Executive Director Martha Phillips Testimony before the Senate Committee on Aging

Thank you for inviting the Concord Coalition to testify today on the revenue side of the Social Security reform debate.

The Concord Coalition is a bipartisan, nonprofit grass roots organization with members and chapters nationwide dedicated to generationally responsible fiscal policy and long term economic growth. Its co-chairs are former senators Warren Rudman (R-NH) and Sam Nunn (D-GA).

Concord is heartened that on-budget deficits are now smaller than they ever have been since Social Security was taken off budget in 1983 and on-budget surpluses now appear possible. This is attributable not only to years of Congressional fiscal discipline and hard political work but also to the strong growth in revenues as the nation enjoys the longest peacetime economic expansion in our country's history.

Unfortunately, today's healthy fiscal condition is unlikely to continue indefinitely. Nor is its favorable demographic pattern. Today we have a relatively small number of retirees compared to a relative large number of working age citizens. This is about to change, dramatically and permanently.

The large baby boom generation is poised to begin getting Social Security benefits in less than a decade. At the same time, longevity is also increasing, perhaps faster than official projections note. These twin pressures of increasing longevity and the baby boomers' retirement will result in an increase in the portion of elderly from about 12 percent of the nation's population today to 20 to 24 percent by 2040. Today's preschoolers, who will be working age taxpayers when that time comes, will find it a struggle to finance Social Security, Medicare, and large portions of Medicaid--the chief income security and health insurance supports for elderly Americans.

That is why Concord believes it is urgent to use the current political, fiscal, economic and demographic windows of opportunity to undertake reforms now of these programs. Waiting until changing demographics overtake us can only make the task more difficult and the impact on individuals more abrupt and painful. There is widespread consensus that, even though Medicare is a more serious and difficult problem to solve in the long run, Social Security reform is do-able, ripe for debate, and should be at the top of the agenda in 1999.

Reduced to its fundamentals, the problem ahead for Social Security is that the growing real benefits promised under current law cannot be financed by the revenue structure that is now in place. Options to solve that problem fall into two boxes: reducing future benefits from promised levels or increasing the revenues flowing into the system. Some suggest that proposals to establish individually-owned Social Security investment accounts constitute a third box. But these individual-account proposals, reduced to their fundamentals, nevertheless rely on some combination of increased revenues from taxes and/or investment to support retirement benefits as well as a scaling back of benefits offered under the traditional Social Security program.

My testimony today focuses on issues associated with increasing revenues used to finance Social Security and, specifically, the tax aspects of the Social Security reform debate.

The Concord Coalition strongly urges you to avoid relying exclusively on tax increases to close the gap between future benefits and expected revenues.

Closing the gap by relying solely on taxes would constitute an impossibly burdensome tax increase, would have negative labor force consequences, would be generationally unfair, and would use resources that might be needed for other purposes in the future, including dealing with intractable Medicare problems.

It is frequently suggested that a tax rate increase of "merely 2.2 percent" of payroll is all that is required to fix Social Security for the next 75 years. This description of the situation is misleadingly benign.

First, it makes the tax increase seem a lot smaller than it would be. Two percent of anything doesn't seem like much. But when one considers that the entire employer/employee payroll tax supporting Social Security retirement, survivors, dependents and disability benefits is 12.4 percent, adding another 2.2 percent means an increase in the payroll tax of close to 18 percent. For a family with $35,000 income, the increase would raise total FICA, including the 2.9 percent that supports Medicare, from $5355 annually to $6125, an increase of $770. A tax cut of $500 is often described as a meaningful and substantial amount; so is a tax increase of half again as much.

Second, increasing the payroll tax by 2.2 percentage points wouldn't do the job. The 2.2 percent is based on what is needed today to close the long-term 75-year actuarial deficit. Seventy-five years sounds like a long time, and securing the future of Social Security for 75 years seems like a cautious, prudent thing to do. But there's a catch. The 75-year approach that is so frequently used does not ask how much it would take to make the program sound in the 75th year. Instead, this calculation assumes that surplus revenues collected today will be invested in safe government securities so that both principal and interest can be used later when large deficits occur. This calculation therefore assumes that in the 76th year, no assets will remain and that revenues from payroll tax and taxation of benefits will be sufficient to pay only 75 percent of benefits. So this calculation turns out not to be prudent or cautious at all.

