JOBS, COMPETITIVENESS,AND ENVIRONMENTALREGULATION:WHAT ARE THE REAL ISSUES?

Robert Repetto

W O R L D R E S O U R C E S I N S T I T U T E

JOBS, COMPETITIVENESS, ANDENVIRONMENTAL REGULATION:What Are the Real Issues?

Robert Repetto

W O R L D R E S O U R C E S I N S T I T U T E

March 1995

Library of Congress Cataloging-in-Publication Data

Repetto, Robert C.Jobs, competitiveness, and environmental regulation : what are the real issues? / Robert Repetto.

p. cm.

Includes bibiliographical references.

ISBN 1-56973-030-X1. Environmental policy—Economic aspects—United States. 2. Environmental protection—Economic

aspects—United States. 3. Competition, International. 4. Labor supply—United States.I. Title

HC110.E5R42 1995

3637'00973—dc20 95-1197CIP

Kathleen CourrierPublications Director

Brooks BelfordMarketing Manager

Hyacinth BillingsProduction Manager

Great Northern Railway, Sam Fields, AFL-CIOCover Photos

Each World Resources Institute Report represents a timely, scholarly treatment of a subject of public concern. WRI takes re-sponsibility for choosing the study topics and guaranteeing its authors and researchers freedom of inquiry. It also solicits andresponds to the guidance of advisory panels and expert reviewers. Unless otherwise stated, however, all the interpretationand findings set forth in WRI publications are those of the authors.

Copyright © 1995 World Resources Institute. All rights reserved.Printed on recycled paper

CONTENTS

ACKNOWLEDGMENTS v

FOREWORD vii

I. WHAT ARE THE ISSUES? 1A. Dislocation of Trade and Investment 1B. Job Loss 2

II. ARE THESE ISSUES WORTH WORRYING ABOUT? 3A. The "Competitiveness" Issue in Principle 3B. The "Competitiveness" Issue in Fact 5

III. HOW ENVIRONMENTAL PERFORMANCE AFFECTS COMMERCIAL SUCCESS 11

IV. JOBS AND THE ENVIRONMENT 21

V. THE REAL ISSUE 25

VI. GETTING BETTER RESULTS FROM ENVIRONMENTAL SPENDING 29

APPENDIX 35

NOTES 39

LIST OF TABLES

Table 1. Share in Total World Exports of Manufactures and of Environmentally SensitiveGoods, Selected Industrial Countries, 1970-90 6

Table 2. United States' Exports and Imports, Aggregate Figures for 1990-92 7Table 3. U.S. Direct Foreign Investment, by Region, 1992 ($ million) 8Table 4. Measures of Environmental and Market Performance 14

LIST OF FIGURES

Figure l.A Simple Correlations: Toxic Emissions 15Figure l.B Simple Correlations: Water-borne Emissions 16Figure l.C Simple Correlations: Air Particulate Emissions 17Figure 2.A Partial Correlations: Toxic Emissions 18Figure 2.B Partial Correlations: Water-borne Emissions 19Figure 2.C Partial Correlations: Air Particulate Emissions 20Figure 3 Employment and Pollution Control Expenditures by Major Industrial Sector 23

APPENDIXTable IA. Correlations between Profitability and Air-borne Emissions 35Table IB. Correlations between Profitability and Water-borne Emissions 36Table IC. Correlations between Profitability and Toxic Releases 37

ACKNOWLEDGMENTS

The author's sincere thanks go to Manjula Singh Morgenstern, Richard Schmalensee, Roger Dower,for valuable research assistance and data analysis, Walt Reid, and Jonathan Lash were greatlyand to David Wheeler of the World Bank and Bob appreciated.McGuckin of the U.S. Bureau of the Census for their In addition, the author and WRI are grateful forcooperation. Thanks also to colleagues and former generous financial support for this study from thecolleagues at WRI who helped in various ways in the Joyce Foundation and the American Conservationwork underlying this report: Tine Nielsen, Maggie Association.Powell, Rob Gramlich, Kathleen Courrier andHyacinth Billings. Helpful comments from Richard R.R.

FOREWORD

Americans want a clean environment and havesaid that they are willing to pay for it, but the cost ishigh in terms of jobs and industrial competitive-ness—right?

Wrong. Contrary to their own expectations, econ-omists have found little evidence that the costs of en-vironmental protection have affected the competitive-ness or profitability of U.S. firms or reduced thenumber of jobs in the economy.

In Jobs, Competitiveness, and Environmental Reg-ulation: What are the Real Issues?, WRI economistRobert Repetto summarizes the results of recent stud-ies and examines important new data from thousandsof U.S. industrial operations. He compares environ-mental performance with profitability and finds thatplants with poor environmental records are generallynot more profitable than cleaner ones in the same in-dustry, even controlling for their age, size, and tech-nology. This is true in "dirty" pollution-intensive in-dustries as well as in clean ones.

Environmental protection is not free. We paysome $200 billion a year to avoid the ravages of pol-lution and the destruction of natural resources.

But the real question is not how to resolve theconflict between our economic goals and our envi-ronmental goals. Indeed, that conflict is largely imagi-nary. The real question is how to get more environ-mental protection for less money. How can weenhance the incentives and opportunities for techno-logical and process changes that help the economyand the environment? The answer is that we mustchange the way we pursue our goals. Today's rigid,one-size-fits-all, command-and-control regulationsblock progress more than they spur it. They were de-signed to compel compliance by reluctant industries,but now they restrain the introduction of new envi-ronmental technologies.

Instead of throwing hurdles across the path, arational regulatory system would set environmentalgoals and then allow regulated industries to meetthem in the most efficient way, and it would offer in-dustries economic incentives to continually improve

their environmental performance. If the 104th Con-gress takes this tack, Americans across the politicalspectrum will support a large-scale regulatory reformeffort.

In some cases, fees would be the most efficientway to deter environmental damages. Dr. Repettosuggests fee-based strategies to curb automotive airpollution, a growing problem as more and more dri-vers log more and more miles. Raising parkingcharges can discourage solo commuting. Peggingregistration fees to tailpipe emissions can help get thedirtiest cars off the road. Charging higher tolls duringrush-hour can reduce traffic congestion. As Dr.Repetto notes, fees are so much more efficient thancurrent regulation that even some industries favorthem.

Jobs, Competitiveness, and Environmental Regu-lation extends the policy analyses and recommenda-tions set forth in such previous WRI studies as GreenFees-. How a Tax Shift Can Work for the Environmentand the Economy, A New Generation of Environmen-tal Leadership: Action for the Environment and theEconomy, and Paying the Farm Bill: U.S. AgriculturalPolicy and the Transition to Sustainable Agriculture.By detailing ways that the United States can improveenvironmental quality at less cost to consumers, busi-nesses, and taxpayers, WRI's economists point theway toward an economically and environmentallysustainable future.

We would like to thank The Joyce Foundationand the American Conservation Association for theirfinancial support of this study. To both, we aredeeply grateful.

Jonathan LashPresidentWorld Resources Institute

I. WHAT ARE THE ISSUES?

A. DISLOCATION OF TRADE ANDINVESTMENT

The U.S. economy has improved in the last sev-eral years: employment is up, and productivity hasaccelerated; the federal deficit is lower, and eco-nomic growth has been relatively strong. Nonethe-less, progress in environmental protection is at astandstill. Almost all reauthorizations and new enact-ments of environmental legislation stalled in the re-cently ended 103rd session of Congress. The newmajority in the 104th Congress proposes measuresthat would severely curtail environmental protectionat the federal level. Many businessmen, labor union-ists, politicians, and ordinary citizens fear that Amer-ica can't afford the costs of stronger environmentalprotection, that regulatory burdens are underminingour competitive position internationally, destroyingjobs at home, and dragging down productivity andgrowth.

The United States does spend more on environ-mental protection, absolutely and as a percentage ofgross domestic product, than any other country.1 In1990, the percentage had already reached 2.1 and thetrend is still upward. A complex web of environmen-tal laws and regulations—thousands of pages ofdense, obscure, and sometimes vague language—hasgrown up piecemeal over the past twenty-five years.Various state and federal courts and agencies inter-pret and enforce these requirements, sometimes in-consistently. Regulations limit industry's choice oftechnologies, product design and mix, plant location,and other important production decisions. Firms mustallocate investment and operating funds to reduceenvironmental impacts, with scant hope of recoveringall these expenditures through materials or energysavings or higher product prices. In addition to directcompliance costs, industries face delays and uncer-tainties in dealing with regulatory requirements.2

It is argued that environmental regulations im-pose costs and restrictions on industries in the UnitedStates that rivals in other countries do not face, and

thus put American firms at a competitive disadvan-tage.3 Especially in pollution-intensive sectors thathave been heavily impacted by regulatory require-ments, such as chemicals and petrochemicals, pulpand paper, metals and metal products, and trans-portation equipment, this competitive disadvantagehas allegedly contributed to a loss of America's mar-ket share in world trade. Moreover, it is claimed, inorder to escape the burdens of environmental regula-tion, U.S. and other multinational companies have lo-cated new plants in other countries where environ-mental costs are lower and regulations less stringent.

These concerns have spilled over into the policyarena. They have led to legislative proposals thatwould impose countervailing duties on imports fromcountries with weak environmental standards, inorder to offset the putative cost disadvantages U.S.firms face.4 Related proposals call for using anti-dumping provisions of U.S trade law against foreignproducers who fail to incorporate environmentalcosts fully into their export prices.5 Concerns aboutMexico's relatively lax environmental regulations fig-ured prominently in the NAFTA debate and resultedin elaborate safeguard mechanisms to ensure that dif-ferences in environmental standards would not dis-tort trade flows.6 The same worries have spawned avigorous debate over whether GATT rules should beamended or reinterpreted to allow the United Statesto apply trade penalties based on the environmentaleffects of production processes and methods used inother countries—an idea that excites grave fears inother parts of the world of disguised protectionism orlarge-country pressures to adopt excessively strict en-vironmental standards, or both.7

Concerns over competitive disadvantage manifestthemselves also in strong pressures, especially amongOECD countries, to harmonize their environmentalstandards—not only those that apply to the character-istics of traded commodities but also those that gov-ern the methods used to produce such goods. Pres-sures for harmonization of environmental standardsraise a host of issues: harmonization toward what

level? How can pressures to adopt the weakest stan-dards be avoided? Must state and local governmentsrelinquish their standard-setting prerogatives? Howare various national interests to be represented in in-ternational standard-setting processes? These issuesarise primarily because harmonization is thought nec-essary to "level the playing field" and avoid competi-tive disadvantage.

Is all this necessary? Must we pay a heavy pricein international trade and investment for environmen-tal protection, or can we have our cake and eat ittoo? The counter-argument, articulated first by busi-ness school professor Michael Porter, asserts thatstringent environmental regulations may lead firms todevelop new, less-polluting and more efficient prod-ucts and manufacturing processes. Such innovationsgive firms that have responded creatively to regula-tion a competitive advantage over sluggardly rivals asenvironmental standards tighten worldwide.

"Ultimately, nations succeed in particular in-dustries because their home environment isthe most forward-looking, dynamic, andchallenging... Strict government regulationscan promote competitive advantage by stim-ulating and upgrading domestic demand.Stringent standards for product performance,product safety, and environmental impactpressure companies to improve quality, up-grade technology, and provide features thatrespond to consumer and social demands.Easing standards, however tempting, iscounterproductive."8

A somewhat different counterargument hinges onthe rapid growth in markets for goods and servicesthat "solve" environmental problems. According torecent surveys, these "green" industries, which sellpollution monitoring and abatement equipment, engi-neering and construction services, and a variety ofproducts with environmentally superior characteris-tics, have already reached almost 200 billion dollarsin sales annually in the industrialized countries alone,and are expected to expand even more rapidly in the

newly industrialized countries where environmentalconditions demand increased attention.9 Countrieswith more stringent environmental standards in theirhome markets will allegedly develop a competitive ad-vantage in these "green" industries, offsetting whateverdisadvantage those standards impose on the "dirty" in-dustries. Both counterarguments suggest that our rela-tively strict environmental standards are likely to con-fer benefits on American industry in the long-run.

B. JOB LOSS

A parallel debate revolves around the effect ofenvironmental regulations on the employment rate.Business spokesmen frequently argue that stricter en-vironmental standards will force them to close downfactories or move them overseas. Restrictions on nat-ural resource use, such as limits on timber harvestingon public lands, are attacked because they reduceemployment along with production. Labor unionistsalso fear job losses if environmental regulations raiseproduction costs or restrict supply. Many economistssubscribe to a more sophisticated version of this ar-gument, pointing out that diverting capital to investin pollution-control equipment instead of capacityexpansion or productivity improvement also limitsthe growth of output and employment over time.10

The usual riposte is that environmental protec-tion actually creates more jobs than are lost: limitsput on the extraction of natural resources maythreaten jobs in extractive industries, but will save orcreate jobs in recreation industries and in footloosehigh-tech industries attracted to a high-quality envi-ronment. Environmental regulations that require pol-lution abatement or raise energy prices create jobs inindustries supplying pollution-control or energy-con-servation equipment and services. Since these indus-tries are more labor-intensive than the heavily pollut-ing industries (e.g., energy supply, basic metals, andchemicals) it is argued that greater expenditures onenvironmental protection will create jobs on balance,even if it's at the expense of employment in the pol-luting sectors.

B

II. ARE THESE ISSUES WORTH WORRYING ABOUT?

A. THE "COMPETITIVENESS" ISSUEIN PRINCIPLE

The proposition that differential environmentalstandards lead to loss of competitiveness and em-ployment is so obvious to many businessmen, laborleaders, and politicians that it is regarded as ax-iomatic. Its validity needs no demonstration: if U.Sfirms are forced to incur costs that their internationalrivals are not and these costs are not matched bymarket benefits, then profitability or market sharewill suffer, so output and employment will bereduced.

Despite its plausibility, the proposition is flawedboth in principle and in fact. In principle, competi-tiveness—manifested as an increase in net exports ina single industry or in all industries together—is not avalid economic objective. The quest for competitive-ness rests on topsy-turvy mercantilist principles thatequate exports with economic advantage and importswith economic harm. From this standpoint, if theUnited States became increasingly "competitive" in allsectors, we'd export a great deal and import little.But, there is no reason to export except to trade forthings we want and cannot make as well or ascheaply at home. Exporting just to amass foreign cur-rencies or other financial assets without eventuallyimporting in return makes no sense. From an eco-nomic and an environmental perspective, the lessproduction needed to support any standard of con-sumption, the better. If countries can acquire whatthey want at lower real cost through internationaltrade, they're better off.

