Oaxaca_rif command and jmpierce2

[Aleksandra Anic]

Dear all,

I use FFL methodology to decompose gender pay gap using recentered influence function beyond the mean (i.e. across distribuion). User-written command oaxaca_rif is written by Fernando Rios Avila. Is it possible to do decomposition by gender and time at the same time. I have two datasamples, for 2014 and 2018, and I see that both explained and unexplained part of the gap changed, I want to investigate what contributed to those changes between two years, something similar to user written command jmpierce2 by Ben Jann which uses JMP 1991 and 1993 methodology.

I have not seen so far papers combining OB-RIF by two groups and two samples.

Here are the referecnes:
Ben Jann, 2005. "JMPIERCE2: Stata module to compute trend decomposition of outcome differentials," Statistical Software Components S444804, Boston College Department of Economics, revised 27 Aug 2008.
Juhn, Chinhui, Kevin M. Murphy, Brooks Pierce (1991). Accounting for the Slowdown in Black-White Wage Convergence. Pp. 107-143 in: Workers and Their Wages, ed. by Marvin Kosters, Washington, DC: AEI Press.

Firpo, S. P., N. M. Fortin, and T. Lemieux. 2018. Decomposing wage distributions using recentered influence function regressions. Econometrics 6: 28. https://doi.org/10.3390/econometrics6020028.
Rios-Avila, F. 2020. Recentered influence functions (RIFs) in Stata: RIF regression and RIF decomposition. Stata Journal, 20(1), 51-94. https://doi.org/10.1177/1536867X20909690.


Thank you in advance.
Best regards,
Aleksandra

[FernandoRios]

Hi Aleksandra
Unfortunately, no that isnt possible. Oaxaca_rif is based on Oaxaca, which only allows for comparing 2 groups.
Perhaps something you could do, is using xtoaxaca (another user written command) which deals with time and groups decomposition. The difference will be for you to create the RIFS for the variable of interest.
Just make sure your sample has no missing data in dep or indep variables.
example:

egen time_gender = group(time gender)
** 2 groups across T periods

egen rif_gini = rifvar(y) , gini over(time_gender)

And then use rif_gini as your dep variable.

HTH