Revised March 26, 2001
Tax Cuts Vs. Spending Increases:
Is There a Basis for Chairman Greenspan's Preference for Tax Cuts?
by Jason Furman and Peter Orszag
View PDF of full report
View HTML of factsheet
View PDF of factsheet
If you cannot access the file through the link, right-click on the underlined text, click "Save Link As," download to your directory, and open the document in Adobe Acrobat Reader.
In testimony before the Senate Budget Committee on January 25, Federal Reserve Board chairman Alan Greenspan argued that tax cuts are preferable to spending increases as a means of avoiding "excessive" surpluses.(1) This paper examines Chairman Greenspan's argument.
If one were to accept Chairman Greenspan's proposition that projected surpluses are excessive and need to be reduced (a proposition that is highly debatable and that a number of leading economists have questioned), the implication would be that the government should either cut taxes or expand programs. This analysis evaluates Chairman Greenspan's seemingly automatic preference for tax cuts and his apparent suggestion that tax cuts are virtually always preferable to program initiatives.
This analysis finds little evidence to justify an automatic preference for tax cuts on other than ideological grounds. It evaluates Chairman Greenspan's preference for tax cuts using three criteria.
- Short-run impact on the economy. Program increases and tax cuts produce roughly the same macroeconomic impact. Both have similar impacts on the short-term economy, national savings, and government debt. There is not a basis here for an automatic preference for either tax cuts or program increases.
- Long-run economic and social benefits. In the long run, it is unclear whether tax cuts in general or program increases in general more effectively serve economic and social objectives: the answer depends on the type of tax cut and the type of program increase. It is not possible, for example, to state absolutely that reductions in marginal tax rates are more important for the country than a Medicare prescription drug benefit or improvements in the nation's physical infrastructure. These crucial decisions for the country must ultimately be decided in the political arena.
- Political viability of future adjustments. Chairman Greenspan's principal argument for tax cuts over spending increases is his political assessment: He argues that there is a tendency for spending to grow beyond what was originally planned and that it is easier, if the surplus forecasts prove incorrect, to reverse tax cuts than spending increases. He has offered no empirical evidence for this assertion, however, and the assertion appears to be mistaken. Some program increases particularly in the discretionary spending area appear to be easier to reverse than tax cuts. Increases in some means-tested and other entitlement programs may be reversible as well.
Short-Run Impact on the Economy
The Federal Reserve's primary responsibility is the short-run economy, and in particular, managing the trade-off between inflation and unemployment. Chairman Greenspan acknowledges that his preference for tax cuts over spending increases is not related to their short-run impact on the economy. Both tax cuts and spending increases have similar short-run impacts on the economy, inflation, unemployment, and growth. Furthermore, spending increases and tax cuts have similar effects on national savings and government debt. Using the conventional criteria that Federal Reserve forecasters employ, there is no reason to prefer tax cuts to increases in programs.
If the objective is to use tax cuts or spending increases to stimulate the economy, it is difficult to argue one way or the other that tax cuts would be more effective. On the one hand, the impact of a tax cut on the economy would, if anything, be smaller than a spending increase in the short run, because a small percentage of the tax cut would be saved rather than spent and hence would not be injected into the economy quickly. On the other hand, tax cuts can be implemented relatively quickly once they have been passed; a change in tax withholding tables can be accomplished in a matter of weeks. It may take longer for funds to flow as a result of spending increases, depending on the type of spending increase. Expansions in programs that require bidding by private-sector firms, for example, would require longer lead times before cash was expended than expansions in benefit programs. For some spending programs, several months may elapse after a spending increase is approved before the cash actually flows from the government.
Long-Run Economic and Social Benefits
Given the limited resources in any budget, an important question is how the benefits of tax cuts compare to the benefits of alternative uses of the surplus. Answering this question requires evidence about the impact of tax cuts and spending increases on economic and social outcomes. More importantly, it involves fundamental choices about values.
Advocates of lower marginal tax rates often cite several benefits, including increasing work incentives, encouraging more savings, and rewarding entrepreneurship. The evidence for each of these links is mixed. Numerous economic studies have found that lower tax rates do little to encourage work effort by working-aged males who already are generally working full-time but do provide additional incentives for women to enter the labor force or men to delay partial or full retirement.(2) The evidence of the impact of lower taxes on savings is debated, with some studies finding an increase in savings and others finding no effect.
