Executive Director Martha Phillip's Testimony before the Senate Committee on Aging: Social Security Reform Evaluating "Privatization" Plans

The Problem:

A rapidly aging population in the United States, as in most other industrialized nations, poses serious long term problems for the Social Security, Medicare and Medicaid programs. The huge baby boom generation will begin reaching retirement age in little more than a decade. By 2040, after these 76 million boomers are fully retired, the number of seniors will be at least double what it is today and there will be only two working taxpayers for every retiree.

The 1983 Social Security reforms trimmed benefits and speeded up payroll tax increases in order to amass program surpluses. It was expected that these surpluses would increase net national savings and thereby encourage investments that would boost economic growth and help deal with the inevitable pressures that the retirement of the baby boomers would pose.

What the 1983 reformers did not anticipate was continual large peacetime budget deficits that have resulted in $4 trillion of federal borrowing since that year. The Social Security surpluses are being used to finance consumption-driven budget deficits, not for increasing net national savings, investment and economic growth.

Interlocking solutions:

Balancing the federal budget, and then moving into budget surpluses roughly equal to the Social Security Trust Fund annual surpluses, would increase the pool of capital available for investment in increasing the rate of economic growth.

A larger, more vigorous economy in the next century means more resources will be available to provide, though some combination of public and private means, an adequate retirement for the baby boom generation without requiring excessive, or generationally unfair, contributions from youth and working age citizens.

Politically and practically, to balance the federal budget and keep it in balance, everything must be on the table -- including Social Security, to complete the circle. The long term financial commitments implicit in the structure of the current program are unsustainable. The issue is not whether the program will be revised, but rather when and how.

The longer we wait to address these problems, the more limited and, ultimately, the more draconian the choices and options will be. Many proposals require a substantial lead time in order to have a chance to achieve full impact by the time significant numbers of baby boomers reach retirement age.

Ironically, the current presidential campaign, and most House and Senate campaigns, are maintaining a pact of silence on the Social Security question. Far too many candidates respond, when Social Security is raised, by pledging not to touch it. Yet these are unavoidable issues. "Not touching it" is the one option we do NOT have.

Proposals for Reform

The problems ahead for Social Security can be tackled in a variety of ways. Most proposals currently in circulation include one or more of the following:


  • Raise payroll taxes to levels required to continue financing benefits on a pay as you go basis or to partially finance the projected shortfall.


  • Increase gradually the age for full retirement benefits beyond the age 67 under current law. Age increases are an indirect way to reduce benefits.


  • Reduce initial benefit levels for those with a history of higher earnings or for those who have far above average retirement incomes. Means testing could take the form of changes in the benefit determination calculations, a comprehensive entitlements means test that would explicitly reduce benefit payments for well-to-do retirees, or inclusion of all benefit amounts in income for purposes of the federal income tax.


  • Reduce annual cost of living adjustments, either across the board or in a way that protects those with lowest benefit levels.


  • Invest Social Security Trust Fund assets in higher yielding private market securities.


  • Require workers to build private retirement savings in lieu of or in addition to their Social Security benefits.

A number of proposals involving the last, the so-called "privatization option," have generated considerable attention and enthusiasm in the last year. Part of its attraction may stem from the fact that, at least at first blush, many of these proposals appear to involve neither sacrifice nor losers. Is some form of privatization the answer to the Social Security dilemma? Or is it too good to be true?

The transition from the existing under-funded system to one that depends wholly or substantially on private savings is the greatest hurdle. At least one generation and perhaps several will be required to continue paying for the retirement of their elders while also setting aside reserves to pay for their own retirement. This double burden can be spread out, but it cannot be avoided altogether.

A number of think tanks, academics and legislators are beginning to examine privatization, or "thrift" plans, as they are sometimes called, and several have been drafted into bill form and introduced. The Concord Coalition does not plan to offer a specific proposal of its own. However, we do believe that as these plans are discussed, promoted and considered, they should be measured against a number of evaluation criteria or questions. These are outlined below:


Macroeconomic/Fiscal Concerns

1. Would the reform increase net national savings and income or merely swap ownership of assets?


Shifting ownership of Social Security retirement savings from government accounts to private investments would require the government to borrow more from private markets to cover its annual budget deficits.