Such an approach makes sense for programs such as, for example, unemployment compensation, where the business cycle results in alternating periods of high and low unemployment. States build up reserves during times of low unemployment and then use their reserves to finance benefits when unemployment is high.

But Social Security's projected path is quite different. Instead of recurring cycles of lean and plenty, Social Security's path is one of surpluses in the years just ahead, followed by a steady, unbroken, long-term decline to ever greater annual deficits. By 2032, the year Social Security is projected to use up the last of its bonds and interest, the OASDI trust funds will be running an annual operating deficit of more than $240 billion measured in 1998 dollars. This certainly puts the concept of 75-year "solvency" in a whole new light. The $240 billion is to be supplied, actuarial projections assume, by cashing in the last remaining bonds held by the system. The following year, the deficit will be even larger, but there won't be any more bonds. At that point it won't matter much that, back in 1999 or 2002, the program enjoyed annual surpluses. Those surpluses won't help pay the bills in 2032 or in 2074.1

Suppose that we did agree to increase payroll taxes, starting right away, by 2.2 percent. When we get to, say 2040, benefits are expected to cost 18.13 percent of payroll but revenues, even with the additional 2.2 percentage points, would be only 15.38 percent of payroll, leaving a gap of 2.73 percent of payroll. And this is, of course, exclusive of any Medicare shortfalls we might also be experiencing.

How much would payroll taxes have to be increased to put the program into balance 75 years from now? Last April, the Social Security Trustees' Annual Report indicated that, on the intermediate path, OASDI benefits would cost 19.79 percent of payroll in 2075--6.43 percent of payroll more than income into the trust funds from payroll tax and taxation of benefits. Since the intermediate path makes assumptions about longevity gains that may be too conservative and assumptions about growth in labor productivity that may be too optimistic, the cost of the current program could even be larger.

Annual OASDI Operating Balanceas Percentage of Payroll2
 Current LawWith Additional 2.2%

The bottom line is that if you wanted to solve the entire problem by increasing payroll tax rates, not only for 75 years but for the indefinite future, an increase of 4.7 percent would be required, not 2.2 percent. This would be a 38 percent increase over the current rate and would put the total payroll tax rate at 17.1 percent3 (assuming that current-law provisions that tax benefits for higher income retirees would also continue to supply revenue to the program.)

Payroll tax is a heavy burden:

Another reason to avoid raising payroll tax rates above their current levels is that payroll taxes are already the largest tax owed by three-quarters of American workers (counting the employer's share but not the 2.9 percent Medicare tax.) Many in Congress are urging large income tax cuts this year to give tax relief to beleaguered taxpayers. But for most families, it's the payroll tax that is the largest burden. Many millions of American tax filers are not taxpayers because, after taking into account their dependents, deductions, tax credits and other factors they do not owe any tax under our progressive system of taxation. But we seldom focus on the fact that a great many of those who are deemed not to have enough income to owe federal income taxes nevertheless must contribute 15.3 percent of their earnings in payroll taxes.

Over time, the flat payroll tax has come to provide a much larger share of federal revenues than it did in past decades. While personal and corporate income taxes have declined as a share of overall revenue and a percent of GDP, the regressive payroll tax has increased. In fact, it has increased about 3 percentage points per decade.

Income, Income Tax, and Payroll Tax Liability(for a Median Income Family)
Median income family of four6,75612,80827,61944,61549,531
Income tax liability7361,3593,7924,4124,947
Payroll tax liability(employer, employee shares)3009363,7006,8267,578
Income tax +/- payroll tax43642393-2,414-2,631
Source: Joint Committee on Taxation, JCS-8-97

The payroll tax is a flat tax. Some people think flat taxes are fair; after all everyone pays the same rate, and what could be more fair than that? But in thinking about raising payroll tax rates, it's important to remember that unlike the progressive income tax, which at least attempts to relate tax burden to people's ability to pay, the payroll tax is blind. It doesn't take into account whether the worker who earns $35,000 a year has a spouse who earns twice as much, a spouse who stays home to care for a handicapped child, or no spouse at all. It doesn't care whether the worker supports only himself or herself or has a dozen dependents.

Generational equity:

Raising payroll taxes puts the burden of saving Social Security squarely on the shoulders of those who are working today and the generations that follow them. The generation that has already retired is not affected by this change.

