Revised April 11, 2002
SOCIAL SECURITY AND THE TAX CUT
The 75-Year Cost of the Tax Cut Is More than Twice
as Large as the Long-Term Deficit in Social Security
by Peter Orszag, Richard Kogan, and Robert Greenstein
PDF of this 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.
Restoring long-term solvency to Social Security and ensuring a sustainable long-run fiscal policy for the United States are issues of major importance. To help illuminate these issues, this analysis examines and compares the fiscal dimensions of two major items: the projected long-term deficit in Social Security and the long-term cost of the tax cut enacted last June (assuming that the provisions of the tax cut are extended beyond their scheduled expiration dates).
From listening to various pundits and policymakers, many Americans may believe that the tax cut is modest in size while the long-term Social Security shortfall is enormous. Senator Phil Gramm, for example, has said: "This is not a huge, irresponsible tax cut, this is a modest tax cut....this is a prudent, responsible tax cut." Last year's interim report of the President's Social Security commission stated that the Social Security shortfall is of a magnitude that threatens "astronomical levels of borrowing." Michael Tanner of the Cato Institute, one of the most ardent and most widely quoted privatization proponents, has likened Social Security's condition to that of the Titanic, while David John of the Heritage Foundation has written that Social Security faces a "monsoon."
As this analysis shows, the long-term size of the tax cut is more than double the entire long-term Social Security shortfall. The tax cut is not as modest as its proponents often claim, while the Social Security shortfall although a significant problem that must be addressed is not as gargantuan as often portrayed by those seeking radical changes in Social Security.
When this analysis was first issued on August 2, 2001, the Administration attempted to refute its findings on the relative magnitude of the tax cut and the Social Security shortfall. The Administration's arguments were unpersuasive (see box on page 3). Moreover, even the Administration's "refutation" conceded that over the next 75 years, the revenue loss from the tax cut is fully as large as the shortfall in Social Security.
Because of the tax cut and because the projections of large and growing surpluses made last spring have turned out to be too optimistic for other reasons as well resources no longer exist outside of Social Security that could assist in restoring solvency to this program. Cancelling some provisions of the tax cut before they take effect in future years could help to provide such resources.
The Size of the Tax Cut and the Social Security Shortfall
Over the next 75 years - the period used to measure Social Security's financing gap - the revenue loss from the tax cut will be $5 trillion larger than the entire Social Security shortfall.
According to the official estimates that the Social Security actuaries and trustees issued in March 2002, the projected long-term deficit in Social Security over the next 75 years the period used for measuring long-term solvency equals 1.87 percent of the wages, salaries, and self-employment income that will be subject to the payroll tax during this period, or $3.7 trillion in present value. (Present value is the amount today that, with interest, would exactly cover these future costs.) The trustees' report also shows that, measured as a share of the economy, the Social Security shortfall equals 0.72 percent of the Gross Domestic Product over the next 75 years.(1)
Cost of Tax Cut and Size of Social Security Shortfall Over 75 Years As Share of GDP Present Value Social Security Shortfall
Tax Cut 1.68% $8.7 trillion
To measure the long-term cost of the tax cut, we take the Congressional Budget Office's most recent estimate of the cost of the tax cut in 2011 if all of its provisions are extended, and assume that these costs will remain constant as a share of GDP after 2011. Assuming that the cost of tax cuts will remain constant as a share of GDP once the tax cuts are fully in effect is the standard approach that the Congressional Budget Office, the Office of Management and Budget, and the General Accounting Office all use when preparing long-term fiscal projections. In this case, such an approach is likely to understate long-term revenue losses because the costs of several provisions of the tax bill, such as the estate tax repeal and the introduction of "Roth 401(k) pension plans," are virtually certain to grow faster than GDP for a number of years after 2011.(2) For this reason, our estimates of the long-term costs of the tax cut are likely to be conservative.
The projected cost of the tax cut over 75 years amounts to 1.68 percent of GDP, or $8.7 trillion in present value. Thus, the cost of the tax cut over the next 75 years is more than twice as large as the long-term deficit in Social Security, as shown in the table above.(3)
In other words, if the tax cut takes full effect as scheduled and continues after 2010, its long-term costs will substantially exceed the 75-year deficit within Social Security. In fact, if the tax cut were scaled back so that three-fifths of it took effect while the funds from the other two-fifths of the tax cut were used instead to strengthen Social Security, the entire 75-year deficit in Social Security could be eliminated.
