The role of citizens' initiatives figures prominently in contemporary debates on constitutional change. It is widely believed that permitting initiatives should improve the congruence between citizen preferences and policy outcomes across the spectrum of issues on which initiatives may be placed. This paper investigates the theoretical basis for this view. It begins by identifying three basic reasons why electoral competition may not, by itself, be sufficient to ensure congruence on specific issues. Each reason relies critically on the fact that citizens have only one vote to cast for candidates who have responsibility for choosing a bundle of issues. It then shows how allowing initiatives permits the unbundling of specific issues which improves congruence when the three reasons apply. Important caveats to this logic are also presented.