Graduate Students

George Derpanopoulos

Contact Information

Office  Not Available
My dissertation, “Networks and Dictatorships: Essays on Economic Predation, Coups, and Dictators’ Fate”, adopts a network approach to study three central topics in authoritarian politics: (i) economic predation by dictators on elites, (ii) military coups, and (iii) dictators’ post-tenure fate.

Essay 1: "Elites, Financial Networks, and Constraints on Dictators: Evidence from the Panama Papers" (email for draft; presentations: APSA 2017, MPSA 2017, UCLA Graduate Student Conference 2017)

Abstract: A large literature argues that dictatorships can achieve high levels of economic growth if dictators can commit to not expropriate elites. Extant research has focused on the role of formal institutions—legislatures and parties—in helping elites constrain dictators’ predation. I complement this literature by documenting the role of an informal institution, financial networks, in constraining the dictator. I argue that dense financial ties among elites serve to diffuse private information on the economy, hence facilitating elites’ monitoring—if the dictator reneges on his commitment, informed elites are able to infer and punish his defection. This credible threat deters the dictator from predation and commits him to sharing rents with elites. Accordingly, I hypothesize that dictatorships where elites’ financial network allows for greater diffusion of private information enjoy stronger property rights. To test my claims, I uncover networks of elites’ co-ownership of offshore companies—a strong type of financial tie—using the largest leak of financial information to date, the Panama Papers. A statistical analysis of all cases during 1990-2015 supports my hypothesis.

Essay 2: "Factionalism, Communication, and Coordination: The Role of Military Officer Networks in Coups" (in progress)

Abtract: Why do some military officers participate in coups, while others stay in the barracks? Why are some coup attempts successful, while others fail? How does military factionalism affect coup dynamics? I tackle these questions through a formal model of coordination under limited information. In my model, officers only participate in coups they believe will succeed—coups where the expected number of participants exceeds their belief about regime strength. These beliefs are private, but officers communicate them to other officers in their social network. I model how the structure of the officer network affects individual participation in a coup and, ultimately, the coup’s incidence and success. A feature of my model is that coup attempts require the existence of a faction/clique–a group of officers who are all connected to each other–whose size exceeds regime strength. This is because, within a faction, there is common knowledge that all of its members expect they can coordinate to topple the regime. I explore the effect of factionalism–a reduction in average faction size–on the incidence of coup attempts. My model explains why the literature has been unable to show a consistent statistical relationship between factionalism and coups: because the effect of factionalism varies by regime strength–positive for relatively weak regimes, negative for relatively strong regimes–empirical analyses that pool regimes of differing strength will find a null effect. I test my theory using novel archival data on Greece (1945-1967). I estimate officers’ social network using information on their education and assigned positions–two officers are coded as connected if they overlapped in their studies or assignments.

Essay 3: "Violent Leader Change, Exit Options, and Dictators’ Fate: The Role of International Relations" (in progress)

Why do some dictators that are deposed through violence go into exile or secure amnesty from prosecution, while others are imprisoned or executed? Answering this question can explain why dictators from countries experiencing ongoing violence often refuse to abandon power—if they cannot secure exile or amnesty, they can delay imprisonment or execution by clinging on to a collapsing throne. I underline the importance of international relations and dyadic factors between countries in shaping dictators’ fate. I argue that a dictator (from a country) more densely interconnected with other leaders (countries) can expect a better post-tenure fate. Two types of dyadic factors shape a dictator’s fate after office. First, the dictator’s personal ties to other countries; namely, the number of foreign leaders that are his friends or relatives and the wealth he holds abroad. Second, ties between the dictator’s country and other countries; specifically, shared culture (ethnicity, language, religion), military and diplomatic alliances, trade, and geographic distance. Ties can have demand- and/or supply-side effects. The former reflect an ousted dictator’s willingness to seek exile in a country that he is connected to. Supply-side effects capture a country’s willingness to host an ousted dictator whose country it is connected to, or to pressure that country to grant the dictator amnesty. Ties between dictators(hips) and other countries become more important in an era of expanding jurisdiction for international courts. Paradoxically, by imposing costs on member-states that protect ousted dictators, particularly the most repressive ones, international courts might make these dictators even more hesitant to give up power, thereby prolonging their repression. In such cases, ties may have a welfare-improving role: foreign leaders and countries strongly connected to an ousted dictator and his country may still find it rational to defy international courts and protect that dictator, hence incentivizing him to exit office. I find support for my argument through a statistical analysis of all relevant cases since 1945.


MS Statistics, UCLA (expected 2017); MPhil Economics, University of Cambridge (2011); BSc Economics and Politics, University of Bristol (2010)

Fields of Study

Comparative Politics: Authoritarian Regimes, Political Economy | Methodology: Social Network Analysis, Machine Learning


I study comparative politics and political methodology. In the former subfield, my research interest revolve around authoritarian politics, particularly the determinants of regime stability, leader turnover, and coups. In my dissertation, I use a network approach to model authoritarian regimes, and test the implications of my theory using originally collected cross-national data. In political methodology, I am interested in network analysis and machine learning, with a focus on network games and regularization, respectively.

Grants and Awards

  • Humane Studies Fellowship, Institute for Humane Studies (2017-2018, 2016-2017)
  • Adam Smith Fellowship (Research Track), Mercatus Center (2016-2017)
  • PhD Scholarship, Institute for Humane Studies (2015-2016)
  • Adam Smith Fellowship, Mercatus Center (2015-2016
  • Dissertation Year Fellowship, UCLA Graduate Division (2017-2018)
  • Summer Fellowship, UCLA Dept. Political Science (2017, 2016, 2015)
  • Graduate Research Mentorship, UCLA Graduate Division (2014-2015)
  • Graduate Summer Research Mentorship, UCLA Graduate Division (2014, 2013)
  • Graduate Fellowship, UCLA Graduate Division (2012-2013)


  • American Political Science Association Annual Conference (2017)
  • Midwestern Political Science Association Annual Conference (2017, 2016)
  • UCLA Graduate Student Conference in Comparative Politics (2017, 2015


Barbara Geddes



Working Papers

  • Occam’s Lasso: Parsimonious Theory-Building Needs Machine Learning (in progress)