Racing to the Bottom? Foreign Investment and Air Pollution ...sabrash/110/Chem... · wheeler / air pollution in developing countries 227 investment (FDI) among developing countries

JOURNAL OF ENVIRONMENT & DEVELOPMENTWheeler / AIR POLLUTION IN DEVELOPING COUNTRIES

Racing to the Bottom?Foreign Investment and Air Pollution

in Developing Countries

DAVID WHEELER

Critics of free trade have raised the specter of a “race to the bottom” in whichenvironmental standards collapse because polluters threaten to relocate to“pollution havens” in the developing world. Proponents of this view advocatehigh, globally uniform standards enforced by punitive trade measures thatneutralize the cost advantage of would-be pollution havens. To test the race-to-the-bottom model, the author analyzes recent air quality trends in the UnitedStates and the three largest recipients of foreign investment in the developingworld: China, Brazil, and Mexico. The evidence clearly contradicts the model’scentral prediction: The most dangerous form of air pollution has actuallydeclined in major cities of all four countries during the era of globalization.Citing recent research, the author argues that the race-to-the-bottom model isflawed because its basic assumptions misrepresent the political economy of pol-lution control in developing countries.

Could globalization trigger an environmental “race to the bottom” inwhich competition for investment and jobs relentlessly degrades envi-ronmental standards? Because billions of people subsist on less than U.S.$2 a day, it would be cavalier to dismiss this threat lightly. Indeed, therace-to-the-bottom model provokes widespread concern because itsunderlying assumptions have an air of plausibility.1 In the race-to-the-

225

Journal of Environment & Development, Vol. 10, No. 3, September 2001 225-245© 2001 Sage Publications

1. Daly (2000) recently provided a forceful statement of the race-to-the-bottom model.In the United States, political opponents of the North American Free Trade Agreement(NAFTA) and the World Trade Organization (WTO) frequently invoke elements of themodel. For example, Congressman David Bonior (1999) recently offered the following cri-tique of the WTO:

Many of the world’s political and economic leaders have adopted our ways of discussing theglobal economy. They’re warning against “a race to the bottom.” They’re calling for “putting ahuman face on the global economy.” And they say they’re supporting labor and environmentalstandards in trade agreements. . . . Make no mistake about what is at stake in the deliberationsand decisions of the WTO. . . . The WTO, as currently structured, threatens to undo internation-ally everything we have achieved nationally—every environmental protection, every con-sumer safeguard, every labor victory.

In a similar vein, Ralph Nader’s presidential campaign has offered this statement ontrade and the environment on its Web site:

Among the most fetid examples of political cowardice and collusion between elected represen-tatives and big business of the past thirty-five years are the passage of the North American Free

bottom world, decent environmental standards impose high costs onpolluters in high-income economies. To remain competitive, these firmsrelocate to low-income countries whose people are desperate for jobsand income. Local governments ignore regulations in order to promoteinvestment and economic growth, allowing businesses to minimizecosts by polluting with impunity. Driven by shareholders to maximizeprofits, international firms follow suit. Rising capital outflows force gov-ernments in high-income countries to begin relaxing environmentalstandards, but this proves fruitless because the poorest countries haveno environmental standards at all. As the ensuing race to the bottomaccelerates, all countries converge to the hellish pollution levels thatafflict the poorest.

Proponents of this catastrophe model have a straightforward preven-tive recommendation: high, globally uniform environmental standardsand, for countries that are unwilling or unable to enforce them, tariffs orother restrictions on imports of their pollution-intensive products thatneutralize their cost advantage as pollution havens. Proponents of freetrade naturally view these prescriptions as anathema, arguing that theirmain impact would be the denial of jobs and income to the world’s poor-est people.

Testing the Race-to-the-Bottom Model

What does the evidence suggest? Pollution control does use scarceresources, so polluting activities in high-income economies have higherregulatory costs than in developing countries (Jaffe, Peterson, Portney, &Stavins, 1995; Mani & Wheeler, 1998). This creates a source of “compara-tive advantage” for the latter that might induce some highly pollutingindustries to relocate. Empirically, the question turns on whether differ-ences in pollution control costs outweigh other locational considerationsand whether internationally mobile activities actually pollute more indeveloping countries. Numerous studies have suggested that pollutioncontrol costs are not major determinants of relocation (Albrecht, 1998;Eskeland & Harrison, 1997; Janicke, Binder, & Monch, 1997; Levinson,1997; Tobey, 1990; Van Beers & Van den Bergh, 1997) and that mostOrganisation for Economic Co-operation and Development (OECD)–

226 JOURNAL OF ENVIRONMENT & DEVELOPMENT

Trade Agreement . . . and the revised General Agreement on Tariffs and Trade . . . into federallaw. These agreements have little to do with the benefits of trade for citizens of member coun-tries. The agreements were designed, largely by corporate lobbyists, as a “pull-down” mecha-nism and to facilitate the movement of capital across national boundaries. Such one-dimensional monetized logic tramples long-standing efforts around the world—some verysuccessful—to protect the environment because environmental safeguards are very often con-sidered “non-tariff barriers to trade” and thus become targets for removal. Five years of WTOoperation have made clear what a grave threat the trade organization is to the world environ-ment. (Nader 2000)

based multinationals maintain near-uniform environmental standardsin their national and international plants (Dowell, Hart, & Yeung, 2000).However, the evidence also suggests that pollution havens can emergein extreme cases (Xing & Kolstad, 1995). During the 1970s, for example,environmental regulation tightened dramatically in the OECD econo-mies with no countervailing change in developing countries. The regu-latory cost differential was apparently sufficient to generate a significantsurge in production and exports of pollution-intensive products fromdeveloping countries. Since then, however, regulatory changes in thelatter have narrowed the gap and apparently stopped the net migrationof polluting industries. In the present, relatively stable situation, devel-oping countries’ imports from high-income economies remain more pol-lution intensive than their exports to those economies (Albrecht, 1998;Mani & Wheeler, 1998).

A simpler, more direct test of the race-to-the-bottom model is alsopossible because the model yields such a simple, robust prediction: Afterdecades of increasing capital mobility and economic liberalization, therace to the bottom should be under way, and pollution should be increas-ing everywhere. It should be rising in poor countries because they arepollution havens and in high-income economies because they are relax-ing standards to remain cost competitive. Trends in available pollutiondata provide a reasonable basis for testing these propositions.

