[Alex Miller]

Carlo:

- I tried adding i.county but as you can see 3 of my county dummies get dropped and I don't know why...

. xtgls BEVsalesshare2 $varying $invariant i.county i.date, panels(hetero) corr(ar1)
note: 16.county omitted because of collinearity
note: 17.county omitted because of collinearity
note: 18.county omitted because of collinearity
note: 683.date omitted because of collinearity

Cross-sectional time-series FGLS regression

Coefficients:  generalized least squares
Panels:        heteroskedastic
Correlation:   common AR(1) coefficient for all panels  (0.5292)

Estimated covariances      =        18          Number of obs     =      1,296
Estimated autocorrelations =         1          Number of groups  =         18
Estimated coefficients     =        97          Time periods      =         72
                                                Wald chi2(96)     =    4355.63
                                                Prob > chi2       =     0.0000

-----------------------------------------------------------------------------------
   BEVsalesshare2 |      Coef.   Std. Err.      z    P>|z|     [95% Conf. Interval]
------------------+----------------------------------------------------------------
            HOVkm |   .0028243   .0008236     3.43   0.001       .00121    .0044386
            NTPKM |   1.044655   .1909987     5.47   0.000     .6703044    1.419006
            NFPKM |  -.0000351   .0001897    -0.18   0.853    -.0004069    .0003368
        lCHRoadKm |   .0005701   .0001764     3.23   0.001     .0002244    .0009158
          DGPrice |   1.099708   .4603553     2.39   0.017     .1974284    2.001988
      EnergyPrice |  -5.30e-06   .0000247    -0.21   0.830    -.0000537    .0000431
       Arbeitslos |   .0141821   .0031306     4.53   0.000     .0080463     .020318
             AVKT |   .0000247   .0000133     1.85   0.064    -1.41e-06    .0000508
        Einkommen |   7.54e-07   3.72e-07     2.03   0.043     2.47e-08    1.48e-06
        KuestenKm |  -.0000429   .0000545    -0.79   0.431    -.0001498    .0000639
         MuMShare |    .000518   .0015161     0.34   0.733    -.0024536    .0034896
             Temp |   .0837146   .0260164     3.22   0.001     .0327234    .1347059
                  |
           county |
      Aust-Agder  |   .0140983   .0395476     0.36   0.721    -.0634135    .0916101
        Buskerud  |   .1760323   .1127569     1.56   0.118    -.0449672    .3970318
        Finnmark  |   .6700852   .4240885     1.58   0.114     -.161113    1.501283
         Hedmark  |   .2401618   .1081435     2.22   0.026     .0282044    .4521192
       Hordaland  |   .3127543   .3366671     0.93   0.353    -.3471011    .9726097
 Møre og Romsdal  |   .2334566   .2968067     0.79   0.432    -.3482738     .815187
        Nordland  |   1.127708   1.181876     0.95   0.340    -1.188728    3.444143
         Oppland  |    .282466   .1378638     2.05   0.040      .012258    .5526741
            Oslo  |  -.2552385   .0953817    -2.68   0.007    -.4421831   -.0682939
        Rogaland  |  -.0416909   .1000655    -0.42   0.677    -.2378157    .1544339
Sogn og Fjordane  |   .1905085   .1986498     0.96   0.338     -.198838    .5798549
        Telemark  |   .1625748   .0780885     2.08   0.037     .0095242    .3156255
           Troms  |   .4430188   .2773593     1.60   0.110    -.1005955    .9866332
       Trøndelag  |   .6389879   .6288288     1.02   0.310    -.5934938     1.87147
      Vest-Agder  |          0  (omitted)
        Vestfold  |          0  (omitted)
         Østfold  |          0  (omitted)
                  |
             date |
             613  |  -.2268591   .0927098    -2.45   0.014    -.4085669   -.0451513
             614  |  -.7043166   .2928438    -2.41   0.016     -1.27828   -.1303532
             615  |  -.9169508   .3799831    -2.41   0.016    -1.661704   -.1721976
             616  |  -.7581243   .3127974    -2.42   0.015    -1.371196   -.1450527
             617  |  -.7775726   .3195716    -2.43   0.015    -1.403921   -.1512237
             618  |  -.7862758   .3209708    -2.45   0.014    -1.415367   -.1571845
             619  |  -.6270099   .2564727    -2.44   0.014    -1.129687   -.1243327
             620  |  -.5250393   .2154985    -2.44   0.015    -.9474087     -.10267
             621  |   -.527263   .2138185    -2.47   0.014    -.9463397   -.1081864
             622  |  -.5390297   .2244603    -2.40   0.016    -.9789638   -.0990955
             623  |   -.503813   .2058997    -2.45   0.014     -.907369   -.1002571
             624  |  -1.004212   .4080189    -2.46   0.014    -1.803914   -.2045092
             625  |  -1.231644   .5068825    -2.43   0.015    -2.225115   -.2381725
             626  |  -1.415411   .5805272    -2.44   0.015    -2.553224   -.2775987
             627  |  -1.691162   .6953116    -2.43   0.015    -3.053948   -.3283766
