statusxyq.cbg.163.com

20173rdInternationalConferenceonSocialScienceandManagement(ICSSM2017)ISBN:978-1-60595-445-5TheResearchontheEffectsofInternettoChineseEconomicGrowth:BasedontheEmpiricalAnalysisofProvincialPanelDataRen-qiangXIE1,a,*,Fu-lanYE1,Zhen-yuXU11ResearchCenterforScienceTechnologyandSociety,FuzhouUniversityofInternationalStudiesandTrade,Fuzhou,Fujian,Chinaacbg1567@163.
com*CorrespondingauthorKeywords:InternetDevelopment,EconomicGrowth,PanelDataModel.
Abstract.
ThearticlemakesanempiricalanalysiswithEviews7.
0.
Weselected2010-2014in31provincialregionsGDP,Internetdevelopmentlevelindex,fixedassetinvestmentandemploymentpopulationindex.
WeusetheextendedDouglasfunctiontoestablishapaneldataregressionmodeltostudythecontributionoftheInternetdevelopmenttoChina'seconomicgrowth.
Basedontheresults,weputforwardsomepolicyrecommendations.
IntroductionTheInterneteconomyandeconomicgrowtharereallyrelevant,anditisnecessarytostudytherelationshipbetweenthem.
IfwecangettheroleoftheInternetineconomicgrowth,wecanfurtherpromoteChina'seconomicgrowththroughtheInternet.
FewscholarsusepaneldatatostudytheroleofInternetineconomicgrowth.
ThispaperusespaneldatatoanalyzetheimpactofInternetresourcesoneconomicgrowth,andhopestoprovidesomepracticalsuggestionsforthetransformationofeconomicdevelopmentmode.
1.
ModelIntroduction1.
1PanelmodelInstatisticsandeconometrics,thepaneldatareferstomulti-dimensionaldatafrequentlyinvolvingmeasurementsovertime.
Paneldatacontainobservationsofmultiplephenomenaobtainedovermultipletimeperiodsforthesamefirmsorindividuals.
Inotherwords,itisthesametimeinthetimeandcrosssectionofthethree-dimensionaldata,alsoknownastimeseriesandcrosssectionaldata,whichisthecross-sectionofindividualsatdifferentpointsintherepeatedobservationdata.
Themixedregressionmodelisthatallcoefficientsdonotchangewithtimeandindividuals.
Thereisnosignificantdifferencebetweentheindividualandthecrosssectionintermsoftimeandcrosssection.
Atthispointwecanestimatetheparametersofthepanelmodelwiththeordinaryleastsquaresmethod.
Suchastheformula1.
,1,2,,;t1,2,,TαβεitititYXiN(1)Thefixedeffectsmodelisdividedinto3types,respectively,suchasformula2,3and4.
1,2,,;t1,2,,TitiititYXiNαβε2)Randomeffectsmodelcanbeusedtodescribethelossofinformationduetothelackofmodel.
Seeformula5.
01,2,,;t1,2,,TititititYXiNααγβε5)1,2,,;t1,2,,TitiititYXiNαβε3)01,2,,;t1,2,,TititititYXiNααγβε4)1.
2ExtendedCobb-DouglasmodelTheCobb–Douglasfunctionisaparticularfunctionalformoftheproductionfunction,widelyusedtorepresentthetechnologicalrelationshipbetweentheamountsoftwoormoreinputs,particularlyphysicalcapitalandlabor,andtheamountofoutputthatcanbeproducedbythoseinputs.
Sometimesthetermhasamorerestrictedmeaning,requiringthatthefunctiondisplayconstantreturnstoscale.
Basedonthepaneldata,thelinearregressionoftheDouglasfunctionisobtained:itititYAKLαβ=(6)0ititAAR=(7)Will(3-7)into(3-6):0ititititYAKLαβγ=(8)Onbothsidesoftheequalsignlogarithm(3-8):0ln1,2,3,.
.
.
,31;2010,2011,.
.
.
,2014γαβξ==itititititYLnALnLnKLnLit(9)(Ystandsforoutput,RstandsfortheInternet,Kstandsforcapital,andLstandsforlabor)2.
IndexandData2.