Climatic and economic factors cause pollution to vary considerablyfrom year to year, so trend analysis requires an extended series of moni-toring data. For comparison of environmental conditions in large urbanregions, air pollution measures are generally more reliable and compa-rable than water pollution data. Among widely measured air pollutants,the international health community currently believes the most damag-ing to be suspended particulate matter (SPM; dust). Numerous healthstudies in low- and high-income countries have associated high concen-trations of suspended particulates with higher than normal rates ofdeath and illness from cardiopulmonary problems. Over time, healthresearch has narrowed its focus from all SPM to particles less than 10microns in diameter (PM-10) and, most recently, to particles whosediameters are less than 2.5 microns (PM-2.5). Atmospheric monitoring isadjusting to these findings, but PM-2.5 readings remain scarce in low-income countries. Because all particulate concentration measures arecorrelated, however, even SPM measures provide useful informationabout pollution that severely damages human health.

Air quality monitoring is routine in high-income countries, but itremains uneven in the developing world. China, Mexico, and Brazil pro-vide notable exceptions. During the past two decades, these three rap-idly industrializing countries have begun monitoring and reportingSPM, PM-10, and other forms of air pollution in a number of industrialcenters. In addition, they had the top three shares of foreign direct

Wheeler / AIR POLLUTION IN DEVELOPING COUNTRIES 227

investment (FDI) among developing countries throughout the 1990s.China’s average share was 28%, and the averages for Mexico and Brazilwere 9% and 7%, respectively. As Figure 1 shows, their combined shareof FDI was nearly 60% of the total for developing countries in 1998. If therace-to-the-bottom model is correct, urban SPM and PM-10 levelsshould be rising in all three countries. In the United States, proponents ofthe race-to-the-bottom model have been particularly vocal about moreliberal trade agreements with Mexico and China. It would seem reason-able to trace this concern to deteriorating air quality in U.S. cities becauseU.S. industrial imports from all three countries have been expanding fordecades.

To test the race-to-the-bottom prediction, Figures 2 to 4 present SPMand PM-10 monitoring data along with FDI statistics for China, Mexico,and São Paulo State, which is Brazil’s dominant industrial region. Forcomparison, Figure 5 displays PM-10 monitoring data from five U.S.metropolitan areas (Los Angeles, Houston, Chicago, Atlanta, and NewYork). The Chinese series is the average annual SPM reading for morethan 50 cities, reflecting the extensive coverage of China’s air monitoringnetwork. The Mexican data reflect SPM readings in Mexico City, and theBrazilian data are drawn from PM-10 measures for the industrial regionof Cubatão in São Paulo State. Mexico City is by far the largest industrialcenter in Mexico, and Cubatão has traditionally been a center for pollution-intensive industry in São Paulo.2 These two regions, along with Los


0

10

20

30

40

50

60

1982 1984 1986 1988 1990 1992 1994 1996 1998

% o

f T

ota

l FD

I in

Dev

elo

pin

g C

ou

ntr

ies

Figure 1: Foreign Direct Investment in China, Brazil, and MexicoSource: World Development Indicators (World Bank).Note: FDI = foreign direct investment.

Angeles, suffer from geographic and climatic conditions that make themnatural traps for air pollution.

The foreign investment data in Figures 1 to 4 provide a compellingpicture of expansion in China, Mexico, and Brazil during the past twodecades. However, Figures 2 to 5 show no sign of a race to the bottom;trends in particulate pollution are downward in all four countries.Despite China’s poverty and rapid industrialization, its major urbanareas have experienced significant declines in SPM. During the periodfrom 1987 to 1995, the average concentration fell from nearly 500 µg/m3

to somewhat more than 300 µg/m3. After rising in the early 1990s, Mex-ico City’s percentage of SPM readings above standards fell to historicallows in the latter part of the decade (the period during which NAFTAwas implemented). In Cubatão, Brazil, the average PM-10 concentrationfell from 155 µg/m3 in 1984 to approximately 80 µg/m3 in 1998.3 In theUnited States, PM-10 concentrations declined in all five metropolitan


ForeignDirectInvestm

ent

(1998$USBillion)

300

350

400

450

500

1987 1989 1991 1993 1995

Par

ticu

late

Air

Po

lluti

on

(SP

M -

ug

/m3 )

10

20

30

40

0

SPM FDI

Figure 2: China: Urban Air Pollution Versus Foreign Direct Investment (FDI)Sources: SPM data are from the China Environmental Monitoring Center. FDI data arefrom World Development Indicators (World Bank).Note: SPM = suspended particulate matter.

2. U.S. data are from the Environmental Protection Agency, Office of Air QualityPlanning and Standards (2001). Brazilian data are from the Pollution Control Agency ofSão Paulo State (Companhia de Tecnologia de Saneamento Ambiental, 1990, 1994, 1996,2001). Chinese data are from the China Environmental Monitoring Center and Wheeler(1999). Mexican data are from the National Environment Ministry (2001).

3. Comparable PM-10 data for the São Paulo metropolitan area are not available. How-ever, air quality reports from Companhia de Tecnologia de Saneamento Ambiental, SãoPaulo State’s Environmental Sanitation Agency, show that average SPM levels in SãoPaulo fell by 52% during the period from 1980 to 1995.

areas during the period from 1988 to 1997. The sharpest reduction (32%)was recorded in Los Angeles, the largest U.S. urban-industrial area in theregion most affected by NAFTA.

These results strongly contradict the race-to-the-bottom model.Instead of racing toward the bottom, major urban areas in China, Brazil,Mexico, and the United States have all experienced significant improve-


2.0

4.5

7.0

9.5

12.0

14.5

10

20

30

40

50

60

1989 1991 1993 1995 1997

% o

f S

PM

Rea

din

gs

Ab

ove

Sta

nd

ard

s

Fo

reig

n D

irec

t In

vest

men

t(1

998

$US

Bill

ion

)

SPMFDI

Figure 3: Air Pollution in Mexico City Versus Foreign Direct Investment in MexicoSources: FDI data are from World Development Indicators (World Bank). SPM data arefrom the National Environment Ministry (2001).Note: SPM = suspended particulate matter, FDI = foreign direct investment.