             628  |  -1.156301    .471875    -2.45   0.014    -2.081159   -.2314433
             629  |  -.7842419   .3155263    -2.49   0.013    -1.402662   -.1658216
             630  |  -.7195585   .2878142    -2.50   0.012    -1.283664    -.155453
             631  |  -1.075721   .4389351    -2.45   0.014    -1.936018   -.2154239
             632  |  -1.390086   .5753903    -2.42   0.016    -2.517831   -.2623422
             633  |  -1.125633   .4606353    -2.44   0.015    -2.028462   -.2228048
             634  |  -.7319151   .2952611    -2.48   0.013    -1.310616    -.153214
             635  |  -.6073547   .2428528    -2.50   0.012    -1.083338   -.1313719
             636  |  -.7969532   .3108218    -2.56   0.010    -1.406153   -.1877537
             637  |  -1.018591   .4054745    -2.51   0.012    -1.813306   -.2238753
             638  |  -.9531387   .3779906    -2.52   0.012    -1.693987   -.2122908
             639  |  -.6762004   .2634221    -2.57   0.010    -1.192498   -.1599026
             640  |  -.7782525   .3045332    -2.56   0.011    -1.375127   -.1813784
             641  |  -.8192766   .3225763    -2.54   0.011    -1.451515   -.1870387
             642  |  -1.158212   .4628405    -2.50   0.012    -2.065363   -.2510613
             643  |  -1.279468   .5228701    -2.45   0.014    -2.304274    -.254661
             644  |  -1.348491   .5573389    -2.42   0.016    -2.440855   -.2561263
             645  |  -1.210291   .4978799    -2.43   0.015    -2.186117   -.2344639
             646  |  -.9993405   .4216226    -2.37   0.018    -1.825706   -.1729753
             647  |  -1.049846   .4448822    -2.36   0.018    -1.921799   -.1778926
             648  |  -1.287348   .5263805    -2.45   0.014    -2.319035   -.2556609
             649  |  -1.198639   .4966977    -2.41   0.016    -2.172149   -.2251295
             650  |  -.9450781   .4137151    -2.28   0.022    -1.755945   -.1342113
             651  |  -.8831236   .3606346    -2.45   0.014    -1.589954   -.1762928
             652  |   -1.17132   .4822397    -2.43   0.015    -2.116493   -.2261478
             653  |  -1.035592   .4314637    -2.40   0.016    -1.881246   -.1899391
             654  |  -1.336239   .5584903    -2.39   0.017     -2.43086   -.2416178
             655  |   -1.22443   .5195855    -2.36   0.018    -2.242799   -.2060608
             656  |  -1.062353   .4391422    -2.42   0.016    -1.923056   -.2016503
             657  |  -.7360511   .3011323    -2.44   0.015     -1.32626   -.1458426
             658  |  -.7580307   .3142412    -2.41   0.016    -1.373932   -.1421293
             659  |  -.1000405   .0416224    -2.40   0.016    -.1816189   -.0184621
             660  |  -.0430431   .0259607    -1.66   0.097    -.0939252    .0078389
             661  |   .2606484   .1016282     2.56   0.010     .0614607     .459836
             662  |  -.0926133   .0665767    -1.39   0.164    -.2231013    .0378747
             663  |  -.4448154   .1882968    -2.36   0.018    -.8138704   -.0757604
             664  |   .0602728   .0270416     2.23   0.026     .0072723    .1132733
             665  |  -.6300406   .2730304    -2.31   0.021     -1.16517    -.094911
             666  |  -.7713944   .3188387    -2.42   0.016    -1.396307   -.1464819
             667  |   .1157233   .0462074     2.50   0.012     .0251584    .2062881
             668  |   .2200686   .0892745     2.47   0.014     .0450938    .3950434
             669  |   .3255079   .1350015     2.41   0.016     .0609099     .590106
             670  |   .3304627   .1340658     2.46   0.014     .0676986    .5932268
             671  |   .6034622   .2515191     2.40   0.016     .1104939    1.096431
             672  |   .6775598   .2859752     2.37   0.018     .1170587    1.238061
             673  |   1.231374   .5205094     2.37   0.018     .2111947    2.251554
             674  |   .4974773   .2005191     2.48   0.013     .1044672    .8904875
             675  |    .640154   .2762957     2.32   0.021     .0986243    1.181684
             676  |   .7074929   .3129547     2.26   0.024     .0941128    1.320873
             677  |   .0726837   .0374787     1.94   0.052    -.0007733    .1461407
             678  |   .0530418    .046525     1.14   0.254    -.0381455     .144229
             679  |   .4668917   .2002696     2.33   0.020     .0743705    .8594129
             680  |   .6216121   .2533763     2.45   0.014     .1250037    1.118221
             681  |    .727005   .3106338     2.34   0.019      .118174    1.335836
             682  |   .2855514    .110083     2.59   0.009     .0697926    .5013101
             683  |          0  (omitted)
                  |
            _cons |  -15.44177    5.96521    -2.59   0.010    -27.13336   -3.750171
-----------------------------------------------------------------------------------

