When you do a Blinder-Oaxaca decomposition, you get a breakdown of the percent difference between two groups (say disabled and not disabled) that is attributable to the difference in their characteristics and the percent difference attributable to the difference in the coefficients.

I am doing this for a binary dependent variable. If I want to see what the impact of disability is on the dependent variable, why is this different than just estimating the model for the whole population and having disability as an independent variable? Wouldn't the odds ratio for disability give me the percent of the gap between disabled and non-disabled due to the coefficients in the Blinder-Oaxaca decomposition?


Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Browse other questions tagged or ask your own question.