تجارت خارجی و کاهش آلودگی در سوئد: تجزیه و تحلیل تجزیه شدن اثرات ساختاری و تکنولوژیکی بلند مدت

This article examines whether there exists any causal relationship between foreign trade and declining pollution in developed countries. In other words, do developed countries outsource their problems to less developed countries rather than solve them? The case study is the Swedish economy and the two environmental indicators employed are energy consumption and CO2 emissions. No causal relationships are found, since Sweden has long been a net exporter of embodied energy and CO2 and continues to be so after 1970, when energy consumption stabilizes and CO2 emissions decline. In addition, the ratios of net exported energy and CO2 to total consumption remain stable, which means there were no effects on the energy intensity or CO2 intensity either. These results suggest that internal forces, like efficiency improvements, changed consumption patterns and transformation of the energy system, have been crucial for relative environmental improvement in Sweden, while foreign trade has played no role.

According to a majority of economists, international trade contributes to an increase in total world production, which will be distributed among countries according to their relative strength on the international markets, expressed in the relative prices of their products. The world economy benefits from this. There is, however, warranted concern among many scholars about the possible detrimental effects of trade on the environment and on working conditions in developing countries. Laxer regulation of working conditions and of environmental standards in developing countries compared to that in the developed countries may stimulate the establishment of enterprises, which are neither friendly to human health nor to the environment. The idea that the location of industry is determined by environmental regulations is called the pollution haven hypothesis (Copeland and Taylor, 2004). Several studies have been conducted that attempt to show such effects with mixed results. One reason why such effects are not found more frequently is that environmental costs hardly ever surpass 4% of production costs and therefore could not have a decisive impact on localization (Xepadeas and De Zeeuw, 1999). There is, however, support in the most recent studies (1999–2004) for a pollution haven effect, i.e. that lax environmental regulations play a marginal role for location decisions, but not for the pollution haven hypothesis that lowering trade barriers leads to a shift in the location of polluting activity to the regions with the weakest regulation (Copeland and Taylor, 2004). Still there is some economic logic in developing countries specializing in labor-intensive and sometimes also natural-resource-intensive production. This could naturally imply that some environmental improvement in developed countries is achieved at the expense of worsened conditions elsewhere in the world. Against this stands the fact that it also makes sense, according to trade theory, that developed countries specialize in capital intensive production, which is generally more energy demanding and polluting than labor-intensive production. According to the Environmental Kuznets Curve (EKC) hypothesis, the negative environmental effects of economic development increase until a certain income is reached. Thereafter the environmental impact, usually measured as pollution emissions, will decrease (Panayotou, 1992). Suspicion is warranted that this downward slope of the curve, or the environmental improvement, is somehow related to international trade (Stern et al., 1996). It may very well be that as countries develop they reach another position in the world economic system, which means that instead of exporting natural-resource-intensive goods they import such goods (Hermele, 1995). This would imply decreasing energy demand and concomitant pollutants in those countries. For this system to continue to function, there is a need for other countries, at a lower stage of development, to enter the international trade arena, and to become net exporters of products that cause pollutant emissions and/or put high pressure on finite natural resources. This means that there are not necessarily any environmental gains in the world system if some developed countries show the EKC trajectory. If this geographical shift, rather than elimination, of environmental problems is an important feature of development, this has adverse consequences both for developing countries and for the global environment: the latter because some problems like the greenhouse effect are truly global in scope. It may, of course, instead be the case that the downward slope of the EKC stems from internal causes in the developed countries. Higher income leads to changed demand for goods and services, which may give rise to consumption in that country that is less material intensive and more knowledge intensive. In addition, people at a higher level of income, with more knowledge about environmental problems and perhaps also more effective democratic institutions, may devote resources to cleaner production. Coupled with technical change on the production side, these demand side changes may cause an EKC (Kriström, 2000). The case does not necessarily have to be simple. It does not have to be either foreign trade or internal causes that explain the downward slope of the EKC; it may well be a combination. Therefore, the issue of a possible causal relationship between trade patterns and pollution merits attention, and if such a relationship can be established there is a need for a quantification of these effects compared to the effects of internal forces. Some of the previous research, which has related international trade to the EKC, has done this under the a priori assumption that international trade stimulates economic growth, which in turn increases environmental problems. Environmental problems do, however, decline during the later stages of the growth process, according to the EKC. The impact of trade on the environment is hence estimated as the net result of growth (causing increasing emissions) and the EKC (causing declining emissions) (Dean, 2002). This kind of research does not examine whether there is any causal relationship between trade and the EKC or not. Researchers who have addressed the question of a possible causal relationship between trade and the EKC path have reached different conclusions. Antweiler (1996) constructs a pollution index for 1987 and finds that highly developed countries export more products that cause emissions than they import. Hence, the idea of environmental dumping in less developed countries is not supported at all. Agras and Chapman (1999) use an econometric model where they find energy prices to be the single most important explanatory variable for energy/GDP and carbon dioxide/GDP ratios, while trade, to their surprise, is insignificant. Chapman and Suri (1998), on the contrary, find trade to be an important explanatory variable for the EKC, when they use a combined time-series and cross-section model, and test for factors determining energy–GDP ratios. They draw the conclusion that trade causes both the upward and downward slope of the EKC. Muridian et al. (2002) provide some support for the impact of trade. They study embodied emissions in imports and exports for 18 industrialized countries in the period 1976–1994 in their trade with developing countries. However, a clear cross-section trend between embodied pollution in exports and imports was not found. As pointed out by the authors, this is probably due to differences in the composition of production among the countries. Cole (2004) finds some evidence of trade effects on the EKC when studying 10 air and water pollutants, but these effects do not seem to be widespread or large. Most of these studies are based on econometric techniques, which have the advantage of covering several countries, which in turn allow general conclusions. The main disadvantage is that the adverse environmental effects of trade are not addressed directly, but are instead approached through related indicators, like the export/production and import/production ratio for manufacturing industry (in economic terms) or trade openness defined as the ratio of the sum of imports and exports to GNP. This is mainly since economic data are readily accessible, while data relating these economic figures to environmental indicators, i.e. detailed historical environmental accounts, are in short supply. Our article contributes to the discussion by focusing on one country, namely Sweden, over a longer period than previous studies, combining the results of our two independent studies (Kander, 2002; Lindmark, 2001) of energy–GDP ratios and CO2–GDP ratios, and directly addressing the crucial question of directions in the balances of trade with respect to energy and CO2. This is done by comparing trade statistics and production statistics at a fairly detailed level and calculating the net effect of trade. Kander (2002) compared energy in trade for three benchmarks between 1955 and 1987. The time period was here extended to 2000. These energy studies only take direct energy consumption of the final production into account, and not the energy demands of the entire production chains. Carbon dioxide is to a large extent determined by the composition of energy carriers and we also address the issue of possible bias in trade with respect to energy carriers. For this we use the input–output analyses of Lindmark (2001) with a first benchmark in 1913 and a last one in 1995. The approach we use in this study focuses only on the exchange of manufactured products with other countries. Another, related issue is the role of direct foreign investments. Swedish capital is used to establish industrial enterprises abroad and foreign capital is invested in Swedish companies. We do not account for the balance in such exchange of direct investments and whether it has some environmental bias.

