April 19, 2001
The Bush Tax Cut Is Now about the Same Size As the Reagan Tax Cuts
by Peter R. Orszag(1)
View PDF version of this report
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.
Proponents of President Bush's proposed tax cut have argued that it is much smaller than the 1981 Reagan tax cut and other historical tax cuts.(2) Some Bush Administration officials and Members of Congress have echoed these claims and suggested this shows the proposed tax cut is of a responsible size. Careful examination, however, shows these arguments reflect apples-to-oranges comparisons. Furthermore, both new Joint Tax Committee estimates and changes that the House of Representatives has made in the Bush tax cut have raised the tax cut's cost. On a comparable basis, the Bush tax cut, when fully in effect, is now about the same size as the Reagan tax cuts.
If the cost of the Reagan tax cut is adjusted for the impact of inflation and the subsequent 1982 tax increase (which scaled back the 1981 tax cut), the net tax cut is about the same size as a share of the economy (2.1 percent of GDP) as the proposed Bush tax cut would be (2.0 to 2.3 percent of GDP), rather than being several times the size of the Bush tax cut. Furthermore, the Reagan tax cut occurred when marginal tax rates were higher than today. A reduction in marginal tax rates is therefore not as significant today as in 1981. Finally, the Reagan tax cut was a major factor in generating large budget deficits from which the nation took more than decade and a half to recover.
Inflation and the Revenue Baseline
Before 1985, frequent tax cuts were necessary just to prevent large tax increases over time because the tax code was not indexed to inflation. The result was a natural upward "creep" in tax collections over time, as ongoing inflation pushed individuals into higher tax brackets. To see how this worked, assume that the current tax code was not indexed to inflation. If a taxpayer's income merely kept pace with inflation, his or her purchasing power and standard of living would not increase. The taxpayer's tax liability, however, would increase because the standard deduction, the personal exemption, and other features of the tax code would not be adjusted for inflation. The taxpayers' tax liability would rise over time for a second reason as well the tax rate brackets would not be adjusted for inflation, so a taxpayer could be pushed into a higher tax bracket if his or her wages simply remained even with inflation.
For these reasons, under a tax code that is not indexed, taxpayers pay a higher percentage of income in taxes with each passing year even when their income gains merely keep them even with inflation or lag behind inflation. By contrast, when the tax code is indexed as it is today taxpayers' tax liabilities do not increase unless their incomes rise faster than inflation.
The lack of indexing in the tax code before 1985 consequently produced an automatic upward creep in tax collections over time. As a 1998 Treasury Department paper noted, "Without indexation, bracket creep occurs, which increases federal revenue as a percentage of GDP without any legislative action....In fact, when inflation is relatively high and bracket creep is particularly intense, as it was through much of the 1970's, policy makers have to cut taxes repeatedly to maintain the desired level of taxes."(3) In other words, regular tax cuts were necessary just to keep taxes steady as a percentage of taxpayers' incomes and to avert tax increases over time.
During this period, current law thus contained built-in tax increases at any point in time. Policymakers cut taxes every few years to offset much or all of the tax increases that otherwise would occur, but the Congressional Budget Office was forced in constructing its revenue baseline to assume that taxes would rise over time as a share of taxpayers' incomes, because the baseline reflected current law. Under this system, CBO "scored" legislation that merely kept tax burdens steady as a tax cut, even though families would feel no benefit from it: Their taxes would not change as a share of their income. The baseline against which the Reagan tax cut (the Economic Recovery Tax Act of 1981) and other tax cuts were measured thus was an artificially inflated baseline as a result of the lack of indexing in the tax code.
Percentage of ERTA Revenue Reduction Due Solely to
Effect of Inflation after October 1, 1981 on Baseline
Year 1982 28% 1984 32% 1987 45% Source: Congressional Budget Office, Baseline Budget Projections for Fiscal Years 1983-1987, February 1982, Tables 11and 12, Figure 5, and author's calculations.
