Multistate RDF solves fiscal volatility

Therefore, it may be appropriate, or at least predictable, for a state to perform some of the redistributive function of taxation, but costly for it to perform the insurance function. Given that states use, and likely always will use, redistributive programs such as a progressive income tax, we should ask whether there are ways to limit those insurance costs and thereby mitigate some of the painful revenue volatility faced by states, due in part to their use of income taxes.

Because the costs of limited state-level progressive income taxation are related to insurance, rather than redistribution, there is an obvious way to mitigate those costs, at least theoretically: reinsurance. If it is possible to reinsure the state-level risks of a progressive income tax at the national level, then it would be possible in theory to have our cake and eat (at least some of it) too. We could have some state-level redistribution while still having largely national-level income insurance.

In particular, the most direct theoretical solution would be an actual pooled reinsurance fund, also known as a multistate rainy-day fund (RDF). 14 A multistate RDF could pool the risks of state revenue volatility and provide some of the national risk-pooling benefits of a federal income tax. Many states already have their own independent [*93] RDFs, 15 which are essentially a form of self-insurance. The multistate RDF would improve on these, much as pooled health or casualty insurance for individuals is superior to self-insurance.

There has been relatively little work addressing the theory and design of a multistate RDF. What little work there is tends to be limited to basic policy proposals cast as second-best options for managing state fiscal volatility; 16 here, by contrast, I argue that, properly conceived, a multistate RDF would actually be part of an appropriate risk-management system for the combined federal/state tax-and-transfer system.

The legal literature tends to be dismissive of a multistate approach, mostly due to the classic insurance problems of adverse selection 17 and moral hazard. 18 This Article suggests some design features to mitigate those problems. In particular, a key insight of this Article is that some moral hazard, in the form of increasing state reliance on income taxes, is a less of a threat when a purpose of the RDF is precisely to help support state-level redistribution. In the social insurance context, some moral hazard can be "good," in the sense that the behavior being motivated, while costly, is ultimately desirable. 19 Just as health insurance coverage leads to more health care, 20 state revenue insurance could lead to more redistribution. Therefore, moral hazard becomes less of a bug and more of a feature (though one that still requires close management).