Geospatial Model for Residential Property Tax Collection using the Census Block Groups as Homogenous Zones

Abstract

This investigative project proposes a tax collection model based on a property value homogeneous area using the U.S. Census block group as the base region for residential tax type. Geospatial technology was used to compare market value sales with owner-occupied dwellings Census data applying geostatistics, spatial join visualization, and spatial distribution. The developed model uses the Census block group as geographic region with the market property sales as baseline for current fiscal economy. A sample of 288 records of 5,614, for years 2013 – 2018, was set aside to calculate the Root Mean Square Error after comparing each set of data. This model confers an improvement in out-of-sample prediction accuracy of up to 5% proving that market value sales resembles current fiscal economy reality. Key Terms ⎯ geospatial technologies, homogeneous zone, market value, real estate appraisal process