Preoccupation with "competitiveness" reflects thealmost total dominance of producer interests overconsumer interests in trade policy, which is thereforeinveterately mercantilist. In the Uruguay Round andin previous trade negotiations under the GeneralAgreement on Tariffs and Trade (GATT) or in othersettings, when a country lowers its barriers to im-ports, it is viewed as a "concession" to foreign coun-tries, although the main beneficiaries are consumers

in the country that cuts its tariffs. If the United Statesis a net importer of a particular good, then the valueof its consumption exceeds that of its domestic pro-duction, and the benefits of a lower price to con-sumers are typically much greater than the harmdone to domestic producers.11 Thus, the U.S. econ-omy also gains overall. Yet, mercantilist trade policyholds that a country that lowers its import barriers inits own interests deserves to be "compensated" by itsforeign trading partners by similar cuts in their importbarriers.

The idea that extremely poor peoplecan better afford to pay inflated pricesfor inferior quality merchandise wouldbe laughable if not so tragic. By interna-tional agreement, the developing coun-tries were given discretion to shootthemselves in the foot.

This way of looking at trade policy reflects pro-ducer interests completely. Still, since successive ne-gotiating rounds have lowered trade barriers substan-tially, this perspective might be dismissed as a quaintbut innocuous convention in the specialized jargonof multilateral trade policy—were it not for the enor-mous damage the underlying assumptions havedone, especially in the developing world. For theGATT granted the developing countries—becausethey are poor—special dispensation to establish andmaintain higher levels of protection for their produc-ers than their richer trading partners do. What adiplomatic triumph that was! The idea that extremelypoor people can better afford to pay inflated pricesfor inferior quality merchandise would be laughableif not so tragic. By international agreement, the

developing countries were given discretion to shootthemselves in the foot. Of course, most developingcountry governments jumped at the invitation to pro-tect their domestic producers, and so created monop-olistic, inefficient, and technologically backward in-dustries. At the same time, they reduced theirpeoples' living standards. Only in recent years—andmostly through unilateral action rather than throughmultilateral negotiation—are developing countrieslowering the barriers to international commerce thathave lowered their real incomes and retarded theirgrowth.

The dominance of producer interests in tradepolicy perhaps explains why such international tradeorganizations as the GATT have had so much moretrouble incorporating environmental objectives intotheir operations than other inter-governmental orga-nizations have. It has been the custom and privilegeof producers everywhere, but especially in the devel-oping and formerly socialist economies, to external-ize some of their production costs by dumping virtu-ally all their wastes—however toxic—outside thefactory gate into the most convenient water body, airstream, or vacant lot. Few developing or transitionaleconomies devote even one third of one percent oftotal income to environmental control. The resultingpollution exacts a heavy toll on people's health andwelfare, as well as on surrounding enterprises depen-dent on increasingly degraded natural systems.12

These real economic damages, which total 1 or 2percent of GDP in the industrialized countries, canreach 4 percent of GDP or more in the newly indus-trializing and resource-dependent economies.13 Over-whelmingly, these environmental damages are borneby domestic residents and firms in the form of illhealth, reduced productivity and higher costs. Apartfrom the greenhouse gases and ozone-depletingCFCs, few pollutants cross international boundaries.

Yet, so dominant are producer interests in tradepolicy that developing countries complain that theireconomies would become less competitive if they en-acted and enforced measures to reduce the injuriessuffered by their own populations and natural systems.Simultaneously, producer interests in the developedcountries complain of unfair competition from importsproduced under weaker environmental standards innewly industrializing nations. In fact, by allowing their

firms to externalize significant production costs bydumping their wastes indiscriminately, developingcountry governments are subsidizing consumers inrich countries at the expense of their own populationsand national economies. Since firms don't have toincur the costs of pollution control, those costs are notreflected in the prices of exported commodities.Therefore, consumers in the importing countries don'thave to pay any share of the environmental controlcosts. Nonetheless, rich country governments, reflect-ing producer interests, complain of damage from un-fair competition, and developing countiy governmentscomplain of "Northern over-consumption" but resistthe measures that would make those consumers paytheir way. If exporting countries sell their wares belowcost by failing to internalize environmental damagesinto producers' costs and prices, then the importingcountry is the gainer and the exporting country is theloser.14 But one would never guess that by listening tothe trade policy debate.

If exporting countries sell their waresbelow cost by failing to internalize envi-ronmental damages into producers'costs and prices, then the importingcountry is the gainer and the exportingcountry is the loser. But one wouldnever guess that by listening to the tradepolicy debate.

In any case, the effects of environmental regula-tions on trade shouldn't be judged at the level of theindividual firm or even the individual industry. Busi-nessmen care about the fortunes of their own firms,but public policy must be constructed on a broaderframe. If one firm lacks the technological or manage-rial capability to meet an environmental standard effi-ciently, then another firm in the same industry maygain market share at its expense. Governmentsshould not (but often do) tailor policy to the leastcapable of firms within an industry.

Moreover, many of the pollution control costs aswell as the costs of environmental damage originat-ing in a single industry are diffused throughout theeconomy over time. Even though in the short run thepolluting firm may pay the costs of abatement, mostof those costs are eventually passed along to cus-tomers. If the polluting firm produces capital goodsor intermediates, these customers are other firms.15

Analogously, most environmental damages from un-controlled pollution are borne not by the offendingfirm but by other households and enterprises. Healthdamages lower productivity and raise health carecosts throughout the economy. Chemical and oilspills drive up insurance rates for all firms, not justthe careless ones. Air and water pollution from basicindustries raise costs or reduce profits in such unre-lated industries as agriculture, forestry, fishing,tourism and outdoor recreation. Consequently, the ef-fects of environmental regulation must be evaluatedat the level of the economy as a whole.

Unfortunately, the models and methodologiesnow used to do that are fatally flawed. Empiricalmacroeconomic models used by leading academiceconomists and economic consulting firms to esti-mate the economic effects of environmental regula-tions completely omit the damages that pollution andother environmental impacts impose on consumersand even on producers. They only include the costsof pollution abatement. Naturally, they conclude thatenvironmental regulations impose an overall cost onthe economy. President Truman, tired of economicadvisors who always said "On the one hand, this...and on the other hand, that...." once beseeched hisstaff to find him a one-armed economist. Wishes aredangerous—there's always the chance they'll begranted. Now most macroeconomists look only at thecosts of reducing environmental damages and ignorethe costs of not reducing those damages.16

Several economists have also estimated the ef-fects of environmental regulation on productivitygrowth at the industry level. In principle, such stud-ies are more valid than those that focus merely ontrade dislocations. Yet, estimates of productivity im-pacts also measure only the effects of regulation onindustry costs, but don't account for the reductions inpollution damages attributable to those regulations.17

For example, if an electric utility switches to low-sul-

phur coal in response to the Clean Air Act, its esti-mated productivity declines because low-sulphur coalcosts more per BTU and generates no more electric-ity. Nowhere do the productivity estimates reflect thereduced damages from respiratory disease or fromacid deposition on forests and materials. Productivitymeasurements that include both the costs and bene-fits of environmental regulations lead to dramaticallydifferent conclusions. Environmental regulations maywell raise the rate of productivity growth, if theirbenefits exceed their costs.18

B. THE "COMPETITIVENESS"ISSUE IN FACT

Any significant change in a country's export costswould lead over time to an adjustment in the ex-change rate or in real wage levels to maintain thebalance of international payments, so efforts to lookat the effects of environmental regulations have hadto try, in principle, to hold these variables constant.In practice, economists have investigated the compet-itiveness issue by looking at

a) whether highly regulated industries suffer ad-verse trends in net exports relative to lightlyregulated industries;

b) whether production of highly regulated indus-tries moves abroad to less regulated countries;

c) whether U.S. firms in highly regulated indus-tries invest overseas in less regulated coun-tries;

d) whether such basic indicators as productivityare adversely affected in highly regulatedindustries.19

Economists who have reviewed the research onthis subject, which includes a number of careful andingenious studies, find scant evidence that environ-mental regulation has had adverse effects by any ofthese measures. The reason why most efforts to findadverse effects have come up empty is evident fromthe historical data. Consider exports from industriesheavily impacted by environmental regulations in theindustrialized countries, relative to other exports fromthose countries. The industries that spend most tocomply with environmental regulations are pulp andpaper, petroleum products, organic and inorganicchemicals, coalmining, fertilizer, cement, ferrous and

non-ferrous metals, metal manufactures, and woodmanufactures such as veneers and plywood.20 A re-cent World Bank report reviewed trends in worldtrade in these products from 1970 to 1990, a periodin which most industrial countries put their environ-mental regulations into effect. The report found that"Contrary to common perceptions, higher environ-mental standards in developed countries have nottended to lower their international competitiveness.There has been little systematic relationship betweenhigher environmental standards and competitivenessin environmentally sensitive goods (those that in-curred the highest pollution abatement and controlcosts...)."21

In fact, as the data in Table 1 show, the countrieswith tight environmental standards have had moreexport success in these environmentally sensitive in-dustries than in manufacturing industries as a wholeor in their entire range of industrial and agriculturalexport products. Between 1970 and 1990, the indus-trial countries' overall share in world exports de-clined from 74.3 to 72.7 percent, mainly because therest of the world experienced faster economic growthand now contributes a larger share of world outputthan before. The industrial countries' share of world

Contrary to widespread perceptions,the industries heavily affected by envi-ronmental regulations did relativelywell in international trade.

exports of manufactured goods declined even more(from 91-3 to 81.3 percent), largely because the com-position of expenditures and output in the rich coun-tries shifted toward services while that of the coun-tries in the early stages of industrialization shiftedaway from agriculture toward manufactures. How-ever, within the category of manufactured exports,the share of the advanced countries in exports in in-dustries that experience the highest pollution controlcosts has actually declined by very little (just from81.3 to 81.1 percent). The sectors in which the indus-trial countries markedly lost their comparative advan-tage were not those heavily affected by environmen-tal regulations but rather those in which labor costsare a large fraction of total costs, such as textiles,

Table 1. Share in Total World Exports of Manufactures and of Environmentally Sensitive Goods,Selected Industrial Countries, 1970-90.

Regions/Countries

Industrial Countries, of which,AustriaFinlandNorwaySwedenGermanyJapanUnited States

Total1970

74.31.0

0.8

0.8

2.3

11.76.6

14.5

Exports1990

72.71.30.8

1.0

1.8

12.28.8

11.4

All Manufactures1970 1990

91.31.3

0.90.8

2.9

17.210.216.9

81.31.6

0.90.52.0

15.211.812.3

EnvironmentallySensitive1970

81.31.32.1

1.9

4.012.18.0

11.6

Industries1990

81.12.0

2.4

1.7

3.413.88.0

10.1

Source: Piritta Sorsa, 1994, Table 2 and Annex Table 2.

apparel, footwear, and other light manufactures. Con-trary to widespread perceptions, the industries heav-ily affected by environmental regulations did rela-tively well in international trade.

Among the industrial countries, the United Stateswas no exception. As evident in Table 1, our share inworld exports has declined along with our fallingshare in world output, and our share in manufacturedexports has declined considerably faster. But, withinthe manufacturing sector, the decline in our share inexports of environmentally sensitive products hasbeen much less than the average. In other words, theindustries most affected by regulations have per-formed relatively well in international trade over aperiod in which regulatory compliance costs havebeen rising. Of course, these trends don't implycausality. They merely suggest that other, more pow-erful forces have been at work reshaping the worldeconomy. They also show why statistical studies havenot been able to show any consistent link betweenenvironmental regulation and trade performance.

The diversity in the experiences of industrializedcountries reinforces the point. Germany, for example,which in many respects has tighter environmentalstandards than the United States does, actually in-creased its export share in environmentally sensitivegoods while losing market share in manufactures as a

whole. Japan, whose industries are typically less pol-luting than their U.S. counterparts, held its own inthe sectors most affected by regulation. When theperformance of individual industries within the envi-ronmentally sensitive group is examined, the diver-sity of experience increases further: the U.S., forexample, seemed to strengthen its comparative ad-vantage in 17 of 38 individual environmentally sensi-tive industries, and lose ground in the rest.22 Clearly,important factors other than regulation are at work.

A broad look at the U.S. trade balance with othercountries and regions also casts doubt on the tradeimpacts of differences in pollution control costs. Overthe early 1990s, the U.S. had an overall trade deficit,which reflected our macroeconomic imbalance. Therest of the world was lending us money to finance theexcess of our total consumption over our aggregateproduction, which meant that the United States had tohave an import surplus. However, our trade deficitwith Japan was relatively large, although Japan's envi-ronmental standards are stricter than our own in mostrespects.23 The United States maintained a trade sur-plus with Mexico, even though Mexico's environmen-tal standards were significantly weaker than ours. Ingeneral, as Table 2 illustrates, the pattern of U.S. tradedeficits had no relation to the environmental stan-dards of our trading partners relative to our own.

Table 2. United States' Exports and Imports. Aggregate Figures for 1990-92.

Canada

Japan

Germany

Other Industrialized CountriesAfrica, total

Asia, excl. JapanMexico

Other Western Hemisphere, totalE. Europe & F.S.U.

Exports1990-92

258,261

144,49661,252

224,375

18,589202,458

102,24990,921

10,845

Imports1990-92

288,808

287,561

85,591259,455

46,064

338,92398,549

106,2473,295

Exports * Imports

0.89420.50250.71560.86480.40350.59741.03750.85583.2914

Source: Directory of Trade Statistics Yearbook, IMF, 1993.

Looking at investment flows to less developedcountries doesn't change the picture. The data on di-rect foreign investment provide no support for thecontention that multinational companies are relocat-ing environmentally sensitive industries in countrieswith weak regulations. It is true that direct foreign in-vestment in developing countries has increasedsharply since the mid-1980s after collapsing duringthe debt crisis in the first half of the 1980s. For exam-ple, by 1992, the developing and transitionaleconomies received nearly half—45 percent—of U.S.direct investment abroad (USDIA). But a muchsmaller proportion of that direct investment went intothe environmentally sensitive industries (petroleumand gas, chemicals and related products, and primaryor fabricated metals) than was the case for U.S. directinvestment abroad in the already developed countrieswith relatively tight environmental standards. Table 3shows that 24 percent of USDIA into the advancedcountries went into pollution-intensive sectors, butonly 5 percent of USDIA into the less developedeconomies went into those sectors. Of the total directforeign investment in pollution-intensive industries,84 percent went to other developed countries, com-pared to 49 percent of overseas investment in otherindustries. To the extent that the advanced countriesseem to be exporting their "dirty" industries, theyseem predominantly to be sending them to eachother, not to the less developed economies.