0

2

4

6

8

10

1985 1987 1989 1991 1993 1995 199780

95

110

125

140

155

Par

ticu

late

Air

Po

lluti

on

(PM

-10

-- u

g/m

3 )

ForeignDirectInvestm

ent

(1998$USBillion)

PM-10 FDI

Figure 4: Brazil: Air Pollution in Cubatão Versus Foreign Direct Investment in SãoPaulo State

Sources: PM-10 data are from Companhia de Tecnologia de Saneamento Ambiental. FDIdata are from the São Paulo State System for Data Analysis. World Development Indicators(World Bank).Note: PM-10 = particulate matter with particles less than 10 microns in diameter,FDI = foreign direct investment.

ments in air quality. The improvements in Los Angeles and Mexico Cityare particularly noteworthy because they are the dominant industrialcenters in the region most strongly affected by NAFTA.

Problems With theRace-to-the-Bottom Model

The race-to-the-bottom model’s basic assumptions must be flawed,because its predictions are inconsistent with urban air pollution trendsin three of the developing world’s major industrial powers. In fact,empirical research has undermined all of these assumptions.

Pollution control is not a critical cost factor for most private firms. Researchin both high- and low-income countries suggests that pollution controldoes not impose high costs on business firms. Jaffe et al. (1995) and oth-ers have shown that compliance costs for OECD industries are surpris-ingly small, despite the use of command-and-control regulations thatare economically inefficient. These results suggest that differential pol-lution control costs do not provide OECD firms with strong incentives tomove offshore. Firms in developing countries frequently have evenlower costs because the labor and materials used for pollution controlare less costly than in the OECD economies. Big polluters also have


20

25

30

35

40

45

50

55

60

1988 1989 1990 1991 1992 1993 1994 1995 1996 1997

Los Angeles

Chicago

Houston

Atlanta

New York

Par

ticu

late

Air

Po

lluti

on

(PM

-10

-- u

g/m

3 )

Figure 5: Air Pollution in U.S. Metropolitan Areas, 1988 to 1997Source: Environmental Protection Agency, Office of Air Quality Planning and Standards(2001).Note: PM-10 = particulate matter with particles less than 10 microns in diameter.

lower average control costs per unit of pollution because abatement issubject to scale economies. Figure 6 displays recent econometric esti-mates of control costs for sulfur dioxide (SO2) air pollution in large Chi-nese factories (Dasgupta, Wang, & Wheeler, 1997).4 For non-state-ownedenterprises, costs of a few dollars per ton are typical until control ratesrise above 70%. As Figure 6 shows, state-owned enterprises have muchhigher costs because they are operated less efficiently. The average costof pollution control has therefore declined as China has moved awayfrom state ownership during the era of liberalization. In Colombia, anew pollution charge program has sharply reduced organic water pollu-tion by large factories. Colombian factory managers have found thatcleaning up is cheaper than paying charges, even when they are set atrelatively low levels. No participating factory seems to have experi-enced financial difficulties in the process (Wheeler, 1999). Similar con-clusions have emerged from studies of regulation and control costs inMalaysia (Jha, Markandya, & Vossenaar, 1999; Khalid & Braden, 1993).

Low-income communities penalize dangerous polluters, even when formalregulation is weak or absent. Abundant evidence from Asia and LatinAmerica shows that neighboring communities can strongly influencefactories’ environmental performance (see Dasgupta, Lucas, & Wheeler,


0

50

100

150

200

250

300

0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9

Rate of Abatement

$US

/To

n

Non-SOESOE

Figure 6: Air Pollution Control Costs in ChinaSource: Wheeler (1999).Note: SOE = state-owned enterprises.

4. Xu, Gao, Dockery, and Chen (1994) have shown that atmospheric SO2 concentrationsare highly correlated with damage from respiratory disease in China. SO2 and other oxidesof sulfur combine with oxygen to form sulfates and with water vapor to form aerosols ofsulfurous and sulfuric acid. Much of the health damage from SO2 seems to come from fineparticulates (PM-2.5) in the form of sulfates.

1998; Hartman, Huq, & Wheeler, 1997; Hettige, Dasgupta, & Wheeler,2000; Hettige, Huq, Pargal, & Wheeler, 1996; Huq & Wheeler, 1992;Pargal & Wheeler, 1996). Where formal regulators are present, commu-nities use the political process to influence the strictness of enforcement.Where regulators are absent or ineffective, nongovernmental organiza-tions (NGOs) and community groups (including religious institutions,social organizations, citizens’ movements, and politicians) pursue infor-mal regulation based on convincing polluters to conform to socialnorms. Although these groups vary from region to region, the patterneverywhere is similar: Factories negotiate directly with local actors inresponse to threats of social, political, or physical sanctions if they fail tocompensate the community or reduce emissions.

Indeed, communities sometimes resort to extreme measures whensufficiently provoked. In the Asian Survey, Robert Cribb (1990)recounted an Indonesian incident “reported from Banjaran near Jakartain 1980 when local farmers burned a government-owned chemical fac-tory that had been polluting their irrigation channels.” In a similar vein,Mark Clifford (1990) reported in the Far Eastern Economic Review thatcommunity action prevented the opening of a chemical complex inSouth Korea until appropriate pollution control equipment wasinstalled.

When factories respond directly to communities, the results may bearlittle resemblance to the dictates of formal regulation. For example,Cribb (1990) also cited the case of a cement factory in Jakarta, Indonesia,that, without admitting liability for the dust it generates, “compensateslocal people with an ex gratia payment of Rp. 5,000 and a tin of evapo-rated milk every month.” Agarwal, Chopra, and Sharma (1982)described a situation in India where, confronted by community com-plaints, an Indian paper mill installed pollution abatement equipment,and to compensate residents for remaining damage, the mill also con-structed a Hindu temple. If all else fails, community action can also trig-ger the physical removal of the problem. In Rio de Janeiro, Brazil, forexample, a neighborhood association protest against a polluting tanneryled managers to relocate it to the city’s outskirts (Stotz, 1991).