In 1981, the size of the Reagan tax cut was measured using this inflated baseline. That increased the apparent size of the tax cut, since the tax cut was measured relative to a baseline that assumed significant tax increases. As the Congressional Budget Office noted when the Reagan tax cut was first proposed, "While the Administration proposal would reduce revenues by large amounts in those years, it is important to keep in mind that, without a tax cut, income taxes rise continually because of the effects of inflation on the graduated income tax rate schedule...a large share of the Administration's proposed tax cut would simply offset these tax increases [emphasis added]."(4) Table 1 shows the percentage of the Reagan tax cut's cost that was due solely to measuring the tax cut against a revenue baseline that assumed future tax increases as a result of the effects of inflation. As the table indicates, by 1987, some 45 percent of the cost of the Reagan tax cut simply reflected the effects of inflation on the baseline. In other words, when measured relative to a baseline that adjusted for inflation, the cost of the Reagan tax cut that year would be only modestly more than half as large.
Since 1985, the tax code has been indexed to inflation, and the baseline consequently no longer includes large, automatic tax increases over time.(5) In other words, ongoing inflation no longer causes a large upward movement in taxes over time in the baseline. Under the current baseline, any tax cut that reduces tax burdens on families by a given amount would be scored as costing less than the tax cut would cost if it were measured against a baseline that did not include indexing.
Comparing the Bush tax cut to the Reagan tax cut thus is misleading: A large component of the Reagan tax cut merely prevented an increase in taxes that would have otherwise occurred because of the lack of indexing in the tax code.
The 1982 Tax Increase
The 1981 tax cut was excessive, a conclusion to which David Stockman (then the director of the Office of Management and Budget) and others in the Reagan administration came not long after its enactment. As a result, the Reagan administration worked to scale back the tax cut one year later. The Tax Equity and Fiscal Responsibility Act of 1982 (TEFRA) increased revenue by closing some loopholes broadened in the 1981 act, altering depreciation deductions, tightening safe harbor leasing rules, and making several other changes. As CBO noted, these "tax increases partly offset the revenue effects of ERTA [the 1981 act] by offsetting almost two-thirds of the ERTA corporate income tax reductions and about 10 percent of the ERTA individual income tax reductions."(6) The net cost of ERTA and TEFRA is a more appropriate measure of the Reagan tax cuts than the cost of ERTA alone.
Comparing the Net Size of the Reagan Tax Cuts and the Proposed Bush Tax Cut
Various estimates are available of the size of the Reagan tax cuts as a share of the economy. The National Taxpayers Union, for example, cites the cost of ERTA as 3.3 percent of GDP. The Treasury Department paper cited above estimates a cost for ERTA four years after enactment of 4.15 percent of GDP.(7) The Congressional Budget Office, in 1983, estimated the cost of ERTA would be 5.6 percent of GDP in 1988.(8) We use the CBO estimate as a basis for comparison with the proposed Bush tax cut.
Table 2 Percentage of GDP ERTA 1981 5.6% Minus: 40 percent adjustment for impact of inflation on baseline -2.2% Equals: ERTA cost against indexed baseline 3.4% Minus: TEFRA 1982 increase -1.2% Equals: Net cost of Reagan tax cuts (as % of GDP) 2.1%
As explained, these figures should be adjusted for the impact of inflation on the revenue baseline and for the partial reversal of the 1981 tax cuts enacted in 1982. Table 1 indicates that the share of the revenue cost due solely to inflation in the baseline was 45 percent in 1987. To be conservative, we assume that 40 percent of the cost of ERTA was due to the impact of inflation on the revenue baseline and that 60 percent thus was a true tax cut.(9) Also, according to CBO, the revenue increase from TEFRA amounted to about 1.2 percent of GDP.(10) Table 2 displays these adjustments to the cost estimates. The net result is that the adjusted cost of the Reagan tax cuts amounted to 2.1 percent of GDP.
Cost of Bush Tax Cuts as a Percentage of GDP
Initial estimates from the Joint Committee on Taxation and CBO suggested that the Bush tax cut would amount to 1.5 percent of GDP in 2010.(11) Congressional action to date has increased the cost of certain tax cuts, however, as have new estimates from the Joint Tax Committee.
For example, the House of Representatives has taken a different and considerably more costly approach to tax relief for married couples than the President's proposal does. In addition, the Joint Tax Committee's estimate of the cost of repealing the estate tax has increased, because the estimate now reflects the important interaction between the income tax code and the estate and gift tax. The Joint Tax Committee explains that it has recently adjusted its estimates to take into account the "significant revenue effects that result from a variety of income tax avoidance opportunities made possible by the repeal of the estate and gift tax."(12) That is, in the absence of estate and gift taxes, high-income families would have more flexibility to avoid income taxes.
