Revised November 6, 2001
Budget Cuts vs. Tax Increases at the State Level:
Is One More Counter-Productive than the Other During a Recession?
by Peter Orszag and Joseph Stiglitz(1)
PDF of report
If you cannot access the files through the links, right-click on the underlined text, click "Save Link As," download to your directory, and open the document in Adobe Acrobat Reader.
States are suffering substantial fiscal stress as a result of the recent economic slowdown. Mississippi, Ohio and South Carolina enacted broad-based spending cuts and eight states raised taxes in enacting their fiscal year 2002 budgets. A number of states are slated to consider additional budget cuts in special legislative sessions that have been called for that purpose. In addition, several governors including those in Colorado, Georgia, Iowa, Maryland, North Carolina, and Vermont have used their executive authority to reduce spending. Many more states are expected to initiate budget cuts when state legislatures reconvene this winter and confront budgets that have fallen out of balance as a result of the downturn.(2)
In all states except Vermont, some form of balanced budget rule forces such counter-productive fiscal policies: When the state enters a recession, revenue naturally falls and expenditures rise. The balanced budget rules then force the state to reduce spending, raise taxes, or some combination thereof, which is counter-productive since it exacerbates the economic slowdown.
Some state policy-makers apparently believe that from a macroeconomic perspective, reducing spending is preferable to raising taxes. For example, the Hartford Courant reported that members of the Connecticut Economic Conference Board have said, "The worst action the state could take would be to raise taxes during a recession; the best course would be to cut spending."(3) Similarly, media reports have suggested that in Florida, "Gov. Jeb Bush, while insisting he prefers to avoid deep cuts in education or health care, says the only thing off the table at this point is a tax increase. Gov. Bush says he believes raising taxes only would exacerbate economic woes by damping consumer and business spending."(4)
Despite these claims, economic analysis suggests that tax increases would not in general be more harmful to the economy than spending reductions. Indeed, in the short run (which is the period of concern during a downturn), the adverse impact of a tax increase on the economy may, if anything, be smaller than the adverse impact of a spending reduction, because some of the tax increase would result in reduced saving rather than reduced consumption. For example, if taxes increase by $1, consumption may fall by 90 cents and saving may fall by 10 cents. Since a tax increase does not reduce consumption on a dollar-for-dollar basis, its negative impact on the economy is attenuated in the short run. Some types of spending reductions, however, would reduce demand in the economy on a dollar-for-dollar basis and therefore would be more harmful to the economy than a tax increase.
In analyzing the economic impact of spending reductions, it is important to draw a distinction between transfer programs (such as unemployment insurance and Social Security) and direct government spending on goods and services (such as purchasing military equipment or building roads). Basic economy theory suggests that direct spending reductions will generate more adverse consequences for the economy in the short run than either a tax increase or a transfer program reduction. The reason is that some of any tax increase or transfer payment reduction would reduce saving rather than consumption, lessening its impact on the economy in the short run, whereas the full amount of government spending on goods and services would directly reduce consumption.
A reduction in government spending on goods and services is thus likely to be more harmful to the economy in the short run than an increase in taxes or a reduction in transfer program spending. Within the sphere of changes to taxes and transfer programs, the impact on the economy depends primarily on the propensity to consume that is, on how much of an additional dollar of income is spent rather than saved among those who receive the transfer payments or pay the taxes. The more that the tax increases or transfer reductions are focused on those with lower propensities to consume (that is, on those who spend less and save more of each additional dollar of income), the less damage is done to the weakened economy.(5) Since higher-income families tend to have lower propensities to consume than lower-income families, the least damaging approach in the short run involves tax increases concentrated on higher-income families.(6) Reductions in transfer payments to lower-income families would generally be more harmful to the economy than increases in taxes on higher-income families, since lower-income families are more likely to spend any additional income than higher-income families. Indeed, since the recipients of transfer payments typically spend virtually their entire income, the negative impact of reductions in transfer payments is likely to be nearly as great as a reduction in direct government spending on goods and services.
