jannsz06 发表于 2012-2-22 15:58 
或者在你的命令后附加选项加入nomata 就可以了

我现在可以做出结果来了,用xtabond2时出现这种结果:
xtabond2 tfp l.tfp exp ds l(0/1).( cap lab wag) yr*, gmm(l.(tfp exp ds cap wag lab)) iv(yr* v212,eq(level)) robust small
Favoring space over speed. To switch, type or click on mata: mata set matafavor speed, perm.
yr2007 dropped due to collinearity
Warning: Two-step estimated covariance matrix of moments is singular.
Using a generalized inverse to calculate robust weighting matrix for Hansen test.
Difference-in-Sargan statistics may be negative.
Dynamic panel-data estimation, one-step system GMM
------------------------------------------------------------------------------
Group variable: id Number of obs = 25912
Time variable : year Number of groups = 3239
Number of instruments = 267 Obs per group: min = 8
F(16, 3238) = 606.66 avg = 8.00
Prob > F = 0.000 max = 8
------------------------------------------------------------------------------
| Robust
tfp | Coef. Std. Err. t P>|t| [95% Conf. Interval]
-------------+----------------------------------------------------------------
tfp |
L1. | .1178575 .0101662 11.59 0.000 .0979247 .1377902
~~~~~~~~~~~
然后我有用stata中的xtdpdsys 做,结果又是这样的:
xtdpdsys tfp exp ds x, lags(2) maxldep(4) maxlags(4) twostep pre(wag, lagstruct(1,2)) endog(cap l
> ab, lagstruct(2,2)) artests(2)
System dynamic panel-data estimation Number of obs = 25912
Group variable: id Number of groups = 3239
Time variable: year
Obs per group: min = 8
avg = 8
max = 8
Number of instruments = 112 Wald chi2(13) = 1985.92
Prob > chi2 = 0.0000
Two-step results
------------------------------------------------------------------------------
tfp | Coef. Std. Err. z P>|z| [95% Conf. Interval]
-------------+----------------------------------------------------------------
tfp |
L1. | -.1493898 .020698 -7.22 0.000 -.1899571 -.1088225
L2. | -.098065 .0176222 -5.56 0.000 -.1326039 -.063526
wag |
--. | 1.211775 .1119757 10.82 0.000 .9923069 1.431244
L1. | -.0310314 .0732519 -0.42 0.672 -.1746026 .1125398
cap |
--. | 2.027527 .2150829 9.43 0.000 1.605972 2.449082
L1. | -.789193 .1912774 -4.13 0.000 -1.16409 -.4142963
L2. | -1.009526 .1720639 -5.87 0.000 -1.346765 -.6722869
lab |
--. | .8132374 .2897639 2.81 0.005 .2453107 1.381164
L1. | -.0381389 .2215211 -0.17 0.863 -.4723124 .3960345
L2. | -.4914825 .1215177 -4.04 0.000 -.7296528 -.2533122
exp | .1561342 .1118842 1.40 0.163 -.0631548 .3754232
ds | .2202115 .1078105 2.04 0.041 .0089068 .4315162
x | -.0137616 .0092327 -1.49 0.136 -.0318574 .0043343
_cons | -.2406626 1.109699 -0.22 0.828 -2.415633 1.934308
------------------------------------------------------------------------------
Warning: gmm two-step standard errors are biased; robust standard
errors are recommended.
Instruments for differenced equation
GMM-type: L(2/5).tfp L(1/4).L.wag L(2/4).L2.cap L(2/4).L2.lab
Standard: D.exp D.ds D.x
Instruments for level equation
GMM-type: LD.tfp LD.wag L3D.cap L3D.lab
Standard: _cons
.
其中Warning: gmm two-step standard errors are biased; robust standard
errors are recommended.
这句话说明了什么问题呀?是不是都说明了我的工具变量选择不好呀?真的谢谢楼主的帮助了~~~~