Stata The Stata listserver
[Date Prev][Date Next][Thread Prev][Thread Next][Date index][Thread index]

st: 2SLS fixed effects estimation and time effects


From   "Katarina Lynch" <[email protected]>
To   [email protected]
Subject   st: 2SLS fixed effects estimation and time effects
Date   Tue, 21 Dec 2004 10:20:17 +0000


Dear all,

I run a command

xi: xtivreg y x1 x2 x3 i.year, fe

and get very large standard errors, p-value of the F-test (null: no fixed effects) = 1

however,

xi: xtreg y x1 x2 x3 i.year, fe

gives more sensible results. Why does the fixed effects estimation with instrumental variables when time effects are controlled for, give no better results than in OLS? I have a feeling that time effects are incompatible with xtivreg, fe because fixed effects captures within variation over time. However, in OLS case it is working fine. I would very much appreciate any help! Instrumental variables are lagged values of the regressors.

Below is the output:

xi: xtivreg Y X1 X2 X3 X4 i.year (X5 X6 X7=X5_lag X6_lag

 X7_lag), fe
i.year _Iyear_1960-1990 (naturally coded; _Iyear_1960 omitted)

Fixed-effects (within) IV regression Number of obs = 123
Group variable: id2 Number of groups = 54

R-sq: within = . Obs per group: min = 1
between = 0.0386 avg = 2.3
overall = 0.0655 max = 3

Wald chi2(8) = 275.75
corr(u_i, Xb) = -0.8763 Prob > chi2 = 0.0000

------------------------------------------------------------------------------
Y| Coef. Std. Err. z P>|z| [95% Conf. Interval]
-------------+----------------------------------------------------------------
X5 | 9.052149 12.92154 0.70 0.484 -16.2736 34.3779
X6| -7.896397 15.88667 -0.50 0.619 -39.0337 23.24091
X7 | -42.29215 64.92772 -0.65 0.515 -169.5481 84.96384
X1 | -6.873546 3.581599 -1.92 0.055 -13.89335 .1462598
X2 | 2.607123 5.567398 0.47 0.640 -8.304776 13.51902
X3| 9.689091 16.8018 0.58 0.564 -23.24184 42.62002
X4 | (dropped)
_Iyear_1970 | -10.75312 11.70917 -0.92 0.358 -33.70266 12.19643
_Iyear_1980 | -3.804827 4.229946 -0.90 0.368 -12.09537 4.485716
_Iyear_1990 | (dropped)
_cons | 152.8336 179.9555 0.85 0.396 -199.8727 505.5399
-------------+----------------------------------------------------------------
sigma_u | 7.7108873
sigma_e | 4.4741294
rho | .7481261 (fraction of variance due to u_i)
------------------------------------------------------------------------------
F test that all u_i=0: F(53,61) = 0.28 Prob > F = 1.0000
------------------------------------------------------------------------------
Instrumented: X5 X6 X7
Instruments: X1 X2 X3 X4 _Iyear_1970 _Iyear_1980 _Iyear_1990
X5_lag X6_lag X7_lag

.
.

_________________________________________________________________
Wenn�s mal schnell gehen soll. MSN Suche. http://search.msn.de/ Jetzt kostenlos nutzen und keine Zeit mehr verlieren.

*
* For searches and help try:
* http://www.stata.com/support/faqs/res/findit.html
* http://www.stata.com/support/statalist/faq
* http://www.ats.ucla.edu/stat/stata/




© Copyright 1996–2024 StataCorp LLC   |   Terms of use   |   Privacy   |   Contact us   |   What's new   |   Site index