Some plans propose issuing staggering amounts of federal debt to pay off existing implicit obligations, leading to huge debt service costs followed by payment of principal when bonds are redeemed. Issuing explicit government debt (recognition bonds, consols, etc.) to replace debt implicit in the current un- or under-funded system does not change total national savings or capital.

Investing some or all of government-owned Social Security assets in stock market or other private investments does not increase net savings to economy but could benefit the trust fund to extent that government profits from the higher amounts that can currently be earned in the private market.

Some maintain that although a plan may increase the level of federal red ink for a number of years, that is acceptable so long as genuine national savings are increasing by several times that amount.

Increasing net savings would require either reducing federal deficits by trimming non-Social Security accounts (and here the proposals become very vague), or requiring additional mandatory contributions (which may look and feel very much like tax increases to many workers). Increasing savings necessarily means reducing current consumption. A dollar saved cannot simultaneously be a dollar spent.

Eventually, when the productivity-enhancing results of increased savings kick in, consumption can return to, and perhaps exceed, previous levels at the same time that substantial savings are being tucked away. But at the beginning of this process, it is inescapable that increased savings means decreased consumption, no matter how the trade-offs are explained or masked.

Household savings behavior: Feldstein and others show strong evidence that the current system reduces propensity to save by insuring against old age income loss, longevity, and inflation. Yet Mitchell & Zeldes (Wharton School) suggest that for younger people who now doubt they will get full Social Security benefits, the decreased uncertainty of a funded system would reduce their precautionary saving. Mandating additional contributions would counteract this.


2. Does the proposed reform retain flexibility to make adjustments down the road if events do not match current projections?


"Recognition bonds" lock in current implicit commitments by formalizing the implicit liabilities -- issuing debt in the present value amount of what workers/retirees would receive under today's program, formally and explicitly converting those expectations into hard cash (or the equivalent in government obligations.)

But Congress has changed Social Security benefits many times in the past and is likely to do so again in response to the impact of baby boomer retirement. People's actual future benefits therefore may not necessarily turn out to be as much as the benefits they would get from the program as it is structured today.

Instead of locking in today's expected benefits by issuing recognition bonds to individual workers/beneficiaries, a reform plan could, instead, issue general debt to the public at large in order to raise any funds needed to meet transition costs. This would preserve the possibility of future reductions in real Social Security benefits to make it more affordable to pay off the debt.

3. Does the plan provide adequate financing?


Relying on the spread today between return to private stocks and return to Treasury debt is risky. If the government goes into the arbitrage business by buying up large quantities of stocks and selling treasury debt to the public to make up the shortfall, this will tend to close the current yield spread between the two kinds of securities.

What safeguards are needed, if any, for paying benefits when the onset of retirement coincides with a down market?

Has the financing been calculated using realistic, even conservative, assumptions regarding rates of return, present value of liabilities, behavioral responses and other factors? Plans that depend on implausibly high real rates of return are suspect.

4. What are the tax consequences of the reform plan?


Should the remaining stream of benefits from "tail" of current system be fully taxable? If so, should the tax revenues be plowed back into retirement benefits? for current retirees? for current workers?

What taxes should apply when private savings are drawn out of the new plan upon retirement? upon disability? upon death? If lump sum? If annuity?

Should contributions by workers/employers be deductible? If so, then should benefits be taxable when received? If contributions are deductible, what impact would this have on IRAs and other tax-advantaged savings?

If a major tax reform is adopted by Congress, how would the reformed Social Security/private savings scheme work? Under a national sales tax or VAT? Under a USA (Nunn-Domenici) tax reform that permits unlimited deductions for savings?


