THE TRADE AND ENVIRONMENT DEBATE
1.1 Globalization and the Trade versus Environment Debate
"It was the best of times, it was the worst of times." This line, written by Charles Dickens over 100 years ago, captures the present-day divide between supporters and critics of globalization. During the 1990s, North America and much of Europe enjoyed its longest peacetime expansion, unemployment rates hit historic lows, and real income growth in much of the developing world soared. To many these are the fruits of globalization. But this same decade saw little progress in addressing climate change, a decline in fish and tropical forest stocks, and by some measures, rising inequality in the world distribution of income. To many others, these are its costs.
Debates over "globalization" have been going on for some time. But nowhere has the divide between the two views of globalization been more apparent than in recent discussions concerning trade liberalization and the environment.
For the last ten years environmentalists and the trade policy community have squared off over the environmental consequences of liberalized trade.1 This debate was fueled by negotiations over the North American Free Trade Agreement and the Uruguay round of General Agreement on Tariffs and Trade (GATT) negotiations, both of which occurred at a time when concerns over global warming, species extinction, and industrial pollution were rising. The debate intensified with the creation of the World Trade Organization (WTO) and proposals for future rounds of trade negotiations.
Trade negotiators saw the WTO as a step forward because of its improved dispute settlement procedures and because it closed loopholes in previous trade agreements. Environmentalists, however, were disturbed by the intrusion of trade agreements into what many thought were purely domestic matters. Perhaps not surprisingly, then, an attempt to initiate a new round of multilateral trade negotiations in Seattle became a flashpoint for growing unrest with globalization and efforts at further trade liberalization.
The purpose of this book is to study the interaction between international trade and the environment using both economic theory and empirical analysis. Our objective is to move the discussion forward by developing useful theory and devising methods to help in the empirical estimation of key magnitudes. We hope to enlist readers in further discussion and evaluation of trade's environmental effects, and this book is designed to equip them with the tools for doing so. In the end, differences of opinion will of course remain because the effects of international trade on the environment are still not fully understood. But we hope to give researchers and policymakers a common language and framework within which to discuss, debate, and continue their investigations.
1.2 Two Questions and a Preview of Our Answers
Throughout the book, we focus on two key channels through which trade can affect the environment. The first is via its effects on the level, or scale, of economic activity. If trade spurs economic activity, then the pure income-generating effects of trade may be harmful to the environment. The second channel is via a composition effect--a change in the mix of economic activity in countries, caused by trade. Many environmentalists are concerned that trade may lead to a shifting of polluting industry from rich to poor countries, and this global composition effect may also raise world pollution.
As we will see, both of these channels are more complex than these simple arguments suggest, and this leads to the two key questions that unify the book:
How does the increase in economic activity induced by international trade affect the environment?
Many in the "deep green" environmental movement believe unfettered access to world markets is necessarily harmful to the environment. While arguments differ in the details, the primary objection is that international trade leads to a greater scale of economic activity--be it transportation services, more production of goods and services, or more consumption--and that economic activity per se harms the environment. In this view, if international trade stimulates economic activity and if this activity is inherently environmentally damaging, then so too is international trade.
This concern forces us to think about the link between economic activity and environmental quality more generally and not just with regard to international trade. It leads us to define a measure of the scale of economic activity and to then link this measure to changes in the economy brought about by trade liberalization. But since any change in the scale of economic activity also affects incomes, we must also take into account the impact of income gains on environmental regulation.
If higher real incomes generate a greater ability and willingness to implement and enforce environmental regulations, then the logical chain linking trade liberalization to environmental destruction is broken. A trade liberalization that raises the scale of economic activity will then also lower the dirtiness of production techniques, and its full environmental impact can only be resolved through careful empirical investigation.
We address this first question by using our theoretical framework to separate the impact of economic growth on the environment from that caused by trade liberalization.2 We show that the relative strength of scale versus technique effects depends on how government policy is formed and how quickly it changes in response to new conditions. We also set out the theoretical conditions under which either the scale or the technique effect is stronger. But since theory alone cannot determine the answers to our questions, we also present empirical work undertaken using a large crosscountry data set on measured sulfur dioxide concentrations in over 100 major cities in the world.
