September 30, 2002
THE SHAW Social Security PROPOSAL:
THE ROLE OF MASSIVE GENERAL REVENUE TRANSFERS
by Kilolo Kijakazi, Richard Kogan, and Robert Greenstein
View PDF of this report
To access the PDF files, right-click on the underlined text, click "Save Link As," download to your directory, and open document in Adobe Acrobat Reader.
In December 2001, the President’s Commission on Strengthening Social Security issued three proposals to replace a portion of traditional Social Security benefits with private accounts. Two of these proposals would restore long-term Social Security solvency.
Both of those plans also would reduce traditional Social Security benefits substantially and, for many beneficiaries, leave combined benefits from Social Security and private accounts below the benefit levels provided under the current-law benefit structure. The Commission’s plans illustrated the fact that diverting revenue from the Social Security Trust Fund to individual accounts would require reducing traditional Social Security benefits by larger amounts than would be necessary if funds were not diverted in this manner.
Most Members of Congress who support conversion of a portion of Social Security to private accounts have distanced themselves from these plans, in part out of concerns over the impact on public opinion of proposals that would leave large numbers of beneficiaries with lower overall benefits than under the current Social Security benefit structure. Within days of the Commission’s final action in adopting its proposals, several House leaders issued alternative Social Security plans that they said would restore Social Security solvency and partially convert Social Security to individual accounts with no benefit reductions relative to the benefits scheduled under current law and no payroll tax increases.
One of these plans was introduced by Rep. Clay Shaw, the Chairman of the Subcommittee on Social Security of the House Ways and Means Committee. The Shaw bill (H.R. 3497) is a significant revision of a plan released in April 1999 by Rep. Shaw and then-Rep. Bill Archer but never formally introduced. The Shaw plan would restore solvency while increasing benefits, and with no change in the payroll tax. This analysis examines the Shaw plan.
How was Rep. Shaw able to develop a plan that would preserve the Social Security benefit levels of all beneficiaries and avoid any payroll tax increases when the Commission, after eight months of deliberation, was unable to produce plans that did not entail substantial benefit cuts relative to the current-law benefit structure? The answer is straightforward. The Shaw plan relies upon massive and unprecedented general revenue transfers.
The Commission plans, themselves, entail large general revenue transfers. In fact, a principal criticism and key weakness of the Commission proposals is that, as an analysis by Peter Diamond (an M.I.T. economist widely considered one of the world’s leading experts on retirement systems) and Brookings senior fellow Peter Orszag demonstrates, these proposals depend upon multi-trillion dollar transfers from the rest of the budget without offering any way for the rest of the budget to come up with those funds. Yet the general revenue transfers in the Shaw plan greatly exceed those in the Commission proposals.
Data from the official analysis of the Shaw plan that the Social Security actuaries issued indicates that the general revenue transfers required under the Shaw plan over the next 75 years would be one-third larger than the level of transfers that would be required to restore Social Security solvency fully for the next 75 years without making any changes in the Social Security system.
- The actuaries projected last year that restoring Social Security solvency with no changes to Social Security would entail general revenue infusions totaling $3.4 trillion in “net present value.” (“Net present value,” as used here, is the amount that, if transferred to the Social Security Trust Fund today, would be sufficient — along with the interest it would earn — to enable the Trust Fund to pay full benefits for the next 75 years. “Net present value” is the standard measure that economists and fiscal analysts use in measuring costs over long time periods such as 75 years.)
- Policymakers and analysts from across the political spectrum, including many policymakers who favor partially replacing Social Security with private accounts, have warned that general revenue transfers of this magnitude would be fiscally irresponsible, particularly if they are not accompanied by deep cuts in other federal programs or substantial tax increases to free up the necessary funds.
General Revenue Funding Required Over 75 Years Under the Shaw Plan
Amount of Transfers Required to Restore Solvency with No Changes in Social Security
Transfers Required Under the Shaw Plan
In Net Present Value
As a Share of GDP
- The Shaw plan entails general revenue transfers totaling $4.5 trillion in net present value over the next 75 years, or one-third more than the $3.4 trillion in transfers that would be required to restore solvency for the next 75 years without making any changes in the system.
It thus should come as no surprise that the Shaw plan is able to restore solvency while avoiding hard choices and including no benefit reductions or payroll tax increases (and including some benefit increases). The plan relies on transfers of unprecedented sums from the rest of the budget.
