Wednesday, August 17, 2005

Testing the Pollution Havens Hypothesis

As nations trade more, do poorer nations specialize in exporting dirty goods? This would allow richer nations to specialize in cleaner service industries and might explain why we see richer nations having lower levels of air pollution than poorer nations.

While many of my students have an intuitive belief that this argument holds, the empirical data supporting this claim is not so strong. My favorite explanation for why the evidence is not strong is that domestic pollution havens exist. In the United States, environmental regulation is not uniformly enforced. Under the Clean Air Act, areas assigned to non-attainment status face stricter regulation than areas that are cleaner. Dirty companies can avoid political risk and exchange rate risk by locating in a "domestic" pollution haven. The good news in terms of public health is that these tend to be sparsely populated areas. Thus, if a dirty plant raises local pollution levels but only 8 people live nearby could the social costs of this activity be that high?

A second explanation for why the international pollution havens hypothesis literature finds mixed results has been emphasized by Brian Copeland and Scott Taylor. They point out that "dirty industries" also happen to be capital intensive industries. Think of industries such as steel production or oil refining. These are not labor intensive industries. So what?

The "Factor Endowment Hypothesis" actually predicts that richer nations will
specialize in exporting dirty goods (because these industries are capital intensive and richer nations have the capital) while the "pollution haven hypothesis" posits that poorer nations with more lax regulation will specialize in these industries. Given that these two hypotheses work in opposite directions, it should not be surprising that so many empirical studies of the pollution havens hypothesis yield mush.

Perhaps not surprisingly, when Yutaka Yoshino and I looked at world trading patterns over the years 1980 to 1997 we found that middle income nations were the pollution havens. These nations were exporting a lot of dirty goods relative to their exports of clean goods. These nations have some capital but are likely to have more lax regulation than richer nations. We wrote a paper that was part of a bigger special issue on Pollution Havens.

Our paper focused on a simple idea that didn't work out as well as I had hoped. More and more nations are participating in Regional Trade Agreements such as NAFTA. In a world of regionalized trade, I wanted to test whether relative income matters. If a middle class nation is the poorest nation in its trading bloc, is it a growing pollution haven? To be honest we did not find strong evidence supporting this claim.

Some recent empirical research has found evidence that increased regulation in one nation induces a "displacement" effect such that the dirty activity migrates to less regulated areas. Here is a quote from a nice paper by Levinson and Taylor (2004).
"This paper uses both theory and empirical work to examine the effect of environmental regulations on trade flows. We develop a simple economic model to demonstrate how unobserved heterogeneity, endogeneity and aggregation issues bias measurements of the relationship between regulatory costs and trade. We apply an estimating equation derived from the model to data on U.S. regulations and net trade flows among the U.S., Canada, and Mexico, for 130 manufacturing industries from 1977 to 1986. Our results indicate that industries whose abatement costs increased most experienced the largest increases in net imports. For the 20 industries hardest hit by regulation, the change in net imports we ascribe to the increase in regulatory costs amounts to more than half of the total increase in trade volume over the period." by Levinson and Taylor

When I talk to people about the pollution haven hypothesis, they often point to Europe and ask whether Romania will be a pollution haven for France and Italy. These nations are physically close to each other. I think that it is no accident that the EU nations are requiring Eastern European nations to adopt strict environmental regulations as a requirement for entry in this trading bloc.

While people often wonder whether Africa will be a pollution haven for the rest of the world, geography does matter. In previous research, I found that heavy to ship manufacturing goods where less likely to be exported further distances back to the United States. In a world of zero transportation costs, certain nations might be more likely to be pollution havens but this is not our world!

More frustrating for empirical economists is that we have mainly looked at manufacturing trade patterns. Ideally, I would like to see trade in garbage, nuclear waste, mining and commodities and other natural capital measures.