Why I get some missing values in the Hausman test results (in the right most column)?

[Ijaz Ahmad]

I have estimated fixed, random, hybrid, and correlated random effects models following Reinhard Schunck's article "Within and between estimates in random-effects models: Advantages and drawbacks of correlated random effects and hybrid models." However, when I run the Hausman test, I get some dots in the right-most column in the following result. My panel data has 50 clusters and 17 periods and is strongly balanced.

Any advice would be appreciated.

. hausman fe re

                 ---- Coefficients ----
             |      (b)          (B)            (b-B)     sqrt(diag(V_b-V_B))
             |       fe           re         Difference       Std. err.
-------------+----------------------------------------------------------------
     econfre |     .735193      .652026         .083167        .0628192
prop_pvt_enr |    3.931047      2.89601        1.035037        .4946274
legis_profes |    .0014851     .0059278       -.0044427        .0039628
        year |
       2006  |    .1999835     .2203892       -.0204056               .
       2007  |    .4463056     .4628292       -.0165236               .
       2008  |    .7932149     .8093098       -.0160949               .
       2009  |    1.046555     1.015324        .0312309        .0343462
       2010  |    1.046008     1.026366         .019642        .0383504
       2011  |     1.18577     1.182135        .0036346        .0263476
       2012  |    1.159966     1.185721       -.0257553         .009921
       2013  |    1.395991     1.418851       -.0228602        .0078753
       2014  |    1.181323     1.227975       -.0466517        .0100741
       2015  |    1.245597     1.284881       -.0392841               .
       2016  |    1.271927     1.304402       -.0324754               .
       2017  |    1.174128     1.205104       -.0309764               .
       2018  |    1.143925     1.176945       -.0330199               .
       2019  |    1.248591     1.269823       -.0212323               .
       2020  |    1.639967     1.681629       -.0416624               .
       2021  |    1.400029     1.435361       -.0353313               .
------------------------------------------------------------------------------
                          b = Consistent under H0 and Ha; obtained from xtreg.
           B = Inconsistent under Ha, efficient under H0; obtained from xtreg.

Test of H0: Difference in coefficients not systematic

   chi2(19) = (b-B)'[(V_b-V_B)^(-1)](b-B)
            =   6.94
Prob > chi2 = 0.9945
(V_b-V_B is not positive definite)

.
end of do-file

However, I do not get these dots when I run fe and re-models before the Hausman test without including the year-fixed effects. See the following results:

. hausman modelFE3 modelRE3

                 ---- Coefficients ----
             |      (b)          (B)            (b-B)     sqrt(diag(V_b-V_B))
             |    modelFE3     modelRE3      Difference       Std. err.
-------------+----------------------------------------------------------------
     econfre |    .6612818     .5860479        .0752338         .038665
prop_pvt_enr |    6.010547     4.245311        1.765236        .5620877
legis_profes |   -.0047386     .0008039       -.0055425        .0041265
------------------------------------------------------------------------------
                          b = Consistent under H0 and Ha; obtained from xtreg.
           B = Inconsistent under Ha, efficient under H0; obtained from xtreg.

Test of H0: Difference in coefficients not systematic

    chi2(3) = (b-B)'[(V_b-V_B)^(-1)](b-B)
            =  13.18
Prob > chi2 = 0.0043

.
end of do-file

Does anyone know why I am getting these dots? I am not sure whether the information I provide is sufficient to understand my problem. Any advice would be appreciated.

Ijaz