This is ironic since retirement benefits paid to today's beneficiaries provide an excellent return on the taxes they paid in during their working lives. In contrast, young people entering the workforce today can look forward to meager returns on their payroll taxes, and many will actually receive a negative rate of return. Eugene Steuerle and Jon Bakija calculated that after taking into account spousal and survivor benefits and the possibility of dying before retirement age, the real internal rate of return will decline dramatically under the current program for younger cohorts (assuming that full benefits somehow will be paid after 2032.) An average-income one-earner couple in their late 70s today (born in 1925) enjoys a 5.66 rate of return, considerably more than their grandchildren (born in 1975) who can count on a 3.41 rate of return for a one-earner couple or 2.34 percent for a more typical two-earner couple.4 Geoffrey Kollmann has calculated that it took 2.6 years for an average earner, with a dependent spouse, who retired at age 65 in 1980 to recover the combined employee-employer OASI taxes; the same worker retiring in 2020 would need 20.2 years to recover his or her taxes and a maximum earner with dependent spouse would not recover the taxes for nearly 35 years--assuming he lived to 100.5

Today's oldest retirees are the lucky beneficiaries of the start-up period for the Social Security system when payroll tax burdens were low, while young people now entering the workforce are getting into a mature system. Thus, it is unrealistic to expect that younger workers will ever get the rates of returns their parents enjoyed. But that's no justification for making the deal worse for workers in the future. Generational fairness argues against raising payroll taxes for younger workers: it would worsen their rate of return but leave unaffected those who are already receiving benefits, regardless of their wealth.

It is true that Social Security includes disability protection as well as the retirement, dependent and survivors benefits used in the rate-of-return calculations cited above. But, even taking disability coverage into account, the return is poor for today's workers and getting worse. Increasing taxes would make the deal even worse for younger workers. Indeed, it is the poor rate of return that younger workers expect to get that has fueled much of the interest in establishing individual accounts that can take advantage of the higher rates of returns associated with market investments.

Medicare challenges still loom:

The Social Security program does not exist in a vacuum. Making Social Security sustainable would not be nearly so difficult if it were not for the even greater dilemma of addressing projected Medicare shortfalls. For seniors, adequate and affordable health insurance is just as important a part of retirement security as income, and for some, perhaps, more important.

The outlook for the Medicare program is affected not only by the trend toward an older population but also by the prospect of increasing per-capita expenditures as medicine becomes more intensive and a greater portion of beneficiaries are in the over age-85 group that uses medical services to a greater extent than "young" retirees still in their late 60s. On the current path, the question is not whether Medicare eventually will cost more than Social Security, but when. What's more, official projections are, if anything, overly optimistic. They include an assumption that Medicare costs will gradually stop climbing at their current rate, although no one knows how, or when, this de-escalation will occur.

Conventional cost-cutting measures, such as managing care, reducing payments to providers, increasing premiums and co-payments, and combating waste and fraud will not suffice to keep Medicare shortfalls in check. A strong possibility exists that additional revenues will be needed to finance Medicare in the future, particularly if Congress decides to broaden Medicare coverage to include prescriptions and other services routinely included in employer-provided health insurance.

The bottom line is that revenues dedicated to making Social Security sustainable cannot also be used to shore up Medicare. In an ideal world, Congress might choose to resolve the more difficult Medicare dilemma first, and then move on to the "easier" problems facing Social Security. In the real world, Social Security is first up. Care should be taken in solving the Social Security problem not to use resources that will be needed in the future for Medicare. This means that the revenues going into the Medicare HI Trust Fund from taxation of Social Security benefits should not be reclaimed by Social Security.

To the extent that revenues are increased, what are the options?

The case against relying exclusively on a FICA rate increase to solve the Social Security problem is compelling. However, if political leaders conclude that some increase in revenue into the traditional Social Security program will be part of the solution that ultimately is cobbled together, what are the options? They include:

  • raising the tax rate by a small amount,
  • raising the earnings base on which the tax rate is levied,
  • increasing the tax levied on benefits received by retirees, and
  • expanding the base beyond earnings, or even abandoning payroll tax in favor of some other tax base.

Settling on an acceptable set of revenue increases as part of the ultimate Social Security "fix" won't be any easier than agreeing on ways to reduce future benefits. Indeed, there are fierce advocates for reducing payroll taxes and taxes on benefits. But unless future benefits are brought into line with future tax revenues, the Social Security program will not have been made sustainable. "Reforms" that rely on promises that future generations of workers will make good on benefits pledged by today's politicians will not make the program more affordable in the future. Worse, if such reforms give the impression that the Social Security Trust Fund is flush with surplus resources, future Congresses might be tempted to increase promised benefits still further, as has happened so many times in the past.

