t from the ac-tions taken. Among them: a price-fi xing in-dictment against bakers analyzed by Craig Newmark of North Carolina State Universi-ty, and a consent decree that prohibited air-lines from announcing the end dates of their

fare promotions analyzed by Steven Morrison (Northeastern University) and me.

In 2001, Sotheby’s and Christie’s settled a price-fi xing suit alleging that in 1995 the two auction houses began conspiring to elevate their commissions. Apparently, the settlement had little effect on consumers because Sothe-by’s 2003 revenues (mainly from commis-sions) as a percentage of its auction revenues actually exceeded its 1993 revenues as a per-centage of auction sales. To be sure, it is pos-sible to identify some successful price-fi xing prosecutions. But no serious academic study has shown that these cases have led to signifi -cantly lower prices for a protracted period.

Retrospective assessments of specific

price-fixing prosecutions have also found

that consumers did not benefit

from the actions taken.

ap

ix/a

la

my


Justice Department enforcement may be ineffective because it primarily prosecutes fi rms engaged in activities that involve other goals besides raising prices. For example, Sproul suggests that a cartel may reduce costs through shared advertising and re-search, which may reduce prices, not increase them. Another possibility is that a cartel may be pursuing equity goals.

For example, in 1991 the Justice Depart-ment brought suit against MIT and the Ivy League schools, which had been coordinat-ing their scholarship decisions. The schools claimed that this “overlap process” enabled them to concentrate scarce fi nancial aid on needy students by making it possible for them to agree to not use scholarships to compete for students on the basis of merit. The gov-ernment, for its part, claimed that the schools were conspiring to reduce total aid and there-by to raise net revenues.

The evidence suggests the government erred. Dennis Carlton (University of Chica-go), Gustavo Bamberger (Lexecon Consul-tants) and Roy Epstein (unaffi liated) found that the overlap process did not have a statis-tically signifi cant effect on total student aid, but did result in a larger share being award-ed to low-income students. Thus, from most peoples’ perspective, the challenge to the overlap process was counterproductive.

MergersOn their face, mergers may harm or bene-fi t consumers. Those that enable fi rms to ac-

quire market power result in higher pric-es, while those that generate operational and managerial effi ciencies reduce costs and thereby make it possible to lower prices. In some markets, a stronger merged entity may also stimulate competition and reduce mark-ups. The effectiveness of merger policy there-fore depends on how well the antitrust au-thorities distinguish pro-competitive mergers

from anticompetitive ones.Dartmouth economist B. Espen Eckbo’s

analysis of mergers that were challenged by the Justice Department or FTC suggests they were not generally anticompetitive, and would have generated effi ciency gains had they been allowed to go through. My own research with Robert Crandall of Brookings confi rms this fi nding: antitrust authorities have primari-ly attacked mergers that would enhance effi -ciency, either by blocking them in court or by allowing them to proceed only if the merg-er partners agreed to conditions that raised their costs.

DeterrenceThe strongest argument in support of anti-trust policy is that it deters fi rms from engag-ing in illegal practices that would ultimate-ly raise prices. Unfortunately, the benefi cial effects of deterrence – the price-fi xing ar-rangement that never takes place, the merg-er to monopoly that is not consummated, the predatory strategy that is not attempted – are diffi cult to observe.

International comparisons have therefore

Antitrust authorities have primarily attacked

mergers that would enhance efficiency, either by

blocking them in court or by allowing them to


been used to shed some light on the deter-rence issue. The late George Stigler of the University of Chicago compared market con-centration in specifi c industries in the United States with the same industries in England, which at the time of his study (1966) did not have a policy against concentration. He con-cluded that the Sherman Act has had a very modest effect in reducing U.S. concentration. A quarter century later Eckbo explored the question by estimating whether the probabil-ity that a horizontal merger was anticompeti-tive was higher in Canada, where (until 1985)

mergers were effectively unconstrained. He found no evidence the U.S. antitrust laws were deterring anticompetitive mergers.

