Plus, most of the entries are a mess of "spouse" jobs. e.g. They have 8k receptionists in the top 1% and 7k laborers. Dropping the lowest income household earners that weren't required in order to reach 1% would focus the data better.