If I understood right, you need to merge different sub-data. You should create a single identification number (the key) because your current geofips is not unique to each observation.

One possibility is by doing a grouping of existent variables. You should do this in each sub-data (you can use a loop). For instance:

egen id = group(geofips gename year)

Then, you should sort your data and save it.

sort id

save ResearchGate1.dta, replace

Let's assume that ResearchGate1 is for unemployment. And ResearchGate2 for incomes.

Then, you are able to merge your data. If you open ResearchGate1, then you can merge it with ResearchGate2 and so on.

merge 1:1 id using "ResearchGate2.dta"

If I understood right, this may it help you,

Best

Gonzalo Durán S.

I see now what I was missing.

Thank you so much.

Best

Gonzalo Durán S. has a good solution - however be cautious to ensure that in < ResearchGate1.dta> represents the same distinct set of and as in . This is because sorts the data by the variable list and assigns each distinct group an integer value. If one dataset includes distinct groups that the other does not, you likely will not be merging your data correctly.

A way you can ensure you are merging correctly is by merging with these variables: . If I understand correctly, it sounds like this should be a 1:1 match. If you are getting any error suggesting otherwise, you will need to review your data to understand why this is the case.

Jeffrey Cisyk I thought about that. But in this particular case I can assure that the sets of variables represent the same group in both, because they were part of the same data base.

I separeted them in order to organize it as a panel.

Thanks for the advice though.