Firms and individuals convicted of price-fi xing are subject to criminal penalties and also vulnerable to private suits for treble damages. Recently, the Department of Jus-tice has attempted to strengthen deterrence

by imposing higher fi nes on corporations for price-fi xing and offering lenient settlement terms for fi rms that disclose their roles in conspiracies and cooperate with the govern-ment. These policies may be thought to bene-fi t consumers. But Bruce Kobayashi of George Mason University cautions that they may ac-tually lead to over-deterrence, which would induce excessive investments in monitoring and prevention, raise production costs and result in higher consumer prices.

In any event, some fi rms apparently have not been deterred from colluding because the

Justice Department continues to bring price-fi xing cases. While it is possible that Justice has succeeded in deterring the most serious instances of price-fi xing and has therefore been increasingly prosecuting marginal cases, we can’t document the claim.

A Credibility CheckReaders may question whether the available evidence leads to uniformly critical conclu-sions of market failure policies in general and of antitrust policy in particular. I maintain that it does, but acknowledge that my assess-ment is not shared by all who have written on the topic. It is therefore useful to consider how those who believe that antitrust policy has had positive effects support their position empirically.

In response to my work with Robert Cran-dall, Jonathan Baker, the former director of the FTC’s Bureau of Economics, offered a cost-benefi t analysis of antitrust enforcement as support for the desirability of the activity. He argued that the annual costs of enforce-ment were small (roughly $1 billion) and

proceed only if the merger partners agreed to

conditions that raised their costs.

©c

la

ve

r c

ar

ro

ll/a

ge

fo

to

sto

ck


tk

that the potential benefi ts were likely to be as much as $100 billion a year. But Baker did not obtain his estimate of benefi ts from empirical assessments of the effects of actual antitrust cases. Instead, he claimed that they could be large because price-fi xing conspiracies were quite costly to consumers – for example, col-lusion among vitamin producers cost con-sumers at least $100 million.

However, in the absence of hard evidence that antitrust actions have actually reduced the costs of collusion to consumers or bene-fi ted them in other ways, it is inappropriate to claim that the consumer gains from such actions exceed enforcement costs – especially because these costs are likely to approach several billion dollars annually when one accounts for lawyers’ fees, payments to con-sultants, opportunity costs of managers and employees who participate in a fi rm’s defense and so on.

Mikhail Kouliavtsev of Philadelphia Uni-versity defends antitrust policy on similar grounds, citing evidence that antitrust policy had deterred mergers between 1959 and 1972 that would have been harmful to consumers. But this evidence was based on the assump-

tion that any increase in concentration in a product market increases profi t margins. De-spite this rather heroic assumption, the 59 horizontal merger cases that provided the basis for Kouliavtsev’s favorable assessment generated gains in market effi ciency that were smaller than the enforcement costs created by the action. Thus, it is diffi cult to maintain a favorable assessment of antitrust on cost-benefi t grounds.

economic regulationBeginning with the 1887 Interstate Com-merce Act that regulated the railroads, the federal government has controlled pricing, entry and exit in industries where compe-tition allegedly was not workable because scale economies would lead fi rms to under-cut one another’s prices until they were either all bankrupt or the industry was monopo-lized. During the 1970s the federal govern-ment began to deregulate large parts of the transportation, communications, energy and fi nancial industries because it became clear that economic regulation was, in fact, imped-ing competition that could benefi t consum-ers. Today, federal price regulations are largely

67

tk

confi ned to agricultural commodities and in-ternational trade of selected products – nei-ther of which is believed to invoke natural monopoly considerations.

