Fixing multicollinearity in panel data regression

Hey guys.

It's a big question in short format; I have data from 42 countries over 12 years and one of my independent variables is GDP (transformed by ln). I have a few other macro level variables and I would like to run some pooled models but I run into trouble - I have a lot of multicollinearity between my independent variables.

Does anyone know of methods to fix this and still keep the pooled models?