lS IW fie 1d al ds n-ve le s. at 1-1-:o \-!1, Jl e e J February 1996 ECONOMIC GROWTH AND ENVIRONMENTAL QUALITY 23 fl l'l"J:I<-UI Appln atums 6(I). 1996, pp 23-24 ,. I 496 ~y the Eco log • cal Soc1e1y of Amenca We have the greatest respect for the authors of the Science article (Arrow et al. I 995) and we agree with their argum ents . However, we would like to offer some points of clarification and to broaden the base of the discussion. The Science article focused on the inverted U relationship between environmental quality and gross domestic product (GDP). But this empirical re-lationship adopts a trivial definition of environmental quality: emissions of specific pollutants, compared across vast differences in culture, resources, and scales. A subset of pollutants in a limited number of places cannot be accepted as surrogates for the complex in-teractions between economic growth and the environ-ment on which that growth depends. Policy discussions and scientific debates often focus on a simplified concept of environmental interactions. Accepting these simple concepts places advocates of growth in a s trong position to argue for continued de-velopment. At one extreme, the environment is reduced to extractable resources. The environmentalist argues that economic growth will exhaust extractable re-sources, such as, oil and iron ore. This argument faces a forceful counterargument: most extractable resources are substitutable. Substitutes may be other extractables, human capital, or human-made capital. Don 't worry about running out of coal-there is fission, fusion, or biomass fuel crops. The point can be argued, of course. because substitutability varies widely across extracta-' Manuscnpt received I 0 July 1995. t For reprints of this Forum, see footnote I, p. 12. ' The Oak Ridge National Laboratory is managed by Lock-heed Martin Energy Systems, Inc. , under subcontract number DE-AC05-840R21400 for the U.S. Department of Energy. Environmental Sc1ences Division Publication Number 4454. ble resources and experience has shown that substi-tutes, such as nuclear power, may create as many prob-lems as they solve. But if one accepts that the envi-ronmental problem can be reduced to extractable re-sources, the advocates of growth s tart the debate from a strong position. At another extreme, environmental issues are re-duced to amenities-nice places to have a pic nic. In thi s case, the environment has aesthetic value but no direct involvement in the economic process. You are unlikely to win the argument if you maintain that aes-thetics are more important than jobs. If you want to preserve Spotted Owls s imply because they are pretty. you are likely to be hit in the back of the head with an axe handle. In other cases, analysis focuses on direct physical and health effects of pollution. For example, the as-sessment of acid precipitation considered direct dam -age to buildings, agricultural crops, visibility, and trout populations. But eliminating direct impacts of pollut-ants w ill not prevent gradual, indirect effects such as climate change or loss of the ozone layer. Underlying all of these simplifications are basic eco-system services: cleaning the water. purifying the air. decomposing wastes, maintaining CO, balance. per-mitting recovery from natural disturbances, filtering ultraviolet radiation , and providing sources of nev. medicines. These ecosystem functions are not s ub'>li -tutable and not just amenities that can be traded for short-term economic gains. Discussions of economic growth often ASSUME stable, resilient ecosystems that will continue to provide these life-support services. But economic development will impose in creasing stress on ecosystems. even if wealthy nations can be expected