The data seems very weird for my home county (Sanders, MT). A few factors might contribute to this: 1. There are a lot of troubled youth 'programs' there. Weal

The data seems very weird for my home county (Sanders, MT). A few factors might contribute to this: 1. There are a lot of troubled youth 'programs' there. Wealthy parents send their children to a depopulated part of Montana to keep them out of trouble, and then they return to the advantages of their family's income. One member of my high school's graduating class works for his dad's investment bank, for instance. 2. The data might not include people who don't graduate high shcool(?) or it might not include dishwashers and convenience store clerks. I'd like to know how the data was gathered. 3. Small sample sizes which include outliers give distorted results. One early-stage Dropbox engineer makes a big difference when the entire county graduated 300 people from high school that year.