I am using pooled cross sectional survey data to estimate prefernce for redistribution. I have 33 countries for 6 survey rounds covering the periods 2000-2013.
I would like to estimate the effects of country-specific variables such as the Gini index and I would like to know how the crisis influenced preference for redistribution. Currently, I am using a random intercept model estimating country-specific intercepts. E.g.:
xi: xtmixed prefredistri ib5.income unempl12m selfempl eduyrscentered agec agec2
female married childever discrimination reli Union leftori socialmob ib3.health
i.year GDPlag || cntry:, mle covar(unstructured)
Including year dummies would estimate the fixed effects per year which have not been controlled for, for each country individually - and not overall time-constant year effects - right? Thus changes in GDP could not be interpreted for each year right? Am I right? How should I best incorporate the time dimension in the multilevel model and get some inference about time effects?
I would be highly thankful for help.
