The Pollution Haven Hypothesis (PHH)

1. The Pollution Haven Hypothesis (PHH) November 8, 2007

2. The main question • Does trade liberalization worsen environmental problems? Many people think that the answer is “yes”, and base their opposition to trade liberalization on this belief. They also worry that trade makes it more difficult to impose environmental laws – and that trade makes these laws less effective. • How do we approach these issues? We need to know exactly what we mean by the question(s), and we need some theory to help organize our thoughts.

3. Outline of lecture • Describe Pollution haven hypothesis (PHH) • Review theory of the second best and the Principal of targeting. • Review reasons why countries trade. • Discuss examples in which increased trade or foreign direct investment (FDI) can harm or benefit the environment. Discuss welfare effect. • Explain relation between a country’s environmental laws and that country’s comparative advantage in “dirty goods”. • The next set of lectures will discuss the empirical issues associate with the PHH.

4. Important concepts or terms in these lectures; some of these are familiar • Pollution haven hypothesis • Theory of the second best and Principle of targeting • Factor endowments and the Heckscher-Ohlin-Samuelson model. • Rybczynski effect • Environmental Kuznets’ curve • leakage

5. Three versions of the Pollution Haven Hypothesis (PHH) • 1. At the margin, pollution control costs have some effect on investment decisions and trade flows. (Non-controversial) • 2. Pollution control costs are important enough to measurably influence trade and investment. (A difficult empirical question – and the one that we are interested in.) • 3. Countries set their environmental standards below socially efficient level in order to attract investment or to promote its exports. (This is a difficult empirical and conceptual issue, which the empirical literature does not address.) • #3 unlikely to be true unless #2 is also true.

6. Relevance of TSB when #3 is true • Suppose that environmental laws are too weak and that trade liberalization leads to greater environmental damage. • Trade liberalization might either increase or decrease welfare – recall example from previous lecture on TSB. • Liberalization (e.g. a reduction in trade restriction) increases “standard” gains from trade – which in the partial equilibrium model is the net increase in the sum of consumer and producer surplus. But liberalization might (or might not!) exacerbate environmental distortion. (Higher gains from trade is the benefit and higher environmental damage – in some cases -- is the cost of trade liberalization.) • Are the benefits greater or less than costs? Remember that welfare costs tend to be proportional to “square of the distortion”. (See notes on “gains from trade, partial equilibrium.)

7. Principle of Targeting (again) • The economist’s instinctive response to “second best” arguments for trade restrictions: Principle of Targeting. Trade policy is a poor policy tool for protecting the environment. Environmental policy is the proper tool. • Environmentalist’s instinctive response: We’ll use any tool that is available.

8. Reasons (other than policy differences) for the relation between trade and pollution • (a) Wealth differences cause environmental policy differences, leading to comparative advantage (CA). • (b) Different relative factor endowments associated with pollution and CA • (c) FDI can cause changes in relative factor endowments, leading to production and trade effects (the “Rybczynski effect”). • (d) Trade can promote growth which can affect the environment (the Environmental Kuznets’ Curve). • (e) Technology transfer and the environment

9. “Thought experiments” on the environmental effects of trade liberalization • We cannot do actual experiments to determine the environmental effects of trade liberalization. • In a “thought experiment” we essentially tell ourselves different stories in an effort to think through the relation between trade and the environment. • These experiments (obviously) cannot tell us the truth, but sometimes they are able to help us think more clearly about the evidence that we see. • The next dozen or so slides are examples of thought experiments, guided by economic theory.

10. Wealth and environmental policy • Is “environmental quality” a luxury good? Are the rich willing to spend a larger share of budget on environmental quality? • Nuemayer (pg. 44) cites survey evidence showing lack of correlation between expressed concern for environment and GNP. In other surveys poorer people express a higher “willingness to pay” (as share of income) for environmental quality, compared to rich. • Regardless of level of concern, poor countries may be in a weaker position to pay for clean environment, because of competing needs.

11. (a) The environmental effects of trade driven by differences in wealth • Suppose that two countries have same relative productivities, but one country richer (e.g. because of absolute advantage). • Suppose that pollution standards are set at socially optimal levels (so no market imperfection). Greater wealth leads to higher environmental standards. Poor country has CA in dirty industry because its environmental protection is weaker. • Trade causes production of pollution-intensive good to shift to poorer country with weaker (but optimal) environmental laws. Trade increases aggregate pollution because (under plausible circumstances) production of both goods increases, and more of the dirty good is being produced in the region with weaker environmental protection; trade also increases welfare. (There are no distortions in this example.) • In other words, the observation that an increase in trade increases pollution does not imply that presence of a market failure.

12. Review relation between relative factor endowments, comparative advantage, and pollution • Differences in relative factor endowments are a standard basis for CA. (HOS model, previous set of slides) • Suppose cloth is more capital intensive than food (K/L ratio higher in cloth sector). • The country with higher K/L ratio tends to have CA in cloth. (Other considerations might also be important.) • If the cloth sector is pollution-intensive, country with higher K/L ratio tends to have a CA in the dirty sector.

13. (b) Environmental effect of trade driven by differences in factor-endowments (example) • Suppose that (i) North has stricter environmental standards and therefore uses cleaner production methods. (ii) It is relatively well endowed with capital, and the capital intensive industry is also pollution intensive. ((i) and (ii) are offsetting effects. • Suppose that the capital effect (ii) dominates so North has CA in pollution intensive industry. • When countries trade, production of pollution intensive good shifts to North. • Since North uses cleaner methods (assumption i above), aggregate pollution declines. • Welfare increases because of gains from trade. There are no market failures in this example.