To the extent that the advanced coun-tries seem to be exporting their "dirty"industries, they seem predominantly tobe sending them to each other, not to theless developed economies.

This phenomenon is corroborated by trends indeveloping countries that are major recipients ofdirect foreign investment and keep statistics by sectorof destination. In Nigeria, Hong Kong, Korea, Malay-sia, Philippines, Singapore, Taiwan, Thailand, Ar-gentina, Brazil, Colombia, and Venezuela together,and in each one individually except Venezuela, thestock of inward foreign direct investment in the pol-lution-intensive industries represents a smaller shareof total foreign direct investment now than in the1960s or early 1970s, despite the fact that environ-mental regulations have tightened in the countriesmaking the foreign investments.24 This implies thatsince 1970 foreign direct investment has increasedmuch faster in other sectors than in the pollution-in-tensive industries. The multinational companies thathave really been raising their stakes rapidly in the

Table 3. U.S. Direct Foreign

Receiving Region

Developed CountriesDeveloping CountriesTotalDeveloped Countries + Total

Investment,

Petroleum&Gas

(1)

171

-327-156n.a.

by Region,

Sector

Chemicals(2)

4,0701,0075,0770.80

1992 ($ million).

Metals(3)

503247750

0.67

Subtotal(4)

4,744927

5,6710.84

All otherSectors

(5)

15,35916,09231,4510.49

Total(6)

20,10317,01937,1220.54

(4) + (6)

0.2360.0580.15

Source: "U.S. Direct Investment Abroad," U.S. Department of Commerce, Survey of Current Business, July 1993;Table 12.4, p. 104.

developing world include consumer products compa-nies such as Coca-Cola, service companies such asCiticorp, and makers of apparel, appliances and otherlabor-intensive products or components.

To be sure, the share of the developing countriesin world production and trade in the pollution-inten-sive sectors has grown, but this is not necessarily be-cause of differences in environmental standards.25

Demand for these products has grown faster in therapidly industrializing countries of Asia and LatinAmerica. Production has followed the growth of de-mand. The relocation of production in these basic in-dustries to the newly industrializing countries canalso be attributed to the well-known "product cycle"described decades ago by Raymond Vernon andothers.26 As industries mature with respect to productand process designs, their outputs become more like"commodities" subject primarily to price competition,which induces migration to low-cost producing coun-tries. Cost advantages may stem from lower wages ormaterials costs. Advanced countries maintain compar-ative advantage in technologically sophisticated in-dustries and in new products designed for high-in-come consumers. The product cycle can readilyexplain the modest gains the developing countrieshave made in basic chemicals, metals, pulp andpaper, and other polluting industries.

In the face of these basic trends in internationaltrade and investment, there's little wonder that econo-metric investigations find scant evidence that differ-ences in environmental regulations affect patterns oftrade, foreign investment or industrial location. JudithDean, a professor at the School of Advanced Interna-tional Studies at Johns Hopkins University, surveyedan extensive economics research literature datingmostly from the 1970s and 1980s. Her conclusion:

"More stringent regulations in one countryare thought to result in a loss of competitive-ness, and perhaps in industrial flight and thedevelopment of pollution havens. The manyempirical studies that have attempted to testthese hypotheses have shown no evidence tosupport them."27

Other experts have gone over the same ground. A re-cent OECD volume summarizing a symposium ontrade and environment concluded:

"Empirical studies show that the costs of pol-lution control are a small part of total costsin most sectors and that nearly all the OECDcountries have introduced similar environ-mental measures at roughly the same time.Environmental measures have not been thesource of significant cost differentials amongthe major competitors and have had minimaleffects on overall trade between OECD andnon-OECD countries."28

A still more recent literature review by economistsfrom Harvard University, the National Bureau of Eco-nomic Research, and Resources for the Future drewvirtually the same conclusion:

"We assess the evidence and find that thereis little to document the view that environ-mental regulations have had a measurablyadverse effect on competitiveness. Althoughthe long-run social costs of environmentalregulation may be significant, including ad-verse effects on productivity, studies attempt-ing to measure the effect of environmentalregulation on net exports, overall tradeflows, and plant location decisions have pro-duced estimates that are small, statistically in-significant, or not robust.. ."29

These economists also find little evidence to sup-port Michael Porter's counter-hypothesis thatstricter regulations actually improve internationalcompetitiveness.

A well-known environmental lawyer in a recentlaw review article has provided a somewhat morepessimistic reading of essentially the same body ofevidence, but his conclusions were based mainly onthe presumption of unmeasured costs of environmen-tal regulation in addition to pollution control costs,such as legal expenses, regulatory delays and uncer-tainties. Such costs undoubtedly exist in the UnitedStates, largely as the result of our litigious, adversar-ial, command-and-control approach to regulation, butthe author takes little account of the very significantoverall regulatory delays and uncertainties facing pri-vate investors in less developed countries.30

Studies have also investigated whether differ-ences in the stringency of environmental regulations

from state to state within the United States have hada measurable effect on the location of new industrialplants. The answer is generally no. Other factorsdominate.31 This is a more sensitive test of the impactof environmental factors on investment decisions.States do differ in the stringency of their emissionsstandards and in the resources they put into enforce-ment of environmental regulations. Other locationalcosts probably vary less among regions within thecountry than between the United States and foreigncountries. So, if environmental factors don't affect lo-cational decisions within the United States, they areunlikely to affect investment decisions internationally.

In summary, the many economists who have in-vestigated the impact of environmental standards ontrade and investment and those who have reviewedthe research literature have consistently found thatregulatory differences among jurisdictions have nosignificant impact on the direction or magnitude oftrade and investment flows, even in industries whosecompliance costs are relatively high. These findingsare perfectly consistent with the basic facts presentedabove on trends in North-South trade and investmentover the past twenty years, which give no indicationthat countries with more stringent standards have suf-fered a loss of international competitiveness.

III. HOW ENVIRONMENTAL PERFORMANCE AFFECTSCOMMERCIAL SUCCESS

If "competitiveness"—the ability to sell in compe-tition with foreign producers—is not a good indicatorof commercial success, then what is? In a competitiveeconomy, profitability is a much better measure. Itencompasses success in the domestic as well as inthe international market and reflects costs of produc-tion along with sales volume. Profitability, literally"the bottom line" in a market economy, captures allthe factors influencing the success of the enterprise,while export sales measure only one aspect of suc-cess. For this reason, the "competitiveness" issue isbetter posed in a different form:

Do establishments with superior environ-mental performance tend to be more or lessprofitable than establishments with inferiorenvironmental performance within the sameindustry?

This question focuses on actual environmentalperformance rather than on regulatory "stringency,"which can't be defined or measured. Comparing legalrequirements won't do: strict regulatory standardsaren't stringent if they're not enforced. Moreover, U.S.federal environmental regulations now fill 16 vol-umes, so finding a single summary measure of regu-latory stringency is virtually impossible. Comparingthe stringency of regulations in different countries iseven more difficult, since regulations are multi-dimensional and countries' administrative approachesvary widely. One country may be tougher on certainforms of pollution and laxer on others. Furthermore,within any industry some firms will be operating wellwithin their permitted emissions while others may beout of compliance. Using a firm's expenditures onpollution control as a surrogate indicator confusesthe issue, because inefficient firms will probablyspend more to comply with the same regulationsthan efficient ones will. What matters is their actualenvironmental performance. The right question iswhether firms whose environmental performance isbetter than their competitors within the industry aremore or less successful in the marketplace.

The right question is whether firmswhose environmental performance isbetter than their competitors within theindustry are more or less successful inthe marketplace.

Environmental performance is measured by emis-sions per unit of shipments. If industrial processesare viewed in thermodynamic terms as the transfor-mation of materials and energy from crude into us-able forms, then the ratio of waste products to useful,salable outputs is one measure of the efficiency ofthe process. Since all materials that enter an industrialprocess must come out again in some form as physi-cal outputs, because matter is neither created nor de-stroyed, emissions per unit of shipments reflects theratio of useful "good" outputs to useless "bad" out-puts. Viewed in this way, it makes more sense to hy-pothesize that industrial processes that transform alarger fraction of the energy and materials they useinto salable forms might be more profitable.32

The standard hypothesis, of course, is that betterenvironmental performance comes at a cost, so firmsthat divert resources to reduce their emissions be-yond the point at which waste recovery just pays foritself must sacrifice some profits. Under this hypo-thesis, environmental performance and profitabilityshould be inversely related. The competing "PorterHypothesis" holds that once firms are motivated toseek out solutions to environmental problems—byregulations or other pressures—they typically findpreviously overlooked cost-saving opportunities toimprove processes, reduce wastes, or redesignproducts.

Economists view with enormous skepticism thehypothesis that firms typically overlook opportunities

to reduce costs or improve product quality.33 One ofthe most important insights in economics is that mar-ket competition continually pressures firms to maxi-mize profits by reducing their operating costs and im-proving their products. This explains why firms in amarket economy are more efficient in providinggoods and services than organizations not subject tomarket competition—the U.S. Congress, for example.However, in their formal analyses, economists takethis insight a step further and stipulate that mostfirms throughout the economy have already opti-mized their operations, an assumption that absolutelydumfounds anybody who has actually worked insidea corporation for more than a week. This extraordi-nary assumption is analytically convenient: econo-mists can say much more about some observed be-havior if it reflects the maximum attainable value ofsome objective—such as profitability—than if it is justpart of a general muddling along. However, if it weretrue that companies typically operate at maximum ef-ficiency, it would be hard to understand exactly whatthe hordes of management consultants swarmingaround them are being paid to do. To take a specificexample, it would be hard to understand how theFord Motor Company, after watching their Japaneserivals at work, could achieve radical cost savings inproducing new models—after almost a century in thebusiness—by starting to have their designers talkwith their manufacturing engineers and marketing ex-perts while the designs are being worked out.34

Alternative models of organizational and manage-rial behavior featuring bounded rationality and adap-tive decision-making, "satisficing" behavior, principal-agent problems and other incentive failures withinthe organization can help explain why firms don'toperate as efficiently as possible. Economists havehelped develop these models.35 Such models havebeen applied to environmental issues to explain whyfirms that agreed to cooperate with EPA's voluntary"Green Lights" program by investing in cost-effectiveenergy-saving investments have been able to findmany projects that earn relatively high rates of return,projects that presumably were available before thecompanies joined the program.36 But, in most analyti-cal work, economists treat these inefficiencies as spe-cial cases. In this investigation, the Porter hypothesisis reflected in the possibility that firms with superior

environmental performance also achieve superiorprofitability within their industries.

The empirical tests of these competing hypothe-ses make use of a relatively new database generatedby the U.S. Census Bureau's Center for EconomicStudies, the Longitudinal Research Database (LRD).37

Taking advantage of new possibilities in data pro-cessing and retrieval, this database merges records onindividual industrial establishments from six censusesof manufactures and twelve or more annual surveysof manufactures. Each census covers more than200,000 large manufacturing establishments, and con-tains detailed information on each establishment's lo-cation and ownership, its inputs of materials, energy,labor, and capital and its outputs of products and ser-vices. The Annual Survey of Manufactures is a muchsmaller stratified sample designed to include mostlarge establishments in surveys taken periodically innon-census years. It contains most of the same infor-mation collected in the censuses plus detailed infor-mation on assets, investments, depreciation, andother costs.

Parts of this large core database have beenmerged, establishment by establishment, with infor-mation from other sources, including databases onemissions and pollution control expenditures bymanufacturing firms. For the 1987 census year, LRDhas been combined with EPA's Toxic Release Inven-tory, which provides information on the releases anddischarges of over 300 toxic substances,38 the Na-tional Emissions Data Systems, which gives informa-tion on the discharge of non-toxic effluents into sur-face waters, and the Aerometric Information RetrievalSystem, which documents the atmospheric release ofpollutants regulated under the Clean Air Act. LRDdata have also been merged with information fromthe Commerce Department's Pollution Abatementand Control Expenditure surveys.3^ The result is adatabase encompassing thousands of manufacturingestablishments (the exact number depending onwhich environmental data are being matched to theLRD data) and containing detailed information onemissions, production costs, sales, and revenues.Using this database, it was possible to investigatewhether firms with superior environmental perfor-mance were more or less profitable than theircompetitors.

In measuring environmental performance, thetoxic release data have been kept distinct from infor-mation on conventional pollutant releases into airand surface waters. Separate emissions-to-shipmentsratios have been calculated for all three, in order toavoid reducing drastically the size of the sample ofestablishments that could be used in the analysis.Relatively few establishments could be matched fromall four datasets. Also, looking at airborne, water-borne, and toxic emissions separately reduced the al-ready difficult problems of aggregating emissions ofvarious substances. The Toxic Release Inventory wasaggregated into total pounds released into all media,including transfers to treatment works, ignoring thewidely differing toxicities and characteristics of vari-ous substances. Water pollutants included BOD (bio-logical oxygen demand) in kilograms per day, andTSS (total suspended solids) in kilograms per day.Separate ratios of effluents to shipments were com-puted for each measure, but the results reported laterare based on a combined ratio that added BOD andTSS together, then divided the sum by the establish-ment's total shipments. Air pollution was measuredby the ratio of particulate emissions to total ship-ments. Although the same firms are not representedin all comparisons, these measures give a fairly com-prehensive picture of the environmental performanceof manufacturing establishments.

Specialized (5-digit Standard Industrial Classifica-tion code) industrial sectors that produce a relativelynarrow range of homogeneous products were se-lected for study. The SIC classifies industries evenmore narrowly (7-digit or 9-digit codes) but furtherdisaggregation would have limited the sample sizesdrastically. Even at the 5-digit level of classification,had industrial sectors that include firms producing awide range of products ("miscellaneous inorganicchemicals," for example) been selected, comparisonsof establishments making very different products withquite different materials and technologies would havebeen inevitable. Comparing the emissions per unit ofshipment among such establishments would havebeen no more meaningful than comparing applesand oranges. Confining the investigation to special-ized sectors with relatively homogeneous productlines reduced one possible source of spurious varia-tion in the findings. Sectors with homogeneous prod-

uct lines were chosen on the basis of several addi-tional criteria: first, to represent a wide range of man-ufacturing industries; second, to include sectors thathave significant environmental impacts and incur rel-atively large environmental control costs; and third,to include sectors with sufficient numbers of estab-lishments in the matched database to allow meaning-ful comparisons across plants.

