Social Security's Office of the Chief Actuary provides an overview on the current and projected financial condition of the Disability Insurance program.
The Social Security Disability Insurance (DI) program is financed through one of the two trust funds established for Social Security. The other fund is for the Old-Age and Survivor's Insurance (OASI) program. While these trust funds are technically separate, the financial outlook of the programs has typically been viewed on a combined basis for two reasons. First, the benefits provided under the two programs are closely interrelated. Second, historically, when only one of the two programs has approached financial problems in the near term, the Congress has tended to borrow from or reallocate taxes from the other program to avoid an imminent crisis, thus more equalizing the financial outlook for the two programs. For these reasons, this paper will consider the OASI financial outlook as well.
The financial outlook for the DI and OASI programs, both separate and combined, are described in detail in the annual report of the Board of Trustees to the Congress. This report provides specifically the financial outlook for the programs assuming that no changes are made to the Social Security Act, which schedules future benefit and tax levels. The report thus provides the Congress an indication of the expected extent to which currently scheduled taxes are more or less than needed to finance the scheduled benefits. It also indicates when changes may be needed to assure adequate financing for the program. The trustees report has been produced every year, starting in 1941. Copies of all reports, through the latest report issued on May 1, 2006, can be obtained at http://www.socialsecurity.gov/OACT/pubs.html.
The latest trustees report (2006) indicates that the DI Trust Fund is projected under the trustees intermediate assumptions to remain solvent until 2025, during which year the trust fund is projected to become exhausted. If these assumptions are realized, and there is no Congressional action between now and 2025, monthly benefits scheduled in the law would no longer be payable in full on a timely basis after the exhaustion of the trust fund assets. However, even in this case, continuing tax income for the program would be sufficient to pay 82 percent of scheduled benefits for the remainder of 2025. Due to projected further increases in the cost of the program above the expected tax income, the percentage of scheduled benefits that would be payable would decline very gradually, reaching 75 percent for 2080.
The 2006 trustees report indicates somewhat better financial status for the separate OASI program, with solvency through 2041 and trust fund exhaustion in the following year under the intermediate assumptions. On a combined basis, the OASI and DI Trust Funds would remain solvent through 2039, with assets exhausted in 2040. At the time of trust fund exhaustion, 74 percent of scheduled OASDI benefits would be payable. With no change in the law, 70 percent of currently scheduled OASDI benefits would be payable in 2080.
Whether the financial outlook of the DI program is considered in isolation or on a combined basis with the larger OASI program, currently scheduled tax income is expected to be insufficient to provide for timely payment of full benefits throughout the long-range 75-year projection period. Thus, it will be necessary for changes to be made in either the taxes or benefits scheduled in current law, for both the DI and OASI programs. The balance of this paper describes a little about the history and nature of the financial status for the Social Security programs, and the potential approaches that are available for strengthening and improving their financial outlook.
The Social Security program was established in law in 1935, payroll tax collections started in 1937, and monthly retirement benefits started in 1940. Based on legislation in 1956, monthly benefits from the DI Trust Fund started in 1957. For both programs, the number of beneficiaries grew gradually as workers attained insured status, became more aware of the available benefits, and adjusted their expectations. By about 1975, benefit enrollments reached maturity for both programs, in the sense that individuals of all ages then had the full opportunity to become insured. This maturity is illustrated by the ratio of OASDI covered workers to beneficiaries (Chart 1). This ratio was initially quite high when essentially all current workers were covered, but only those with sufficient recent earnings could receive benefits. By 1975, this ratio had declined to 3.2 workers per beneficiary, and it has remained close to that level ever since.
Beginning in 2008, the "baby-boom generation" (those born from 1946 through 1965) will be starting to receive Social Security retirement benefits. The combination of their leaving the workforce and becoming beneficiaries coincides with the substantial reduction in the level of the worker-to-beneficiary ratio, from about 3.3 currently to 2.2 by 2030. In fact, the principle reason for the decline in this ratio is not the retirement of the "boomers" per se, but the relatively low birth rates that these generations have had and subsequent generations are expected to have. While the total fertility rate averaged 3.3 children per woman between 1946 and 1965, it declined to 2.0 by 1972 and has remained around that level since. This relatively low birth rate is expected to continue into the future, as indicated by the intermediate assumption of the 2006 Trustees Report. The persistent lower birth rate means that not only the ratio of workers to beneficiaries will decline, but that this ratio will remain at a new lower level. Smaller additional declines in the ratio continue after 2030 as the life expectancy is projected to continue rising. But this is only a gradual and relatively small effect compared to the large and permanent shift due to the decline in birth rates.
Another way to look at these projections is to reverse the ratio, resulting in the number of beneficiaries per 100 workers. Chart 2 illustrates this ratio separately for the DI and OASI programs as well as for the combined program. It is evident from Chart 2 that the cost of the DI program is less than for the OASI program and that growth in DI cost will also be less in the future. The principal reason why the growth in beneficiaries per 100 workers is so much less for the DI program after 2026 is that the highest age for disabled worker benefits will be fixed at 66 thereafter. Therefore, increases in longevity after 2026 will have a significant effect on the number of aged OASI beneficiaries and much less effect on the number of DI beneficiaries.
