08/01/2009 05:12 am ET | Updated May 25, 2011

We Could Get the Third Stimulus Wrong

As state governments across the country grappled this week with painful decisions about tax increases and service cuts to close budget deficits, stimulus watchers raised serious concerns about the deleterious effects that these deficit-closing measures will have on economic recovery. Stephanie Kelton pointed out that "Jobs that are being created (or saved) through the left hand of the Obama stimulus package are disappearing at least as rapidly as the right hand slashes billions from state budgets." Indeed, state and local government purchases declined $78.8 billion in the last six months.

These claims are used to argue for additional fiscal assistance for state governments. One popular means of providing this assistance is the now-defunct General Revenue Sharing, a Nixon administration program ended by Reagan that directed federal funds to state and (at that time primarily) local governments with essentially no strings attached. James Galbraith was an early proponent of resurrecting GRS to mitigate the economic downturn.

As much as state fiscal relief does stimulate the economy and as much as a third stimulus might be necessary, GRS has two important flaws that point out a larger problem with the stimulus package: how we can effectively target stimulus funds to the people and institutions that need them most.

First, GRS's allocation formula is problematic. The Wall Street Journal suggests today that so far stimulus aid has been insufficiently targeted to the states with the most need, that is, to those with the highest unemployment rates. An earlier San Francisco Federal Reserve research note elaborated on exactly which portions of the stimulus package are more and less targeted to needy states. The parts of the stimulus that are well-targeted are those with formulas that proxy for "rapid reversals in economic fortunes." This is true of the increased federal matching grant for Medicaid, which uses the rise in the state's unemployment rate as one variable in the formula. The parts that are rather poorly targeted are those with formulas based primarily on population. This is true of the state Fiscal Stabilization Fund, which uses population as its primary variable.

A January Congressional Research Service report on GRS suggests that the grant program would be only modestly successful at targeting aid to needy states. Of the six states with budget deficits more than 20% of their general funds -- Arizona, California, Alabama, Florida, Illinois, and Rhode Island -- only three would receive more per capita than the country as a whole from a hypothetical GRS allocation of $40 billion. This hypothetical allocation would provide California with less than 25% of the funds it needs to close its budget gap; Rhode Island would receive around 30%.

GRS's relationship to state unemployment is no better. Five of the ten states (including D.C.) with the highest unemployment rates would receive less than the country's per capita average from the hypothetical GRS allocation of $40 billion. In contrast, eight of the ten states with the lowest unemployment rates would receive more than the country's per capita average. GRS would certainly help states suffering from budget shortfalls. But its reliance on population, taxes, and income rather than on variables more directly associated with the economic downturn -- such as the unemployment rate or, as John Judis might suggest, the number of unemployed -- makes it a less-than-ideal allocation mechanism.

The second problem with GRS is that it undermines the very component that is indispensable to an effective stimulus package: coordinated spending between federal, state, and local governments. By allocating funds directly to state governments with few strings attached, GRS gives federal officials little, if any, say in how funds will be expended. Rather than utilize federal resources -- ranging from experience with grant programs to less susceptibility to parochial interests -- to maximize the investment potential of federal funds, GRS promotes spending decisions based on local politics and short-termism.

The need for federal aid for states is indisputable. But there are better ways to target stimulus funds than with a grant program that will help the states in most need only modestly and will damage the already tenuous coordination between federal, state, and local officials that is critical to economic recovery.

After all, the third stimulus package should be the best so far.