Study On The Impact Of International Trade On China’s Environmental Pollution

Economic globalization makes the environme ntal problems become an international problem gradually. With continuous economic de velopment, the environmental pollution has become more and more serious in lots of countries, especially in developing countries. As the largest developing country, C hina has paid huge cost of environmental pollution during its high-speed development. Since C hina has become more and more polluted, the pollution scope is also expanding, from the most developed east area to middle and west areas, from sky to ocean, from surface to underground, referring air, water, earth pollutions of different degrees. These pollution not only threathen the survival and health, but also bring heavy economic costs. The conflict between trade and environment is from the contradiction between the growing need for resources by expanding trade and limited supply of resources. In the stage of C hina’s comprehensive deepening of reform and expanding of openness, it is meaningful to study the relationship between trade a nd environmental pollution, to understand the big environmental cost from trade correctly, to promote the coordinated development of trade and environment.Based on previous studies, this paper makes a comprehensive and in-depth theoretical and empirical analysis about the impact of international trade on environment. We firstly built theoretical frame by constructing the general equilibrium model of trade and environment pollution, and then discuss specific channels through which trade affects environmental pollution by mechanism analysis. Except for the traditional scale effect, composition effect and technique effect, we also add income effect, market competing effect and reallocation effect. In analysis of reallocation effect, we assume that the heterogeneity of enterprises shows in environmental pollution characteristics, which expand the hypothesis that all enterprises are homogeneous. During empirical research, choosing industrial pollutant and suspended particulate emissions as our key objects of study, using differenct levels of sample data, applying quantitative analysis and comparative research methods, based on the international comparison of the environmental effect of trade, we focus on the impact of Chinese foreign trade on environmental pollution and put forward the corresponding trade and environment policy suggestions to coordinate the trade and environment.The main parts of this paper consist of 4 chapters. In chapter 3, we build a n open economy general equilibrium model consisting of two sectors and study the various channels of trade affecting environmental pollution. Trade could affect a nation’s environmental condition by affecting its economic activity scale, product structure, clean technology diffusion, which we call scale effect, composition effect and technique effect. At the same time, if we assume that consumers with different income level have different expectation for environment, trade can affect the need of environment by affecting income level. In market competing effect, we stress mainly on the power resources of enterprises’ technology innovation when they face tremendous pressure in international competition, which will be the power and support for achieving energy-saving and emission reduction. In the description of the reallocation effect, without the original hypothesis of enterprises homogeneity, we assume that enterprises differ not only in productivity, but also in emission efficiency, pollution control input and so on. There is a negative correlation between productivity and pollution intensity. Trade then sends resources to more productive export enterprises, i.e. more environmental efficient enterprises. Even though the pollution intensity of individual enterprise does not change, the average industry pollution intensity falls down because of trade, which will affect environmental pollution subsequently.Chapter 4 to chapter 6 is empirical study. We divide into three levels. First, we investigate the impact of trade on environmental pollution using international sample data in chapter 4. We choose 40 countries’ panel data who account for 50% of world ’s trade volume to regress. In control variables, except for income level per capita, population, proportion of industrial added value, FDI, we also control the emission related variables like enery consumption per GDP unit, proportion of fossil, along with time fixed effects and individual fixed effects. The results show that trade is beneficial to environmental condition of developed countries, but not for developing countries. Therefore, the impact of trade on environemtal pollution is different with different country type. And different pollutants and emission indexes will have different conclusions. In order to compare the gap between China and other countries, we make China as the opposition group of other sample countries, the regression results show that the increasing of trade-GDP ratio will reduce SO2 per capita and intensity, but raise the NOX per capita and intensity. C hina’s conclusion is different with general developing countries. Therefore, it is necessary to study the impact of trade on China’s environmental pollution from more