论文首页哲学论文经济论文法学论文教育论文文学论文历史论文理学论文工学论文医学论文管理论文艺术论文 |
人力资本投入与GDP
组长 马俊
引言
社会经济的发展是与多种多样的因素相关的,传统观念认为影响生产的决定性因素就是资本的投入与劳动者的数量。这种观点在资本主义早期得到了事实的佐证,但随着时代的进步,这种理论却无法解决许多新问题。近年来,人力资本理论由定性到定量的发展得到了越来越多经济学家的关注,
研究目的
作为人力资源管理专业的学生,我们希望能利用本专业知识与计量经济学相结合,通过对柯克道格拉斯函数的检验,分析人力资本投入在经济发展中的作用,据以观察我国在人力资本投资上的政策方向,预测经济发展方向,进行政策评价。
分析
1.柯克道格拉斯模型的检验
柯克道格拉斯函数的核心是认为,生产的决定性因素就是资本投入量与劳动者的数量,据此我们统计了1990至2002年全国固定资产投资额和就业人数:
obs GDP(亿元) 固定资产投资额(亿元) 就业人数(亿人)
1990 18598.40 4449.290 6.390900
1991 21662.50 5508.800 6.479900
1992 26651.90 7854.980 6.555400
1993 34560.50 12457.88 6.637300
1994 46670.00 16370.33 6.719900
1995 57494.90 20019.26 6.794700
1996 66850.50 22974.03 6.885000
1997 73142.70 24941.11 6.960000
1998 76967.20 28406.17 6.995700
1999 80579.40 29854.71 7.058600
2000 88228.10 32917.73 7.115000
2001 94346.40 37213.49 7.302500
2002 102398.0 43499.91 7.374000
据此,我们用eviews软件根据柯克道格拉斯函数对影响GDP的因素进行分析,为避免多重公线性,用分别用GDP,固定资产投资额和就业人数的对数值ly,lx ,lz代替y x z
Dependent Variable: LY 本文来自中国科教评价网
Method: Least Squares
Date: 05/20/04 Time: 13:26
Sample: 1990 2002
Included observations: 13
Variable Coefficient Std. Error t-Statistic Prob.
C 1.806738 1.837126 0.983459 0.3486
LX 0.730456 0.081339 8.980388 0.0000
LZ 0.994193 1.343659 0.739915 0.4764
R-squared 0.991580 Mean dependent var 10.87716
Adjusted R-squared 0.989896 S.D. dependent var 0.582647
S.E. of regression 0.058567 Akaike info criterion -2.638121
Sum squared resid 0.034301 Schwarz criterion -2.507748
Log likelihood 20.14779 F-statistic 588.8253
Durbin-Watson stat 0.906606 Prob(F-statistic) 0.000000
得表达式
LY = 1.80673763 + 0.7304560978*LX + 0.9941926462*LZ
模型总体拟合度较好,F统计值显著。但我们知道,固定资本投入到产出有一个过程,具有滞后性,因此用阿尔蒙法进行滞后,并试图消除其造成的多重共线性。经分析,以滞后三阶为最优:
Dependent Variable: LY
Method: Least Squares
Date: 05/20/04 Time: 13:48
Sample(adjusted): 1993 2002
Included observations: 10 after adjusting endpoints
Variable Coefficient Std. Error t-Statistic Prob.
C 2.592632 1.275232 2.033067 0.0977
LZ 1.478120 1.370459 1.078558 0.3300
PDL01 0.404223 0.049484 8.168727 0.0004
PDL02 0.200412 0.161523 1.240768 0.2697
PDL03 -0.241153 0.103723 -2.324971 0.0676
R-squared 0.998926 Mean dependent var 11.13983
Dependent Variable: LY
Method: Least Squares
Date: 05/20/04 Time: 14:17
Sample(adjusted): 1993 2002
Included observations: 10 after adjusting endpoints
Variable Coefficient Std. Error t-Statistic Prob. (科教论文网 Lw.nsEAc.com编辑整理)
C 1.720045 4.163169 0.413158 0.7196
LZ 1.591358 2.460885 0.646661 0.5842
PDL01 0.367198 0.185682 1.977560 0.1866
PDL02 0.060922 0.275179 0.221389 0.8453
PDL03 -0.143292 0.195252 -0.733884 0.5394
PDL04 0.029352 0.194188 0.151152 0.8937
PDL05 0.067112 0.329218 0.203853 0.8573
PDL06 -0.062945 0.186173 -0.338100 0.7675
R-squared 0.999265 Mean dependent var 11.13983
Adjusted R-squared 0.996693 S.D. dependent var 0.336436
S.E. of regression 0.019347 Akaike info criterion -5.061963
Sum squared resid 0.000749 Schwarz criterion -4.819895
Log likelihood 33.30981 F-statistic 388.4971
Durbin-Watson stat 2.879180 Prob(F-statistic) 0.002570
Lag Distribution of LX i Coefficient Std. Error T-Statistic
. * | 0 0.16298 0.42717 0.38155
. *| 1 0.36720 0.18568 1.97756
. * | 2 0.28483 0.16230 1.75498
* . | 3 -0.08413 0.25202 -0.33382
Sum of Lags 0.73088 0.42473 1.72082
Lag Distribution of LS i Coefficient Std. Error T-Statistic
* . | 0 -0.10070 0.46525 -0.21645 内容来自www.nseac.com
. * | 1 0.02935 0.19419 0.15115
. *| 2 0.03352 0.24247 0.13824
* . | 3 -0.08820 0.12359 -0.71371
Sum of Lags -0.12604 0.40657 -0.31000
由此我们得到该模型的表达式:
LY = 1.720044891 + 1.591358377*LZ + 0.1629841247*LX + 0.367198033*LX(-1) + 0.2848271534*LX(-2) - 0.08412851405*LX(-3) - 0.1007048014*LS + 0.029352059*LS(-1) + 0.03351895422*LS(-2) - 0.08820411579*LS(-3)
模型检验
ARCH Test:
F-statistic 0.737454 Probability 0.595832
Obs*R-squared 2.971116 Probability 0.396100
Test Equation:
Dependent Variable: RESID^2
Method: Least Squares
Date: 05/28/04 Time: 21:28
Sample(adjusted): 1996 2002
Included observations: 7 after adjusting endpoints
Variable Coefficient Std. Error t-Statistic Prob.
C 0.000146 8.85E-05 1.651497 0.1972
RESID^2(-1) 0.004550 0.561467 0.008104 0.9940
RESID^2(-2) -0.527388 0.487722 -1.081330 0.3587
RESID^2(-3) -0.346473 0.631646 -0.548523 0.6215
R-squared 0.424445 Mean dependent var 6.99E-05
Adjusted R-squared -0.151110 S.D. dependent var 9.54E-05
S.E. of regression 0.000102 Akaike info criterion -15.24072
Sum squared resid 3.14E-08 Schwarz criterion -15.27163
Log likelihood 57.34251 F-statistic 0.737454