Why Wald Ch2 and Prob>ch2 value are not reported for a random effect model with clustered errors?

[Mohamud Hussein]

Hi members,

I run a random effects model with standards errors cluster on my panel id. The ouput does not report any Wald Ch2 and Prob>ch2 values.

Stata manauls says that "The VCE you have just estimated is does have not of sufficient rank to perform the model test", and indicates that, although estimated coefficients may be fine, I "need to consider carefully whether any of the reported standard errors mean anything".

I am not quite sure what this means? Are estimated beta and SEs good enough for reporting or I need to explore an alternative model, see attached output?

Can you please advice me on this?

Thanks,
mohamud

[Clyde Schechter]

You are more likely to get a helpful answer is you show us the actual output. Copy it from the Results window (or your log file) and paste it into a code block on this forum in a new post on this thread [See FAQ if you don't know how to do this].

[Mohamud Hussein]

Clyde- sorry, I tried to attach the output for some reason did not succeed. Here is my ouput:


. xtreg PriceKg i.b2.cool branv packz prodc prodt Month retai, re vce(cluster SKUCode)

Random-effects GLS regression Number of obs = 1009
Group variable: SKUCode Number of groups = 19

R-sq: within = 0.1684 Obs per group: min = 28
between = 0.7066 avg = 53.1
overall = 0.4162 max = 60

Wald chi2(9) = .
corr(u_i, X) = 0 (assumed) Prob > chi2 = .

(Std. Err. adjusted for 19 clusters in SKUCode)

Robust
PriceKg Coef. Std. Err. z P>z [95% Conf. Interval]

cool
1 -.2862552 .8629927 -0.33 0.740 -1.97769 1.405179
3 -.0438983 .7750043 -0.06 0.955 -1.562879 1.475082
4 .9265256 .5208154 1.78 0.075 -.0942539 1.947305
5 3.584071 .702803 5.10 0.000 2.206602 4.961539

branv .6495303 .2859727 2.27 0.023 .0890341 1.210027
packz .3103391 .1842407 1.68 0.092 -.0507661 .6714443
prodc -2.360135 .8975332 -2.63 0.009 -4.119268 -.6010024
prodt .3636645 .1571556 2.31 0.021 .0556452 .6716839
Month .0501156 .0116846 4.29 0.000 .0272142 .073017
retai -.6052335 .209236 -2.89 0.004 -1.015328 -.1951385
_cons 4.756011 1.618619 2.94 0.003 1.583575 7.928447

sigma_u 1.0284748
sigma_e 2.0245792
rho .20512425 (fraction of variance due to u_i)

[Clyde Schechter]

The usual cause of this error is that the number of coefficients in the model exceeds the number of clusters when using the cluster robust VCE. But you have 19 clusters, which is in theory enough to estimate 10 coefficients and a constant (although not enough that I'd place much confidence in the results). The other way this can happen is if there is a variable that only takes on non-zero values in a single cluster. One of the categories of the cool variable would be a good candidate for this problem. (Remember that what counts is how many values there are in the estimation sample: so additional non-zero observations when other model variables have a missing value don't help you here.) If you identify such a cluster, you might want to consider either excluding it from analysis, excluding the variable that the cluster is problematic for from the analysis, or combining that cluster with one of the other clusters--if there is a way to do so that doesn't do violence to the meaning of the clustering variable.

[Carlo Lazzaro]

Mohamud:
pasting what you typed and what Stata gave you back in between code delimiters (#-icon among the options of the A-icon Advanced editor) will ease your attempt to post all the details you deem relevant for helpful replies.

[Mohamud Hussein]

Many thanks Clyde.

I checked whether Wald ch2 and prob>ch2 were appeared when eliminated the entire cool variable and can confirm that this variable is the one causing the problem as you anticipated.

However, I am not quite sure whether I can drop the category which contains only zeros for 3 out of my 20 panel variable (SKUcode).

Cool codes for a declared country origin label for 20 food items under study. The zero in this case indicates that a country of origin has not been declared for a particular item. If I eliminate the 3 category with zeros as advised that means I will reduce the data size significantly but more importantly would also excludes a relavant category as I am assessing the willingness to pay for a country of origin in content both undeclared and declared origin.

