Trustees Report- 1996

F. ACTUARIAL ESTIMATES

The required information for the fiscal year that ended September 30, 1995, is presented in section II.C of this report. Estimates of the operations and status of the trust funds during fiscal years 1996-2005 are presented in this section. In addition, similar estimates for calendar years 1996-2005 are presented. A description of the actuarial status of the trust funds over the next 75 years, including long-range estimates of program income and program costs over that period, is also included in this section. The methods used to estimate the short-range operations of the trust funds and the long-range actuarial status are described in section II.H.

A number of different measures are useful in evaluating the financial status of the trust funds over the next 75 years. In addition to actuarial balance, and summarized income and cost rates, which are described in detail below, these measures include (1) the levels of future annual income and outgo, both in terms of dollars and relative to annual taxable earnings or payroll, including the pattern and ultimate values of such levels; (2) the annual differences between income and outgo, i.e., the annual balances, in dollars and relative to taxable payroll; (3) the size of future fund accumulations, in dollars and relative to future annual expenditures; and (4) the year in which trust fund exhaustion is estimated to occur. Estimates of all these indicators are presented in this section or in the appendices of this report. However, more attention is focused on certain elements of these measures, as described below.

In the short range, the adequacy of the trust fund level is generally measured by the "trust fund ratio," which is defined to be the assets at the beginning of the year expressed as a percentage of the outgo during the year. (For the years 1984-90, the assets at the beginning of the year also included advance tax transfers for the month of January. Assets at the beginning of subsequent years include advance tax transfers only if such transfers are needed to enable the timely payment of benefits.) The trust fund ratio represents the proportion of a year's outgo which can be paid with the funds available at the beginning of the year. During periods when trust fund disbursements exceed income, as might happen during an economic recession, trust fund assets are used to meet the shortfall. In the event of recurring shortfalls for an extended period, the trust funds can allow sufficient time for the development, enactment, and implementation of legislation to restore financial stability to the program.

The test of financial adequacy over the short-range projection period (the next 10 years), is applicable to each of the OASI and DI Trust Funds, separately, as well as to the combined funds. The requirements of this test are as follows: If the estimated trust fund ratio for a fund is at least 100 percent at the beginning of the projection period, then it must be projected to remain at or above 100 percent throughout the 10-year projection period. Alternatively, if the ratio is initially less than 100 percent, then it must be projected to reach a level of at least 100 percent by the beginning of the sixth year and to remain at or above 100 percent throughout the remainder of the 10-year period. In addition, the fund's estimated assets at the beginning of each month of the 10-year period must be sufficient to cover that month's disbursements. This test is applied on the basis of the intermediate (alternative II) estimates. Failure to meet this test by either trust fund is an indication that solvency of the program over the next 10 years is in question and that Congressional action is needed to improve the short-range financial adequacy of the program.

Basic to the discussion of the long-range actuarial status are the concepts of "income rate" and "cost rate," each of which is expressed as a percentage of taxable payroll. The annual income rate is the ratio of income from revenues (payroll tax contributions and income from the taxation of benefits) to the OASDI taxable payroll for the year. The OASDI taxable payroll consists of the total earnings which are subject to OASDI taxes, with some relatively small adjustments. Because the taxable payroll reflects these adjustments, the annual income rate can be defined to be the sum of the OASDI combined employee-employer contribution rate (or the payroll-tax rate) scheduled in the law and the rate of income from taxation of benefits (which is, in turn, expressed as a percentage of taxable payroll). As such, it excludes reimbursements from the general fund of the Treasury for the costs associated with special monthly payments to certain uninsured persons who attained age 72 before 1968 and who have fewer than 3 quarters of coverage, transfers under the interfund borrowing provisions, and net investment income.

The annual cost rate is the ratio of the cost (or outgo, expenditures, or disbursements) of the program to the taxable payroll for the year. In this context, the outgo is defined to include benefit payments, special monthly payments to certain uninsured persons who have 3 or more quarters of coverage (and whose payments are therefore not reimbursable from the general fund of the Treasury), administrative expenses, net transfers from the trust funds to the Railroad Retirement program under the financial-interchange provisions, and payments for vocational rehabilitation services for disabled beneficiaries; it excludes special monthly payments to certain uninsured persons whose payments are reimbursable from the general fund of the Treasury (as described above), and transfers under the interfund borrowing provisions. For any year, the income rate minus the cost rate is referred to as the "balance" for the year. (In this context, the term "balance" does not represent the assets of the trust funds, which are sometimes referred to as the "balance" in the trust funds.)