Rising income strengthens regulation. Countries regulate pollutionmore strictly as they get wealthier for three main reasons. First, pollutiondamage gets higher priority after rising wealth has financed basicinvestments in health and education. Second, higher income societieshave stronger regulatory institutions because technical personnel aremore plentiful, and budgets for monitoring and enforcement activitiesare more generous. Third, higher income and education empower localcommunities to enforce higher environmental standards, whateverstance is taken by the national government (Dasgupta & Wheeler, 1996;Pargal & Wheeler, 1996). The result is a very close relationship between



national pollution regulation and per capita income, as illustrated inFigure 7.5

Of course, regulation is only one determinant of pollution, which isthe focus of this article. Numerous theoretical articles have explored theoverall relationship between economic development and environmen-tal quality (Gruver, 1976; John & Pecchenino, 1994; John, Pecchenino,Schimmelpfennig, & Schreft, 1995; Lopez, 1994; McConnell, 1997;Selden & Song, 1995). This theoretical work has focused particularly onestablishing conditions for the existence of an environmental Kuznetscurve (EKC) or “inverted-U” relationship between environmental qual-ity and economic development. In the first stage of industrialization,pollution in the EKC world grows rapidly because regulation is weak,communities are too poor to pay for abatement, and people are far moreinterested in jobs and income than clean air and water. The balance

1

1.2

1.4

1.6

1.8

2

2.2

2.4

2.6

2 2.5 3 3.5 4 4.5

Log (Income Per Capita)

Log

(A

ir P

ollu

tion

Reg

ulat

ion)

Figure 7: Air Pollution Regulation and Per Capita IncomeSource: Dasgupta, Mody, Roy, and Wheeler (2001).

5. The measure of regulation in Figure 7 was developed from an analysis of countryreports on environmental management submitted to the United Nations. For a detaileddiscussion, see Dasgupta, Mody, Roy, & Wheeler (2001).

shifts as income rises because leading sectors become cleaner, peoplevalue the environment more highly, and regulatory institutions becomemore effective and responsive to a greater demand for environmentalquality. Along the curve, pollution levels off in the middle-income rangeand then falls toward preindustrial levels in wealthy societies. Numer-ous empirical studies have tested the EKC model by regressing cross-country measures of ambient air and water quality on various polyno-mial specifications of per capita income (Cole, Rayner, & Bates, 1997;Grossman & Krueger, 1995; Hettige, Lucas, & Wheeler, 1992; Holtz-Eatkin & Selden, 1995; Horvath, 1997; Panayotou, 1993, 1995; Rock, 1996;Selden & Song, 1995; Shafik, 1994; Shafik & Bandyopadhyay, 1992; Stern,Auld, Common, & Sanyal, 1998).6 The regressions, typically fitted tocross-sectional observations, suggest turning points in the range of U.S.$5,000 to U.S. $8,000 for several air and water pollutants.

These EKC estimates contrast sharply with the implications of therace-to-the-bottom model as well as the findings reported in this article.The race-to-the-bottom model implies that existing EKC estimates aretoo optimistic because they suggest that developing countries will even-tually reduce pollution by raising income. If a race to the bottom hasbegun, the conventional EKC is simply a snapshot of a dynamic processthat will raise the income-pollution relationship to a horizontal line atthe pollution level of the poorest, least regulated country.

In contrast, the results reported in this article suggest that conven-tional EKC estimates are too pessimistic. For example, the average urbanpollution index for China cited in Figure 2 comes from a more detailedstudy cited in Wheeler (1999), which shows that particulate pollutionhas declined since 1987 in both the coastal cities (the main recipients offoreign investment) and the interior cities of the country’s northernregion. According to recent World Bank data (expressed in constant 1995U.S. dollars), China’s gross national product (GNP) per capita grew fromU.S. $259 in 1985 to U.S. $711 in 1998. China’s apparent “turning point”for pollution (approximately U.S. $300) contrasts sharply with conven-tional EKC estimates in the range of U.S. $5,000 to U.S. $8,000. During thesame period, the GNPs per capita of Brazil and Mexico have been in arange near U.S. $4,000. During a period of rapid reduction in urban airpollution, both countries have experienced modest income growth to alevel that remains below the EKC’s conventional turning point range.

Local businesses control pollution because abatement reduces costs.Although public spirit moves a notable minority of firms to control pol-lution, most managers are bound by pressures from markets and share-holders. Through a variety of channels, regulatory and market forcesinduce managers to reduce costs by controlling pollution. Where formal


6. See Stern (1998) for an excellent overview of EKC studies.

regulation is well developed, financial penalties for excessive pollutioncan include charges, fines, and revenue losses from plant closures.Where formal regulation is not present, local communities can exactpenalties through political, social, and economic channels. Marketagents can also play an important role. Bankers may refuse to extendcredit because they are worried about liability; consumers may avoid theproducts of firms that are known to be heavy polluters. In response tothese factors, cost-minimizing managers will reduce pollution to thepoint at which the marginal cost of abatement equals the marginal pen-alty for polluting.

Polluting emissions also reflect managers’ technology decisions. Pol-lutants are unmarketed production residuals whose disposal createsenvironmental damage. Improved technologies that waste fewer rawmaterials therefore have an environmental advantage that complementstheir cost advantage. In the OECD countries, innovations induced bystricter regulations have generated significantly cleaner technologiesthat are available at incremental cost to producers in developing coun-tries. Even in weakly regulated economies, many firms have adoptedcleaner technologies simply because they are more profitable. Empiricalstudies have shown that firms in relatively open developing economiesadopt such technologies more quickly (Birdsall & Wheeler, 1993; Huq,Martin, & Wheeler, 1993).

Large multinational firms generally adhere to OECD environmental stan-dards in their developing-country operations. Multinational firms operateunder close scrutiny from consumers and environmental NGOs in theOECD economies. Although the influence of these groups is wellknown, recent research has suggested that investors also play an impor-tant role in encouraging clean production. Heavy emissions may signalto investors that a firm’s production techniques are inefficient. Investorsalso weigh potential financial losses from regulatory penalties and lia-bility settlements. Several studies have confirmed that the U.S. andCanadian stock markets react significantly to environmental news, gen-erating gains from good news and losses from bad news in the range of1% to 2% (see Hamilton, 1995; Klassen & McLaughlin, 1996; Lanoie &Laplante, 1994; Lanoie, Laplante, & Roy, 1997; Muoghalu, Robison, &Glascock, 1990).