14. The point of previous two examples • There is no “theoretical” reason to assume that increased trade either increases or decreases levels of pollution. • The relation between trade and pollution is an empirical, not a theoretical issue. • Regardless of whether trade increases or decreases pollution, there are positive gains from trade (in the absence of offsetting distortions – the TSB.)

15. (c) The effect of changes in factor endowments • There are many reasons for changes in levels of factors, e.g. FDI, migration. What are the environmental effects of these changes? • The “Rybczynski effect” (a theorem). Suppose that cloth is relatively capital intensive, compared to food. At constant prices, an increase in capital increases production of cloth and decreases production of food. (Illustrate graphically) • If cloth is pollution-intensive, the increase in capital increases pollution. • FDI can cause changes in relative factor endowments, leading to production, trade and environmental effects.

16. Illustration of Rybczynski theorem • Suppose that one unit of cloth production uses one unit of capital and one unit of labor, and suppose that one unit of food production uses ½ unit of capital and 2 units of labor. • Define C and F as number of units of cloth and food that economy produces, and suppose that the economy has 10 units of capital and 15 units of labor available. • The full-employment conditions for capital is: 1C + ½ F = 10; the full employment condition for labor is 1C + 2F = 15. Next slide shows the graph of these equations. They are both satisfied if C = 6.66 = F. • What happens if we increase the number of units of capital by 10%, to 11? Then F= 2.66 and C = 9.66. Food production declines and cloth production increases by more than 10%.

17. Solid lines show graphs of full employment conditions. The intersection of graphs is the equilibrium production point. Dashed line shows effect of increasing capital: an increase in capital increases cloth production and decreases food production, at constant prices. • cloth Full employment of labor Full employment of capital food

18. (d) Economic growth and the environment • The “Environmental Kuznets’ Curve” (EKC): inverted U relation between income and environmental damage. • Environmental damage depends on (i) scale, (ii) composition, and (iii) technique. • (i) Higher output associated with more damage. • (ii) A change in composition such as shift from agriculture to industry to services worsens and then improves environment. (Much too simple a description.) • (iii) More modern industrial techniques tend to be less polluting.

19. Empirics and the EKC • There is a huge literature on EKC, establishing inverted U-shaped relation between environmental variables and income. (Other variables included, e.g. measures of inequality, “openness to trade”). • Recent studies challenge earlier EKC findings. • (The positive relation between “openness” and growth is widely accepted by economists, but empirical evidence is mixed – recall comments at beginning of the semester.)

20. (e) Globalization, technology transfer, and the environment • Multinational firms (MNFs) “likely” to use modern, less polluting plants in foreign subsidiaries. • It is cost-effective for them to use same technology in different countries, even where environmental laws are weaker. • Reputation is important to MNFs (but we frequently see examples where MNF break environmental or safety laws, e.g recent BP case). • Empirical evidence show that more open countries likely to be early adopters of modern methods (e.g. steel production).

21. The main points, again • There are lots of reasons why countries trade. Differing environmental laws may be important in some sectors, but probably have small aggregate impact. • There are many ways in which trade and FDI can impact the environment. There is no theoretical presumption that a particular effect dominates. • Absent other distortions, trade increases welfare, even if it harms the environment. • In the presence of other distortions (e.g. excessively weak laws) the TSB kicks in. • The presence of other distortions might increase the total gains from trade, or make them negative. • I emphasized the example in which trade increases the cost of other (e.g. environmental) distortions. But trade might reduce these costs. • The Principal of Targeting: trade policy is seldom an efficient way of achieving environmental objectives.

22. Just in case you thought that things were not sufficiently complicated… • Here is one more example intended to convince you that strict environmental laws (e.g. high pollution taxes) might or might not decrease a country’s comparative advantage in the “dirty good”. • Consider a two sector (HOS) model – the dirty good and the clean good – and two countries. • The countries are identical except that Home has stricter environmental policy (never mind why). For simplicity, suppose that Home uses a higher pollution tax, whereas Foreign does not tax pollution. • Does Foreign (the country with zero environmental taxes) necessarily have the comparative advantage in the dirty good?

23. As with most economic questions…the answer is “It depends.” • Here is a model in which Foreign does have CA in dirty good: Pollution abatement uses capital and labor in the same ratio as in production of the dirty good. • In this case, part of the capital and labor in the dirty sector is diverted to abatement activities. • The dirty sector in Home diverts a greater fraction of these factors to abatement, relative to Foreign (because of the pollution tax in Home). • Home produces less dirty good per “bundle” of capital and labor (because some of that capital and labor is devoted to abatement), relative to Foreign. Foreign’s production cost of dirty good is less than Home’s, so Foreign has CA in dirty good.

24. But here is a model that goes the other way • Suppose that the activity “abatement” (or “cleanup”) uses only capital. So this activity is more capital intensive than either production of clean or of dirty good. Suppose that the production of the clean good is more capital intensive than production of the dirty good. • The abatement activity (encouraged by the pollution tax) uses capital, thereby reducing the amount of amount of capital available for use in the clean and dirty sectors. • The tax has two effects: As in the first example, it increases production costs of dirty good in Home, eroding Home’s CA in the dirty good. The second effect comes from the Rybczynski theorem: the lower level of capital tends to promote production of the dirty good and decrease production of the clean good. • If the second effect outweighs the first effect, the tax on pollution may cause Home to have a comparative advantage in the dirty good.

25. The point of this example • Some relations that seem “obvious” may not be correct. • It may seem “obvious” that if Home increases its environmental taxes, this change erodes any CA that Home might have in the dirty good. • Although this relation might certainly be true, there are circumstances where it is not true. • Don’t be so sure that you know how policies affect economic outcomes. • The real world too complicated to be explained by any single (useful) theory. Still, theory helps us to organize ideas (and questions); theory provides a lens through which to study the world. Sometimes the lens creates more distortion than clarity, but that is the nature of scholarship.