Federalized Party Systems and Subnational Party Competition: Theory and an Empirical Application to Argentina

Abstract

In 1949 V.O. Key wrote about the importance of state-level one-party in the southern United States for organizing local authoritarian rule in a nationally competitive party system.1 Key's study documented a phenomenon that continues to pose theoretical puzzles to contemporary scholarship on party systems: the simultaneous existence of competitive party politics and noncompetitive party politics in one national party system. In addition to documenting U.S. party system dynamics at the subnational level that were distinct from those at the national level, Key also uncovered important institutional interac tions between noncompetitive state party systems and the competitive national party system. These findings (and many others that followed about U.S. state party politics) pro vided significant possibilities for theory building about parties and party systems. How ever, this theoretical promise was stifled by two subsequent developments in political science. The first was the impermeability of boundaries between American and com parative politics, which relieved Americanist scholars of the burdens of generalization and comparative theory builders of the burden of paying close attention to U.S evidence. The second was the theoretical development of comparative literatures on party systems, whose most influential scholars overlooked or rejected the incorporation of subnational contexts into their theorizing about party systems. As a result, scholars of American politics developed an extensive empirical literature on state party politics while com parative theorizing about parties and party systems remained oblivious to the theoretical implications of this trend. Today there is new interest in how and why the quality of democracy varies across subnational territorial units of countries.2 Party system dynamics are a crucial piece of the puzzle. However, the comparative literature on parties and party systems offers few theoretical tools to scholars interested in this topic. This is because in that theoretical tradition party systems are conceived of and measured nationally. Their systemic prop erties are assessed at the national level, and the indicators used to measure those proper ties (usually votes for national offices or seats in national legislatures) are national. This practice has created a situation of conceptual and measurement incompleteness that hinders new discoveries in the study of party competition across jurisdictional bound aries of the nation-state. 21