• SUPPORT CATO
• FOR STUDENTS
• FOR MEDIA
• FREE NEWSLETTERS

• HOME

• ABOUT

• BLOG

• BOOKSTORE

• EVENTS

• EXPERTS

• MULTIMEDIA

• PUBLICATIONS

Books

Reviews & Journals

Public Filings

Policy Studies

Opinion & Commentary

• RESEARCH AREAS

• SPEAKERS BUREAU

• SUPPORT CATO

June 15, 1998
Social Security Choice Paper no. 11

The Myth of the 2.2 Percent Solution

by Neil Howe and Richard Jackson

This paper is based on an earlier work of the authors that appeared in the Concord Coalition's Facing Facts 3, no. 1 (January 10, 1997).

---

Some defenders of the current Social Security system suggest that a "mere" 2.2 percentage point increase in the payroll tax would be sufficient to solve the system’s financial problems. Of course, "mere" is a relative term. If enacted, that tax hike would cost roughly $75 billion in fiscal year 1998 -- the equivalent of a 10 percent increase in everyone’s personal income taxes. But more important, the suggestion that Social Security’s problems can be solved by a 2.2 percentage point increase in the payroll tax is seriously misleading.

The claim of a 2.2 percent solution is based on a measure of Social Security’s fiscal health, called actuarial balance, which counts surpluses accumulated in the program’s trust funds as genuine savings. But the trust funds contain nothing but a stack of Treasury IOUs that will require additional taxes or borrowing from the public to redeem. The more accurate measure of Social Security’s fiscal health is the program’s operating balance -- that is, its annual earmarked tax revenues minus its annual outlays. That balance is projected to turn negative in 2013 and widen to an annual deficit of $734 billion, or 4.8 percent of payroll, by 2031, the last full year the trust funds are technically "solvent." Even if the 2.2 percent solution were enacted, Social Security would still face large and steadily growing operating deficits starting in 2020.

Entirely apart from the savings fallacy, there are other serious problems with the 2.2 percent solution:

• The 2.2 percent figure assumes that the reform would take effect at the beginning of calendar year 1998. Thus it is already out of date. Each year Congress waits, the magnitude of the tax hike required to balance the Social Security trust funds rises.

• The solution is not permanent. In fact, an additional tax hike would be required every year to keep the trust funds in balance over a full 75 years.

• The 2.2 percent solution is based on the Social Security trustees’ "intermediate" projection. If we accept the trustees’ "high-cost" projection, whose key economic and demographic assumptions more closely reflect historical experience, the necessary tax hikes would be at least twice as large.

• The 2.2 percent solution focuses only on the program’s solvency. But simply raising taxes (or cutting benefits), while it may balance the Social Security trust funds, does nothing to redress the program’s other underlying problems, most notably the declining rate of return for young workers.

Ultimately, the real issue in Social Security reform is, not how to meet some official solvency test, but how to ensure Social Security’s economic sustainability and generational equity. This can best be done by transitioning to a funded system of personally owned accounts.

Introduction

Each year, Social Security’s trustees calculate a measure of the program’s long-term fiscal health called its "actuarial balance." In 1998 that balance showed a shortfall equal to 2.2 percent of taxable payroll. In theory, this is the amount by which Congress would have to raise taxes or cut benefits, starting this year, to keep the Social Security trust funds in balance over the next 75 years.

Those who want to preserve the status quo regularly trot out the 2.2 percent figure as evidence of how small the Social Security problem is. How can anyone talk seriously about a crisis, they say, when that crisis could be solved by a "mere" 2.2 percentage point increase in the payroll tax? To bolster this argument, defenders of the current system point to the official status accorded trust-fund accounting in government reports, including that of the recent Social Security Advisory Council. While the council was divided about reform options, all members endorsed actuarial balance as the basic measure of Social Security’s long-term fiscal health.

Such unanimity is puzzling. Actuarial balance not only greatly understates the true future fiscal burden posed by Social Security, it ignores the program’s impact on national savings and generational equity. Ironically, these are precisely the concerns that led 5 of the Advisory Council’s 13 members to advocate moving from the current pay-as-you-go system to a funded system of personally owned accounts.