We have demonstrated that Sweden was a substantial net exporter of energy in goods between 1955 and 2000. Without foreign trade, Swedish energy consumption would have been lower than it actually was in this period. In addition, we have stressed that the important variable for an assessment of the possible impact of trade on changes in energy consumption or changes in energy intensity is not the energy balance in trade per se, whether this is positive or negative, but rather changes in this balance. Possible changes can be investigated by examining the ratio of net exported energy to total energy. An increase in this ratio during a certain period implies that trade in that period contributed to increasing energy consumption and perhaps also increasing energy intensity, while a decrease has the opposite effect. The ratio of net exported energy to total energy increased between 1970 and 1987, but decreased back to the 1970 level in 2000. Hence foreign trade cannot explain any of the stabilization of energy consumption or decline in energy intensity between 1970 and 2000, at least not at the level of aggregation at which this study is performed or with the methods employed. The study uses a pretty fine sectoral division of the economy, but homogeneity within these groups may still be questioned. The method investigates the effects of foreign trade on energy consumption in principle, looking at structures of imports and exports, but does not take actual technology of specific countries into account, and that may in reality differ. Emissions of carbon dioxide are logically determined by the composition of energy carriers and the amounts of energy consumption. In a real historical perspective, energy carrier composition plays the largest role. We have examined whether there are any biases in the Swedish foreign trade with respect to CO2 and hence energy carriers. To examine this carefully, we have conducted studies of the entire production chains, using input–output tables. The result is that Sweden was a net exporter of CO2 in the period 1913–1995. Compared to total emissions, the net export of emissions showed a decline between 1913 and 1955, which in fact means a contribution to lower emissions, so in this period Sweden benefited somewhat from her foreign trade in keeping emissions down. But the balance changed and between 1955 and 1995 the contribution of net exports to total emissions increased, so in this period Sweden had no advantage from foreign trade in keeping emissions down. Just as in the case of energy exports, this means that changes in foreign trade cannot explain the decrease in CO2 emissions that took place in Sweden since the early 1970s.