Does the U.S. Outsource Polluting Industries?

Featured in print Digest

The cleaner U.S. manufacturing composition is not offset by dirtier imports. Rather, the composition of imports has also become cleaner.

Polluting industries' share of U.S. manufacturing output has declined in recent decades. That is good news for environmental quality in the United States. However, the fact that the decline has coincided with falling trade barriers has given rise to suspicions that the United States has outsourced its polluting industrial processes to developing countries. In Trade Liberalization and Pollution Havens (NBER Working Paper No. 10585), authors Josh Ederington, Arik Levinson, and Jenny Minier ultimately refute the notion that domestic manufacturing is cleaner today because trade agreements have allowed the United States to use "pollution havens" in the developing world to do its dirty work.

"We find no evidence that domestic production of pollution-intensive goods in the U.S. is being replaced by imports from overseas," they state. On one hand, Ederington, Levinson, and Minier understand how the "casual observer" could see a link between lower tariffs and a cleaner U.S. manufacturing sector. After all, while U.S. manufacturing was growing by 51 percent -- even as total emissions of many pollutants were growing at half that rate or actually dropping -- U.S. tariffs were falling dramatically. Between 1978 and 1994, tariffs on manufactured products were reduced 50 percent on average. But if trade liberalization were simply allowing dirty industries to relocate offshore, then the proportion of U.S. imports produced by pollution-intensive industries would rise as tariffs fell. Yet the opposite occurred.

Imports overall grew by 318 percent during the period. But according to World Bank data that characterizes industries by their pollution intensity, imports of goods manufactured in highly polluting processes grew at a much slower rate. In other words, just as the U.S. manufacturing sector was growing while simultaneously shifting toward clean industries, the same thing was happening to our imports: they were rising, but the percentage of goods coming from polluting industries was going down. "The cleaner U.S. manufacturing composition is not offset by dirtier imports," the authors write. "Rather, the composition of imports has also become cleaner."

Importantly, their conclusion holds firm even when they limit the analysis to trade with developing countries. Imports from developing countries grew by 344 percent, but imports of pollution-intensive goods from developing countries grew much more slowly.

Ederington, Levinson, and Minier believe that, if anything, lower tariffs might be actually slowing the U.S. shift to cleaner industries, because imports of goods made from pollution-intensive processes have not kept pace with the overall rise in imports. Thus, surprisingly, one potential environmental consequence of tariff reductions is that U.S. industries are dirtier than they otherwise would be.

This refutation of the conventional wisdom -- that tariff reductions reduce pollution in rich countries by sending the problem to poor countries -- has considerable implications for both public perceptions of trade agreements and the manner in which they are negotiated. The general concern that trade liberalization is causing developed countries to outsource dirty manufacturing processes to the developing world has sparked heated protests against the World Trade Organization. Furthermore, the perceived link has prompted U.S. officials to require detailed "environmental reviews" of trade agreements as part of the negotiating process.

Ederington, Levinson, and Minier note that some may find fault with their findings by arguing that past tariff reductions are not representative of probable future trade agreements. If future trade agreements reduce U.S. tariffs on polluting goods by more than tariffs on clean goods, then the expected patterns may emerge. Polluting industries will relocate away from the United States, where pollution will be reduced at the expense of environmental quality elsewhere.

But even when the authors consider a hypothetical future in which tariffs on all manufactured goods are dropped from current levels down to zero, they still find no evidence that the change would spark a disproportionate shift in polluting industries from the United States to developing countries. "Once again, we find little basis for the fear that tariff reductions themselves will generate large shifts in the composition of U.S. manufacturers away from polluting industries."

-- Matthew Davis