Overall, the evidence is not consistent with the belief that the level of taxes or government spending has a large effect on economic growth. There is no clear link between periods of low taxes and high growth. The strongest period of growth in U.S. history was the 1960s when the top marginal rate was 70 percent or higher. More recently, economic growth in the 1990s was quite strong, despite the 1993 increase of the top marginal tax rate from 31 percent to 39.6 percent.(3) In addition, after-tax income gains have been significantly larger among the top five percent of tax filers the only group affected by the increase in marginal tax rates in the 1990s than among the rest of the population.(4) More detailed economic research also finds no evidence that countries with lower tax rates or higher levels of government spending experience stronger economic growth.(5)
A recent study by Jonathan Gruber of M.I.T. and Emmanuel Saez of Harvard provides the most robust evidence on the effects of tax cuts. It uses a broader array of data and statistical techniques than most other analyses of the impact of tax cuts. Gruber and Saez find only a modest increase in economic activity resulting from reductions in tax rates, and they do not find the effect to be statistically significant.(6)
The estimates in the Gruber-Saez study would suggest that the Bush tax cut would result in only a small (if any) increase in the size of the economy over the next ten years and that the economy would benefit more if the surplus were used instead to increase national saving to a greater degree by reducing more of the debt.(7)
In making budgetary choices, any potential increase in the economy resulting from tax cuts needs to be compared to what would occur under alternative uses of these funds, which include debt reduction and program initiatives. There is no general presumption in favor of either tax cuts or spending increases as a mechanism for boosting economic growth and improving living conditions in the long run; one must evaluate specific tax and program proposals on their individual merits. For example, do we automatically prefer reductions in marginal income tax rates to a Medicare prescription drug benefit, expansions in Head Start, increases in housing vouchers, improvements in the nation's airports, reductions in classroom size, increased student aid for higher education, or other possible uses of the surplus?
A number of potential uses of the surplus for program initiatives have substantial economic and social benefits, including in some cases, reducing the need for more costly government spending in the future. For example:
- Smaller class sizes. Princeton economics professor Alan Krueger and Diane Whitmore examined the evidence from a Tennessee experiment that randomly assigned 11,600 elementary school students to different class sizes and found that smaller classes led to higher middle school test scores, a greater likelihood of taking the SAT, and higher SAT scores. Overall, they concluded that the real (i.e., inflation-adjusted) rate or return from lowering class size from 22 to 15 was 5.5 percent.(8) In other words, a dollar spent reducing class sizes today raises real incomes in the future by more than $5.
- Head Start and preschool. The best experimental evidence on the impact of intensive pre-schools comes from the Perry Preschool Experiment in the early 1960s, which found that an additional $1 invested in an intensive pre-school program not only substantially increased high-school graduation rates but also saved $4.75 in future expenditures on special education, public assistance, and crime.(9) More recent evidence from experts on Head Start at UCLA finds suggestive evidence that Head Start can increase the probability of completing high school and attending college, increase earnings for younger workers, and lower crime rates.(10)
- Research and development. Evidence shows that the social rate of return (that is, the amount by which society benefits) from research and development substantially exceeds the private rate of return (that is, the amount by which the organization conducting the research benefits), and as a result, that public investment in research and development often is necessary and has a high payoff.
In addition, whatever the economic evidence, decisions regarding how to expend the projected surpluses (to the extent that surpluses will be consumed rather than saved) should reflect the nation's fundamental values. Paul Krugman, one of the nation's leading economists, recently wrote that in arguing for tax cuts rather than spending increases, "Mr. Greenspan was out of bounds. Since when is it the Fed's business to say that we should have a tax cut rather than, say, a new prescription drug benefit or for that matter a missile defense system?"(11) A Washington Post editorial similarly noted, "This is a prosperous era, but one in which poverty and income inequality are still high; the baby boomers are about to retire, with all the cost shifts that implies; a seventh of the country is still without health insurance; the defense budget needs to be increased; so, too, arguably, does spending for some domestic purposes. In an era such as that, how much sense does it make to offer large-scale tax relief mainly to the better-off? The question is at least as much a social as an economic issue..."(12)
Interestingly, Chairman Greenspan himself noted that some spending programs may have significant long-run returns for the economy. He stated in response to a question from Senator Corzine in a Senate Banking Committee hearing on February 13, 2001, "Obviously, there are innumerable types of activities which you can engage in which expenditure projects do have a very clear rate of return in the sense of what they can do to the economy...I just say that it is important to scrub all expenditure programs to be sure that they are efficient, effective and they work."(13)
Political Viability of Future Adjustments
Chairman Greenspan's main arguments for tax cuts over spending increases are not based as much on economics as on politics. Greenspan put forward two propositions in his testimony:
- Spending tends to grow while tax cuts are limited: "history illustrates the difficulty of keeping spending in check.. .In contrast to most spending programs, tax reductions have downside limits."(14)
- Should the current surplus projections prove too rosy, it is easier to reverse a tax cut than a spending increase.