The plan does not indicate from whence this money would come. Long-term budget projections issued by the General Accounting Office and the Congressional Budget Office, as well as the long-term projections of independent analysts, show very large long-term budget shortfalls outside Social Security. The retirement of the baby-boom generation will lead to substantial increases in costs for Medicare and the long-term-care component of Medicaid, while the aging of the population will slow labor-force growth and hence the growth of the economy. Recent estimates suggest that the long-term budgetary shortfall outside Social Security is at least twice as large as the long-term Social Security shortfall.
In short, no funds are available in the budget to make the massive transfers the Shaw plan entails. Where the funds would be found is entirely unclear. If the Shaw plan were enacted, the general revenue transfers it would require would represent a “magic asterisk” of historic proportions. Budget cuts or tax increases amounting to $100 billion in this year, and growing in size with the economy in subsequent years, would be needed to offset the costs of the general revenue transfers envisioned by the Shaw plan.
To get a sense of the magnitude of these costs, the complete elimination of all education and veterans programs would fall short of covering the required general revenue transfers. The box on page 8 provides examples of the dimensions of the budget cuts that would be needed to produce $100 billion a year.
The Shaw plan also suffers from other weaknesses. Under the plan, the revenue transfers would not go directly from the general fund of the Treasury to the Social Security Trust Fund. Rather, they would take a more circuitous route. General revenues would first be paid to personal accounts, called “Social Security Guarantee Accounts.” The amounts in those accounts would then be invested in stocks and corporate bonds. Later, after individuals started to collect Social Security retirement or disability benefits, all but five percent of the amounts in their accounts would be transferred over time to the Social Security Trust Fund.
Because the general fund amounts would be invested in the market, the Social Security actuaries’ analysis of the Shaw plan stresses that the projection that the plan would restore long-term solvency depends critically on the stock market performing well over the next 75 years. If the market turned out to rise by significantly less than the actuaries assumed in their assessment of the plan, solvency would not be restored.
Restoration of solvency also would depend upon the political sustainability over time of a key provision of the plan that may prove highly controversial — a provision that effectively imposes a 95 percent tax rate on the personal accounts the plan establishes. Under the plan, five percent of the amount in an individual’s account would be paid to the individual in a lump sum at the time that the individual began to draw Social Security benefits. The other 95 percent would be transferred over time to the Social Security Trust Fund to help it cover the costs of paying Social Security benefits, rather than (except in a small number of cases) going to the beneficiary (or the beneficiary’s heirs). If political pressures led to the 95 percent figure being significantly reduced and more of the account balances going to beneficiaries and heirs, Social Security solvency would be impaired.
Also of importance, the personal accounts the plan would establish would eventually dominate the stock and bond markets. They ultimately would account about half of the total amounts invested in these markets. There is question as to whether it would be wise for these quasi-private, quasi-governmental accounts to dominate the market in this fashion.
Finally, the plan contains perverse incentives for people who become disabled or reach retirement age and do not expect to live that long to decline to begin receiving Social Security benefits. When people died before beginning to draw benefits, their accounts would be passed on in full to their heirs. Once someone had received any Social Security benefits (including as little as one month of benefits), the balance remaining in the account would, over time, be deposited in the Social Security Trust Fund; none of it would go to heirs. This would create an incentive for those in failing health to avoid drawing Social Security benefits if they could afford to do so. Those who could not afford to do so would not have this option.
This analysis first provides a description of the Shaw plan. It then examines these and several related issues.
How the Shaw Plan Works
The Shaw proposal, which is known as the “Social Security Guarantee Plus Plan,” would establish a system of voluntary personal accounts. These accounts would be designed to guarantee that participants receive a benefit at least as large as the Social Security benefit they would receive under current law. Although the system would be voluntary, it is likely that all workers would participate. No one could do worse than under current law, and beneficiaries would generally receive more than under current law.
Unlike the plans that the President’s Social Security Commission developed last year, the Shaw plan would not divert Social Security payroll tax revenue from the Social Security Trust Fund to personal accounts. Instead, the accounts would be funded entirely through general revenue from the rest of the budget.
Each year, the federal government would place in an account for each participating worker an amount equal to a specified percentage of the worker’s wages or salary. The amount would equal three percent of roughly a worker’s first $18,000 in wages and salary (in 2002), plus two percent of any wages and salary between about $18,000 and the maximum amount of wage and salary income that is subject to the Social Security payroll tax, currently $84,900. The $18,000 and $84,900 thresholds would be adjusted annually to reflect average wage growth in the U.S. economy.
