Testimony for the House Ways and Means Subcommittee
on Social Security
May 24, 2005
Stephen C. Goss, Chief Actuary
Social Security Administration
Chairman McCrery, ranking member Levin, and members of the subcommittee, thank you very much for the opportunity to talk with you today about the Social Security Trustees Reports and the future financial status of the Social Security program.
Annual Reports from the Board of Trustees to the Congress on the financial condition of the Old-Age, Survivors, and Disability Insurance program have been prepared continuously starting with 1941. These reports are required by law to include an assessment of the "actuarial status" of the trust funds. This assessment has been used by the Congress numerous times since 1941 as the basis for modifying the program to either alter the scope and nature of the program, or to improve the financial status of the program.
The Office of the Actuary at the Social Security Administration prepares the projections used in these reports as well as projections of the effects of proposals to change the program, and has done so continuously since the inception of the program in 1935. The Office has always operated on a non-partisan basis providing objective estimates to the Trustees, the Administration, and the Congress. As you know, while the Office of the Actuary resides within the Social Security Administration, it operates on an independent basis, particularly regarding work for the Congress, including this Subcommittee, the full Ways and Means Committee, and the Senate Finance Committee. Our work for the Congress is always done on a confidential basis during the development of a proposal for changing Social Security, and remains confidential unless and until the requestor is prepared to go public with the proposal. Our current Commissioner, Jo Anne Barnhart, as well as former Commissioner Kenneth Apfel have strongly supported the independence of the Office of the Actuary, well understanding the importance of this independence to the credibility of our work.
Today I would like to speak about three aspects of our analysis of the actuarial status of the Social Security program under current law for the Trustees Report. These are (1) the basic status of financing and solvency over the 75-year long-range period as reported in the 2005 Trustees Report and changes from the prior report, (2) the principal assumptions used in the projections and how they are driving the projected financial status, and (3) some of the possible legislative changes that are available to improve the actuarial status of the program.
(1) Where We Are-The Basic Actuarial Status of the Social Security Program
In the 2005 Trustees Report, the intermediate projections indicate that the annual excess of tax income over program cost will begin to decline in 2009, and in 2017 cost will exceed tax income. At that point the accumulated trust fund assets of about $2.4 trillion in present value will begin to be used to augment tax income so that benefits scheduled in current law will continue to be paid in full. These assets are, by law, invested wholly in securities backed by the full faith and credit of the United States Government, and have always been redeemed when needed. While there is no question that these securities will be redeemed when needed, this redemption will require the Federal Government to increase taxes, lower other expenditures, or issue publicly-held debt in amounts equal to the net redemptions by the trust funds.
If no changes are made, it is projected that the combined trust fund assets would become exhausted in 2041 and the program would no longer be considered to be solvent. This means that we would no longer be able to fully pay benefits scheduled in current law on a timely basis. Instead, we would be able to provide 74 percent of scheduled benefits with continuing tax revenues. After 2041, program cost is projected to continue growing faster than tax income. By 2079, 68 percent of scheduled benefits are expected to be payable if no changes are made.
Changes from the 2004 to the 2005 Trustees Reports were small. No changes in the principal economic or demographic ultimate assumptions were made. The estimate that the years of expected trust fund exhaustion and cost exceeding tax income will be one year sooner was largely the result of unexpectedly high growth in prices that was not matched by a similar unexpected increase in average earnings levels. This effect resulted in lower annual program cash-flow surpluses or higher deficits through 2024.
However, for years after 2024, other changes, principally in the methods we use for the projections resulted in somewhat lower annual cash-flow deficits for the program. For the year 2078, the estimated annual cash-flow deficit was reduced from 5.91 to 5.66 percent of taxable payroll. Thus, on balance, the long-range actuarial status of the Social Security program is essentially unchanged in the 2005 Trustees Report.