AgricultureEconomic regulation of agriculture amounts to an array of subsidies, including subsidies for being a farmer, subsidies to cover loss-es on what is grown and subsidies for taking land out of production. In 2000, government assistance constituted more than 100 per-cent of net farm income in eight states. Sub-sidies mainly go to big corporations and the richest farmers. The Environmental Working Group, a Washington-based nonprofi t advo-cacy group, reported that in 2003 the top 6 percent of recipients collected 55 percent of all subsidies. Although subsidies used to be inversely related to farm prices (that is, sub-sidies increased when farm prices fell), that relationship has not held in recent years. For example, according to the U.S. Department of Agriculture, farm earnings in 2004 reached a record $74 billion, while direct government payments were as high as they were in years

when farmers earned a fraction of that sum.Historically, government intervention in

agriculture was motivated by the “farm prob-lem” – that is, the low earnings of most farm-ers and the great instability of income from farming. But Bruce Gardner of the Univer-sity of Maryland found that farm and non-farm incomes converged in the late 1960s, and average farm incomes actually exceeded non-farm incomes by the 1980s. Farm and non-farm labor markets have become much more integrated, with some family members living on farms and commuting to non-farm jobs. On average, farmers have also become wealthier than non-farmers.

The offi cial rationale for farm regulation is therefore a chimera. Yet governmental assis-tance to agriculture continues – and has even grown – in recent years. Price support pro-grams for some agricultural commodities were initially created under the 1933 Agricul-tural Adjustment Act, which has been revised every few years since its inception. Currently, eligible commodities include wheat, corn, cotton, peanuts, dairy products, sugar and rice. (Longstanding tobacco price supports

Farm earnings in 2004 reached a record $74 billion, while direct government payments were

as high as they were in years when farmers

earned a fraction of that sum.


ended as of the 2004 crop year, with full com-pensation paid to land owners.) The agricul-tural sector also idles much of its land as part of the Department of Agriculture’s acreage reduction authority.

Prices for milk are set by the federal gov-ernment in accordance with the Milk Mar-keting Order of 1937. This program effective-ly raises the price of fl uid milk and lowers the price of milk going into dairy products.

Government regulation (subsidization) of these markets cannot be justifi ed on econom-ic grounds. According to a variety of studies, commodity price support programs transfer income from consumers to producers, and in the process generate annual net deadweight losses between $3 billion and $12.4 billion. A study by Peter Helmberger of the University of Wisconsin estimates that the dairy support program generates deadweight losses in the $1 billion-a-year range.

Amendments to the farm bill in 1996 were designed to scale back the costs of the pro-grams (including their restrictions on plant-ing) and to wean farmers from taxpayer sup-port. But Congress yielded to pleas from farm lobbyists and provided annual “emergency” loans that raised the program’s costs. In May 2003, President Bush signed a 10-year $190 billion farm subsidy bill that, among other things, increased payments to farmers who produce wheat and corn and introduced an expensive new milk price support program.

International TradeInstead of garnering direct subsidies from the U.S. Treasury, some industries make use of indirect federal support in the form of import quotas and tariffs that restrict the availabil-ity of foreign competitors’ automobiles and light trucks, steel, textiles and apparel, chemi-cals, dairy products and sugar. In theory, pro-tection for an “infant industry” – an industry whose costs are likely to fall sharply (and to competitive levels) if it is given time to gain a toehold in the market – could be justifi ed for a limited period. Tariffs could also be jus-tifi ed if they increased national welfare at the expense of foreign fi rms – for example, by requiring such large economies of scale that foreign fi rms couldn’t compete.

However, the evidence points to a very dif-ferent outcome: trade protection has mainly generated gains to established U.S. industries at far greater expense to consumers. Draw-ing on several studies, Robert Feenstra of the University of California (Davis) puts the an-nual net cost to the United States of tariffs and quotas between $12 billion and $18 bil-lion. To add insult to injury, import prices are raised by the 1920 Jones Act, which increases the cost of transporting all goods by requiring that shipping within the United States be on U.S.-built and U.S.-owned vessels.

International airline service is, for the most part, regulated through bilateral agree-ments. My own research with Steven Mor-rison suggests that fares between the United States and other countries would, on average, fall by one-quarter if they were determined in a market environment that paralleled deregu-lated competition on U.S. domestic routes.