angles.Referring to the C hina problem, we first do research through C hina provincial level panel data, using both static model and dynamic mode method to verify the effect of trade on environmental pollution. In order to answer the question that whethe r this effect will be disturbed by choosing different pollutants、areas and trade modes, we choose industrial SO2, three industrial wastes and comprehensive index of environmental pollution building therefrom to proxy environmental condition, and make analyses of the east、middle and west regions and different trade modes. In addition, considering the limitation using only trade-GDP ratio to proxy trade openness, we calculate the amended trade-GDP ratio through method of Shen(2005). IV method and system GMM method are used to solve the endogeneity problem derived from the reverse causality between trade and environment. The results of 2SLS using IV and system GMM are very different with OLS results which prove that it is absolutely necessary to solve the endogeneity problem.The results show that the impact of trade on China’s environmental pollution is uncertain and complex. Conclusion will be different using different pollutants, indexes, areas and trade modes. In detail, in stastic panel data regression, increasing of trade-GDP ratio is in favour of reducing industrial SO2 intensity, three industrial wastes intensities and pollutant discharge fee weighted comprehensive index 1,but rasing arithmetic mean comprehensive index 2. In the sub regional regression, the sign of trade elasticity of different regions are the same only when the dependent variables are industrial solid waste intensity and comprehensive index 2. When using other 4 pollutants as dependent variables, trade is beneficial to western environmental pollution, but unfavourable for the east and middle areas. We find from regression of different trade modes that all kinds of pollutants’ emission level drop with the increasing of conventional trade export-GDP ratio, but the emission level of industrial SO2, gas and solid waste intensity and index 2 step up with the increasing of processing trade export-GDP ratio. Therefore, from the existing results, conventional trade is good for environment, but processing trade may be bad for it. We only use industrial SO2 index in dynamic panel data because of data completeness. The results shows trade will reduce the industrial SO2 emission.In stastic panel data, we creatively use an instrumental variable of exchange rate shock, which is the weighted average exchange rate of China provinces’ largest trade partner countries, as proxy of trade-GDP ratio to regress with 2SLS. In economic research, exchange rate is generally regarded as the exogenous variable. Especially after 1997’s Asian financial crises, all governments began to strengthen macro-control of the country’s exchange rate, which is a good reason why exchange rate can be treated as exogenous variable.Another advantage of our IV is that as a countr y level variable, individual province’s trade is not enough to change the exchange rate. That means the individual enterprise of each province can not forcast the exchange rate change. We also use the weighted exchange rate to increase the rationality of IV.In dynamic panel data, we use interior IV to regress by system GMM. Because the lag dependent variable joine in explanatory variables, we need GMM method. Althouth the level GMM, differential GMM and system GMM all could be used, but a lot of papers have proved that system GMM performs better, so we choose system GMM. From the long term elasticity, we find that the beneficial effect of trade on industrial SO2 is more obvious in the long run. In the dynamic sub region regression, we find that trade can only improve the environmental condition of west area, but may deteriorate to east and middle areas, which keep the same conclusion with stastic panel data.In order to study this problem form more micro level, we use China’s city level data in chapter 6. We first use 74 cities cross-section data. When choosing pollutant, this chapter uses PM2.5, which has received extensive attention recently. PM2.5 is called fine particles in China. Its primary source is the burning of coal, diesel oil and industrial produc tion. It also can be formed by chemical reaction of other pollutants. So it is very typical and innovative to choose PM2.5 as our study object. PM2.5 has become the priority in pollution prevention because its major hazard to human health. It has very impo rtant practical significance to study whether trade will increase the PM2.5 emission. The regression shows a negative result.In cross section data regression, we use in the two stage IV method by Blanchard and Perotti(2002). When choosing the IVs for PM2.5 in the first stage regression, according to Chen et al.