Environmental performance varies remarkablyeven among establishments in narrowly defined in-dustrial lines, such as makers of printed circuitboardsor ready-mix concrete. A common measure of vari-ability is the coefficient of variation, which is the ratioof the standard deviation of a variable to its mean.Across all the industries examined in this study, themedian value of this measure was 1.7: the standarddeviation of environmental performance among es-tablishments in the same industry was typically sev-enty percent larger than the average of the individualestablishment's emissions-to-output ratios.

Two measures of profitability were constructedfrom the LRD data. The first is the gross operatingmargin, defined as the difference between the totalvalue of shipments and total operating costs (includ-ing labor, materials, energy, rental, and contractcosts), expressed as a fraction of the total value ofshipments. The second is the net return as a fractionof the end-of-year book value of fixed capital. Thenet return is simply the difference between the totalvalue of shipments and total operating costs, minusannual depreciation. Neither of the two is a perfectmeasure of profitability. Gross operating margin,which excludes capital costs, would be higher incapital-intensive firms than in less capital-intensivefirms, even if the two were equally profitable. Thenet return on book value reflects a user charge onowned capital, but such factors as taxes and inflationwould make this measure diverge from a true returnon invested capital. However, comparing these mea-sures only across establishments within narrowly de-fined industrial segments minimizes these distortions.Establishments within a single narrowly-defined in-dustry are likely to be similar in capital-intensity andto face similar inflationary trends and tax regimes.Table 4 summarizes the various measures of envi-ronmental and economic performance used in thisanalysis.

Table 4. Measures of Environmental and Market Performance.

A. Measures of Environmental Performance

l.a Total Toxic Releases per dollar of shipments

2.a Total Airborne Particulate Emissions per dollarof shipments

3.a Biological Oxygen Demand (BOD) plus TotalSuspended Solids (TSS) per dollar of shipments

B. Measures of Profitability

l.b Gross Operating Margin: (Total value ofshipments less total operating costs) dividedby total value of shipments

2.b Net Return on Book Value: (Total value ofshipments less total operating costs lessannual depreciation) divided by book valueof invested capital

What do the results of this exercise show? Corre-lations between environmental performance wouldbe positive under the standard hypothesis, negativeunder the competing "Porter hypothesis." The de-tailed findings are laid out in Appendix Tables IA-ICbut can be comprehended more readily by looking atFigures 1A-1C. In each of these graphs, the two mea-sures of profitability are represented on the axes—gross margin on the horizontal and net return on thevertical axis. Each point represents the correlation co-efficient between environmental performance and thetwo measures of profitability. Thus, if environmentalperformance in a particular industry is positively cor-related with both measures of profitability, the indus-try will be represented by a point in the upper rightquadrant of the graph. The further away from the ori-gin of the graph in both dimensions, the closer thecorrelations. If the industry's environmental perfor-mance is negatively correlated with both measures ofprofitability, it will be represented by a point in thelower left quadrant. If the correlation with gross mar-gin is positive, but that with net return on capital isnegative, the point will fall in the lower right; if thecorrelations are reversed, the point will be in theupper left. In general, there is no tendency for supe-rior profitability to be correlated with greater emis-sions per unit of output.

It should be emphasized that each "point" in agraph summarizes the association between environ-

mental and market performance across many differ-ent establishments within an industrial sector. Thenumber of establishments included in each industryranges from a minimum of 10 to a maximum of 429,as reported in the Appendix tables. For reasons ofconfidentiality, data on individual establishments arenot divulged, so reported findings combine data oneach establishment in the sample into an aggregatecorrelation coefficient. However, readers should beaware that the analysis covered thousands of manu-facturing plants. Data on toxic releases and profitabil-ity were combined for 1,936 individual establish-ments, for example.

If the data from the individual plants are samplesof the establishments in their industries, could thecorrelation coefficients have arisen by chance if thetrue correlation between environmental and eco-nomic performance were actually zero? Because thenumber of establishments for which data werematched differed in the various industries, signifi-cance tests were calculated for each industry's corre-lation coefficients. The results are depicted graphi-cally by using a square to indicate a pair ofcorrelations of which at least one was highly unlikelyto have arisen solely by chancea and a circle to de-

a Formally, a "significant" correlation was defined as onethat would not have arisen by chance more than one timein twenty if the true correlation with environmental perfor-mance were zero.

B_

Figure 1.A. Simple Correlations: Toxic EmissionsSimple Correlation Coefficients between Total Toxic Releases per Dollar of Shipments and (a) Gross OperatingMargin and (b) Net Return on Book Value of Invested Capital

1

0.8

0.6

0.4

0.2

0

-0.2

-0.4

-0.6

-0.8

-1

NOM/BVC

GOM/TVS

-1 -0.8 -0.6 -0.4 -0.2 0 0.2 0.4 0.6 0.8 1

Notes: Toxic emissions = total TRI releases per dollar of shipmentsNOM/BVC = Net Operating Margin as a proportion of the Book Value of CapitalGOM/TVS = Gross Operating Margin as a proportion of the Total Value of Shipments• Significantly different from zero with 5 percent probability# Not significantly different from zero

pict a pair of correlations neither of which differedsignificantly from zero.

Looking first at the correlations between manu-facturing plants' profitability and their toxic emissions(relative to shipments)—which draw on the largestdatabase for making comparisons—one finds thatpoints are not clustered in the upper right: there isno tendency for higher toxic emissions to be associ-ated with higher profitability. For all but one of theindustries in which such an association exists, thecorrelation is weak and could have arisen by chance.There are twice as many industries for which the cor-

Across the entire range of industries,correlations between profitability andthe intensity of toxic releases are weak.

relation between profitability and toxic emissions isnegative, implying that plants in those industries with

Figure 1.B. Simple Correlations: Water-borne EmissionsSimple Correlation Coefficients between Water Pollution per Dollar of Sales and (a) Gross Operating Marginand (b) Net Return on Book Value of Invested Capital

1

0.8

0.6

0.4•

0.2

-0.2 •

-0.4

-0.6

-0.8

NOM/BVC

- -

•

- -

•

•

- -

GOM/TVS

-0.8 -0.6 -0.4 -0.2 0.2 0.4 0.6 0.8

Notes: Water-borne emissions = BOD plus total suspended solids per dollar of shipmentsNOM/BVC = Net Operating Margin as a proportion of the Book Value of CapitalGOM/TVS = Gross Operating Margin as a proportion of the Total Value of Shipments• Significantly different from zero with 5 percent probability• Not significantly different from zero

superior environmental performance (low toxic emis-sions relative to shipments) are more profitable. Inseveral of these industries, including sectors such asmetalplating and industrial detergents that have sig-nificant environmental control problems, the associa-tion is strong. However, across the entire range of in-dustries, correlations between profitability and theintensity of toxic releases are weak.

The same general conclusions are borne out bycorrelations between waterborne emissions and air-borne particulate emissions, respectively, and mea-sures of profitability. Figures IB and 1C reveal no

tendency for profitability to be positively correlatedwith emissions-intensity for either conventional waterpollutants, such as BOD and total suspended solids,or for airborne particulate emissions. It is at leastequally likely for plants with superior environmentalperformance to be more profitable. Overall, the asso-ciations are weak: other factors are determining theeconomic performance of individual manufacturingestablishments.

"Yes," skeptics will say, "and it is precisely theseother factors that mask the true effects of environ-mental control costs on profits. That's why simple

B-

Figure 1.C. Simple Correlations: Air Particulate EmissionsSimple Correlation Coefficients between Air Particulate Emissions per Dollar of Sales and (a) Gross OperatingMargin and (b) Net Return on Book Value of Invested Capital

0.8

0.6

0.4

0.2

0

-0.2

-0.4

-0.6

-0.8

-1

NOM/BVC

GOM/TVS

-0.8 -0.6 -0.4 -0.2 0.2 0.4 0.6 0.!

Notes: Air particulate emissions = particulate emissions per dollar of shipmentsNOM/BVC = Net Operating Margin as a proportion of the Book Value of CapitalGOM/TVS = Gross Operating Margin as a proportion of the Total Value of Shipments• Significantly different from zero with 5 percent probability• Not significantly different from zero

correlations don't reveal much." Older plants, for ex-ample, are probably both dirtier and have higher pro-duction costs, because they embody outmodedprocess technologies, because they are hard to retro-fit with efficient pollution control equipment, and be-cause they require a lot of maintenance to preventleaks and emissions. Larger plants probably achieveeconomies of scale, both in producing their primaryoutputs and in treating or controlling effluents. Differ-ences like these, which may affect both profitabilityand pollution-intensity among plants in an industry,could create spurious correlations between economic

and environmental performance or mask whatevertrue associations may exist.

To guard against this possibility, the analysis wasrepeated using partial correlation analysis instead ofsimple correlations. Partial correlation analysis is atechnique for exploring the association between twovariables while removing the influence of extraneousvariables that influence both. In effect, it is a way ofcontrolling statistically for the effects of extraneousvariables, or "holding them constant." Partial correla-tion coefficients were calculated for the same indus-tries and establishments as before, controlling for a)

the age of the plant, b) the scale of production, mea-sured by the total value of shipments, and c) theamount of recent investment in plant and equipment,as a fraction of book value of capital.

The results are presented graphically in Figures2A-2C and in Appendix Tables IA-1C. What is strikingis how small the overall differences are between thesimple and partial correlations. Even when the ageand scale of a plant and the amount of recent invest-ment in plant and equipment are taken into account,there is no overall tendency for plants with superiorenvironmental performance to be less profitable.

Even when the age and scale of a plantand the amount of recent investment inplant and equipment are taken into ac-count, there is no overall tendency forplants with superior environmentalperformance to be less profitable.

Figure 2.A. Partial Correlations: Toxic EmissionsPartial Correlation Coefficient Controllings for (a) Scale, (b) Age of Plant, and (c) Amount of RecentEquipment Investment

1

0.8

0.6

0.4

0.2

0

-0.2

-0.4

-0.6

-0.8

NOM/BVC

• *

GOM/TVS

-0.8 -0.6 -0.4 -0.2 0.2 0.4 0.6 0.8

Notes: Toxic emissions = total TRI releases per dollar of shipmentsNOM/BVC = Net Operating Margin as a proportion of the Book Value of Capital

GOM/TVS = Gross Operating Margin as a proportion of the Total Value of Shipments

• Significantly different from zero with 5 percent probability• Not significantly different from zero

H -

Across the thousands of plants in the sample, it is atleast equally likely, and perhaps somewhat morelikely, for plants with lower emissions—relative toproduction—to achieve higher operating margins andreturns on invested capital. By and large, however,the associations are weak. In the last analysis, otherfactors influence profitability more strongly.

This body of evidence bears directly on the rela-tionship between environmental performance andcompetitiveness. The results fully support earlier find-ings based on international trade and investmentflows. There is simply no evidence that superior en-

vironmental performance puts firms at a market dis-advantage or adversely affects market performance.The implications are important for both public andprivate policy. Some important ramifications are high-lighted below:

• Environmental regulations need not and shouldnot be weakened or relaxed so as to under-mine their environmental objectives out offear of adverse effects on industries' marketperformance.

• In the international arena, there is no need forcountervailing tariffs, anti-dumping duties or

Figure 2.B. Partial Correlations: Water-borne EmissionsPartial Correlation Coefficient Controllings for (a) Scale, (b) Age of Plant, and (c) Amount of RecentEquipment Investment

1

0.8

0.6

0.4

0.2

0

-0.2

-0.4

-0.6

-0.8

-1

NOM/BVC

GOM/TVS

-0.8 -0.6 -0.4 -0.2 0 0.2 0.4 0.6 0.8 1

Notes: Water-borne emissions = BOD plus total suspended solids per dollar of shipmentsNOM/BVC = Net Operating Margin as a proportion of the Book Value of CapitalGOM/TVS = Gross Operating Margin as a proportion of the Total Value of Shipments• Significantly different from zero with 5 percent probability• Not significantly different from zero

Figure 2.C. Partial Correlations: Air Particulate EmissionsPartial Correlation Control Settings for (a) Scale, (b) Age of Plant, and (c) Amount of Recent EquipmentInvestment

1

0.8

0.6

0.4

0.2

0

-0.2

-0.4

-0.6

-0.8

NOM/BVC

GOM/TVS

-0.E -0.6 -0.4 -0.2 0 0.2 0.4 0.6 0.8

Notes: Air particulate emissions = particulate emissions per dollar of shipmentsNOM/BVC = Net Operating Margin as a proportion of the Book Value of CapitalGOM/TVS = Gross Operating Margin as a proportion of the Total Value of Shipments• Significantly different from zero with 5 percent probability• Not significantly different from zero

HI

other trade penalties directed at imports fromcountries with weaker environmental standardsthan our own, at least when the resulting envi-ronmental impacts are confined within theirown borders. Nor must environmental stan-dards necessarily be "harmonized" to preventcompetitive dislocations.Within the United States, since superior envi-ronmental performance makes firms no less

profitable, institutional and fiduciary investorsshould not expect to earn lower portfolio re-turns if they invest in the stocks of firms withsuperior environmental performance withineach industry. Similarly, environmentallyscreened stock portfolios that avoid the worstperforming firms within each industry shouldnot expect to achieve lower average returns asa result.

IV. JOBS AND THE ENVIRONMENT

Similar conceptual confusion and factual misun-derstandings derail the debate over the effects of en-vironmental regulations on employment. Some main-tain that regulations destroy jobs in the regulatedindustries; others retort that regulations increase em-ployment in environmentally benign industries.40 Ofcourse, both groups are correct, narrowly speaking.Virtually any expenditure, however foolish or unpro-ductive, will generate employment. That's why thelast line of defense for the most egregious pork-bar-rel spending is that it creates a certain number ofjobs. The Corps of Engineers generated employmentwhen draining our nation's wetlands; it will createjobs again when restoring stream flows and undoingthe damage its previous projects have done. That'sclose to digging holes in the ground and filling themin again, but it creates jobs.

Shifting resources from producing chemicals tocleaning up waste sites costs some jobs in chemicalsindustries and creates some jobs in remediation firms.The important question is not whether employmentincreases or decreases on balance, but whethercleaning up waste sites is worth the costs. The rele-vant costs are all the resources, not just the laborinput, that could be used to generate other neededgoods and services.

To focus this more sharply, think about the ex-penditures people make privately to improve theirown environments. If people buy home water filtersto improve the quality of their drinking water, no-body worries about destroying jobs. Instead, it is as-sumed that jobs will be created in the water filter in-dustry. Of course, the people who buy water filtersmust spend less of their incomes on other things—movie tickets, for example—so employment will alsogo down in the movie industry. Who knows (orcares) whether employment per dollar of sales isgreater in the water filter industry or in the movie in-dustry? People assume that they can buy the goodsand services they prefer and markets—includinglabor markets—will adjust to this allocation ofresources.

