TAKING A STANCE ON WAR: ANALYZING HOW UNITED STATES HOUSE MEMBERS NAVIGATE PARTY, PERSONAL IDEOLOGY, AND CONSTITUENT INTERESTS

Abstract

When politicians face tough legislative decisions, they need to respond to a number of interest groups before acting. Namely, three major factors of consideration are constituent interests, party stance, and personal ideology. Using messaging data, such as from congressional debates or social media posts, I make claims about the degree to which House members respond to these three factors of consideration. I use messages from politicians about wars with United States’ involvement, specifically Iraq and Afghanistan, as a proxy for constituent representation. Observing the veteran populations in each congressional district acts as a way to quantifiably measure the varying constituent interests regarding war. There is analytical evidence that veterans have more robust opinions about war, thus allowing for the messaging of politicians to act as a proxy for district representation. Further, I use district contestation as a second variable to understand drivers of messaging frequency. The data suggests that representing highly contested districts motivates a change in messaging frequency, meaning that the consideration of reelection plays a factor in political behavior. However, when pairing veteran populations with a congressional district’s degree of electoral contestation, a significant relationship is not found. Further, it appears that those House members who served in the military speak about war significantly more than their fellow civilian members. At pivotal moments in war, an increase of responses from parties as a whole is observed. This finding lacks specific individuals driving the increase of responses, suggesting that the actions come from the entire party. Observing these three pieces of data, work to support the claim that House members respond to their party and their own personal ideology rather than constituent base when deciding how to message those they represent.