The Central Fallacy

Proponents of the "2.2 percent solution" rest their case on two facts. First, without any change in current taxes or benefits, the Social Security trust funds are projected to remain solvent until the year 2032.¹ Second, if Congress were to close the program’s actuarial deficit of 2.2 percent of payroll, the trust funds would be solvent for the next 75 years, or through the year 2072.² Either way, this "solvency" has a precise statutory definition: It means that the Social Security trust funds are projected to possess sufficient assets, and hence budget authority, to cover current-law benefits until these dates.³

The problem is that this statutory definition of solvency ignores economic reality. It assumes that trust-fund surpluses accumulated in prior years constitute genuine economic savings that can be drawn down to cover trust-fund deficits incurred in future years. But the reality is, they don’t: The surpluses are spent each year by Congress on current cash outlays; the assets held by Social Security consist of nothing but Treasury IOUs. When it is time for the trust funds to redeem these IOUs, Congress must increase taxes, cut other spending, or borrow more from the public to raise the cash. Actuarial solvency thus rests on nothing more than Congress’s promise to spend money that it has yet to raise. The open-ended appropriation that Congress has granted Social Security expresses its long-term support for the program, but it does not say anything meaningful about how the long-term cost will be funded -- or, indeed, whether the cost is affordable at all.

A much better indicator of Social Security’s future cost burden is its annual operating balance -- that is, the projected yearly gap between its outlays and its earmarked tax revenues. The amount by which that balance falls is the amount of new cash Congress will have to raise if promised benefits are to be paid. Under current law, Social Security’s operating balance is projected to remain roughly constant (in current dollars) over the next decade and then decline in every year after 2005. The balance will turn negative in 2013 and widen to an annual deficit of $734 billion, or 4.8 percent of payroll, by 2031, the last full year the trust funds are technically "solvent." Even if the 2.2 percent solution were enacted, Social Security would still face large and steadily growing operating deficits starting in 2020.

Figure 1 shows these operating-balance projections, both with and without the 2.2 percent solution. A cursory glance reveals a year-by-year outlook that is strikingly at odds with the benign future suggested by the actuarial-balance measure.

Figure 1
Annual Social Security (OASDI) Operating Balance:
Intermediate Projection with and without 2.2% Solution

Proponents of the 2.2 percent solution will counter that trust-fund IOUs do indeed constitute genuine economic savings -- and that it is the operating-balance measure that is misleading. Why? Because, say the proponents, when the trust funds run a surplus, the Treasury doesn’t have to borrow as much from the public to pay for current cash outlays from the rest of the budget. The bonds that the Treasury will have to sell to the public tomorrow to cover operating deficits are nothing more than the bonds (plus interest) that the Treasury, thanks to the trust funds, doesn’t have to borrow from the public today. Thus, Social Security won’t add anything to the publicly held debt all the way up to its date of insolvency. To proponents of the 2.2 percent solution, it doesn’t matter whether trust-fund surpluses are accompanied by overall budget surpluses. (In fact, the federal budget has thus far been running large yearly deficits and is unlikely to achieve more than modest surpluses in the coming decade.) All that matters is that, absent Social Security, the overall deficit would be (dollar for dollar) larger.

This line of reasoning assumes that our political system tracks and targets some desired balance in the federal budget excluding Social Security -- which for simplicity we will call the on-budget⁴ -- rather than in the unified budget.⁵ Although this is an interesting proposition, there is little evidence to support it. Over the past 15 years, the era in which we built up nearly all of today’s trust-fund assets, there have been many plans and processes aimed at balancing the budget. But all of them aimed at balancing the unified budget; none aimed at balancing the on-budget, that is, the unified budget excluding Social Security. To be sure, numbers for the on-budget and federal funds balance are published, and Congress does have to vote periodically on the "gross" debt ceiling. But aside from such exceptions, virtually the entire gamut of budget rules and targets now used by the Executive Branch and Congress focuses on one and only one definition of budget balance -- namely, the unified balance.

