I've been calculating the Gini coefficients for sets of countries. Originally I decomposed the Gini based on income sources (employment and self employment):
-descogini income employ self-
This gave me the contribution of each source, as well as the overall or total income Gini. After creating the group variable Sector, I also decomposed inequality on a sector basis using:
- ineqdeco income, by(Sector) -
This gave the Theil and other breakdowns I was looking for, but it also gave me the overall Gini for the dataset. Finally I just looked at income only:
-ineqdeco income-
This gave the Gini for the set as well. While they are similar, the results differ slightly in each case. Which one is correct?
I also looked at the Gini for each sector. First I acted as if sectors were an income sources as I didn't know the correct code, and used:
-descogini Income a f g h ...ru-
where a ...ru are NACE sector codes. This gave me the Gini/Gk for each sector. However the results here differ considerably to the sector breakdown using - ineqdeco income, by(Sector) -
The sector breakdown using - ineqdeco income, by(Sector) - does return the same result as when I use this command to look at sector inequalities individually using - ineqdeco income a - or -ineqdeco f - etc.
Again which way is correct? I would appreciate any suggestions.
-descogini income employ self-
This gave me the contribution of each source, as well as the overall or total income Gini. After creating the group variable Sector, I also decomposed inequality on a sector basis using:
- ineqdeco income, by(Sector) -
This gave the Theil and other breakdowns I was looking for, but it also gave me the overall Gini for the dataset. Finally I just looked at income only:
-ineqdeco income-
This gave the Gini for the set as well. While they are similar, the results differ slightly in each case. Which one is correct?
I also looked at the Gini for each sector. First I acted as if sectors were an income sources as I didn't know the correct code, and used:
-descogini Income a f g h ...ru-
where a ...ru are NACE sector codes. This gave me the Gini/Gk for each sector. However the results here differ considerably to the sector breakdown using - ineqdeco income, by(Sector) -
The sector breakdown using - ineqdeco income, by(Sector) - does return the same result as when I use this command to look at sector inequalities individually using - ineqdeco income a - or -ineqdeco f - etc.
Again which way is correct? I would appreciate any suggestions.