These figures highlight the inconsistency in the rhetoric of those who portray Social Security as facing an enormous financial chasm while touting the tax cut as modest and prudent.
The figures on the relative size of the Social Security shortfall and the tax cut also show the fundamental inconsistency in the rhetoric of policymakers, interest groups, and others (including some Administration officials) who portray Social Security as facing an enormous financial chasm that threatens the nation's long-term fiscal health while touting the tax cut as modest and prudent.
We should emphasize that we would not recommend canceling 40 percent of the tax cut and placing all of the freed-up resources in Social Security. The nation will face serious financial strains when the baby boomers retire in large numbers. The long-term financing shortfall in Medicare is larger than that in Social Security, and the nation also is likely to face needs in the decades ahead that will require resources in other areas, including areas relating to children, the environment, the large number of Americans without health insurance, the lack of a Medicare prescription drug benefit, and the uncertain costs of homeland security, as well as other problems that inevitably will arise in the future but that we cannot foresee today. A balanced long-term fiscal policy is likely to entail some changes in Social Security to reduce its future claims on the budget, rather than simply providing it with whatever resources are needed from the rest of the budget to close its entire long-term financing shortfall.
The Administration's Refutation Is Not Convincing
When this analysis was first issued on August 2, 2001, the Bush Administration responded by saying that the cost of the tax cut is only one percent of GDP (rather than 1.68 percent) while the Social Security shortfall is, likewise, close to one percent of GDP (rather than 0.7 percent). Although the Administration itself thus acknowledged that the revenue loss from the tax cut is fully as large as the Social Security shortfall, the 1.0 percent of GDP figure that it used for both estimates is not valid its Social Security estimate differed from the traditional measure issued by the Social Security Trustees, while its tax cut estimate failed to include the cost of at least three provisions of the tax-cut law.
Under the intermediate projection prepared by the highly respected Chief Actuary at the Social Security Administration and published in the 2001 Social Security Trustees report, the Social Security shortfall was projected to equal 0.7 percent of GDP, virtually identical to the trustees' current estimate of 0.72 percent. The Center relies on these published figures. The Administration, by contrast, claimed that the Social Security shortfall equaled about one percent of GDP. It did so primarily by ignoring the assets of the Social Security Trust Fund. Such an assumption contradicts the long-established practice of the Social Security actuaries and trustees in evaluating the long-term imbalance within Social Security; the actuaries and trustees appropriately count the Trust Fund's $1.2 trillion in assets, since these assets clearly are available to help finance Social Security benefits. But even if one adopts the assumption the Administration did and ignores the Trust Fund's assets, the resulting restatement of the Social Security imbalance over the next 75 years (at 1.0 percent of GDP) is still much smaller than the cost of the tax cut (at 1.68 percent of GDP).
As noted, the Administration estimated the cost of the tax cut to be only 1.0 percent of GDP. It did so by looking solely at the cost of the tax cut, as enacted, in 2010, rather than at the cost of the tax cut when fully phased in and with all of its provisions extended. Under the Administration's estimating approach, the provisions of the tax cut that are artificially slated to expire in 2004, 2005, and 2006 are assumed to die rather than to be extended including a provision scheduled to expire in 2004 that protects millions of taxpayers from being subject to the mushrooming individual Alternative Minimum Tax. The Administration's estimate that the tax cut would cost 1.0 percent of GDP thus assumed that 35.5 million taxpayers would be subject to the AMT in 2010, as compared with 1.4 million in 2001, and that the AMT would cancel out significant parts of the tax cut for large numbers of taxpayers. No credible observer believes Congress will simply allow this AMT-relief provision to expire in 2004. Similarly, the Administration's approach excluded the large cost of repealing the estate tax, a cost that only shows up in years after 2010. Under the tax-cut legislation enacted last year, the estate tax is not repealed until 2010. As tax estimators know, the cost of repealing the estate tax shows up only a year or two after the year in which it is repealed because there is normally a lag of a year or so between the time an individual dies and the time the estate is settled and tax is paid on it. (Even when ignoring the real costs of the tax cut, the Administration massaged its figures; CBO estimates that if all the provisions of the tax cut expire on schedule, it will still cost 1.2 percent of GDP in 2010.)
