Iam currently writing for my undergraduated thesis. I have large panel data in year 2007 and 2014. my dependen variable is change in consumption.
My independen variable is :
1. interaction between change in health condition in dummy form with dummy variable for distance 0-1km
2. interaction between change in health condition in dummy form with dummy variable for distance 1-10km
3. interaction between change in health condition in dummy form with dummy variable for distance >10km

I perform command for declaring panel using
xtset pidlink year
and regression using RE
xtreg consumption distance1 distance2 distance3 demographic, RE
the result is all the variable omitted due to collinearity, is there any other solution to solve this problem? is it okey with me makes my independent variable for distance like that?

Thank you in advance