According to a recent study of toxic polluters (Konar & Cohen, 1997),firms whose bad press has the greatest impact on stock prices subse-quently reduce emissions the most. Similar effects have been identifiedby recent research on new stock markets in Argentina, Chile, Mexico,and the Philippines (Dasgupta, Laplante, & Mamingi, 1997). In fact, theresponses are much larger than those reported for U.S. and Canadianfirms: Gains average 20% in response to good news and losses rangefrom 4% to 15% in the wake of bad news.


Multinationals have responded to such factors. A recent study of 89U.S.-based manufacturing and mining multinationals with branches indeveloping countries found that nearly 60% adhere to a stringent inter-nal standard that reflects OECD norms, and the others enforce local stan-dards (Dowell et al., 2000). Controlling for other factors (e.g., physicalassets, capital structure), the study found that firms with uniform inter-nal standards had an average market value U.S. $10.4 billion higher thantheir counterparts. To illustrate the implications for local environmentalquality, Figure 8 reports results from a careful audit of Indonesian facto-ries undertaken in 1995 (Afsah & Vincent, 1997). Almost 70% of domesticplants failed to comply with Indonesian water pollution regulations,whereas approximately 80% of the multinational plants were fullycompliant.

Implications of the Evidence

A large body of evidence suggests that the predictions of the race-to-the-bottom model are inaccurate because its assumptions are not realis-tic. Although pollution control costs matter to factory owners and man-agers, they are generally not a critical factor in location decisions. Inaddition, emissions are strongly affected by the increased availability ofclean technologies and the ubiquity of penalties for polluting. Evenwhere formal regulation is weak or absent, local communities usenumerous informal channels to penalize polluters when they suffer


0

20

40

60

80

100

Domestic Multinational

Plant Ownership

Superior

Compliant

Poor

Very BadCu

mu

lati

ve %

Figure 8: Indonesia: Plant Ownership and Compliance With Pollution RegulationsSource: Afsah and Vincent (1997).

from severe environmental damage. At the national level, governmentsdisplay a remarkably consistent tendency to tighten regulation asincomes grow. Within countries, regional differences in income and edu-cation also produce variations in community-based enforcement ofenvironmental norms. Among multinationals, scrutiny from customersand investors has led the majority of firms to standardize their environ-mental performance on OECD norms. The rest subscribe to local norms,which rise over time with income.

In this more realistic view of the world, an environmental race to thebottom appears extremely unlikely. In fact, the converse appears tooccur as “the bottom” rises with economic growth. The poorest societiespersistently improve their environmental quality as investmentincreases employment and income. Mutually reinforcing feedbackmechanisms at the local, national, and international levels producesincreasing pressures for pollution control as societies develop.

Although the evidence suggests that globalization has been generallycompatible with pollution reduction, several caveats are in order. First,to invert Keynes’s maxim, “in the short run, a lot of us might be dead.”Under rapid liberalization, a sudden increase of industrial investmentcould create pockets of severe pollution before national governments orlocal communities could respond effectively. This could occur even if“clean” multinationals expanded locally because domestic firms wouldalso be attracted by increased production and export opportunities. Sec-ond, the findings on particulate pollution in this article may not hold forother damaging pollutants. Comparable monitoring data on water pol-lution, pollution from heavy metals, and other contaminants are stillvery scarce in developing countries. Recent research also suggests thatthe situation may be quite different for global pollutants such as carbondioxide (CO2) that have no direct impact on the emitting society. Theymay escape the regulatory forces whose effects on local pollution aredescribed in this article. For example, the results of de Bruyn (1997) sug-gest that CO2 emissions are positively affected by economic growth.

Third, the apparent absence of a race to the bottom in particulate pol-lution does not imply that globalization has not had a chilling effect onregulation. PM-10 readings have declined in U.S. cities since 1985, butthey might have fallen more quickly in the absence of concerns aboutinvestment and jobs. Similarly, air quality improvements in China,Brazil, and Mexico might have been more rapid if policy makers had notbeen concerned about pollution abatement costs.7 These possibilitiescannot be tested directly because they involve a counterfactual condi-tion. Nevertheless, regulatory chill may well have retarded environmen-tal progress.


7. My thanks to an anonymous referee for raising this issue.

Fourth, communities’ capacities to control pollution formally orinformally depend on the quality of available information about emis-sions sources and damages. Some dangerous pollutants can be seen orsmelled, but others cannot be detected without specialized equipment.So, information gaps may well lead to much higher levels of contamina-tion than local communities would tolerate if they were better informed.

Fifth, globalization is almost certain to produce an increase in averagepollution intensity (emissions/output) as developing countries increasetheir shares of world industrial production. High-income countries havestricter formal and informal regulations than low-income countries, soproduction in the latter has higher pollution intensity (on average;exceptions were noted in the previous section). To illustrate, Figure 9summarizes a recent econometric result from a cross-country analysis oforganic water pollution: Relative to the highest income countries, thepoorest countries have an approximately tenfold differential in pollu-tion intensity. The biggest improvements come relatively early in thegrowth process, as countries grow from approximately U.S. $500 percapita to U.S. $3,000 per capita. After that, the marginal improvementslevel off considerably (Mani & Wheeler, 1998). By implication, shifting aunit of production from a high-income country to a poor economy willsubstantially increase the pollution intensity of that unit.

The air monitoring data in Figures 2 to 5 reveal the consequences ofdifferential pollution intensity. In the late 1990s, Cubatão’s PM-10 con-centration was approximately 80 µg/m3, whereas PM-10 concentrationsin most U.S. cities were below 30 µg/m3. After a decade of decline,China’s urban SPM readings were still approximately 300 µg/m3 in


0

10

20

30

40

50

60

70

80

90

100

0 2000 4000 6000 8000 10000 12000 14000 16000 18000 20000

Per Capita Income ($US)

Po

lluti

on

Inte

nsi

ty In

dex

Figure 9: Industrial Pollution Intensity and Economic DevelopmentSource: Mani and Wheeler (1998).

