When estimating census variables on custom geographies using tongfen_estimate, either current area-based interpolation or enhanced using dasymmetric methods, there is a question where people in different categories live. We can exploit spatial autocorrelation across neighbouring regions to estimate the likelihood that people of certain characteristics cluster in different parts of a census polygon to enhance precision of the estimate.
When estimating census variables on custom geographies using tongfen_estimate, either current area-based interpolation or enhanced using dasymmetric methods, there is a question where people in different categories live. We can exploit spatial autocorrelation across neighbouring regions to estimate the likelihood that people of certain characteristics cluster in different parts of a census polygon to enhance precision of the estimate.