Taking into account cost of the three tax bills approved by the House (H.R. 3, H.R. 6, and H.R. 8), as estimated by the Joint Committee on Taxation, plus the cost shown in the Administration's budget for the remaining components of its tax package, and adding the Joint Tax Committee's estimate of the cost of modifying the Alternative Minimum Tax to prevent 15 million additional taxpayers from becoming subject to the AMT because of the Bush plan a cost most observers agree will eventually be incurred brings the total cost of the proposed tax cuts (excluding interest) to 2.0 percent of GDP in 2011 and 2.3 percent of GDP between 2012 and 2021.(13) (The estate tax would not be repealed until 2011 under the House legislation, so examining the cost of the overall tax cut in 2012 or 2013 provides a better indication of its long-term impact than the 2010 cost. To estimate the size of the tax cuts after 2011, we assumed that once a tax cut is fully in effect, its cost would generally remain constant as a share of the economy.)(14)
The Bush tax cut thus is about the same size as the adjusted Reagan tax cut. The Bush tax cut would cost 2.0 to 2.3 percent of GDP; the adjusted Reagan tax cuts cost about 2.1 percent of GDP.
The Legacy of the Reagan Tax Cut
Even if the Bush tax cut represented only a modest proportion of the properly measured Reagan tax cut, the basic logic of the comparison would be problematic. The Reagan tax cut does not represent a valid basis for evaluating what size tax cut is fiscally responsible.
Even with the subsequent tax increase in 1982, the 1981 tax cut imposed a damaging fiscal legacy on the nation. The unified budget deficit rose from $74 billion in 1980 to $221 billion in 1986 and a peak of $290 billion in 1992. As a percentage of GDP, the deficit (adjusted for the economy's business cycle) rose from 0.7 percent in 1980 to a peak of 4.8 percent in 1986.
Some advocates for the Bush tax proposal have argued that the Reagan tax cuts were not a cause of the large budget deficits during the 1980s. Rather, they argue, the problem arose because of large spending increases. This argument is not supported by the evidence. CBO produces estimates of revenues and outlays that adjust for the state of the business cycle. These figures indicate that, adjusted for the state of the business cycle, revenue fell from 19.4 percent of GDP in 1981 to 16.9 percent of GDP in 1986 and 17.3 percent in 1987. Outlays rose from 19.9 percent of GDP in 1981 to 21.7 percent in 1986 and 20.6 percent in 1987. Between 1981 and 1987, revenues fell three times as much as a percentage of GDP as spending increased. (Revenues declined by 2.1 percent of GDP while outlays increased by 0.7 percent of GDP.) Furthermore, all of the increase in outlays relative to GDP was due to interest outlays, which themselves mostly reflected the tax cuts. Excluding interest, outlays actually fell: Non-interest outlays fell from 19.7 percent of GDP in 1981 to 19.4 percent in 1986 and 18.8 percent in 1987. These data clearly demonstrate that the Reagan tax cuts played a substantial role in the budget deficits of the 1980s and early 1990s.
The Kennedy Tax Cut
Some proponents of the proposed Bush tax cut have invoked the 1964 Kennedy tax cuts as a justification. The proponents argue that the Bush tax cut is the same size as and designed in a similar fashion to the Kennedy tax cut. Neither of these propositions is entirely accurate.
First, according to the Treasury Department, the Revenue Act of 1964 reduced revenue by 1.6 percent of GDP. This cost of the 1964 tax cut is estimated against an unindexed baseline, as the unadjusted cost of the Reagan tax cut is. The true cost of the 1964 tax cut is thus somewhat less than 1.6 percent, since part of the tax cut merely offset a tax increase assumed in the baseline. The proportion of the 1964 tax cut attributable to this factor is likely to be relatively small, however, because inflation was low in the early 1960s. Nonetheless, even using the 1.6 percent of GDP estimate, the Kennedy tax cut was somewhat smaller than the proposed Bush tax cut.
