Transitions in County-Level Poverty in the Rural United States: A Comparison of LISA Techniques

Matthew Brooks , Pennsylvania State University

This analysis compares three different spatial statistical techniques to understand how the geographic distribution of poverty has changed in the rural United States since 1990, and if there is evidence for county-level diffusion effects. Specifically, the three featured techniques are based on LISA (Local Indicators of Spatial Association) statistics and differ in how they report changes in the distribution of poverty. The three techniques are bivariate LISA, cluster transitions, and diffusion transitions. Data is taken from IPUMS-NHGIS. Several regional clusters of persistent poverty are identified, with regions such as the Mississippi delta and Appalachia being consistent across techniques. However, some regional differences in poverty are only present in a single method. Evidence also suggest that there is a diffusion effect for poverty in Southeastern metropolitan fringe counties. Overall results from this analysis both confirm literature-identified spatial trends as well as highlight new trends and regional clusters which require additional research.

See paper

 Presented in Session 3. Population, Development, & the Environment; Data & Methods; Applied Demography