Path independent inequality measures

http://www.sciencedirect.com/science?_ob=ArticleURL&_udi=B6WJ3-45FCBSK-23&_user=984616&_rdoc=1&_fmt=&_orig=search&_sort=d&view=c&_acct=C000049824&_version=1&_urlVersion=0&_userid=984616&md5=5bd07d5825126723d2a79f17691c76bb
http://dx.doi.org/10.1006%2Fjeth.1999.2565

This paper explores a natural decomposition requirement for inequality measures, motivated by Shorrocks and Anand, which we call path independent decomposability. Given a collection of population subgroups, overall inequality is to be decomposed into between-group and within-group terms based on a notion of a representative income of a population (such as the mean income). Between-group inequality is the measure applied to the smoothed distribution, which replaces each income in a subgroup with its representative income. Within-group inequality is the measure applied to the standardized distribution which re-scales subgroup distributions to a common representative income level. Path independence then requires overall inequality to be the sum of the within and between-group terms.

Author(s): Foster, James E., Shneyerov, Artyom A. Originator(s): Vanderbilt University. Department of Economics
Resource added in: 27/12/2000
Objective: This paper derives the entire class of (relative) inequality measures satisfying path independent decomposability -- a single parameter family containing both the "second." Theil measure (also known as mean logarithmic deviation) and the variance of logarithms.
Available languages: English
Equity, 35144, Income, Measuring, Methods and Indicators
  • Share
  • Suggest tag
    [X]
    Suggestions
    Separated by comma
    [X]
    Thank you for your suggestion.
    Communication problem. Please try again later.
  • Report error
    [X]
    Reason
    New Link (Optional)
    [X]
    Thank you for your report.
    Communication problem. Please try again later.
comments powered by Disqus

More related