Notwithstanding these costs, the United States and its trading partners have gradually reduced trade restrictions outside of agricul-ture by linking policy to multilateral agree-ments. But political considerations can still

Trade protection has mainly generated gains

to established U.S.

industries at far greater

expense to consumers.

pr

ev

iou

s p

ag

e:

ric

ha

rd

os

bo

ur

ne

/b

lu

e p

ea

rl p

ho

to

gr

ap

hic

/a

la

my;

op

po

sit

e p

ag

e:

©r

ob

cr

an

da

ll/th

e i

ma

ge

wo

rk

s


tk


lead to backsliding. Upon taking offi ce, Presi-dent Bush expressed a desire to end all tariffs on consumer goods. He then chose to sup-port new tariffs of 8-30 percent on a broad range of steel products imported from Japan, South Korea, Taiwan, China, Russia and Eu-rope. Bush subsequently lifted the tariffs, but indicated that he might impose them again.

Quotas on textiles and apparels ended worldwide in 2004. However, the Bush admin-istration reacted to the fl ood of Chinese cloth-ing imports in the early months of open trade by threatening new quotas on cotton shirts, trousers and underwear from China. Faced with that threat, China “voluntarily” limited clothing exports for three more years.

the big picture Government policy to correct market fail-

ures is characterized by two major fl aws that cost the U.S. economy hundreds of billions of dollars annually. First, intervention has cre-ated economic ineffi ciencies where signifi -cant market failures do not appear to exist; antitrust laws and economic regulations have raised fi rms’ costs and generated economic windfalls for various interests at the expense of consumers. To be sure, antitrust enforce-ment may be deterring anticompetitive be-havior – but this is only a surmise. A possi-ble benefi t of intervention in some markets is that the public has escaped exposure to some dangerous products, such as harmful drugs. But once again, there is little evidence that the benefi ts of such regulation have exceed-ed the costs.

Second, in situations where market fail-ure does occur, government policy has ei-

It appears

that in at least

some instances,

buyers and sellers have greater incentives to

correct market failures than does the government.

ju

stin

le

igh

to

n/a

la

my


ther achieved expensive successes by correct-ing these failures in a less-than-optimal way – for example, by making the environment safer through extremely costly regulation of pollutants. And, in some cases it has actually made bad situations worse by imposing rem-edies that are more costly than the problems they solve.

What explains the prevalence of govern-ment failure when policymakers are pre-sumably trying to correct market failures? In some cases, intervention may be counter-productive because the policies are fl awed or poorly implemented – think of price controls in the face of shortages – or because policy-makers see their job as serving one interest group at the expense of the public.

Charles Wolf of RAND argues that the very existence of government failure suggests offi cials lack appropriate incentives to recon-cile the costs and benefi ts of intervention with its societal costs and benefi ts. In con-trast, at least in some instances, buyers and sellers have greater incentives to correct mar-ket failures than does the government.

Market RobustnessMarket failure is less common and less costly than might be expected because fl awed mar-kets tend to repair themselves. For example, the lure of easy profi t typically draws new entry to markets that are less than fully com-petitive, and competition often develops in markets with entry barriers such as econo-mies of scale or bottlenecks created by pro-tected intellectual property like patents. Eco-nomic deregulation suggests that experience is more instructive than theory on the ques-tion of whether effective competition will de-velop in particular markets.

Market failure from imperfect information is limited because consumers are able to make informed choices by drawing on professional

product reviews and word of mouth. For ex-ample, my own research with Fred Manner-ing of Purdue University found that consum-ers’ adoption of airbag-equipped automobiles during the 1990s was spurred by their friends’ experiences with airbags and media reports about experiences that other motorists had with airbags. By the same token, my research with John Calfee and Randolph Stempski (both from the American Enterprise Insti-tute) found that the increasing demand for cholesterol-reducing drugs could in part be attributed to the spread of word-of-mouth news about successful treatments.

The advent of the Internet has given con-sumers another way to become informed about the quality of products and services and to troll multiple sellers for lower prices. Judith Chevalier and Dinah Mayzlin of Yale con-fi rmed that consumers pay attention to book reviews that are available at Amazon.com and Barnesandnoble.com. Web-based airline fares have made the lowest fares available to almost every traveler. Moreover, research by Florian Zettelmeyer (UC-Berkeley), Fiona Scott Morton (Yale) and Jorge Silva-Risso (UC-Riverside) concluded that, collectively, consumers using the Internet to purchase cars are saving some $200 million a year because they are better informed when they negotiate with salespeople.