Other suggested option for merging the offending category with another one is not feasible too in this case as undeclared origin is different from declared country.

Perhaps I need to explore an alternative model?

Thanks,
Mohamud

[Sinem Ates]

Dear All,

I have the same problem of missing Wald chi2 and prob>chi2 with the random effect model below. Could you please advise me how I can find the cause of the problem?

Thank you in advance.

xtreg PRICEENDOFQUARTER_w BVPS_w EPS_w NON_SI_ALL SIZE_w LEV_w ROE_w i.INDG DATE2 DATE3 DATE4 DATE5 DATE6 DATE7 DATE8 DATE9 DATE10 DA
> TE11 DATE12 DATE13 DATE14 DATE15 DATE16 DATE17 DATE18 DATE19, re robust

Random-effects GLS regression                   Number of obs     =      6,579
Group variable: ID                              Number of groups  =        409

R-sq:                                           Obs per group:
     within  = 0.1713                                         min =          1
     between = 0.5657                                         avg =       16.1
     overall = 0.5764                                         max =         19

                                                Wald chi2(82)     =          .
corr(u_i, X)   = 0 (assumed)                    Prob > chi2       =          .

                                   (Std. Err. adjusted for 409 clusters in ID)
------------------------------------------------------------------------------
             |               Robust
PRICEENDOF~w |      Coef.   Std. Err.      z    P>|z|     [95% Conf. Interval]
-------------+----------------------------------------------------------------
      BVPS_w |   .8919972   .2118278     4.21   0.000     .4768223    1.307172
       EPS_w |   1.108857   .5716956     1.94   0.052    -.0116461    2.229359
  NON_SI_ALL |  -6.543971   2.328437    -2.81   0.005    -11.10762   -1.980319
      SIZE_w |  -.3893046   .4819263    -0.81   0.419    -1.333863    .5552536
       LEV_w |   .8643213   1.101401     0.78   0.433    -1.294384    3.023027
       ROE_w |   -2.04782   .8253725    -2.48   0.013     -3.66552   -.4301196
             |
        INDG |
         31  |  -4.090169   3.761615    -1.09   0.277     -11.4628    3.282461
         33  |  -6.151265   4.143106    -1.48   0.138     -14.2716    1.969075
         34  |  -7.929094   3.906076    -2.03   0.042    -15.58486    -.273327
         35  |  -7.276811   4.497647    -1.62   0.106    -16.09204    1.538416
         37  |  -7.054528   3.957894    -1.78   0.075    -14.81186    .7028017
         39  |  -8.369118   3.779185    -2.21   0.027    -15.77619   -.9620511
         41  |  -8.807136    3.88224    -2.27   0.023    -16.41619   -1.198085
         43  |  -6.909256   4.184706    -1.65   0.099    -15.11113    1.292618
         45  |  -2.581687   5.529887    -0.47   0.641    -13.42007    8.256693
         50  |   11.26811   18.53219     0.61   0.543    -25.05431    47.59053
         51  |  -8.890667    4.05465    -2.19   0.028    -16.83764   -.9436989
         53  |  -7.995907   3.968383    -2.01   0.044     -15.7738    -.218019
         54  |  -8.160738   3.745324    -2.18   0.029    -15.50144   -.8200376
         55  |    .135205   7.528914     0.02   0.986    -14.62119     14.8916
         56  |  -6.044666   4.036187    -1.50   0.134    -13.95545    1.866115
         57  |  -8.337951   4.008646    -2.08   0.038    -16.19475   -.4811482
         58  |  -.5998357   7.178546    -0.08   0.933    -14.66953    13.46986
         59  |   -7.60933   3.906017    -1.95   0.051    -15.26498    .0463223
         60  |  -5.730984   4.648281    -1.23   0.218    -14.84145     3.37948
         62  |  -8.273364   4.144884    -2.00   0.046    -16.39719   -.1495407
         63  |   22.93586   17.22974     1.33   0.183    -10.83382    56.70554
         64  |  -3.176404   4.964211    -0.64   0.522    -12.90608     6.55327
         65  |   7.379138   5.273016     1.40   0.162    -2.955783    17.71406
         67  |  -8.799142   5.571442    -1.58   0.114    -19.71897    2.120684
         69  |   -8.00599   3.981803    -2.01   0.044    -15.81018      -.2018
         70  |   5.636822   13.26535     0.42   0.671    -20.36279    31.63643
         71  |  -3.728197   5.977415    -0.62   0.533    -15.44372    7.987322
         72  |  -7.561192   4.136759    -1.83   0.068    -15.66909    .5467066
         74  |  -9.616042   3.905686    -2.46   0.014    -17.27105   -1.961037
         75  |  -10.84639   4.350778    -2.49   0.013    -19.37376   -2.319022
         77  |   -10.4228   3.996894    -2.61   0.009    -18.25657   -2.589037
         80  |  -5.162223   5.668116    -0.91   0.362    -16.27153     5.94708
         82  |  -8.325002    3.79173    -2.20   0.028    -15.75666   -.8933474
         84  |  -8.946675     3.8319    -2.33   0.020    -16.45706   -1.436289
