An Islamic Conference on Violence and Extremism in Beirut

When Groups Fall Apart: Measuring Transnational Polarization with Twitter from the Arab Uprisings


Scholars continue to disagree as to what extent international social connections act as a conduit to influence contentious politics within states. To answer this question, we provide the first rigorous and real-time measure of transnational ideological diffusion across sectarian groups by employing a novel statistical method and new data to capture the transnational dynamics of polarization after the Arab Uprisings of 2011. As authoritarian governments fell, populations in several states polarized between secularists and Islamists over what kind of regime was to replace the ousted one. To examine these endogenous processes, we collected a comprehensive dataset on elite and citizen Twitter accounts in Cairo and Alexandria (Egypt) and Tunis (Tunisia) for a ten-month period during the critical year of 2013. Given the difficulty in directly measuring polarization, we also developed a new model, item response theory-vector autoregression (IRT-VAR), that allows us to incorporate measurement uncertainty while providing over-time estimates of transnational polarization. We show through our model that following catalytic events like regime ousters (such as the military coup against the Muslim Brotherhood in Egypt), we can separate the direct effects of these events on group polarization within a country from indirect transnational feedback happening through the channel of social media.

Paper link: []