Despite attempts to paper over the dispute, political scientists in the pluralist tra dition disagree sharply with public and social choice theorists about the importance of institutions and with William Riker in particular who argues in Liberalism against Pop ulism that the liberal institutions of indirect democracy ought to be preferred to those of populist democracy. This essay reconsiders this dispute in light of two ideas unavailable to Riker at the time. The first, offered by Russell Hardin, is that constitutions can more usefully be conceptualized as coordinating devices as opposed to social contracts. The importance of this idea is that it allows for a more theoretically satisfying view of the way that constitutions become self-enforcing. The second idea, which derives from the various applications of concepts such as the uncoveted set, argues that although institutions such as the direct election of president are subject to the usual inabilities that concern social choice theorists, those instabilities do not imply that "anything can happen" -instead, final outcomes will be constrained, where the severity of those constraints depend on institutional details. We maintain that these ideas strengthen Riker's argument about the importance of such constitutional devices as the separation of powers, bicameralism, the executive veto, and scheduled elections, as well as the view that federalism is an important component of the institutions that stabilize the American political system.We conclude with the proposition that the American Civil War should not be regarded as a constitutional failure, but rather as a success. In framing a government which is to be administered by men over men, the great difficulty lies in this: You must first enable the government to controul the governed; and in the next place, oblige it to controul itself. Riker bases his analysis on the research of social choice theorists such as Arrow (1963), Plott (1967, Davis and Hinich (1966), McKelvey (1976), and Schofield (1978. This research reveals two things. First, coherence (in the form of a well-defined social preference order over feasible policy outcomes) in majoritarian processes without impediments to direct citizen control of policy requires a nearly-impossible-to-achieve balance of citizen preferences on salient issues, sufficient uniformity of tastes and perceptions such that preferences on all issues (potential or otherwise) can be mapped to a single "ideological" dimension, or the existence of a unique salient issue. Second, democratic procedures such as two candidate elections and committee voting are arbitrary in the sense that outcomes and the identities of winners and losers depend on procedural details, including the order with which election candidates announce their campaign platforms, the order with which alternatives are considered in a legislative voting agenda, the identities of persons who have access to an agenda, and the list of feasible policy alternatives.From these results Riker constructs the argument ...