In short, the Administration's estimate that the cost of the tax cut is 1.0 percent of GDP relied upon gimmicks embedded in the tax bill to make the bill's cost appear lower than it actually is. Paul Krugman, the Princeton economist, wrote in the August 21 New York Times that the Administration's attempts to counter these Center estimates were deceptive and unsuccessful and that "the [C]enter's estimate matches those of the I.M.F. and other independent organizations." (For a more complete analysis of the weaknesses of the Administration's claims, see "Administration Critique of Center Analysis Does Not Withstand Scrutiny," Center on Budget and Policy Priorities, August 3, 2001.)
Providing resources from the rest of the budget to close a portion (rather than all) of the Social Security shortfall, however in conjunction with other Social Security reforms is likely to be essential if any reform plan to restore long-term solvency is to have hope of being enacted. Otherwise, the Social Security benefit cuts and payroll tax increases that will be required as part of any solvency plan are likely to be too large for such a plan to be politically viable.
The recommendations adopted by the President's Social Security commission illustrate this point. One of the three proposals advanced by the commission did little to restore 75-year solvency to the Social Security program. The other two did, but only with substantial cuts in guaranteed Social Security benefits and major adverse consequences for the rest of the budget. These other two proposals would cause a deterioration in the unified budget of more than $1 trillion over the next decade and between $1.9 trillion and $2.2 trillion during the decade from 2013 to 2022 (assuming that all eligible workers participate in the individual accounts). Moreover, the adverse budgetary consequences would persist for decades. Members of the commission were asked where these vast budgetary resources would come from, especially since surpluses outside of Social Security have now been replaced by deficits for the foreseeable future. They were unable to identify a way to finance the provision of these large sums. We do not concur with the commission's proposals, but we would note that scaling back some provisions of the tax cut that have not yet taken effect could help provide the sizeable general-fund resources the commission counts on.
The relative magnitudes of the long-term deficit in Social Security and the long-term revenue loss resulting from the tax cut highlight an important question: Given the demographic and other challenges that lie ahead, is a tax cut that ultimately will provide approximately 35 percent of its benefits to the most affluent one percent of the population the best use of the bulk of the surplus that had been projected outside Social Security and Medicare Hospital Insurance?
General Fund Assistance to Social Security
As alluded to above, the tax cut is likely to make Social Security reform considerably more difficult, if not impossible, for the foreseeable future. The tax cut consumes non-Social Security resources that are likely to be essential to the development of a politically viable package of reforms to restore Social Security solvency.
Transfers from the non-Social Security budget are likely to be crucial to the political viability both of Social Security plans that include individual accounts and of plans that do not. Without such transfers, individual accounts will have to be financed from existing Social Security revenue. Diverting revenue from the Social Security Trust Fund into individual accounts, however, would exacerbate Social Security's projected long-term deficit by reducing the revenue available to the system. Restoring long-term balance to the Social Security system while shifting revenue from the Trust Fund to individual accounts requires larger reductions in Social Security benefits (relative to the benefits that would be paid under the benefit formula in current law) than otherwise would be needed.
An analysis by one of the authors of this analysis and three leading economists and Social Security experts Henry Aaron, Alan Blinder, Alicia Munnell, and Peter Orszag found that if payroll tax revenues equaling two percent of wages were shifted from Social Security to individual accounts and Social Security benefits were maintained at current-law levels for people currently 55 and older, guaranteed Social Security benefits for workers 30 and under would have to be cut more than 50 percent.(4) Including the income projected from individual accounts, the overall retirement income for such workers (their reduced Social Security benefits plus the retirement income they would receive from the individual accounts) would average 20 percent below current-law levels, with some workers losing considerably more than that, if stock market returns were as high in future decades as promoters of private accounts predict. As these figures suggest, the magnitude of the reductions in Social Security benefits that would be necessitated by action to create individual accounts without securing additional revenue from the non-Social Security budget is likely to doom individual account plans that lack another revenue source.
