تجزیه و تحلیل اقتصادی از معیارهای غرب آمریکا میانه روند تولید آلاینده گازهای گلخانه ای 1970-2000

From 1970 to 2000, U.S. economic output doubled but emissions of four criteria pollutants from economic activity—CO, NOx, VOC, and SO2—decreased by 20%. Understanding what factors have contributed to this pollution reduction in the U.S. as a whole, as well as in various regions within the country, has important policy implications. A recently developed regional environmental–econometric input–output model for the Midwestern states of the U.S. has been used to examine the causes of pollution reduction in this regional economy over a thirty-year period. Simulations conducted with this model suggest that, for the rate of growth experienced over the period, technological improvement has dominated economic structural change in the reduction of pollutant emissions. On average, technological improvement has accounted for approximately 80% of emissions reduction, while economic structural change explains the remaining 20% of the decrease. Our analysis suggests that, while much remains to be done in reducing emissions in both developed and developing countries, policies that are informed by an understanding of the role of structural change and which promote the adoption of more recently developed technologies may contribute substantially to sustainable development.

The 1970 Clean Air Act and the establishment of the United State Environmental Protection Agency (USEPA) initiated a new era of protecting human health and the ecosystem from harmful air pollution in the U.S. Since then, emissions of airborne so-called ‘criteria pollutants’ in the country (http://www.epa.gov/air/urbanair/) have decreased markedly. For example, from 1970 to 2002, emissions of carbon monoxide (CO), nitrogen oxides (NOx), volatile organic compounds (VOC), and sulfur dioxide (SO2) decreased by 48%, 17%, 51%, and 52%, respectively ( USEPA, 2003), while U.S. economic output doubled. These reductions are believed to have been achieved largely through technological advances induced by strict regulations and market forces. On the other hand, economic structural change, or the change in the composition of an economy's industrial output, may also have contributed to reduced emissions ( Casler and Blair, 1997, Selden et al., 1999, Kimmel et al., 2002 and Tao et al., 2007). Selden et al. (1999) analyzed the U.S. emissions trends from 1970 to 1990 and found out that the economic structural change reduced emissions at a smaller scale than technology (including energy efficiency and other technique) did. Kimmel et al. (2002) concluded that the decreases in CO, SO2 and NOx emissions in the capital of Estonia were due largely to the restructuring of the Estonia economy and shift away from the usage of sulfur-rich fuel. The causes of U.S. economic structural change are manifold. First, demand for services has grown faster than that for manufactured goods; second, productivity in manufacturing has increased more than that in services, leading to higher wages in manufacturing and a substitution away from labor to service inputs in manufacturing production, hence to higher output from the service sector; and third, in the era of globalization, relocation of some manufacturing production out of the country has resulted in services accounting for a larger share of gross product in the entire economy. These structural changes may have effectively reduced pollution in regions that outsourced production by heavily polluting industries while the goods produced in these sectors were obtained through international trade generating increased pollutant emissions in other parts of the world. Answering the question of what accounts for reductions in some pollutant emissions of US industry—technology vs. economic structural change—has important policy implications (Levinson, 2008). If economic structural changes, not technological improvements, account for the majority of pollution reductions in the U.S., then the U.S. pattern of industrial production is not sustainable and cannot be replicated elsewhere in the world, since the least developed countries may never be able to import pollution-intensive goods from even poorer countries. Levinson (2008) has analyzed the causes of pollutant emissions reductions in the U.S. manufacturing sectors and pointed out that approximately 75% of the reduction has been due to improved technologies and that changes in the composition of output have accounted for the remaining 25%. Since the 1980s, there have been other studies focusing on the impacts of structural and technological changes on energy consumption and its related gas emissions, particularly carbon dioxide (CO2), using various decomposition methods (see, e.g., Dietzenbacher and Los, 1998, Ang and Zhang, 2000, Hoekstra and van den Bergh, 2002 and Hoekstra and van den Bergh, 2003). Ang and Zhang (2000) reviewed the decomposition methods developed and applied from the late 1970s to 2000 in analysis of energy and environmental emissions. Among all methods, they found that the revised Laspeyres method and the log mean divisia method showed superiority to others in that they were complete decompositions and no residual was produced. Despite the progress made in analyzing the relationship between changes in industrial structure and technology and CO2 emissions, little has been done in the way of systematic analysis of the relationship between pollutant emissions reduction and regional economic structure as a whole. In this paper, we report on research in which a newly developed regional environmental–econometric input–output model has been used to explore the sources of and changes in pollutant emissions generated by economic activity in the Midwestern U.S. The emissions selected for study come from a wide range of sources, including those resulting from energy consumption, material usage, and beyond. We also explore how technological and structural changes have contributed to reductions in emissions in this region between 1970 and 2000. In the next section, the Midwest model will be presented; Section 3 provides the results and their interpretation, and the paper concludes with some summary comments in the final section.

