ORIGINAL PAPER
麻腮风疫苗不良反应
Spatial heterogeneity of economic development and industrial pollution in urban China
Canfei He •Zhiji Huang •Xinyue Ye
Published online:7May 2013
痴ÓSpringer-Verlag Berlin Heidelberg 2013
Abstract The relationship between economic develop-ment and environmental pollution has been widely studied in the context of the environmental Kuznets curve.This study applies the three-dimension framework of density,division,and distance propod by the World Bank to identify the spatial heterogeneity of development and pollution in urban China.An inverted U relationship is detected between density and industrial SO 2emission,while a cubic relationship is found between density and industrial SO 2/soot emission intensity.The statistical sig-nificance of division indicates that the pollution haven hypothesis holds in the western region and cities in the periphery.The environmental implication of distance is that the industrial pollution is largely concentrated in the national and regional cores.
Keywords Density ÁDistance ÁDivision ÁIndustrial pollution ÁUrban China
1Introduction
As Chowdhury (2012)argued,‘‘anthropogenic environ-mental change and the future of development are two of the
greatest challenges facing human society today …Public and scientific attention to international environmental and climate change issues has rin at unprecedented rates in recent years.’’Since 1978,China has witnesd both remarkable performance of economic growth and contin-uous environmental degradation.The exceptional eco-nomic performance has been at the cost of environmental deterioration over the same time period in China (Economy 2005).The overall environmental quality is worning in China,with environmental pollution equivalent in mone-tary terms to an estimated annual loss of 3–8%of GDP.In addition,environmental pollution has caud rious health issues and economic loss.It might jeopardize eco-nomic development and sustainability.Recently,anti-pol-lution protests have successfully stopped the PX projects in Xiamen,Dalian,and Ningbo.In 2012,environmental protests occurring in Qidong of Jiangsu Province and Shifang of Sichuan Province have turned into rious social conflicts in China.The pro-growth a
nd resource intensive strategy has certainly led to the environmental deterioration and thus have raid environmental awareness in the nation (Jahiel 1998;He et al.2012a ).Searching for the appro-priate equilibrium between economic growth and envi-ronmental protection is an emerging and pressing issue in its unprecedented uneven socioeconomic and environ-mental transformation.
China changes from a socialist regime of planned economy to a more market-bad profit-eking entrepre-neurial state (Ye and Wei 2005,2012;Cao and Ye 2012;Yue et al.2012;Wu et al.2012).Specifically,the ‘‘politically correct and economically rewarding’’entre-preneurial activities are encouraged,admired,and pursued by government agencies and individuals.Meanwhile,the development–environment relationship has been widely investigated bad on the norm of the environmental
C.He ÁZ.Huang
College of Urban and Environmental Sciences,Peking
University-Lincoln Institute Center for Urban Development and Land Policy,Peking University,Beijing 100871,China e-mail:hecanfei@urban.pku.edu
X.Ye (&)
Department of Geography,Kent State University,Kent,OH 44242,USA
e-mail:xye5@kent.edu
Stoch Environ Res Risk Asss (2014)28:767–781DOI 10.1007/s00477-013-0736-8
Kuznets curve(EKC)(Shen and Hashimoto2004;He 2010;Song et al.2008).Yet,little connsus exists on the findings of EKC in China.Economic growth has both harmful effects on environmental quality and beneficial effects.The harmful effects come via the scale effects and the beneficial effects come via shifts toward more envi-ronmentally friendly ctors and cleaner production tech-niques(Grossman and Krueger1991).
佛手瓜怎样种植This paper will asss environmental pollution from a three-dimension framework propod by the World Development Report2009,‘‘Reshaping Global Economic Geography’’(World Bank2009).The three dimensions include density,division,and distance.Density refers to the economic mass per unit of land area or the geographic compactness.Division refers to economic disintegration from the global economy.Distance refers to the ea for goods,rvices,labor,capital,information,and ideas to traver across space.This framework has been success-fully applied in some ca studies.For instance,Hector and Gabriel(2009)analyzed the relationship between the three dimensions and we
lfare in11Latin American coun-tries,finding a positive relationship between density and welfare along with a negative relationship for the rest two. Roberts and Goh(2011)explained the regional differences of labor productivity in Chongqing.
Following the introduction,the cond part will provide a conceptual framework and develop the rearch hypoth-esis.The third ction will specify the variables and models.We will then prent the statistical analysis and conclude with a summary of keyfindings.
