State redistribution is bad, exacerbates the consequences to recessions because it’s reliant on volatile income tax for revenue,

This Article takes a new look at the tax assignment question and argues that the literature has largely misidentified the costs of state redistributive tax-and-transfer systems. The relevant costs, at least within the range of plausible current policies, are not from tax migration and crippling state tax competition, but rather from poor risk management - suboptimal insurance against income shocks, both for states and their residents. Thus states have some space to perform redistribution, but insurance against low income should be addressed nationally. The problem is that the central tool for both redistribution and income insurance is the same: the progressive income tax.

A redistributive tax-and-transfer system, and a progressfive income tax in particular, not only funds the operations of government and provides for income redistribution, but also acts as a form of insurance against uncertain future income. This aspect of a progressive income tax is often forgotten in analyzing the tax, and that is particularly true in the fiscal federalism literature. 3 Because a progressive income tax provides for relatively high taxes when incomes are high, and for low taxes, exemptions, and transfers when incomes are low, it acts to reduce the volatility in annual after-tax income for taxpayers, and reducing volatility is the classic function of insurance. Thus one can distinguish the redistributive function of an income tax - redistributing income from those with high ability or endowment to those with low ability or endowment - from the insurance function of an income tax - a given taxpayer paying some premium today that is a function of the expected value of potential transfers to her in the future.

In considering the question of where best to assign the progressive income tax in a federal system, one can thus ask what levels of government can best perform each of these two functions - redistribution and insurance. This Article suggests that the answer may be different for each function. Contrary to the classic fiscal federalism literature, there are good reasons for some limited redistribution to be assigned to the states. This is not so for the insurance function, however.

First, regarding redistribution, theoretical work shows that relaxing some of the assumptions in the early work of Richard Musgrave, Wallace Oates, Charles Tiebout, and others leads to the result that redistribution need not be exclusively federal. 4 Furthermore, recent empirical work shows that the risks of tax migration actually may be quite limited, meaning that states may not pay a high economic cost [*91] for implementing some limited redistribution. 5 With the burdens on state budgets increasing, as more and more social insurance and other duties are devolved to the states, this is good news - news that the fiscal federalism literature has not yet fully absorbed.

Second, regarding insurance, even if there is little potential for hurting economic output and state revenues due to tax migration, other costs remain. In particular, a state and its residents are exposed to greater risk and volatility from a state progressive income tax. At the state government level, reliance on an income tax exposes a state to greater revenue volatility, which can exacerbate the costs of an economic downturn, as we have seen. The volatility problem is well understood, and has prompted many calls for states to have better counter-cyclical budget policy. 6

A related cost, which has not previously been articulated in the literature, is the risk to individual state residents. Since a progressive income tax is partly insurance against uncertain future income, insuring at only the state level is simply not as efficient as insuring at the federal level. Because income shocks are often regionally correlated, a regionally defined insurer - that is, a state - is less equipped to smooth out income fluctuations than the federal government is. By pooling the risks of income shocks only at the state level, a state income tax misses out on the actuarial benefits of national risk-pooling present in a federal income tax.

The costs of such suboptimal insurance were displayed during the Great Recession. Following the 2008 financial crisis, state tax revenues dropped dramatically, 7 even while the demands on state social insurance programs rose. 8 This combination had a brutal effect on state budgets, state residents, and the overall economy: Nearly 750,000 state and local jobs were cut, 9 state services shrank, 10 schools [*92] increased class sizes and cut programs, 11 and so on. These effects not only likely slowed U.S. economic growth 12 but also had a profound and lasting human cost. And they can be traced largely to the risk-shifting properties of an income tax, and the revenue volatility that comes with that risk. 13 Thus, there is a strong argument for assigning the insurance function of a progressive income tax to the federal government, which is better able to absorb that risk.