My topic has changed (again) since my last blog post and since the fall conference. I have narrowed my topic to evaluate the effect of military experience on congressional roll call voting. The number of Senators and Representatives with military experience has been steadily declining since WWII. Do veterans in congress vote in a significantly different way than non-veterans? What can we expect from the continued declining proportion of veteran legislators? The new goal of my paper is to provide an initial answer to these questions.
I have narrowed my focus to the 111th Congress (2009-2010). My analysis so far has shown that military experience does not have a statistically significant effect on a legislator’s propensity to increase/decrease military spending or support/oppose a more assertive US military presence abroad. This would seem to indicate that veterans are not more or less likely to be war hawks than their non-veteran peers. Party affiliation is the most powerful predictor of a legislator’s military spending views. This is in-line with previous literature on roll-call voting behavior, which has shown that party affiliation is the strongest predictor on core party issues. This has traditionally meant strong military spending and lower taxes for Republicans, and entitlements for Democrats.
So far, I have found that military experience does affect one area of voting: veteran’s benefits. I regressed a legislator’s score from the Iraq and Afghanistan Veterans of American report card (which tracks key votes on veterans issues) on the legislator’s age, tenure, religion, gender, party, race, and military experience. The dummy variable for military experience was both positive and statistically significant: if your representative or senator has served in the military, then they are more likely to support generous veteran benefits. Veterans affairs has typically been a non-partisan issue (at least compared to vote on Social Security, healthcare, and taxation). This means that party influence should have a smaller effect on the roll-call vote outcome. This allows the representative to vote more in line with their personal beliefs and their constituent’s opinions.
Moving forward, I plan to check the effect of military experience on a greater number of dependent variables. I will also expand the size of my dataset to include the 109th, 110th, and 112th congressional sessions, providing I have the time. I am looking for feedback on my empirical method (e.g. what else should I include in the regression), possible holes in my analysis, and other topic I should cover in the rest of my paper.