Raising FICA rates:

FICA is a flat tax, so any increase in the rate will apply to every worker, regardless of economic circumstances. The burden of this tax on lower wage workers, on single mothers striving to maintain middle class incomes, on youth seeking part-time employment and struggling young families is already great. (Even though the Earned Income Credit helps offset some or all of this burden, very few workers see the EIC in their paychecks and many do not file the tax returns necessary to benefit from the credit.) Increasing the FICA burden should be considered a last resort. Indeed, President Clinton, at the forum Concord and AARP hosted in Kansas City in April and again at a forum we hosted in August in Albuquerque specifically ruled out FICA rate increases.

In addition to the regressive nature of this tax and the fact that it is the largest of the "big four" taxes6 that three-quarters of working age families pay, there are labor force considerations as well. The payroll tax has a negative effect on work incentives, wages, and job creation.

FICA taxes are divided evenly between employees and their employers. But just because employers appear to pay half doesn't mean that employees don't ultimately bear the burden of both halves of the tax. Indeed, the economic effect of the FICA tax is the same as if employees were paying the entire 12.4 percent out of their paychecks. This is because employers can devote only so much to hiring labor and still make adequate profits. The more that goes for payroll taxes, or for fringe benefits, the less that is available to be paid directly in wages. Because of FICA, it costs an employer 12.4 percent more to hire an employee than that employee takes home in wages. This wedge has an impact on hiring, pay and job creation decisions.

If the payroll tax rate were increased, it is unlikely that workers' pay would drop overnight in response. In today's hot labor market, pay probably wouldn't drop at all. But the results nevertheless would be felt in time as pay increases came more slowly or, in some cases, did not come at all. And the additional payroll tax could be the last straw tipping some employers to calculate that it would be more profitable to install machines rather than hire more human workers.

Another negative impact of a FICA increase is that entry level jobs would become more expensive for employers to create, and this would tend to suppress job creation. With states trying to help millions of single mothers move from welfare into the work force, almost always at entry levels, and growing numbers of high school and college graduates poised to enter the work force, a FICA rate increase would be a move in the wrong direction. At higher wage levels, employers can slow down pay increases and promotions to gradually offset the impact of FICA rate increases. But near the minimum wage, employers cannot pass on the burden of payroll taxes increases to their employees.

Raise wage base:

If raising the FICA tax rate would have such undesirable consequences for average and lower-wage workers, what about raising the tax base on which the tax is levied?

FICA taxes are levied this year on the first $72,600 of wages earned by each worker. Of course, the vast majority of American workers do not have wages this high, so they are personally indifferent as to where the tax base cap is set. However, currently about 14 percent of wages earned in covered employment exceed the cap and therefore are not subject to FICA. This 14 percent of wages is earned by about six percent of the 140 million workers in the nation who have some wages in excess of the FICA wage cap. They are the ones who would be affected by an increase in the tax.

Since the tax base ceiling is indexed to average wages, and since earnings at the top of the income distribution have been growing faster than average earnings, the share of all earnings that is taxable has consequently declined. Some have suggested raising the tax base cap from its current level of about 86 percent of covered wages back to its "historic" level of 90 percent. However, it is just not true that 90 percent is the historic average. In fact, it's closer to the historic high water mark.

Although the ceiling on FICA wages has declined as a percentage of covered earnings lately, it seldom has been as high as 90 percent. The long-term historic average is just under 84 percent since 1937. The average over the last 50 years since 1949, has been 82.6 percent. Even the last 20 years, since 1979, the average has been 87.7 percent. During the 1950s and 60s, the taxable share of earnings fluctuated between 71 and 82 percent. In the 1970s, Congress passed legislation that sharply raised, and indexed, the tax base. But even so, the share of earnings beneath the cap has remained less than 90 percent in all but two years, the recession recovery years of 1982 and 1983. In this perspective, today's 86 percent level isn't terribly out of kilter with the program's history. And to argue that 90 percent is simply the historic level is dead wrong.