This change may also be seen in the estimates for the actuarial deficit and other measures of the unfunded obligation for the program over the 75-year long-range valuation period. The actuarial deficit expresses the magnitude of expected net future shortfalls over the entire period as a percentage of the taxable payroll over the entire period. Normally this actuarial deficit is expected to increase by 0.07 percent of payroll solely due to the shift in the 75-year period from one report to the next. In fact, the actuarial deficit increased from 1.89 to only 1.92 percent of payroll for the 2005 Report, consistent with the small reduction in projected cash-flow deficits for 2025 and later. The 75-year net shortfall may also be expressed as a percentage of the GDP over the entire period. For the 2005 Report, the net shortfall as a percent of GDP is projected at 0.6 percent of GDP over the period, again slightly lower than projected for the 2004 Report.
Finally, the shortfall can also be viewed in the form of an aggregate dollar amount in present discounted value to the beginning of the valuation period, or January 1, 2005. In this form, the net shortfall over the next 75 years is estimated at $4 trillion present value dollars. This amount is larger than the estimated unfunded obligation of $3.7 trillion present value dollars reported in the 2004 Trustees Report largely because the valuation date, that is the date to which net shortfalls are discounted, is one year later in time. By discounting the annual shortfalls for each future year to 2005 rather than to 2004, the present value amount is measured as 5 to 6 percent greater in the new report. What is critical to note in these measures of unfunded obligation is that they represent the net shortfall for the 75-year period as a whole, and thus must be met with changes that will be applied over the 75-year period as a whole. Expressing the unfunded obligation as a percent of taxable payroll or GDP better illustrates the magnitude of the changes that will be needed.
It should also be noted that in assessing the actuarial status of the Social Security program, more than just attaining solvency throughout the 75-year projection period is considered. When this first goal is met, an additional criterion for achieving "sustainable solvency" should be considered. This additional requirement asks that the level of the trust fund assets be projected to be stable or rising as a percent of annual program cost at the end of the period. When both 75-year solvency and this additional requirement are met, the program may be expected to continue to be solvent for the foreseeable future, under the assumptions used in the projection. Meeting these criteria further suggests that even if actual experience in the future varies from the assumptions to a degree, only small modifications are likely to be needed to maintain adequate financing for the program.
The Office of the Actuary has provided an assessment of the degree to which each comprehensive proposal achieves the criteria for sustainable solvency in scoring starting in the middle 1990s. Providing this additional assessment has helped to lead to the development of numerous proposals that meet the criteria for sustainable solvency under the Trustees intermediate assumptions over the last 10 years. The Trustees Reports have also referred to the criteria for sustainable solvency since 1999.
(2) The Principal Assumptions for the Trustees Report
Projections of future cost and income for Social Security are driven by a number of principal economic and demographic assumptions that are selected by the Trustees each year. The process for this selection each year starts with analysis and recommendations provided to the Trustees by the Office of the Actuary. This analysis and the recommendations are discussed extensively and final assumptions are adopted by the Trustees, generally very close to those recommended by the actuaries. In fact, this process has resulted in remarkably consistent assumptions over the years, and across Administrations. Changes in ultimate assumptions are made incrementally and only after evidence supporting change is fully discussed and analyzed. See Table C, attached . (To access documents with this logo, you will need the Adobe Acrobat reader software. If you do not already have this software, see our page on downloading and printing Portable Document Format () documents.)
Maintaining consistent and objective assumptions for the Trustees Report has been aided by two important requirements imposed by the Congress in the law. First, the law requires the inclusion of two Public Trustees on the Board, one effectively representing each major political party. Over the years, the Public Trustees have always worked together and have had a major and positive influence on the Board. The second requirement is that the Chief Actuary is required to provide a statement included in the report indicating whether the assumptions and methods used are individually and collectively reasonable. I am happy to report that my statement in the 2005 Report indicated that the assumptions and methods are reasonable.
The principal economic assumptions include real-wage and productivity growth assumptions, price growth, interest rate, and employment rate assumptions. Productivity growth provides the basis for average wage growth. The intermediate assumptions include an ultimate assumption of 1.6 percent average annual growth in total economy productivity, equal to the average growth rate over the last four complete economic cycles, from 1966 to 2000. The average annual real growth rate in the average wage was 1.15 percent over the same period, also very close to the ultimate real wage differential of 1.1 percent. The ultimate real interest assumption on long-term Treasury bonds is assumed to average 3 percent, or slightly below the average yield of 3.4 percent over the last four complete economic cycles.