The sum of emissions from four criteria air pollutants, CO, NOx, SO2, and VOC declined 20% between 1970 and 2000. Meanwhile, total U.S. economic output increased by nearly 100%. Emissions should have increased by 100% if pollution had increased one-for-one with output for the same time period. The large reduction in emissions of criteria pollutants introduces an interesting yet important question as to what has contributed most to this development. Conventional wisdom says an improvement in technology, represented by a change in emissions intensity (I) in this study, can reduce emissions markedly. On the other hand, on-going globalization and economic structural change (S), i.e., the shift of economic activity toward cleaner sectors, may also have played roles in the U.S. criteria pollutant clean-up. Determining which factor has been the more important driver of change has important policy implications. If the clean-up in the U.S. has resulted mainly from technological advances, the U.S. pattern of industrialization can be replicated in other parts of the world through technology dissemination and penetration—say, through programs promoting the “clean development mechanism.” If, however, structural change in the composition of output by the U.S. economy—e.g., outsourcing production by heavily polluting industries—has been the determining factor of the clean-up, the path is not replicable, since eventually there will be no country left to take heavy polluted industries in the production-chain. In this study, a recently developed regional environmental–econometric input–output model was employed to investigate factors contributing to reductions in criteria pollutant emissions in the regional economy of the Midwestern U.S. Specifically, we have considered how much of these reductions can be attributed to technological change (the I effect) and how much to economic structural change or the changing composition of regional GDP (the S effect). The results show that, over the 30-year period from 1970 to 2000, production in the Midwestern U.S. economy almost doubled with a greater concentration of activity being in cleaner sectors and decreased shares of agriculture, construction, and mining in aggregate production (Table 3). Among manufacturing sectors, those with higher pollution potentials, e.g., sectors 06 (Primary Metal Products) and 07 (Fabricated Metal Products), also saw substantial reductions in their shares of output (more than 20% in comparison with the respective 1970s level). During the same time period, the emissions intensity factor I decreased substantially (more than 50% between 1970 and 2000) for all manufacturing and service sectors ( Table 4). Meanwhile, the emissions intensity factors for agriculture and construction sectors increased by 200% and 65%, respectively. The change in emissions of the four criteria pollutants studied in the case of the Midwestern U.S. is the collective result of changes in economic scale, technology, and structure. In order to quantify the contribution of each factor to overall changes in these emissions, we applied the revised Laspeyres decomposition method. Our findings (and common sense) suggest that, other things being equal, economic growth (Q) should always increase criteria pollutant emissions (see Table 5). If I and S remained at their 1970s values, the Q effect alone would have increased the Midwestern U.S. criteria pollutant emissions by 63% between 1970 and 2000. This increase in emissions potential, however, was largely offset by emissions reductions that resulted from technological change (the I effect), and which, together with a change in the composition of output (the S effect), decreased the production of criteria pollutants from the entire Midwestern U.S. economy over the three decades. Neutralizing the Q effect (i.e., comparing emissions between base year and the years with concurrent economic growth but fixed I and S), we quantified the relative contributions of the I and S effects to criteria emission changes over time. With all four criteria pollutants combined, we found the I effect contributed at least 75% to reduced pollution emissions from the Midwestern U.S. economy, while the S effect accounted for up to 25% ( Fig. 4). These figures generally hold for individual pollutants of NOx, SO2, and VOC as well. However, the I and S effects on CO production deviate from this pattern with the S effect contributing at least 50% of overall CO emissions change since the early 1980s. Within the manufacturing sectors, the result is consistent with a dominant I effect (more than 70% averaged over 30 years) even for CO emissions changes. Our analysis suggests that, while much remains to be done in reducing pollutant emissions by both developed and developing countries, and while it is perhaps unreasonable to expect that there will be no limit to emissions reductions from adoption of more recently developed technologies, the adoption of such technologies as have been employed in the U.S., may nonetheless contribute to more sustainable development ( Hardner et al., 1999). There are a few caveats that should be raised concerning the findings of this study. First, it should be noted that this study has only accounted for criteria pollutant emissions from economic activity, which produced approximately 46% of total combined 1999 Midwestern U.S. emissions of CO, NOx, SO2, and VOC. Emissions from on-road mobile and residential sources (e.g., residential heating/cooling and backyard burning) were excluded from the analysis since they could not be assigned to any industrial sector in the MW-REIM model. In particular, emissions from on-road mobile sources accounted for more than 80% of the remaining 54% pollution in the study period. From 1970 to 2000, Americans drove 157% more miles (Office of Highway Policy Information, 2008, http://www.fhwa.dot.gov/policy/ohpi/hss/hsspubs.cfm) but emissions from driving vehicles decreased by 61%. This decrease represents a 75% reduction in on-road mobile emissions that would have been generated if the same 1970s emissions intensity factors had been in force for year 2000; the decrease is almost entirely due to technological improvements induced by more stringent environmental regulations and market forces (http://www.epa.gov/oms/hwy.htm). This reduction in transport-related emissions, taken together with the findings on emissions reduction discussed above, gives cause for some optimism for a future with a cleaner environment built on technological improvement. Second, the current MW-REIM model does not include commodity price effects and this omission may bias the relative contributions of I effects vs. S effects. For example, increases in real energy prices can lead to substitution away from energy intensive and polluting inputs towards other inputs for a given state of technology. This price factor definitely would impact the I and S effects defined in the study. A next-generation regional environmental–econometric input–output model should take commodity price effects into account and thus provide a better quantification of each contributor to criteria pollutant emissions changes. Finally, CO2, an important green house gas, was not included in the analysis since it was not designated as a criteria pollutant by the U.S. EPA. During the same time period of 1970 to 2000, however, U.S. CO2 emissions from fossil fuel burning increased by 30% (http://cdiac.ornl.gov/ftp/trends/emissions/usa.dat). Clearly, there is still a long way to go to reduce CO2 emissions resulting from human activity. Continued experimentation with innovative policy in concert with improved technology may make the goal of sustainable development achievable.