County-level poverty estimates for the contiguous United States, 2001, 2005

Efforts to estimate various sociodemographic variables in small geographical areas are proving difficult with the replacement of the Census long form with the American Community Survey (ACS). Researchers interested in sub-national demographic processes have generally relied on Census long form data products in order to answer research questions. ACS data products promise to begin providing up-to-date profiles of the nation's population and economy; however, unit and item-level non-response in the ACS have left researchers with gaps in sub-national coverage resulting in unstable and unreliable estimates for basic demographic measures. Borrowing information from neighboring areas and across time with a spatiotemporal smoothing process based on Bayesian statistical methods, it is possible to generate more stable and accurate estimates of rates for geographic areas not represented in the ACS. This research assesses this spatiotemporal smoothing process in its ability to derive estimates of poverty rates at the county level for the contiguous United States. These estimates are then compared to more traditional estimates from the Census, and error rates are calculated to evaluate the practical application of this smoothing method. The resulting summary choropleth map displays the Bayesian estimates of county-level poverty at a scale of 1 to 12,000,000 along with summary choropleth maps of the more traditional estimates at a scale of 1 to 37,000,000 for 2001 and 2005. Error rates indicate that the Bayesian estimates of county-level poverty produced by our succinct model produce results similar to more complex traditional estimates produced by the Census.