However, if you argue that a reason to raise the wage base is that growing wage inequality has resulted in the base covering a lower percent of earnings, then there is some merit--if you believe that the difficult post-recession 1981 to 1984 years were the model to follow. Even then, you should consider the latest statistics that indicate wages for the lowest paid part of the work force finally are beginning to be pulled up by the current extraordinary business cycle and are growing faster than wages at other pay levels. It may turn out that the percent of covered payroll may halt or even reverse its recent decline.

Suppose you decide to raise the cap to 90 percent and hold it there. How much of the problem would this solve? About one-fifth of Social Security's long-term trust fund deficit. If you want to solve more of the problem this way, consider that you would have to more than double the cap to get even half the savings you could get by eliminating the cap altogether. To get two-thirds of the savings, you would have to triple the ceiling on the tax base.

Pegging the base to 90 percent of the covered earnings in 2000 would mean increasing the base from today's $72,600 to about $100,000. Obviously this will be a popular option for the vast majority of American workers whose earnings are less than that amount, and they would see such an increase as equitable. But suppose you decide to eliminate the cap altogether and tax 100 percent of earnings. Here are some likely results:

  • Benefits in the future would grow to huge amounts since benefits are related to taxable wages. The benefit formula could, of course, be changed so this wouldn't happen, but that would make Social Security an even worse deal than it already is for high earners. If benefits were permitted to grow, they still would be a lousy return on the money paid in. Even so, benefits would be huge in absolute terms. Ultimately the effect would be to increase the cost of Social Security so that six-figure monthly checks could be mailed to the likes of Bill Gates, Warren Buffett and Ross Perot.
  • Even at these high levels, many with consistently high earnings would get back less in nominal dollars than they paid into Social Security. The analogy is to putting your money in the bank, and instead of getting interest, the bank takes money out of your account.
  • As we learned all too well when income tax marginal rates were quite high, people with high earnings, self-employed as well as others, become extremely creative in finding ways to re-characterize their incomes and restructure their businesses to convert as much as possible of their earned income into income from investment. Tax avoidance is a time-honored American tradition, and eliminating the FICA cap would bring it back with a vengeance.
  • The self-employed would be particularly affected. While self-employment earnings comprise only 7 percent of all covered earnings, they comprise nearly 20 percent of earnings above $100,000. Much of this income represents a return to capital invested in small businesses. If you raise the cap substantially, prepare to see many of these businesses hire lawyers to figure out ways around the cap.

Many are calling for not merely raising the cap on taxable wages but eliminating it altogether. President Clinton ruled that out in the Social Security Forum in Kansas City last April, saying it would solve only a portion of the problem but at the cost of tremendously changing the whole Social Security system because there no longer would be a correlation between taxes paid in and benefits received.

Nevertheless, some see eliminating the cap as the reverse of affluence testing on the benefit side. But there's a fundamental difference. Both reforms would move Social Security away from individual equity. But an affluence test would do so by emphasizing social adequacy more, while hiking the tax base would emphasize it less. Both reforms would help bridge Social Security's long-term deficit. But an affluence test would do so by decreasing the system's long-term cost, while hiking the tax base would do so by increasing it.

Some people argue that means testing Social Security by reducing benefits for affluent retirees would weaken the collective bonds that have been such an important part of the Social Security tradition in America. I would argue that eliminating the cap on wages is a far surer way to destroy that bond. The tax levels that some people would be required to pay are far more likely than means testing benefits to create resentment and opposition.

Taxing benefits:

One way to bring benefits into line with revenues is to affluence-test the benefits. The Concord Coalition has long advocated this approach on grounds of equity and fairness. Only those who could best afford to have their benefits reduced would be affected. Public opinion polls show greater support for this option than for almost any other, and it is favored by rich and not-rich alike. This single change would go a long way toward solving the problem. Appendix III of the 1994 Advisory Council on Social Security Report indicated that the plan Concord advocates would improve the long-range OASDI Actuarial balance by 1.65 percent of payroll. In its 1997 volume on Spending and Revenue Options for Reducing the Deficit, the Congressional Budget Office calculated savings from reducing entitlements (Social Security, Medicare and other non-means tested entitlements) to middle and high-income families at between $50 billion and $60 billion annually. The budget savings from reducing just Social Security would approach $35 billion annually. Despite widespread public support and the large savings that it would achieve, policy makers have not embraced this proposal.

