What accounts for the clean-up of US manufacturing: technology or international trade?

Arik Levinson 02 January 2008



Antiglobalisation protesters display signs denouncing international trade's role in polluting the environment.1 Pundits write Op-Ed pieces cautioning that increased trade has environmental costs.2 And a majority of Americans agree that "freer trade puts the United States at a disadvantage because of our high ... environmental standards".3

Are they correct? Over the past thirty years, while the real value of US manufacturing output has increased by more than 70 percent, the total annual air pollution emitted by US manufacturers declined substantially, by 58 percent for the sum of four common air pollutants.4

One explanation for the clean-up of US manufacturing is that the protesters are correct, and that thanks to freer trade, the US now imports polluting goods it once produced domestically, and concentrates domestic manufacturing on goods less likely to incur environmental regulatory costs. Of course, there is an alternative explanation: thanks to improved technology (cleaner fuels, end-of-pipe abatement, process changes, etc.) US manufacturers may now be able to produce more output using less pollution. Which of these explanations, trade or technology, accounts for the dramatic clean-up of US manufacturing pollution?

Most of the economic research to date has focused on related causal questions:5 Have US environmental regulations driven polluting industries overseas? Are polluting industries more likely to locate where environmental standards are less strict? Have lower tariffs enabled polluting industries to relocate from the US to less strict locales? These are important questions, but they are difficult to answer for several reasons. First, measuring the strictness of environmental standards poses difficult conceptual challenges. Some regulations may be strict but unenforced, while others are lax but tightly enforced. Regulations may mandate technology adoption, limit effluent quantities, or impose taxes. Ranking these multidimensional laws is difficult if not impossible.

Second, these questions of causality face the problem that international trade and environmental regulations are determined simultaneously. Polluting industries may be more likely to locate in lax jurisdictions, but jurisdictions that attract lots of polluting industries may impose tight standards. Disentangling the effect of regulations on trade from the simultaneous effect of trade on regulations requires long time series' of data and instrumental variables analyses. Recent empirical work using these techniques does seem to find that those industries that have seen the largest increase in pollution abatement costs in the US have also seen the largest increase in net imports, which suggests that increased trade may be responsible for some of the clean-up of US manufacturing.6

Whether or not US regulations have pushed US manufacturers overseas, this begs the question we began with. How much of the clean-up of US manufacturing can possibly be explained by increased imports of polluting goods, as opposed to technology improvements? That question turns out to be much simpler to answer, and may well be more important.

Grossman and Krueger (1993) laid out what has now become a standard framework for thinking about the effects of trade on the environment. Pollution can be thought of as the product of three components: the overall size of the economy ("scale"), the mix of sectors comprising the economy ("composition"), and the technologies employed in production and abatement ("technique"). Line (1) in Figure 1 plots the overall scale of the manufacturing sector, which grew by 71 percent in real terms from 1972 to 2001. If pollution increased one-for-one with manufacturing output, then pollution from US manufacturing would also have increased 71 percent. This is the scale effect.

Line (2) in Figure 1 depicts the actual pollution from US manufacturers, which declined by 58 percent.7 The difference between lines (1) and (2) depicts the dramatic clean-up of US manufacturing. It must be accounted for by some combination of improved technology and changes in industrial composition (which itself may be the result of increased imports of polluting goods).

To calculate how much of the clean-up of US manufacturing can be explained by technology, I use a dataset created by the US Environmental Protection Agency (EPA) that reports for 1997, the amount of pollution emitted per dollar of value shipped by each of the 470 six-digit North American Industry Classification System (NAICS) codes that comprise the manufacturing sector.8 I take these emissions factors as an estimate of the state of "technology" as of 1997. In each year from 1972 to 2001, I multiply each industry's 1997 emissions factor by its concurrent real value of shipments, which tells me how much each industry would have emitted in every year, if in every year each industry had used its 1997 technologies. For each year I then sum across all manufacturing industries, which tells me how much pollution would have been emitted by the entire manufacturing sector if it had its concurrent scale and composition but the 1997 technologies. Scale this series so that 1972 equals 100, and it becomes line (3) in Figure 1.

Line (3) plots the combined scale and composition effects of changes in manufacturing on pollution emissions, holding technique constant. It increases by 17 percent over the 30-year period. The difference between line (1) -- the scale of manufacturing-- and line (3) -- the scale and composition -- is due the shift in US manufacturing towards less polluting goods. It accounts for 40 percent of the overall reduction in pollution from manufacturing. The other 60 percent, represented by the difference between lines (2) and (3), must be due to improvements in technology or "technique."

At this point, we are halfway to answering the question posed. We know that most of the clean-up of US manufacturing comes from improved technology, not from changing composition. If international trade were driving the improvement, most would come from changing composition. Since composition only accounts for 40 percent of the gap between lines (1) and (3), we know that trade can account for at most 40 percent of the clean-up. Some of that 40 percent, however, could be accounted for by changing domestic consumption patterns.

