Over the past couple of decades, technical models, both statistical, machine learning and combinations of these methods, for forecasting various forms of political conflict, including protest, violent substate conflict, and even coups, have become surprisingly common in policy and NGO communities, particularly in Europe, though not, curiously, in US academia. These methods, working with readily available, if noisy, open source data, use a number of familiar predictive analytical approaches such as logit models in the statistical realm and random forests in the machine learning, and consistently outperform human analysts. This talk will first review the current state of the field, with a particular emphasis on why current models work whereas prior to 2005 there was little consistent success with the problems, and then present some challenges that remain unresolved. The talk will assume familiarity with general social science quantitative approaches, but not with the details of specific technical approaches: lots of graphics, a couple tables, no equations.
Philip Schrodt is a senior research scientist at the statistical consulting firm Parus Analytics. He received an M.A. in mathematics and a Ph.D. in political science from Indiana University in 1976, and has held permanent academic positions at Pennsylvania State University, the University of Kansas, and Northwestern University. He has also held research appointments in the United Kingdom and Norway, and has taught and done field research in the Middle East. Dr. Schrodt’s major areas of research are quantitative models of political conflict and computational political methodology.
