Files
Abstract
The 2000 presidential election renewed calls to abolish the Electoral College and elect the President by popular vote. However, the Electoral College also affects how campaigns allocate resources like money and time. Previous research has revealed that candidates allocate more resources to battleground states with unpredictable outcomes. My research goes beyond Shaw (1999, 2004 and 2006) by adding several control variables and testing for super battleground states. Such battleground states have many electoral votes. Consequently, these models more effectively include size and competitiveness. Likewise, the control variables mean that the model is specified more accurately. This research uses negative binomial count models and pooled OLS regression for the last five elections. These relationships remain even with control variables. Tests also suggest the existence of super battleground states. However, the findings are inconclusive due to apparent multicollinearity. This research has significant normative implications concerning voter turnout and states having equal roles.