Question regarding specifications

[Moon Choi]

Hello, I'm a toddler in econometrics.
I got a lot of help by reading the posts from this forum, thank you all.
So I decided to join the forum to ask some of my questions.

Q: I'm currently working on a paper to examine the effect of variables on Gini Coefficient.
I'm using panel data, t=16 with n=544, time-fixed with time dummies.
And here is my question, when comparing specifications, is it okay to divide my sample into 2 (say, advanced economy and developing economy) and compare the coefficients with each other? and with a model including all?
Or do I need to omit, or add an explanatory variable because fixed effect already accounts for differences.
Thank you for reading my question.

[Phil Bromiley]

First, you'll increase your chances of a helpful answer by following the FAQ on asking questions - provide Stata code in code deimiters, Stata output, and sample data using dataex. Also, try to reduce the code you include to what is needed to demonstrate your problem. You should also have searched for the major terms in your question from the command window of Stata (findit majorterm) and in the subject index in the documentation (under index at the bottom of the list of documentation).

If you believe the parameters on the x's should vary, it makes sense to test for that variation. Rather than dividing the sample into two, I'd be inclined to use a dummy for the difference (advanced means d=1, developing means d=0) and then interact this with your x's. Look at the factor variable notation.

Some forms of difference (e.g., a dummy without being interacted) will not estimate if all firms are only in one kind of country.