1IndexselectionThispaperisbasedonthedevelopmentstatusoftheInternetreleasecalendarChinesereport,andconsideringthetimeandeasytocollectdata.
WefinallychoosetheprovincesandautonomousregionsofthenumberofInternetusers,thenumberofsitesandthenumberofdomainnamesofthesethreeindicatorstobuildasetofdatatorepresentthelevelofdevelopmentoftheInternetIndex.
GDPintheregiontorepresenttheoutputofY,thetotalamountofinvestmentinfixedassetsinvariousregionstorepresentthecapitalfactorK,thetotalnumberofemploymentintheregiontorepresentthelaborforceL.
2.
2DataprocessingTheempiricalanalysisoftheoriginaldataisthemainsourceofChineseBureauofstatisticsreleased2011-2015"ChineseStatisticalYearbook",the"2011-2015ChinaInternetdevelopmentstatisticsreport",aswellastheprovinces,municipalitiesandautonomousregions,thestatisticalyearbookofstatisticsbulletin.
Thispaperselectstheindexarea2010-2014yearsofChina'sprovinces,municipalitiesandautonomousregions,GDP,totalinvestmentinfixedassets,theemploymentpopulationandthedevelopmentoftheInternet,whichhasover5year.
Thenumberofthepaneldataof31.
CalculationandanalysisoftheprocessusetheEXCEL,Eviews7.
2andothersoftwaretocomplete.
Thenumberofsites,thenumberofusersandthenumberofdomainnamestotakethenaturallogarithm,werenamedI1,I2,andI3.
Thestandardizedvariableswerenamed*1I,*2I,*3I.
UsingGPCAmethod,weputthe31provinces,municipalitiesandautonomousregionsin2010byahorizontalarrayof3indicatorstoformadatatable,sothereare5suchdatatablesfrom2010-2014.
Weputthese5datatablesfromtoptobottom,andthismatrixisaglobaldatawithboththecrosssectionandthetimeseries.
Then,wecanmakeaprincipalcomponentanalysisofthisdatasheet:Itcanbeseenfromtheresultsthatthecumulativecontributionrateofthefirstprincipalcomponentis91.
75%,whichisgreaterthan90%.
ThefirstprincipalcomponentisnormalizedandtheresultsareusedastheInternetdevelopmentindexR.
Then,thefourvariablesofGDP,fixedcapitalinvestment,employmentandInternetdevelopmentindexweretreatedbylogarithm.
Table1.
Contributionrateandexpressionofeachprincipalcomponent.
principalcomponentContributionrateCumulativecontributionrateConcreteexpressionFirstprincipalcomponent0.
91750.
9175***11230.
55940.
588870.
5834FIII=++Secondprincipalcomponents0.
06770.
9852***21230.
82470.
32580.
4620FIII=Thirdprincipalcomponents0.
01481.
0000***31230.
08190.
739670.
6679FIII=+3.
EmpiricalAnalysis3.
1UnitroottestInthispaper,wewillmakeaunitroottestforeachvariableinthepaneldata.
Weusetwokindsofunitroottestmethods,namelythesameunitroottestLLCandtheADF-Fishertestofdifferentunitroottest.
Ascanbeseenfromtable2,theoriginalvaluesofthefourvariablesintheprocessoftesting,inadditiontolnRandlnL,arerejectedbytheoriginalhypothesis.
Thatis,withouttheexistenceofaflatrootvariable.
However,afterthefirstorderdifference,lnRandlnLcanstronglyrejectthehypothesisthatthereisaunitroot.
Thatis,thefirstorderdifferencebetweenlnRandlnLisstable.
Therefore,thepaneldataoftheabovevariablescanbedeterminedasoneorderstationarydata.
Althoughtheexperimentaldatahasbeentestedstability,consideringtheinstabilityofpaneldatabasedonthedata,thecointegrationteststudiesthelong-termequilibriumrelationshipbetweenthevariablestoensurescientificempiricalresults.
Table2.
Unitroottestforeachvariable.
variableHorizontalvalueFirstorderdifferencevalueLLCADF-FisherLLCADF-FisherlnY-8.
5251(0.
0466)97.
4242(0.
0016)----lnR8.
0692(1.
0000)17.
5959(1.
