• 城镇人均收入与人均通讯消费分析 不要轻易放弃。学习成长的路上,我们长路漫漫,只因学无止境。


    城镇人均支出与人均通信生产剖析

    [择要]本文旨在与对1992——2004年我国人均支出对人均通信生产的影响。起首,咱们综合了几种关于支出和生产的次要现实概念;进而咱们树立了现实模子。而后,搜集了相干的数据,哄骗EVIEWS软件对计量模子举行了参数估量和检讨,并加以批改。最后,咱们对所得的剖析了局作了经济意思的剖析。

    [关键词]? 城镇家庭人均支出(人均现实支出) 人均通信生产

    一? 提出问题

    跟着经济的生长,群众糊口程度的进步,人际交往的需要,对信息的需要也成逐步回升的趋向。九十年代以来,我国通信事业有了较大的生长,从“中国电信”一家独霸全国,生长到今天的“铁通”“联通”“网通”等瓜分全国。改革开放以来的经济在从企图向市场转型的进程中,群众的生产程度、布局都产生了很大转变。由于出生避世,引进外洋进步前辈的技巧、借鉴本国进步前辈的 运营和管理经验,增进我国电信业的全方位生长壮大。同时也拓宽融资渠道,有利于引进外资,也有利于改良资金布局。跟着市场经济的生长,以及九十年代前期我国对工资布局作了很大的调解,使得我国人均支出不论是从程度仍是布局上来讲都有了很大的转变。从而咱们发觉以上的转变足以以影响通信生产。针对这类现象,咱们搜集了1992——2004年间城镇家庭人均支出,人均通信生产。

    ?二.经济现实陈述

    ?西方经济学中关于生产与支出决议关连的有关现实假说

    凯恩斯相对支出假说

    ?对??

    有(1),即会随支出的而增进 ,但其增量小于支出增量。

    ?(2),即

    由? 可知

    有,即支出的均匀生产偏向递加。

    相对支出假说下的生产函数通常采用线性方式,

    此时,函数合乎假说和

    三 样本数据搜集

    本模子使用光阴序列数据,Yt=α +βXt+Ut,Y为人均通信生产,单元元,Xt为城镇家庭人均支出,单元元。

    数据来源于国度统计局网站()。在经由大量剖析比拟后咱们采用了所取样本数据见

    表1,

    ?Y?X

    1992?10.62000?2031.530

    1993?28.27000?2583.160

    1994?62.85000?3502.310

    1995?87.97000?4279.020

    1996?102.9500?4844.780

    1997?121.5400?5188.540

    1998?142.4000?5449.500

    1999?173.7000?5864.700

    2000?232.8000?6295.910

    2001?281.5000?6868.900

    2002?358.8000?8177.400

    2003?424.0100?9061.220

    2004?454.6000?10128.50

    四.安稳性的检

    (一)

    表二? X:

    ADF demo Statistic? 1.222472???? 1%?? Critical Value*?-4.1366

    ????? 5%?? Critical Value?-3.1483

    ????? 10% Critical Value?-2.7180

    *MacKinnon critical values for rejection of hypothesis of a unit root.

    ????

    ????

    Augmented Dickey-Fuller demo Equation

    Dependent Variable: D(SER01)

    Method: Least Squares

    Date: 06/03/05?? Time: 19:40

    Sample(adjusted): 1993 2004

    Included observations: 12 after adjusting endpoints

    Variable?Coefficient?Std. Error?t-Statistic?Prob.?

    SER01(-1)?0.054515?0.044594?1.222472?0.2496

    C?383.3354?254.7627?1.504677?0.1633

    R-squared?0.130014???? Mean dependent var?674.7475

    Adjusted R-squared?0.043015???? S.D. dependent var?318.2914

    S.E. of regression?311.3704???? Akaike info criterion?14.47086

    Sum squared resid?969515.2???? Schwarz criterion?14.55167

    Log likelihood?-84.82513???? F-statistic?1.494438

    Durbin-Watson stat?1.117371???? Prob(F-statistic)?0.249560

    /1.222472/

    ????? /-3.1483/

    ????? /-2.7180/

    谢绝原假定,不经由进程检讨,证实是不安稳的。

    表三 Y:

    ADF demo Statistic? 2.333367???? 1%?? Critical Value*?-4.1366

    ????? 5%?? Critical Value?-3.1483

    ????? 10% Critical Value?-2.7180

    *MacKinnon critical values for rejection of hypothesis of a unit root.

