Decomposing the Theil coefficient of inequality

[Robert Sweeney]

I am trying to decompose the Theil coefficient of income inequality into between and within components on a sector (manufacturing, retail, etc.) basis. Sector variable names follow EU NACE sector coding (a, b ..u). That is, my variable containing observations on income in agriculature is labelled 'a'. 'Income' is the variable containing all income observations in the year. As per a previous thread, I generated a group variable called sector:

'egen sector = group(a be f g h i j k ln o p q ru)'

then I input the command which apparently should give me my Theil and its breakdown

'ineqdeco Income, by(sector)'

It gives the Theil GE(1) below:

'Generalized Entropy indices GE(a), where a = income difference
sensitivity parameter, and Gini coefficient

----------------------------------------------------------------------
All obs | GE(-1) GE(0) GE(1) GE(2) Gini
----------+-----------------------------------------------------------
| 73.49310 0.38037 0.26989 0.33007 0.39041
----------------------------------------------------------------------

It also gives some subgroup stats, which don't really make sense to me, and they go on for ages until cut off.

'Subgroup summary statistics, for each subgroup k = 1,...,K:'


-------------------------------------------------------------------------------------
group(a |
be f g h |
i j k ln |
o p q ru) | Popn. share Mean Relative mean Income share log(mean)
----------+--------------------------------------------------------------------------
1 | 0.00588 1.00000 0.00004 0.00000 0.00000
2 | 0.00588 40.00000 0.00167 0.00001 3.68888
3 | 0.00588 127.70000 0.00532 0.00003 4.84968
4 | 0.00588 300.00000 0.01250 0.00007 5.70378
5 | 0.00588 2027.40000 0.08445 0.00050 7.61451
6 | 0.00588 3000.00000 0.12496 0.00074 8.00637
......'

It also gives me the following (among other things), which I suppose should give what I'm looking for

'Subgroup indices: GE_k(a) and Gini_k

----------------------------------------------------------------------
group(a |
be f g h |
i j k ln |
o p q ru) | GE(-1) GE(0) GE(1) GE(2) Gini
----------+-----------------------------------------------------------
1 | 0.00000 0.00000 0.00000 0.00000 0.00000
2 | 0.00000 0.00000 0.00000 0.00000 0.00000
3 | 0.00000 0.00000 0.00000 0.00000 0.00000
4 | 0.00000 0.00000 0.00000 0.00000 0.00000
5 | 0.00000 0.00000 0.00000 0.00000 0.00000
.....'

Does anyone have any idea why it isn't working, and how to get the within/between breakdown?

Thanks.

[Stephen Jenkins]

Have you got the latest version of the program? -ssc install ineqdeco, replace-

PS please use CODE delimiters to report Stata input and output (see Forum FAQ for more about this)

[Robert Sweeney]

Thank you. I made a mistake in creating group variable. Problem solved.