It's not very complicated. You really simply spread out all the loans on a map and see if there's a difference in the coverage. It could be anybody. It could be an academic, or it could be.... I mean, the private insurers' data is basically public, so you get the data from the two, you merge them, and you see if there are differences.