(2013), we choose the latitude distance between cities and Huai River to proxy PM2.5 in view of China’s Huai River policy, which provides central heating in cities north of the Huai River but not in the south via the burning of coal which is the main reason of PM2.5. We find that one percentage increase in PM2.5 decreases trade-GDP ratio by 2.008 percentage points. We use the first stage regression residuals to instrument trade openness in the second stage regression in order to solve the reverse causality effect. The result shows that one percentage increase in trade-GDP ratio increases PM2.5 by 0.378 percentage points as well, which means trade is bad for environmental pollution, while the OLS estimate is much different. O ur analysis tells that the effects of trade on environment will be substantially undervalued, reaching wrong conclusion if we do not consider the reverse simultaneous bias. The elasticity of export dependence and heavy industry export dependence reach to about 0.65, while the elasticity of intensive sectors export dependence is as high as 0.795. So the mechanism analysis shows that exports of pollution intensive sectors lead to the pollution.We also estimate the casual effect of trade on PM2.5 on the basis of 334 cities’ panel data. In order to control the endogeneity problem, we make empirical analysis based on different model settings, static and dynamic panel model, and implement different IV strategies, like lagged explanatory variable, geographic distance as external IV and system GMM to control endogenous variable.The results show that PM2.5 emission increases with greater trade-GDP ratio, no matter by static or dynamic model. In 2SLS regression, we find that a one percentage increase in trade-GDP ratio decreases the emission of PM2.5 by 0.06-0.25 percentage points. The lagged dependent variable coefficient is significantly positive, which means the emission is continuous. The GMM result also shows that trade has more influence on the coming environmental pollution.Based on the theoretical and empirical research, we put forward some policy suggestions from trade and environment sides. The trade policy suggestions include implementing different trade policies in different regions of C hina, coordinating trade policy and environmental policy,accelerating industrial structure adjustment and upgrad ing of east and middle areas, increasing the openness of C hina’s western regions, rasing environmental protection threshold in undertaking the industrial transfer process, improving actively the structure of import and export commodities, playing the "going out" strategy, increasing the introduction of environmental protection equipment and cleaner technology, adusting pattern of trade, speeding up the transformation and upgrading of processing trade, making policies of international trade from the long-term angle and so on. The environment policies include suiting local conditions when making environmental policies, using different standards in the national scope, making different emission standards according to different pollutants, making a comprehensive consideration of different indexes of the same pollutant, paying attention to the continuity of environmental policies and so on.This paper draws the following conclusions.First, generally speaking, the impact of trade on environmental pollution is uncertain. Second, the impact of international trade on environmental pollution of different types of countries is different. Trade tends to improve the situation of environmental pollution in developed countries. Third, the impact of international trade on pollutants of different types in C hina is different. Fourth, trade tends to reduce the environmental pollution level in western area of C hina, but tends to increase pollution level in eastern and middle areas. Fifth, conventional trade is not bad for C hina’s environmental pollution, but processing trade may increase emissions of some pollutants. Sixth, compared to the entire international trade, export should bear the main responsibility for the deterioration of air pollution in our city. Seventh, different measuring indexes may influce the conclusion. Eighth, long-term effect of trade on environmental pollution is greater than the short-term effect, and pollution has the characteristics of path dependence.The main innovation of this paper is that we comprehensively analyze the main channels from which trade can affect environmental pollution, while do the empirical research from multi levels, multi angles and multi methods. We use international data, C hina’s province level and city level data.In city level data, the sample number is 334, and we use the enterprise level data in mechanism analysis. In the perspective of analysis, we not only consider the different measurement standard of trade and pollution, but also different areas and trade modes. In research methods, we try several empirical methods, for example, the IV method, system GMM method. In the selection of external IV tools, we use weighted average exchange rate of China provinces’ largest trade partner countries as proxy of trade-GDP ratio. We also use IV based on heating policy differences between two sides of huai river which is a special case of China.In the future research on the trade effect of environmental pollution, further study can be extended by the following aspects. First, test by more pollution of different types. Second, find a more scientific method to construct the comprehensive index of environmental pollution. Third, test the impact of trade on environmental pollution from angle of industries and nationalities. Fourth, consider the influence of international trade agreements and environmental agreement. Fifth, study the specific impacts of various channels.