This article describes efforts to develop an exploratory agent-based model as a tool for studying decision making in political regimes such as Iraq, North Korea, and Syria. Our hybrid of the landscape metaphor and the rule-based system approach captures the trade-offs leaders face in balancing components of a utility function, plus risk profiles that allow departures from conventional utility maximization. Two simple experiments concerning succession demonstrate the surprising compromises both leaders and elites are willing to make, as well as the instability of these bargains.2008 Wiley Periodicals, Inc. Complexity 14: [36][37][38][39][40][41][42][43][44] 2008 1. INTRODUCTION A utocratic governments have long drawn interest from both scholars and government analysts, seeking to understand better how they operate and potentially to anticipate the sorts of decisions they are likely to make in domestic and foreign policy contexts. Over recent years, specific attention has been paid to countries such as Afghanistan, Burma, China, Cuba, Iran, Iraq, Libya, North Korea, Syria, and Zimbabwe. During previous eras, the focus at different points was on the fascist systems in Germany and Italy; the Communist monopolies of the Soviet Bloc; and military juntas and other dictatorships across Latin America, as well as in Africa and Asia. A common feature of these governments is severe restrictions on openness, pluralism and competition in many spheres of society.Comprehending the dynamics of such closed political regimes presents a challenge for traditional analytical methods. As a general matter, these regimes exhibit outcomes that are difficult to explain or trace back to their causes. In particular, the actions of their leaders often appear to be conditional on subtle variations in circumstances, obscure, unpredictable, or even illogical. When one factors in the secretive, insular nature of these regimes and the intricacies of and differences This is an invited paper for the Special Issue-Security and Complexity, Guest Editor: Jürgen Scheffran. among the constituent actors and the settings in which they operate, it can be difficult to achieve adequate explanatory leverage with statistical, mathematical, and game-theoretical techniques, informed by the standard means of collecting quantitative and qualitative primary data. We opt instead for an analytical strategy that involves considering closed regimes as complex adaptive systems (CAS). In our view, this alternative makes sense because these regimes have the hallmarks of CAS, which include (a) heterogeneous agents, (b) interacting with each other and their environment in nonrandom ways, (c) often using non-linear rules to decide what actions to take, (d) both creating and responding to positive and negative feedback processes, and (e) sometimes adapting and learning to achieve better results over time. Various types of computer simulations permit one to capture these features and to explore their association with the dynamic properties of the system.For purposes of stud...