When not impeded by regulation, adver-tising can enable fi rms to overcome the inef-fi ciencies created by the lack of property rights in most information. Among the posi-tive externalities associated with advertising are better information about diet and health, opportunities to improve health through drug therapy and the dangers of smoking.

Firms that produce faulty products or pro-vide unsatisfactory services damage their rep-utations and incur fi nancial costs. Ginger Jin (University of Maryland), Andrew Kato


(University of Maryland) and John List (Uni-versity of Chicago) point out that “certifi ca-tion markets” can provide information to re-duce this possibility. For example, Underwrit-ers Laboratories certifi es the reliability of consumer and industrial products, Moody’s and Standard & Poor’s report bond ratings to inform investors about risk, and accounting companies audit fi nancial reports for public corporations in an effort to deter and expose fraud.

Firms have also instituted compensation policies to rectify cases where their products do not meet expectations. For example, when airlines oversell fl ights, they typically com-pensate travelers who voluntarily agree to be bumped more than they compensate travelers who are involuntarily bumped and are con-sequently eligible for compensation under government regulations. The industry policy evolved in response to a government rule that increased penalties for involuntary bumping. The volunteer system increased effi ciency by delaying people who care least about wait-ing for the next departure, while still enabling carriers to overbook at a higher rate and fl y with fewer empty seats.

Finally, fi rms have strong internal incen-tives to maintain safe workplaces. Employ-ers with work environments that pose in-herent risks must pay higher wages to attract workers. In general, fi rms also benefi t from safe workplaces because production costs are lower, employees’ productivity is higher, workers’ compensation insurance premiums are reduced and fi rms are more likely to avoid costly litigation.

Market forces can even help lower some external costs. People’s choices of where to live and work refl ect sorting to reduce the costs of congestion, airplane noise, air pollu-tion, street crime and the like. My work with John Calfee concluded that workers with long commutes valued their time less than workers with shorter commutes, implying that com-muters who dislike congestion the most re-duce its cost by living closer to their work-places. By the same token, my research with

Steven Morrison and Tara Watson (Williams College) concluded that people who have a high tolerance for noise tend to live closer to airport fl ight paths and require less compen-sation to be “made whole” – as refl ected in lower housing prices – than people who have a low tolerance for noise.

Other market forces besides households’ self-sorting help remedy externalities. Fish farms, for example, are a market response to declining stocks in public fi sheries. In ad-dition, cooperation in public fi sheries has helped to raise the catch in the past decade. For example, James Acheson of the University of Maine notes that lobstermen have adopt-ed the practice of cutting a notch in the tail of a female lobster carrying masses of roe, so other lobstermen would know that these lob-sters are to be left alone as breeding stock.

Some recent work even suggests that mar-kets may tend to correct multiple sources of failure. For example, Scott Taylor (Universi-ty of Calgary) along with Werner Antweiler and Brian Copeland of the University of Brit-ish Columbia found that freer trade enhanced product market competition and increased

Some policies may have produced benefits

when they were initially implemented,


GDP, which led to public demands for less polluting production methods. Specifi cally, the authors found that sulfur dioxide emis-sions fell by roughly the same amount in per-centage terms that national income rose from more open international markets.

Government Agency ShortcomingsThe policy problems that government agen-cies face today are generally more challenging than the problems of earlier eras. And often the agencies aren’t up to the task, contribut-ing to policy failures by their shortsighted-

ness, infl exibility and confl icts of interest. Some policies may have produced benefi ts

when they were initially implemented, but policymakers have lacked the vision or will to modify them when they lose value. For exam-ple, agricultural price supports are essential-ly New Deal programs developed in the most tumultuous period in modern American eco-nomic history. Current agriculture policy should refl ect the economic realities of the times and eliminate efforts to redistribute in-come through subsidies and price regulation.