Account-holders would select an account manager from a list of “certified account managers,” selected by a “Social Security Guarantee Board,” which the Social Security trustees would establish. The certified account managers would consist of banks, brokerage firms, insurance companies, investment advisors, and the like. Each account manager would be required to invest 60 percent of the funds under that firm’s management in stock mutual funds that track broadly representative stock indices the Social Security Guarantee Board had chosen, and to invest the other 40 percent of the funds in mutual funds consisting of broadly representative portfolios of corporate bonds. Individual account-holders could not elect to depart from this 60/40 ratio, nor could they invest in individual corporations or in mutual funds that did anything other than track broad market indices. Each account manager would pool the accounts that firm would manage and invest them collectively. Pooled investment is the approach taken by the Thrift Savings Plan for federal employees.
The accounts would be charged directly for the administrative costs of operating this system, including the costs and profits of the account managers and the costs of the Social Security Guarantee Board that would oversee the system. Once the system of individual accounts had been in effect for five years, administrative fees would be limited to one-quarter of one percent of the account balances each year. There would be no limit on the administrative fees during the first five years.
Upon retiring or becoming disabled, a worker would receive five percent of his or her account balance in a lump-sum payment. The other 95 percent of the account balance would essentially be converted to an annuity — that is, a monthly payment to the beneficiary until the beneficiary died. The amount of the annuity would be based on average life expectancy and on the assumption that the account balance that would remain after each monthly payment was made would continue being invested 60 percent in stocks and 40 percent in bonds. For married workers, a joint annuity would be calculated. The annuity amounts would be adjusted annually for inflation.
Workers would not, however, actually receive these monthly annuity payments. Instead, the payments would be transferred directly to the Social Security Trust Fund. The Trust Fund would pay each beneficiary the monthly Social Security benefit amount that the beneficiary was slated to receive under the regular Social Security benefit structure. The monthly annuity payments would help the Trust Fund cover the costs of paying regular Social Security benefits, rather than accruing to the individual account-holders.
Most workers would receive their normal Social Security benefit amount, regardless of how well or poorly their account had done. The only circumstance in which the size of the annuity payment from a beneficiary’s account would affect the size of the monthly check the beneficiary received would be in those cases where the monthly annuity payment made on behalf of the beneficiary exceeded the regular Social Security benefit the beneficiary was slated to get. In such cases, the beneficiary would receive a monthly payment from the Social Security Trust Fund equal to the individual’s full monthly annuity amount, rather than his or her Social Security benefit amount. Based on the analysis of the plan conducted by the Social Security actuaries, however, such cases would be quite rare.
Thus, most workers would receive a monthly Social Security benefit exactly equal to their scheduled Social Security benefit, plus a lump-sum payment equal to five percent of the amount in their account at the time they began receiving Social Security disability or retirement benefits. In essence, beneficiaries would receive five percent of their accounts, with the other 95 percent going to help defray the costs of paying regular Social Security benefits.
The Shaw plan also has a few other features. As noted, the accounts of people who die before beginning to draw Social Security benefits would be added to their estates. The accounts of people who die after beginning to draw Social Security benefits would be transferred to the Trust Fund. In addition, the plan includes several benefit improvements geared toward women. These improvements do not affect the basic structure of the plan and are described in an appendix at the end of this paper.
The Financing of the Plan
An analysis of the plan issued by the Office of the Chief Actuary of the Social Security Administration shows that very large amounts of general revenue would be required to finance the individual accounts. Over 75 years, these transfers would amount to $4.5 trillion in net present value. This is $1.1 trillion more than would be needed to maintain Social Security solvency for the next 75 years without any changes in the Social Security system.
These transfers would not be paid back to the general fund at any point. To the contrary, the general fund transfers would continue to be made in perpetuity; the new deposits made each year into the individual accounts would continue being financed out of general revenues. These transfers ultimately would result in an infusion of funds into the Social Security Trust Fund, as the annuity payments from the individual accounts were deposited in the Trust Fund to help cover the ongoing costs of paying Social Security benefits. But the general fund would not be compensated and would be depleted on a permanent basis.