The important question is not whetheremployment increases or decreases onbalance, but whether cleaning up wastesites is worth the costs.

Now suppose that people sensibly conclude itwould be much cheaper if one giant water filter werepurchased for the whole community, instead of eachhousehold buying a mini-filter for itself, and the townvotes to tax itself to pay for a water treatment plant.Again, employment rises in the giant water filter in-dustry. But, since people have to cut down on otherspending to pay the additional taxes, employmentgoes down in industries producing the other goodsand services that people had been spending theirmoney on.a How it balances out is anybody's guess,but that's not the relevant question. What's upper-most is whether the community thinks clean water isworth the cost.

But suppose the government requires the town'sfactory to install the water filter on its exhaust pipebefore it dumps its wastewater in the community'swater supply. Won't factory workers lose their jobs?Quite possibly. When the factory raises its prices tocover the costs of the filter, employment is likely tosuffer. But, as before, employment goes up in thewater filter industry. On balance, industrial employ-ment might rise or fall, but the real issue is whetherthe voters think that protecting the water supply isworth the cost.

a Of course, if the community chooses to raise taxes onpayrolls and incomes, thereby discouraging some of itscitizens from working as hard, employment will go downmore than it would if the community taxed other activi-ties—downtown parking or trash disposal, for example.More on that later.

Petroleum refining, chemicals manufac-turing, pulp and paper, and primarymetals—the environmentally sensitiveindustries in which pollution abatementcosts represent a relatively large frac-tion of output value—are all among theindustries with the fewest employeesper million dollars in shipments.

In fact, pollution-control expenditures purchasegoods and services from a broad cross-section ofAmerican industry: energy, construction, chemicals,machinery and transport equipment, rubber, stoneand glass, instruments, engineering and other servicesindustries, and more.41 These expenditures also paythe salaries of a wide range of skilled and semi-skilledworkers. Consequently, the direct employment gener-ated per dollar of expenditure on pollution control isquite similar to the average employment generated bya dollar of sales across all of American industry. A re-cent EPA study calculated that pollution-control ex-penditures in 1991 directly generated about 744,000jobs. 2 Compared to the $107.9 billion dollars spenton pollution control in that year, this implies a directemployment content of 6.9 jobs per million dollars ofexpenditure (in 1986 dollars).43 This estimate is inthe same range as others: the OECD estimated 10 jobsper $1,000,000 in expenditures in 1990; the Environ-mental Business Journal estimated 7.6 jobs per$l,000,000.44 Numbers vary slightly depending onhow environmental expenditures were defined andhow estimates were made. The main point is that thepollution-control industry, however defined, is aboutas labor-intensive as U.S. industry as a whole. In 1991,18.45 million employees in manufacturing industriesproduced $2.82 trillion in total shipments, for a ratioof 6.5 jobs per million dollars in sales. Consequently,one would not expect that a shift in expenditures to-ward pollution control, even at the expense of manu-facturing output, would have a significant direct im-

pact on employment. Gains and losses would roughlybalance out.

In fact, this conclusion is probably unduly pes-simistic. The industries that generate the most pollu-tion and incur the highest pollution-control expendi-tures are by no means the most labor-intensive.Figure 3 demonstrates just the opposite: petroleumrefining, chemicals manufacturing, pulp and paper,and primary metals—the environmentally sensitiveindustries in which pollution abatement costs repre-sent a relatively large fraction of output value—are allamong the industries with the fewest employees permillion dollars in shipments. These sectors are notonly less labor-intensive than manufacturing in gen-eral, they are also less labor-intensive than the pollu-tion-control industry. It follows that increasing expen-ditures on pollution control is unlikely to reduceoverall employment in the short run. In the longerterm, any second-order effects would almost certainlybe undetectable among the more powerful macro-economic secular and cyclical forces that drive unem-ployment up and down.

The real issue is not the environment vs. jobs.The issue is what we want our economy to produce.If we want it to produce a clean environment alongwith other goods and services, the industries thatcontribute to a clean environment will have higheroutput and employment; those that do damage to theenvironment will have less. While jobs in particularindustries may rise or fall, total employment will notbe systematically affected.

What we want the economy to produce is con-tinually changing, and industries expand and contractas a result. As sales of personal computers haverisen, sales of portable typewriters have declined.Jobs for typewriter repairmen have disappearedwhile opportunities for computer programmers havemultiplied. No doubt this shift has created hardshipsfor some households, but no politician or lobbyisthas said, "The U.S. economy can't afford to have per-sonal computers because it will destroy jobs in thetypewriter industry." Yet, they routinely claim that wecan't afford clean air because it will destroy jobs inthe coalmining or some other industry.

Those who complain about the effect of environ-mental controls on employment are usually thinkingabout particular jobs, in their own firms, industries

Figure 3. Employment and Pollution-Control Expenditures by Major Industrial Sector

Apparel

Leather

Furniture

Textile mill products

Printing/Publishing

Lumber

Fabricated Metal

Electronic/Electric Equipment

Stone, Clay, Glass

Industrial machinery

Instruments

Manufacturing, average

Transportation

Paper

Primary metal Industries

Food

Chemicals

Tobacco

Petroleum

Rubber

• 1990 Employment per $ Million Gross Outputn Abatement Costs as Percentage of Output Value

or communities—and well they should. Losing a jobhurts the individuals affected and their families and,if clustered in particular localities, those communitiesas well.45 The role of public policy, however, is notto guarantee particular jobs but to ease the transitionfrom declining to expanding industries—through un-employment compensation and retraining programs,by giving people opportunities to acquire the skillsand resources they need for greater occupational mo-bility, by macroeconomic policies that maintain highaggregate employment, and through measures thatmoderate abrupt economic shocks.

In a market economy, people don't have entitle-ments to particular jobs. This is obvious in the private

sector: each year many companies downsize theirworkforces in response to declining demand or tech-nological changes; other companies expand. Shiftslike these in annual employment dwarf those re-motely attributable to environmental policy. If theprivate sector does not give workers entitlements toparticular jobs, neither can the public sector. How-ever, workers can expect government to maintainhigh aggregate employment and to help them findother jobs. Americans overwhelmingly agree that theeconomy should produce a clean environment, evenif it means producing a little less of something else. Aspurious fear that a clean environment means higherunemployment should not stand in the way.

V. THE REAL ISSUE

The real issue with environmental spending, aswith all spending, is not jobs or "competitiveness,"but whether we're getting good value for our money.Are the resources devoted to environmental protec-tion buying significant improvements in environmen-tal quality, or are they being frittered away with littleto show? People want unpolluted air and water, safeand healthy neighborhoods and workplaces, and un-degraded natural resources. They've demonstratedover and over again that they're willing to pay forthem. Two decades ago, for example, the automakersclaimed vehemently that if buyers were forced to paythe two or three hundred dollars it would cost to putcatalytic converters on new cars, consumers wouldbalk and the industry would collapse. Since then,have buyers been heard complaining about the costof catalytic converters? In fact, after improving thequality of its manufacturing systems and its cars,which are now far cleaner and more fuel-efficientthan they were 20 years ago, Detroit just had its mostprofitable year in postwar history.

The real issue with environmentalspending, as with all spending, is notjobs or "competitiveness," but whetherwe're getting good value for our money.

¥Where people can buy environmental quality for

themselves, they're doing so in increasing numbers.The agricultural chemicals industries has claimed thatif the use of pesticides and chemical fertilizers areregulated more strictly, food prices will rise beyondthe means of middle-class American families. Yet, theorganic foods industry is getting premium prices andexpanding at 25 to 30 percent per year as consumersflock to buy chemical-free foods that they perceive tobe less risky. As a result, mainstream supermarkets

are now stocking organic foods, and more and morefarmers are finding that they can reduce their chemi-cal use and increase their profits.

Although such market trends as these demon-strate that people value a clean, safe environment,their willingness to pay is more often masked by theneed to decide collectively on the quality of the envi-ronment they share. Not everybody can afford a fam-ily estate on the coast of Maine as an escape fromurban summer smog. Most of us have to breathe theair in the cities and suburbs where we live, and it'smore or less the same quality air for the entire urbanpopulation. It's virtually useless for one person tospend the extra money for a cleaner car unless othersdo the same. Without that assurance, people tend tounderspend on environmental quality. That's themain reason why governments have to get involvedin the collective decision-making process.

For similar reasons, not all the benefits from acleaner environment show up in the record of markettransactions. Sure, one of the reasons that summerplaces on the Maine seacoast are too expensive formost of us is that people put a very high value onunspoiled natural resources and a clean environment.However, the discomfort people feel breathing pol-luted air in the city all summer doesn't necessarily re-sult in any market transaction, and so doesn't gener-ate income for anybody. Of course, if things get badenough, people get sick and miss work, expenditureson healthcare increase, and perhaps a few people dieprematurely. Ironically, some of these effects, such asincreased healthcare expenditures, are counted as in-creases in income and output. A good part of the ap-parent economic cost of environmental protectionstems from the incomplete and misleading account-ing methods used to measure income and productiv-ity. If we're getting good value from expenditures onenvironmental protection then (properly measured)income and productivity will rise as a result.

The nub of the issue is that we're not getting asmuch as we should for our expenditures on environ-mental protection, now exceeding two percent of

A good part of the apparent economiccost of environmental protection stemsfrom the incomplete and misleading ac-counting methods used to measure in-come and productivity.

total GDP. This is not news. The inefficiencies in en-vironmental, health, and safety regulations have beenamply documented and analyzed over the past twodecades or more.46 The problems have been exten-sively studied and are well-known.

Regulations aimed at reducing illness and death,whether in the environment or in the workplace,often require expenditures that buy very little incre-mental risk reduction. How much society shouldspend to reduce such risks is a difficult question,47

but there's clearly something wrong when oneregulation buys a reduction in the same risk at a costseveral orders of magnitude greater than another reg-ulation does. This anomaly has been documentedtime and again.48 Why not concentrate efforts moreheavily on the actions that reduce risks at lower costand, in that way, achieve much greater overall im-provements in health and safety for the same totalexpenditure?

For several reasons, inefficiencies of this kindpersist. Many environmental statutes prescribe thatregulated firms install the best available control tech-nology or achieve the maximum achievable emis-sions reduction within the bounds of economic feasi-bility, irrespective of the risk reductions that result.Among other things, this often implies that all emis-sions sources are required to install the same kind ofpollution control equipment, regardless of the num-bers of people exposed or the extent of their expo-sure. It sometimes implies that industries with "deeppockets" or in a better position to pass the costsalong to consumers are made to adopt tighter con-trols, whether or not this extra expenditure is justifiedby additional improvements in health and safety.

Moreover, regulatory decisions are not basedconsistently on an assessment of risks or on a com-

parison of the incremental risk reduction tighter stan-dards would achieve and their costs. Some statutesand regulations require such a comparison; othersare based on the goal of eliminating risks completely,whatever the cost. Even when risk assessment is partof the regulatory decision-making process, methodsand assumptions vary widely. To err on the side ofconservatism, regulators sometimes make wildly un-realistic assumptions about people's likely behaviorand exposure (children living near hazardous wastesites eating a steady diet of dirt day-in and day-out,for example). Sometimes the risk assessment is car-ried out to support a pre-determined regulatory deci-sion, rather than the reverse.49 Because these prob-lems persist, some environmental regulations buyrelatively little improvement in health and safety, cer-tainly much less than could be achieved for the sameexpenditure.

Some environmental regulations buyrelatively little improvement in healthand safety, certainly much less thancould be achieved for the same expen-diture.

Pollution control efforts also require muchgreater expenditures than necessary to improve envi-ronmental quality. Environmental regulations grewup piecemeal, and often imply that waste streams areshunted among air, land and water: to keep pollu-tants out of the atmosphere, exhaust gases are"scrubbed" and the resulting sludges are dumped onland; to avoid putting wastes into landfills, they're in-cinerated, and the exhaust gases go up the chimney;and so on. Industry and regulators are still trying toestablish the integrated, cross-media approach toresiduals management and waste reduction that engi-neers and economists advocated decades ago.50 Suchan approach could substantially reduce the costs andenvironmental impacts of waste management.

One-size-fits-all regulations that necessitate uni-form technological solutions to pollution problems

within an industry drastically raise the costs of com-pliance. They're usually end-of-pipe solutions thatdiscourage innovative product redesign and processchanges that could save money in the long run. Theyignore important differences among firms in thescale, remaining lifetime, location, and design of fa-cilities, though such differences often make otherpollution-control options more cost-effective. As a re-sult, the same amount of expenditure buys drasticallyvarying reductions in emissions in different indus-tries, among emissions sources within industries, andeven within the same facilities. Instead of startingwith the least expensive ways of reducing emissionsand moving gradually up the cost curve until theemission reduction target is achieved, this regulatoryapproach sometimes requires a great deal of expen-diture for relatively meager results. A recent Re-sources for the Future report summarized numerousstudies of the cost-effectiveness of regulation underthe Clean Air Act: most studies show that actual ex-penditures are several times those that would be re-quired to achieve the same goals under a least-costapproach.51

These findings were recently borne out by a de-tailed study of a single facility, the Yorktown refinery.EPA and the refinery's owner, AMOCO, jointly inven-toried all sources of volatile hydrocarbon emissionsand estimated the costs of controlling them. Thestriking conclusion they reached was that 90 percentof the emissions reductions achievable by the applic-able technology-based regulations could be obtainedat about 25 percent of the cost by putting tighter con-trols on emission points that could be eliminatedcheaply, while foregoing some controls that werevery expensive and accomplished little.52 Except thatit was studied carefully, there's nothing peculiarabout the Yorktown refinery; the same potential sav-ings are very likely available at other large industrialfacilities as well. Add to them the potential savingsfrom eliminating large cost discrepancies betweenone facility and another and it's easy to understandthe findings reported in the RFF study.

An important example of the inefficiency thatdrives up the costs of controlling emissions is the"new source bias" of regulations—the tendency toapply much more stringent requirements to new cars,factories, and powerplants than to those already in

use. To make progress with such a strategy, new in-vestments must be held to high (and often costly)standards, while the much larger universe of operat-ing vehicles or factories are allowed to continue emit-ting at much greater rates, even though in many casestheir performance could be improved cheaply. Oldcars that need a tune-up are responsible for a dispro-portionately large share of automobile emissions, forexample, but federal regulations have largely ignoredthem while tightening standards on new model cars,which are much cleaner to start with. This "newsource bias" has a second pernicious effect: by mak-ing new models or plants more expensive relative tothe cost of continuing to operate older equipment, itdiscourages investment, slows down the replacementrate and keeps the older, dirtier and less efficientequipment in use longer.53 This is counter-productiveeconomically as well as environmentally.