Neither of these propositions is supported by the empirical evidence. Moreover, both ignore important differences across types of spending programs and tax reductions, such as the difference between discretionary spending (like food safety or defense spending) that must be enacted annually and mandatory spending (like Social Security or Medicare) where changes in law generally are permanent.
Although some mandatory spending programs have ultimately proven to be significantly more expensive than initially estimated (Medicare is one example), this is not always the case. The Children's Health Insurance Program is an example of a program that has not cost as much as expected, mostly because enrollment rates have been lower than anticipated. More importantly, discretionary spending programs generally do not experience large unanticipated increases in costs. Unlike mandatory spending or tax cuts, discretionary spending programs are reviewed each year within the overall budget context, limiting the potential for their costs to explode. Also, discretionary spending programs are only allowed to spend the amount of money specifically appropriated by Congress; they must live within their annual appropriations. (In contrast, entitlements like Social Security and Medicare are legally obligated to spend whatever it costs to meet their commitments under the law.) Furthermore, there are several examples of discretionary programs that expanded to fulfill a certain need and then contracted again when they served their perceived purpose, such as defense spending during wars and the Cold War.
In addition, the historical record does not suggest that spending programs are inherently difficult to control. According to the Congressional Budget Office, total federal spending has fallen from 22.3 percent of the economy (or the Gross Domestic Product) in 1991 to 18.2 percent in 2000 (see Figure).(15) Total spending is now significantly lower, as a share of the economy, than it was during the 1980s. Such declines in spending relative to the size of the economy are not consistent with assertions that spending increases are more difficult to control than tax cuts.
The latest CBO projections assume continued declines in government spending. Specifically, CBO projects that total federal spending will decline from 18.2 percent of the Gross Domestic Product (GDP) in 2000 to 15.1 percent in 2011. Last year's level 18.2 percent of GDP was already the lowest level of federal spending since 1966, and CBO projects a new 35-year low will be set in 2001, with federal spending falling to 18.0 percent of GDP. If spending declines to 15.1 percent by 2011, as CBO projects, federal spending will constitute a smaller share of the economy than at any time since 1951.
The projections for appropriated spending are equally striking. Under CBO's new projection, appropriations are projected to decline from 6.3 percent of GDP in 2000 and 2001 to 5.1 percent in 2011. The 6.3 percent figure for 2000 and 2001 already is the lowest on record, going back to 1962 (the first year for which these data are available). Appropriations constituted 12.7 percent of GDP in 1962 and did not fall below 7.0 percent of GDP until 1996. In part, the steep decline in appropriated spending reflects the end of the Cold War and the resulting decision to reduce the size of the armed forces. Even so, under CBO's latest projections, domestic appropriations which already are very close to their lowest level on record as a share of the economy, going back to 1962 will continue declining as a share of a GDP through 2011, establishing record lows.
Even last year, with substantial increases in appropriations levels, domestic discretionary spending simply grew at about the same rate as the economy: Such spending has remained constant at 3.1 percent of GDP in each year from 1998 through 2001. Moreover, total discretionary spending (which includes discretionary spending for defense, domestic, and international programs) has edged down as a share of the economy over this period. This evidence runs counter to the perception of Chairman Greenspan, among others, who testified that "The flurry of increases in outlays that occurred near the conclusion of last fall's budget deliberations is troubling, because it makes the previous year's lack of discipline less likely to have been an aberration." While discretionary spending increased in inflation-adjusted terms, domestic discretionary spending has not experienced a "flurry of increases" when it is measured relative to the size of the economy.
Last Year's "Flurry of Spending Increases"
Some policymakers have repeatedly said in recent weeks, citing Congressional Budget Office data, that actions by Congress and the President last year increased spending by $600 billion over ten years. This presumably is what Chairman Greenspan referred to when he spoke of last year's "flurry of spending increases." But the story is not as simple or straightforward as the pronouncements of a $600 billion spending increase would suggest. What the CBO data show are the following.
