做的是2000年到2014年的聚堆收敛问题,相关性检验显示是空间误差模型,同时面板数据具有异方差,开始使用MLE,误差模型结果很差,不知道什么原因,后来文献中说异方差问题是要使用GMM法,但怎样用stata做空间面板模型的广义矩估计,主要我的模型是误差模型,GMM貌似只能做空间自回归问题呀????这是广义矩估计做出的结果
* Spatial Panel Autoregressive Generalized Method of Moments (SPGMM)
==============================================================================
lvgdp = lngdp
------------------------------------------------------------------------------
Sample Size = 372 | Cross Sections Number = 31
Wald Test = 57.1026 | P-Value > Chi2(1) = 0.0000
F-Test = 57.1026 | P-Value > F(1 , 340) = 0.0000
(Buse 1973) R2 = 0.2608 | Raw Moments R2 = 0.9669
(Buse 1973) R2 Adj = 0.1933 | Raw Moments R2 Adj = 0.9639
Root MSE (Sigma) = 0.0171 | Log Likelihood Function = 1001.7905
------------------------------------------------------------------------------
- R2h= 0.1614 R2h Adj= 0.0850 F-Test = 71.22 P-Value > F(1 , 340) 0.0000
- R2v= 0.1378 R2v Adj= 0.0592 F-Test = 59.14 P-Value > F(1 , 340) 0.0000
------------------------------------------------------------------------------
lvgdp | Coef. Std. Err. t P>|t| [95% Conf. Interval]
-------------+----------------------------------------------------------------
lngdp | -.0115129 .0015236 -7.56 0.000 -.0145097 -.0085162
_cons | .1994339 .0147444 13.53 0.000 .1704322 .2284356
------------------------------------------------------------------------------这是空间误差模型结果,估计情况相当差,也不知道什么原因。
MLE Spatial Error Panel Normal Model (SEM)
==============================================================================
lvgdp = lngdp
------------------------------------------------------------------------------
- R2h= . R2h Adj= . F-Test = . P-Value > F(1 , 340) .
- R2v= 0.0000 R2v Adj=-0.0912 F-Test = 0.00 P-Value > F(1 , 340) 1.0000
------------------------------------------------------------------------------
| Coef. Std. Err. z P>|z| [95% Conf. Interval]
-------------+----------------------------------------------------------------
lvgdp |
lngdp | -1.52e-11 . . . . .
_cons | .0877433 . . . . .
-------------+----------------------------------------------------------------
/Lambda | 1 4.74e-13 2.1e+12 0.000 1 1
/Sigma | -3.10e-11 1.20e-12 -25.93 0.000 -3.33e-11 -2.86e-11
------------------------------------------------------------------------------


雷达卡




京公网安备 11010802022788号