By isolating the pure scale and technique effects of trade,3 we estimate that a trade liberalization that raises the scale of economic activity by 1% raises pollution concentrations by 0.25 to 0.5% via the scale effect, but the accompanying increase in incomes per capita drives concentrations down by 1.25-1.5% via a technique effect. These estimates imply a strong policy response to trade-inspired income gains. As we show, they also imply that economic growth created by neutral technological progress will both raise real incomes and improve environmental quality, but economic growth fueled by capital accumulation alone will worsen the environment.
Our estimates and analysis are important in establishing that we need to identify the source of income gains before we predict its environmental consequences. For example, neutral technological progress favors no industry, and we find that it improves environmental quality through its role in raising incomes and tightening techniques of production. And while capital accumulation raises both incomes and scale, it favors the production of dirty, capital-intensive processes. Hence capital accumulation creates an additional effect leading to a worsened environment. Trade liberalization can be environmentally friendly since it brings income gains and will lead some countries to specialize in relatively clean industries. But not all countries can specialize in the production of clean goods, and hence it is important to determine which countries are likely to specialize and export relatively dirty products. This brings us to our second question.
2. How does environmental policy affect a nation's trade pattern?
This second question arises from concerns that dirty industries will leave tightly regulated countries and migrate to countries with lax regulations. As a result, international trade could alter the composition of output across countries, leading poor countries with relatively weak environmental regulation to specialize in the production of dirty goods while rich and tightly regulated countries specialize in clean goods. Even if trade liberalization had no effect on the scale of economic activity or on the dirtiness of the techniques of production, it could create pollution havens in the developing world by altering the composition of their output toward dirty goods.
Dirty industry migration is a serious concern because it would imply that poor less developed countries are bearing the pollution burdens of rich developed country consumption. Despite the claims of some economists that this may well be efficient, it would be unpalatable to many. Dirty industry migration may also raise concerns about competitiveness and cause regulatory chill in the developed countries which could slow down ongoing efforts to raise environmental protection. At worst it could usher in a worldwide race to the bottom in environmental protection as nations relax standards to forestall dirty industry migration.
These are valid concerns that need to be seriously addressed. To examine dirty industry migration and the creation of pollution havens we investigate how differences across countries in both environmental regulations and other country characteristics interact to determine a country's trade pattern. We also have to isolate changes in the composition of output created by trade liberalization from those created by more mundane sources such as taste changes or ongoing growth in a nation's productive capacity. To do so we employ our theoretical framework to isolate the composition effect of trade liberalization on pollution.
Isolating changes in the composition of output created by trade is critical to examining whether lax-regulation countries are destined to become pollution havens. While this may appear obvious to some, it too is an empirical question. While developed and less developed countries differ widely in the stringency of their environmental regulations, they also differ widely in average education levels, in available infrastructure, and in capital equipment per person. If these other differences are significant determinants of production costs, then it is no longer clear that lax regulations alone create a cost advantage in dirty good production.
In fact, our empirical work indicates that greater access to international markets creates only relatively small changes in pollution via the composition effect, and that conventional determinants of production costs are more important in determining "international competitiveness" than are differences in meeting the costs of environmental regulation. We find little evidence for dirty industry migration, or the pollution haven hypothesis. Our empirical results suggest just the opposite: that relatively rich developed countries have a comparative advantage in dirty goods. As a result, freer international trade shifts dirty good production from lax-regulation countries to more stringent-regulation countries. If this is correct, then the global composition effect of trade lowers world pollution. Combining our estimates of scale, composition, and technique effects created by a trade liberalization yields a surprising conclusion: freer trade appears to be good for the environment--at least for the average country in our sample and for the pollutant we consider.
While this result will be of interest to policymakers because of the importance of sulfur dioxide pollution and its close connection to other equally noxious pollutants, the most important contribution of this book is the theoretical framework it presents, the methods it espouses, and its discussion of competing hypothesis linking international trade to environmental outcomes.
1.3 Our Method of Analysis
We do not present an exhaustive account or review of all that is known about international trade and the environment, but instead develop a simplified but cohesive framework to investigate the relationship. We show that our framework is useful for understanding the links between economic growth and the environment, useful in disentangling the many different motives for trade in environmentally damaging goods, and useful as a springboard for empirical work that estimates the environmental consequences of liberalized trade. The strength of the book lies in the consistent application of our theoretical approach to various questions and our integration of this approach with empirical work.
