[PEN-L:9487] Re: Measurement of inequality

[Rosser Jr, John Barkley]

     Two widely used and reported measures are the quintile 
and decile ratios.  They are respectively the ratio of what 
the top fifth or tenth of the distribution receives 
(income, wealth, whatever) to the bottom fifth or tenth 
receives.  These numbers are fairly widely available for 
many countries and periods of time, at least as available 
as Gini coefficients.  More theoretical measures are the 
Dalton and Atkinson indexes..  But the Dalton depends on an 
assumed utility function form, and the Atkinson one puts 
strong assumptions on the utility function as well.  The 
Atkinson presumably measures a loss of welfare due to 
deviations from a presumed "equally distributed equivalent 
income."   Whew!  I am not aware of any empirical estimates 
based on either of these measures.
Barkley Rosser
On Mon, 14 Apr 1997 11:22:29 -0700 (PDT) "Bove, Roger E." 
<[EMAIL PROTECTED]> wrote:

> 
>      A friend has asked me about alternatives to the Gini coefficient. I 
> seem to remember a paper titled 'On the Measurement of Inequality', but I 
> can't find it in my files. Anyone have suggestions?
>                               Roger

-- 
Rosser Jr, John Barkley
[EMAIL PROTECTED]