Variance contribution testing

I have a dataset that has geographic data (county-level data of cancer incidence) which has a large variance. For each county, I have a list of county-level characteristics (socioeconomics, environmental, etc - e.g. obesity).

I want to understand the contribution of the county characteristics to the geographic variation in cancer incidence. What test should I use?
I was thinking of using linear regression, but many of these county-level characteristics are colinear (e.g. unemployment and education, etc)



Not a robit
So all variables are at the county level?

What do you mean by "has a large variance"?

It may be beneficial to look to the oncology or epidemiology literature to see if similar study control for a covariance structure between tangential counties.