Some readers may prefer different methods to discuss different issues, or wish for a presentation less encumbered by formal theory. We chose to be constrained by our theoretical framework because much of the current debate is not constrained by either theory or empirical work, and this has produced an astonishingly high ratio of rhetoric to results. Strong ideologies are at play here. There are those for whom free trade is a goal in and of itself and who speak of it in reverential tones; and there are also those who scorn any exercise to evaluate trade's environmental impact because they view its costs as clearly self-evident. This is an area of public policy debate sorely in need of guidance from further theory and empirical work. Accordingly, our goal is to introduce the reader to a new set of issues and build understanding. To do so we use theory to identify the main forces at work, motivate our empirical approach, and constrain the conclusions that we can draw.
While the appeal of doing "international environmental economics" has surely been present before, various difficulties face those who attempt it. The key difficulty is in introducing public goods into a general equilibrium model rich enough to explore the implications of trade, yet tractable enough to examine policy and to serve as the basis for empirical work.
We have attempted to strike a constructive balance between model complexity and tractability. Our analysis is marked by three characteristics: we adopt general equilibrium models; we assume environmental policy can be altered in light of changed economic conditions; and we adopt relatively simple economic models where the economy is aggregated into only two sectors. Throughout we examine only local pollution, although our methods can be extended to consider transboundary or global pollution.4
General equilibrium methods are necessary because they allow us to consider the full implications of international trade on the environment. For example, a partial equilibrium model of a clean industry might predict that the clean industry expands as a result of trade liberalization. Since the clean industry does not pollute, the partial equilibrium model does not predict any effects on the environment. But a general equilibrium analysis recognizes that as the clean industry expands, it must draw resources from other parts of the economy. If these other parts of the economy pollute, then the expansion of a clean industry can lead to a fall in the country's pollution.
In addition, a general equilibrium analysis allows us to grapple with income effects, which have played a central role in the debate over the effects of trade and growth on the environment. If trade stimulates a dirty industry, it will tend to increase pollution. But if trade also raises real incomes via general equilibrium effects, then it will increase the demand for environmental quality, and this can have a dampening effect on the increase in pollution via an endogenous policy response.
Some of the tools we employ are commonly used in competitive general equilibrium trade theory, but will be unfamiliar to many readers.5 In an attempt to engage those unfamiliar with general equilibrium methods, we develop special cases of our models in some chapters to provide explicit solutions and derivations of main results. The examples we provide have the twin purpose of clarifying and simplifying the more general analysis we present elsewhere, and on occasion demonstrating surprising counterexamples.
Endogenous policy is necessary if we are to capture the response of pollution policy to rising incomes and changing prices brought about by free trade. Much of the literature has focused on the role of differences in environmental policy between rich and poor countries in influencing the trade pattern and in determining the effects of trade on the environment. But if policy differences are caused in part by income differences, and if trade affects incomes, then we cannot evaluate the long-run effects of trade liberalization without taking into account the effects of changes in income and relative prices on environmental policy.
To facilitate a simple analysis of endogenous policy, we assume that the government maximizes an objective function that reflects the weighed sum of preferences of agents in the economy. In many cases, we assume all agents are identical; however, we will also consider a simple political economy model where the government places different weights on "Brown" consumers who benefit from dirty good production than it places on "Green" consumers who benefit mainly from clean good production.
While endogenous and fully responsive policy is a useful benchmark, it may not always reflect real-world conditions. Environmental economics was born out of the recognition of market failure and imperfect policy. Accordingly, wherever possible we present our analysis first under the assumption that policy is rigid and imperfect, and then present the case of optimal and flexible policy. This is useful not only because it may represent a reasonable representation of short- versus long-run outcomes, but also because the rigid policy analysis is often a pedagogically useful precursor to the analysis of optimal policy.
