The determinants of aggregated electricity intensity in China.docx
《The determinants of aggregated electricity intensity in China.docx》由会员分享,可在线阅读,更多相关《The determinants of aggregated electricity intensity in China.docx(20页珍藏版)》请在冰豆网上搜索。
ThedeterminantsofaggregatedelectricityintensityinChina
ThedeterminantsofaggregatedelectricityintensityinChina
ZHADonglan
CollegeofEconomicsandManagement,NanjingUniversityofAeronauticsandAstronautics,29YudaoStreet,210016Nanjing,China
ABSTRACTThisstudyprovidestechnologicalchange,factordemandandinter-factorsubstitutabilitymeasuresforchinaelectricityindustry.Weuseindividualfuelpricedataandtranslogfactorcostfunctionapproachtoestimatetotalfactorcostfunctionsandfuelshareequations.TheimpliedpriceandMorishimasubstitutionelasticitiesforinter-factorareobtained.Byestimatingthesubstitutionbetweenfactors,ourpurposeistoexplorethedrivingforcestoaggregateelectricityintensity.Theresultssuggestthatenergyissubstitutableforcapitalandcomplementaryforlaborregardingcross-priceelasticity.CapitalisMorishimasubstitutiontoenergyandlaboriscomplementtoforenergy.Factorsubstitution,budgetandtechnologyeffectscontributetotheyear-to-yearvolatilityofelectricityintensityintheproportion48%:
41.99%:
9.89%.
Keywords:
China,Electricityintensity,Factorsubstitution
1.Introduction
Withprominentdecreaseof69percentofChinaenergyintensityin1985-2002andareversalsince2003,avastmajorityofstudiesattempttoexplainthisphenomenon.Thedebateofthedrivingforcestothepersistentdeclineofenergyintensityhasbeenrevolvingaroundstructuralshiftsandefficiencyimprovements.JustasCromptonandWu(2005)pointedoutthatChina'sdecliningenergyintensityisoftenexplainedintermsoftechnologicalandstructuralfactors.
Atthebeginningof1990s,researchersarguedthatthestructuralshiftsawayfrommoreenergy-intensiveindustrialsub-sectorsplayedthemostimportantroleinChina(e.g.,Smile,1990;Kambara,1992).Sincethen,theliteraturesusuallyacceptedtheviewthatthecontinuousdeclineinenergyintensitywasmainlyattributedtotheenhancedefficiencyofenergyuseofindustrysectors,whilethestructuralshiftswithinthemanufacturingsub-sectorsorfromprimarytosecondaryortertiaryindustryplayonlyanominalrolefrom1980totheearly1990s(SitonandLevine1994;Zhang,2003).Garbaccioetal.(1999)evenfoundthatstructuralchangeactuallyincreasedenergyintensitybetween1987and1992.Sincethemid-1990s,efficiencyimprovementshavebeenparticularlymarkedinenergy-intensiveindustriessuchasmetallurgy,cement,paper,textiles,oilandcoalprocessing,andelectricalpowergeneration(Philip,2009).Contrarytotheabovementionedstudiesfocusingonthedeclineofenergyintensity,thereareliteraturesspecializingontheunusualphenomenonofenergyintensityincreasingsince2003.MaandStern(2008)indicatedthattheincreaseinenergyintensitysince2000couldbeexplainedbynegativetechnologicalprogress.Zhaoetal.(2010)foundthatthemostimportantdriverbehindChina’senergyintensityincreaseduring1998-2006wastherapiddevelopmentinenergy-intensiveindustries.
Thecommonfeaturethepreviousstudieshaveisthattheyusuallyapplytheindexdecompositionmethodwhichisusedtotracktheenergyintensity/efficiencytrendandassessthefulfillmentofenergyefficiencytargets.Ashortcomingofpurelydescriptivedecompositionstudiesisthatitcannotexplainwhatisbehindenergyefficiencyimprovementsandenergy-savingstructuralchange(WelschandOchsen,2005).Toourunderstanding,notonlytheenergyefficiencyimprovementscouldbecausedbytechnologyadvancementandfactorssubstitution,butalsobepossiblebytheenergypricechanging.Thisviewissupportedbyotherresearchers.Employingasetofpaneldataforapproximately2500ofChina’smostenergyintensivelargeandmedium-sizedindustrialenterprisesduring1997-1999,Fisher-Vanden(2004)obtainedthatrisingrelativeenergypricesaccountfor54.4%ofthedeclineinmeasuredaggregateenergyintensity.ByreviewingofthederegulationofenergypricesinChinabetween1985and2004,HangandTu(2007)impliedthatanincreaseofrelativepricesofdifferentenergytypesleadstoadecreaseincoalintensity,oilintensity,andaggregateenergyintensity.ShiandPolenske(2005)foundanegativeown-priceelasticityofenergyintensity,aprice-inducementeffectonenergyefficiencyimprovement,andagreatersensitivityoftheindustrysector,comparedtotheoveralleconomy.Zhaoetal.(2010)arguedthatlowenergypriceshavedirectlycontributedtohighindustrialenergyconsumptionandindirectlytotheheavyindustrialstructurealthoughtheydidn’tempiricallygetwhethertheenergypriceshaveeffectonenergyintensity.