While there is widespread agreement that strong patent laws in the 19th century spurred economic innovation and growth, a decline in the rate of patenting (adjusted for popula-tion growth) began at the end of that century even though the country continued to inno-vate and grow. It is possible that patents be-came less essential for fostering innovation as greater competition caused fi rms to increase their proprietary innovative activities.

Antitrust offers another example of short-sighted policy. After taking several years to re-solve a monopolization case, the Justice De-

partment and the FTC often fi nd that market conditions have changed. But they are unable to adjust to the presence of new sources of competition that have undermined the ratio-nale for bringing a case in the fi rst place. In-deed, my own research with Robert Crandall suggests the antitrust authorities have found it diffi cult to formulate consent decrees in monopoly and merger cases that benefi t con-sumers in the long run.

Even a casual review of the research shows that the list of agencies that have failed to ad-just their policies appropriately is a long one.

• The Justice Department and the FTC have litigated price-fi xing and mergers that are not harmful to consumers.

• Standards set by the Occupational Safety and Health Administration focus on machin-ery safety, ignoring the critical interactions between labor, equipment and the work-place environment.

• The Mine Safety and Health Administration has compromised the potential benefi ts from regulating coal mine safety by misal-locating funds.

• The Consumer Product Safety Commission’s priority rankings are at variance with cost-benefi t ratios.

• Both the Transportation Department and the Food and Drug Administration have failed to anticipate consumers’ offsetting behavior that has undermined the effi cacy of their policies.

• The Environmental Protection Agency’s command-and-control regulation pays insuffi cient attention to costs, and is often based on implausibly high values of lives saved.

but policymakers have lacked the vision or will

to modify them when they lose value.


• Most agencies that address external costs either eschew prices as a mechanism for enhancing effi ciency or, as in the case of fuel, cigarette and alcohol taxes, set them with little regard to estimates of the societal costs. Still another source of ineffi ciency is agen-

cy policy that confl icts with either its own ob-jectives or those of another agency. For exam-ple, the Securities and Exchange Commission is responsible for preventing companies from misleading the public about their fi nancial conditions. But as Lawrence White of New York University notes, the SEC has enacted entry regulations that have stifl ed competi-tion among independent credit-rating agen-cies that could provide investors with valuable assessments of companies’ creditworthiness. Thus, regulation has reduced the effectiveness of a private certifi cation market by thwarting competition.

The Transportation Department seeks to encourage fuel economy and promote safety. But the EPA’s emissions regulations increase vehicle weight and reduce fuel economy, while loopholes in the Transportation De-partment’s fuel-economy standards increase the share of relatively wasteful and dangerous light trucks on the road.

The U.S. Forest Service is responsible for preserving public lands in their natural state, implying that the agency should allow fi res to consume mature wilderness because fi re is needed to regenerate many wild species. However, as Robert H. Nelson of the Univer-sity of Maryland notes, the service has con-tained such fi res, allowing plant life that is

able to reproduce without fi res to overwhelm indigenous species. Recently, the Forest Ser-vice has permitted landscape-wide logging to reduce forest fi res – a practice that does not promote healthy forest growth.

When viewed in their totality, market-fail-ure policies often confl ict: policymakers want to use the antitrust laws to promote domes-tic competition, but they enact trade poli-cies that are not justifi ed on market-failure grounds and make it more diffi cult for for-eign fi rms to compete with U.S. fi rms. Policy-makers seek to lower the cost of transporta-tion externalities, yet they subsidize road use, and so forth.

political forcesIf it is so easy to identify fl aws in the design and implementation of policies directed at reducing market failure, why don’t policy-makers address these fl aws and develop more effi cient policies? George Stigler’s answer is that the primary intent of government inter-vention is to redistribute income to favored interests. Thus, it would be unreasonable to expect that government would place a high priority on effi ciency.

