This week, Dave Altig and John Robertson of the Federal Reserve Bank of Atlanta published a fascinating post on increased volatility in the revenues of state and local governments at the state and local level. Arpit Gupta kindly emailed the link. They draw on the work of UCLA law professor Kirk Stark, who argues in a recent article in the Virginia Tax Review that the federal government has contributed to this rising volatility in a number of ways:
[T]he basic picture is that under current law, the federal government generally favors the adoption of state individual and corporate income taxes (by virtue of both the administrative benefits associated with base conformity and the price effects associated with federal income tax deductibility) and to a lesser degree property taxes (where there are no base conformity benefits but there is a positive price effect due to deductibility). Additionally, federal law currently disfavors the adoption of general sales taxes (by virtue of the lack of any base conformity benefits and the usual lack of any price effects from deductibility). As for the distribution of tax burdens among income classes, federal law currently gives states an incentive to concentrate the tax burden on those individuals who are best positioned to take advantage of federal subsidies made available through income tax deductibility. Thus, all else equal, states should favor the use of deductible taxes imposed on federal itemizers (subject to the AMT caveat described above), as well as taxes on high-income taxpayers who likely face the highest federal marginal tax rates. …
Current federal law can be viewed as contributing to state revenue volatility in several respects: First, the inclusion of corporate income and non-wage personal income in the federal base offers up an easy, almost irresistible tax base for state governments. Second, the general federal preference for more progressive tax systems (because of the differential value of deductibility) likely contributes to the volatility of state revenue. Third, federal rules relating to the taxation of internet and mail-order sales generally inhibit the development of broad-based personal consumption taxes in the states. In short, federal law generally favors the adoption of tax structures that contribute to — perhaps even exacerbate — revenue volatility. [Emphasis added]
Stark then offers an illuminating discussion of how these federal incentives shaped tax reform in California, a state known for its high level of revenue volatility, before offering options for reform. His normative starting point for reform is that the federal government should strive to respect the fiscal sovereignty of state and local governments:
Adherence to a principle of base composition neutrality might have both negative implications for federal policy (e.g., the federal government should not confer differential subsidies for different types of taxes) as well as positive implications (e.g., the federal government should strive to provide comparable administrative support for sales and property taxes as it provides for income taxes). The overarching principle is one of deference to, and respect for, the political choices of subnational governments.
This leads Stark to my favorite tax reform:
Perhaps the most straightforward federal policy change recommended by the analysis above would be to reform or repeal the deduction for state and local taxes. Because the deduction favors the adoption of progressive income taxes, and such taxes have been implicated in the revenue volatility states experience, a rethinking of this longstanding subsidy would seem to be in order. Repealing the federal subsidy does not guarantee that state and local governments would reduce their reliance on more volatile revenue sources. In addition, it bears noting that outright repeal would also involve elimination of the deduction for property taxes — a relatively stable source of revenue. Nevertheless, unless state and local governments are wholly unresponsive to the price effects conferred by deductibility (and empirical studies suggest that they are in fact responsive to those effects), it is likely that repealing § 164 would result in a substitution away from currently favored taxes.
One less dramatic alternative would involve offering a flat-rate credit.
Stark also recommends (wisely, in my view) centralizing particularly volatile tax bases, like corporate income and individual income taxes on nonwage income, e.g., capital gains. The reason is simple: “mobile tax bases should generally be assigned to higher levels of government.” Moreover, centralization could involve earmarking some of the revenue for state and local governments:
Congress could impose a special surcharge on capital gains income, the proceeds of which would be earmarked for federal grants to the states. The object is not necessarily to reduce the tax burden on investment income, or even to reduce the volatility of tax payments. Rather, the idea is to shift the revenue volatility to the level of government that is better positioned to handle it.
Right now, as Stark explains, the federal investment in defining the personal and corporate income tax bases makes it attractive for state governments to “free-ride”; Stark suggests that the federal government offer administrative support for non-income taxes as well:
A federal agency charged with developing state-of-the-art expertise on valuation techniques, compliance efforts, and other features of property tax administration might offer a useful counterweight to the benefits for income tax administration currently generated by the Service.
Suffice it to say, this might not go down well with all conservatives. But the ultimate goal is to help make the states less likely to go bust and demand bailouts, which seems quite sensible.
The Stark approach would, in encouraging the adoption of a more stable revenue base for state and local governments, make the tax system at the subnational level somewhat more regressive, which would strengthen the case for maintaining a somewhat more progressive tax system at the national level.