Income tax receipts are highly cyclical, but sales taxes are pretty cyclical too, and performed especially badly in the recent downturn. Source: Center on Budget and Policy Priorities.
Income tax receipts are highly cyclical, but sales taxes are pretty cyclical too, and performed especially badly in the recent downturn. Source: Center on Budget and Policy Priorities.

Revenue volatility is a real problem for states and cities. Recessions cause tax revenue to decline and demand for services to rise. Because states and cities must (roughly) balance their budgets each year, this leads them to cut spending and raise taxes at the least opportune time in the economic cycle. Stable sources of tax revenue that reduce these pressures are desirable.

Elizabeth McNichol of the Center on Budget and Policy Priorities, a liberal economic think tank in Washington D.C., has a new paper urging states not to shift from income taxes to sales taxes as a strategy for addressing revenue volatility. She makes four main points: Sales taxes, while less volatile than income taxes, still can experience significant declines in recessions (see the chart); sales taxes face a long-run revenue insufficiency problem, as an eroding tax base means that receipts do not keep up with economic growth; sales taxes are regressive; and there are better strategies available to manage volatility.

You can think of McNichol's paper as a belated indictment of Louisiana Governor Bobby Jindal’s now-dead tax reform plan, which would have repealed the income tax while expanding sales tax to more items and raising the rate. McNichol's diagnosis is right for much the same reason that Jindal’s plan was ill-conceived. But I'd offer a different set of strategies for addressing volatility than McNichol does.

McNichol calls first for better management of state reserve funds: Divert more revenue to these when times are good and draw on them when revenue is weak. This is a consensus view, but unfortunately difficult to put into practice. Politicians, whose terms of office are not infinite, are always tempted to underfund reserves so they can hand out goodies, whether in the form of tax cuts or higher spending.

On designing taxes to reduce volatility, she suggests a few things: diversifying taxes, including more reliance on severance (natural resource extraction) taxes; not interfering with local property taxes; and temporary income tax increases and rebates to smooth out the revenue trend. The first and last suggestions are problematic.

McNichol says "severance taxes, while highly volatile, are likely to be counter cyclical." But the Federal Reserve Bank of Kansas City study she cites on this point did not reach a statistically significant finding on the counter-cyclicality of severance taxes. We have just gone through an economic cycle where severance taxes were strongly counter-cyclical: Natural resource prices were strong and production was rising, while the overall economy was weak. But declines in natural resource prices can drag down both severance tax receipts and the broader economy in resource-rich states at the same time: Think Texas during the 1980s oil bust. As such, increased reliance on narrow and volatile taxes like severance taxes could expose states to greater fiscal risk.

As for temporarily raising income taxes in recessions and sending rebates when the economy is strong, there are two problems with this strategy. One is that it is procyclical: Tax increases in recessions are harmful to the economy, while rebates sent when revenue is soaring provide stimulus when it is least needed. Tax increases and rebates also may become more permanent than they are intended to be: The withholding of property tax rebates in New Jersey has proven politically challenging over the last few years, when the state can ill-afford to pay them, and New York’s "temporary" income tax increase on high earners is looking more and more permanent.

McNichol's second suggestion is much better, and I would expand on it. Property taxes are great. They are a highly stable source of revenue, only modestly regressive and less damaging to the economy than income or sales taxes. Property taxes are efficient for the same reason they are hated: They are extremely difficult to avoid. The old adage is "tax something and get less of it," but there’s no way to increase or decrease the amount of land.

Taxes on improvements (buildings and the like) do distort the economy, but not that much because once you’ve built a building, you can’t move it in response to unfavorable tax policy.

Most states should be encouraging localities to rely more on property tax to finance themselves. Liberals are sometimes squeamish about this approach, seeing it as a way to offer excellent local government services only to jurisdictions with high property values. But some states (like Massachusetts and New Jersey) show that it is possible to rely on property tax as the sole major local tax source while using a progressive local aid formula to shore up local budgets in poor jurisdictions. California, for example, could learn a lot from this.

And states should consider statewide property tax as a stable revenue source to offset income and sales tax volatility. A statewide property tax would be more progressive than local property taxes, as it would effectively be a way to redistribute from high property value jurisdictions to low property value jurisdictions. And it would be a way to reduce revenue volatility without reducing progressivity, if the new property tax came as an offset for sales tax instead of income tax.

(Josh Barro is lead writer for the Ticker. Follow him on Twitter.)