Conservatives Must Rethink Spending Caps

In 1992, Colorado voters enacted the Taxpayer Bill of Rights ("TABOR"), a state constitutional provision that limits spending increases to inflation plus population growth, unless voters approve a greater increase. This reform touched off a wave of efforts by anti-tax activists to enact similar limits in other states. However, 17 years later, Colorado stands alone as the only state with a TABOR-like spending limit.

Last week, Maine residents voted on a TABOR-inspired measure called the Maine Tax Relief Initiative. Proponents argued that the initiative would control outsized growth in government spending and let taxpayers keep more of their money; opponents (chiefly, public employee unions) alleged that the initiative would require cuts in basic services.

The Maine vote did not break the national trend: voters bought the union arguments and rejected the question by a 3-2 margin, more resoundingly than they'd defeated a similar proposition three years ago. Voters in Washington defeated a similar proposal last week, as did Oregon and Nebraska electorates in 2006. In the last two decades, California and Washington have eviscerated spending limits that formerly contained government growth.

Maybe the critics have a point. The problem with the TABOR resolutions that fiscal conservatives keep proposing is that they seek to shrink government as a share of the economy, every year, forever, meaning that government cannot maintain current service levels without voter overrides. A refined spending cap that uses a looser index for expense growth could avoid this perverse outcome while still restraining government --and allow proponents to better rebut scare tactics at the ballot box.

While other states have previously and subsequently capped taxes or expenditures, Colorado's TABOR stands out for its stringency as a spending cap. This is in part due to TABOR's use of inflation plus population growth as a spending increase index. Unless voters pass overrides, TABOR holds spending per capita fixed in real terms.

Since enactment, Colorado has seen notably slow growth in state and local expenditures as a share of income. On this measure, Colorado stood at 98% of the national average in 1990, but dropped to 87% by 2006, despite a voter-approved suspension of TABOR for several years this decade.

Because the economy grows in real terms over time, TABOR requires consistent reductions in government spending as a share of the economy. From 1967 to 2006, American population growth plus inflation totaled 797%. But personal income growth over this period was 1,639%, reflecting the fact that real income per capita grew 94%. If every state had enacted a TABOR-style limit on expenditures in 1967 and no overrides passed, state and local expenditures would have fallen from 16.4% of income to 8.4% over that period, even as they remained fixed in real, per capita terms.

It's hard to believe this is what voters want, to cut government so drastically relative to economic growth. It's more likely that voters want to use some of their increased productivity to grow government in real terms over time, just like private consumption grows -- which would indicate that personal income is the most appropriate index for a spending cap.

Michigan has had a spending cap in effect since 1978, but unlike TABOR it is indexed to personal income. The provision also includes a rainy day fund, which the state must replenish when income is rising quickly and can draw down when the economy slows. Dr. Barry Poulson of the University of Colorado, a key proponent of TABOR-style reforms, notes that this reform has been more effective than TABOR in smoothing budget cycles.

One thing a personal income-based cap won't do is shrink state and local government as a share of personal income. That government will grow at an appropriate pace provides little comfort if we start from a baseline that is too high. And indeed, Poulson notes that the Michigan reform has been ineffective in shrinking government. Where spending is already too high, a hybrid reform is called for.

States where government starts off too large should enact spending caps that closely restrict spending until it reaches a specified percentage of personal income, after which the cap would be linked to personal income. (The personal income measure should be a multi-year average, to avoid cyclicality in government spending.) This reform would produce an inverted hockey stick graph of spending as a share of income: falling until it reaches the target, then flat thereafter.

New York, where taxes per capita are 160% of the national average, would be well-advised to adopt such a reform. A state with a limited public sector, like South Dakota, might judge that its size of government is already appropriate and that there is no need for an initial shrink-down period.

Proponents of small government will be skeptical of the idea that a less-stringent spending limit is a preferable alternative to TABOR. If voters see TABOR as too strict, they can pass annual overrides to grow government faster. Why build real government growth into the system?

There are several responses to this. One is that TABOR is premised on the idea that institutions matter, and that a spending cap will have real effects on the size of government. In light of that, we should not set an overly strict limit with the expectation that voters will tinker with spending levels later.

Second, extremely stringent spending limits may not be sustainable. While Michigan's less-strict reform has survived for 30 years, Washington and California did away with stronger spending caps only about a decade after their enactments, as voters came to believe that the caps were reducing the quality of government services. Colorado's TABOR has been under constant political fire and was significantly weakened by referendum earlier this decade. A cap that sharply reduces spending may be a victim of its own success.

Third, voters have no apparent appetite for TABOR as currently structured. All five TABOR measures advanced in the last four years went down to defeat, with "no" votes ranging from 54% to 71%. Allowing growth with personal income would enable proponents to better rebut claims that a cap will degrade public services, making enactment more likely.

This last fact should make a personal income-linked cap especially appealing to beleaguered TABOR supporters. After all, a spending cap can't be strong or effective if it never becomes law.