I'm try to decompose inequality/dispersion ratios (top 10% to bottom 10% or top 20% to bottom 20%) according to subgroups. I would like to say something like e.g. males are responsible for x% and females y% of the inequality as measured by e.g. top to bottom decile. I know GINI and Theil indices can be decomposed according to between and within group inequalities. Of course, dispersion ratios ignore the within component.

The decile ratio, R, could be described thus: R = (xa + xb + .. xalpha + ya + yb + ... ybeta)/(x1 + x2 + ..xn + y1 + y2 + ....ym) ;

where xa-to-xalpha, ya-to-ybeta are men and women in the top decile and x1-to-xn, y1-to-ym are men and women in the bottom decile

R = (xa + xb + ..xalpha)/(x1 + ..ym) + (ya + yb + ..ym)/(x1 + ..ym)

That is, I can decompose according to the contribution to men and women who make it into the top decile, but I cannot decompose the the overall contribution of of males and females. Can this be done? Thanks

  • $\begingroup$ I have read that dispersion ratios cannot be decomposed: scholar.google.com/… $\endgroup$
    – R.S.
    Dec 6, 2017 at 17:30

1 Answer 1


Dispersion ratios cannot be meaningfully decomposed: https://scholar.google.com/scholar?biw=1440&bih=769&um=1&ie=UTF-8&lr&cites=13319421342172587982


Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge that you have read and understand our privacy policy and code of conduct.

Not the answer you're looking for? Browse other questions tagged or ask your own question.