0000)-32.
2265(0.
0000)134.
6160(0.
0000)lnK-12.
2907(0.
0000)90.
2357(0.
0111)----lnL-3.
2021(0.
0007)76.
9563(0.
0956)-4.
5346(0.
0000)83.
3353(0.
0248)3.
2CointegrationtestThispaperchoosesKao(1999)presentedinpaneldatacointegrationmethodtotesttheexistenceofseveralvariablescointegration.
KaotestistestedbytheEngle-Grangerdevelopment,theoriginalhypothesisis:thereisnocointegrationrelationshipbetweenvariables.
Table3.
Cointegrationtestbetweenvariables.
variableKAOtestT-ADFtestvalueADFprob.
lnY,lnK-6.
7717400.
0000lnR,lnL-2.
6349830.
0042lnY,lnK,lnR,lnL-7.
0918150.
0000Fromthestationarytest,thatoflnYandlnKsequenceisstable,andlnRandlnLsequenceisthefirst-orderdifferencestationary,sothispaperrespectivelyonlnYandlnKsequence,lnRsequenceandlnLsequence,fourofthethreesetsofDataCointegrationtest.
Accordingtotable3testresults,fromthestationarytest,thesethreegroupsofADFPvaluesare0.
0042,0and0,farlessthanthesignificancelevel0.
05,sowerejectthenullhypothesis.
Thereisacointegrationrelationshipbetweenthesefourvariables.
Thereisalong-termstableequilibriumrelationshipbetweenGDP,Internetindex,fixedcapitalinvestmentandemploymentpopulation.
3.
3ModelestimationandselectionThispaperhascarriedontheregressionofthreemodelsforpaneldata,andthedefaultchoiceofindividualfixedeffectsmodelinfixedeffectmodel,gettheregressionresultsintable4.
WeshoulduseFtesttodeterminewhetherthemixedregressionmodelortheindividualfixedeffectmodel.
Fstatisticsareasfollows:Amongthem,N=31,T=5,K=3,whichcancalculatethevalueoftheFstatistic14.
5105,greaterthanthecriticalvalueofF(30122),sotheoriginalhypothesisisrejected.
Thatis,theindividualfixedeffectmodelcanbeused.
Table4.
Comparisonofregressionmodels.
MixedregressionmodelIndividualfixedeffectmodelRandomeffectsmodelConstantterm10.
2135(0.
0000)6.
3840(0.
0081)9.
9916(0.
0000)Ln(R)2.
3416(0.
0000)0.
7491(0.
0002)1.
5545(0.
0000)LN(K)0.
5786(0.
0000)0.
4922(0.
0000)0.
4408(0.
0000)Ln(L)0.
0241(0.
4817)0.
4488(0.
0058)0.
3907(0.
0000)2R0.
97430.
99690.
9285SSE3.
84740.
46030.
7990So,inthechoiceoffixedeffectmodelandrandomeffectmodel,weusetheHousemantest.
Inthetest,iftheteststatisticPvalueislessthanthelevelofsignificance,werejectthenullhypothesis.
Themodelshouldbesetupasafixedeffectmodel.
Ifwedonotrejectthehypothesis,themodelshouldbesetupasfromarandomeffectsmodel.
Thetable5isHousemantestresults.
HstatisticvalueofPwas0,farlessthanthesignificantwater0.
5,sowerejectthenullhypothesis.
Themodelshouldbesetupasafixedeffectmodel.
Basedontheabove,thispaperdeterminesthefinalregressionmodelforindividualfixedregressionmodel.
Table5.
HausmanTest.
TestSummaryChi-Sq.
StatisticChi-Sq.
d.
f.
Prob.
Cross-sectionrandom62.
050330.
00004.
ConclusionandSuggestion4.
1ConclusionByselecting31provinces,city,autonomousregion5yearsofdatameasurementtest,wefinallydeterminetheindividualfixedeffectmodel.
ThismodelanalyzestherelationshipbetweentheInternetandtheeconomicgrowthbasedontheempiricalresultswhichareasfollows:(1)ThroughtheADFunitroottestandcointegrationtest,weconcludethatthereisacointegrationrelationshipbetweenthefourvariablesunderthesignificanceofthe5%,namelythereisalong-termstablerelationship.