Finally, throughout the book, we limit the complexity of our models. Previous work in this area has sometimes been hampered by a well-intentioned effort to use very general models. This can often lead to complicated analysis in which many of the results are ambiguous. Our goal in the book is to introduce the reader to a new set of issues and build understanding. Relatively simple economic models can do just that--they shed light on different questions, they provide insight, and they guide--but not completely determine--the direction of empirical work. There are surely more general formulations that will overturn some of our results and introduce other complications, and we encourage readers to pursue these advances. But given the severe limitations researchers face in obtaining environmental data, adding further theoretical nuances will be for naught if we lack data necessary to explore their strength or validity. At present, getting the simple logic right seems a worthy goal. We leave it to readers and future researchers to elaborate on our findings.
1.4 Plan of the Book
We begin by developing our two-sector general equilibrium pollution and trade model in chapter 2. One special case of the model is a simplified version of the pollution haven model in Copeland and Taylor (1994). Another special case gives us the canonical Heckscher-Ohlin model that links relative factor abundance to international trading patterns. Chapter 2 is a difficult one, as it develops in detail the model's theoretical structure, starting with a consideration of pollution as a joint product of output and ending with a discussion of optimal pollution policy using the constructs of pollution demand and supply. Readers may wish to skip certain sections of this chapter, safe in the knowledge they may return to clarify a derivation or definition that they find puzzling in a later chapter.
After laying out the theoretical foundation in chapter 2 we proceed in chapter 3 to investigate the environmental Kuznets curve. We start with an examination of growth because the environmental consequences of trade and growth share some common features and any credible analysis of the effects of trade on the environment has to disentangle changes in pollution caused by trade from those caused by growth.
This chapter demonstrates that our simple pollution demand and supply framework can be employed to discuss the most commonly cited explanations for the environmental Kuznets curve. But this chapter is more than review and exercise. We present new theoretical results detailing the conditions under which strong policy responses alone can generate an EKC, new results linking increasing returns to abatement (at the industry level) to the EKC, and new results regarding the link between the sources of economic growth and environmental quality.
Following our discussion of growth, we proceed to trade liberalization in chapter 4. We identify the scale, composition, and technique effects created when a country liberalizes its trade and find that the positive and normative effects of trade liberalization depend critically on whether a country is a dirty good importer or exporter. We consider both a reduction in transport costs and a fall in tariffs as the motivations for further trade and introduce a simple political economy model as well.
In chapters 5 and 6 we turn to a discussion of the pattern of trade. We start in chapter 5 by demonstrating that differences across countries in pollution policy alone can lead to the pollution haven hypothesis. But since policy differences should arise endogenously from more basic assumptions about country characteristics, we then link differences in the stringency of pollution regulation to cross-country differences in institutions, in income levels, and in the fragility of environments. In each case, we demonstrate how differences in country characteristics lead to differences in the stringency of regulation and hence trade. Although in each case free trade results in the lax-regulation countries exporting dirty products, we show that this trade need not be either welfare reducing or environmentally damaging.
In chapter 6 we broaden the potential motives for trade to demonstrate that differences across countries in other characteristics can also influence relative production costs. We focus on differences across countries in their capital stocks and labor forces because these differences are at the basis of the factor endowments hypothesis. We examine how differences in income levels and factor endowments interact to determine the pattern of trade. This chapter contains several new results. We demonstrate that rich but capital-abundant countries may export dirty goods, while poor and capital-scarce countries export clean goods. We show that when differences in other country characteristics lead to cost advantages that overwhelm the pollution haven motives for trade, many of the dire consequences of international trade disappear. We demonstrate that world pollution can fall with trade, that imperfectly regulated and poor developing countries must both gain from trade and see an improvement in their environment, and that pollution may fall in both rich and poor countries with trade liberalization. All of these results are in direct opposition to predictions of the pollution haven hypothesis.
In chapter 7 we present empirical work estimating the strength of scale, composition, and technique effects. We employ many of the results developed in earlier chapters to develop a simple reduced form estimating equation linking pollution concentrations to country characteristics, and measures of both openness to trade and comparative advantage. This equation is then estimated on a large panel data set containing measures of sulfur dioxide pollution concentrations drawn from cities in 44 developed and developing countries over the 1971 to 1996 period.
Chapter 8 presents a short conclusion and suggestions for future research. Some of these suggestions are topics that we had hoped to include but did not because of limited time or space. Others are suggestions for further empirical or theoretical work addressing unanswered questions regarding linkages between trade and the environment.
Return to Book Description
File created: 8/7/2007
Questions and comments to: email@example.com
Princeton University Press