2A conceptual framework
Some recent theoretical studies have incorporated envi-ronmental pollution into the classical new economic geography model(Rauscher2008;Lange and Quaas2007). Pollution is a centrifugal force which will parate the consumers from the producers.Meanwhile,pollution might trigger stringent environmental regulation to rai produc-tion costs and reduce the economic competitiveness of the regions which have implemented such regulations.As a conquence,environmental pollution will reshape the spatial distribution of economic activities,which in turn largely determines the geographical pattern of environ-mental pollution.
Existing studies have confirmed the significance of many economic determinants in pollution emissio
n, including per capita GDP,foreign investment,industrial structure,ownership structure of local economy,capital abundance per labor,industrial GDP density,trade density, and population density(He2009,2010;He et al.2012a). The EKC literature propos an inver U-shaped relationship between economic growth and pollution (Grossman and Krueger1991).Empirical studies have investigated industrial pollution across Chine provinces and cities in terms of the EKC.A number of studies have confirmed the prence of EKC for pollutants,such as SO2 emission(Shen and Hashimoto2004;Shen2006;Song et al.2008;He2009,2010).
The pollution haven hypothesis suggests that pollution intensive industries will migrate to regions which are weakly environmentally regulated.Dean et al.(2009)and He et al.(2008)provided empirical evidence to support the pollution haven hypothesis in China.Some studies further revealed the importance of the institutional factors.For instance,pollution charges,environmental subsidies,and the bargaining power of factories and community pressures might also influence the emission behaviors of industrial polluters in China(Wang2002;Wang and Wheeler2005; Wang et al.2003;Wang and Jin2007).In addition,He et al.(2012a)found that economic liberalization and decentralization have been harmful to the urban environ-ment.In another study,He et al.(2012a,b)reported that low capability to and strong resistances to enforce stringent environmental regulations have contributed t
o the envi-ronmental degradation in urban China.Tho studies have advanced the knowledge of environmental pollution.The current rearch will link economic geography to the spa-tial pattern of environmental pollution in urban China following a three-dimension framework.
2.1Density and environmental pollution
Economic activities tend to concentrate in cities sincefirms would benefit from agglomeration effects,resulting in higher population and economic density.Economically, density results in increasing returns and will positively contribute to the productivity as revealed by a number of empirical studies.For instance,Ciccone and Hall(1996) found that doubling employment density will increa the average labor productivity by about6%.This externality in turn would attract more people and business.
The relationship between density and pollution is, however,conditionally interdependent and inconclusive. Much scholarly attention has been devoted to using density to explain the variation in pollution emission or pollution intensity(Panayotou1997;Magnani2000;Selden and Song1994;Taskin and Zaim2000;Zeng and Eastin2007; Shen2006;He et al.2008).Some studies report that pol-lution increas with density(Zeng and Eastin2007),while others suggest a negative impact of density on p
ollution (Shen2006;He et al.2008).The association remains uncertain between density and pollution.Different depen-dent variables and the samples ud in the analysis lead to inconclusive and mixedfindings.
Theoretically,there exists a nonlinear relationship between density and pollution(Panayotou1997).Hence, the association between density and pollution might change qualitatively under various circumstances.When density is lower,there may be fewer polluters and thereby less pol-lution emission.Lower-density regions may also have few concerns with environmental regulation,leading to higher-pollution intensity(Selden and Song1994).As density increas,more polluters would concentrate in a given area,resulting in higher pollution.However,when density and associated pollution reach a certain level,significant social pressure might emerge from local community and nongovernmental environmental organizations.Con-quently,local governments may implement more stringent environmental regulations,in order to force polluters to relocate.In that ca,pollution may decreas with density. Thus,an inverted U relationship is expected between density and pollution.However,if the pressure cannot directly change the environmental behavior of polluters or the density is associated with pollution intensive industries, pollution will continue to increa with density.A linear positive relationship between density and pollution would then be expected.
2.2Division and environmental pollution
英语b级单词
Division indicates how a city is engaged with globalization. Urban China has been increasingly integrated with the world economy through foreign investment and trade. Theoretically,the net environmental effect of global inte-gration is not clear.The pollution haven hypothesis pro-pos that pollution intensive industries will migrate to regions which are weakly environmentally regulated in order to save costs.Meanwhile,China may hold its lax environmental policies to give producers an advantage in the competitive international market,generating the so-called eco-dumping hypothesis(Christmann and Taylor 2001).Both hypothes assume that integration into the global economy would increa pollution in urban China. On the contrary,participation in the global economy might reduce pollution through technique and structural effects (Shin2004).For example,international trade can encour-age technological and managerial innovations beneficial to environmental improvement.Foreign investments enable favorable conditions for the diffusion of international environmental norms and standards by creating opportu-nities and necessities for environmental institution-building and policy process(Shin2004).Exports may encourage governments to ensure the compliance of their exporting products with the environmental quality standards of importing countries or international standards.
