Voting Behavior in Presidential Elections, 1992 - 2008
The animation of elections returns between 1992 and 2008 provides an overview of voting behavior in the United States. The maps do not account for candidates outside the Democrat and Republican parties, thereby masking significant support for third-party candidate Ross Perot in the 1992 and 1996 elections. Accordingly, these maps appear to be washed out in the light, low-percentage hues on the map. Candidates Bob Dole and President Bill Clinton re-established nationwide support for the Republican and Democratic parties in the 1996 elections, partially diminishing Perot’s appeal. Clinton ultimately emerged as the victor, having secured a substantial margin of the popular vote in the northeast, the upper Midwest, the Pacific coast and much of the South. The pivotal 2000 elections indicate a surge in Republican rural support that resulted in the close, hotly-debated contest between candidates George W. Bush and John Kerry. While the Democrats maintained support in the most populous counties in the country in 2000, they were unable to achieve election. In the 2004 elections, the popularity of war-time President Bush continued the transformation of the formerly Democratic South of the bygone Clinton years into a Republican stronghold. Although Barack Obama won every region in the 2008 elections, gaining strong increases in Democratic support nationwide, he was unable to re-secure support in the Great Plains and the South, which went to Republican candidate John McCain.
A cluster analysis distills the information in the animation frames into a single map by grouping counties with similar trends in voting behavior between 1992 and 2008. In this case, the clusters were formed by combining the mean percent voting margins for each presidential year in the study period (1992, 1996, 2000, 2004 and 2008). The resulting map shows a persistence of Republican support in the Midwest and Mountain regions, which are colored by red hues. Strong Democratic support along the coasts and in the Great Lakes region, indicated by the blue hues, is also clear. A mosaic of voting trends emerges east of the Mississippi River, presumably from various rural-urban-suburban differences. Interestingly, this cluster analysis provides support for gradual polarization of party support, as seen in the increasing disparity between vote margins in the clusters over time. However, it does not address voting shifts that occurred in the past five presidential elections. A second cluster analysis is required for this purpose.