How to aggregate structure and composition effect of all categorical variables when using RIF composition

[Jie Deng]

Hi all,
When I run oaxaca_rif to decompose the determinants of wage inequality, following by Rios-Avila(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.), and(1319457, 1319457, Recentered Influence Functions in Stata: Methods for Analyzing the Determinants of Poverty and Inequality (April 26, 2019). Levy Economics Institute, Working Paper 927 (2019), Available at SSRN: https://ssrn.com/abstract=3378811 or http://dx.doi.org/10.2139/ssrn.3378811) I find that I don't know how to aggregate the effect of the categorical variables like following picture.

My code is: oaxaca_rif lnwage female marriage age working_type educ_* ind_* occ_* own_* , by(year23) rif(gini) wgt(1) relax robust . Where the educ_* ind_* occ_* own_* are all dummy variables that I create for this.

[FernandoRios]

Look into the helpfile of oaxaca_rif and the one from Oaxaca.
it’s a matter of using parenthesis and give a name to all variables to be aggregated
hth

[Jie Deng]

Thank you so much Fernandorios, that helps me a lot. Can I ask another two question about RIF regression and composition.
1. Categorical variables in composition, if by year, there is a significant effect in composition effect in industry A. How can I explain this by year and omitted group? I find it really difficult.
2. The difference between RIF over(year) and oaxaca_rif , by(year), If I want to know what factor contribute to the gini index change between two year , can I do both of them?