How do you detect and handle multicollinearity in a dataset with hundreds of potential policy indicators?