Thepaneldatacanbereturned.
Intheprocess,FtestandHousemantestwereusedtodeterminetheindividualfixedeffectsmodelandpaneldatamodel.
(2)Theresultsfromthemodelregressionmodelareeffective.
ThedevelopmentoftheInternethasapositiveeffectoneconomicgrowth,andwiththeincreasinglevelsofInternettechnologies,itscontributiontoeconomicgrowthisgrowing.
(3)WeanalysisoftheInternet'simpactoneconomicgrowthfromthethreeindustriesandfinaldemandpointofview.
Thereispositiveinfluenceandnegativeinfluence.
ThecombinationoftheseeffectsisnotdifficulttospeculatethatonthewholeInternetstillhasapositiveimpactoneconomicgrowth,whichisconsistentwiththeempiricalresults.
4.
2SuggestionTheimprovementofthesysteminthenetworkcontactadvantagetoChina'seconomicgrowthcontributionratiowillbeincreasinglylarge.
ThecurrentsituationofthedevelopmentoftheInternetthroughout,someempiricalconclusionsbasedontheabove,weputforwardthefollowingsuggestions:First,toimprovetheInternetinfrastructure,optimizingtheoperationmode.
TheimportanttargetforfurtherimplementationoftheNationalBroadbandPlanistoimprovethelevelofbroadbandnetworks.
Weshouldincreasecapitalinvestmentandpolicysupporttofurtherimprovetheinfrastructureconstruction,strengthentheinternationalcooperationofcybersource,reducenetworkoperationcost,andmakefulluseofInternetresourcesandpositiveinfluenceoneconomicgrowth.
Second,weshouldacceleratetheconstructionofthemobileInternet,andexplorenewimpetustoitsdevelopment.
WeshouldspeeduptheconstructionofthemobileInternetandwirelessnetworks,bythenetwork,terminalandotherrestrictedaccessgroupsandregionsprovidethepossibilityoftheuseoftheInternettoimprovenetworkcoverage,andimprovethemobileclientindustrychain,thenewdirectionoftheminingdevelopmentoftheInternet.
Addnewvigortoit.
Third,weshouldpromotethe"Internetplus"plan,theformationofanewformofInterneteconomy,and"Internetpluscore"forinnovation.
Onlytheintroductionofadvancedtalentandmanagementmodel,tocreateagoodatmospherefordevelopment,wecanstimulatemoreinnovativethinking.
Weshouldacceleratethe"Internetplus"plan,promotethecombinationofInternetandtraditionaleconomy,soastoreleasethehugepotentialoftheInterneteconomy.
AcceleratethepaceofeconomicdevelopmentoftheInternet,soastobetterplaytheroleoftheInternettopromotesustainedeconomicgrowth.
Fourth,therationalallocationofresources,promotethebalanceddevelopmentoftheRegionalInternet.
WhentheeconomyleveldifferencesleadtodifferencesinthelevelofInternetdevelopment,thegovernmentisakeyregulatoryroleinthemiddle.
Weshouldaccordingtotheactualsituationofeachregion,formulatecorrespondingpoliciesandgiveappropriatesubsidies,strengthentheconstructionofInternetresources,andpromotetheharmoniousdevelopmentofsocietyandeconomy[1].
Fifth,acceleratetheimprovementofrelevantsystems,andcreateahealthyInternetenvironment.
InternetCoshouldactivelycooperatewiththecommonwaytoregulatethesafetyoftheindustry.
Thegovernmentshouldspeedupthenetworksecurityofthespeciallegislation,perfectinformationsecuritysystem.
Citizensshouldstrengthentheirownpersonalqualities,startfromtheirown,andtherationaluseoftheInternet,tocreateaharmoniousnetworkenvironmenttomakeacontribution[2].
Reference[1]ZhangTing,JinJianglei.
AnalysisoftheImpactofInternetDevelopmentonChina'sRegionalEconomy[J].
ProductivityResearch,2014,(9):62-64.
[2]JiangWei.
InternetPlusElectronicCommerce:toPromotetheCoordinatedDevelopmentofRegionalEconomyinChina[J].
WorldTelecommunications,2015,(5):52-56.