This raises the question of where the general fund would secure the resources to make these large transfer to the individual accounts. No such funds appear to be available; under likely budget policies (i.e., policies that reflect the President’s defense and homeland security increases, continuation of at least a substantial share of the tax cut, etc.), the non-Social Security budget is in deficit as far as the eye can see. Analyses by the Congressional Budget Office, the General Accounting Office, economists at the Brookings Institution, and others project that the deficits eventually will reach alarming dimensions as a result of the baby-boomers’ retirement, continuing increases in health care costs, and the tax cut enacted last year if it is made permanent. (See box below.)
As a result, policymakers would essentially face three choices in finding the trillions of dollars to transfer to the individual accounts: 1) impose deep cuts on the rest of the budget; 2) impose large tax increases; or 3) run even larger deficits outside Social Security than those currently projected, with such deficits likely to continue for decades. The Shaw plan gives no indication of which path should be taken, leaving it for future policymakers to grapple with how to finance the large transfers to the individual accounts.
To be sure, the plan would restore Social Security solvency both for the next 75 years and after that. At the end of the 75-year period, the Trust Fund would have substantial balances. The Trust Fund would reach solvency and build these sizeable balances because the Trust Fund would receive quite large amounts over time from the individual accounts.
The Trust Fund would particularly benefit from a form of “arbitrage.” As a result of the large transfers that would have to be made from the general fund to the individual accounts, the amounts that the Treasury would have to borrow from outside investors would be substantially greater than would otherwise be the case. The federal government would be borrowing at the Treasury bond rate to finance the transfers to the individual accounts. The funds deposited in the individual accounts would then be invested in stocks and corporate bonds that presumably would yield an average rate of return that would exceed the Treasury rate over time. The individual accounts thus would capture the “spread” between the Treasury rate and the stock-and-bond rate and would ultimately transfer most of these gains to the Social Security Trust Fund. This type of transaction — borrowing at one rate and using the borrowed funds to secure a higher rate of return — is a form of arbitrage.
The Magnitude of the General Revenue Transfers that Would Be Required
The Shaw plan would increase the long-term fiscal imbalance outside of Social Security by $4.5 trillion in net present value or by 0.9 percent of GDP. These technical figures may not be readily comprehensible to the public.
One way to make these figures more understandable is to note that the general fund transfers called for in the Shaw plan could be financed by an immediate, permanent tax increase that would initially equal $100 billion a year and grow in size at the same rate as the economy in subsequent years. (Such a tax increase would do nothing to close the existing long-term imbalance in the budget outside Social Security but would cover the cost of the general fund transfers and thereby prevent the long-term imbalance outside Social Security from getting worse than it already is.)
Alternatively, the cost of the general fund transfers under the Shaw plan could be covered by budget cuts initially totaling $100 billion a year, with the $100 billion level of these costs growing larger in subsequent years at the same rate as the economy. A budget cut of $100 billion a year would be greater than the combined budgets of the departments of education and veterans affairs. Other ways to cut the federal budget by $100 billion per year include, for example:
- cutting federal health care programs except Medicare in half — that is, a 50 percent reduction in Medicaid, the State Children’s Health Insurance Program, veterans health programs (including veterans’ hospitals), military health, Indian health, community health centers, maternal and child health grants, and health assistance to people with AIDS/HIV;
- cutting all federal low-income budget programs in half except the health programs — that is, a 50 percent reduction in the Supplemental Security Income program for the elderly and disabled poor, the food stamp program, the school lunch and other child nutrition programs, the Supplemental Nutrition Program for Women, Infants, and Children (WIC), Pell Grants (which help lower-income students afford to go to college), student loans, Title 1 education for the disadvantaged, housing vouchers for low-income families and individuals, homeless grants, housing for the elderly and handicapped, public housing, foster care, family support payments, child support enforcement, and various other programs;
- cutting the defense budget by one-quarter;
- eliminating all funding for veterans and military retirees, including all veterans pension, disability compensation, and health programs; or
- eliminating all funding for highways, railroads, mass transit, conservation, anti-pollution programs, national parks and other natural resources, the FBI, the DEA, the INS, and all law enforcement and the judiciary generally.
Securing resources for the Social Security Trust Fund through arbitrage may not be inappropriate if the scope of the arbitrage is limited. The scope of the market investments and arbitrage under the Shaw plan, however, is immense. By 2060, the personal accounts that the plan establishes would own about half of all holdings in the U.S. stock and bond markets. The dominance of these quasi-private, quasi-governmental accounts in the financial markets could raise troubling questions related to corporate governance and other matters.