Benefits, Protections, Generational Fairness

5. How do those who are already retired or about to retire fare under the reform?


Ninety-five percent of the current system's benefits go to retired workers, elderly widows/ers, and elderly spouses. Are their benefits maintained, reduced on a progressive basis, reduced selectively, or reduced across the board?

Five percent of the current system's benefits go to child survivors, non-elderly disabled widows aged 50-60, and children aged 18 and under and permanently disabled children of retired workers. How are these beneficiaries provided for in the reform plan?

Some earlier plans suggested by academics would cash out current retirees and replace their stream of benefits with recognition bonds with which they could purchase annuities. This assumes that the private annuity market could and would offer a product, despite the likely adverse selection that would occur.

6. How does the reform plan deal with disability?


Some provision for disability is needed under any reform plan because it requires a long time to build up enough savings/assets to purchase an annuity. Few workers becoming disabled at, say, age 40, would have saved enough in private retirement accounts to fund adequate disability coverage.

Private disability insurance exists, so some argue that this coverage could be handled either publicly or privately. Some plans retain the disability portion of OASDI and privatize only OASI. If the private disability option is selected, should disability insurance be mandatory? Only for workers with dependents?

7. What are the inter-generational tradeoffs? How does the reform plan deal with future cohorts of retirees?


Three groups are involved: those who are retired or about to retire, boomers, and Xers and "Millennials." Each presents a different problem.

A reform plan should make future cohorts of beneficiaries unambiguously better off than they would be under the current system. This should be measured in terms of


  • adequacy (both wage replacement and absolute dollar amounts)
  • rate of return: people should get a better deal for their money

Both these criteria should be measured without "smoke and mirrors" using realistic, even conservative, rates of return.

8. How does the reform plan deal with lower-income workers?


Low income workers now get a very favorable deal under Social Security in terms of the progressive benefit structure that provides a higher wage replacement for lower incomes than for higher.

Some reform plans would subsidize contributions to private savings for lower-income workers since their personal contributions alone would not provide adequate benefits unless they earned an extraordinarily high rate of return.

Other plans have two tiers. The bottom, basic tier provides universal modest benefits to make sure that lower income workers have adequate retirement incomes. However, making this basic tier universal is far more costly than means-tested low-income support.

Although low income workers get progressively higher benefits, on average they live fewer years after retirement and thus receive less in total than the progressive benefit structure would indicate. Some argue, therefore, that no additional protections or supplements are required in order to provide the same benefits for lower income workers that the current system provides.

Also, to the extent that low-income workers do not spend all their accrued assets under a reform plan, they can bequeath the remainder to their heirs, unlike the current system.

9. How does the reform plan deal with people who make poor investments and end up with inadequate retirement resources?


Some people will be unlucky. Some plans call for a public means-tested guarantee to back up the privately invested contribution accounts. Does the offering of indexed Treasury bonds eliminate the need for such a safety net?


Administrative/Fiduciary Concerns

10. How does the reform plan deal with investment decisions?


Does the plan have a sensible set of regulations having to do with investment decisions, including diversification requirements and risk/safety requirements?

Do the plans offer a government default option, permitting workers to default to a "generic" investment plan operated by the federal government (like the federal workers' current "Thrift Savings Plan.")

Investor education is essential. Workers and retirees must become financially literate. How is this provided for?

Objective reporting to the public of essential features (yield, risk, etc.) of various investment plans should complement investor education. Should objective reports be provided by the federal government? By a "consumer reports"-type independent board?


11. How does the reform plan restrict the use of the privately saved funds?


What restrictions are there on taking money out of the retirement savings plan? Are there any hardship exceptions that should be made? Terminal illness? Serious illness? Home purchase? Education? Unemployment? A red convertible?

Is it acceptable to permit withdrawals at any age so long as the worker uses the first dollars to purchase an annuity sufficient to provide poverty-level retirement benefits for worker and spouse?

12. What are the administrative costs under the reform system?


They will always be higher than under the current system. How can they minimized?

Must there be a "switchboard" entity to track contributions, investments, etc?