The existing studies have not generated consistent results regarding the relationship between foreign investment(or exports)and pollution in China.For instance,Christmann and Taylor(2001)found that foreign ownership and exports to developed countries increa the lf-regulation of environmental performance in China.Foreignfirms per-form in a more environmentally friendly manner than state-owned and privately ownedfirms.He et al.(2008),how-ever,reported evidence to support the pollution haven hypothesis in China,which indicates that foreign invest-ment and exports would harm the environment.The impact of division on environmental pollution may depend on the local economic development level.In the developed coastal region,environmental regulations are more stringently implemented(Van Rooij2010).Foreign investments may bring advanced technologies and environmental manage-ment.Exporters are more likely directed to developed countries.Hence,integration with global economy would benefit the urban environment in the developed regions. Facing geographical disadvantages in the globalization era, the underdeveloped inland China would discount the enforcement of environmental regulations and create pollution havens for foreign investors and exporters (He et al.2012a).
2.3Distance and environmental pollution
Distance quantifies how far a city is from the economic cores or higher-density areas.New economic
geography indicates that economic activities concentrate in the eco-nomic cores,which have the best market accessibility and large market potential.The core areas are also the rela-tively developed regions within a country.Studies report that market potential has significantly impacted economic growth in urban China(Au and Henderson2006;Hering and Poncet2010).Foreignfirms are also attracted to cities with large market potential in China(He2003).Cities significantly differ regarding the distance to economic cores,which has important environmental implications.In the underdeveloped regions,economic activities are polarized in relatively developed cities,which may host high value added pollution intensive industries.The geo-graphical proximity to economic cores may be associated with more pollution.However,industrial polluters may be forced to stay away from the economic cores in the developed regions such as the Yangtze-and the Pearl River Deltas.
下水道工In addition,distance would moderate the associations between density(or division)and environmental pollution. When cities far away from national and regional economic cores are denly populated or more involved in the global economy,they are more likely to have pollution.Tho distant cities are relatively underdeveloped and are less willing to enforce stringent environmental regulations. Moreover,they may lack the capability and suffer less
pressure to implement environmental regulations(He et al. 2012a,b).Using weaker environmental reg
ulation as their comparative advantages,distant cities would be attractive to pollution intensive industries.
3Model specification and variables
This study would empirically test the impacts of economic geography on environmental pollution.The panel data model is defined as follows:
大树的特点POLLUTION it¼aþb1Density itþb2Density2it
þb3Density3itþb4Distance iþb5Division it
þb6DensityÃDistanceþb7Division
ÃDistanceþb X itþg iþc tþe it;ð1Þwhere i and t stand for city and year,POLLUTION it is the pollution intensity or pollution emission in city i and year t.Density it,Distance i and Division it reprent the proxies for density,division,and distance of Chine cities.This study explores the cubic relationship between density and pollution,as well as the interactions between distance and density(or division)on environmental pollution.X it is a vector of controlling variables.g i is the time constant city-specific effect while c t is the unobrved time effect.This study covers285cities,including prefecture
cities,sub-provincial cities and centrally administered cities of Bei-jing,Tianjin,Shanghai,and Chongqing.Due to lack of data,Lhasa and Karamay are excluded.The study period is between2005and2010.The Urban Statistical Yearbook of China has provided data for the three pollution emissions since2004at the prefecture level cities,but with some missing data.The year2005collected data for all prefec-ture level cities.The year2010is the most recent year in which data are available.The period2005–2010makes a perfect panel data structure.
Bad on a panel data of prefecture level cities during 2005–2010,this study tests the impacts of density,dis-tance,and division on industrial pollution emission and intensity.For each dependent variable,we run three ts of model estimations.First,controlling density,distance, division,and other variables,the squared density and cubic density are ud to explore the nonlinear relationship between density and industrial pollution.Second,the interactions between distance and density/division are employed to examine how distance moderates the impacts of density and division.Third,the interactions between division and city location dummy(EAST and CENTER) are included to investigate how city locations condition the influence of division on industrial pollution.We apply the random effect models to estimate the parameters while controlling time dummies(TDummy)indicating the year of2006–2010to make the estimations robust.The three equations are as follows:
POLLUTION it¼aþb1Density itþb2Density2it
þb3Density3itþb4Distance iþb5Division it
þb X itþu TDummy tþg iþe it;ð2Þ
POLLUTION it¼aþb1Density itþb2Distance i
þb3Division itþb4DensityÃDistance
þb5DivisionÃDistanceþb X it
þu TDummy tþg iþe it;ð3Þ
POLLUTION it¼aþb1Density itþb2Distance i
þb3Division itþb4DivisionÃEast
þb5DivisionÃCenterþb X it
þu TDummy tþg iþe it:ð4Þ
A log-linear functional form is adopted to transform a likely nonlinear relationship between pollution emission and the explanatory variables into a linear one.It also reduces the outliers,nonnormality and heteroscedasticity among the residuals.In addition,the regression coefficients are measures of the elasticity of pollution emission with respect to the explanatory variables.The dependent and independent variables are discusd below.