However, due to the indexing and other features of Social Security, program cost is not greatly sensitive to variation in economic assumptions. The major effects on program cost relative to the base of taxable earnings are in the demographic changes that confront the program.
The principal demographic assumptions include birth rates, death rates, and immigration. The ultimate rate of decline in death rates for individuals over age 65 is about the same as for the average of the last century, and considerably faster than for the last 20 years. Immigration is assumed to be at roughly the average level over the last 20 years or so.
But birth rates are the principal reason that the cost of the Social Security program as a percentage of the taxable payroll will shift to a new higher level over the next 25 years. The "total fertility rate" or the average number of children women have was about 3.3 children per woman during the baby-boom years from 1946 through 1965. By 1972, however, the total fertility rate dropped to 2 children per woman and has stayed at about that level ever since. The ultimate assumption is for an average total fertility rate of 1.95 for the future.
This shift in the total fertility rate is directly responsible for the shift in the ratio of workers to beneficiaries that is projected to occur between about 2010 and 2030.
This ratio has remained constant at about 3.3 workers per beneficiary since 1975, when the Social Security program matured in its coverage of the population. Had the total fertility rate stayed at 3 or higher, the current 12.4 percent payroll tax rate would be adequate to finance currently scheduled benefits and we would not be discussing future shortfalls. But due to the shift in birth rates over 30 years ago, we will see the ratio of workers to beneficiaries drop to 2.2 by 2030 and 2.0 by 2040. It is this shift that makes the current law 12.4 percent tax rate insufficient to fully finance the currently scheduled benefits in the long run.
Directly reflecting the decline in the projected worker to beneficiary ratio between 2010 and 2030, is the increase in the Social Security cost rate, as a percent of taxable payroll, over the same period. During this period the cost rate is projected to shift from a level that is now well below the current tax rate of 12.4 percent to a level that is well above it.
Continuing but much more gradual decreases in the worker to beneficiary ratio and increases in the cost rate are projected after 2030 based on expected future increases in life expectancy. But these are modest in comparison with the shift in the cost rate that will result from the decline in birth rates after 1965.
(3) Possible Legislative Changes to Improve the Actuarial Status of Social Security
The projected net shortfall in financing for Social Security over the next 75 years could be met by an average reduction in benefits of 13 percent or an average increase in tax revenue of 15 percent over the period. But the timing of the expected shortfalls is important. Most proposals being considered would confirm the pay-as-you-go nature of the financing of Social Security by targeting changes to years after trust fund exhaustion in amounts roughly equal to the projected annual shortfalls.
As mentioned earlier, the annual cash-flow shortfall for the year 2079 is projected to be about 5.7 percent of taxable payroll. To meet this annual shortfall in 2079 would require benefits that were then nearly one third lower than are currently scheduled, or revenue that is nearly 50 percent higher than currently scheduled, or some combination of the two. Greatly reducing or eliminating this annual shortfall for 2079 will be necessary if sustainable solvency is to be achieved.
Several changes to lower scheduled benefits, by slowing the projected growth, have been considered. These include changes in the normal retirement age, and modifications of the basic benefit formula. Benefit formula changes include general "price indexing" of benefits across future generations, and "progressive indexing" which would provide for larger percentage reductions for higher earners, thus making the current benefit formula more progressive.
Potential changes to increase revenue for Social Security in the future include increasing the taxable maximum amount, increasing taxation of benefits, and increasing payroll tax rates. Additional revenue could also be generated by modifying the pay-as-you go nature of Social Security financing to include more substantial advance funding.
Many combinations of the provisions mentioned above, as well as a large number of other possible provisions could restore long-range solvency, and sustainable solvency for the Social Security program. Changes will be needed well before the expected date of trust fund exhaustion in 2041. By enacting needed changes sooner, we will have more options to consider, be able to phase changes in more gradually, and give affected individuals more advance notice.
The Trustees Reports required by law have played a fundamental role in informing the Congress and the Administration of the actuarial status of the program, and the magnitude of changes that may be needed for the future. The Office of the Actuary has been and will continue to be available to the Congress and the Administration for objective and non-partisan estimates both of the current status of the program, but also for possible changes to Social Security.
Thank you very much for the opportunity to present these remarks. I look forward to trying to answer any questions that you may have.