Document Type



Doctor of Philosophy


Political Science, American Politics

Date of Defense


Graduate Advisor

David Kimball


Baybeck, Brady

Robertson, Dave

Stein, Lana


The purpose of this study is to explore whether statewide geography influences a candidate’s electoral position. More specifically, what role does geography play in candidate emergence and success on the statewide level? Conventional wisdom holds that geography does matter in candidate emergence and success. Beyond the anecdotal, however, there has been little research in this area. I theorize that general attitudes about rural identity and, conversely, urbanity affect electoral outcomes based upon the population composition of the state. To explore these questions, I created a typology characterizing states as rural or urban, based upon an index capturing rural cultural and physical attributes of each state, and as concentrated or dispersed, a measure of the proximity of urban areas within a state. I coded the hometowns of candidates from 1948-2008, including a classification as to whether their home county was rural or urban on a relative scale. Using this data, I implemented a negative binomial regression to consider the likelihood of candidate emergence. In this model, I found that rurality is not a negative predictor of candidate emergence in most state types, with urban states being the exception. I then implemented a logit model to estimate the likelihood of winning with the traditional predictors of candidate success and within my theoretical framework. Relative to senate races, if candidates can get into the race, the negative effects of hometown lose significance. In gubernatorial races, coming from a rural area actually increases the likelihood of getting elected in rural, concentrated, and dispersed classified states. In sum, political geography matters and has implications on both candidate emergence and success. These findings and the use of this typology add an important component for future research. Simply put, geography should not be ignored in politics.

OCLC Number