Other Financing Issues
Two other financing issues warrant mention. The actuaries’ assessment of the plan is based upon an assumption that the individual accounts would earn an average “real” rate of return (i.e., the rate of return after inflation is taken into account) of 5.05 percent per year. The actual rate of return could, of course, be higher or lower. If it proved to be significantly lower, the amounts paid from the accounts to the Social Security Trust Fund would not be sufficient to restore 75-year solvency.
A second issue is whether the feature of the plan under which 95 percent of the account balances would essentially redound to the benefit of the Social Security Trust Fund, rather than to individual account-holders, would be politically sustainable over time. This might depend, in part, on how the accounts were explained and understood. If the public came to view the accounts as their personal accounts, then attacks on the transfers to the Trust Fund as representing a confiscatory 95 percent tax rate on the accounts could prove politically potent. The transfer to the Trust Fund of all balances remaining in an account when a beneficiary died could come under similar attack. If these features of the plan could not be preserved, Social Security solvency would be jeopardized.
The Long-term Fiscal Imbalance and the Shaw Plan
CBO, GAO, OMB, and other analysts periodically publish long-term projections of the federal budget. Using these data, the 75-year fiscal imbalance in the federal budget as a whole can be converted to a single figure, which CBO calls the “fiscal gap.”
Recent CBO long-term analyses suggest that the fiscal gap in the total federal budget is approximately $11 trillion to $13 trillion over 75 years in net present value, if the fiscal gap is calculated using the same general approach as the Social Security actuaries use in their analyses. The long-term imbalance in the budget as a whole thus is very large, with somewhat less than one-third of the imbalance occurring within Social Security and somewhat more than two-thirds occurring outside Social Security. The Shaw plan would eliminate the $3.4 trillion Social Security shortfall, but only by adding another $4.5 trillion to the long-term imbalance in the rest of the budget.
The Plan’s Perverse Incentive
Under the Shaw plan, there would be only a limited potential for account balances to be passed on to heirs. As noted, the amounts remaining in an account would be forgone at death unless an account-holder died before beginning to draw any Social Security benefits. In that case, the account would be bequeathed to heirs.
Under such a system, it is inevitable that those who qualify for disability or retirement benefits but do not expect to live a long time would weigh whether to begin drawing benefits at all. This could lead to undesirable effects.
Individuals who did not expect to live for more than a few years and could afford to do so could decide to forgo Social Security benefits and to use their Social Security Guarantee accounts to enlarge their estates. By contrast, retirees and disabled individuals with limited life expectancies but few other assets would not have this luxury; they generally would need Social Security benefits to support themselves for the period they remained alive.
Such approaches are inconsistent with the basic purpose of Social Security. Moreover, to the extent that such practices became quite widespread and the Trust Fund consequently received less in payment from the individual accounts than the actuaries projected, Social Security solvency could be compromised.
The Shaw plan represents an attempt to introduce personal accounts in a painless manner. It is presented as restoring solvency, protecting and even increasing benefits and assuring that no one is worse off, and avoiding any increase in payroll taxes. It is a quintessential “free lunch” approach to Social Security.
As with virtually everything else in life, however, there is no “free lunch.” The plan accomplishes this goal only through massive general revenue transfers, for which the plan provides no financing. The general revenue transfers under the plan are so great that they exceed by one-third the general revenue transfers that would be required if policymakers were to adopt the fiscally imprudent course of simply transferring enough money from the rest of the budget to pay full Social Security benefits for the next 75 years without making any changes in the Social Security system.
Provisions Targeted to Women
The Shaw plan includes several benefit enhancements that would primarily benefit women. These include the following changes.
- Increasing the elderly survivors benefit. Under current law, a married couple may receive 100 percent of the higher earner’s benefits plus a spouse benefit equal to 50 percent of the higher earner’s benefit. Alternatively, the spouse may receive benefits based on his or her own earnings history if this would result in benefits higher than the spouse benefit. When one spouse dies (typically the husband), the surviving spouse (typically the wife) receives 100 percent of her own benefit or 100 percent of the deceased spouse’s benefit, whichever is higher. This represents a decline in benefits of one-third to one-half compared to the couple’s combined benefit when both spouses were alive.
The Shaw plan would add an additional option: an elderly survivor could receive a benefit equaling 75 percent of the couple’s combined benefit, up to a maximum amount set at the benefit level that the average retiree receives. Capping the benefit that could be provided under this option at the average retiree’s benefit level holds down costs by targeting the benefit increase on surviving spouses who receive low or moderate benefits. Surviving spouses whose benefits under current law would be less than the cap (i.e., less than the average retiree’s benefit) would gain.