Another method for reducing net benefits to retirees with the highest incomes would be to make entitlements fully subject to individual income tax.7 Currently, taxation of benefits generates $9 billion for the OASDI programs and another $5 billion for the Medicare Part A program. If the thresholds were eliminated and up to 85 percent of all benefits were includable in federally taxable income, amounts equal to about 0.21 percent of payroll would be raised. Thus, this option would not go as far as explicit means testing toward making Social Security sustainable over the long term.

Nevertheless, there are several reasons to give this option serious consideration. First, unlike an explicit means test, for which a new administrative structure would have to be established, increased taxation of benefits could be accomplished with the federal income tax system currently in place.

Yes, it's true that taxing benefits more would worsen workers' return on the amounts they paid in over the years. But the deal would be worsened only for those who could afford it. High income beneficiaries might gripe, but equity would be against them as soon as they compared their circumstances to working age people at their same income level. Take, for example, two married couples residing in Fairfax, Virginia, in 1998. One is a 35-year-old working couple with one small child, $30,000 in self-employment income, and a condo worth $100,000. Their bill for the "big four" taxes--federal and state income taxes, federal payroll taxes, and local real estate taxes--comes to $7,906.

Now consider their neighbors down the hall--a 70-year-old retired couple with the same $30,000 in income, split evenly between Social Security benefits and taxable investment returns, and the same $100,000 condo. How much does this retired couple pay in the big four taxes? Nothing. They owe no FICA tax since none of their income comes from earnings. With their personal exemption and total exclusion for Social Security benefits, they owe no federal income tax. And they owe no state and local taxes because Fairfax County, like many jurisdictions, waives property taxes for seniors (but not the young) beneath certain income levels.

Now imagine that the same couples have $75,000 incomes and $200,000 condos. (Assume that the retired couple gets maximum Social Security benefits and that the rest of their income is taxable investment returns.) In this example, the working couple pays $26,101 for the big four taxes while the retired couple pays just $10,642. Of that, the extra 85-percent Social Security tax tier accounts for exactly $1,446.

Although some seniors complain today that any of their Social Security is subject to income tax, most understand that the vast discrepancies between their burden and that of younger families with identical income and greater expenses are impossible to defend.

Shortly before last fall's elections, there was brief talk about reducing the federal income tax on Social Security benefits. Fortunately that idea, which would have been a move in the wrong direction, was dropped once the election was past.

Going Outside the Box

Because changes in taxes or benefits are so politically contentious, there is an understandable temptation to look outside the box for some other way to bring more income into the system, in order to sustain the current program without reducing anyone's promised benefits. The President proposed one such plan in his State of the Union address; dozens of legislators, economists and study panels have offered other plans.

The President proposed crediting general revenues to the Social Security Trust Fund to enable Social Security to continue paying benefits for an additional 20 years, even though cash flow still would turn negative in 2013. This complicated proposal would dedicate 62 percent, or $2.8 trillion of anticipated unified budget surpluses over the next 15 years to the Social Security Trust Funds, in addition to the approximately $2.7 trillion in surpluses the Trust Funds are expected to accrue under current law.

President Clinton also proposed gaining additional income from investing 20 percent of the transferred $2.8 trillion in financial markets. Over time, investment of Trust Fund assets in common stocks will almost certainly yield more income for the Social Security program than investments in Treasury securities. The Administration is counting on the higher yields from private sector investment to achieve an additional six years of trust fund solvency.

Finally, the President proposes using about 12 percent of the anticipated surpluses, or $536 billion over 15 years, to finance Universal Savings Accounts in which individuals would receive a flat contribution from the government with additional matching government contributions based on an unspecified progressive formula.

Nothing in the President's plan would make the Social Security program less costly in the future than it would be under present law. In other words, it doesn't change Social Security's bottom line: under current law, the Trustees project that Social Security and Medicare by 2040 will cost nearly twice what they do today as a percentage of workers' payroll. In fact, the President proposes two expansions of current Social Security benefits and an expansion of the Medicare program to include prescription drug benefits. In contrast, many of the plans proposed by Members of Congress and other groups spell out specifically what choices their sponsors would make in order to scale back the future growth in Social Security benefits. The President alludes to the need to do this in order to make Social Security sustainable indefinitely but declines to outline specific options.

