This Article critiques the prevailing justification for subsidies for the charitable sector and suggests a new alternative. Existing rationales are based on an economic model that assumes a single government whose decisions are guided by a single median voter. I argue that this theory is unpersuasive when translated to federal systems, such as the United States, in which there may instead be thousands of competing local governments.
I then attempt to construct a theory of the charitable sector that takes account of interactions between charity, local government, and national government. In this revised account, charity is most important when federalism mechanisms break down. For example, frictions on exit produce too little jurisdictional competition, and excessively easy exit produces too much competition—a race to the bottom. In both these cases, the quality of the resulting government services is predictably low, so that charity can be expected to outperform rival governments. Even if not, the threat of the charitable alternative may supply competition that is otherwise missing from the market for government services. These conclusions also have implications for the law of charitable organizations, as I detail.