Mean Log Deviation (MLD) decomposition

[Ahmed Zohirul ISLAM]

I am analyzing the level of income inequality between high-skilled and low-skilled populations using New Zealand census data. I will use Mean Log Deviation (MLD) decomposition technique to decompose within-group and between-group inequality.
What commands should I use to perform the MLD decomposition to analyze within-group and between-group contributions? I would like to understand how it works and replicate it in my analysis

[Stephen Jenkins]

Try -ineqdeco- (available from SSC). That assumes you have unit-record data.

In 2018 I examined a University of Waikato PhD thesis by Omoniyi B. Alimi about "Inequality within and between New Zealand Urban Areas" which applied MLD decompositions to NZ Census data. To address grouping (banding) in the earnings variable, -rpme- (on SSC) was used. I recommend you look at that thesis.

[Ahmed Zohirul ISLAM]

Thank you very much Professor Jenkins for your advice. Hope to get your support in the future.

[Ahmed Zohirul ISLAM]

I have read Omoniyi B Alimi's thesis. He studied regional level data. Since, he analyzed grouping (banding) income variables, he used rpme.

In my case, I am analyzing national level data. Statistics New Zealand publishes midpoints for each income categories for the national level. So, I have unit-record data.

I used ineqdeco command to analyze the level of inequality. To decompose the level of inequality, I used the following command-
ineqdeco varname, by (groupvar). The command is working to decompose within-group and between-group inequality.

Now the question is - Can I use the ineqdeco command to decompose the change in inequality between two periods? What would be the command in this case?

Thank you so much in advance.

[Stephen Jenkins]

Please see my response at https://www.statalist.org/forums/for...all-ge-measure. Also, please do not ask the same question in multiple threads

[Ahmed Zohirul ISLAM]

Dear Professor Stephen, please accept my sincere apologies.

Thank you so much for replying to my question elsewhere.