    ????

    ????

    Augmented Dickey-Fuller demo Equation

    Dependent Variable: D(SER02)

    Method: Least Squares

    Date: 06/03/05?? Time: 19:35

    Sample(adjusted): 1993 2004

    Included observations: 12 after adjusting endpoints

    Variable?Coefficient?Std. Error?t-Statistic?Prob.?

    SER02(-1)?0.094160?0.040354?2.333367?0.0418

    C?21.08984?8.483771?2.485904?0.0322

    R-squared?0.352525???? Mean dependent var?36.99833

    Adjusted R-squared?0.287777???? S.D. dependent var?20.72437

    S.E. of regression?17.48998???? Akaike info criterion?8.712145

    Sum squared resid?3058.993???? Schwarz criterion?8.792963

    Log likelihood?-50.27287???? F-statistic?5.444602

    Durbin-Watson stat?1.344857???? Prob(F-statistic)?0.041810

    /2.333367/

    ????????? / -3.1483 /

    ??????? /-2.7180/

    谢绝原假定,不经由进程检讨,证实是不安稳的。

    (二)举行协整性检讨

    天生ET=X-(α+ρ y)

    表四

    1992?2178.858

    1993?2719.898

    1994?3618.3

    1995?4379.938

    1996?4936.71

    1997?5269.316

    1998?5517.76

    1999?5914.18

    2000?6309.93

    2001?6853.7

    2002?8115.82

    2003?8960.514

    2004?10009.44

    检讨ET的安稳性

    DW=0.462912

    在明显性程度为0.05和0.1下经由进程程度型检讨。

    也就是说,咱们要在如下的检讨顶用0.05和0.1的明显性程度对咱们的数据举行估量和检讨。

    五.参数估量与检讨

    (一)将样本数据导入Eviews软件举行OLS估量,失掉输出了局如下:

    表五

    Dependent Variable: SER02

    Method: Least Squares

    Date: 06/01/05?? Time: 20:41

    Sample: 1992 2004

    Included observations: 13

    Variable?Coefficient?Std. Error?t-Statistic?Prob.?

    SER01?0.060313?0.003423?17.62135?0.0000

    C?-153.6739?21.10424?-7.281659?0.0000

    R-squared?0.965787???? Mean dependent var?190.9238

    Adjusted R-squared?0.962676???? S.D. dependent var?148.0898

    S.E. of regression?28.60996???? Akaike info criterion?9.686025

    Sum squared resid?9003.829???? Schwarz criterion?9.772940

    Log likelihood?-60.95916???? F-statistic?310.5121

    Durbin-Watson stat?0.462912???? Prob(F-statistic)?0.000000

    (二)模子的检讨

    1.经济意思的检讨

    经由上面的剖析咱们在现实上已晓得,人均支出X与城镇住民人均通信生产Y的增进是正的线形关连,这与现实中X与Y同向转变是相符的。当人们的支出不竭添加的同时,食物所占比例随之降低,其余生产所占比例有所回升,这是合乎咱们家庭生产的习气的。

    2.统计揣度检讨

    从估量的了局能够看出,可决系数为0.965787,模子拟合情形比拟抱负,系数明显性检讨T统计量为:17.62135。在给定明显性程度为0.05的情形下,查T散布表在自由度为N-2=11下的临界值为2.201,由于17.62135大于2.201,以是经由进程T检讨谢绝原假定。表白人均支出X对城镇住民人均通信生产有明显影响。

    3.计量经济检讨

    (1)由于咱们树立的模子惟独一个说明变量,以是不具有多重共线性。

    (2)异方差?? 图一

    由图可知,必然具有异方差。

    由于是光阴序类数据,咱们采用ARCH检讨

    表六

    ARCH demo:

    F-statistic?1.205990???? Probability?0.385049

    Obs*R-squared?3.761678???? Probability?0.288375

    ????

    demo Equation:

    Dependent Variable: RESID^2

    Method: Least Squares

    Date: 06/01/05?? Time: 20:52

    Sample(adjusted): 1995 2004

    Included observations: 10 after adjusting endpoints

    Variable?Coefficient?Std. Error?t-Statistic?Prob.?

