oaxaca rif

[Aleksandra Anic]

Dear all,

I have a question regarding user written command oaxaca_rif by Firpo, Sergio, Fortin, Nicole M. and Lemieux, Tomas (2018). Decomposing Wage Distributions Using Recentered Influence Function Regressions Econometrics, 6(2)28.

I cannot figure out if it is possible to see the specification for probit/logit used for reweighting (rwprobit ( ))? I know that I can run probit separately, but I am not sure if it will give me the same estimation results as it is inside oaxaca_rif command.

Thank you in advacne.

Best wishes,
Aleksandra

[Stephen Jenkins]

Let's get the acknowledgements right for a start. oaxaca_rif is by Fernando Rios-Avila, and part of larger package of cummunity-contributed commands by him:

SJ-20-1 st0588 . . . . . . . . Recentered influence functions (RIFs) in Stata
. . . . . . . . . . . . . . . . . . . . . . . . . . . . F. Rios-Avila
(help rifvar, rifhdreg, rifsureg2, oaxaca_rif, uqreg, hvar,
rifsureg if installed)
Q1/20 SJ 20(1):51--94
provides recentered influence functions (RIFs) that analyze
unconditional partial effects on quantiles in a regression
analysis framework (unconditional quantile regressions) to
facilitate the use of RIFs in the analysis of outcome
distributions: create RIFs for a large set of distributional
statistics, estimate RIF regressions enabling the use of
high-dimensional fixed effects, and provide Oaxaca-Blinder
decomposition analysis (RIF decompositions)

Have you read Fernando's article to see if it answers your question? (I recall that there is a discussion paper/preprint version available if you cannot access the SJ version.)

[FernandoRios]

Thank you Stephen!
also, Aleksandra, if you type "ereturn list" after oaxaca_rif, it should provide you with the coefficients and standard errors associated with the logit/probit model that you estimated.
you can also use the option, noisily, to display those results.

[Aleksandra Anic]

Dear Fernando,

Frist of all, I appologise for my error in the post regarding acknowledgements right. I get an error when adding noisly option, option noisly not allowed. Here is my command.

oaxaca_rif lyhw $demo $emp $occup y2-y4, by(female) rif(q(10)) wgt(0) rwprobit($demo $emp $occup $children y2-y4) noisly

I see now that there are coefficients in eretrun list for probit, e(b_probit) and e(V_probit), but I prefer regular output, since it is easily readable and there are also other information such as p values, pseudo R2, etc.

Could you please help me with noisly option.

Thank you very much.

Best,
Aleksandra

[Aleksandra Anic]

Thank you Stephen!
also, Aleksandra, if you type "ereturn list" after oaxaca_rif, it should provide you with the coefficients and standard errors associated with the logit/probit model that you estimated.
you can also use the option, noisily, to display those results.

Dear Fernando,

Frist of all, I appologise for my error in the post regarding acknowledgements right. I get an error when adding noisly option, option noisly not allowed. Here is my command.

oaxaca_rif lyhw $demo $emp $occup y2-y4, by(female) rif(q(10)) wgt(0) rwprobit($demo $emp $occup $children y2-y4) noisly

I see now that there are coefficients in eretrun list for probit, e(b_probit) and e(V_probit), but I prefer regular output, since it is easily readable and there are also other information such as p values, pseudo R2, etc.

Could you please help me with noisly option.

Thank you very much.

Best,
Aleksandra

[FernandoRios]

Hi Aleksandra,
That should work.
make sure you have the last version of "rif" installed in your computer.
Let me know if that doesnt fixed that issue.
Fernando

[Aleksandra Anic]

Hi Fernando,

I updated rif ado file, but it still does not work. The same error appears.

Best,
Aleksandra

[Stephen Jenkins]

Please show us the results of

which oaxaca_rif.ado, all

Long shot: you might have multiple versions on your computer, and the old version still being used

[FernandoRios]

the other option,
in your command line you typed "noisly", but it should be "noisily".

[Aleksandra Anic]

Dear Fernando and Stephen,

Thank you very much for your help. It was a typo. Now it works. I apologize for disturbing.

Thanks a lot.

Best wishes,
Aleksandra