Who is Most Likely to Oppose Federal Tuition-Free College Policies? Investigating Variable Interactions of Sentiments to America’s College Promise

With the conclusion of the 2016 election, Americans were questioning if race and gender identity differences are as prevalent as the election suggests. Attempting to answer that question as it pertains to higher education policy and drawing inspiration from Social Identity Theory, this research utilized thousands of social media comments to analyze the likelihood of standing against the tuition-free policy, America’s College Promise, as determined by source, gender, and race and subsequent variable interactions. To investigate these likelihoods a binomial logistic regression model was calculated. Using marginal estimates, results suggest that separately race and gender are influential factors and of the four sources examined comments from the Fox News source was clearly different than the other three. For most interactions, race is the most dominant influence followed by gender – until interacting with the Fox News source. Next, Bag of Words models were generated to capture tokens (words and phrases) associated to source, gender, and race - and for variable interactions. Uncovered tokens illustrate several obvious differences between political identities and provides nuance to findings and discussion presented. This research concludes by discussing the importance of findings as it relates to intersections of crafting higher education policy and understanding identity differences