The tax cut is likely to make Social Security reform more difficult because it consumes resources likely to be essential to the development of a politically viable reform plan that restores long-term solvency.
In short, regardless of whether Social Security reform includes individual accounts, transfers from the non-Social Security budget are almost certain to be essential to the development of a politically acceptable reform plan. Such transfers are not likely to be possible without creating or increasing deficits outside the Social Security and Medicare Hospital Insurance trust funds, unless the tax cut is scaled back rather than extended in its current form.
The remainder of this analysis presents in greater detail the projections of the relative sizes of the long-term deficit in Social Security and the revenue loss from the tax cut.
The 75-year Deficit Within Social Security
As is well known, Social Security currently owns assets Treasury bonds backed by the full faith and credit of the U.S. government totaling more than $1 trillion. In addition, Social Security is currently running annual surpluses of roughly $150 billion, and these surpluses are expected to increase in size for a decade. According to the current projections of the Social Security Trustees, annual Social Security tax revenue (which does not include interest on the bonds the Trust Fund holds) will fall below Social Security benefit expenditures starting in 2017, but Social Security as a whole will run a surplus of more than $310 billion in 2017 because it will earn interest income on the bonds it holds. The Trustees expect Social Security to remain in surplus until 2027, even with the increase in the cost of benefits that will occur as the "baby boom" generation retires. At that time, the Trust Fund's assets will total $7.2 trillion (or $3.5 trillion if measured in today's dollars).
The Social Security actuaries calculate, however, that those assets, along with the interest on them and future Social Security revenue, will be insufficient to cover all of Social Security's future costs. Over the 75-year period used for long-term Social Security planning, the shortfall is projected to be $3.7 trillion.(5) In other words, if Social Security currently had $4.9 trillion in assets rather than the $1.2 trillion it now holds, projected Social Security revenues plus the expanded trust fund reserves (and the interest the reserves would earn) would cover projected costs for the next 75 years.
An equivalent measure of the long-term deficit under Social Security is the actuaries' projection that the system faces a projected 75-year imbalance equal to 0.72 percent of the Gross Domestic Product. In other words, if Social Security had additional revenue equal to 0.72 percent of GDP each year, its 75-year deficit would be eliminated. While this shortfall is far from trivial, it is not insurmountable. Last June, the International Monetary Fund concluded that "the long-term financing problems of Social Security are not large, especially compared with those in several other industrial countries, and could be addressed through relatively small adjustments in the program's parameters provided they are implemented quickly."(6)
The 75-year Cost of the Tax Cut
Budget policies are not commonly discussed in terms of their costs over 75 years, in part because the resulting figures would be mind-numbing. But it is instructive to do so, given the concerns over the long-term health of the federal budget that are being emphasized in the Social Security debate.
The Deficit Within Social Security and the Cost of the Tax Cut, Measured in Perpetuity
It is possible to examine the size of the deficit in Social Security in perpetuity (rather than over 75 years) and the cost of the tax cut in perpetuity. Such a comparison can be made by using the same methodology as described here to estimate the permanent cost of the tax cut, and by using figures from the Social Security actuaries to estimate the permanent Social Security deficit. In both cases, the projection horizon is extended far beyond 75 years.
Calculations of costs in perpetuity are subject to even more uncertainty than the already uncertain estimates for 75 years, or even for 10 years. Birth, death, and productivity rates a century or several centuries from now are highly speculative. We would not recommend basing analyses or making policy decisions on specific estimates of costs in perpetuity.
The present value of the cost of the tax cut in perpetuity, estimated as above but extending the analysis beyond 75 years, equals $11.8 trillion (in 2002 dollars). Last year, the Social Security actuaries estimated that the present value of the cost of transforming Social Security from a primarily pay-as-you-go system to a fully funded system would amount to $11.7 trillion (in 2001 dollars).* This cost is approximately equal to the projected deficit in Social Security in perpetuity.** (This $11.7 trillion figure also is the cost that would have to be paid to transform Social Security fully into a system of individual accounts.) With the Social Security trustees' new projection of the Social Security shortfall over the next 75 years (1.87 percent of payroll) being nearly identical to the shortfall they projected a year ago (1.86 percent of payroll), an updated estimate of the size of the shortfall in perpetuity would yield a number very close to last year's $11.7 trillion figure, plus one year of inflation (to express the estimate in 2002 dollars).
