How is a Blizzard Like Unemployment Insurance?, or, Should States Save for Disasters?

Living in New England, it's easy to get distracted by snow.  I started blogging last week about unemployment insurance, but I want to divert for a bit to talk about some recent news coverage about how many state budgets are being further strained by winter weather cleanup costs.  As it turns out, the two pose the same question: to what extent should state or local governments save up to pay for dealing with big, uncertain events? 

One thing we know for sure is that states in fact don't save.  More than 30 states are expected to have to borrow money from the federal government to cover the costs of paying unemployment insurance benefits this year.  And although nearly every state has a "rainy day fund" to cover unexpected shortfalls (whether climatological or otherwise), none of them save enough to actually do much good. 

This shouldn't be surprising.  Mobile voters probably don't want to invest too heavily in preparing for contingencies they might not be around to see.  And politicians probably don't want to raise taxes now in order to hand money over to the person who defeats them in the next election.  Although in both cases there are stories one could tell that would cut the other way--state savings should increase home values, and surplus funds lower public borrowing costs--it seems likely that either officials or their constituents will be more grasshopper than ant.    

Problem is, all we know for sure is that states don't save; we don't know for sure why.  And it turns out we probably need to know that in order to decide whether states should save, or whether instead the federal government should save for them.  (An example of the latter,  as I mentioned in my earlier post, is that the federal government can lend money to states that can't afford to pay their UI claims.)

Why does it matter?  Well, if states can't save because of exit pressures, then fiscal disasters look a lot like other kinds of redistribution: it's more efficient to do it at the federal level, where you don't drive people and firms from one place to another just because of the local tax policy.  On the other hand, if the problem is officials, then federal savings don't have this benefit, and create a crucial cost: they induce moral hazard among local officials, who know that they will be bailed out of any serious mistakes.

So, long story short, I don't think it's clear whether states should save up for storms real or metaphorical.  What do you think?