The same emphasis on unified balance dominates the recent bipartisan commitment to balancing the federal budget by 2002. Once again, no one is talking about balancing the on-budget. No one, that is, is aiming for a unified budget surplus of $139 billion in 2002.⁶

One might, perhaps, hypothesize that the nation’s leaders are now implicitly "targeting" a $139 billion on-budget deficit. If so, one would expect them to focus solely on changes in projections for the on-budget balance. But they don’t. Up until 1998, projections for the on-budget were widely regarded as footnotes—when they were calculated at all. As former chairman of the Council of Economic Advisers Herbert Stein once remarked, "You must get 300 pages into the budget before finding a table with the surplus in Social Security and the deficit in the rest of the budget shown separately."⁷ This year, the administration’s proposed Social Security "reserve" for the first time put the spotlight on the on-budget balance. Even this proposal, however, continues to leave Social Security’s trust-fund surpluses unbudgeted. Despite the rhetoric, notes economist Martin Feldstein, "In reality, saving Social Security comes last."⁸ If leaders were truly intent on saving Social Security surpluses, one might also expect that debates over changes in Social Security outlays and revenues would take place in isolation from debates over changes in the rest of the budget. But of course there is no such isolation, either on the outlay or on the revenue side.⁹

The evidence that the unified budget is what matters is indeed overwhelming. The unified budget is what gets cited in nearly every media story and political speech. The unified budget is what defines the constitutional boundary between citizen and state.¹⁰ And the unified budget is what economists have always favored -- which is why the Executive Branch formally adopted it in the 1960s,¹¹ why Congress has been reluctant to abandon it in the 1990s,¹² and why it is used by governments and national accounting systems worldwide. Even more persuasive than the consensus of economists is the verdict of the federal agencies that are themselves in charge of evaluating accounting and financing issues. According to the Congres-sional Budget Office,

The government’s ability to finance a given schedule of Social Security benefits is more closely related to the future strength of the economy than to the solvency of the Social Security trust funds. This is because the assets of the Social Security trust funds do not represent any real stock of resources set aside to pay for benefits in the future. Benefits due in a given year must be financed either by government revenues collected in that year or by borrowing.¹³

According to the General Accounting Office,

The budgetary reality is that the increased payroll taxes enacted in the 1983 amendments are being used to finance the current operations of government. This will continue to be true so long as we run on-budget deficits.¹⁴

According to the Congressional Research Service,

Perhaps the biggest misperception is that the Social Security trust funds represent actual resources to be used for future benefit payments, rather than what is in reality a promise by the government to take the steps necessary to secure resources from the economy at that time. The accumulation of these securities in itself does not guarantee economic growth nor will it necessarily mitigate the problems associated with the rapid increase in Social Security’s cost when the baby boom retires in the next century.¹⁵

According to the House Committee on Ways and Means,

When the government issues a security to one of its own accounts, it hasn’t purchased anything or established a claim against some other person or entity. It is simply creating an IOU from one of its accounts to another. Hence, the building up of federal securities in a federal trust fund—like that of Social Security—is not a means in and of itself for the government to accumulate assets.¹⁶

And according to the Social Security Administration,

In effect, the general account is spending the OASDI surpluses to finance general account deficits. . . . In this situation, the trust fund more accurately represents a stack of IOUs to be presented to future generations for payment, rather than a buildup of resources to fund future benefits.¹⁷

Federal Reserve Board chairman Alan Greenspan has repeatedly emphasized that a program’s fiscal burden must be judged by its future impact on the yearly unified balance—not by some long-term trust-fund balance.¹⁸ Sen. Daniel Patrick Moynihan (D-N.Y.), who was instrumental in putting together the 1983 Social Security Reform Act that has given rise to today’s trust-fund surpluses, has since disowned his own handiwork. Remarking on Congress’s failure even to attempt to achieve an on-budget balance, Moynihan has noted that these trust-fund surpluses have simply allowed Congress to tax less and spend more than it otherwise would. Trust-fund accounting, he concludes, amounts to "thievery."¹⁹ (The late Sen. John Heinz (R-Penn.), asked whether Moynihan’s description of the trust funds was accurate, is said to have replied, "Certainly not. What is going on is not thievery. It is embezzlement.")²⁰

