Gender pay gap - decomposition and OLS regression

[Natalie Becker]

Dear Stata users,

I want to regress firm-level adjusted gender pay gaps derived for employees working in non-managerial positions using Ñopo decomposition against a set of firm characteristics, including the share of women workers and women managers.

I used "nopomatch" command to do Nopo decomposition but I do not know how to save the results of the firm-level adjusted gender pay gap to use it in the regression. I would appreciate every suggestion.

Thank you!

Natalie

[Eric Makela]

Hi Natalie,

So from reading the documentation of 'nopomatch', it seems the command creates variables that would seemlingly be equal across each group created through the 'reportby' option. Am I understanding correctly, that your data are a cross-section of companies, with each company represented in a row?

[Natalie Becker]

Hi Eric,

Thank you for responding. I use a cross-section and matched employer-employee dataset. There are 1,763 companies and ~30,000 employees in the dataset.

The 'nopomatch' command creates two variables '_supp' and '_match' which are not raw gender pay gap. I used 'reportby' option in the 'nopomatch' to decompose the gender pay gap on the firm level but these results are not saved as variables.

I would like to use a raw gender pay gap from this decomposition as a dependent variable in the OLS regression but I do not know how to save it as a variable.