Dear Stata Community,

I would like to hear your opinion on the results I get from the reverse causality tests I performed (via the commands pvar & pvargranger). I am using Stata16. My panel dataset contains 17 countries (with t=6, 1995 - 2020), and a total of 102 observations. Here is the data of the three variables I want to perform a reverse causality test (lngdp , as my DV, and z_exp_diss and z_rd as my IVs) :

Code:
* Example generated by -dataex-. For more info, type help dataex
clear
input float(ln_gdppc z_exp_diss z_rd z_labourcost)
 7.722362   1.3693258  -1.0572331   -3.1591444
 7.390948   1.3693258   -.8968285   -1.1011491
 8.268708   1.0252694  -1.0343182      -.70354
 8.826501    .5967307   -.7364238     .4561532
 8.861625    .3057935   .13434455     1.313958
 9.193006  -.03913696  -.11771997    2.2822099
8.4923935    .6811494   .13434455   -1.3257246
 8.484998    .6758959   .13434455     -.423741
 9.261784   .24520718 -.071890056   -.07031069
 9.543186   .08851427   -.3468695     .4561532
 9.374306  -.15563676  -.11771997    .18739893
 9.611536   -.3543932   .50098383    .24630398
 9.633082   1.6185626   -1.584277     -.788216
 9.574174    1.242349  -1.5155323    -.5673221
   10.125    .7579433   -1.194723    .02172846
10.342505     .644174  -1.0343182     .4561532
10.057654   1.7487774   -.9426584    .18739893
 10.23489   1.7039355   -.5989341     .2757565
 8.669764   -.9285728 -.026060145   -1.2152777
 8.704343   -1.470508   .50098383    -.3942885
 9.505296   -1.584609    .6155586    .11744916
 9.901489  -1.3211894   1.0051129     .4561532
 9.788621   -1.395671   2.3570952     .6255053
10.064526  -1.2370006    2.402925    1.3176396
8.0490465    .7375261   -.7364238    -1.796965
 8.312864   1.1607805   -.6676789    -1.152691
 9.250176      .45456   .06559968    -.6777691
 9.601762    .2032704    1.555072     .4561532
 9.771226    .2759098   1.3030072      1.24769
10.074213 -.004973041   1.6467316    2.0981317
 9.469571   1.4492502  -1.0801481   -1.2410486
 9.396235   1.1072044   -.7593387    -.6888137
10.023567     .699696   -.7135088 -.0040425034
10.200541    .9931969   -.6676789     .4561532
 9.807405   1.1554782   .18017446    .13217543
 9.882394   1.0336165    .8676231     .1910805
 8.410656  -.06580974   -.4156144   -2.0362668
 8.439076   -.0849603  -.23229474    -.9097077
 9.323721   -.3819059   .06559968   .014365326
 9.487337   -.4923444    .5468137     .4561532
   9.4499   -1.281711   1.0280278     .7838126
 9.725068  -1.3347082    1.348837    1.4133604
 9.936175   -1.348037   .08851463    -.6740875
 9.907858  -1.1452198   .24891932    -.4568751
 10.37484  -1.1431822   .34057915    .02172846
10.491288  -1.1320345    .7530484     .4561532
10.316598  -1.0410514   1.0280278     .6070974
10.411156   -.1259437   1.2800924     .7433153
 7.752862    .9413027  -1.0572331    -1.679155
 8.118978   1.6185786  -1.0343182   -1.1858251
 8.932632   1.0744585   -.8280836    -.6740875
 9.339922    .4698162   -.6447639     .4561532
 9.530582    .6044782    -.621849    1.1851033
 9.788576   .54994786  -.57601905    2.1680813
 7.681465    .7674782  -1.0572331    -1.263138
 8.099624    1.209263   -.6905938    -.3795622
 8.968876    .8577559   -.3239546   -.09240008
 9.391884    .7559402   -.2552097     .4561532
 9.565089   .51060766   .34057915     .7838126
 9.883381    .1518277   .24891932     2.002411
 9.195986   1.7571434  -1.4697024    -.3169756
 9.252665   2.0901248  -1.4697024    -.3169756
  9.67333   1.0573235   -.8280836     .1027229
 9.989627   1.4017817   -.6905938     .4561532
 10.12349      1.2024   -.3926994      .780131
10.302955    .8918662   -.6905938    1.4354497
  8.21252   .29601064    -.621849    -1.288909
 8.412156   -.5381761  -.57601905   .018046891
8.9898815  -1.0286179   -.7593387   .007002194
 9.442484  -1.1793464   -.3926994     .4561532
 9.439744   -1.034704   .24891932     .5997343
 9.660938  -1.0364012    .9821979     1.008388
 9.374274    .2080747   -.8509985    -.7256294
 9.349907   -.3067259   -.3926994    -.2102102
 9.840181   -.5954473   -.3010396    .24630398
10.021213  -.16994613    1.486327     .4561532
  9.86487  -.12019895    .7988783    .27207494
 10.05415   -.4888581   1.1655176    .55923706
 7.408698   1.1293633   -.3010396   -3.1333735
 7.414517   1.1625406   -1.194723   -2.0878088
 8.437701    .7149008  -1.1030631    -.7072216
 9.013605   -.4342212   -.9884883     .4561532
 9.101546     -.68262   -.9197434     .3162537
9.4664955  -1.0684882   -.9426584    1.6416174
 8.480348   -.1563294  .019769765   -1.2263223
 8.596586   -.7502491  -.57601905   -.39797005
 9.366126  -1.1660045   -.9197434    .09904134
 9.726201   -.6055823   -.6447639     .4561532
 9.699594   -.6734886    .6155586     .6181421
 9.866112  -.54155046  -.14063492    1.2845055
 9.280841   -.9585403    1.371752    -1.425127
  9.23027   -.9140266   1.0738577    -.8066239
 9.803607  -1.2356193   1.2113475    -.1549867
10.065162   -1.439919     2.65499     .4561532
 9.946631  -1.4153608    2.998714     .4451085
 10.16378  -1.3464882   2.6320746     .8206282
 9.646785   -1.071029  -.27812466    -.7366741
 9.596491   -1.164435 -.026060145    -.4090147
 10.18185   -1.296438    .4780689   .003320628
 10.32557  -1.1727965   1.0738577     .4561532
end