Actions taken last year increased projected expenditures on federal programs by $434 billion over ten years when compared to the CBO baseline that (for non-entitlement programs) adjusts these programs for inflation only and not for growth in the U.S. population. The $434 billion increase in projected program expenditures will result in $118 billion in additional interest payments on the debt over ten years. That brings the total cost to $552 billiona (which some policymakers have rounded to $600 billion).
Some $368 billion or 85 percent of the increased program expenditures consist of increases in discretionary (i.e., non-entitlement) programs. This $368 billion amount is the increase over a baseline that does not adjust discretionary program spending for population growth and that consequently assumes real reductions in government services per U.S. resident. Comparing the spending levels approved last year to a baseline for discretionary spending that does reflect the effects of population growth (as well as inflation) an approach the White House uses when measuring the rate of spending growth in Texas during George W. Bush's tenure as governorb would show the increases to be somewhat smaller.
Still another way to measure changes in federal spending is to track spending as a share of the economy. This year, total federal spending declined as a share of the economy falling to its lowest level as a share of the economy since 1966 despite the "flurry of spending increases."
It also may be noted that nearly one-third of these spending increases came in defense spending. Moreover, 55 percent of the increase in discretionary spending occurred in three areas that the new Bush administration budget proposes to increase further defense, education, health research and related health programs. Indeed, some of those who cite the "$600 billion spending increase" as evidence that a large tax cut is needed to prevent spending from growing too much simultaneously argue that spending did not grow enough in the principal areas in which spending was increased.
a According to CBO, an additional $9 billion in interest payments will result from small tax cuts enacted last year.
b The White House, "Press Briefing by Ari Fleischer," February 27, 2001, p.12.
Spending programs thus do not necessarily explode. Tax cuts, furthermore, do not necessarily have natural "downside limits." New tax loopholes, for example, may be used increasingly over time, resulting in substantially higher tax expenditures than initially estimated. Recent articles in the journal Tax Notes and in the New York Times suggest that elimination of the estate tax may generate significant increases in other tax avoidance strategies, so that the revenue loss from eliminating the estate tax may be much larger than currently anticipated and could burgeon over time as more taxpayers learn of and use the loopholes available to them.(16) Also, some tax cuts simply cannot be scaled back or eliminated because people have made lifetime financial decisions based on them such as Roth IRAs, in which people have incurred tax liability to save money today because they have been promised tax benefits in the future.
Another issue that Chairman Greenspan raised is what would happen if large increases in spending or reductions in taxes are enacted this year, but the budget situation subsequently deteriorates, creating a need for these changes to be partially or completely reversed in the future. From a political perspective, it is unclear that it is any easier to raise taxes than to reduce spending if current projections turn out to be overly optimistic.
To be sure, there was bipartisan support in 1982 for rolling back some of the large 1981 Reagan tax cut. But the first President Bush paid a substantial political price for the tax increases enacted in 1990, and many political observers believe the 1993 tax increases contributed to the Democratic loss of Congress in the 1994 election. In the aftermath of these two events, enacting tax increases appears to have become a considerably more difficult political task. Furthermore, there are a number of examples of reductions in spending, including entitlement spending. Reductions in Medicare were enacted in the 1990, 1993, and 1997 budgets. Reductions in means-tested programs such as food stamps, the Supplemental Security Income program, the Social Services Block Grant, and child nutrition programs were enacted in both 1981 and 1996. Even the school lunch program was cut considerably in 1980 and 1981, and the unemployment insurance program also was cut substantially in the early 1980s. These examples demonstrate that spending on various entitlement programs can be rolled back. Discretionary spending is even more flexible. Past experience is replete with examples of nominal freezes or other measures that partially reversed previous increases in various areas of discretionary spending.
In short, evidence does not support the proposition that spending programs are inherently more difficult to "keep in check" than tax cuts.
Even if one accepts the argument that the government should not accumulate private assets and the only way to avoid this accumulation is to enact policies today to dissipate the projected surplus arguments that other Center papers and economists such as Alice Rivlin have suggested are of dubious merit Chairman Greenspan's solution to this issue is problematic. There is no general reason to believe tax cuts are automatically preferable to spending increases.
According to the principal economic criteria the short-run impact on the economy and the long-run impact on economic efficiency there is no reason automatically to prefer tax cuts to spending increases. Both policies have similar short-run impacts. Over the longer run, the effects depend significantly on the types of tax cuts and program increases.