In other words, the projected cost of the tax cut in perpetuity and the projected cost of the Social Security shortfall in perpetuity are about the same close to $12 trillion in present value. Shifting the focus beyond 75 years consequently does not alter the basic finding of this analysis that the long-term cost of the tax cut is at least as large as the long-term deficit in Social Security.
* Office of the Chief Actuary, "Unfunded Obligations and Selected Transition Costs for the Combined Old-Age and Survivors Insurance and Disability Insurance (OASDI) Programs," April 5, 2001.
** For further discussion of the connection between the cost of transforming Social Security to a fully funded system and the deficit in Social Security in perpetuity, see John Geanakoplos, Olivia S. Mitchell, and Stephen P. Zeldes, "Social Security's Money Worth,"in Olivia S. Mitchell, Robert J. Myers, and Howard Young, Prospects for Social Security Reform (University of Pennsylvania Press: Philadelphia, 1999).
To calculate the long-term costs of the tax cut, we use estimates of the tax cut supplied by the Joint Committee on Taxation (JCT), the official tax estimator for Congress. The tax cut includes several provisions that expire before 2010, and all of its other provisions expire in 2010. Administration officials and other prominent supporters of the tax cut have made clear that they expect the tax cut to continue and that those who oppose its continuation will be portrayed as seeking to impose hefty tax increases on the American people. The CBO estimates used here show the costs that will occur in 2011 if the provisions are made permanent law. The CBO included these estimates in its annual report of January, 2002.(7)
To project the cost of the tax cut beyond 2011 (the last year for which JCT estimates are available), we assume it will remain a constant share of the economy thereafter. Based on the conservative assumption that the tax cut will remain a constant share of the economy from 2011 on, the cost of the tax over the next 75 years amounts to 1.68 percent of GDP over that period.(8) In dollar terms, the long-term cost of the tax cut amounts to $8.7 trillion in present value.(9) The cost of the tax cut thus is more than twice as large as the long-term deficit in Social Security, which amounts to 0.72 percent of GDP, or $3.7 trillion in present value.
Measured over the next 75 years, the costs of the tax cut, if extended permanently, are more than twice as large as the shortfall in Social Security. While the Administration recently wrote that it is "impossible to afford" the current Social Security system "without large tax increases," this analysis makes clear that what the Administration calls large tax increases are less than half the size of the tax cut the Administration pushed through last year.(10) Policymakers concerned about both the long-term fiscal health of the nation and the restoration of long-term Social Security solvency would do well to examine options for canceling some of the scheduled tax cuts before they take effect (particularly provisions narrowly targeted on those with the highest incomes) and using a portion of the resources as a down-payment in restoring solvency to the Social Security system. Canceling part of the tax cut could, if all goes well, provide the resources for transferring some general revenues to Social Security. Such transfers are likely to be an essential ingredient of a sound Social Security reform package that makes changes in the Social Security program.
Without the resources consumed by the tax cut, the Administration and Congress are likely to have an exceedingly difficult time in fashioning Social Security proposals that both avoid very large benefit cuts and achieve solvency over 75 years. In addition, if the tax cuts take effect as scheduled and are continued after 2010, as the Administration proposes in its current budget, the long-term drain on the budget will exceed the long-term benefit to the budget of eliminating the entire Social Security shortfall.
1. Under the Social Security actuaries' intermediate projections, the projected 75-year deficit amounts to 1.87 percent of taxable payroll. Over this 75-year-period, taxable payroll will amount to 38.4 percent of the Gross Domestic Product when both are expressed in present value. As a result, the 75-year imbalance amounts to 0.72 percent of GDP, which is equal to 1.87 percent of taxable payroll multiplied by 38.4 percent. The figure of 0.72 percent of GDP appears in Table VI.E5 on page 164 of the Trustees Report of March 26, 2002.