1995, far higher than readings in OECD cities. This differential may havecreated the mistaken view that globalization is creating a race to the bot-tom, even though air quality seems to be improving in countries at allincome levels. If current trends continue, the bottom will continue torise, and international average pollution intensities will begin to fall atsome point in the future. The critical point for the globalization debate isthat cities in developing countries have actually improved their environ-mental quality during a period of rapid liberalization and industrialgrowth.

Conclusions and Policy Implications

Alarge body of evidence suggests that for locally damaging air pollu-tion at least, there will be no environmental race to the bottom for twomain reasons. First, communities in developing countries are neitherpassive agents nor focused exclusively on material gain. Empoweredwith good information about the benefits and costs of environmentalprotection, they will act to protect their own interests. As their incomesand education levels improve, they will control pollution more strictly.Second, consumers and investors assign significant value to environ-mental performance, and if they are well informed, their market deci-sions will provide powerful incentives to reduce pollution. On bothcounts, the most plausible long-run forecast is for rising, not falling,environmental quality in both high- and low-income economies. A sig-nificant gap may remain, but pollution damage should decline in poorcountries as they develop.

Although this news is good for the global environment in the longrun, adjustment to a cleaner world is not likely to be smooth. Countrieswhose economic policies induce a rapid expansion of income andemployment may also experience severe environmental damage unlessdirect measures are taken to accelerate regulation’s positive long-runresponse to income growth. On the macropolicy front, the persistence ofa regulatory gap between rich and poor countries may lead to continu-ing controversy over trade policy and international assistance strategy.OECD interest groups that support protectionist measures for other rea-sons may continue to invoke the race-to-the-bottom model, relying on acommon misperception that the regulatory gap automatically implies arace to the bottom. Despite clear evidence to the contrary, they may con-tinue to argue that a race to the bottom can only be avoided through theenforcement of uniform environmental standards in all countries. Lack-ing any direct means of creating such a regime, interest groups mayargue for trade restrictions and aid cutbacks until poor countries close


the gap. The available evidence suggests that such measures will retard,not advance, the day when the gap actually disappears.

In summary, the basic assumptions of the race-to-the-bottom modelare contradicted by a large body of empirical research. Its flaws invali-date its main conclusion (the inevitability of the race to the bottom) andits main policy prescription (uniform standards enforced by any meansnecessary). Abandoning this theory, however, does not imply that poorcountries must resign themselves to bad environmental quality for anextended period. Several recent benefit-cost analyses have made per-suasive cases for stricter pollution control, even in very low-incomeeconomies. In China, for example, a recent study has shown that the eco-nomic returns to pollution abatement would justify significant tighten-ing of regulation (Dasgupta et al., 1997). Similar studies in Indonesia(Calkins, 1993) and Brazil (Von Amsberg, 1997) have produced similarconclusions.

How can environmental quality improvement be accelerated in theera of globalization? Recent international experience has identified fourkeys to rapid progress:

1. Sustained support for programs that provide public, easily accessible informationabout polluters, pollution damages, local environmental quality, and the cost ofpollution abatement. Such programs significantly improve the abilities oflocal communities to protect themselves, national regulators to enforcedecent environmental standards, and market agents to reward cleanfirms and punish heavy polluters. International institutions such as theWorld Bank have begun supporting this idea in collaborative programswith environmental agencies in Indonesia, the Philippines, China, Thai-land, Vietnam, Mexico, Colombia, Brazil, and elsewhere.8

2. Sustained support for development of stronger regulatory institutions and cost-effective measures to reduce pollution. Sustained support is critical becauseinstitutional development takes time. Although private sector clean-production initiatives can play a valuable role, only public sector institu-tions can protect society’s general interest in a cleaner environment.

3. Rejection of trade and aid sanctions as levers to force closure of the regulatory gapbetween low- and high-income countries. First, such sanctions are unjustbecause they fail to discriminate between clean and dirty firms in theaffected countries. Numerous studies have shown that factories withworld-class standards are operating even in the poorest countries (Afsah &Vincent, 1997; Hartman et al., 1997; Huq & Wheeler, 1992; Wheeler et al.,1999). Second, such blunt instruments will inevitably penalize workers inpoor countries by reducing opportunities for jobs and higher wages.Finally, they will not work anyway. As noted in previous sections of thisarticle, poor countries have weaker regulations and higher pollutionintensities for a host of reasons. Governments of low-income countries


8. For more information about these programs, see New Ideas in Pollution Regulation(2001).

could not deliver on promises of OECD-level regulation, even if theywere willing to make them.

4. Willingness by the World Bank, the International Monetary Fund, and otherinstitutions to take explicit account of environmental risks in the design andimplementation of adjustment operations and other economic reform programs.Rapid structural change could inflict severe pollution damage on somelocalities unless public environmental information and regulation keeppace with changing conditions. Willingness to recognize and compensatefor this risk with targeted programs will greatly strengthen the credibilityof international institutions that support continued economic liberaliza-tion in the name of sustainable development.

Manuscript submitted November 14, 2000; revised manuscript accepted for publicationMarch 19, 2001.

References

Afsah, S., & Vincent, J. R. (1997, March). Putting pressure on polluters: Indonesia’s PROPERprogram (A Case Study for the Harvard Institute for International Development 1997Asia Environmental Economics Policy Seminar). Cambridge, MA: Harvard Institutefor International Development.

Agarwal, A., Chopra, R., & Sharma, K. (1982). The state of India’s environment, 1982. NewDelhi, India: Centre for Science and Environment.

Albrecht, J. (1998). Environmental policy and inward investment position of U.S. dirtyindustries. Intereconomics, 33(4), 186-194.

Birdsall, N., & Wheeler, D. (1993). Trade policy and industrial pollution in Latin America:Where are the pollution havens? Journal of Environment & Development, 2, 137-149.

