Per population they do if you would read the bottom. Its taken by how many kids went D1 divided by how many kids actually played football in the state over the last 4 year period. You don't have to believe me but take that up with the NCAA who released the Study
You’re the one who posted the original picture with no context. It is very sensible that a smaller sample size puts out more per capita. Basic statistics suggest that’s the case for all case study’s using population and per-capita.