Looking at Maps with Tribal Lands in Mind

Re-reading Cornell and Kalt's "Two Approaches" got me thinking about the heterogeneity of institutions and circumstance across tribal areas in the U.S. Whenever I see maps in an article like this:

http://www.nytimes.com/interactive/2016/01/07/us/drug-overdose-deaths-in-the-us.html

I always try to mentally compared them to maps like this:

https://nationalmap.gov/small_scale/printable/images/pdf/fedlands/BIA_2.pdf

Is there anything going on here?  White Earth and Rosebud (and maybe Colville?) sort of stick out on the negative side, while other areas of tribal land (Cheyenne River, Eastern Cherokee?) seem to maybe be better off than surrounding areas. This is central challenge in doing research on Indian Country as a whole.  The average effect of living in Indian Country for a problem like this maybe zero, but the average may mask a lot of variation.