Abstract
This article re-examines standard methodology used to eliminate the impact of changes in age composition and age-income profiles on trends in income inequality. Attempts at such adjustment rely primarily on the approach developed in Paglin (1975) or on subsequent refinements of the Paglin measure (Formby et al., 1989). The present study identifies fundamental shortcomings in the approach and, hence, in associated estimates. To illuminate the degree of imprecision associated with the Paglin-type measure, the paper employs an alternative age-adjustment procedure based on an Entropy measure of dispersion discussed in Shorrocks (1980) and Mookherjee and Shorrocks (1982). The Entropy measure overcomes the problems with Paglin's approach and allows age-related changes to be accurately identified. Comparing age-adjusted inequality changes between 1980 and 1986 calculated in Formby et al. (1989), which are based on the Paglin-type measure, with those based on the Entropy measure reveals that the former approach overstates the importance of age by over 60 percent.
Get full access to this article
View all access options for this article.