تمرکززدایی و مشاغل حرفه ای سیاسی Decentralization and political career concerns

1. Introduction The literature on federalism has evolved through what Qian and Weingast (1997) and Weingast (2009) refer to as two generations of development. The first generation treats each governmental unit as a benevolent social planner, and examines the costs and benefits of decentralization in terms of scale economies, inter-regional spillovers, heterogeneity across regions, etc. The second generation recognizes the incentive problems of politicians at different governmental units, and examines how the degree and the form of decentralization affect their incentives. Despite these developments, much of the second-generation literature on federalism assumes that politicians at subnational governments are surrogates of their corresponding regions, and represent their regions (however imperfectly due to incentive problems) when they interact with other politicians to make collective decisions at the national government (Besley and Coate, 2003; Coate and Knight, 2007; Knight, 2004; Luelfesmann et al., 2015). Such an assumption is appropriate when we study legislators in a democracy (such as senators in the U.S.), but less so when the politicians concerned are administrators, whose political career paths typically start at some subnational governments and end (hopefully) at the national one. For such politicians, their political career concerns are part and parcel of their incentives (Myerson, 2006). In an autocracy like China, where politicians at subnational governments are not subject to electoral checks and balances, such political career concerns are arguably even the dominant, if not the only, forces that incentivize these politicians to perform (Li and Zhou, 2005; Xu, 2011). The omission of political career concerns hence renders the secondgeneration literature on federalism especially inadequate in studying the degree of decentralization in an autocracy.