only the estimated coefficient

[Daniel Gomez]

Hi,

Which could be the reason that Stata shows me only the estimated coefficient?:


------------------------------------------------------------------------------
| Robust
ln_TRADE | Coef. Std. Err. t P>|t| [95% Conf. Interval]
-------------+----------------------------------------------------------------

ln_DIST | -.7334557 . . . . .
CNTG | 1.421494 . . . . .
LANG | -3.674322 . . . . .
CLNY | 6.189377 . . . . .
_cons | 17.52462 . . . . .
--------------------------------------------------------------------------------------


Thanks!

[Maarten Buis]

I don't understand that question. If I try to parse your sentence literally it means Stata shows you more than the estimated coefficients (there is a double negation in that sentence), but I am not sure that is what you mean. You did not tell us what the command is that you used. You did not tell us what you expected.

[Carlo Lazzaro]

Daniel:
the trivial explanation rests on the assumption that your data do not vary:

. set obs 10
number of observations (_N) was 0, now 10

. g A=10

. reg A

      Source |       SS           df       MS      Number of obs   =        10
-------------+----------------------------------   F(0, 9)         =      0.00
       Model |           0         0           .   Prob > F        =         .
    Residual |           0         9           0   R-squared       =         .
-------------+----------------------------------   Adj R-squared   =         .
       Total |           0         9           0   Root MSE        =         0

------------------------------------------------------------------------------
           A |      Coef.   Std. Err.      t    P>|t|     [95% Conf. Interval]
-------------+----------------------------------------------------------------
       _cons |         10          .        .       .            .           .
------------------------------------------------------------------------------

. reg A, r

Linear regression                               Number of obs     =         10
                                                F(0, 9)           =       0.00
                                                Prob > F          =          .
                                                R-squared         =          .
                                                Root MSE          =          0

------------------------------------------------------------------------------
             |               Robust
           A |      Coef.   Std. Err.      t    P>|t|     [95% Conf. Interval]
-------------+----------------------------------------------------------------
       _cons |         10          .        .       .            .           .
------------------------------------------------------------------------------

.

If that were not the case, please post what you typed and what Stata gave you back (via CODE delimiters, please).
You could also share with interested listers an example/excerpt of your data via -dataex-.

[Daniel Gomez]

Thanks, I just edited the question...

I am using this command: regress ln_TRADE EXPORTER_TIME_FE* IMPORTER_TIME_FE* ln_DIST CNTG LANG CLNY RTA , cluster(pairid)

It is a gravity equation with fixed effects: exporter-time and importer-time fixed effects.
I would like to know p -values.

[Carlo Lazzaro]

Daniel:
is the issue still alive when you use default standard errors vs clustered ones?

[Daniel Gomez]

Carlo Lazzaro , Yes. with standard errors is the same. Stata only reports the coefficients.

[Carlo Lazzaro]

Daniel:
could you please share an example/excerpt of your data via -dataex-?

[Daniel Gomez]

I am trying to estimate a gravity equation with exporter-time and importer-time fixed effects using panel data, but I would like to focus only on one country (country AAA).
The model works when I consider all the countries.
I am not sure if I am creating well the fixed effects:

__________________________________________________ __________________________________________________ _____________________________________

* Create exporter-time fixed effects
egen exp_time = group(exporter year)
quietly tabulate exp_time, gen(EXPORTER_TIME_FE)

* Create importer-time fixed effects
egen imp_time = group(importer year)
quietly tabulate imp_time, gen(IMPORTER_TIME_FE)


regress ln_TRADE EXPORTER_TIME_FE* IMPORTER_TIME_FE* ln_DIST CNTG LANG CLNY RTA if exporter == "AAA" | importer == "AAA", cluster(pairid)

[Daniel Gomez]

* Example generated by -dataex-. To install: ssc install dataex
clear
input str3(exporter importer) int year double(TRADE CNTG LANG CLNY RTA pairid)
"AFG" "AGO" 1991 0 0 0 0 0 170
"AFG" "AGO" 1995 0 0 0 0 0 170
"AFG" "AGO" 1999 0 0 0 0 0 170
"AFG" "AGO" 2003 0 0 0 0 0 170
"AFG" "AGO" 2007 176.303 0 0 0 0 170
"AFG" "AGO" 2011 280.397 0 0 0 0 170
"AFG" "AGO" 2015 305.721 0 0 0 0 170
"AFG" "ALB" 1991 0 0 0 0 0 339
"AFG" "ALB" 1995 0 0 0 0 0 339
"AFG" "ALB" 1999 0 0 0 0 0 339
"AFG" "ALB" 2003 0 0 0 0 0 339
"AFG" "ALB" 2007 0 0 0 0 0 339
"AFG" "ALB" 2011 0 0 0 0 0 339
"AFG" "ALB" 2015 6.941 0 0 0 0 339
"AFG" "ARG" 1991 0 0 0 0 0 508
"AFG" "ARG" 1995 0 0 0 0 0 508
"AFG" "ARG" 1999 0 0 0 0 0 508
"AFG" "ARG" 2003 137.218 0 0 0 0 508
"AFG" "ARG" 2007 266.024 0 0 0 0 508
"AFG" "ARG" 2011 8.311 0 0 0 0 508
"AFG" "ARG" 2015 6.5 0 0 0 0 508
"AFG" "ARM" 1991 0 0 0 0 0 677
"AFG" "ARM" 1995 0 0 0 0 0 677
"AFG" "ARM" 1999 0 0 0 0 0 677
"AFG" "ARM" 2003 0 0 0 0 0 677
"AFG" "ARM" 2007 0 0 0 0 0 677
"AFG" "ARM" 2011 2.844 0 0 0 0 677
"AFG" "ARM" 2015 .016 0 0 0 0 677
"AFG" "ATG" 1991 0 0 0 0 0 846
"AFG" "ATG" 1995 0 0 0 0 0 846
"AFG" "ATG" 1999 30.529 0 0 0 0 846
"AFG" "ATG" 2003 0 0 0 0 0 846
"AFG" "ATG" 2007 .044 0 0 0 0 846
"AFG" "ATG" 2011 .927 0 0 0 0 846
"AFG" "ATG" 2015 0 0 0 0 0 846
"AFG" "AUS" 1991 214.328 0 0 0 0 1015
"AFG" "AUS" 1995 41.639 0 0 0 0 1015
"AFG" "AUS" 1999 33.711 0 0 0 0 1015
"AFG" "AUS" 2003 103.291 0 0 0 0 1015
"AFG" "AUS" 2007 110.605 0 0 0 0 1015
"AFG" "AUS" 2011 755.399 0 0 0 0 1015
"AFG" "AUS" 2015 1044.682 0 0 0 0 1015
"AFG" "AUT" 1991 1541.949 0 0 0 0 1184
"AFG" "AUT" 1995 421.421 0 0 0 0 1184
"AFG" "AUT" 1999 126.268 0 0 0 0 1184
"AFG" "AUT" 2003 398.829 0 0 0 0 1184
"AFG" "AUT" 2007 253.302 0 0 0 0 1184
"AFG" "AUT" 2011 149.792 0 0 0 0 1184
"AFG" "AUT" 2015 211.206 0 0 0 0 1184

[Phil Bromiley]

First, if you want fixed effects, xtset your data and use xtreg.
Second, when you use importer/time and exporter/time fixed effects, you've created so many fixed effects that you simply have too many parameters. What you do is

xtset Exporter time
xtreg y x i.time, fe

Adding fixed effects for each Importer is also possible using reghdfe. But interacting time with Exporter and Importer and then using them as fixed effects creates far too many fixed effects.