Thispaperalsoattemptstoinvestigatethedeterminantsofenergyintensity.However,wefocusonelectricpowerindustry.Toourunderstanding,fewliteratureshavefocusedonpowersectoronthisissue.Therestofthepaperproceedsasfollows.ThefirstpartofthispaperoutlinestheelectricityproductionandconsumptioninChina.Followingthis,themethodologyanddataaredevelopedtocalculatetheinterfactorsubstitutabilityanddecomposeChina’schangingelectricityintensitytoascertainthedrivingforces.Theempiricalfindingsandtheinterpretationoftheresultsarepresentedinthefourthsection.ThelastsectionprovidesconclusionsanddiscussestheimplicationsofourempiricalresultsforincreasingChina’selectricityefficiency.
2.ElectricityproductionandconsumptioninChina
Electricityisoneofthemostimportantandbasicneedsoftodaycommunity.AccordingtoCSY(2008),netelectricitygenerationinChinain2007totalled3281TWh,83percentofwhichwasthermalpower,followedbyhydropowerwith14.8%.Thethermalpowerwasgeneratedthroughthecombustionoffossilfuels.Coalplaysthelargestroleingeneration,followedbypetroleumandthennaturalgas,witheachrepresenting97.46percent,1.38percent,and1.16percentof2007’stotalnetgeneration,respectively.Itisobviousthatfossilfuelcombustiondominatesingeneratingpower,despiteitsnegativeenvironmentalexternalities.Moreover,coal-firedpowergenerationwillbeinastageofstabledevelopmentuntilatleast2020.
Before2003,theStatePowerCorporation(SPC)dominatedtheproductionandsupplyofelectricpowerinChinaandownedalltransmissionnetworksandalargeproportionofthedistributionfacilities.Attheendof2002,thegeneratingassetsoftheSPCweredivestedtofiveindependentpowerproducerstoincreasethecompetitivenessofthepowerindustryandrevamppricingmechanisms.Despitetheremarkablegrowthofelectricitygenerationexperiencedfrom1985to2007,thecurrentbalancebetweenpowersupplyanddemandinChinaisstillatalowlevel.Since1998,demandforelectricityhasacceleratedbeyondwhatmanyeconomistshadinitiallypredicted.Consequently,anumberofseriouseconomicproblemshavebeguntoemerge.
TheaveragepowernetworkenergylossrateinChinawas6.97%in2007,whichmeanselectricityof229TWhwaslostinpowergridin2007.Lossesintransmissionin2007were206TWh,accountedfor6.72%ofconsumptionbyusage.ItindicatesthatChina’spowertransmissionsystemremainsunder-improvedandtherearemuchpotentialinenhancingtheratesofelectricityutilizationefficiency
Withrespecttoelectricityconsumption,since1980,itincreasedfast.From1985to2007,itgrewatannualaveragingnearly6.87%,versus4.43%perannualforcoaland5.46%perannualforpetroleum.Thereasonsarenumerous,andmainlyledbystronggrowthinindustrialactivityandincreasingpenetrationsofelectricalequipment,spacecoolingandotherelectricappliancesinbothindustrialandresidentialsectors.Fig.1showstheconsumptionstructureofelectricitybysectorsin1985and2007.Industrycontributesmostofthenation’stotalelectricitydemandalthoughitdecreasedfrom79.74%to75.3%.Theresidentialsector'scontributiontototalnationalelectricitydemandrosefrom5.4%to11.07%duringthesameperiodtimes(23years)(CSY).Itwaslargelyduetotheincreasingpenetrationofelectricalapparatussuchastelevision,spacecoolingtothefamilywithhigherincome.
Fig1.Electricityconsumptionstructure
3.Methodologyanddata
Thetranslogproductionfunction(TPF)developedbyChristensenetal.(1973)introducesinteractiontermsandcanbeestimatedinasymmetricsystemofderivedfactorshareequationsthatimproveestimationpropertiesrelativetoasingleequation.Ageneralprobleminestimatingaproductionfunctionisthatinputfactorsarelikelyendogenous,therebyviolatingabasicnecessaryconditionforordinaryleastsquarestobeunbiased.Byusingfactorpricesinacostfunctionframeworkthisparticularproblemismostlikelycircumvented,somostempiricalstudiesestimateatranslogcostfunctioninsteadofatranslogproductionfunction(Koetse,2008).
Ourpointofdepartureistoestimatethefactorsubstitution.Followingthepreviousstudiesonfactorsubstitution,weusethetypicalapproachreflectingfactorsubstitutabilitywithatranslogcostfunction.Theunitcostfunctioncanbestatedas:
(1)
WhereLnindicatesthenaturallogarithm;Cmeanstheequilibriumtotalcost;Pit(Pjt)isthepriceofinputfactori(j)attimet;tisatimevariabletocapturetechnicalchange.
Linearhomogeneityinputprices,aninherentfeatureofanycostfunctionrequiresthefollowingregularityconditions:
and
forj=1,…n.
(2)
ApplyingtheShephard’slemma,wecanderivealinearexpressionoftheshareofoverallcostattributabletoeachfactori:
(3)
Here,
isappointedasdistributionparametersand
assubstitutionparameters(Christensenetal.,1973).Theformermeasureshowtheco