计划生育政策的储蓄与增长效应-理论与中国的经验分析

:*汪伟:本文研究了由计划生育所引发的中国人口急剧转变所带来的储蓄与增长效应在生育受到计划生育政策约束的制度环境下,本文通过一个简单的三期世代交替模型讨论了出生率人口增长率人口年龄结构等人口变量对储蓄和经济增长的影响本文的理论研究发现,出生率(人口增长率)下降会提高国民储蓄率与人均GDP的增长率,人口老龄化对储蓄率影响不确定本文运用中国19892007年的分省面板数据进行了计量分析,经验结论与模型的理论预测相一致,人口变量的变化能够解释人均GDP增长率的绝大部分变化,也能很好解释储蓄率的变化本文的估计结果对替代性的人口变量的选取估计方法的选择以及估计函数形式的设定都是非常稳健的,计划生育的储蓄与增长效应的确非常显著:计划生育人口转变储蓄经济增长*,,:200434,:[e-mail]!(:10CJL014),(:B802)211!2010!,2010!,,,,96%,,,,,(,2005)70,,!,,1970581209018(,2003),2000122,(,2009)1949,36∀,,19781825∀801988233∀,5060!,2008121∀,194920∀,,197866∀,,194935,8065,200773,63201010#,!!,11,,表1中国人均收入增长率储蓄率出生率死亡率人口增长率变化时间段196019701970198019801990199020002000200819892008人均GDP增长率27%63%82%87%93%87%国民储蓄率268%328%349%406%439%424%粗出生率(∀)335245212176126156粗死亡率(∀)1147167666666人口增长率(∀)2211741451106090:WorldBank,WorldBankOnlineDatabase,2009,,,:,,(Malthus,1798);,,(Boserup,1981);,(Srinivasan,1988;GalorandWeil,1999),20,,,,,,!(BloomandWilliamson,1998),,,(Modigliani,1970;HigginsandWilliamson,1997;ModiglianiandCao,2004),,,,(MasonandLee,2006)(),(ModiglianiandBrumberg,1954),,,,,Zhangetal.(2001),,,Bloometal.(2003)64::#∃%∃%∃%,WorldBankOnlineDatabase,Bloometal.(2007),,,,,,(BeckerandLewis,1973;BarroandBecker,1989;Beckeretal.,1990),,,,,,Zhangetal.(2001),,(2008),,,,,,,,!!,,2000,,10%,,(2009),,Lietal(2007),Zhangetal(2003),Zhangetal(2003),,,,,,,19892007()N1t,N2tN0t,,,t,,,N0t=65201010tN1tt(t1),,,N2t=t-1N1t-1,,wt,:c1t+tc0t+st=wt+(1-vt-1)Rtst-1t-1(1)Rt+1st=c2t+1(2),c1t,c2t+1,,,c0t,Rt+1tt+1:U=logc1t+vtlogc2t+1+1-!t∀logc0t(3),∀,!0,,:c0t=-!t∀c1t,c1t=#c,t[wt+(1-vt-1)Rtst-1t-1](4)c2t+1=vtRt+1c1t,st=#s,t[wt+(1-vt-1)Rtst-1t-1](5)#c,t=1(1+t+1-!t∀),#s,t=t(1+t+1-!t∀)#c,t,#s,tLietal(2007),,(Romer,1986),:Yt=AK∃t(k%tNt)1-∃,A0,0∃1,0&%&1Yt,Kt,kt,Nt,kt=kt=KtNt=[Yt(ANt)]1[∃+%(1-∃)],yt=YtNt=f(kt)=A(kt)∃+%(1-∃),,,:wt=(1-∃)Ak∃+%(1-∃)t,Rt=∃Ak(∃-1)(1-%)t(6),:Yt=Ntwt+RtKt,t:Nt-1t-1c2t=Nt-1t-1Rtst-1,Nt-1(1-t-1)Rtst-1,RtKt,NtNt-1=t-1,:Rtst-1=RtKtNt-1=Rtktt-1(7)(5),(6):st=#s,t[1+∃(1-vt-1)(1-∃)]wt(8),:Kt+1,t=Ntst=Nt#s,t[1+∃(1-vt-1)(1-∃)]wt,:kt+1t=A#s,t(1-∃vt-1)k∃+%(1-∃)t()(6)(8),t:StYt=NtstYt=Nt#s,t[1+∃(1-vt-1)(1-∃)](1-∃)f(kt)Ntf(kt)=(1-∃vt-1)#s,t=(1-∃vt-1)t1+t+1-!t∀(9),(StYt)t0,vt-1,,,vt-1,vt66::,,:(StYt)t0,(StYt)t-10,:,,,,;,,,,,,,yt=YtNt=f(kt)=A(kt)∃+%(1-∃),t:gt={ytNtyt-1Nt-1}{[Nt(1+t)+Nt-1t-1][Nt-1(1+t-1)+Nt-2t-2]}-1(10)tnt=[Nt(1+t)+Nt-1t-1][Nt-1(1+t-1)+Nt-2t-2]-1,:gt=(ktkt-1)∃+%(1-∃)(NtNt-1)(11+nt)-1(11),:gt=[A(1-∃vt-2)#s,t-1(Nt-1Nt)]∃+%(1-∃)[Yt-1(ANt-1)](1-∃)(%-1)(NtNt-1)(11+nt)-1(12):log(1+gt)=C+[∃+%(1-∃)]log[(1-∃vt-2)Ts,t-1]-(1-∃)(1-%