This budget reality explains an apparent paradox often posed by defenders of the current Social Security system: If individuals and private pensions are engaged in genuine savings when they invest in Treasury bonds (and they are), why then is such investment by the Social Security trust funds fundamentally different? The reason is that when a private citizen decides to buy a Treasury bond, that purchase doesn’t induce Congress to incur a larger debt than it otherwise would. This is not, strictly speaking, a public versus private distinction: Nothing prevents the federal government from establishing a trust fund that genuinely "funds" claims held "in trust" and that doesn’t allow budgetary double counting. In point of fact, however, few federal trust funds are set up this way.²¹ Most merely function as ad hoc memo accounts whose sole purpose is to accumulate budget authority.

Admittedly, the whole question of how Social Security surpluses do or do not change the overall budget balance (and national savings) will strike many Americans as rather abstract. A more vital question is whether what the defenders of Social Security are advocating for future fiscal policy makes sense, given where our economy and society now find themselves.

In this larger context, the logic of actuarial balance rests on a fantastic proposition -- namely, that America’s lofty savings rate since the mid-1980s, due to the thriftiness of government, is paying in advance for a sumptuous deficit banquet 20 years from now. In other words, because we’re saving more than we ought today it’s OK to save less than we ought tomorrow. Neither half of this proposition is defensible. Clearly, we have not been saving more than we ought to have been over the past 15 years -- an era in which U.S. rates of net national savings (and economy-wide productivity growth) have been well below our historical average and the contemporaneous average for other developed nations. Nor, given the heavy downward pressure on national savings expected to accompany America’s "age wave" beginning in the 2010s, could anyone reasonably advocate additional ways to pull down our savings rate in the future.

In the end, there’s only one possible defense of trust-fund accounting, and that is to insist that Social Security is an entirely separate realm of government—not just another federal agency but a kind of independent republic exempt from the sovereignty of Congress. Thus, if we didn’t save enough publicly in the 1980s and 1990s, it’s the fault of the rest of government. And if we will need to undertake some draconian belt-tightening in the 2020s and 2030s, that too is the fault of the rest of government.

But why should we think of Social Security this way? Not because the contributions are owned or managed by the workers who make them and not because the benefits are economically funded or legally guaranteed. None of this is true. Social Security is and always has been a tax-and-transfer program, shifting income from current workers to current beneficiaries according to benefit rules that Congress can (and often does) change without warning and without regard to prior individual contributions. As it turns out, the most often-cited reason for why Social Security should be considered a self-contained deal is that Congress calls it a trust fund. Thus does the logic become a perfect and empty circle.

Other Problems

Entirely apart from the savings fallacy, there are other serious problems with the 2.2 percent solution:

• The 2.2 percent solution is not small. If carried out entirely on the revenue side, the 2.2 percent solution would mean a tax increase of roughly $75 billion this fiscal year—the equivalent of a 10 percent increase in everyone’s personal income taxes. From FY 1998 through 2002, it would mean an increase of roughly $400 billion. A tax hike of this size is inconceivable in today’s political climate. Certainly, no one would dream of calling it "small" or "bearable" if it were actually proposed.

• It is too late to enact the 2.2 percent solution. The 2.2 percent actuarial deficit figure published in the most recent (1998) Trustees’ Report refers to the 75-year period from 1998 through 2072. Thus, any "solution" based on this figure would have had to have gone into effect at the beginning of calendar year 1998. Clearly it is too late for that, unless Congress resorts to the extreme (and almost unthinkable) expedient of a tax hike that is not only large but retroactive. Practically speaking, several years would have to pass before a savings package could be studied, debated, enacted, and fully implemented. Yet each year that we delay, the share of payroll required to bring the trust funds into balance rises.22 Let’s assume that the earliest Congress could enact a savings package is 2002. This delay would raise the required savings by roughly 0.3 percentage points—to 2.5 percent of payroll.

