Hausman test

[maryam shafqaat]

Hi,

I'm running a panel regression. These are my panel commands, followed by regression commands for random effects and fixed effects and hausman test. But the hausman test says,

chi2(0) = (b-B)'[(V_b-V_B)^(-1)](b-B)
= 0.00
Prob>chi2 = .
(V_b-V_B is not positive definite)


Can someone please guide me? Is there an issue in my data set, panel commands or what?

[Carlo Lazzaro]

Maryam:
what if you invoke the option -sigmamore- available from -hausman-?

[maryam shafqaat]

Carlo Lazzaro I just tried and no luck. Is there an issue with my data?

[maryam shafqaat]

egen idc = group( Distrcts)
move idc Distrcts
xtset idc
xtset idc Years, yearly
xtreg MPI BISPEXP CANTONMENT BANKS AGRICULTURE
xtreg MPI BISPEXP CANTONMENT BANKS AGRICULTURE, fe
estimate store random
estimate store fixed
hausman random fixed

These are my commands

[Carlo Lazzaro]

Maryam:
the main issue is that the correct sequence should be:

hausman fixed random

Besides,

xtset idc

is redundant vs

xtset idc Years, yearly

[maryam shafqaat]

Thank you Cralo, Ill try these.

[Eric de Souza]

You also need to put -estimates store random- soon after the first xtreg command:
xtreg MPI BISPEXP CANTONMENT BANKS AGRICULTURE
estimate store random
xtreg MPI BISPEXP CANTONMENT BANKS AGRICULTURE, fe
estimate store fixed

[maryam shafqaat]

Thanks Eric, will do

[Carlo Lazzaro]

Maryam:
as Eric skillfully noticed, your delay in saving -xtreg,re- (that I baldly overlooked) made -hausman- comparing the same -fe- model (that you actually saved, inadvertently, twice):, as you can see from the following toy-example, that replicates your discouraging (but really frequent even when every step before -hausman- is correct) -hausman- result:

. xtreg ln_wage age, fe

Fixed-effects (within) regression               Number of obs     =     28,510
Group variable: idcode                          Number of groups  =      4,710

R-sq:                                           Obs per group:
     within  = 0.1026                                         min =          1
     between = 0.0877                                         avg =        6.1
     overall = 0.0774                                         max =         15

                                                F(1,23799)        =    2720.20
corr(u_i, Xb)  = 0.0314                         Prob > F          =     0.0000

------------------------------------------------------------------------------
     ln_wage |      Coef.   Std. Err.      t    P>|t|     [95% Conf. Interval]
-------------+----------------------------------------------------------------
         age |   .0181349   .0003477    52.16   0.000     .0174534    .0188164
       _cons |   1.148214   .0102579   111.93   0.000     1.128107     1.16832
-------------+----------------------------------------------------------------
     sigma_u |  .40635023
     sigma_e |  .30349389
         rho |  .64192015   (fraction of variance due to u_i)
------------------------------------------------------------------------------
F test that all u_i=0: F(4709, 23799) = 8.81                 Prob > F = 0.0000

. xtreg ln_wage age, re

Random-effects GLS regression                   Number of obs     =     28,510
Group variable: idcode                          Number of groups  =      4,710

R-sq:                                           Obs per group:
     within  = 0.1026                                         min =          1
     between = 0.0877                                         avg =        6.1
     overall = 0.0774                                         max =         15

                                                Wald chi2(1)      =    3140.35
corr(u_i, X)   = 0 (assumed)                    Prob > chi2       =     0.0000

------------------------------------------------------------------------------
     ln_wage |      Coef.   Std. Err.      z    P>|z|     [95% Conf. Interval]
-------------+----------------------------------------------------------------
         age |   .0185667   .0003313    56.04   0.000     .0179174    .0192161
       _cons |   1.120439   .0112038   100.01   0.000      1.09848    1.142398
-------------+----------------------------------------------------------------
     sigma_u |  .36972456
     sigma_e |  .30349389
         rho |  .59743613   (fraction of variance due to u_i)
------------------------------------------------------------------------------

. estimates store fe

. estimates store re

. hausman fe re

Note: the rank of the differenced variance matrix (0) does not equal the number of coefficients being tested (1); be
        sure this is what you expect, or there may be problems computing the test.  Examine the output of your
        estimators for anything unexpected and possibly consider scaling your variables so that the coefficients are on
        a similar scale.

                 ---- Coefficients ----
             |      (b)          (B)            (b-B)     sqrt(diag(V_b-V_B))
             |       fe           re         Difference          S.E.
-------------+----------------------------------------------------------------
         age |    .0185667     .0185667               0               0
------------------------------------------------------------------------------
                           b = consistent under Ho and Ha; obtained from xtreg
            B = inconsistent under Ha, efficient under Ho; obtained from xtreg

    Test:  Ho:  difference in coefficients not systematic

                  chi2(0) = (b-B)'[(V_b-V_B)^(-1)](b-B)
                          =        0.00
                Prob>chi2 =           .
                (V_b-V_B is not positive definite)

Set aside my need for a new pair of spectacles, what happened remind me of the need to post what one have typed and received from Stata (output tables included), as recommended by the FAQ.

[maryam shafqaat]

Thank you Carlo for the help!