)log[Yt-1(Nt-1(1+t-1)+Nt-2t-2)]-[∃+%(1-∃)]log(1+nt)+(1-∃)(1-%)log{Nt[(Nt(1+t)+Nt-1t-1)]}(13)C=logA,(1-∃vt-2)Ts,t-1(),ntt,Yt-1(Nt-1(1+t-1)+Nt-2t-2),Nt[(Nt(1+t)+Nt-1t-1)t,,()(9)(13),,:tsrit=0+1fertilityit+2opopit+&∋Xit+ui+!it(14)log(1+pgdpgr)it=∋0+∋1log(1+ngrow)it+∋2log(rpgdp)it-1+∋3logtirit-1+∋4logwpratioit+(∋Zit+ui+!it(15),i,t,ui,!it(14),tsr,fertility,opop,X,,10,2(15),log(1+pgdpgr),pgdpgrGDP,log(1+ngrow)log(rpgdp)logtirlogwpratio,ngrowrpgdptirwpratioGDP,Z,∋10,∋20,∋30,∋401989200729(),GDPGDP∃%,20052008∃67201010%,GDP1989,(1-)∃%∃%∃1990%,,1=(-),,,90;,2=(-),,90,19982007表2各变量描述性统计量变量表示观测数均值标准差最小值最大值国民总储蓄率(%)tsr551427798117787011国内总投资率(%)tir551456099023927753人均GDP增长率(%)pgdpgr551978416-370381人均对数实际GDP(元)rpgdp5518150756621049出生率(∀)feritility55114474844702651人口自然增长率(∀)ngrow551819463-3102075成人占总人口比重(%)wpratio551690341559878018成人与抚养人口比wpr551229048149405老人占总人口比重(%)opopratio5517132012961637少数民族人口比重(%)minpopratio337155116970046224人口迁移率1(∀)Migration1551223265-8791421人口迁移率2(%)Migration2290335817-9684306平均受教育年限huc5517241154541091政府支出GDP(%)fisrat55113555764685204工业GDP(%)indrat551379283911206215城市化率(%)urbliz5513282164612158909城乡收入比urrat551268066124476实际利率(%)rint551-050479-170611142,()GDP,,,,,,,,,(),68::,:,31:,,,,1∀,061%;1%,095%,,123GDP4GDP(Modigliani,1970;HigginsandWilliamson,1997;ModiglianiandCao,2004),GDP,2GDP,,3GDPGDP,GDP,,,,:,;,;GDP,;GDPGDP,;,;,4,,,GDPGDPGDP,,,69201010,5,,67,,45,,,,表3储蓄方程面板估计结果I(固定效应)自变量估计1估计2估计3(IV)估计4估计5(IV)估计6估计7(IV)fertility-061-048-049-030-032-031-037(-688)***(-554)***(-522)***(-371)***(-368)***(-269)***(-290)***opopratio095077067151142085085(405)***(339)***(286)***(546)***(492)***(236)**(229)**pgdpgr035051026034033040(655)***(605)***(508)***(358)***(496)***(338)***fisrat-027-026-025-024(-446)***(-392)***(-311)***(-283)***indrat054054049048(106)***(978)***(894)***(810)***urbliz-004-004-013-011(-120)(-106)(-217)**(-175)*urrat-034-070005007(-047)(-088)(006)(008)rint-003-002-018-018(-083)(-030)(-189)*(-177)*inmigration-010-012-018-021(-113)(-122)(-192)*(-209)**是否加入时间哑变量否否否否否是是D!Wu!Hausman检验的p值005600040000R2032037034052049055053:stata100,t,357GDP,GDPGDPGDP,******1%5%10%,1989199819992007,419891998(),;19992007(),,!,70::,,,,,,,(ModiglianiCao,2004),,,,,,,表4储蓄方程面板估计结果((分时段回归,固定效应)自变量198920071989199819992007估计1估计2估计3估计4估计5估计6估计7估计8估计9fertility-050-030-043-039-030-033-013-006-006(-553)***(-352)***(-342)***(-339)***(-235)**(-225)**(-068)(-028)(-025)opopratio062116078140238131053057016(402)***(402)***(211)**(297)***(430)***(197)**(178)*(180)*(038)pgdpgr(-1)04303