ABSTRACT
This chapter examines the 2008 presidential election through the lenses of each of these dimensions. It offers a quantitative analysis using geographically weighted regression, a means of examining the spatially uneven strength of different variables' influence on votes for Barack Obama while statistically controlling for the influence of confounding forces. Religion played an important role in the 2008 presidential election. Once the controls for ethnicity and religion are introduced, the geographically weighted regression (GWR) nonetheless confirms that Obama's support tended to rise steadily with household income in the West more than most parts of the country. Thus, without the controls of income or religion, white support was highest in Northern California and the Pacific Northwest and parts of Virginia near Washington, DC. Once the effects of income and religion are removed, however, this pattern changes significantly, with white support highest in the West Coast and Nevada, the Mississippi Delta, Colorado, southern New England, and the greater Washington, DC region.