The main arguments that Chairman Greenspan advanced were not economic, but rather grounded in his own personal values or based on his political prognostications. Debates over the size and role of government are fundamental issues that must be resolved by the political system. They should not rely too heavily on experts in monetary policy.
Furthermore, history provides little support for Chairman Greenspan's hypothesis that program initiatives are inherently more difficult to reverse than tax reductions. At least for discretionary spending, it is likely that spending is easier to control than tax cuts.
If a substantial fraction of the surplus is committed this year, policymakers should debate the best ways to address the major challenges that America faces, rather than being guided by an ideological preference for either tax cuts or program initiatives.
1. Alan Greenspan, Statement before the Committee on the Budget, United States Senate, January 25, 2001. For other Center analyses of Greenspan's testimony, see Peter Orszag and Robert Greenstein, "The Greenspan Concern over Public Ownership of Private Assets: Can the Social Security Trust Fund Safely Own Such Assets?" Center on Budget and Policy Priorities, February 26, 2001, and Peter Orszag, "The Peril of Zero Debt and the Long-Term Budgetary Outlook: Some Questions Regarding Chairman Greenspan's Recent Testimony," Center on Budget and Policy Priorities, February 22, 2001.
2. The evidence is surveyed in Jerry Hausman, "Taxes and Labor Supply," Handbook of Public Economics: Volume I, 1985.
3. The increase in the marginal tax rate was actually larger than this increase in statutory income tax rates, since the 1993 budget deal also applied the Medicare payroll tax rate to an unlimited amount of earnings. Before 1993, the Medicare payroll tax (2.9 percent of earnings) applied only up to a given level of earnings.
4. Isaac Shapiro, "The Latest IRS Data on After-Tax Income Trends," Center on Budget and Policy Priorities, February 26, 2001.
5. See, for example, Joel Slemrod. "What Do Cross-country Studies Teach About Government Involvement, Prosperity, and Economic Growth?" Brookings Papers on Economic Activity No. 2, 1995: pp. 373-431.
6. Jonathan Gruber and Emmanuel Saez, "The Elasticity of Taxable Income: Evidence and Implications," NBER Working Paper 7512, January 2000.
7. See appendix to Peter Orszag, "The Peril of Zero Debt and the Long-Term Budgetary Outlook: Some Questions Regarding Chairman Greenspan's Recent Testimony," Center on Budget and Policy Priorities, February 22, 2001.
8. Alan Krueger and Diane Whitmore, "The Effect of Attending a Small Class in the Early Grades on College-test Taking and Middle-School Test Results: Evidence from Project Star," NBER Working Paper No. 7656, April 2000.
9. Evidence cited in Council of Economic Advisers, "To Save One Dollar", White Paper, October 1995.
10. Eliana Garces, Duncan Thomas, and Janet Currie, "Longer Term Effects of Head Start," NBER Working Paper No. 8054, December 2000.
11. Paul Krugman, "Et Tu, Alan?" New York Times, January 28, 2001, page A15.
12. "The Greenspan Turnaround," Washington Post, January 28, 2001, page B6.
13. Hearing before the Senate Banking, Housing, and Urban Affairs Committee, February 13, 2001.
14. Alan Greenspan, Statement before the Committee on the Budget, United States Senate, January 25, 2001. For other Center analyses of Greenspan's testimony, see Peter Orszag and Robert Greenstein, "The Greenspan Concern over Public Ownership of Private Assets: Can the Social Security Trust Fund Safely Own Such Assets?" Center on Budget and Policy Priorities, February 26, 2001, and Peter Orszag, "The Peril of Zero Debt and the Long-Term Budgetary Outlook: Some Questions Regarding Chairman Greenspan's Recent Testimony," Center on Budget and Policy Priorities, February 22, 2001.
15. Reduced interest costs on the public debt explains only a small portion of the decline. Excluding net interest payments, spending fell from 19.0 percent of GDP in 1991 to 15.9 percent in 2000, a decline equal to 3.1 percent of GDP. The decline in total spending (including interest costs) was 4.1 percent of GDP.
16. Jonathan G. Blattmachr and Mitchell M. Gans, "Wealth Transfer Tax Repeal: Some Thoughts on Policy and Planning," Tax Notes, January 15, 2001; John Buckley, "Transfer Tax Repeal Proposals: Implications for the Income Tax," Tax Notes, January 22, 2001; and David Cay Johnston, "Questions Raised on New Bush Plan To End Estate Tax," New York Times, January 29, 2001.