    C?1074.229?464.9595?2.310370?0.0602

    RESID^2(-1)?-0.076893?0.516201?-0.148959?0.8865

    RESID^2(-2)?0.133688?0.507889?0.263222?0.8012

    RESID^2(-3)?-0.591737?0.437524?-1.352468?0.2250

    R-squared?0.376168???? Mean dependent var?654.7810

    Adjusted R-squared?0.064252???? S.D. dependent var?505.4297

    S.E. of regression?488.9228???? Akaike info criterion?15.51146

    Sum squared resid?1434273.???? Schwarz criterion?15.63249

    Log likelihood?-73.55730???? F-statistic?1.205990

    Durbin-Watson stat?0.861696???? Prob(F-statistic)?0.385049

    从输出的辅助回归函数中得obs*-squared为3.761678,P=0.288375,

    ??????????????????? ∵3.761678<0.28837

    ??????????????????? ∴以是经由进程检讨谢绝原假定

    表白模子中具有明显的异方差现象。即,跟着光阴的推移,多种因素对其有着影响。如,同需用度单元价钱的转变,通信是产竞争的激烈程度,手机价钱的降低,国度政策的疏导等。

    (3)自相干检讨

    咱们的模子惟独一个说明变量,把其余的影响因素都放在了随机误差项U里。因此必定具有自相干。

    哄骗图示法,由Eviews软件失掉如下了局:

    图二

    由图能够初步判别,此模子有自相干。再哄骗D-W法检讨由DW=0.462912,查DW表,n=13,k’=1,在α==0.05时,查得两个临界值别离为:上限DL=1.010,上限DU=1.331,由于DW统计量为0.462912

    六.? 计量经济参数批改

    按照上述检讨能够失掉,咱们树立的模子具有异方差与自相干,上面举行批改。

    起首是对异方差的批改。

    A.哄骗WLS估量法失掉如下输出了局:

    表七

    Dependent Variable: SER02

    Method: Least Squares

    Date: 06/02/05?? Time: 15:07

    Sample: 1992 2004

    Included observations: 13

    Weighting series: W

    Variable?Coefficient?Std. Error?t-Statistic?Prob.?

    SER01?0.059010?0.000206?285.7880?0.0000

    C?-143.0366?2.026115?-70.59648?0.0000

    Weighted Statistics????

    R-squared?0.999993???? Mean dependent var?350.3346

    Adjusted R-squared?0.999993???? S.D. dependent var?1083.497

    S.E. of regression?2.962703???? Akaike info criterion?5.150719

    Sum squared resid?96.55371???? Schwarz criterion?5.237635

    Log likelihood?-31.47968???? F-statistic?81674.76

    Durbin-Watson stat?1.135523???? Prob(F-statistic)?0.000000

    Unweighted Statistics????

    R-squared?0.964832???? Mean dependent var?190.9238

    Adjusted R-squared?0.961635???? S.D. dependent var?148.0898

    S.E. of regression?29.00618???? Sum squared resid?9254.943

    Durbin-Watson stat?0.435347???

    剖析:

    ??? R=0.999993?? T=285.7880〉2.201

    B.再用对数变换法,将变量X,Y互换成LNX,LNY。用OLS法对LY,LX回归,失掉了局如下:

    表八

    Dependent Variable: LY

    Method: Least Squares

    Date: 06/02/05?? Time: 12:45

    Sample: 1992 2004

    Included observations: 13

    Variable?Coefficient?Std. Error?t-Statistic?Prob.?