Second, the distribution of Kennedy tax cut was dramatically different from the proposed Bush tax cut. According to estimates of the Kennedy tax cut that the Joint Committee on Internal Revenue Taxation made at the time, the bottom 40 percent of the income distribution received 18 percent of the personal income tax cut from the 1964 act, and the bottom 85 percent received 59 percent of the tax cut. The top 2.4 percent of the income distribution received 17.4 percent of the tax cut, and the top 0.4 percent of the income distribution received six percent of it. By contrast, under the proposed Bush personal income tax reductions, the bottom 40 percent of the income distribution would receive only six percent of the tax cuts, and the bottom 80 percent of the distribution would secure 38 percent of the tax reductions. At the other end of the income spectrum, the top one percent of the distribution would receive 31 percent of the personal income tax cuts.*
The Kennedy tax cut thus differed significantly from the proposed Bush tax cut. It was both somewhat less expensive and much less unevenly distributed.
* The estimates of the distributional impact of the Bush tax cut noted here reflect the Bush tax cut as modified by action to date in the House of Representatives. These estimates use Citizens for Tax Justice estimates of the distributional effects of the proposed changes in personal income taxes when the changes are phased in fully.
Both the distributional estimates for the Kennedy tax cut and those for the Bush tax cut ignore the effects of reductions in corporate and estate taxes. It is possible to estimate the effects of the Bush corporate and estate tax cuts, based on Treasury Department estimates of the incidence of these taxes. When this is done, the top one percent of taxpayers are found to receive an estimated 39 percent of the total Bush tax cut package.
There are no distributional estimates available of the Kennedy corporate tax reductions. Since the Kennedy corporate tax cuts comprised a smaller share of the Kennedy tax package than the Bush corporate and estate tax cuts ultimately would comprise of the Bush tax package, the conclusion stated above that the Kennedy tax cut was less unevenly distributed than the Bush proposal would be even stronger if the distributional impacts of the entire Kennedy and Bush tax packages were compared.
The result of those deficits was that the federal government was forced to borrow massively, and debt held by the public rose from 25.7 percent of GDP at the end of 1980 to 49.5 percent at the end of 1993. Federal debt per household was more than $15,000 higher in 1993 than if debt had remained at the same level (relative to the size of the economy) as it was in 1980. Interest payments on the debt (in constant dollars) were more than $9,000 higher per household in total between 1981 and 1993 than if debt had remained at its 1980 level as a share of the economy. The test of fiscal responsibility today should not be this historical record of deficits and debt that was partly engendered by the Reagan tax cuts.
The Level of Marginal Tax Rates
Marginal tax rates were higher in 1981 than they are today. In 1981, the maximum marginal income tax rate was 70 percent on unearned income and 50 percent on wage and salary income. Today, the maximum statutory marginal income tax rate is 39.6 percent.(15) A reduction in marginal tax rates is therefore less important today than in 1981.
The cost of the Bush tax cut has risen over the past several weeks, due to modifications made by the House of Representatives and updated estimates from the Joint Tax Committee. On a comparable basis, the proposed Bush tax cut, when fully in effect, is now about the same size as the Reagan tax cuts. Arguments to the contrary are based on a way of measuring the size of the Reagan tax cuts that has two shortcomings: that way of measuring the Reagan tax cuts overlooks the partially offsetting 1982 tax increase, and it counts measures that simply keep taxes from rising as constituting large tax cuts. In addition, justifying the Bush tax cut by comparing it to the Reagan tax cut implicitly ignores the fiscal problems to which the Reagan tax cuts were a significant contributor. After adjustment is made for the impact of inflation on the revenue baseline before 1985 and for action that scaled back the 1981 tax cut just one year later, the Reagan tax cut is about the same size as the proposed Bush tax cut.
1. Peter R. Orszag is President of Sebago Associates, Inc., an economics consulting firm. He has previously served as Special Assistant to the President for Economic Policy, as Senior Economist on the Council of Economic Advisers, and as a member of the economics faculty at the University of California, Berkeley. In August 2001, he will join the Brookings Institution as a senior fellow in economic studies.
2. Eric V. Schlecht, "History vs. Hysteria: Correcting the Hyperbole Surrounding the Bush Tax Cut Plan," National Taxpayers Union, NTU Issue Brief 111, January 24, 2001.
3. Jerry Tempalski, "Revenue Effects of Major Tax Bills," U.S. Department of the Treasury, OTA Working Paper 81, December 1998, page 4.
4. Congressional Budget Office, "An Analysis of President Reagan's Budget Revisions for Fiscal Year 1982," March 1981, page 19.