Bonior, D. (1999, December). Defending democracy in the new global economy (Statement to anAFL-CIO conference on workers’ rights, trade development, and the World TradeOrganization, Seattle, WA) [Online]. Available: http://www.ourfuture.org/readarticle.asp?ID=548

Calkins, R. (1994). Indonesia: Environment and development. Washington, DC: World Bank.Clifford, M. (1990, October 18). Kicking up a stink: South Korean government reels from

anti-pollution backlash. Far Eastern Economic Review, 72-73.Cole, M. A., Rayner, A. J., & Bates, J. M. (1997). The environmental Kuznets curve: An

empirical analysis. Environment and Development Economics, 2(4), 401-416.Companhia de Tecnologia de Saneamento Ambiental (CETESB). (1986). Restoring the Serra

do Mar. São Paulo, Brazil: Author.Companhia de Tecnologia de Saneamento Ambiental (CETESB). (1990). Cubatão: A change

of air. São Paulo, Brazil: Author.Companhia de Tecnologia de Saneamento Ambiental (CETESB). (1994). Acao da CETESB

em Cubatão: Situacao em Junho de 1994. São Paulo, Brazil: Author.Companhia de Tecnologia de Saneamento Ambiental (CETESB). (2001). Home page

[Online]. São Paulo, Brazil: Author. Available: http://www.cetesb.sp.gov.brCribb, R. (1990). The politics of pollution control in Indonesia. Asian Survey, 30, 1123-1135.Daly, H. (2000, August). Globalization. Paper presented at the 50th Anniversary Conference

of the Aspen Institute, Aspen, Colorado.Dasgupta, S., Laplante, B., & Mamingi, N. (1997, October). Capital market responses to envi-

ronmental performance in developing countries (World Bank Development ResearchGroup Working Paper No. 1909). Washington, DC: World Bank.


Dasgupta, S., Lucas, R., & Wheeler, D. (1998, December). Small manufacturing plants, pollu-tion and poverty: New evidence from Brazil and Mexico (World Bank DevelopmentResearch Group Working Paper No. 2029). Washington, DC: World Bank.

Dasgupta, S., Mody, A., Roy, S., & Wheeler, D. (2001). Environmental regulation and devel-opment: A cross-country empirical analysis. Oxford Development Studies, 29(2).

Dasgupta, S., Wang, H., & Wheeler, D. (1997, October). Surviving success: Policy reform andthe future of industrial pollution in China (World Bank Policy Research DepartmentWorking Paper No. 1856). Washington, DC: World Bank.

Dasgupta, S., & Wheeler, D. (1996, November). Citizen complaints as environmental indica-tors: Evidence from China (World Bank Policy Research Department Working Paper No.1704). Washington, DC: World Bank.

de Bruyn, S. M. (1997). Explaining the environmental Kuznets curve: Structural changeand international agreements in reducing sulphur emissions. Environment and Develop-ment Economics, 2(4), 485-503.

Dowell, G., Hart, S., & Yeung, B. (2000). Do corporate global environmental standards cre-ate or destroy market value? Management Science, 46(8) 1059-1074.

Environmental Protection Agency, Office of Air Quality Planning and Standards. (2001).1997 national air quality and emissions trends report: List of tables [Online]. Washington,DC: Author. Available: http://www.epa.gov/oar/aqtrnd97/tables.html

Eskeland, G., & Harrison, A. (1997). Moving to greener pastures? Multinationals and the pollu-tion-haven hypothesis (World Bank Policy Research Working Paper No. 1744). Washing-ton, DC: World Bank.

Grossman, G. M., & Krueger, A. B. (1995). Economic growth and the environment. Quar-terly Journal of Economics, 110(2), 353-377.

Gruver, G. W. (1976). Optimal investment in pollution control capital in a neoclassicalgrowth context. Journal of Environmental Economics and Management, 5, 165-177.

Hamilton, J. (1995). Pollution as news: Media and stock market reactions to the toxicrelease inventory data. Journal of Environmental Economics and Management, 28, 98-103.

Hartman, R., Huq, M., & Wheeler, D. (1997, January). Why paper mills clean up: Determinantsof pollution abatement in four Asian countries (World Bank Policy Research DepartmentWorking Paper No. 1710). Washington, DC: World Bank.

Hettige, H., Huq, M., Pargal, S., & Wheeler, D. (1996). Determinants of pollution abatementin developing countries: Evidence from South and Southeast Asia. World Development,24(12), 1891-1904.

Hettige, H., Lucas, R.E.B., & Wheeler, D. (1992). The toxic intensity of industrial produc-tion: Global patterns, trends and trade policy. American Economic Review Papers and Pro-ceedings, 82, 478-481.

Hettige, M., Dasgupta, S., & Wheeler, D. (2000). What improves environmental compli-ance? Evidence from Mexican industry. Journal of Environmental Economics and Manage-ment, 39(1), 39-66.

Holtz-Eakin, D., & Selden, T. M. (1995). Stoking the fires? CO2 emissions and economicgrowth. Journal of Public Economics, 57, 85-101.

Horvath, R. J. (1997). Energy consumption and the environmental Kuznets curve debate. Sydney,Australia: University of Sydney, Department of Geography.

Huq, M., Martin, P., & Wheeler, D., (1993, January). Process change, economic policy, andindustrial pollution: Cross-country evidence from the wood pulp and steel industries. Paperpresented at the annual meetings of the American Economic Association, Anaheim,California.

Huq, M., & Wheeler, D. (1992). Pollution reduction without formal regulation: Evidence fromBangladesh (World Bank Environment Department Working Paper No. 1993-39). Wash-ington, DC: World Bank.

Jaffe, A., Peterson, S., Portney, P., & Stavins, R. (1995). Environmental regulation and thecompetitiveness of U.S. manufacturing: What does the evidence tell us? Journal of Eco-nomic Literature, 33, 132-163.


Janicke, M., Binder, M., & Monch, H. (1997). “Dirty industries”: Patterns of change inindustrial countries. Environmental and Resource Economics, 9, 467-491.

Jha, V., Markandya, A., & Vossenaar, R. (1999). Reconciling trade and the environment: Lessonsfrom case studies in developing countries. Cheltenham, UK: Edward Elgar.

John, A., & Pecchenino, R. (1994). An overlapping generations model of growth and theenvironment. Economic Journal, 104, 1393-1410.

John, A., Pecchenino, R., Schimmelpfennig, D., & Schreft, S. (1995). Short-lived agents andthe long-lived environment. Journal of Public Economics, 58, 127-141.

Khalid, R., & Braden, J. B. (1993). Welfare effects of environmental regulation in an openeconomy: The case of Malaysian palm oil. Journal of Agricultural Economics, 44, 25-37.

