Notice: On April 23, 2014, Statalist moved from an email list to a forum, based at statalist.org.

# st: xttest0 confusion

 From Humaira Asad To STATA HELP Subject st: xttest0 confusion Date Wed, 11 May 2011 18:24:51 +0000

```

Hi,

I have estimated the folowing model using re vce(cluster country) and then run the BP test which is significant. This means the individual country effects are random. But when I estimate serial scorrelation in the residuals there is serial correlation present. What should I do, it indicates endogeneity in the regressors? If yes why BP test is showing country effects are random? Confused?

xi: xtreg l_gini_u l_gdp_gr l_pcrdbgdp l_ls l_kg l_1_inf legal_or  pcrdbgdpls i.year, re vce(cluster cn_
> no)
i.year            _Iyear_1965-2010    (naturally coded; _Iyear_1965 omitted)
Random-effects GLS regression                   Number of obs      =       487
Group variable: cn_no                           Number of groups   =        84
R-sq:  within  = 0.1530                         Obs per group: min =         1
between = 0.5219                                        avg =       5.8
overall = 0.4163                                        max =        10
Random effects u_i ~ Gaussian                   Wald chi2(16)      =    153.49
corr(u_i, X)       = 0 (assumed)                Prob > chi2        =    0.0000
(Std. Err. adjusted for 84 clusters in cn_no)
------------------------------------------------------------------------------
|               Robust
l_gini_u |      Coef.   Std. Err.      z    P>|z|     [95% Conf. Interval]
-------------+----------------------------------------------------------------
l_gdp_gr |   .0057638   .0124442     0.46   0.643    -.0186264    .0301541
l_pcrdbgdp |   .3042015   .0936341     3.25   0.001     .1206821    .4877208
l_ls |  -.2380501   .0567163    -4.20   0.000    -.3492119   -.1268882
l_kg |  -.0429575   .0335356    -1.28   0.200     -.108686    .0227709
l_1_inf |   .0226364   .0100428     2.25   0.024     .0029528      .04232
legal_or |  -.0846046   .0145304    -5.82   0.000    -.1130836   -.0561256
pcrdbgdpls |  -.0937477   .0251026    -3.73   0.000    -.1429478   -.0445475
_Iyear_1970 |   -.062607   .0310871    -2.01   0.044    -.1235367   -.0016773
_Iyear_1975 |  -.0810071    .044773    -1.81   0.070    -.1687604    .0067463
_Iyear_1980 |  -.1041635    .051172    -2.04   0.042    -.2044588   -.0038682
_Iyear_1985 |   -.150318   .0520214    -2.89   0.004    -.2522781   -.0483579
_Iyear_1990 |   -.120455   .0492264    -2.45   0.014    -.2169369   -.0239731
_Iyear_1995 |  -.0623324   .0510912    -1.22   0.222    -.1624692    .0378044
_Iyear_2000 |  -.0422021     .05261    -0.80   0.422    -.1453159    .0609117
_Iyear_2005 |  -.0487584   .0543264    -0.90   0.369    -.1552362    .0577194
_Iyear_2010 |  -.0483711   .0573857    -0.84   0.399    -.1608451    .0641028
_cons |   4.834027   .2165074    22.33   0.000      4.40968    5.258374
-------------+----------------------------------------------------------------
sigma_u |  .12337294
sigma_e |  .12342285
rho |  .49979776   (fraction of variance due to u_i)
------------------------------------------------------------------------------
. xttest0
Breusch and Pagan Lagrangian multiplier test for random effects
l_gini_u[cn_no,t] = Xb + u[cn_no] + e[cn_no,t]
Estimated results:
|       Var     sd = sqrt(Var)
---------+-----------------------------
l_gini_u |   .0678958       .2605682
e |   .0152332       .1234228
u |   .0152209       .1233729
Test:   Var(u) = 0
chi2(1) =   461.53
Prob > chi2 =     0.0000