5. The indexation of the tax code, effective in 1985, was a component of the 1981 act.
6. Congressional Budget Office, "Baseline Budget Projections for Fiscal Years 1984-1988," February 1983, page 27.
7. Tempalski, op. cit., Table 2.
8. Congressional Budget Office, "Baseline Budget Projections for Fiscal Years 1984-1988," February 1983, Table 11.
9. It is worth noting that the CBO analysis dividing the cost of the tax cut between the effect of inflation on the baseline and the remainder after taking such inflation into account was based on expected inflation as of the early 1980's. Actual inflation turned out to be somewhat lower than expected, which would affect the division of the total cost of the tax cuts into the two components. Analysis using actual inflation instead of expected inflation, however, is unavailable. In addition, the analysis using expected inflation reflects what policy-makers had expected at the time of passage of the legislation. That perspective may be the most relevant for evaluating the relative size of the Reagan and proposed Bush tax cuts.
10. In 1983, one component of TEFRA 1982 was reversed: the withholding of tax from interest and dividends was repealed in the Interest and Dividend Tax Compliance Act of 1983. The cost of that act (0.04 percent of GDP) is subtracted from the TEFRA 1982 cost estimate in the figures use here. Note also that correcting for the effect of inflation on the baseline would, if anything, raise the revenue gain due to TEFRA: The revenue increase from TEFRA would be larger if the baseline had been adjusted for the effects of inflation and baseline revenues had consequently been lower than in the official baseline.
11. This 1.5 percent-of-GDP figure compares the cost estimate for the Bush plan that the Joint Committee on Taxation produced last year to the GDP projection the Joint Committee used in generating that estimate. Since then, the CBO has revised upward its estimate of projected GDP in 2010. The CBO revisions, however, also affect the cost of the proposed Bush tax plan.
12. Lindy Paull, "Memorandum to John Buckley: Estate and Gift Tax Estimates," Joint Committee on Taxation, March 26, 2001.
13. Since the Reagan revenue estimates are measured relative to an inflation-adjusted baseline, two objections to the Bush figures are possible. First, the cost of repealing the estate tax is measured against a baseline in which the estate tax exemption is not indexed to inflation after 2006. The cost of the estate tax repeal may therefore be exaggerated relative to an inflation-indexed baseline. However, even though the exempt amount under current law (and therefore under the baseline) is not indexed after 2006, it still is higher in nominal terms in 2011 than the current level plus inflation. Under current law, the exempt amount will increase from $675,000 this year to $1 million by 2006. Adjusting for inflation using CBO's economic projections, this year's level of $675,000 is equivalent to $893,000 in 2011. The statutory level of $1 million is thus higher than the inflation-adjusted level even as of 2011. The second objection is that the cost of preventing the proposed tax cut from increasing the number of taxpayers on the AMT may be incorrect, depending on precisely how the AMT is adjusted for inflation within the inflation-adjusted baseline. Such effects, however, are likely to be relatively small.
14. The one exception to this assumption is the estate tax. For more information regarding the estimation of the revenue costs after 2011, see Richard Kogan, Joel Friedman, and Robert Greenstein, "Cost of Tax Cut Would More than Double to $5 Trillion in Second Ten Years," Center on Budget and Policy Priorities, April 4, 2001.
15. In practice, the effective marginal tax rate differential may be modestly smaller than this difference in statutory rates. In particular, the 2.9 percent payroll tax that funds the Hospital Insurance component of Medicare now applies to an unlimited amount of earnings; before 1993, it applied only up to a given threshold of earnings. In addition, the phase-out of itemized deductions (which applies to married taxpayers with adjusted gross income of more than $128,950 in 2000) and of personal exemptions (which applies to married taxpayers with adjusted gross income of more than $193,400 in 2000) may raise the effective marginal income tax rate above 39.6 percent for some taxpayers, depending on their level of deductions. Two points are worth noting. First, the personal exemption phase-out ended at an adjusted gross income of $315,900 in 2000. The 39.6 percent bracket began at a taxable income level of $288,350. This makes it unlikely that many taxpayers in the 39.6 percent bracket are affected by the phase-out of the personal exemptions. The Joint Tax Committee recently stated that "generally no taxpayers in the 39.6-percent statutory bracket would be affected." See Joint Committee on Taxation, "Overview of Present Law and Economic Analysis Relating to Marginal Tax Rates and the President's Individual Income Tax Rate Proposals," March 6, 2001, page 18. Second, the phase-out of itemized deductions raises marginal tax rates only slightly (by approximately one percentage point for those in the 39.6 percent marginal bracket).