Hello everyone,

First of all, I would like to apologize in advance if I commit any mistake on this post or I do not fully respect the policy of Statalist. I will try to do my best. Comments and questions are more than welcome.

I am studying the rise of populism in Spain since 2000 at a province level.

I am using the following equation: Yijt=a+b1(gdp_growth)it+b2it+b3(Edu)+b4(Elec)j+ai+ at+Eijt

Yijt=share of the populist vote in province i, type of election j, and time t, j=g if the election is national or if it is regional.

There are three dummies for the four types of elections I have in my database: First takes value 1 if it is General Election, 0 otherwise; the second takes value 1 if it is regional and 0 otherwise; the third takes value 1 if it is European and 0 otherwise and the fourth is not included because we assume if the rest is 0 it is local election.

If I use xtset command the following error pops up.

Code:
.
xtset id year
repeated time values within panel
id=province

I have this problem because for some provinces, in the same year two elections were celebrated and STATA cannot read it as a panel data. For example, for id 1 I have one year repeated 2009, because on that year general elections and European elections were held on the same year. My total number of observations is 813.

Could you please help me with some ideas regarding what commands do you recommend me to use?

Thank you very much for your time and consideration. Best wishes, Josué.