This is the most significant of the benefit improvements in the plan, although its cost is not large. Over 75 years, its cost equals slightly less than one-tenth of one percent of taxable payroll.
- Child-care credits. Workers who take time out of the labor force to care for a child (of their own or their spouse) under age seven would receive child care credits for up to five caregiving years. The amount of the credit would be based on the formula that is used for the calculation of Social Security disability benefits; the credit amount for a worker would equal 25 percent of the average earnings that would be used to calculate disability benefits for that worker if the worker had been disabled during the caregiving period. (If the worker had some earnings during the caregiving period but the earnings were less than the credit amount, she would receive the credit amount.) This is the second most significant of these benefit enhancements in terms of cost, but nonetheless is a modest change. Its cost equals six one-hundredths of one percent of taxable payroll over 75 years.
- Liberalizing rules for receipt of surviving spouse benefits by disabled individuals. The measure would eliminate the requirement that disabled widows or widowers must be at least 50 years old to receive these benefits. (These survivors would lose their eligibility for these benefits if they remarried before age 50.) In addition, eligibility for these benefits would no longer be limited to surviving spouses who became disabled within seven years of their spouse’s death (or within seven years of the date when benefits for a surviving spouse who is caring for young children became payable), as is required under present law. The seven-year requirement would be dropped. These improvements would affect a very small number of individuals; their cost would be negligible, totaling less than one one-hundredth of one percent of taxable payroll over 75 years.
- Relaxing requirements for divorced spouse benefits. Currently, individuals who become divorced must wait two years after their divorce to be eligible for spouse benefits if the former spouse has not begun to collect benefits. The intent of this requirement is to discourage couples from divorcing merely to gain access to spousal benefits that otherwise would be unavailable until the worker begins to collect benefits. Under the Shaw plan, if a worker remarries during that two-year period, the former spouse becomes eligible for spouse benefits at that time. In such as case, it is evident that the divorce was not obtained simply to gain access to spouse benefits. This improvement is even smaller, with a cost of less than five-one-thousandths of one percent of taxable payroll over 75 years.
- Reduction in the “Government Pension Offset.” This provision of current law affects individuals (most commonly women) who receive a spouse benefit or a surviving spouse benefit from Social Security and also receive a pension from employment not covered by Social Security (typically, from state or local government employment). Their Social Security benefit is reduced by two-thirds of their pension benefit. This two-thirds reduction, known as the “government pension offset,” is designed to avoid inequities and prevent double-dipping. The Shaw plan would limit the offset in Social Security benefits to one-third of the pension benefit, which would increase Social Security benefits for those affected by the government pension offset but result in greater inequities in benefits. The cost of this provision also is small — two one-hundredths of one percent of taxable payroll over 75 years.
 For a detailed examination of the Commission proposals, see Peter A. Diamond and Peter R. Orszag “Reducing Benefits and Subsidizing Individual Accounts: An Analysis of the Plans Proposed by the President’s Commission to Strengthen Social Security,” Center on Budget and Policy Priorities and The Century Foundation, June 2002.
 The other plan unveiled shortly after the Commission’s final action that is presented by its sponsors as restoring solvency and establishing personal accounts without any reduction in benefits or increase in payroll taxes is a proposal introduced by House Majority Leader Dick Armey and Rep. Jim DeMint.
 Diamond and Orszag, op. cit.
 Stephen C. Goss, Chief Actuary, Social Security Administration, and Alice H. Wade, Deputy Chief Actuary, “OASDI Financial Effects of the ‘Social Security Guarantee Plus Plan’ INFORMATION,” Memorandum to Representative Clay Shaw, Chairman Subcommittee on Social Security, December 13, 2001.
 The estimate that the long-term actuarial imbalance in the Social Security Trust Fund equals $3.4 trillion in net present value is the estimate that the Social Security actuaries issued in conjunction with the 2001 Social Security Trustees’ report. The 2001 Trustees’ report also stated that the Trust Fund was out of balance by 1.86 percent of taxable payroll over 75 years. The 1.86-percent-of-payroll figure equals the $3.4 trillion net-present-value imbalance divided by the net present value of taxable payroll over the next 75 years. In other words, the $3.4 trillion net present value figure and the 1.86 percent of payroll figure simply are two ways of presenting the same calculation of the long-term imbalance.