    LX?2.299547?0.099757?23.05151?0.0000

    C?-14.83014?0.854821?-17.34881?0.0000

    R-squared?0.979719???? Mean dependent var?4.847307

    Adjusted R-squared?0.977875???? S.D. dependent var?1.093940

    S.E. of regression?0.162718???? Akaike info criterion?-0.652959

    Sum squared resid?0.291248???? Schwarz criterion?-0.566044

    Log likelihood?6.244234???? F-statistic?531.3721

    Durbin-Watson stat?1.058521???? Prob(F-statistic)?0.000000

    剖析:

    ???? R=0.979719? T=23.05151〉2.201

    比拟两种方式,能够发觉X,Y在非对数线性回归下拟和效果更好,可决系数更大,且T统计量也较好。咱们将模子的表达式基本上能够确定为:Yt=α+βXt+Ut。

    (2)其次是对自相干举行批改。

    哄骗对数线性回归批改并举行迭代,得出如下了局:

    A.表九

    Dependent Variable: LY

    Method: Least Squares

    Date: 06/05/05?? Time: 21:39

    Sample(adjusted): 1993 2004

    Included observations: 12 after adjusting endpoints

    Convergence achieved after 4 iterations

    Variable?Coefficient?Std. Error?t-Statistic?Prob.?

    LX?1.843468?0.288465?6.390602?0.0001

    C?-10.81583?2.558273?-4.227785?0.0022

    AR(1)?0.425312?0.216569?1.963862?0.0811

    R-squared?0.989343???? Mean dependent var?5.054355

    Adjusted R-squared?0.986975???? S.D. dependent var?0.835189

    S.E. of regression?0.095318???? Akaike info criterion?-1.650875

    Sum squared resid?0.081770???? Schwarz criterion?-1.529649

    Log likelihood?12.90525???? F-statistic?417.7612

    Durbin-Watson stat?0.908943???? Prob(F-statistic)?0.000000

    Inverted AR Roots??????? .43

    DW=0.908943自相干不失掉批改,以是模子不成能是对数模子,进一步能够确定模子方式为Yt=α+βXt+Ut。

    B.? ρ=1-DW/2?? DW=0.435347(由表七批改后的数据可知)

    由表二可得ρ=0.7823

    表十

    Dependent Variable: DY

    Method: Least Squares

    Date: 06/02/05?? Time: 15:24

    Sample(adjusted): 1993 2004

    Included observations: 12 after adjusting endpoints

    Variable?Coefficient?Std. Error?t-Statistic?Prob.?

    DX?0.063330?0.007917?7.999430?0.0000

    C?-44.73836?15.90432?-2.812969?0.0184

    R-squared?0.864848???? Mean dependent var?75.85703

    Adjusted R-squared?0.851333???? S.D. dependent var?45.52532

    S.E. of regression?17.55336???? Akaike info criterion?8.719380

    Sum squared resid?3081.204???? Schwarz criterion?8.800197

    Log likelihood?-50.31628???? F-statistic?63.99087

    Durbin-Watson stat?1.182893???? Prob(F-statistic)?0.000012

    DW=1.182893,在0.05的明显性程度下,不克不及谢绝原假定的区间内(DL=1.010,DU=1.331)以是不克不及说批改了自相干性。

    C.间接使用跌代法

    表十一

    Dependent Variable: SER02

    Method: Least Squares

    Date: 06/02/05?? Time: 14:22

    Sample(adjusted): 1993 2004

    Included observations: 12 after adjusting endpoints

    Convergence achieved after 4 iterations

    Variable?Coefficient?Std. Error?t-Statistic?Prob.?

