Estimation of gravity model of trade using PPML, fixed effects

[homa haddad]

Dear all,

I want to estimate the gravity model using the ppml. To control for the MRT I have set exporter and importer time varying fixed effects using this command:

tab gdp_exp, gen(exporteryear_)
tab gdp_imp, gen(importeryear_)

For the PPML command I have:

ppml import loggdpimp loggdpexp logpimp logpexp logdist Dsanc Dlan Dbor exporteryear_* importeryear_*

Unfortunately I get no results using this command on stata. It takes forever and Stata seems to run the regression, the only thing I get is this note:

note: checking the existence of the estimates

My question is, are the exporter and importer dummies correct? Do I have an alternative to control for MRTs using another command?

[Joao Santos Silva]

Homa,

Try the command ppml_panle_sg; it is much faster for this kind of application.

Best wishes,

Joao

[homa haddad]

Dear Joao Santos Silva ,

thank you for your response. I ran the command:

ppml_panel_sg import loggdpimp loggdpexp logpimp logpexp logdist Dsanc Dlan Dbor, ex(exporter) im(importer)y(year) nopair

and got this error:

insufficient observations
r(2001);

Here is the sum of my data:

. sum

    Variable |        Obs        Mean    Std. Dev.       Min        Max
-------------+---------------------------------------------------------
        year |      9,512        2002     8.36704       1988       2016
    country1 |          0
     iso_imp |          0
     iso_exp |          0
      import |      9,512    1.80e+08    1.12e+09        1.1   3.03e+10
-------------+---------------------------------------------------------
     gdp_imp |      9,512    2.59e+11    8.11e+11   1080.561   1.86e+13
     gdp_exp |      9,512    2.60e+11    8.11e+11        107   1.86e+13
       p_imp |      9,512    5.23e+07    9.60e+07      68755   1.38e+09
       p_exp |      9,512    5.23e+07    9.60e+07      68755   1.38e+09
        dist |      9,512    5961.116    3963.541    540.446   17141.34
-------------+---------------------------------------------------------
         lan |        116           1           0          1          1
         bor |        406           1           0          1          1
         san |      1,795    .2674095    .4427311          0          1
   loggdpimp |      9,512    24.90255    2.124178   6.985236    30.5555
   loggdpexp |      9,512    24.91128    2.172856   4.672829    30.5555
-------------+---------------------------------------------------------
     logpimp |      9,512    16.96712     1.61164    11.1383   21.04438
     logpexp |      9,512    16.96712     1.61164    11.1383   21.04438
     logdist |      9,512    8.448982     .744238   6.292395   9.749249
       Dsanc |      9,512    .0504626     .218909          0          1
        Dlan |      9,512    .0121951    .1097619          0          1
-------------+---------------------------------------------------------
        Dbor |      9,512    .0426829    .2021519          0          1
     country |      9,512       164.5    94.68998          1        328
    importer |      9,512    76.53963    34.40382          1        165
    exporter |      9,512    76.53963    34.40382          1        165

I don't understand where the error comes from. What do you think sir?

[Joao Santos Silva]

Dear Homa,

I suggest you contact the authors of the command.

Best wishes,

Joao

[homa haddad]

Ok, I'll do it. Thank you Joao Santos Silva . Best, Homa

[homa haddad]

Dear Tom Zylkin,

I do not have much experience using ppml_panle_sg. Would you please take a look at my data and the problem I'm facing? It would be awesome! Thank you in advance.

[homa haddad]

Dear Joao Santos Silva,

I applied some tricks and could run the ppml command with importer and exporter dummies! Here is the command:

ppml import loggdpimp loggdpexp logpimp logpexp logdist Dsanc Dlan Dbor Dimp* Dexp* year_*

And here is the result: ( only the coefficients of interest for simplicity's sake)

Number of parameters: 361
Number of observations: 9512
Number of observations dropped: 0
Pseudo log-likelihood: -2.633e+11
R-squared: .928788
------------------------------------------------------------------------------
             |               Robust
      import |      Coef.   Std. Err.      z    P>|z|     [95% Conf. Interval]
-------------+----------------------------------------------------------------
   loggdpimp |   1.132046   .0883858    12.81   0.000     .9588129    1.305279
   loggdpexp |   1.231915   .0704439    17.49   0.000     1.093848    1.369983
     logpimp |    .578808   .1488387     3.89   0.000     .2870896    .8705264
     logpexp |   .3734077   .2619459     1.43   0.154    -.1399968    .8868122
     logdist |  -1.089109   .2862648    -3.80   0.000    -1.650178   -.5280402
       Dsanc |  -.4244898   .0745165    -5.70   0.000    -.5705394   -.2784402
        Dlan |   1.591854   .3413182     4.66   0.000     .9228828    2.260825
        Dbor |   -.904043   .3603001    -2.51   0.012    -1.610218   -.1978678
.
.
.
.
 year_29 |   .0327447    .131005     0.25   0.803    -.2240205    .2895098
       _cons |  -49.51446   5.858231    -8.45   0.000    -60.99638   -38.03254
------------------------------------------------------------------------------
Number of regressors dropped to ensure that the estimates exist: 7
Dropped variables:  Dimp9 Dimp59 Dimp148 Dexp1 Dexp70 Dexp83 year_25
Option strict is off

The coefficients have a higher level than coefficients resulting from the OLS regression. Since in your paper "the log of gravity" the PPML estimation got smaller coefficients than the OLS, should I be concerned about my results? Thank you for your help.

[Tom Zylkin]

Dear Tom Zylkin,

I do not have much experience using ppml_panle_sg. Would you please take a look at my data and the problem I'm facing? It would be awesome! Thank you in advance.