It appears that the President proposes to transfer specified amounts of general revenues to Social Security whether or not the anticipated budget surpluses actually materialize. Although nothing currently on the horizon suggests that the surpluses will not be as large as expected, budget "surprises" frequently occur. The current business expansion is now in record territory, having run 96 months, longer than any previous peacetime expansion. Even an average size recession could erase or substantially reduce the projected surpluses. The Congressional Budget Office has sketched several alternative, plausible, scenarios outlining ways that domestic and international factors could lead to recession.

The President's proposal has been described as reducing the publicly held debt. The Concord Coalition favors reducing the publicly held debt. Debt reduction would free capital for investment in making the economy more productive in the future and, in addition to raising future standards of living, would enable tomorrow's taxpayers to better afford the burden of a large elderly population. However, although the President's proposal would reduce the publicly held debt below the level where it stands today, it would raise the debt compared to where it would stand under current law. In 2004, for example, if budget surpluses were permitted to stand, rather than being claimed for spending increases or tax cuts, the debt would decline from $3.67 trillion at the end of this year to $2.93 trillion at the end of 2004. Under the President's proposal, the publicly held debt would decline to only $3.3 trillion. Thus the President's proposal leaves the publicly held debt some $362 billion higher after five years than under current law.

Essentially, the President's confusing general revenue swap is an effort to make budget surpluses disappear so they aren't claimed for other competing purposes in the form of spending increases or tax cuts. Recent events confirm that nothing has changed, even under divided two-party government, with regard to the willingness of Congress and the White House to join together to use budget surpluses rather than reduce the debt. Earmarking the surplus for Social Security is an attempt to prevent a repeat of last year's end game "emergency" spending glut and use surplus dollars for long term saving rather than immediate consumption.

A decade's experience with off-budget Social Security balances amply proved that merely moving Trust Fund surpluses off-budget does not prevent them from being used to finance sizeable on-budget deficits. An alternative way to deploy the coming surpluses would be to move them out of the budget. This could be accomplished by transferring ownership by using the surpluses to fund individual retirement savings accounts.

The President's proposal to establish individual retirement savings accounts to augment the traditional Social Security program is a variation of this idea. Whereas most proposals for retirement savings accounts would incorporate the accounts into the Social Security program, the President proposes keeping the accounts entirely separate. Unlike many individual account proposals, the President would not fund the accounts through FICA taxes, nor would they be geared to a specified percentage of workers' earnings.

An alternative to the President's individual account proposal would be to earmark a specified percentage of the unified budget surplus (or the entire Social Security surplus if smaller) for direct transfer into individual workers' retirement savings accounts. Like the President's proposal, this plan would reserve the surpluses for long-term savings rather than letting them be consumed for various short-term purposes. In effect, each worker would receive a tax credit in the form of a retirement savings account deposit. The money could not be withdrawn for any purpose other than retirement or death.

Using surplus funds to finance individually owned retirement accounts would do nothing to address the long-term unfunded liabilities of the Social Security program. Ideally, those difficult choices would be made in the same legislation that established the individual accounts. Even if the Congress and the White House were unable to agree on ways to hold down the growth in future benefits, using surpluses to finance individual accounts nevertheless would be a good step to take. After several years, as people began to amass a sizeable stake in their individual retirement accounts, the politics of bringing the traditional program into balance would become more favorable. It might even become possible to divert some money now going to retirement income into Medicare to finance retirement health insurance.

Using surplus funds to finance individual accounts does raise some concerns. One is that care must be taken not to do this in a way that establishes a new stream of entitlements that may not be affordable or sustainable in the future. On-budget surpluses still have not actually materialized, and if and when they do, they may not continue indefinitely. A recession of the severity of the 1991-1992 recession would turn projected surpluses into deficits for several years, for example.

One approach to avoid the "new entitlement" problem would be to devote a percentage of any budget surplus to the individual accounts rather than a dollar amount or a percentage of payroll. This would have the advantage of making the amount going into individual accounts each year contingent on the size of the surplus, if any, in the previous year. It could also create a pressure in favor of maintaining budget surpluses so that individuals would receive an annual infusion of new funds into their accounts.

Administratively, the most efficient means of making deposits into individual accounts would be to have the Treasury do this directly. In effect, a "tax credit" would be transferred into each worker's account and invested as designated, with a default option used for workers who failed to make any designation. Employers would not have to be involved in collecting, transmitting or accounting for the deposits, nor would they have any responsibility for the designation of investment choices. Administrative costs could be held to minimal levels by permitting Treasury to combine individuals' deposits into large transfers to designated private sector investment funds.