Now, before doing the reverse causality test, I run a fixed effects model and a pooled OLS model. (The command for the fe model was:

Code:
 xtreg ln_gdppc z_exp_diss z_rd z_labourcost z_invest z_hc,  fe cluster(country)
In the models my independent variable z_exp_diss (which measures the export structure of a country) was positively correlated and signficant with GDP p.c. (ln_gdppc). The results of the tests (pvar and pvargranger) suggest me however, that there is no causal effect of z_exp_diss on ln_gdppc. (Before running pvar, I specfied xtsset country year, delta (5) ) Is there something that I maybe did wrong? Are there any other tests that I could perform? Here are the commands:

Code:
pvar ln_gdppc z_exp_diss z_rd
 pvargranger


Code:
Panel vector autoregresssion



GMM Estimation

Final GMM Criterion Q(b) =  8.91e-33
Initial weight matrix: Identity
GMM weight matrix:     Robust
                                                   No. of obs      =        68
                                                   No. of panels   =        17
                                                   Ave. no. of T   =     4.000


------------------------------------------------------------------------------
             |      Coef.   Std. Err.      z    P>|z|     [95% Conf. Interval]
-------------+----------------------------------------------------------------
ln_gdppc     |
    ln_gdppc |
         L1. |   .0597939   .8042858     0.07   0.941    -1.516577    1.636165
             |
  z_exp_diss |
         L1. |  -.9220663   .8075951    -1.14   0.254    -2.504924     .660791
             |
        z_rd |
         L1. |   .2887812   .3659468     0.79   0.430    -.4284613    1.006024
-------------+----------------------------------------------------------------
z_exp_diss   |
    ln_gdppc |
         L1. |   .0740402    .769195     0.10   0.923    -1.433554    1.581635
             |
  z_exp_diss |
         L1. |   1.193323   .7442247     1.60   0.109    -.2653302    2.651977
             |
        z_rd |
         L1. |    .161285   .3361554     0.48   0.631    -.4975675    .8201375
-------------+----------------------------------------------------------------
z_rd         |
    ln_gdppc |
         L1. |  -.0983861   1.350602    -0.07   0.942    -2.745517    2.548745
             |
  z_exp_diss |
         L1. |  -.3242725   1.424901    -0.23   0.820    -3.117026    2.468481
             |
        z_rd |
         L1. |   .9076021   .6355367     1.43   0.153    -.3380269    2.153231
------------------------------------------------------------------------------
Instruments : l(1/1).(ln_gdppc z_exp_diss z_rd)


Code:
pvargranger

  panel VAR-Granger causality Wald test
    Ho: Excluded variable does not Granger-cause Equation variable
    Ha: Excluded variable Granger-causes Equation variable

  +------------------------------------------------------+
  |  Equation \ Excluded |    chi2     df   Prob > chi2  |
  |----------------------+-------------------------------|
  |ln_gdppc              |                               |
  |           z_exp_diss |      1.304    1        0.254  |
  |                 z_rd |      0.623    1        0.430  |
  |                  ALL |      1.688    2        0.430  |
  |----------------------+-------------------------------|
  |z_exp_diss            |                               |
  |             ln_gdppc |      0.009    1        0.923  |
  |                 z_rd |      0.230    1        0.631  |
  |                  ALL |      2.879    2        0.237  |
  |----------------------+-------------------------------|
  |z_rd                  |                               |
  |             ln_gdppc |      0.005    1        0.942  |
  |           z_exp_diss |      0.052    1        0.820  |
  |                  ALL |      0.770    2        0.681  |
  +------------------------------------------------------+


Maybe somebody can help me with the interpretation. Thank you so much !



Warm regards


Valentina