I ran xtregress for comparison and it did just fine. Would anyone help me identify what was wrong with the process? Thank you!
Additional information: I use STATA 15.0 for Windows
Code:
. spxtregress lgdp_pcp y2005 y2006 y2007
> dead_prc
> y2005_dead_prc y2006_dead_prc y2007_dead_prc
> year, re errorlag(wmat);
(294 observations)
(126 observations excluded due to missing values)
(168 observations used)
(data contain 21 panels (places) )
(weighting matrix defines 21 places)
note: lgdp_pcp:y2005 omitted because of collinearity
note: lgdp_pcp:y2005_dead_prc omitted because of collinearity
note: lgdp_pcp:y2007_dead_prc omitted because of collinearity
note: lgdp_pcp:_cons omitted because of collinearity
Fitting starting values:
Iteration 0: log likelihood = 57.911505
Iteration 1: log likelihood = 57.94042
Iteration 2: log likelihood = 57.940496
Iteration 3: log likelihood = 57.940496
Optimizing concentrated log likelihood:
initial: log likelihood = 26.544279
improve: log likelihood = 26.544279
rescale: log likelihood = 26.544279
rescale eq: log likelihood = 37.92344
Iteration 0: log likelihood = 37.92344
Iteration 1: log likelihood = 42.978376
Iteration 2: log likelihood = 43.020721
Iteration 3: log likelihood = 43.020755
Iteration 4: log likelihood = 43.020755
Optimizing unconcentrated log likelihood:
Iteration 0: log likelihood = 43.020755
Iteration 1: log likelihood = 43.020755 (backed up)
Random-effects spatial regression Number of obs = 168
Group variable: _ID Number of groups = 21
Obs per group = 8
Wald chi2(5) = 489.17
Prob > chi2 = 0.0000
Log likelihood = 43.0208 Pseudo R2 = 0.0944
--------------------------------------------------------------------------------
lgdp_pcp | Coef. Std. Err. z P>|z| [95% Conf. Interval]
---------------+----------------------------------------------------------------
lgdp_pcp |
y2005 | 0 (omitted)
y2006 | 19.68171 11.2663 1.75 0.081 -2.399835 41.76326
y2007 | 19.68406 11.28865 1.74 0.081 -2.441293 41.80941
dead_prc | .0115508 .0070841 1.63 0.103 -.0023337 .0254353
y2005_dead_prc | 0 (omitted)
y2006_dead_prc | -.0012556 .0032023 -0.39 0.695 -.007532 .0050208
y2007_dead_prc | 0 (omitted)
year | -.0089161 .0056161 -1.59 0.112 -.0199234 .0020913
_cons | 0 (omitted)
---------------+----------------------------------------------------------------
wmat |
e.lgdp_pcp | 2.14346 .1849234 11.59 0.000 1.781016 2.505903
---------------+----------------------------------------------------------------
/sigma_u | .377704 .0592981 .2776618 .5137918
/sigma_e | .1512049 .0088053 .1348952 .1694864
--------------------------------------------------------------------------------
Wald test of spatial terms: chi2(1) = 134.35 Prob > chi2 = 0.0000
. xtreg lgdp_pcp y2005 y2006 y2007
> dead_prc
> y2005_dead_prc y2006_dead_prc y2007_dead_prc
> year, re;
Random-effects GLS regression Number of obs = 257
Group variable: _ID Number of groups = 21
R-sq: Obs per group:
within = 0.2149 min = 8
between = 0.0072 avg = 12.2
overall = 0.0781 max = 14
Wald chi2(8) = 63.04
corr(u_i, X) = 0 (assumed) Prob > chi2 = 0.0000
--------------------------------------------------------------------------------
lgdp_pcp | Coef. Std. Err. z P>|z| [95% Conf. Interval]
---------------+----------------------------------------------------------------
y2005 | .1517669 .094186 1.61 0.107 -.0328343 .336368
y2006 | .1831134 .0954493 1.92 0.055 -.0039638 .3701905
y2007 | .0616214 .1022244 0.60 0.547 -.1387349 .2619776
dead_prc | .000789 .0099109 0.08 0.937 -.018636 .0202141
y2005_dead_prc | -.0021991 .0063445 -0.35 0.729 -.0146341 .0102359
y2006_dead_prc | .0058923 .0063114 0.93 0.351 -.0064777 .0182624
y2007_dead_prc | .0109464 .0034607 3.16 0.002 .0041636 .0177293
year | .0240922 .011762 2.05 0.041 .0010392 .0471453
_cons | -46.72111 23.55381 -1.98 0.047 -92.88572 -.5564968
---------------+----------------------------------------------------------------
sigma_u | .52632943
sigma_e | .31472541
rho | .73661597 (fraction of variance due to u_i)
--------------------------------------------------------------------------------
No comments:
Post a Comment