Can you come up with a reason the CDC's Behavioral Risk Factor Surveillance System survey would be wrong here? It groups respondents by county of residence. It's a long-running, well-funded, and valuable survey.