Dear Community,

I am working with the European Working Conditions Survey and I would like to calculate a longitudinal effect. Unfortunately, this survey is not panel data.
In my analysis I would like to include a total of 30 countries at two different survey times. My problem is that I could take time as a grouping variable, but then I would have only two groups.

Now I found a publication that could solve my problem. However, I do not understand how to implement this in Stata.

On the lower level all individuals of the two surveys are considered. On the level above, the indiviudals are nested in "country-years", after which 2 time points * 30 countries = 60 "country-years" are used as grouping variables.

How do I get the mixed command to take these "country-years" as grouping? For example, Stata needs to know that the same country should be compared.

I hope that I could explain my problem reasonably well and I hope that someone can help me.

Thank you very much in advance.

Kind regards
Susan