         86  |  -8.633917    3.90072    -2.21   0.027    -16.27919   -.9886457
         87  |  -8.846921   4.511313    -1.96   0.050    -17.68893   -.0049104
         88  |   8.412477   8.146606     1.03   0.302    -7.554578    24.37953
         90  |  -7.533813   3.956791    -1.90   0.057    -15.28898    .2213561
         92  |  -8.885393   4.648166    -1.91   0.056    -17.99563     .224844
         93  |  -9.307036   3.846315    -2.42   0.016    -16.84567   -1.768398
         94  |  -7.760339   4.025873    -1.93   0.054     -15.6509     .130227
         95  |  -8.122615   3.784774    -2.15   0.032    -15.54064   -.7045943
         96  |  -7.161707   3.881206    -1.85   0.065    -14.76873    .4453169
         97  |  -9.115836   4.043552    -2.25   0.024    -17.04105    -1.19062
        101  |  -12.28656   4.213538    -2.92   0.004    -20.54494   -4.028177
        102  |  -10.64283   3.896835    -2.73   0.006    -18.28049   -3.005176
        104  |  -11.45759   4.210918    -2.72   0.007    -19.71084   -3.204341
        106  |  -.9673605     4.4311    -0.22   0.827    -9.652158    7.717437
        107  |  -7.300037   3.774419    -1.93   0.053    -14.69776    .0976885
        111  |  -9.055243   3.835772    -2.36   0.018    -16.57322   -1.537267
        112  |  -6.873581   4.419683    -1.56   0.120      -15.536    1.788839
        113  |  -9.884155   3.906565    -2.53   0.011    -17.54088   -2.227428
        114  |  -4.476619   5.557725    -0.81   0.421    -15.36956    6.416322
        117  |   19.53413   14.80936     1.32   0.187    -9.491675    48.55994
        119  |   2.788215   4.084507     0.68   0.495    -5.217272     10.7937
        120  |  -7.882104   3.785852    -2.08   0.037    -15.30224   -.4619712
        126  |  -7.505419   4.110076    -1.83   0.068    -15.56102    .5501812
        127  |  -13.88023   3.759113    -3.69   0.000    -21.24795     -6.5125
        129  |  -11.22021   5.078208    -2.21   0.027    -21.17331   -1.267103
        132  |  -7.830173   4.046324    -1.94   0.053    -15.76082     .100475
        141  |  -8.810928   3.769013    -2.34   0.019    -16.19806   -1.423798
        142  |  -9.294045   4.216365    -2.20   0.028    -17.55797   -1.030122
        143  |  -7.907237   4.032082    -1.96   0.050    -15.80997   -.0045009
        150  |  -8.749552    3.97115    -2.20   0.028    -16.53286   -.9662407
        156  |  -2.776763   3.824697    -0.73   0.468    -10.27303    4.719504
        157  |  -8.802144   3.944942    -2.23   0.026    -16.53409   -1.070201
        160  |  -9.414771    3.77758    -2.49   0.013    -16.81869   -2.010851
        161  |  -6.237782   4.928043    -1.27   0.206    -15.89657    3.421004
        162  |  -7.272383   4.011217    -1.81   0.070    -15.13422    .5894575
        163  |  -10.48828   3.813872    -2.75   0.006    -17.96333   -3.013232
        166  |  -10.21665   3.731764    -2.74   0.006    -17.53078   -2.902532
        167  |  -6.877915   3.829593    -1.80   0.072    -14.38378    .6279499
        169  |  -12.29671   4.315027    -2.85   0.004    -20.75401   -3.839411
             |
       DATE2 |   .9400408   .2426848     3.87   0.000     .4643873    1.415694
       DATE3 |   .3056869   .2427755     1.26   0.208    -.1701443    .7815181
       DATE4 |   1.333358   .3186978     4.18   0.000     .7087222    1.957995
       DATE5 |   1.331914   .3735602     3.57   0.000     .5997495    2.064079
       DATE6 |   1.547164   .4306687     3.59   0.000     .7030684    2.391259
       DATE7 |   .9151938   .4201599     2.18   0.029     .0916956    1.738692
       DATE8 |   .8429149   .4211134     2.00   0.045     .0175479    1.668282
       DATE9 |   1.508224    .432611     3.49   0.000     .6603219    2.356126
      DATE10 |    .855066    .418458     2.04   0.041     .0349033    1.675229
      DATE11 |   .8823469   .4042084     2.18   0.029     .0901131    1.674581
      DATE12 |   .9111223   .4336018     2.10   0.036     .0612785    1.760966
      DATE13 |    1.74371   .4895226     3.56   0.000     .7842633    2.703157
      DATE14 |   2.145244   .4694482     4.57   0.000     1.225142    3.065345
      DATE15 |   2.155148    .463622     4.65   0.000     1.246466    3.063831
      DATE16 |     2.5098   .4826357     5.20   0.000     1.563851    3.455748
      DATE17 |   2.855015   .5120923     5.58   0.000     1.851333    3.858698
      DATE18 |   1.447598   .4884295     2.96   0.003     .4902938    2.404902
      DATE19 |   .9693504   .4993351     1.94   0.052    -.0093284    1.948029
       _cons |   18.71972   8.127826     2.30   0.021     2.789474    34.64997
-------------+----------------------------------------------------------------
     sigma_u |  11.214496
     sigma_e |  3.9781354
         rho |  .88823004   (fraction of variance due to u_i)
------------------------------------------------------------------------------