For states interested in the impact only on their own economy rather than the national economy, the arguments made above are even stronger. In particular, the government spending that would be reduced if direct spending programs are cut is often concentrated among local businesses. (This can cause distortions in the long run, but it bolsters the local economy in the short run.) By contrast, the spending by individuals and businesses that would be affected by tax increases often is less concentrated among local producers since part of the decline in purchases that would occur if taxes were raised would be a decline in the purchase of goods produced out of state. Thus, more of the reduction in purchases that results from tax increases than from government budget cuts falls on out-of-state goods (relative to in-state goods), lessening the adverse impact of a tax increase on the state economy. Reductions in direct government spending consequently could have a larger adverse impact on a state's economy than tax increases, which have a stronger adverse impact on out-of-state goods and services.
In addition, higher-income families appear to consume relatively more goods and services produced in other regions of the country (or abroad) than lower-income families do.(7) Compared to lower-income families, higher-income families therefore have much lower propensities to consume local goods, both because they have lower propensities to consume overall and because locally produced goods constitute a smaller share of what they purchase. A tax increase concentrated on higher-income families thus is likely to have a smaller adverse impact on the state economy than other budget balancing alternatives. Similarly, a reduction in transfer payments to lower-income families would have a larger adverse impact on the local economy than a tax increase for higher-income residents.
The conclusion is that, if anything, tax increases on higher-income families are the least damaging mechanism for closing state fiscal deficits in the short run. Reductions in government spending on goods and services, or reductions in transfer payments to lower-income families, are likely to be more damaging to the economy in the short run than tax increases focused on higher-income families. In any case, in terms of how counter-productive they are, there is no automatic preference for spending reductions rather than tax increases.
It is worth emphasizing that any state spending reductions or tax increases are counter-productive at this time: they restrain the economy at a time when it is already slowing. Given the existence of balanced budget rules at the state level, some form of federal fiscal relief to states is therefore warranted.(8)
1. Peter Orszag is the Joseph A. Pechman Senior Fellow in Tax and Fiscal Policy at the Brookings Institution. Joseph Stiglitz is Professor of Economics at Columbia University and one of the recipients of the 2001 Nobel Prize in Economics.
2. Iris Lav, "State Fiscal Problems Could Weaken Federal Stimulus Efforts," Center on Budget and Policy Priorities, October 4, 2001.
3. Hartford Courant, October 7, 2001.
4. Wall Street Journal, October 24, 2001.
5. Differences in propensities to consume may be even more important in the context of temporary changes in tax or transfer programs. Lower-income families are more likely to be liquidity constrained - that is, forced to consume only out of income, rather than out of their assets or by borrowing than higher-income families. A temporary reduction in transfer programs (which would disproportionately affect lower-income families) would therefore reduce consumption by more than a temporary increase in taxes for high-income families, since the tax increase would have little effect on lifetime income for high-income families and could thus have little effect on their consumption (since they could temporarily offset the impact of the higher taxes by borrowing or by consuming some of their assets).
6. Dynan, Skinner and Zeldes (2001) show that, in several different data sets, average propensities to consume out of current and permanent income fall as those income measures rise. Parker (1999) uses data from the Consumer Expenditure Survey and finds that the marginal propensity to consume out of transitory income at low levels of resources (which for most low-income households is effectively current income) is much higher than the marginal propensity to consume out of transitory income for very high-income households. McCarthy (1995) uses data from the Panel Survey of Income Dynamics and shows that the marginal propensity to consume out of idiosyncratic income shocks is larger for low-wealth households than for high-wealth households. The weight of the evidence suggests that lower- and middle-income households do have higher propensities to consume out of available resources. See Karen E. Dynan, Jonathan Skinner, and Stephen P. Zeldes, "Do the Rich Save More?" NBER Working Paper 7906, National Bureau of Economic Research, September 2000; Jonathan Parker, "The Consumption Function Re-estimated," mimeo, August 1999; and Jonathan McCarthy, "Imperfect Insurance and Differing Propensities to Consume Across Households," Journal of Monetary Economics, November 1995, pages 301-27.
7. More precisely, less of the value-added embodied in the goods and services consumed by higher-income families occurs in the local area relative to the goods and services consumed by lower-income families. For a discussion of the geographic profile on different household spending patterns, see, for example, Andrew Bernat and Thomas Johnson, "Distributional Effects of Household Linkages," American Journal of Agricultural Economics, 73(2) May 1991.
8. See Iris Lav, "State Fiscal Problems Could Weaken Federal Stimulus Efforts," Center on Budget and Policy Priorities, October 4, 2001, and Peter R. Orszag, "Evaluating Economic Stimulus Proposals,"Testimony before the Committee on the Budget, United States Senate, October 25, 2001.