Klassen, R. D., & McLaughlin, C. P. (1996). The impact of environmental management onfirm performance. Management Science, 42(8), 1199-1214.

Konar, S., & Cohen, M. (1997). Information as regulation: The effect of community right toknow laws on toxic emissions. Journal of Environmental Economics and Management, 32,109-124.

Lanoie, P., & Laplante, B. (1994). The market response to environmental incidents in Can-ada: A theoretical and empirical analysis. Southern Economic Journal, 60, 657-672.

Lanoie, P., Laplante, B., & Roy, M. (1997, April). Can capital markets create incentives for pollu-tion control? (World Bank Policy Research Department Working Paper No. 1753). Wash-ington, DC: World Bank.

Levinson, A. (1997). Environmental regulations and industry location: International anddomestic evidence. In J. N. Bhagwati & R. E. Hudec (Eds.), Fair trade and harmonization:Prerequisites for free trade? (pp. 429-458). Cambridge, MA: MIT Press.

Lopez, R. (1994). The environment as a factor of production: The effects of economicgrowth and trade liberalization. Journal of Environmental Economics and Management, 27,163-184.

Mani, M., & Wheeler, D. (1998). In search of pollution havens? Dirty industry in the worldeconomy, 1960-1995. Journal of Environment & Development, 7, 215-247.

McConnell, K. E. (1997). Income and the demand for environmental quality. Environmentand Development Economics, 2(4), 383-399.

Muoghalu, M., Robison, D., & Glascock, J. (1990). Hazardous waste lawsuits, stockholderreturns, and deterrence. Southern Economic Journal, 7, 357-370.

Nader 2000. (2000). Environment quotes [Online]. Washington, DC: Author. Available:http://www.votenader.com/issues/environment.html

National Environment Ministry. (2001). Calidad del aire en la zona metropolitana del valle deMéxico [Online]. Mexico City: Author. Available: http://www.ine.gob.mx/dggia/cal_aire/libros/segu_info/mexico.html

New Ideas in Pollution Regulation. (2001). Home page [Online]. Available: http://www.worldbank.org/nipr/

Panayotou, T. (1993). Empirical tests and policy analysis of environmental degradation at differ-ent stages of economic development (Working Paper WP238). Geneva, Switzerland: Inter-national Labor Office, Technology and Employment Programme.

Panayotou, T. (1995). environmental degradation at different stages of economic develop-ment. In I. Ahmed & J. A. Doeleman (Eds.), Beyond Rio: The environmental crisis and sus-tainable livelihoods in the Third World. London: Macmillan.

Pargal, S., & Wheeler, D. (1996). Informal regulation of industrial pollution in developingcountries: Evidence from Indonesia. Journal of Political Economy, 104(6), 1314-1327.

Rock, M. T. (1996). Pollution intensity of GDP and trade policy: Can the World Bank bewrong? World Development, 24, 471-479.

Selden, T., & Song, D. (1995). Neoclassical growth, the J curve for abatement, and theinverted U curve for pollution. Journal of Environmental Economics and Management, 29,162-168.

Shafik, N. (1994). Economic development and environmental quality: An econometricanalysis. Oxford Economic Papers, 46, 757-773.


Shafik, N., & Bandyopadhyay, S. (1992). Economic growth and environmental quality: Timeseries and cross-country evidence (Background Paper for the World Development Report1992). Washington, DC: World Bank.

Stern, D. I. (1998). Progress on the environmental Kuznets curve? Environment and Develop-ment Economics, 3, 175-198.

Stern, D. I., Auld, A., Common, M. S., & Sanyal, K. K. (1998). Is there an environmentalKuznets curve for sulfur? (Working Papers in Ecological Economics, No. 9804). Canberra:Australian National University, Center for Resource and Environmental Studies.

Stotz, E. (1991). Luta pela saúde ambiental: A AMAP contra Cortume Carioca, S.A. Unaexperiencia vitoriosa. In V. V. Valla & E. N. Stotz (Eds.) Participação popular, educação esaúde: Teoria e práctica (pp. 133-160). Rio de Janeiro, Brazil: Relume Dumará.

Tobey, J. A. (1990). The effects of domestic environmental policies on patterns of worldtrade: An empirical test. Kyklos, 43(2), 191-209.

Van Beers, C., & Van den Bergh, J.C.J.M. (1997). An empirical multi-country analysis of theimpact of environmental regulations on trade flows. Kyklos, 50(1), 29-46.

Von Amsberg, J. (1997, June). Brazil: Managing pollution problems, the brown environmentalagenda (World Bank Report #16635-BR). Washington, DC: World Bank.

Wheeler, D. (1999). Greening industry: New roles for communities, markets and governments.New York: Oxford/World Bank.

Xing, Y., & Kolstad, C. (1995, May). Do lax environmental regulations attract foreign invest-ments? (Working Papers in Economics). Santa Barbara, CA: University of California.

Xu, X., Gao, J., Dockery, D., & Chen, Y. (1994). Air pollution and daily mortality in residen-tial areas of Beijing, China. Archives of Environmental Health, 49(4), 216-222.

David Wheeler is lead economist in the Infrastructure/Environment Unit of the World Bank’sDevelopment Research Group. He received his undergraduate degree from Princeton University(1968) and his Ph.D. in economics from the Massachusetts Institute of Technology (MIT) (1974).Before joining the World Bank in 1990, Wheeler was a tenured associate professor of economics atBoston University (1976 to 1990). He has also been a visiting professor at MIT’s Department ofUrban Studies and Planning (1978 to 1979) and the National University of Zaire (1973 to 1975);director of the Development Studies Project in Jakarta, Indonesia (1987 to 1989); and cofounder ofthe Boston Institute for Developing Economies. After joining the World Bank, Wheeler was asked toestablish the Environment Unit in the Development Research Group. As lead economist, he directs ateam that works on environmental policy and research issues in collaboration with policy makersand academics in Brazil, Colombia, Mexico, China, India, Indonesia, the Philippines, and otherdeveloping countries. He has published numerous books and articles on issues related to develop-ment and the environment.


Documents

Racing to the Bottom? Foreign Investment and Air Pollution ...sabrash/110/Chem... · wheeler / air pollution in developing countries 227 investment (FDI) among developing countries