DiogoFerrari.jpg I am a graduate student of Political Science at the University of Michigan, Ann Arbor, with double Major in Methods and Comparative Politics and dual degree in Statistics. My previous training started in the area of Computer Science at the Institute of Mathematics and Computer Science (ICMC) at the University of Sao Paulo (USP). I completed a BA in Social Sciences at USP in 2011 and a MA in Political Science at the University of Sao Paulo in 2013.

My two areas of interest are Methods and Political Economy of Distributive Politics. On the methods side, I am interested in developing empirical (statistical) and formal (analytical) models. I understand empirical modeling as the application of mathematical statistics to develop theory-oriented probabilistic models intended to account for the underlying process that generate the data. I have used this approach in my work, for instance, to develop models to detect fraud in election. Yet in the Methods side, I am also interested in causal inference in the context of social science research, particularly the application of such tools to the study the political economy of distributive politics. The challange here is to have a causal framework that preserves the fruitful idea of Potential Outcome approach and its understanding about identification of causal effect but set it free from the restrictions imposed by its assumptions (SUTVA, etc) and its current limitations to deal with dynamic causation in Social Sciences. In that sense, I am interested in application of Reinforcement Learning and Dynamic Treatment Regimes (DTR), which provide many insights to model social phenomena.

On the substantive side, I am interested in the comparative analysis of the relationship between politics, economic performance, and distribution of income. I have studied the effect of electoral competition and legislative bargain in multiparty systems on the choices of redistributive programs, for instance, on the adoption of programatic interregional redistribution.