• Even if enacted, the 2.2 percent solution would let Social Security’s actuarial balance slip into deficit the very next year—requiring yet another "solution," followed by yet another deficit, and so on. The foregoing calculation assumes that 2072 will forever remain the horizon for trust-fund solvency. In other words, it assumes that while we today would require the trust funds to be in balance over 75 years, our children will be satisfied with 40 years, and our grandchildren will drive over the cliff with their eyes closed. But Social Security has always (and sensibly) been required by law to calculate its balance over a full 75 years. Assuming this remains true, each passing year will add a new higher cost year and subtract a new lower cost year, increasing the projected actuarial deficit. In the late 1990s, this so-called valuation period increment has been adding, each year, roughly 0.08 percentage point to the deficit calculated in the prior year.23 This dynamic is certain to continue for several decades to come.24 Again, assume that Congress cannot enact a solution until 2002. By that year, the required savings will be about 0.5 percentage points higher—that is, it will be 3.0 percent of payroll, not 2.5 percent. And even if this 3.0 percent solution were enacted at this time, the total would keep climbing in later years.

• Even these figures may understate the magnitude of the required savings. The above figures are all based on the trustees’ "intermediate" scenario, which generates the official projections presented to Congress and the public. Many policy experts, however, believe this scenario may be overly optimistic since it assumes faster productivity growth and slower longevity gains than America has actually experienced over the past quarter century. If we accept the trustees’ "high-cost" scenario, whose key economic and demographic assumptions more closely reflect historical experience,25 Social Security’s 2.2 percent of payroll actuarial deficit instantly becomes a 5.4 percent of payroll deficit. And if Congress waited until 2002, a 75-year balance under this scenario would require additional savings of (at least) 2.0 percent of payroll. Let’s call it the "Seven Percent Solution."

The Real Issue

In the end, trust-fund accounting sidesteps the real issue, which is not how to meet some official solvency test but how to ensure Social Security’s economic sustainability and generational equity. It is those concerns that have led a growing number of voices to advocate transitioning to a funded system of personally owned accounts. By linking benefits to genuine economic savings, such a reform not only addresses the fiscal challenge. It would also generate much higher returns on contributions for future retirees, and hence would redress the cascading pattern of generational inequity that now destines each new birth cohort of Social Security participants to receive a worse deal than the last.

---

Notes

1. In this paper, "Social Security" refers to both Old-Age and Survivors Insurance and Disability Insurance; the "Social Security trust funds" refer to the combined trust funds for both programs (OASDI). Unless otherwise indicated, all figures reflect the "intermediate" projections published in the 1998 Annual Report of the Board of Trustees of the Federal Old-Age and Survivors Insurance and Disability Insurance Trust Funds.

2. In other words, the trust funds would be in actuarial balance. Technically, to be "in actuarial balance" means that the trust funds are projected not only to maintain sufficient assets to pay out benefits throughout the 75-year valuation period but also to maintain in every year an asset reserve equal to no less than 100 percent of the current year’s benefit outlays. From the 1972 to the 1987 Trustees Reports, actuarial balance was calculated by means of an "average-cost" method that simply adds up all future balances (plus current trust-fund assets) as a percent of payroll and divides this sum by 75. In 1988 the trustees switched to a "level-cost" method that is similar to the method used before 1972. The level-cost method discounts all future dollar amounts by the projected interest rate.

3. Congress has granted to Social Security a permanent appropriation, which automatically gives it a budget authority (that is, the power to spend money) equal to the assets in its trust funds.

4. Congress currently regards the Social Security and Postal Service trust funds as nominally "off budget." The rest of federal spending is "on budget." Since the Postal Service balance is negligible relative to the Social Security balance, this division effectively divides the budget into Social Security and everything else.

5. The "unified" budget (also known as the "total" or "consolidated" budget) encompasses all spending appropriated by the U.S. Congress.

6. The CBO projects that, in FY 2002, the unified budget balance will be +$69 billion and the on-budget balance will be –$69 billion. The off-budget surplus (consisting entirely of the Social Security trust-fund surplus) is expected to be +$139 billion. See Congressional Budget Office, "The Economic and Budget Outlook: Fiscal Years 1999–2008," January 1998.