    SER01?0.064557?0.005668?11.38928?0.0000

    C?-193.6246?43.82927?-4.417700?0.0017

    AR(1)?0.624654?0.181152?3.448243?0.0073

    R-squared?0.987356???? Mean dependent var?205.9492

    Adjusted R-squared?0.984546???? S.D. dependent var?143.9535

    S.E. of regression?17.89543???? Akaike info criterion?8.819286

    Sum squared resid?2882.219???? Schwarz criterion?8.940513

    Log likelihood?-49.91572???? F-statistic?351.3955

    Durbin-Watson stat?1.340015???? Prob(F-statistic)?0.000000

    Inverted AR Roots??????? .62

    进一步批改自相干,DW=1.240015〉1.182893在0.05的明显性程度下,不克不及谢绝原假定的区间内(DL=1.010,DU=1.331)以是也批改了自相干性。

    七?? .总结

    经由进程以上剖析,咱们失掉如下方程:

    Y= -143.0366+0.059010X

    ????? (2.026115)???? (0.000206)

    ?? T= -70.59648???????? 285.7880

    R-squared=0.999993?? F=81674.76??? DF=13

    该模子的经济意思可说明为:人均支出每增进1个单元,则财政支出均匀增进0.059010

    愧疚的是咱们的模子不是十分的抱负,线性拟和不是很好,这从批改后模子的散点散布图能够看出。

    图三

    从2000年后生长速度有了很大的转变。

    上图中现实的值具有颠簸,咱们只是近似的将其拟和为线性,其中1999年涌现了一个个转折点,这是由于我国在1999年到2000年要面对出生避世的了局,这招致了对斜率参数的明显影响,以及对随机误差的影响。这在很大程度上说明了为何咱们的模子最后涌现了异方差和自相干。

    布景:1.1999——2000,我国各个省市的通信网举行了扩容。加大了用户群。

    ?????? 2.1999——2000,我国通信行业面对出生避世的机遇和挑战。通信改革势在必行。良多处所通信公司对通信的单元价钱作了进一步的调解。同时还针对特殊的生产群体作了特殊的企图,完满了办事。

    ?????? 3.1999——2000,顺手机市场的生长,对通信有了很好的增进作用。

    ?????? 4.网络的生长,使人们逐步的把手机与网络相联系,经由进程手机上彀生产。

    ?如今对咱们的数据举行进一步分段处置。

    ?1992——1999

    ?表十二

    Dependent Variable: SER02

    Method: Least Squares

    Date: 06/02/05?? Time: 16:26

    Sample: 1992 1999

    Included observations: 8

    Variable?Coefficient?Std. Error?t-Statistic?Prob.?

    SER01?0.039526?0.002746?14.39337?0.0000

    C?-75.43015?12.12109?-6.223049?0.0008

    R-squared?0.971853???? Mean dependent var?91.28750

    Adjusted R-squared?0.967162???? S.D. dependent var?55.74647

    S.E. of regression?10.10193???? Akaike info criterion?7.675648

    Sum squared resid?612.2940???? Schwarz criterion?7.695508

    Log likelihood?-28.70259???? F-statistic?207.1690

    Durbin-Watson stat?0.757574???? Prob(F-statistic)?0.000007

    ?R=0.971853? T=14.39337? 比不分段前有了很好的改良。

    ?表十三

    ARCH demo:

    F-statistic?0.131176???? Probability?0.732020

    Obs*R-squared?0.178951???? Probability?0.672276

    ????

    demo Equation:

    Dependent Variable: RESID^2

    Method: Least Squares

    Date: 06/02/05?? Time: 16:35

    Sample(adjusted): 1993 1999

    Included observations: 7 after adjusting endpoints

    Variable?Coefficient?Std. Error?t-Statistic?Prob.?

    C?万博亚洲,新万博娱乐manbetx,万博娱乐国际96.03956?59.12377?1.624382?0.1652

    RESID^2(-1)?-0.298109?0.823092?-0.362182?0.7320

    R-squared?0.025564???? Mean dependent var?82.74366

    Adjusted R-squared?-0.169323???? S.D. dependent var?113.3951

    S.E. of regression?122.6200???? Akaike info criterion?12.69101

    Sum squared resid?75178.38???? Schwarz criterion?12.67556

    Log likelihood?-42.41855???? F-statistic?0.131176

    Durbin-Watson stat?1.493123???? Prob(F-statistic)?0.732020

    ?能够看出不异方差。

    ?然而不克不及够判别有不自相干。然而比起对表五的ARCH检讨要好多。

    2000——2004

    ?表十三

    Dependent Variable: SER02

    Method: Least Squares

    Date: 06/02/05?? Time: 16:29

    Sample: 2000 2004

    Included observations: 5

    Variable?Coefficient?Std. Error?t-Statistic?Prob.?