 In their analysis of the Shaw plan, the actuaries included year-by-year estimates for the next 75 years of the annual payments that the general fund would make to the personal accounts established under the plan. These projections are shown in constant 2001 dollars. Converting these general-fund payments to nominal dollars (using the actuaries’ inflation assumptions) and then to discounted dollars (using the actuaries’ compounded discount rates) shows that the net present value of the general fund payments over 75 years totals $4.5 trillion. That figure, in turn, represents 0.9 percent of long-term GDP. See table 1b of Stephen C. Goss and Alice H. Wade, op. cit.
Because the actuaries’ analysis of the Shaw plan is based on the financial and demographic data in the 2001 Social Security Trustees Report, all figures in this analysis are consistent with those data. In the 2002 report, the Social Security long-term shortfall was revised upward — from $3.4 trillion to $3.6 trillion in net present value. The actuaries’ estimate of the size of the general revenue transfers under the Shaw plan also would likely increase by a modest amount if it were made consistent with the data in the 2002 Trustees report.
 The lower threshold would be set at 21 percent of the maximum amount of wages subject to the payroll tax. In 2002, this would equal $17,829. For ease of exposition, we refer to this level in the text as “roughly $18,000.”
 The actuaries’ analysis of the Shaw plan finds that no one whose personal account earned the average rate of return that the actuaries assumed in their analysis would receive a monthly payment larger than his or her regular Social Security benefit. There would, of course, be some variability in the rates of return that the accounts earned. Even so, the actuaries estimate that the only individuals who would be likely to receive a monthly payment greater than their regular Social Security benefit would be single workers who both earned wages close to or above the maximum level subject to the Social Security payroll tax throughout their working lives and received investment returns on their accounts that exceeded the average rate of return that the actuaries assumed.
 Stephen C. Goss and Alice H. Wade, op. cit.
 This is true regardless of whether the non-Social Security budget otherwise would be in surplus or in deficit. If the non-Social Security budget otherwise would be in surplus, the surplus would be used to pay down the debt, reducing the amount that the government borrows from outside investors. If funds were transferred to personal accounts rather than being used to pay down debt, the debt would be larger than would otherwise be the case. Of course, this point is somewhat academic, since non-Social Security deficits are projected far into the future.
 The actuaries’ memorandum on the Shaw Plan includes projections of the total balances of the Social Security Guarantee Accounts in each year, in constant dollars. Converting these figures to shares of GDP, we see that the SSG account balances would reach 35 percent of GDP by 2028 and 70 percent of GDP by 2056 and keep rising slowly after that. Based on history, these figures would represent approximately one-quarter and one-half of the combined U.S. markets for stocks and corporate bonds, respectively. Specifically, over the period from 1990 through 2000, end-of-year market capitalization averaged 140 percent of GDP. (The markets also totaled 140 percent of GDP in 1996, after the recession of the early 1990s was over but before the boom of the late 1990s took off.) Thus, if the SSG accounts own corporate assets equal to 70 percent of GDP, as the actuaries project, they will probably own about half of the capital markets by the late 2050s.
 The 5.05 percent projected rate of return assumes a 6.5 percent average annual return on stocks, a 3.5 percent average annual return on bonds, and a 0.25 percent annual administrative fee.
 CBPP calculations from CBO data. See Congressional Budget Office, “A 125-Year Picture of the Federal Government’s Share of the Economy, 1950 to 2075,” June 14, 2002. Issued in June, this CBO analysis reflects the budget assumptions that CBO employed in the ten-year budget forecast it released earlier in the year. That forecast has since been superseded by the revised ten-year forecast that CBO issued in late August, which projects lower levels of revenue throughout the ten-year period than CBO’s earlier forecast did. The ten-year projections in CBO’s August forecast imply that even the large estimates of the long-term fiscal gap cited here may be understated by at least $2 trillion.
 In the absence of this offset, spouses who are not covered by Social Security because they are employed by a state or local government that has its own pension system — and does not participate in Social Security — could receive a Social Security spousal benefit as large as the benefit received by many spouses who worked throughout their lives in jobs covered by Social Security and who have no additional pension comparable to a state or local government pension. The government pension offset is designed to take into account the fact that spouses who are not covered by Social Security and pay no Social Security taxes because they participate in an alternative state or local government pension system receive retirement benefits from the alternative pension system that are supposed to substitute in part for Social Security benefits.