Geospatial statistical methods offer a novel approach for measuring public ideology across various political jurisdictions.

Our research utilizes Bayesian universal kriging to develop precise estimates of constituent ideology in 2010, covering all US states and congressional districts. Unlike previous studies, we incorporate Alaska and Hawaii while distinguishing between ZIP codes and ZIP Code Tabulation Areas (ZCTAs). We leverage the 2010 Census for improved population data.

These new district-level measures effectively predict the ideology of elected state legislators, providing strong validation for their accuracy as a measure of constituent views.

Key Findings:

• District-level ideology estimates align well with legislator ideologies

• The technique successfully captures nuanced public opinion patterns across states and districts

• Geographic distinctions enhance measurement precision compared to prior methods