Gini index

[mari ajam]

Hello

How do I calculate the Gini coefficient with the equivalent scale method in stata ?
Please explain the equivalent scale method and per capita income household formula

Thank you

[Marcos Almeida]

Welcome to the Stata Forum / Statalist.

Your query seems to be quite broad. What is more, it gave the impression the you didn't read the FAQ, particularly the topic 3.

That being said, I gather this query may be helpful to you.

Hopefully that helps.

[mari ajam]

Thank you for your help
I have studied the relevant questions, but none have been the Equivalence Scale method, and I want to calculate the Inequality Measures with Equivalence Scales and Adjustment for Household Size and Composition.

[William Lisowski]

It appears likely that you refer to methodology described in

Figini, P., 1998. Inequality Measures, Equivalence Scales and Adjustment for Household Size and Composition [http://www.lisdatacenter.org/wps/liswps/185.pdf], Luxembourg Income StudyWorking Paper No. 185, and Trinity Economic Papers, Technical Paper No. 8/98.

With that said, the output of the Stata command search figini does not return any results, making it seem unlikely that code for this calculation has been implemented in a user-written and shared program.

[mari ajam]

Thank you so much
I have read this article but I have not noticed some of it, for example, in this article
"The OECD scale is a particular case of equation (3), in which a1 = 0.7 and a2 = 0.5. Other scales often used attach values of 0.6 or 0.5 to a1 and 0.4 or 0.3 to a2."
What are these values for other countries? ( example Iran)

[Brina Seidel]

The general problem with using equivalence scales is that no one agrees on what the right scale should be. While some countries have their own equivalence scales, you should only bother using them if your goal is to replicate what that country's statistical office does. There's rarely an objectively good reason to use, say, the Iranian equivalence scale values with Iranian survey data - the Iranian numbers would essentially be arbitrary.

To answer your original question: using the equivalence method and calculating the Gini are two separate steps. First, you should adjust the household or per capita income values in your data using your equivalence scale. Then, you should calculate a Gini using those adjusted income values the same way you would normally calculate a Gini. You do this by sorting individuals by their income, applying survey weights, and counting up the cumulative percentage of total income accruing to the cumulative percentage of the population.

If you have sample data, you can paste a sample here using the dataex command (see Statalist FAQ). I'd be happy to help you code up the calculation after you give it a shot. If the data is at the household level, then you will need a household id, household income, and columns indicating the number of individuals in the household of each class that you're separating out in your equivalence scale (eg children under 5, adult men, or what have you). If the data is at the individual level, you will need an individual id, per capita household income, and columns indicating their age and gender.

[William Lisowski]

This doctoral dissertation from 2012 says that the author found no published equivalence scale in the literature, and on page 104 discusses his methodology and assumptions for a1 and a2.

https://eprints.soas.ac.uk/17359/1/Pourghadiri_3530.pdf

This is pretty far outside my area of competence, and well outside the realm of Stata, so I'll leave it at that. Perhaps another member will be able to address your methodological issues.

Added in edit - and while I was typing this post, indeed,Brina Seidel was able to contribute substantially to the substance.

[Stephen Jenkins]

For a closely related thread, please see https://www.statalist.org/forums/for...ni-coefficient. Mari Agam: please search the Forum for related posts (and read the FAQ).

For a survey of equivalence scales, see e.g. https://onlinelibrary.wiley.com/doi/...1992.tb00538.x and the references therein.

[mari ajam]

Thanks to all , I really appreciate your help.