This article offers a fresh perspective on a familiar and increasingly pressing question for law and policy: How should the costs of caring for dependent people be divided up between the family and society at large? No domestic policy question of our time carries higher stakes or triggers more divisive public discourse. By focusing on underappreciated features of society's cost allocation task, I construct a theoretical framework for structuring the inquiry. The analysis proceeds in three parts. In Part I, I frame society's cost allocation problem. Of central importance is the convergence of societal and familial interests on the goal of appropriately addressing dependence. This shared goal limits the feasible spectrum of policy choices. When coupled with divergent interests about cost bearing, it sets the stage for strategic interactions between families and society as each attempts to bluff the other into bearing a larger share of dependency care costs. As a result of these interactions, dependence burdens left to fall on the family will not necessarily come to rest on the family. In Part II, I detail the strategies that families might employ in an effort to reduce dependency exposure. These strategies produce results ranging from desirable cost reductions to costly deadweight losses. Working through the resulting set of possible familial responses yields a better understanding of the design choices open to society. In Part III, I show how this analysis generates a more useful way of approaching the question of distributing dependence burdens. The paper concludes with some observations on how we might attempt to harness beneficial familial strategies while limiting destructive cost-shifting efforts.