“We Believe in Free Expression...” Reverse-Engineering Twitter’s Content Removal Policies for the Alt-Right

Abstract

Twitter defines its values: “We believe in free expression and think every voice has the power to impact the world.” This values statement is not without caveats; with the proliferation of abusive speech and hate groups on its platform, Twitter has published a series of user content rules to support its values statement so that the greatest number of users can express themselves freely. However, the enforcement of these rules, as well as the number of profiles that are removed due to the violation of stated policies, is unclear. To determine if Twitter is true to its published policies, this thesis describes the mapping of these rules to quantitative findings regarding content removal, specifically for the alt-right, a movement that tests the boundaries of free speech. From the source node of prominent alt-right leader Richard Spencer, I used the Twitter Application Programming Interface (API) to scrape the tweets of almost ten thousand alt-right users, continuously updating this repository of tweets to determine whether or not each user was still active. By running statistical tests and network analysis metrics on the removed versus active user data, I was able to reverse-engineer the ways in which Twitter removes content in practice. Ultimately, this research finds that for the alt-right, Twitter is not honoring its published content removal policies.