Estimating within country and between country - Theil Index

[Sean O'Connor]

Folks,

I'm looking to measure both within and between country inequality. I've been advised to use ineqdeco however, I've no experience with the command.

Firstly, I'm not even sure how to code my data e.g. how to match country level GDP per capita with it's regional values. Is anyone aware of a tutorial or anything online which goes through the various procedure?

Any help appreciated.

[Stephen Jenkins]

Let us suppose that the basic unit in your analysis is the "country" and, for each country, you have a measure of income ("GDP per capita"). Countries can be classified into groups (a.k.a. "regions"). So you may have a dataset with one row per country and variables for country_identifier, income and region. You can then use ineqdeco (on SSC -- you should do as the FAQ says and tell us the provenance of user-written commands) straightforwardly, to decompose global inequality by region. (Within-group inequality is the weighted sum of the inequalities within regions; Between-group inequality is the inequality between regions (inequality of region-mean incomes).) It sounds like you need to do some background reading first. Google for the work of Branko Milanovic, for instance, including his article in the Journal of Economic Inequality 2011, and a recent World Bank working paper (joint with Christoph Lakner). They discuss principles in this area, and provide some estimates. Also see the relevant sections of the chapter by Anand and Segal in the Handbook of Income Distribution, Volume 2, North-Holland, 2015.

[Sean O'Connor]

Thank you Stephen