The Electoral Geography of Provisional Ballots by County: The North Carolina 2008 U.S. Presidential Election

The geography of provisional ballots is a realm of electoral geography that is increasingly important since such ballots can frequently shape electoral outcomes. That said, it is an area of inquiry that remains considerably under-researched. The purpose of this paper is to partially remedy this deficiency by highlighting the geographic significance and overall impact of the 52,000 provisional ballots cast during the 2008 presidential election in North Carolina. The paper focuses on a comprehensive analysis of the spatial distribution of provisional ballots in both absolute and relative terms by county. Based on a regression analysis we highlight the links that exist between the number of provisional ballots cast by county and various predictor variables. The research found that older, well-educated, predominately white counties found in Western and Coastal North Carolina cast fewer provisional ballots than other counties. Furthermore, counties with long agricultural histories and “minority-majorities” located in the northeastern part of the state cast provisional ballots at much higher rates. Abstract: La geografia de los votos provisionales es un ambito de la geografia electoral cada vez mas importante, ya que tales votos con frecuencia pueden influenciar los resultados electorales. Dicho esto, es un area de investigacion que sigue siendo considerablemente poco investigada. El objetivo de este trabajo es remediar parcialmente esta deficiencia poniendo de relieve la importancia geografica y el impacto global/total de los 52.000 votos provisionales de Carolina del Norte durante las elecciones presidenciales de 2008. El documento se centra en un analisis exhaustivo de la distribucion espacial de los votos provisionales tanto en terminos absolutos como relativos por condado. Basado en un analisis de regresion se destacan los vinculos que existen entre el numero de los votos provisionales por condado y diversas variables predictoras/pronosticadoras. La investigacion encontro que los condados mas viejos, bien educados, predominantemente blancos del oeste y de la costa de Carolina del Norte recibieron menos votos provisionales que otros condados. Ademas, los condados con un largo historial agricola y “minoria-mayorias” situados en el noreste del estado tuvieron cantidades mucho mas altas de votos provisionales.