However, the cost of the DI program is still projected to increase significantly as a percent of taxable payroll in the future, because the prevalence of disability in the population has been and is projected to continue increasing. Chart 3 illustrates the prevalence of disability for men and women combined since 1970 and through the long-range projection period of the 2006 Trustees Report. The gross disability prevalence rate is the number of individuals receiving a disabled worker benefit per 1,000 individuals who meet the disability insured status requirements under the program.1 The age-sex-adjusted prevalence rate provides a measure of the prevalence rate without the distracting effects of a changing age distribution of the population and thus provides the best basis for understanding the changing prevalence of disability.
Clearly the prevalence of disability has risen significantly since 1970. By 2005 the age-sex-adjusted prevalence rate nearly doubled from the level for 1970. This was in part due to the maturation of the program, as benefit payments under the program were not available until 1957. The dip in the prevalence rate between about 1980 and 1985 was largely the result of implementation of periodic continuing disability reviews (CDRs) required by the 1980 Amendments to the law. These reviews resulted in the termination of benefits for many individuals, and coincided with reductions in new applications for benefits. After 1985, the prevalence rate began rising again due to changes in the law mandating (1) that termination of benefits due to medical condition would require documentation of improvement in the condition and (2) revisions in the mental impairment listings. The prevalence rate reached a level near 40 disabled workers per 1,000 insured in 2000, and continued rising through 2005.
Disability prevalence is projected to continue rising in the future, but at a much slower pace than in the recent past. The further increases are the result of three factors discussed below: (1) the maturing of increases in disability incidence rates that occurred over the past decade, (2) some projected further increases in disability incidence rates, and (3) projected greater longevity of those becoming disabled in the future.
Disability prevalence rates at any age reflect, among other factors, the disability incidence rates for the population over the past 30 years or so. The level of incidence rates is most important for the prior 10 to 15 years because most disability entitlements start after age 50. Chart 4, from the 2006 Trustees Report, presents the gross and age-sex adjusted disabled worker incidence rates from 1970 to 2015. Incidence rates grew significantly from 1988 to 2005. Future prevalence rates will gradually begin to reflect the recent higher levels of incidence rates over the next 30 years. This maturing of the recent levels of incidence rates will thus tend to increase prevalence rates in the subsequent years.
The second factor affecting projected prevalence rates is the assumed increase in ultimate disabled worker incidence rates above the recent levels. While the incidence rates are projected to decline slightly from 2005 through 2011 as the special disability workload is completed, they are assumed to increase thereafter reaching ultimate levels around 2025 that are significantly higher than the incidence rates for 2005. This is due in part to the improvements in medical interventions which are assumed to improve the survival rates of individuals who suffer both chronic and acute medical conditions that are disabling. While medicine can help maintain health, it also allows those who have suffered impairments to cope with any limitations more effectively, particularly in extending life. Small but meaningful improvements in medical interventions can have significant effects on the survival of individuals who have become disabled increasing their chances to live long enough to qualify for benefits.
The third factor affecting future disabled worker prevalence rates is the projected improvements in death rates after disabled worker benefits have been awarded. The projections in the Trustees Report reflect the assumption that the rate of improvement for disabled workers will be about the same as for the general population. Thus, if the death rate for women declines by 10 percent over some period in the future, the same 10 percent decline is assumed for disabled worker women. Because death rates are much higher for disabled workers than for the general population, however, this results in greater absolute reductions in death rates for the disabled. And these greater absolute reductions result in relative increases in disability prevalence as a percent of the population as the duration of disability increases.
Thus, while the highest age for receipt of disabled worker benefits will stay at 66 after 2026 under current law, projected continued increases in disability prevalence rates will result in some future increase in the cost of the DI program as a percent of taxable payroll. But as shown in Chart 5, the increase in cost rates is projected to be far less for the DI program than for the OASI program. However, the cost of Social Security is generally considered as a whole, and thus the financial status of the OASI and DI programs are generally considered on a combined basis.
"Solvency" of the OASI and DI programs is defined for any point in time as the ability of the trust funds to pay for all scheduled benefits in full and on a timely basis. Solvency is achieved as long as a positive trust fund balance is maintained, that is as long as there are assets in the funds. But while the DI Trust Fund is projected to become exhausted in 2025 under current law, this does not mean that disability benefits would cease at that time. If nothing is done, then it is projected that somewhat less than full scheduled benefits would still be payable starting in 2025.
However, because the OASI Trust Fund is projected to still have substantial assets in 2025, a reallocation of a portion of the OASI payroll tax rate to the DI program would be expected from the Congress, even if no other action had been taken by 2025. With this, the combined OASI and DI programs would then be expected to exhaust their trust funds in 2040. The program would be able to pay about 74 percent of the currently scheduled benefits after exhaustion in 2040 and 70 percent in 2080 (Chart 6).
