悬赏 1 个论坛币 未解决
用两年双向固定效应回归时,一些省份虚拟变量被ommt,第二年被ommit掉了,这不就等于没加时间效应一样吗,请问我该怎么处理这种问题呢?而且我发现不加个体年龄2016就不会被共线掉,加上年龄year2016就会被共线掉(换成年龄平方一样),但是一般不都会有年龄这一控制变量嘛,请问有没有解决办法,还是我不用管这种情况呢?xtset fid year
panel variable: fid (strongly balanced)
time variable: year, 2014 to 2016, but with gaps
delta: 1 unit
tabulate provcd, gen(dummy)
xtreg y x1 huzhu urban14 ln_shouru xiaofei familysize join_market debt age employ2014 jobclass_base badhealth goodhealth
> yanglao yibao dummy1- dummy27 i.year,fe r
note: dummy2 omitted because of collinearity
note: dummy3 omitted because of collinearity
note: dummy6 omitted because of collinearity
note: dummy8 omitted because of collinearity
note: dummy10 omitted because of collinearity
note: dummy19 omitted because of collinearity
note: dummy20 omitted because of collinearity
note: dummy22 omitted because of collinearity
note: dummy23 omitted because of collinearity
note: dummy24 omitted because of collinearity
note: dummy25 omitted because of collinearity
note: dummy26 omitted because of collinearity
note: dummy27 omitted because of collinearity
note: 2016.year omitted because of collinearity
Fixed-effects (within) regression Number of obs = 9,568
Group variable: fid Number of groups = 4,784
R-sq: Obs per group:
within = 0.6668 min = 2
between = 0.4745 avg = 2.0
overall = 0.5365 max = 2
F(24,4783) = .
corr(u_i, Xb) = -0.3573 Prob > F = .
(Std. Err. adjusted for 4,784 clusters in fid)
Robust
y Coef. Std. Err. t P>t [95% Conf. Interval]
x1 -.0127092 .004211 -3.02 0.003 -.0209647 -.0044537
huzhu .0169565 .0075507 2.25 0.025 .0021537 .0317593
urban14 -.0776622 .0363455 -2.14 0.033 -.1489161 -.0064083
ln_shouru .7088656 .0149843 47.31 0.000 .6794896 .7382417
xiaofei -9.38e-06 3.54e-07 -26.54 0.000 -.0000101 -8.69e-06
familysize -.0259619 .0068898 -3.77 0.000 -.039469 -.0124548
join_market -.003551 .0220683 -0.16 0.872 -.046815 .0397129
debt -.0578378 .0135537 -4.27 0.000 -.0844093 -.0312663
age .0026801 .0034166 0.78 0.433 -.0040179 .0093782
employ2014 .0051995 .0166206 0.31 0.754 -.0273846 .0377835
jobclass_base -.0077895 .0188907 -0.41 0.680 -.0448239 .029245
badhealth -.0217827 .0160138 -1.36 0.174 -.053177 .0096116
goodhealth -.0087379 .0163517 -0.53 0.593 -.0407947 .0233188
yanglao -.0049481 .012152 -0.41 0.684 -.0287716 .0188754
yibao .0189818 .0208459 0.91 0.363 -.0218858 .0598494
dummy1 -.0578959 .4055937 -0.14 0.886 -.8530462 .7372545
dummy2 0 (omitted)
dummy3 0 (omitted)
dummy4 .26755 .1732985 1.54 0.123 -.0721948 .6072948
dummy5 .3919417 .0593253 6.61 0.000 .2756369 .5082466
dummy6 0 (omitted)
dummy7 -.116472 .1196302 -0.97 0.330 -.3510022 .1180582
dummy8 0 (omitted)
dummy9 .9015568 .5281672 1.71 0.088 -.133894 1.937008
dummy10 0 (omitted)
dummy11 .1220651 .1305182 0.94 0.350 -.1338107 .3779408
dummy12 .1063478 .0482801 2.20 0.028 .0116966 .2009991
dummy13 .2793704 .0603441 4.63 0.000 .1610682 .3976725
dummy14 .8462843 .252204 3.36 0.001 .3518484 1.34072
dummy15 -.207423 .2828383 -0.73 0.463 -.7619162 .3470702
dummy16 -.1211229 .1172574 -1.03 0.302 -.3510013 .1087556
dummy17 -.0018524 .0329246 -0.06 0.955 -.0663999 .062695
dummy18 .0255594 .2296797 0.11 0.911 -.4247184 .4758373
dummy19 0 (omitted)
dummy20 0 (omitted)
dummy21 .0862908 .2477077 0.35 0.728 -.3993303 .5719118
dummy22 0 (omitted)
dummy23 0 (omitted)
dummy24 0 (omitted)
dummy25 0 (omitted)
dummy26 0 (omitted)
dummy27 0 (omitted)
year
2016 0 (omitted)
_cons -7.056169 .2107651 -33.48 0.000 -7.469366 -6.642973
sigma_u .35043685
sigma_e .31538931
rho .55249223 (fraction of variance due to u_i)