3.1Variables for industrial pollution
This study will deal with two ts of dependent variables. Thefirst t is the annual total pollution emission,which include the industrial SO2emission(TSO2),industrial soot emission(TSOOT),and industrial waste water(TWA-TER).The other is the pollution intensity,which is the total pollution emission divided by gross industrial output at the1998constant price.In1998,the National Statis-tical Bureau adjusted the industrial statistics by only including the industrial enterpris with sales revenues over5million Yuan since the data for smaller enterpris are not reliable.The1998constant price provides a good ba to reduce inflation in industrial statistics.They include industrial SO2emission intensity(RSO2),indus-trial soot emission intensity(RSOOT),and industrial waste water intensity(RWATER).The utilization of industrial pollution is bad on the following consider-ations.First,China is experiencing rapid industrialization and urbanization,so industrial activities are t
he dominant polluters.Thus,the industrial pollution heavily affects environmental quality in urban China.Second,industrial SO2,soot and waste water emissions are largely regulated at stationary sources and would exert significant influence on environmental policies.Finally,official data avail-ability is a major concern.
3.2Proxies for economic geography
This study would introduce economic density,measured as GDP divided by the geographical size of the region.GDP is at the1998constant price.To explore the relationship between density and industrial pollution,we will also introduce the squared density and cubic density.
Distance will be measured as the distance to the national and regional core cities(Distance_Core)and the distance to the capital of the same province(Distance_Cap).To quantify Distance_Core,wefirst identify cities with pop-ulation greater than3million during2000–2010and then measure the shortest highway distance to the identified cities,which include Beijing,Tianjin,Shenyang,Haerbing, Shanghai,Nanjing,Hangzhou,Jinan,Wuhan,Guangzhou, Chongqing,Chengdu,and Xi’an.Geographical proximity to tho cores may discourage the pollution intensive industries.Chine provinces typically favor the develop-ment of provincial capitals by focu
sing on investments and strategic industries.Pollution intensive industries may locate far from the capitals to improve environmental quality of provincial economic cores.Distance_Cap is measured as the highway distance to the provincial capitals and expects positive coefficient.Figure1shows the spatial distribution of lected economic cores in China.
Division is measured as the extent that Chine cities are integrated with the world economy.Chine cities partici-pate in the global economy through utilizing foreign investment and international trade.We introduce the ratio of output by foreignfirms in gross industrial output(FDI),ratio of exports and imports in GDP(TRADE),respectively.
3.3Control variables
Following Cole et al.(2008),this study adopts the pollution demand–supply framework to identify control variables.This framework considers pollution emissions as environ-mental rvices.Pollution demand is defined as an industry’s demand for environmental rvices,while pollution supply is defined as the quantity of pollution that an industry is allowed to emit within a community.We include the ratio of investment for environmental protection in GDP(INVEST). This investment would reduce pollution demand in urban China.Shen(2006)found that investment in environmental protection woul
d significantly reduce pollution in China. The high energy-consuming industries generate the majority of the industrial air pollution in China.The Chine econ-omy is highly dependent upon the production from heavy industries which tend to require high levels of raw material and energy inputs.Energy u is therefore likely to be a strong positive determinant of industrial air pollution.The more energy intensive production would result in a greater demand for pollution.The energy intensity(ENERGY)is ud to indicate the pollution demand.Since the dependent variables are concerned with industrial pollution,we further control the ratio of condary industries in GDP(SECOND). More industries would also lead to more pollution demand.
Pollution supply is determined by polluters and the enforcement of environmental regulations(Cole et al. 2008).Studies have reported that state-owned enterpris (SOEs)are the major polluters since many of them are in the pollution intensive industries and also enjoy more bargain power with local governments(Wang and Jin 2007;Wang and Wheeler2005).The ratio of SOEs in gross output will be included.The enforcement of environmental regulation is often limited by lack of resources and per-sonnel.Strong capability to implement environmental regulation would lower pollution(He et al.2012a,b).We include the number of employees in the EPBs(Environ-mental Protection Bureau)at the county level per10,000 people to quantify the capability of enforcement(RBU-REAU).In addition,s
ocial pressure from local community and nongovernmental organizations(NGOs)would also help reduce pollution supply.We further quantify the pressure using environmental complaints per10,000people (RLETTER)and expect a negative coefficient.