Some have argued – notably, Gary Beck-er of the University of Chicago – that policy-makers try to redistribute income as effi cient-ly as possible. But my own work with Vikram Maheshri of the University of California (Berkeley) suggests that interest groups that receive favors “invest” in their stock of politi-cal capital to secure preferential treatment in the future by strengthening the base of ben-efi ciaries as much as possible. Thus, dead-

Voters may not oppose the redistribution that results from inefficient market-failure policies

as long as they get their particular subsidy.

to

ng

ro

im

ag

e s

to

ck

/a

la

my


tk

weight costs are not necessarily minimized subject to political objectives.

For example, interest groups favoring pub-lic transit see to it that all users pay fares below cost rather than just those who are poor. Truckers are subsidized for their use of the highways instead of paying an effi cient pavement-wear tax. And because the broad base of taxpayers tends to be less aware of their interests than well-organized groups, a lack of public information may also create an imbalance of political pressure that enables economically ineffi cient policies to persist.

Voters may not oppose the redistribution that results from ineffi cient market-failure policies as long as they get their particular subsidy. Consider, for example, Hartwell C. Herring’s July 7, 2002, letter to the New York Times Magazine:

“The Amtrak subsidy, which is minuscule

compared to the bloated farm subsidy and pork-

ridden defense budget, is one of the few direct

benefi ts I get from the federal taxes I pay. I have

a neighbor who is a dairy farmer. He probably

collects more in subsidy from the government

than he pays in taxes. Not me. I am one of the

folks who pick up the tab, and I need my train

service.”

Congress is a powerful conduit for these interests through its appropriations to such agencies as the FTC and the Department of Justice, and through its fi nancing of public enterprises such as highways. Indeed, in some cases members of Congress are also direct benefi ciaries of federal largesse. Between 1996 and 2000, agricultural subsidies were paid to farm operations owned in whole or in part by Senators Charles Grassley (Iowa), Richard Lugar (Indiana), Blanche Lincoln (Ark.) and


Sam Brownback (Kansas), and Representatives Marion Berry (Arkansas), Calvin Dooley (Cal-ifornia), Charles Stenholm (Texas), Dennis Hastert (Illinois) and Bob Stump (Arizona).

Policy is also made by unelected bureau-crats, whose politics often mirrors the in-fl uences affecting Congress and the White House. Thus, federal regulatory agencies such as OSHA, FAA and EPA, and departments in-cluding Transportation and Agriculture are also subject to interest-group pressure that creates deadweight effi ciency losses while redistributing income from less affl uent to more affl uent citizens.

A few studies have tried to establish ex-plicit links between the ways various inter-est groups compromise policies and the in-

effi ciencies created as a byproduct. For ex-ample, my research with Chad Shirley of the RAND Corporation found that certain so-cioeconomic groups (such as high-income

residents) infl uenced urban rail transit fares and service, resulting in sharp deviations from effi cient pricing and service guidelines. Brian Knight of Brown analyzed Congres-sional support for projects that were part of the 1998 Transportation Equity Act for the 21st Century, not surprisingly fi nding that representatives supported highway demon-stration projects to benefi t their constituents. What may surprise, however, is that this polit-ical coalition resulted in a deadweight loss of roughly 96 cents for every dollar spent on the projects – or more than $7 billion.

Looking down the road, the doubling of registered lobbyists in Washington between 2000 and 2005 (the current number: roughly 35,000) indicates that policymakers are likely to be subject to even stronger political pres-sure from interest groups. Indeed, the returns from lobbying have been large, with the re-lationship between clients’ benefi ts and costs reported to approach 100:1 in some instances – that is, groups sometimes have spent just a penny on lobbying to get $1 in favors. Greater competition among lobbyists after the same government largesse may lower returns, but it may also increase the diffi culty of designing and inducing the appropriate incentives to re-duce government failure.

The doubling of registered lobbyists in Washington between 2000 and 2005 indicates that policymakers

are likely to be subject to even stronger political

pressure from

interest groups.

M as

so

cia

te

d p

re

ss

Government Failure Versus Market Failure C€¦ · Government failure, then, arises when government...

Documents

Transcript of Government Failure Versus Market Failure C€¦ · Government failure, then, arises when government...