Dear Homa,
I'd be happy to take a look a closer look at your data to see what is causing the problem with ppml_panel_sg. It looks like you were able to achieve what you were going for with ppml, but please feel free to email me if you ever need to use ppml_panel_sg for your research.
Regards,
Tom

[Joao Santos Silva]

Dear homa haddad,

I believe you are including in your model importer and exporter characteristics that should not be included when you have importer and exporter dummies. Using ppml_panel_sg would help with that. So, maybe you should follow up on Tom Zylkin's kind offer.

Best wishes,

Joao

[homa haddad]

DearJoao Santos Silva ,

I followed your suggestion and discussed the issue I have with Tom Zylkin. I got really helpful advices from him and used the poi2hdfe command, as my dataset was not square. Here is the result:

poi2hdfe import1 loggdpimp loggdpexp logpimp logpexp logdist Dsanc Dlan Dbor year_*, id1(exp) id2(imp)

Dropping exp groups for which import1 is always zeros

Total Number of observations used in the regression -> 9367

Starting Estimation of coefficients

1 dif is -> 263.87923
2 dif is -> .88462613
3 dif is -> .78570317
4 dif is -> .47922415
5 dif is -> .22449422
6 dif is -> .09474546
7 dif is -> .05612353
8 dif is -> .03423796
9 dif is -> .01788093
10 dif is -> .00736692
11 dif is -> .00281278
12 dif is -> .0007891
13 dif is -> .00012874
14 dif is -> .00003796
15 dif is -> .00001659
16 dif is -> 4.337e-06
17 dif is -> 3.080e-07
18 dif is -> 1.542e-09

Coefficients converged after 18 reghdfe calls


 ******* Poisson Regression with Two High-Dimensional Fixed Effects **********

                                                Number of obs     =      9,367
------------------------------------------------------------------------------
             |               Robust
     import1 |      Coef.   Std. Err.      z    P>|z|     [95% Conf. Interval]
-------------+----------------------------------------------------------------
   loggdpimp |   1.132046   .0885372    12.79   0.000     .9585161    1.305576
   loggdpexp |   1.231915   .0705646    17.46   0.000     1.093611    1.370219
     logpimp |    .578808   .1490937     3.88   0.000     .2865898    .8710263
     logpexp |   .3734077   .2623947     1.42   0.155    -.1408765    .8876919
     logdist |          0  (omitted)
       Dsanc |  -.4244898   .0746442    -5.69   0.000    -.5707897     -.27819
        Dlan |          0  (omitted)
        Dbor |          0  (omitted)
      year_1 |    1.59724    .265999     6.00   0.000     1.075891    2.118588
      year_2 |   1.701557   .2586135     6.58   0.000     1.194684     2.20843
      year_3 |   2.152632   .2441069     8.82   0.000     1.674192    2.631073
      year_4 |   1.421192   .2058134     6.91   0.000     1.017805    1.824579
      year_5 |   3.419411   .2805368    12.19   0.000     2.869569    3.969253
      year_6 |   2.893661    .262118    11.04   0.000     2.379919    3.407403
      year_7 |   2.405252   .2452033     9.81   0.000     1.924662    2.885841
      year_8 |   1.846082   .2246121     8.22   0.000      1.40585    2.286313
      year_9 |   1.710722   .2149643     7.96   0.000     1.289399    2.132044
     year_10 |   1.730028   .2098566     8.24   0.000     1.318717     2.14134
     year_11 |    1.70484   .2138328     7.97   0.000     1.285735    2.123944
     year_12 |   1.434204   .2211158     6.49   0.000     1.000825    1.867583
     year_13 |   1.809136   .2048561     8.83   0.000     1.407626    2.210647
     year_14 |    1.73383   .1990534     8.71   0.000     1.343692    2.123967
     year_15 |   1.728628   .1991251     8.68   0.000      1.33835    2.118906
     year_16 |   1.608781   .1768353     9.10   0.000      1.26219    1.955371
     year_17 |   1.428643   .1627806     8.78   0.000     1.109599    1.747687
     year_18 |   1.261635    .149493     8.44   0.000     .9686343    1.554636
     year_19 |   1.122973   .1241564     9.04   0.000     .8796311    1.366315
     year_20 |   .8207016   .1059854     7.74   0.000     .6129739    1.028429
     year_21 |   .7202783   .1085713     6.63   0.000     .5074824    .9330742
     year_22 |   .6247063   .0959396     6.51   0.000     .4366681    .8127445
     year_23 |   .4285563   .0886198     4.84   0.000     .2548647    .6022478
     year_24 |   .1697519    .081832     2.07   0.038     .0093643    .3301396
     year_25 |  -.0448062   .0988734    -0.45   0.650    -.2385945     .148982
     year_26 |          0  (omitted)
     year_27 |   .1756412   .0947305     1.85   0.064    -.0100273    .3613097
     year_28 |    .062862   .1126554     0.56   0.577    -.1579385    .2836625
     year_29 |  -.0120616   .1212996    -0.10   0.921    -.2498043    .2256812

I have two questions. Do you know how I can obtain the R-squared of the regression? I couldn't find any informations regarding to this.
And my second question is: how could I predict the potential trade?

I have only this part, and missing the rest:

 predict poi2hdfe , xb
gen potential_trade = exp(poi2hdfe)

I would be extremely grateful for your response.

Regards, Homa

[Joao Santos Silva]

Dear homa haddad,

I suggest you contact the author of poi2hdfe about these issues.

Best wishes,

Joao