2. The assumption that the tax cut will remain a constant share of GDP after 2011 is likely to be conservative. Before the tax cut was enacted, both income tax revenues and estate tax revenues were projected to grow somewhat faster than the economy. This growth was projected to occur primarily because national income is projected to grow faster than inflation (with the resulting income growth pushing some taxpayers into higher marginal tax brackets even though the brackets are indexed to inflation), and because the amount that was exempt from the estate tax was not indexed for inflation. In addition, some provisions of the tax legislation, such as the creation of Roth 401(k) accounts and the increase in the amount that can be contributed to a Roth IRA, are substantially more costly in the long run than in the short run.
3. Another indication of the conservative nature of the estimate used in this paper that the tax cut has a present value equal to 1.68 or of GDP is that in a forthcoming Brookings Institution paper, Alan Auerbach, William Gale, and Peter Orszag estimate that cost at 1.85 per of GDP. Alan J. Auerbach, William G. Gale, and Peter R. Orszag, "The Budget Outlook and Fiscal Policy Options," Brookings Institution, forthcoming. The difference between the two estimates primarily reflects differences in estimating the costs of the interaction between the tax cut and the Alternative Minimum Tax.
4. Henry J. Aaron, Alan S. Blinder, Alicia H. Munnell, and Peter R. Orszag, "Governor Bush's Individual Account Proposal: Implications for Retirement Benefits," The Century Foundation and the Social Security Network, June 6, 2000.
5. The $3.7 trillion is the net present value of the 75-year Social Security deficit. (This figure can be calculated using the year-by-year data backing up table V1.E7 in the trustees' report. Those data allow the calculation of the present value of GDP over 75 years, which totals $519 trillion under the trustees' projections. The $3.7 trillion figure the present value of the Social Security shortfall equals 0.72 percent of $519 trillion.
6. International Monetary Fund, "Staff Report for the 2001 Article IV Consultation," June 28, 2001, page 30.
7. Congressional Budget Office, The Budget and Economic Outlook: Fiscal Years 2003-2012, January 2002, pp. 47, 65. This analysis also makes use of Joint Committee on Taxation estimates of the cost of addressing problems related to the individual Alternative Minimum Tax that were caused by last year's tax-cut legislation. These JCT estimates, provided last year at the request of Rep. Charles Rangel, reflect costs related to extending a provision of last year's tax-cut legislation scheduled to expire in 2004 that provides relief from the AMT. Through 2004, this provision holds the number of taxpayers subject to the AMT to roughly the number that would have been subject to the AMT under the law in place prior to enactment of the tax-cut legislation. In preparing these estimates of the cost of the tax-cut legislation in 2011 if its provisions are extended, the JCT assumed continuation of this AMT-relief provision in such a manner that the number of taxpayers subject to the AMT would continue to track closely the number of taxpayers who would have been subject to the AMT under prior law. (This approach is likely to understate the cost of addressing problems in the AMT, since under the prior law and hence under the JCT estimates used here the number of taxpayers subject to the AMT still would rise from about 1.5 million in 2001 to more than 20 million in 2011.)
The CBO and JCT figures show that with the AMT and other provisions extended, the tax-cut legislation would cost $1.7 trillion through 2011 (before accounting for additional debt service costs). This figure, which forms the basis for the estimates in this analysis, is lower than a comparable estimate made by the IMF, which concluded that "the total cost of the tax cuts is likely to be higher than the $1.35 trillion estimate. Extending the tax cuts through 2011 and extending the AMT provisions would raise the cost of the package to an estimated $1.9 trillion." (Op. cit., p. 28)
The estimate used here for the cost of the tax cut does not include the cost of extending an array of popular tax credits that are regularly extended for a few years at a time and are virtually certain to continue being renewed. That cost is not included here because the recently enacted tax law does not address the issue of extending these credits.
8. In conducting this analysis, we used the actuaries' estimates of GDP in calculating the amount of the tax cut to assure consistency in our cost estimates.
9. The $8.7 trillion figure is the net present value of the tax cut over the next 75 years, discounted at the same discount rate as the Social Security actuaries use to calculate the 75-year deficit in Social Security.
10. Council of Economic Advisers, Economic Report of the President, February 2002, p. 79.