Political satirists have long poked fun at the missteps of politicians seeking to personally connect with voters. From Andrew Jackson’s whiskey-fueled open houses on Pennsylvania Avenue to Mitt Romney’s comments to the Detroit Economic Club that his wife drives “a couple of Cadillacs,” politicians have gone to great lengths, often unsuccessfully, to find common ground with voters. Politicians base these attempts on the belief that voters seek descriptive representation, an elected body that resembles the population in terms of descriptive characteristics like ethnicity, religious beliefs, and education level. At first glance, descriptive representation seems firmly founded on electoral precedent. In describing Barack Obama’s victory in the 2008 election, political commentators often referred to commonalities between then Senator Obama and certain blocks of American voters as a distinct electoral advantage.
Despite this apparent electoral precedent, it seems fundamentally absurd that Mitt Romney, with a personal wealth of over $200 million and degrees from both Harvard Business School and Harvard Law School, spent a portion of his campaign attempting to convince his conservative base that he loves NASCAR. However, by analyzing the role of descriptive representation, absurdities of this nature become more understandable. I hope to analytically investigate the rationale behind descriptive representation. Specifically, I hope to explain the role of a few descriptive trait of candidates—including education level and overall academic prestige. I will examine the relationship between a candidate’s personal education level and his electoral success in order to analyze the apparent role of descriptive representation within congressional elections.
Of course, this project is not without obstacles. In fact, initial attempts at determining positive relationships between congressional candidates’ education as it compares to their districts’ general education level and their electoral success have fallen short. The biggest obstacle comes from data availability and selection—for one, the availability of data that outlines the relative education level of one district in comparison to another is slim and, further, defining the variable that accounts for academic prestige is complicated and inherently subjective. Additionally, the amount of variance that I seek to capture is relatively low—other, obvious factors will almost certainly supersede descriptive similarities in educational backgrounds (i.e. party identification); thus, in particularly partisan regions, it will likely become increasingly difficult to determine a causal relationship. Despite these obstacles, I believe that alternate approaches to the question of interest could result in more promising results. After all, the causal logic is clear—voters that identify with a candidate will have a higher propensity to vote for that candidate. This causal logic has been verified by previous research regarding genetic qualities like race and gender and, thus, could very well hold with attained attributes like education level and academic prestige.
I am honored to be a part of the Presidential Fellows Program, and look forward to working with all of you.