全媒体时代Finally,we will explore the impacts of interactions between location and division on environmental pollution.We simply introduce two dummies indicating cities in the coastal and central regions(EAST and CENTER).The dependent and independent variables are summarized in Table1.
4Data analysis
4.1Spatial pattern of pollution intensive industries Bad on the pollutant emissions,four pollution intensive industries are analyzed,including power generation
and Fig.1Selected economic cores and provincial capitals in China
heat supply,textile,paper making and-product and chemicalfiber manufacturing.In terms of number offirms, all four industries are relatively disperd in China,both coastal and central China host a large number of polluting plants in tho industries.Power generation and heat sup-ply is the most disperd industry.Plants in this industry not only concentrate along the coastal province such as Liaoning,Shandong,Zhejiang,Fujian,and Guangdong,but also in the inland provinces including Inner Mongolia, Hunan,Sichuan,and Yunnan.Plants in chemicalfiber industry are highly concentrated in the coastal provinces, which host91%of them.The coastal provinces also host 85and74%of plants in paper making and-product,and textile industries.Meanwhile tho two industries have started to diffu to the central China along the Yangtze River.The gross industrial output ems relatively more disperd(Fig.2).This is especially phenomenal for the industries of paper making and-product and power gen-eration and heat supply.Spatial pattern of pollution intensive industries significantly correlate with that of industrial pollution.
4.2Spatial distribution of industrial pollution
Figure3shows the distribution of the total pollution emis-sions in2010,including industrial SO2emissio
n,industrial soot,and industrial waste water.There are veral interestingfindings.First,both industrial SO2and industrial soot emissions heavily concentrate in North China and Northeast China,including Shandong,Henan,Shanxi,He-bei,Inner Monglia,Liaoning,Jilin,and Heilongjiang.Cities in Shandong and Henan demonstrate substantial industrial SO2concentrations.North China hosts many energy intensive and heavy industries,which are the typical pol-luters.Second,due to the industrialization,the Yangtze-and the Pearl River Deltas also have many polluters of industrial SO2and soot.Tho two regions have significantly bene-fited from globalization.They have attracted many foreign firms and developed privately owned enterpris,which are costly to enforce stringent environmental regulations.The recent process of heavy industrialization has introduced heavy industries including machinery,transportation equipment,chemical industry,and petroleum refining, contributing to industrial pollution emission.
Third,compared with industrial SO2and soot,industrial waste water emission is more likely to concentrate in the coast while less likely to occur in Northeast China.It is particularly phenomenal that the concentrations of indus-trial waste water emission are identified in the downstream cities along the major rivers including the Yellow-,the Yangtze-,and the Pearl Rivers.Fourth,some inland major cities such as Chongqing,Urumqi,Guizhou,Baotou, Hohhot,Lanzhou,and Kunming accommodate a large amount of industrial pollution.Tho are the economic
Table1Definition of dependent and independent variables
Variables Definitions Expected sign Time
肉疼是怎么回事RSO2Industrial SO2emission/gross industrial output Dependent2005–2010 RSOOT Industrial soot emission/gross industrial output Dependent2005–2010 RWATER Industrial waste water emission/gross industrial output Dependent2005–2010 TSO2Total industrial SO2emission Dependent2005–2010 TSOOT Total industrial soot emission Dependent2005–2010 TWATER Total industrial waste water Dependent2005–2010 Density GDP per squared km?2005–2010 Distance_Core Shortest highway distance to major cities?2010 Distance_Cap Highway distance to the capital of the same province?2010
FDI Percentage of output by foreignfirms in gross industrial output?2005–2010 TRADE Percentage of imports plus exports in GDP?2005–2010 INVEST Percentage of investment for environmental protection in GDP in the previous year–2004–2009 SECOND Percentage of condary industries in GDP?2005–2010 ENERGY Energy consumption divided by GDP or energy intensity?2005–2010 RSOE Percentage of output by state-owned and state-controlled enterpris in gross industrial output?2005–2010 RBUREAU Number of workers at county level EPB(Environmental Protection Bo
ard)per10,000residents–2005–2010 RLETTER Environmental complaints per10,000residents in the previous year–2004–2009 EAST Dummy for cities in the coastal provinces2010 CENTER Dummy for cities situated in the central provinces2010