Not only are the legal and transactionalcosts of our adversarial approachhigh, the delays and uncertainties haveserious business and financial reper-cussions.

The regulatory process in the United States isalso exceptionally adversarial and litigious, and there-fore cumbersome and time-consuming. Typically, inorder to be able to defend whatever rule they decideon, EPA must compile an elaborate public record ofevidence, comment, and reaction to comments fromall interested parties. Nonetheless, whatever decisionthis process leads them to, EPA is likely to be suedby whichever party feels that its interests have beensacrificed. Quite naturally, this sometimes creates areluctance within the Agency to act, but hesitation isalso likely to generate a lawsuit. Negotiated decisionsand consensus-building on environmental policy arerare in the U.S. compared to other industrial coun-tries.54 Not only are the legal and transactional costsof our adversarial approach high, the delays and un-certainties have serious business and financial reper-

cussions. Most notorious are the problems that havearisen under the Superfund program, intended toclean up dangerous hazardous waste sites. Almost aquarter of all the billions of dollars in expendituresso far have been to pay lawyers' bills, as the compa-nies involved, their bankers, and insurers havebrawled to avoid picking up the tab for past dump-ing. Meanwhile, few sites have actually beenrestored.55

Problems like these should be the focus of con-cern about environmental regulation. The issue is notthat protecting the environment has been creatingunemployment or trade deficits; such claims are erro-neous and won't stand scrutiny. Rather, the issue isthat the United States could be getting much more forthe large sums of money we spend to protect the en-vironment, much more real progress in reducing risksand preventing natural resource degradation.

Rj

VI. GETTING BETTER RESULTS FROM ENVIRONMENTAL SPENDING

How to get better results from spending on envi-ronmental protection is a question that has beenstudied almost to death.56 The ratio of research topractical results is probably as high as in any area ofpublic policy. It is a question that has a sensible an-swer. The accumulated knowledge of good ways toimprove the efficiency of environmental regulationshas been synthesized, summarized, and presentedtime after time—by academic experts, agency re-views, and high-level commissions. The resultingagenda is bi-partisan and establishes common groundamong business and environmental groups. It wasembraced in the report of the National PerformanceReview led by Vice-President Gore,57 and also in theprivately organized report of the National Commis-sion on the Environment, which included former EPAadministrators from one Democratic and three Re-publican administrations.58 Many parts of the agendaare supported by leading business groups59 and byprominent environmental groups.60 Politicians haveheard the prescriptions repeatedly from crediblesources on all sides of environmental issues.61

There is no need to take a wrecking ball to theenvironmental protection system that has been con-structed by Democratic and Republican administra-tions over the past twenty-five years. The Americanpeople don't want that; an overwhelming majority ofvoters think that environmental protection should bemaintained or strengthened. Efforts to undermine en-vironmental protection will just provoke the kind ofbitter political and legal fights that disfigured publicpolicy in the early 1980s. There's a better way.

If Congress and the Clinton Administration wishto work together to improve environmental qualityand reduce the economic cost of doing so, the pathis clear. The potential economic dividends add up toscores of billions of dollars in annual savings to con-sumers, businesses, and taxpayers—resources thatcan be used to generate new investments and higherliving standards. At the same time, environmentalquality can be improved. However, this is not a freelunch. The price is political—challenging some en-

trenched interests in the private sector and some es-tablished administrative routines in the public sector.

For starters, the confusion over the effect of envi-ronmental protection on the economy could begreatly reduced if we corrected the way we keepscore. In a basketball game, if a player scored a goalbut the scorekeeper sometimes decided not to countit and at other times added it to the other team'stotal, how would anybody know who was winning?That's exactly how our current measures of incomeand productivity growth deal with environmental im-provements and damages.2 Modest efforts havebegun in the Department of Commerce and EPA torevise these accounting methodologies. These areworthy of support, because they promise a great dealof illumination at a very modest cost.

Those who believe that markets workbetter than bureaucracies should fullysupport the use of environmental policyinstruments that build the cost of envi-ronmental degradation into the pricestructure.

The most important change, however, is to makewider and more effective use of market incentives inenvironmental policy. Those who believe that mar-kets work better than bureaucracies should fully sup-port the use of environmental policy instruments thatbuild the cost of environmental degradation into theprice structure. Doing so instead of relying predomi-nantly on command-and-control regulatory ap-proaches raises productivity in at least three ways.First, firms can adopt the environmental controls thatare cost-effective for them, rather than following theprescribed technological solutions imbedded in regu-

lations. Second, the economic burdens of environ-mental controls can be redistributed among firms inways that induce those who can clean up relativelyinexpensively to do more. Third, since in a marketeconomy one firm's cost is another firm's opportu-nity, the profit motive can be enlisted more forcefullyto develop new and better methods to deal with en-vironmental problems.

One of the best ways to make an exist-ing regulatory system more flexible andcost-effective is to promote trading inemissions permits or other kinds ofentitlements to the use of naturalresources.

A wide range of market incentive-based policieshas been studied and tried. 3 Each works betterunder certain circumstances than others; no one in-strument is best for all purposes. However, one ofthe best ways to make an existing regulatory systemmore flexible and cost-effective is to promote tradingin emissions permits or other kinds of entitlements tothe use of natural resources.64 This approach allowsthe regulatory authorities to maintain or tighten thelimits on overall emissions, while giving regulatedparties a chance to negotiate arrangements wherebycheap emissions reductions substitute for expensiveones. Such substitutions can typically reduce aggre-gate pollution control costs by thirty percent or more.The flexibility that trading options provides for pollu-tion sources with high compliance costs also reducespressures on the government to weaken standards orissue exemptions.

The Environmental Defense Fund has pioneeredin promoting the use of trading approaches, showingthat they can be useful not only in controlling emis-sions but also in reallocating water supplies from irri-gation to higher-valued municipal and industrial uses,in relocating real estate development from ecologi-cally sensitive lands to more appropriate sites, and in

a variety of other situations. Transferable catch quo-tas, for example, have proven themselves far moreeffective in preventing overfishing and over-invest-ment in fishing gear than the disastrously ineffectualapproaches to fisheries management the UnitedStates has been pursuing.

Some progress has been made in adopting trad-able permits as a policy instrument. One accomplish-ment of the 1990 Clean Air Act Amendments was toallow sources required to reduce sulphur emissionsby 10 million tons in all to trade the reduction re-quirements among themselves, an approach that al-ready had been used successfully to lower the costsof eliminating lead from gasoline. Sulphur trading isexpected to cut several billion dollars off the cost ofthat regulation.

However, progress has been far too slow andlimited. 5 Congress and EPA adopted the so-called"bubble" policy decades ago, which allows firms totrade off tighter controls on some emissions sourcesfor laxer controls on others within the same facility,subject to adequate monitoring, verification, andprogress toward emissions reduction.66 The bubblepolicy was designed precisely to deal with situationslike the Yorktown refinery's. Yet, more than fifteenyears later, state and federal administrators are leeryof this approach and hesitate to make use of it, sacri-ficing huge potential savings. New approaches toemissions trading have overcome some of the designflaws in the early systems—by requiring sellers to im-plement and document their emissions reductionsbelow the regulatory requirements in advance, forexample.67 Businessmen, environmental groups,politicians, and administrators can work together todesign and implement workable trading systems thatpromote environmental improvement at significantsavings.

In some environmental protection programs,however, trading systems are likely to work less wellthan other approaches, such as charging fees to deterenvironmentally damaging activities. For example,when the number of people and firms involved increating the problem is very large and heteroge-neous, creating a parallel market for permit tradingmay be less feasible than using the price mechanismdirectly. In such circumstances, most of the pollutioncomes from sources without regulatory permits. For

m-

When the number of people and firmsinvolved in creating the problem is verylarge and heterogeneous, creating aparallel market for permit trading maybe less feasible than using the pricemechanism directly.

example, automobiles are the source of increasinglysevere environmental problems as the number of ve-hicles on the road and the average number of milesthey travel both increase. It is unlikely that driverscan negotiate trade-offs among themselves even withcellular phones in their cars. Various price mecha-nisms are available to manage these problems. Rais-ing parking charges and eliminating favorable taxtreatment of employer-provided parking can discour-age the solo drive to work. Varying annual vehicleregistration fees according to tested emissions canhelp get the "clunkers" that account for a dispropor-tionately large fraction of emissions off the road."Road pricing," which implies charging tolls that arehigher during rush hours than in off-peak periods,can reduce urban congestion. Raising gas taxes canencourage people to purchase more energy-efficientvehicles and to drive less.

Using fees instead of command-and-control regu-lations to control environmentally damaging behaviorallows the price mechanism to provide incentives forefficiency. Everybody who faces the same fee, for ex-ample, finds it worthwhile to incur more-or-less thesame incremental environmental control cost. Thoseunable to clean up at reasonable cost can temporarilypay the fee instead of spending inordinate amountsto comply with a command-and-control regulation(or taking the matter to court as the less expensiveoption). Fees provide incentives for continual envi-ronmental quality improvement, in contrast to regula-tions, which create a status quo once compliance isachieved.

In the relatively few instances in which the fed-eral government has enacted specific environmental

charges, such as the excise tax on ozone-depletingchlorofluorocarbons, they have promoted the adop-tion of less damaging alternatives, some of whichproved remarkably cost-effective, as the Porter hy-pothesis predicted. For example, the electronics in-dustry, when prodded, found that it could dispensewith CFCs altogether in cleaning its printed circuitboards, substituting aqueous cleaners or fluxless sol-ders that eliminated the need for cleaning. When ap-plied to products that generate environmental prob-lems when used—such as fuels or water—fees orenvironmental surtaxes stimulate consumers to seekout substitutes or find ways to conserve.

The main impediment to the wider use of envi-ronmental charges has been that they collect rev-enue, which generally annoys the industries thathave to pay. Even though the Polluter Pays Principleis accepted in all OECD countries as the basis for en-vironmental policy because it is economically soundand widely regarded as fair, the U.S. Congress indrafting legislation usually gives away the right topollute and hide the costs of environmental controlin technology-based standards. This regulatory ap-proach is so much more costly that even some indus-tries have come out in favor of having environmentalcharges instead. For example, General Motors has ar-gued for promoting fuel efficiency with a gasolinetax, which applies to all vehicles, rather than withCAFE standards, which only hits new car sales. How-ever, what's good for General Motors is apparentlynot good for the politicians.

One would think that the Congress would be de-lighted to find a way to help pay for its promisedmiddle-class and upper-class tax cuts that would im-prove environmental quality, reduce regulatory bur-dens, and make markets work better. Environmentalcharges, unlike virtually all other revenue sources,can actually make the economy more productive.68

At the same time, they offer households and firms theopportunity to reduce their tax bills by behaving inenvironmentally sound ways that are in line withtheir values and ideals. Commuters can avoid payinggasoline taxes by taking the metro, carpooling, or rid-ing their bicycles, for examples. Enacting fees on en-vironmentally damaging products and activities tohelp pay for cuts in more burdensome and distortingtaxes is a "win-win" opportunity too good to pass up.

Environmental charges, unlike virtuallyall other revenue sources, can actuallymake the economy more productive. Atthe same time, they offer householdsand firms the opportunity to reducetheir tax bills by behaving in environ-mentally sound ways that are in linewith their values and ideals.

Unfortunately, that is what Republicans in theHouse of Representatives apparently wish to do. Intheir "Contract with America," they embrace a radi-cally different fiscal approach, proposing that the vic-tims pay the costs of environmental cleanup andbusinesses be granted a legal right to pollute. This in-equitable and injudicious change is imbedded in draftlegislation requiring the federal government to com-pensate property owners when environmental regula-tions reduce the value of investments significantly.Since the only money the federal government has tospend is collected from taxpayers, the effect is toshift the costs of compliance from polluting firms(and ultimately their customers) to the general public.Suppose that risk assessments establish that somechemical in commercial use poses a severe healththreat, and consequently EPA bans or restricts its pro-duction and distribution. Companies that have in-vested in factories to produce this risky chemicalwould have to be compensated by the governmentfor the reduction in the value of their investments, anarrangement that does nothing to encourage firms tobe more responsible. Ultimately, the taxpayer wouldpay the compensation through higher income andpayroll taxes, an outcome that conservative politi-cians supposedly decry because it weakens people'sincentives to succeed. This is a misuse of the fiscalmechanism. Compensating businesses that pollute isnot the way to improve environmental regulation.

Better use of market incentives is one importantway to get better results from environmental spend-

ing, but there are others. Both Carol Browner, theEPA's current Administrator and her predecessor, BillReilly, have pursued initiatives to deal with the air,water, and solid effluents of individual facilities in anintegrated way, to look at the environmental prob-lems of entire industries or regions, and to cooperatewith businesses, environmental groups, and localgovernments in devising sensible solutions. Theseinitiatives are attempts to get away from rigid statute-by-statute, media-by-media regulation and introducesome overall coherence and rationality into theprocess. Environmental groups, corporations, andother stakeholders have shown their willingness toenter into dialogues and partnerships to find bettersolutions to environmental problems. 9 Thesepromising initiatives should be pursued vigorously.

Another important avenue is to use risk assess-ment more effectively in establishing environmentalpolicy and strategy. At present, there is little concor-dance between the resources spent on various envi-ronmental problems and the risks they represent.70

Some of the reasons why this situation persists havebeen discussed above. Risk assessment is fraughtwith difficulties and uncertainties that simple-mindedappeals to "sound science" as the basis for regula-tion cannot sweep away.71 Considerations of fair-ness, caution, and responsiveness to popular con-cerns have a legitimate place in environmentalpolicy. Nonetheless, systematic efforts to comparevarious environmental risks and set priorities accord-ingly can help reduce the extreme misallocations ofresources that regularly occur.72 It doesn't makesense to use risk assessment requirements as a pro-cedural barrier to any regulatory action, but, at aminimum, environmental statutes that seem to pre-clude efforts to take risk reduction into accountshould be amended.

By and large, these opportunities to get more forthe money spent on environmental protection de-pend on changing the approach to environmentalregulation. But such change is just the tip of the ice-berg. Most of our environmental problems are exac-erbated by policy-induced inefficiencies in importantsectors of the economy: agriculture, transportation,industry, and energy. Reducing these inefficiencieswould reduce the need for environmental regulationsand raise productivity throughout the economy.