To estimate how much of the shift in the composition of US manufacturing towards cleaner goods could be explained by international trade, I conduct the following simple experiment. In each year, I multiply each 6-digit industry's net imports (imports minus exports) by the emissions factor reporting how much pollution is generated in the manufacture of that good (including the manufacture of its intermediate inputs). I then aggregate across all industries, which yields the amount of pollution that would have been produced in the US had all of those imported goods been produced domestically, and all of the exported goods been produced overseas. To show the change over time, I subtract from this series its value for 1972 to get line (4) in Figure 1.

Line (4) plots this "no-net-trade-growth" scenario. This line represents what air pollution emissions from US manufacturing would have been if every bit of increased net imports since 1972 had instead been manufactured in the United States, and holding technology fixed as of 1997. The line explains about 70 percent of the composition effect.


What is the bottom line? Increased net imports of polluting goods account for about 70 percent of the composition-related decline in US manufacturing pollution. The composition effect in turn explains about 40 percent of the overall decline in pollution from US manufacturing. Putting these two findings together, international trade can explain at most 28 percent of the clean-up of US manufacturing.

Why should we care?

If the 75% reduction in pollution from US manufacturing resulted from increased international trade, the pundits and protestors might have a case. Environmental improvements might be said to have imposed large, unmeasured environmental costs on the countries from which those goods are imported. And more importantly, the improvements in the US would not be replicable by all countries indefinitely, because the poorest countries in the world will never have even poorer countries from which to import their pollution-intensive goods. The US clean-up would simply have been the result of the US coming out ahead in an environmental zero-sum game, merely shifting pollution to different locations. However, if the US pollution reductions come from technology, nothing suggests those improvements cannot continue indefinitely and be repeated around the world. The analyses here suggest that most the pollution reductions have come from improved technology, that the environmental concerns of antiglobalization protesters have been overblown, and that the pollution reduction achieved by US manufacturing will replicable by other countries in the future.


Cole, M.A. 2004. "US Environmental Load Displacement: Examining Consumption, Regulations and the Role of NAFTA." Ecological Economics 48(4): 439–450.
Grossman, G.M., and A.B. Krueger. 1993. "Environmental Impacts of a North American Free Trade Agreement." in The Mexico-US Free Trade Agreement (edited by P.M. Garber, Cambridge, MA: MIT Press) reprinted in Law and the Environment: An Interdisciplinary Reader (edited by R.V. Percival and D.C. Alevizatos, Philadelphia: Temple University Press, 1997).
Kahn, M.E. 2003. "The Geography of US Pollution Intensive Trade: Evidence from 1959 to 1994." Regional Science and Urban Economics 33(4): 383–400.
Leontief, W. 1970. "Environmental Repercussions and the Economic Structure: An Input-Output Approach." Review of Economics and Statistics 52(3): 262–271.
Levinson, A. "Technology, International Trade, and Pollution from US Manufacturing" NBER working paper no. 13616, November 2007.
Levinson, A., and M.S. Taylor. Forthcoming 2008. "Unmasking the Pollution Haven Effect." International Economic Review.
US EPA.. 2000. "National Air Pollutant Emission Trends, 1900–1998." EPA-454/R-00-002. Research Triangle Park, NC: US EPA Office of Air Quality Planning and Standards.




1 http://www9.georgetown.edu/faculty/aml6/images/WTO_protest.jpg
2 Dorgan, Byron and Sherrod Brown. "How Free Trade Hurts" Washington Post. Saturday, December 23, 2006 p. A21; Myerson, Harold. "Why the Rush on Trade?" Washington Post. Wednesday, November 7, 2007 p. A21.
3 http://www.americans-world.org/digest/global_issues/intertrade/environment.cfm
4 This is the unweighted sum of sulfur dioxide (SO2), nitrogen dioxide (NO2), volatile organic compounds (VOCs) and carbon monoxide (CO). These are the four monitored consistently over this time period by the National Emissions Inventory, and they decreased individually by amounts ranging from 30 percent for NO2 to 66 percent for SO2.
5 Cole (2004 and Kahn (2003) show that imports to the U.S. have shifted towards less polluting goods. An older literature (Leontief 1970) documents the degree to which composition changes in U.S. will have reduced pollution. But neither of these assesses the degree to which technology has reduced pollution in the U.S.
6 See, for example, Levinson and Taylor (2008).
7 Pollution from U.S. manufacturing based on the calculations described in Levinson (2007). I took the U.S. Environmental Protection Agency's "National Emissions Inventory" and summed the disaggregate emissions estimates from those sectors of the economy most closely matching manufacturing.
8 Abt Associates, Inc. 2004. "Trade and Environmental Assessment Model: Model Description." Prepared for U.S. Environmental Protection Agency (EPA), National Center for Environmental Economics. Cambridge, MA: Abt Associates.




Topics:  Environment International trade Productivity and Innovation

Tags:  US, emissions, manufacturing, pollution

Professor of Economics, Georgetown University


CEPR Policy Research