A straightforward, transparent and progressive way to determine how much each worker's account would receive would be to simply divide the portion of the budget surplus set aside for these accounts by the number of workers with at least, say, 1,000 hours of work in the preceding year. Thus every member of the work force would receive the same dollar amount deposited into his or her retirement account. In years when the budget surplus became smaller, so would the universal retirement tax credit. Conversely, if surpluses grew, so would the credits. If as expected, the Social Security surplus was $134 billion in 2001, and if 130 million workers qualified for the credit, everyone's accounts would receive slightly more than $1,000. If, on the other hand, half the expected surplus had been used for other spending or tax changes, everyone's accounts would receive only $500.

One reason that many are beginning to focus on using budget surpluses to finance individual retirement accounts is the difficulty of finding alternative funding sources. Most of the plans proposed thus far either carve out some of the current 12.4 percent payroll tax to finance the accounts, or they add on an incremental payroll tax. Both of these approaches raise problems. The more payroll tax that is carved out, the greater the reductions in future benefit growth must be. And adding on more payroll tax amounts to a tax increase. Theorists can explain why this additional saving is good for the economy as well as for individuals. They can also explain that because the proceeds for this mandatory "contribution" would belong to workers, along with the income generated from compounding investments, the mandatory contributions are really quite different from a tax increase. But since the contributions would in most cases reduce current consumption, most people would decide the add-on contribution was a tax, not matter what it was called. Using budget surpluses to fund retirement accounts offers a way out of this dilemma. However, the surpluses are temporary, and will disappear when the boomers retire and when Social Security costs being escalating dramatically. Therefore, diverting surpluses into retirement accounts can be only a beginning step, not a complete solution to the long term problem.


1 As an aside, the fundamental assumption that surpluses invested in government bonds will provide substantial revenue to the program in the future is perhaps not as sound as it may first appear. It's true that Federal Treasury bonds are one of the safest investments in the world, backed by the full faith and credit of the U.S. government. But the bonds that now are counted as assets of the Social Security program are, at the same time, future liabilities of the government. Converting Social Security's paper assets into cash with which to pay benefits will require the government in the future to squeeze loose the funds out somehow -- by reducing other spending, increasing taxes, or trying to support the shortfall through borrowing (which it won't be able to do for very long given the magnitudes involved.). These unattractive options might be more feasible if we had a robust, growing economy that enjoyed much higher per capita GDP than the levels currently projected for the 2030s and beyond. Such an economy would be spurred through saving the Social Security surpluses accruing now and increasing the stock of capital available for investments that increase economic productivity. However, not once since Social Security surpluses began to build up as a result of the 1983 reforms has the full surplus been saved. Every dollar has been used to finance on-budget deficits of the federal government. Last year, for example, while newspaper headlines blasted out the news of a $70 billion budget surplus, the reality was that on-budget accounts had a $29 billion deficit that was masked by Social Security's $99 billion surplus. On this path, the notion of having an economy rich enough to support using the assets of the Trust Fund to meet benefit obligations in the future is far from certain. It is far more likely that severe fiscal pressure will begin building about the time baby boomers begin retiring and Social Security's income from FICA and income taxes is no longer sufficient to pay for benefits.2 Data from 1998 Annual Report of the Board of Trustees of the Federal Old-Age and Survivors Insurance and Disability Trust Funds.3 Unpublished estimates provided by SSA actuaries, cited in testimony by Jagadeesh Gokhale, Federal Reserve Bank of Cleveland, before the Senate Budget Committee, January 19, 1999.4 C. Eugene Steuerle and Jon M. Bakija, Retooling Social Security for the 21st Century, Urban Institute Press, Washington DC, 1994.5 Geoffrey Kollmann, Social Security: The Relationship of Taxes and Benefits for Past, Present, and Future Retirees, Congressional Research Service Report 95-149 EPW, May 12, 1998.6 Federal and state income taxes, federal payroll taxes, and local real estate taxes7 Under current law, beneficiaries with income of more than $25,000 if single and $32,000 if married must include up to half of their benefits in their taxable income. Revenues from this provision are credited to the OASDI Trust Funds. Those with incomes of more than $34,000 if single and $44,000 if married must include up to 85 percent of their benefits in their taxable income. Revenues from the 85 percent increment are credited to the Medicare HI Trust Fund.