Most of our environmental problemsare exacerbated by policy-induced inef-ficiencies in important sectors of theeconomy: agriculture, transportation,industry, and energy. Reducing these in-efficiencies would reduce the need forenvironmental regulations.

¥One would think that to fulfill their promises to

reduce the deficit and cut taxes, the Administrationand Congress would quickly agree to eliminate awhole range of fiscal subsidies that are inefficient, in-equitable, and environmentally damaging. Take thetransportation industry, for example. Instead of let-ting market forces determine which freight travels bywhich method, government in its wisdom subsidizesthe trucking industry, which pays far less than itsshare of highway costs,73 but then also subsidizes therailroads, inland waterways, and the merchant ma-rine, all the competing transport modes. The net re-sult is a total hash, a good deal of needless environ-mental damage, and a substantial burden on thetaxpayer.

Such is the deference Congress pays to specialinterests that these subsidies persist while programsthat help the genuinely needy are cut. The publiclands belong to the American people—all the people.Instead of collecting fair market value on behalf ofthe citizens from businesses that make use of thepublic lands, Congress defends subsidies to powerfulmining, logging, and agribusiness companies. Al-though crocodile tears are shed for the small rancheror farmer, the fact is that these subsidies go over-whelmingly to big companies and multi-millionaireoperators. As a result, our public lands are over-grazed, deforested, and scarred with abandonedmines still leaching into Western rivers. One way toprotect our natural resources and give the middle-class taxpayer a break would be to put a halt to thesegiveaway deals.

The most important of these opportunities facingthe next Congress will be to restructure our agricul-tural support programs, which have cost the taxpayeran average of 13 billion dollars per year over the pastdecade. Let alone that throughout this entire periodand long before, the average income of farmers ex-ceeded that of non-farmers. Ignore the fact that therichest 5 percent of all farmers received more thanhalf of the payments. Such inequities aside, the agri-cultural support programs badly distort agriculturalmarkets and the use of resources in agriculture. Com-modity support programs induce farmers to plant thesame crops over and over again, maintaining soil fer-tility and suppressing pests through heavy chemicalapplications. The highest payments support the cropsand cropping practices with the worst environmentalcosts of soil erosion and chemical run-off.7

Instead of paying farmers in propor-tion to the acreage of corn, soybeans,and other program crops they plant,and then paying them again to take partof their land out of production, the gov-ernment should reward farmers forprotecting streambanks from erosion,for providing habitat for wild species,for restoring wetland, and for protect-ing the land against wind and watererosion.

Just cutting back on farm subsidies and allowingfarmers more flexibility in their production decisionswould reduce government spending and environ-mental damages substantially, with relatively little im-pact on farm income. If the government continues tosupport farmers' incomes, whether they need help ornot, there are far better ways to do it. Instead of pay-ing farmers in proportion to the acreage of corn, soy-beans, and other program crops they plant, and thenpaying them again to take part of their land out of

production, the government should reward farmersfor protecting streambanks from erosion, for provid-ing habitat for wild species, for restoring wetland,and for protecting the land against wind and watererosion. When farmers undertake such activities, so-ciety benefits, not just the farmers themselves. These"green payments" create the incentives farmers needto provide these social benefits. At the same time,with the government out of commodity markets,farmers would be free to plant the crops that are re-ally the most profitable, agricultural productivitywould rise, and the environmental damages fromagricultural run-off would decline.

In the international arena, there are "win-win"opportunities of comparable importance. Rather thanthreatening our trading partners with trade barriersunless they raise their environmental standards, theUnited States should be offering to reduce trade bar-riers and improve their access to U.S. markets if theydo raise their environmental standards. The UnitedStates gains from trade expansion and loses from

trade restrictions—our own as well as other coun-tries' restrictions. The negative approach hurts usalong with our trading partners. The more positiveapproach worked well in the NAFTA negotiations.Mexico agreed to strengthen its own environmentalprotection policies significantly and to cooperate withthe United States in cleaning up our common borderregion and in protecting dolphins from accidental en-trapment. The carrot of expanded trade worked moreeffectively than the stick of sanctions. The same ap-proach would work well as discussions begin on ex-tending the free trade area in the Americas and creat-ing a vast Pacific free trade area. It could give animportant and positive direction to the new WorldTrade Organization.

These are all opportunities that a RepublicanCongress and a Democratic administration should beable to agree on. They would improve environmentalquality, strengthen the economy, reduce regulatoryburdens, cut the deficit, and help finance popular taxcuts.

Robert Repetto is Vice President and Senior Economist at WRI and Director of the Institute's Program in Eco-nomics and Population. Formerly, he was an associate professor of economics at the School of Public Health atHarvard University and a member of the economics faculty at Harvard's Center for Population Studies.

APPENDIX

Appendix Table IA. Correlations between Profitability and Airborne Emissions8

SICCode

26214

2911129510

32510

3273033219

Name of Number ofIndustry Establishments

Uncoated freesheetGasolinePaving mixturesand blocksBrick and structuralclay tileReady-mix concreteOther grey ironcastings

10

15

37

14

13

15

Gross Operating Marginb

SimpleCorrelation

-.21+ .21

+ .24

-.08-.13

- . 6 9 * e

Partial41

Correlation

-.31+ .19

+ .24

-.34+ .14

-,71*e

Net Return on Capital0

SimpleCorrelation

-.43-.10

+ .15

-.01+ .05

-,69*e

Partiar4

Correlation

-.12-.28

+ .12

-.33+ .46

-.70*e

a. Airborne emissions are measured as particulate emissions per dollar of shipments.b. Gross Operating Margin is defined as the total value of shipments minus current operating costs, expressed

as a fraction of total shipments.c. Net Return on Capital is defined as the total value of shipments minus current operating costs and

depreciation, divided by the book value of capital invested in structures and equipment.d. Partial Correlations control for age of plant, scale of production, and the ratio of investment in plant and

equipment in current year to book value of investment.e. Correlations marked with an asterisk are significantly different from zero at the 5 percent probability level.

Appendix Table IB. Correlations

SIC

Code

202232033226214

26314

28199

28213

28214

28220

2865128697

29111

33123

33541

34710

between

Name of Number of

Profitability and Water-borne Emissions8

Gross Operating Marginb

SimpleIndustry Establishments Correlation

Natural cheese

Canned vegetables

Uncoated free sheet

Recycled paperboard

Inorganic chemicalsThermoplastic resinsand plastics

Thermosetting resins

and plastics

Synthetic rubber

Cyclic intermediatesMisc. cyclic and

acyclic chemicals

GasolineHot rolled sheet and

strip

Extruded aluminumrod

Electroplating, plating

and polish

19

22

19

11

23

35

14

12

12

3455

20

12

28

+ .37-.25+ .01+ .32

+ .43*e

- .27

+ .27+ .19+ .05

+ .02-.11

+ .04

+ .47

- .23

Partial41

Net ReturnSimple

Correlation Correlation

+ .28

- .32

+ .00+ .04

+ .45*e

- .32

+ .13-i -i

+ .06

- .08- .14

+ .01

+ .30

- .20

+ .34

+ .34

+ .09- .14+ .01

-.25

- .06- .16

- .34

- .08

- .05

- .10

+ .12

- .18

on Capitalc

Partiald

Correlation

+ .31+ .34+ .12

- .35+ .10

- .23

- .11

+ .04

- .26

- .14

- .06

- .17

- .01

- .11

a. Water-borne Emissions are measured as BOD plus Total Suspended Solids (TSS) per dollar of totalshipments.

b. Gross Operating Margin is defined as the total value of shipments minus current operating costs, expressedas a fraction of total shipments.

c. Net Return on Capital is defined as the total value of shipments minus current operating costs anddepreciation, divided by the book value of capital invested in structures and equipment.

d. Partial Correlations control for age of plant, scale of production, and the ratio of investment in plant andequipment in current year to book value of investment.

e. Correlations marked with an asterisk are significantly different from zero at the 5 percent probability level.

Appendix Table IC. Correlations between

SIC Name of Number of

Profitability and Toxic Releasesa

Gross Operating Marginb

SimpleCode Industry Establishments Correlation

20223 Natural cheese20863 Bottled carb.

soft drinks20864 Canned carb.

soft drinks24212 Softwood lumber,

rough and dressed24341 Wood kitchen

cabinets25115 Wood bedroom

furniture26214 Uncoated free sheet26314 Recycled paperboard26570 Paperboard packaging28213 Thermoplastic resins,

plastics28214 Thermosetting resins,

plastics28411 Soaps and detergents,

non household28511 Architectural coatings29111 Gasoline

54

62

35

32

38

51473444

116

79

65113100

+ .04

-.14

+ .22

+ .04

-.05

+ .00+ .33*e

-.42*e

+ .17

+ .13

+ .13

_ 3 0 * e

-.08+ .10

Partiald

Correlation

+ .08

-.10

+ .20

+ .03

+ .10

-.01+ .28_ 34*e

+ .21

+ .09

+ .13

-.30*e

-.05+ .08

Net ReturnSimple

Correlation

+ .28

-.20

+ .04

+ .11

-.21

-.05+ .07-.16+ .03

-.08

-.08

-.15-.08-.04

on Capital0

Partiald

Correlation

+ .37

-.14

+ .17

+ .19

+ .20

-.20+ .03-.08-.02

-.06

-.07

-.07-.06-.02

continued on next page

Appendix Table IC. (Continued)

SIC Name of Number ofCode Industry Establishments

30864 Plastic furnishings andfurniture

32210 Glass containers33124 Hot rolled ferrous bars,

shapes and plates33541 Extruded aluminum

rods and bars34111 Steel cans and tinware34112 Aluminum cans34710 Electroplating, plating

and polishings34790 Etching, engraving and

coating36720 Printed circuit boards36741 Integrated microcircuits36913 Lead acid storage

batteries

61

37

42

595162

429

126

7968

52

Gross OperatingSimple

Marginb

Partial*1

Correlation Correlation

+ .06-.15

-.23

- . 1 1

- . 2 1

- . 1 3

_ 34*e

- .08

- .17-.26*e

+ .06

- .09- .19

- .20

- .13- .24- . 1 0

_ 34*6

- .081 Q

-.23

+ .04

Net ReturnSimple

Correlation

+ .05-.20

-.13

-.03-.18-.17

-.07

-.05-.10+ .02

-.08

on Capital0

Partiald

Correlation

-.26-.19

-.12

+ .04-.20-.13

-.08

-.03- .12

-.32*e

-.10

a. Toxic releases are measured as emissions to all media plus transfers for off-site treatment, per dollar of totalshipments.

b. Gross Operating Margin is defined as the total value of shipments minus current operating costs, expressedas a fraction of total shipments.

c. Net Return on Capital is defined as the total value of shipments minus current operating costs anddepreciation, divided by the book value of capital invested in structures and equipment.

d. Partial Correlations control for age of plant, scale of production, and the ratio of investment in plant andequipment in current year to book value of investment.

e. Correlations marked with an asterisk are significantly different from zero at the 5 percent probability level.

NOTES

1. Richard Schmalensee, "The Costs of Environmen-tal Protection," Balancing Economic Growth and

Environmental Goals, (Washington, D.C.: Ameri-can Council for Capital Formation, 1994).

2. Richard B. Stewart, "Environmental Regulationand International Competitiveness," Yale Law

Journal, June 1993, vol. 102, no. 8, pp.2039-2106.

3. Toward Smarter Regulation, The Business Round-table, Washington, D.C., 1994.

4. S.984, 102nd Congress, defined weaker pollutioncontrols as a countervailable subsidy. S.1965,102nd Congress, would have required the Secre-tary of Commerce to levy import fees on productsmanufactured by processes not meeting the stan-dards of the Clean Water Act.

5. S.59, 102nd Congress authorized actions undersection 301 of the Trade Act against policies andpractices of foreign countries that diminish the ef-fectiveness of international conservation agree-ments. HR1445, 103rd Congress, would requireadoption in any trade agreement of the principlethat denial or negation of environmental stan-dards constitute an actionable unfair trade prac-tice. HR4710, 103rd Congress would require fu-ture trade agreements to treat as an actionableunfair trade practice the "denial" of internationallyrecognized environmental standards in the ex-porting country.

6. North American Agreement on Environmental Co-

operation, Government Printing Office, Washing-ton, D.C., September, 1993; Steve Charnovitz,"NAFTA: An Analysis of its Environmental Provi-sions," Environmental Law Reporter, February,vol. 23, 1993.

7. Heraldo Munoz, ed., Environment and Diplo-

macy in the Americas, (Boulder, Colorado: LynneRiener Publishers, 1992).

8. Michael E. Porter, "The Competitive Advantage ofNations," Harvard Business Review, March-April,1990, pp.74, 87.

9'. Development Assistance, Export Promotion, and

Environmental Technology, U.S. Congress, Office

of Technology Assessment, Washington, D.C.,August, 1993.

10. Dale Jorgenson and Peter Wilcoxen, "Impact ofEnvironmental Legislation on U.S. EconomicGrowth, Investment and Capital Costs." U.S. Envi-

ronmental Policy and Economic Growth: How Do

We Fare?, (Washington, D.C.: American Councilfor Capital Formation, 1992); Michael Hazilla andRaymond J. Kopp, "Social Costs of EnvironmentalQuality Regulations: A General EquilibriumAnalysis," Journal of Political Economy, 1990, vol.98, pp.853-873.

11. This is not necessarily true in decreasing returnsor oligopolistic industries. Paul Krugman, ed.,Strategic Trade Policy and the New Lnternational

Economics, (Cambridge: MIT Press, 1986). How-ever, it is true of most actual trade barriers. Costs

and Benefits of Protection, (Paris: OECD, 1985).

12. D. Phantumvanit and Theodore Panayotou, In-

dustrialization and Environmental Quality: Pay-ing the Price, (Bangkok: Thailand DevelopmentResearch Institute, 1990); World Bank, China En-

vironmental Strategy Paper, Washington, D.C.,1982; World Bank, Indonesia Environment and

Development: Challenges for the Future, Washing-ton, D.C., 1993; Werner Baer and Charles Mueller,"Environmental Aspects of Brazil's Economic De-velopment," Brazilian Review, forthcoming.

13. Robert Repetto et al, Wasting Assets: Natural Re-

sources in the National Income Accounts, (Wash-ington, D.C.: World Resources Institute, 1989);World Resources Institute and Tropical SciencesCenter (Costa Rica), Accounts Overdue: Natural

Resource Depreciation in Costa Rica, (Washing-ton, D.C.: World Resources Institute, 1991).