. 
end of do-file

- If I run -xtgls- including -i.county.- and -i.date- is this model similar to a FE or RE model using -xtreg- and -i.date-?

- Additionally, how do I know if I should add -corr(ar1)- or -corr(psar1).-? Is there a way to test for this?

[Carlo Lazzaro]

Alex:
1) one of your county was omitted being the reference category; the remaining two were omitted due to collinearity with other predictors (and there's nothing you can do about that, but changing the specification in the right-hand side of your regression equation). That said, I would check what happen if you replace -i-date- with -c.date-.
2) I would say that your -xtgls- would be (more) similar to an FE vs RE model;
3) the rule of thumb is trying both and see whether a huge difference in standard errors comes alive; it this were the case, go -psar-. Otherwise, you can report both the regressions.

[Iuri Gavronski]

I also had problems with the -xtoverid- routine: "o. operator not allowed r(101);". What I believe fixed the error was a quick patch on xtoverid.ado, around line 230:

    qui di regexm("`inexog'", "(^.*)(o\..*)")
    local inexog = regexs(1)

I don't know how to update the source xtoverid.ado file in the ssc server. If someone reads this, please advise.

This is the location on the script where the change should be made:

[Carlo Lazzaro]

Iuri:
it's true that -xtoverid. user-written command does not support -fvvarlist- notation.
Looking forward to a future patch made by its Authors, in the meantime you can rely on -xi:- prefix to tame the beast (see -help xi:-).

[Iuri Gavronski]

Carlo,
Unfortunately, when you have variables omitted due to collinearity, they get the o. prefix, no matter if you use -xi- or not. Therefore, you get the message "o operator not allowed". The only way is to patch xtoverid.do removing the o. variables. However, the code I pasted earlier only works (I found later) if there is one and only one omitted variable.
The code below, inserted around line 230 in xtoverid.do, fixes the problem, accepting from zero to infinite omitted variables (due to collinearity):

    local hasomitted = regexm("`inexog'", "([A-Za-z0-9_ ]*)(o\..*)")
    if `hasomitted' > 0 local inexog = regexs(1)

Best regards,
Iuri.

[Carlo Lazzaro]

Iuri:
I see the issue now.
It's time for the Authors of -xtoverid- to roll out a new edition of this glorious user-written command.