7. Herbert Stein, "The ‘Threat’ of a Budget Surplus," Wall Street Journal, April 7, 1988.

8. Martin Feldstein, "Let’s Really Save Social Security," Wall Street Journal, February 10, 1998.

9. The charge is frequently made, on the outlay side, that Social Security is "crowding out" other forms of (mainly domestic discretionary) federal spending—a charge made plausible by the fact that total federal spending (as a share of GDP) has been far more stable over the postwar era than any major component of federal spending, such as Social Security. A similar argument could be made on the revenue side, where the substitution of on- and off-budget receipts is sometimes quite explicit. Note, in particular, that the rapidly growing Earned Income Tax Credit is expressly designed to give low-income single workers relief on their personal income taxes equal to their FICA payments.

10. Aside from obligations arising from contracts with individual parties, publicly held Treasury debt is the only type of federal liability that represents constitutionally protected property. Treasury debt held by other federal agencies enjoys no such protection.

11. The consensus among economists, as expressed in the Report of the President’s Commission on Budget Concepts, October 1967, persuaded President Lyndon Johnson to incorporate all federal programs, including trust funds such as Social Security, into a single "unified budget"—which was first presented in the President’s Budget for FY 1969.

12. The 1983 Social Security amendments included a provision to designate Social Security as "off-budget" by FY 1993, a target date that was speeded up (by the original Gramm–Rudman–Hollings legislation in 1985) to FY 1991. In 1994, Congress went a step further by transforming Social Security into an independent agency (as of March 31, 1995). None of these administrative measures, however, has significantly changed how Congress, the media, and most economists look at the budget.

13. Congressional Budget Office, "The Economic Effects of Uncompensated Changes in the Funding of Social Security," April 1991.

14. General Accounting Office, "Social Security: The Trust Fund Reserve Accumulation, the Economy, and the Federal Budget," January 1989.

15. Geoffrey Kollmann, "Social Security: Investing the Surplus," CRS Report for Congress, no. 91–129 EPW, January 1991.

16. U.S. House of Representatives, Committee on Ways and Means, 1996 Greenbook, November 1996.

17. John C. Hambor, Division of Economic Research, Social Security Administration, "Economic Policy, Intergenera-tional Equity, and the Social Security Trust Fund Build-Up," Social Security Bulletin, October 1987.

18. See, for example, Greenspan’s oral testimony before the (Kerrey-Danforth) Bipartisan Commission on Entitlement and Tax Reform, July 15, 1994.

19. Daniel Patrick Moynihan, "It’s a Matter of Trust, Rosty," Washington Post, March 22, 1990.

20. Ibid.

21. Of the (roughly) 120 federal budget accounts classified as "trust funds," only one major account qualifies as a trust fund in the sense described here: the Thrift Savings Plan, a defined contribution pension supplement administered by the Federal Retirement Thrift Investment Board for federal employees under FERS.

22. The reason it rises is that, if Congress delays, it has fewer years to accumulate the surpluses necessary to defray the long-term deficits. Note that this calculation assumes that the horizon of the valuation period remains fixed at 2072.

23. The total actuarial deficit does not always grow by this amount, since other adjustments (including changes in assumptions) can also affect the actuarial deficit from one year to the next. Between 1997 and 1998, for example, the actuarial deficit remained stuck at 2.2 percent. This happened for two reasons: First, the trustees assumed somewhat faster economic growth over the next few years, which increases Social Security’s near-term operating surplus; and second, the trustees raised their real interest rate assumption, which inflates the value of the return to Social Security’s paper trust funds. Together, these changes neutralized (for this year at least) the valuation period increment.

24. The relentless upward drift in the actuarial deficit is a problem well known to experts. It prompted the recent Social Security Advisory Council to recommend that any reform plan not only ensure solvency over the next 75 years but somehow stabilize (if not eliminate) the future actuarial deficits that will reappear each time the valuation period is recalculated.

25. See Richard Jackson, "The Inevitability of Entitlement Reform: An Analysis of the Social Security and Medicare Long-Term Cost Projections," in Final Report to the President: Bipartisan Commission on Entitlement and Tax Reform (Washington: Government Printing Office, January 1995), pp. 121–45.

© 2009 The Cato Institute
Please send comments to webmaster