    SER01?0.059070?0.004796?12.31539?0.0012

    C?-128.4983?39.45793?-3.256591?0.0473

    R-squared?0万博亚洲,新万博娱乐manbetx,万博娱乐国际.980604???? Mean dependent var?350.3420

    Adjusted R-squared?0.974138???? S.D. dependent var?93.43927

    S.E. of regression?15.02651???? Akaike info criterion?8.546684

    Sum squared resid?677.3881???? Schwarz criterion?8.390459

    Log likelihood?-19.36671???? F-statistic?151.6687

    Durbin-Watson stat?2.130183???? Prob(F-statistic)?0.001153

    ?R=0.980604? T=12.31539??? DW=2.130183? 都很好,且不自相干

    ?表十四

    ARCH demo:

    F-statistic?0.116218???? Probability?0.765655

    Obs*R-squared?0.219671???? Probability?0.639291

    ????

    demo Equation:

    Dependent Variable: RESID^2

    Method: Least Squares

    Date: 06/02/05?? Time: 16:44

    Sample(adjusted): 2001 2004

    Included observations: 4 after adjusting endpoints

    Variable?Coefficient?Std. Error?t-Statistic?Prob.?

    C?112.5511?120.5321?0.933785?0.4490

    RESID^2(-1)?0.257122?0.754229?0.340907?0.7657

    R-squared?0.054918???? Mean dependent var?141.2674

    Adjusted R-squared?-0.417624???? S.D. dependent var?144.8157

    S.E. of regression?172.4234???? Akaike info criterion?13.44464

    Sum squared resid?59459.63???? Schwarz criterion?13.13778

    Log likelihood?-24.88927???? F-statistic?0.116218

    Durbin-Watson stat?1.654957???? Prob(F-statistic)?0.765655

    ?可见不异方差。

    ?剖析:与92—99年比拟,00-04年我国生产品经常承接往年国民经济涌现了严重转折后带来的回升惯性,坚持稳中有升,偏旺的优秀态势。旺盛的生产需要对我国抵抗世界经济暖流侵袭,国民经济坚持快捷稳定生长起到重要作用。这些次要由于亚洲金融危机产生后,亚洲各国遍及涌现了减薪或工资冻结。大多数住民的表面支出和现实支出都有所降低。然而我国应答亚洲金融危机期间却采用了大幅度进步城镇低支出者与公职人员支出的十分之举。99-00年间,将国有企业下岗职工基本糊口费,赋闲保险费和城镇住民最低糊口保障程度进步了30%,离退休人员养老金程度进步了30%,机关事业单元职工工资程度进步了30%,并要求各地一次性补发拖欠的国有企业离退休人员兼顾名目内的养老金等一系列启动生产需要的政策,也就合乎了为何我国在经济危机的影响和打击下,99年对92年生产程度总体有了很大进步。这都是由于国度给了相干政策,支出程度有所进步,从政治角度剖析,支出与生产有着明显的影响。从这些转变来看,这很合乎人类汗青生长的规律和生产行为。在解决了饥寒之后开始斟酌其余生产,如精神上的和情趣上的。99年以后,咱们的糊口节拍放慢,对糊口的尺度的要求有所进步。种种迹象表白涌现这类情形是齐全合乎现实经济意思的。

    因此能够以为:在中短期内,当通信政策稳定时、内部经济环境不严重转变的情形下,城镇人均通信生产与人均支出的确具有线性相干关连,能够用最后的模子Yt=α+βXt+Ut举行拟和。在长期中,由于具有不成预知的突发扰动以及经济变量布局性的转变,需要举行批改,也许涌现差别的模子方式。

    ?参考文献

    ?1 《西方经济学》群众大学出版社

    ?2?? ()

    ?3《经济计量学》,上海财经大学出版社

    ?4《现代西方经济学说》,中国经济出版社

    上一篇:生活因自信而精彩字

    下一篇:没有了