14. Graciela Chichilinsky, "North-South Trade and theGlobal Environment," American Economic

Review, vol.84, no.4, September, 1994,

pp.851-874; Kym Anderson and Richard Black-hurst, eds., The Greening of World Trade Issues,

New York, 1992.

15. Hazilla and Kopp, op.cit.

16. The analysis of carbon taxes to combat globalwarming is an important case in point. See U.S.Congress, Congressional Budget Office, Carbon

Charges as a Response to Global Warming: The Ef-

fects of Taxing Fossil Fuels, Washington, D.C.,1990; Dale Jorgenson and Peter Wilcoxen, "Re-ducing U.S. Carbon Emissions: The Cost of Differ-ent Goals," Energy, Growth and the Environment,

vol.7, 1992. For a more balanced treatment,William R. Cline, Global Warming: The Economic

Stakes, (Washington, D.C.: Institute for Interna-tional Economics, 1992); Roy Boyd and KerryKrutilla, "Energy Taxation as a Policy Instrumentto Reduce CO2 Emissions: A Net Benefit Analy-sis," Journal of Environmental Economics and

Management, forthcoming.

17. A.J. Barbera and V.D. McConnell, "The Impact ofEnvironmental Regulations on Industry Productiv-ity: The Direct and Indirect Effects," Journal of

Environmental Economics and Management,

vol.18, 1990, pp.50-65; F.M. Gollop and M.J.Roberts, "Environmental Regulations and Produc-tivity Growth: The Case of the Electric Power In-dustry," Journal of Political Economy, vol.91,1983, pp.654-674; W. Gray, "The Cost of Regula-tion: OSHA, EPA, and the Productivity Slow-down," American Economic Review, vol.77, 1987,pp.998-1006; W. Gray and R.J. Shadbegian, "Envi-ronmental Regulation and Manufacturing Produc-tivity at the Plant Level," Center for EconomicStudies Discussion Paper 93-6, (Washington,D.C.: U.S. Bureau of the Census, 1993).

18. Robert Repetto, "Environmental Productivity andWhy It is So Important," Challenge, vol.33, 1990,Pp.33-38.

19. Adam B. Jaffe, Steven R. Peterson, Paul R. Port-ney, and Robert N. Stavins, "Environmental Regu-lation and International Competitiveness: WhatDoes the Evidence Tell Us?," (Washington, D.C.:Resources for the Future Discussion Paper, De-cember, 1993).

20. Patrick Low and Alexander Yeats, "Do Dirty In-dustries Migrate?," Patrick Low, ed., International

Trade and the Environment, (Washington, D.C.:World Bank, 1992).

21. Piritti Sorsa, Competitiveness and Environmental

Standards, (Washington, D.C.: World Bank Policy

Research Working Paper 1249, 1994); p.i.

22. Ibid, Table 6, p. 11.23 Kopp, Portney, and DeWitt, "International Com-

parisons of Environmental Standards," (Washing-ton, D.C.: Resources for the Future DiscussionPaper, 1990).

24. United Nations, Department of Economic and So-cial Development, World Investment Report, 1992,

New York, 1992, Annex Table 12, p.346.

25. Patrick Low & Alexander Yeats, op.cit.

26. R. Vernon, "International Investment and Interna-tional Trade in the Product Cycle," Quarterly

Journal of Economics, vol.80, May, 1966,

pp.190-207.27. Judith Dean, "Trade and the Environment: A Sur-

vey of the Literature," Patrick Low, op.cit.,

pp.15-28.28. Candice Stevens, "Synthesis Report," Environmen-

tal Policies and Industrial Competitiveness, (Paris:OECD, 1993), p.7.

29. Adam Jaffe, Steven Peterson, Paul Portney, andRobert Stavins, "Environmental Regulation and In-ternational Competitiveness: What Does the Evi-dence Tell Us?," (Washington, D.C.: Resources forthe Future Working Paper, December, 1993); p.i.

30. Richard B. Stewart, "Environmental Regulationand International Competitiveness," Yale Law Re-

view, Vol.102, No.8, June 1993; pp.2039-2106.

31. Timothy Bartik, "The Effects of EnvironmentalRegulation on Business Location in the UnitedStates," Growth and Change, Vol. 19, 1988;

pp.22-44; Jaffe et.al. p. 14; Arik Levinson, "Envi-ronmental Regulations and Manufacturers' Loca-tion Choices," (New York: Columbia University,Dept. of Economics Working Paper, November,1992).

32. Robert U. Ayres, "Industrial Metabolism: Theoryand Policy," Braden R. Allenby and Deanna J.Richards, The Greening of Industrial Ecosystems,

(Washington, D.C.: National Academy Press,1994).

33. Wallace Oates, Karen Palmer, and Paul Portney,"Thinking About the Porter Hypothesis," (Wash-ington, D.C.: Resources for the Future DiscussionPaper, 1993)-

34. "The Car Makers' Recovery Stakes," The Econo-

mist, Oct. 29, 1994, p.73.

35. Paul Milgrom and John Roberts, Economics, Or-

ganization and Management, (New Jersey: Pren-tice-Hall, 1992).

36. Stephen de Canio, "Agency and Control Problemsin U.S. Corporations: The Case of Energy-EfficientProjects," Journal of the Economics of Business,

Vol.1, No.l, 1994, pp.105-123.

37. Robert H. McGuckin and George A. Pascoe Jr.,"The Longitudinal Research Database: Status andResearch Possibilities," Survey of Current

Business, U.S. Department of Commerce, Bureauof Economic Analysis, Vol. 68, no. 11, November1988; pp.30-36.

38. U.S. EPA, Office of Pollution Prevention and Tox-ics, 1992 Toxics Release Inventory, Washington,D.C., 1994 and previous years.

39. Gary L. Rutledge and Christine R. Vogan, "Pollu-tion Abatement and Control Expenditure,"1972-1992, Survey of Current Business, vol.74,no.5, 1994; pp.36-49.

40. Roger H. Bezdek, "Environment and Economy:What's the Bottom Line?," Environment, vol.35,no.7, 1993; pp.7-11.

41. Kusum Ketkar, "Pollution Control and Inputs toProduction," Journal of Environmental Economics

and Management, Vol.1, no.l, 1983; pp.50-59.

42. Deborah Vaughn Nestor and Carl A. Pasurka, Jr.,"The U.S. Environmental Protection Industry,"(Washington, D.C.: U.S. EPA, Office of Policy,Planning and Evaluation, July, 1994).

43. The comparable figure is 6 jobs per $1,000,000 in1991 prices. Alan Carlin, Paul F. Scodari, and DonH. Garner, "Environmental Investments: TheCosts of Cleaning Up," Environment, vol.34, no.2,March, 1992; pp. 12-44.

44. "Redefining the Environmental Industry," Envi-

ronmental Business Journal, vol.6, no.4, April,1991; (Paris: OECD, The OECD Environment In-

dustry: Situation, Prospects and Government Poli-

cies, 1992).

45. Louis S. Jacobsone, Robert J. LaLonde and DanielG. Sullivan, "Earnings Losses of Displaced Work-ers," American Economic Review, vol.83, no.4,1993; pp.685-709.

46 Stephen Breyer, Regulation and its Reform,

(Cambridge, MA.: Harvard University Press,1982); Stephen Breyer, Breaking the Vicious

Circle: Toward Effective Risk Regulation, (Cam-bridge, MA.: Harvard University Press, 1993);Wesley A. Magat, ed., Reform of Environmental

Regulation, (Cambridge, MA.: Ballinger, 1982);Robert Crandall, "Why is the Cost of Environ-mental Regulation So High?" Center for theStudy of American Business, Policy Study 110,Washington University of St. Louis, February,1992; Paul Portney, ed., Public Policies for En-

vironmental Protection, (Baltimore: Resourcesfor the Future, Johns Hopkins Press, 1990).

47. Richard Wilson and Edmund Crouch, Risk/Benefit

Analysis, Ballinger, 1982; William Lowrance, Of

Acceptable Risk; W. Kip Viscusi, Fatal Tradeoffs,

1992.

48. Albert Nichols, "The Regulation of Airborne Ben-zene," Thomas Schelling, ed., Incentives for Envi-

ronmental Protection, (Boston: MIT Press, 1983);A. Myrick Freeman III, "Risk Evaluation in Envi-ronmental Regulation," W.A. Magat, op.cit.;

Tammy O. Tangs, "Five Hundred Life-Saving In-terventions and their Cost-Effectiveness," Risk

Analysis, forthcoming.

49. Lester Lave, ed., Quantitative Risk Assessment in

Regulation, (Washington, D.C.: Brookings Institu-tion, 1982), Ch. 1.

50. Allen V. Kneese and Blair T. Bower, Environmen-

tal Quality and Residuals Management, (Balti-more: Resources for the Future, Johns HopkinsPress, 1979); EPA's current Common Sense Initia-tive attempts to work out integrated cross-mediaapproaches to residuals management within par-ticular industries.

51. Paul R. Portney, ed., Public Policies for Environ-

mental Protection, (Washington, D.C.: Resourcesfor the Future, 1990), p. 72.

52. AMOCO - U.S. EPA Pollution Prevention Project,

Yorktown, Virginia, (Washington, D.C.: Amoco,EPA, June, 1992).

53. Howard Gruenspecht, "Differentiated Regulation:The Case of Auto Emissions Standards," American

Economic Review, vol. 72, no.2, 1990, 328-331;Randy A. Nelson, Michael R. Donihue and ThomasTeitenberg, "Differential Environmental Regulation:Effects on Electric Utility Capital Turnover andEmissions," (Waterville, Maine: Colby College,Dept. of Economics Working Paper, 1990).

_H

54. Rob Coppock, Regulating Chemicals Hazards in

Japan, W. Germany, UK, France, and the EC: A

Comparative Examination, (Washington, D.C.:National Research Council, National AcademyPress, 1986).

55. Robert E. Litan, "Superfund: Assessing the Pro-gram and Options for Reform," Enhancing Envi-

ronmental Quality Through Economic Growth,

(Washington, D.C.: American Council for CapitalFormation, Center for Policy Research, 1993);Roger E. Dower, "Hazardous Wastes," Paul R.Portney, ed., Public Policies for Environmental

Protection, (Washington, D.C.: Resources for theFuture, 1990).

56. Lester Lave & Howard Gruenspecht, "Increasingthe Effectiveness and Efficiency of EnvironmentalDecisions: Benefit-Cost Analysis and EffluentFees: A Critical Review," Journal of the Air and

Waste Management Association, vol.4l, no.5,May, 1991, pp.680-693; Maureen L. Cropper andWallace E. Oates, "Environmental Economics: ASurvey," Journal of Economic Literature, vol.XXX,No.2, June, 1992, pp.675-740.

57. Vice President Al Gore, Creating a Government

That Works Better and Costs Less, Washington,D.C., 1993.

58. Choosing a Sustainable Future, The Report of theNational Commission on the Environment, (Wash-ington, D.C.: Island Press, 1993).

59- Toward Smarter Regulation, The Business Round-table, Washington, D.C., 1994; Stephan Schmid-heiny with the Business Council for SustainableDevelopment, Changing Course, (Cambridge,MA.: MIT Press, 1992).

60. A New Generation of Environmental Leadership:

Action for the Environment and the Economy,

(Washington, D.C.: World Resources Institute,1993).

61. Project 88: Harnessing Market Forces to Protect

our Environment, Washington, D.C., 1988; Project

88—Round II: Incentives for Action: Designing

Market-Based Environmental Strategies, Washing-ton, D.C., 1991.

62. Robert Repetto, "Environmental Productivity andWhy it is So Important," Challenge, 1990; RobertRepetto et al., Wasting Assets, (Washington, D.C.:World Resources Institute, 1989).

63- U.S. EPA, Office of Policy and Program Evalua-tion, Economic Incentives: Options for Environ-

mental Protection, Washington, D.C., 1991;OECD, Managing the Environment: The Role of

Economic Instruments, (Paris: OECD, 1994).

64. Thomas H. Teitenberg, Emissions Trading: An Ex-

ercise in Reforming Pollution Policy, (Washington,D.C.: Resources for the Future, 1985); James T.B.Tripp and Daniel J. Dudek, "Institutional Guide-lines for Designing Successful TransferableRights," Yale Journal of Regulation, vol.6, 1986;pp.369-391.

65. Robert W. Hahn and Gordon L. Hester, "WhereDid All the Markets Go? An Analysis of EPA'sEmissions Trading Program," Yale Journal of Reg-

ulation, vol.6, no.l, 1989; pp. 109-153-

66. Federal Register, December 11, 1979, 71780; Fed-

eral Register, December 4, 1983, 43814; Federal

Register, February 23, 1993, 11110.

67. Richard E. Ayres, "Developing a Market in Emis-sions Credits Incrementally: An 'Open Market'Paradigm for Market-Based Pollution Control,"Environmental Reporter, vol.25, no.31, December2, (Washington, D.C.: Bureau of National Affairs,1994).

68. Robert Repetto, Roger C. Dower, Robin Jenkinsand Jacqueline Geoghegan, Green Fees: How a

Tax Shift Can Work for the Environment and the

Economy, (Washington, D.C.: World ResourcesInstitute, 1992).

69. Partnerships to Progress.- The Report of the Presi-

dent's Commission on Environmental Quality,

(Washington, D.C., 1993).70. U.S. EPA, Office of Policy and Program Evalua-

tion, Unfinished Business: A Comparative Assess-

ment of Environmental Problems, (Washington,D.C.: EPA, 1987); U.S. EPA, Science AdvisoryBoard, Reducing Risk: Setting Priorities and

Strategies for Environmental Protection, (Wash-ington D.C.: EPA, 1990).

71. Lester B. Lave, ed., Quantitative Risk Assessment

in Regulation, (Washington, D.C.: Brookings,1982); The Carnegie Commission on Science,Technology, and Government, Risk and the Envi-

ronment: Improving Environmental Decision-

Making, (New York: Carnegie Corporation,1993).

72. John D. Graham, "Making Sense of Risk: An icy, (Washington, D.C.: Brookings Institution,

Agenda for the Congress," (Boston, Mass.: Center 1989).

for Risk Analysis, Harvard School of Public 74. Paul Faeth et al., Paying the Farm Bill: U.S. Agri-

Health, October, 1994). cultural Policy and the Transition to Sustainable

73- Kenneth Small, C. Winston, and C. Evans, Road Agriculture, (Washington, D.C.: World Resources

Work: New Highway Pricing and Investment Pol- Institute, 199D-

• r