Reflections on Distributive Politics and Universalism

minimal winning size.' Another is that distributive programs organized by strong parties are likely to exhibit greater efficiency (Weingast, Shepsle, and Johnsen 1981; see also Cox and McCubbins 1993). In retrospect, this literature exhibits a number of limitations. First, although it is strongly rooted in case-study style empirical observation of distributive programs, there have been few systematic tests of its hypotheses, such as universalism. Second, theoretical approaches have, by and large, ignored the complexities of distributive politics. Is universalism the sole principle necessary to understand this type of politics? If not, what other principles are relevant? Recent work attempts to rectify both weaknesses. The strength of Stein and Bickers' (1994a) article, Universalism and the Electoral Connection: A Test and