[Carlo Lazzaro]

Sinem:
see: help j_robustsingular.

[Sinem Ates]

Dear Carlo,

Thank you for your mail, I got the point.

Kind regards

[Tosin Brown]

Hi Guys, I'm experiencing the same problem. I'm trying to examine the relationship between corruption on economic growth. My independent variables are population growth , gross capital formation , gdp initial , corruption index ( from transparency international) , education but I'm using a proxy for secondary school enrolment. I tried running a regression using random effects. However , I did not get any results for Prob > chi2 and Wald chi2(3). I will attach my file down below. Please help! .


. xtreg GDPpercapitagrowth Populationgrowthannual Grosscapitalformationcurrent SES CorruptionPerceptionindex,re

Random-effects GLS regression Number of obs = 486
Group variable: countrynum Number of groups = 28

R-sq: Obs per group:
within = 0.0056 min = 4
between = 0.2208 avg = 17.4
overall = 0.0550 max = 21

Wald chi2(3) = .
corr(u_i, X) = 0 (assumed) Prob > chi2 = .

----------------------------------------------------------------------------------------------
GDPpercapitagrowth | Coef. Std. Err. z P>|z| [95% Conf. Interval]
-----------------------------+----------------------------------------------------------------
Populationgrowthannual | .0017312 .3224105 0.01 0.996 -.6301818 .6336442
Grosscapitalformationcurrent | -3.09e-13 4.35e-13 -0.71 0.477 -1.16e-12 5.44e-13
SES | -.0038753 .0123864 -0.31 0.754 -.0281523 .0204016
CorruptionPerceptionindex | -.0137191 .0150687 -0.91 0.363 -.0432533 .015815
_cons | 3.41364 1.104632 3.09 0.002 1.248602 5.578679
-----------------------------+----------------------------------------------------------------
sigma_u | 1.1726838
sigma_e | 2.6569838
rho | .16303829 (fraction of variance due to u_i)
----------------------------------------------------------------------------------------------
Also this message keeps popping up! I have been trying to fix this problem for hours!!


Note: the rank of the differenced variance matrix (3) does not equal the number of coefficients being tested (4); 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.