Because no one would want a sudden 26 percent reduction in all benefits in 2040, we as a nation will need to make changes to the current program before that time. That choice among possible changes will require us to either reduce the level of scheduled benefits or increase the scheduled revenue (potentially in part through higher returns on assets). The magnitude of potential changes range from a reduction in benefits over the next 75 years as a whole of about 13 percent, to an increase in payroll tax revenue over the period of about 15 percent. Of course, these changes do not need to be made uniformly over the 75-year period or across the OASI and DI programs. If benefit reductions or tax-rate increases are phased in gradually so that current beneficiaries receive little or no reduction, or current workers see little or no increase in taxes, then later beneficiaries or workers would need to receive larger changes.
Change is inevitable and choices will need to be made. We should not expect the Social Security program developed to this point to continue to serve best the needs of future populations. The exact extent of future change needed is uncertain. The projections provided here are based on the intermediate "best estimate" assumptions from the 2006 Trustees Report. While changes can be developed, considered, and enacted into law today or tomorrow, we should expect each future generation to make its mark in altering the program in ways that will best serve the needs of the day. However, we cannot ignore the financial shortfalls currently projected. These are based largely on populations already born and are large enough to be extremely likely to occur at some level. Therefore, serious consideration of changes and enactment of changes well before trust fund exhaustion is upon us will be wise. Early enactment of changes will allow for consideration of more options, more advance notice, and a more gradual phase in of the changes. In addition, should changes based on current projections prove to be more than is needed in the future, then the program benefits and revenue can and will be altered again. More frequent program changes in recognition of changing circumstances also allow for smaller and less drastic change.
Improving the financial outlook, or actuarial status of these programs is understood to require some combination of benefit reductions and revenue increases relative to what is scheduled in current law. Most recent comprehensive proposals that rely at least in part on reductions in OASDI benefits to restore long-range solvency for Social Security have specified reduced benefits principally for retirees and other aged beneficiaries. These proposals work on the principle that workers who become disabled, by definition, have little if any choice about working and must apply for the disabled worker benefit. In addition, younger workers who become disabled have had limited time to develop alternative sources of income. Most retirees, on the other hand, have a greater capacity to continue working and to have saved throughout their working years. Thus, a reduction in the monthly benefit offered for retiring at age 62 can be effectively undone by working another year or two to receive a higher benefit with less reduction for early retirement. Under this approach, disabled workers would receive the same DI benefits as in current law, but they might see reductions in their benefits at the time they convert from DI payments as a disabled worker to OASI payments as a retired worker, spouse, or widow.
However, some proposals have extended reductions in the general benefit formula to retired and disabled worker benefits alike. A somewhat smaller percentage benefit reduction suffices to restore long-term solvency for Social Security if the reduction is applied to the disabled as well as retirees and survivors. Regardless of whether DI benefits are reduced directly under a plan, the financial status of the OASDI program as a whole must be restored. Thus, a plan that reduces only OASI benefit payments may need to include a schedule or mechanism for reallocating the combined payroll tax rate between OASI and DI.
While most changes intended to address solvency issues focus primarily on the OASI program, many changes in the DI program are also considered to maintain its currency. Several years ago the substantial gainful activity dollar amount was modified to increase on an automatic basis, to keep pace with growth in the average wage. Changes in the definition of disability and/or the requirements used to determine whether applicants meet the current definition may also be warranted to reflect changing circumstances. For example, the currently proposed regulation to increase the age at which certain vocational criteria may be considered in determining disability has been put forth on the basis that people are, and will be, generally living longer and healthier and so will be able to work somewhat longer. Adoption of this change would not only adapt the program rules to changing circumstances, it would also avoid some of the tendency for the cost of the DI program to grow, in the absence of this change.
Another approach to modify the DI program is to create an environment that is more conducive for disabled worker beneficiaries to return to work, and thus eventually terminate their Social Security DI benefits. The "ticket-to-work" provision is a good example. In general, however, provisions intended to promote return to work for disabled workers are somewhat limited in achieving substantial results by the severity of the disabling condition required to qualify for benefits under the current program. Many beneficiaries have little or no prospect for recovery from their disabling condition. For many others, improvement can take years and very substantial and expensive medical and rehabilitative services.
With the projected year of 2040 in which benefits under the combined OASDI program would first not be payable in full on a timely basis under current law, it may be difficult to motivate the Congress and the American people to make substantial changes to the OASDI program in the near future. But in any case, it is clear that the actuarial shortfall now projected for the DI and OASI programs have a number of potential remedies. The overall program is sustainable largely in its current form, with some changes in both benefit levels and revenue available to the program. Or the program could be changed more fundamentally. Such choices will be made by the American people through their elected representatives in the Congress and the White House. And the program will undoubtedly be modified periodically in the future to reflect the desires of the citizens at that time.
