Legal Commentary

Rick warns of the dangers of federal bailouts of local governments.  But of course the reason bailouts are tempting is because financial crises at lower tiers of government are painful, and not just for the folks in that jurisdiction.  Way back at the beginning of my guest stint, I explained why recessions are hard on state and local governments --- their revenues drop together with the economy --- and why those shortfalls are bad for the national economy.   I also noted that it's tough for states to pluck themselves up by their own bootstraps, since raising taxes and/or borrowing is tough to do at the local level, especially when your neighbors aren't doing it.  So, putting together Rick's post and mine, what we need is a federal support for states, but one that 1. makes states pay part of the bill, so that moral hazard is constrained; and 2. turns off automatically when the state gets out of the recession. 

Turns out we have one of those already, as Jon Klick and I show in our draft paper. 

The solution, bizarrely, is the Alternative Minimum Tax.  (Hah! I got you to click "continue" to read a tax post! )  The federal tax system (in section 164) gives state and local governments a subsidy, by allowing local taxpayers to deduct from their federal taxes the amount of tax they pay to other sovereigns.  (This is also a form of foreign aid, although it's not usually counted as such in federal budgeting).  But since a $1 deduction only reduces the actual federal tax you pay by $1 times your marginal rate, the effect is a matching grant: for every dollar the local taxpayer gives to her government, the feds give her back $.35 or so.  The local government can increase its own taxes to capture some of this subsidy.  

The AMT comes in by disallowing this state-and-local-tax deduction for high-income taxpayers.  So, effectively, when the local economy is doing poorly, and incomes are low, the federal government kicks in a bigger grant.  When the economy heats up again, and incomes rise, the grant drops off.  So state taxes and expenditures should respond to the number of people in the jurisdiction who lose their deductions because of the AMT.   Or at least that's the theory.  Does it actually work that way? 

We find it does.  For example, in states where the % of taxpayers who lose their deductions is in the top 1/6th of the nation, overall spending on social insurance programs is 4% lower.  For medical care, it's dramatically lower: more than 10%. 

But this "automatic stabilizer" effect of the AMT is just a happy accident, so there are lots of ways in which the tax rules aren't really set up to maximize its efficacy.  We suggest some ways to do that, too.