## Atkinson coefficient

Occasional Contributor
Posts: 9

# Atkinson coefficient

Hello,

I need help with programming the Atkinson Coefficient for in-eqaulity.

I couldn't find anything helpful in the internet, i have data which contains id number and income of each people.

I want the program to calculate the Atkinson coefficient, does somebody has a code of this?

Thank you for helping,

Matan.

Valued Guide
Posts: 505

## Re: Atkinson coefficient

If you have IML you do nto need to use my macro

%utl_submit_r64(%nrbquote(
library(ineq);
/* generate vector (of incomes) */
x <- c(541, 1463, 2445, 3438, 4437, 5401, 6392, 8304, 11904, 22261);
/* compute Gini coefficient */
ineq(x);
/* compute Atkinson coefficient with parameter=0.5 */
ineq(x, parameter=0.5, type="Atkinson")
));

[1] 0.4620911
[1] 0.1796591

In response to another post the Whittaker W/M functions are also in R packages

Occasional Contributor
Posts: 9

## Re: Atkinson coefficient

thanks for the help,

but what you gave me is a package for R, and i need to write the code in SAS.

also i don't use the SAS Enterprise Guide but the basic SAS.

thanks.

Contributor
Posts: 20

## Re: Atkinson coefficient

Hi,

Atkinson's inequality measure can be calculated using the formula, say

(yi/ybar)^.5  Atkinson, e=0.5

(yi)^(1/n)  Atkinson, e=1

(yi/ybar)^(-1) Atkinson, e=2

where yi is the income of the each people,

ybar is the mean value of the income of all.

n is the number of people.

e is the weighting parameter.

Regards,

Manohar

Occasional Contributor
Posts: 9

## Re: Atkinson coefficient

Thank you,
So if i understand correctly, for atkinson with parameter 0.5, I don't need to use the number of people in my data set.
Also, if I want paramter 2.5, what formula do I need?

Contributor
Posts: 20

## Re: Atkinson coefficient

Please refer the heading 6.2.3 Atkinson's inequality measures in the attached pdf.

Replying with formula is quite diffcult.

Apply the value of weighting parameter with the formulae gives you the result.

Regards,

Manohar

Discussion stats
• 5 replies
• 554 views
• 0 likes
• 3 in conversation