The long-range actuarial status of the trust funds has generally been summarized by the calculation of the "actuarial balance." The actuarial balance for a specified valuation period is defined as the difference between the summarized income rate and the summarized cost rate over that period. The summarized income rate over a period of years is equal to the ratio of (a) the sum of the trust fund balance at the beginning of the period plus the present value of the total income (excluding interest earnings) during the period, to (b) the present value of the taxable payroll for the years in the period. The summarized cost rate is equal to the ratio of (a) the sum of the present value of the outgo during the period plus the present value of a targeted trust fund level at the end of the period equal to the following year's outgo to (b) the present value of the taxable payroll for the years in the period. A targeted ending trust fund level of 1 year's expenditures is considered to be an adequate reserve for unforeseen contingencies; thus, in addition to the total outgo during the projection period, the summarized cost rate includes the cost of reaching and maintaining a target trust fund ratio of 100 percent through the end of the projection period.

The present-value calculations take account of the effect of interest on future income and outgo. In calculating the present value of future income, for example, the income in each year of the projection period is discounted to the beginning of the period using the interest rate assumed for calculating the interest earnings of the trust funds during the period. Thus, the calculations of the summarized income and cost rates are consistent with the estimates of trust fund operations over the projection period.

If the program is in exact actuarial balance for a particular period (that is, if the actuarial balance is zero), then the present value of estimated future income for all years in the period, plus the beginning trust fund balance, is exactly equal to the present value of estimated future expenditures for all years in the period, plus the present value of targeted trust fund assets at the end of the period in the amount of the next year's estimated outgo. A negative actuarial balance indicates that future estimated income and the beginning trust fund balance together are not sufficient to accumulate to the level of the targeted assets while also covering all estimated expenditures in the period. A positive actuarial balance indicates that in addition to covering all estimated expenditures in the period, the estimated ending trust fund assets are more than the targeted level.

The size of the actuarial balance represents a measure of the program's financial adequacy for the period in question. The actuarial balance can be interpreted as that amount which, if added to the combined employee-employer contribution rate scheduled under present law for each of the next 75 years, would bring the program into exact actuarial balance. Of course, there are any number of different ways to increase taxes or to reduce expenditures, as well as different combinations of such changes, that would have an equivalent effect on the actuarial balance. Any one of these different sets of changes would, therefore, bring the program into exact actuarial balance.

The long-range test of close actuarial balance applies to a set of valuation periods beginning with the first 10 years and continuing through the first 11 years, the first 12 years, etc., up to and including the full 75-year projection period. Under the long-range test, summarized income rates and cost rates are calculated for each of the 66 valuation periods in the full 75-year long-range projection period, with the first of these periods consisting of the next 10 years. Each succeeding period becomes longer by 1 year, culminating with the period consisting of the next 75 years. The long-range test is met if, for each of the 66 time periods, the actuarial balance is not less than zero or is negative by, at most, a specified percentage of the summarized cost rate for the same time period. The percentage allowed for a negative actuarial balance is 5 percent for the full 75-year period. For shorter periods, the allowable percentage begins with zero for the first 10 years and increases uniformly for longer periods, until it reaches the maximum percentage of 5 percent allowed for the 75-year period. The criterion for meeting the test is less stringent for the longer periods in recognition of the greater uncertainty associated with estimates for more distant years.

When a negative actuarial balance in excess of the allowable percentage of the summarized cost rate is projected for one or more of the 66 separate valuation periods, the program fails the long-range test of close actuarial balance. Being out of close actuarial balance indicates that the program is expected to experience financial problems in the future and that ways of improving the financial status of the program should be considered. The sooner the actuarial balance is less than the minimum allowable balance, expressed as a percentage of the summarized cost rate, the more urgent is the need for corrective action. However, it is recognized that necessary changes in program financing or benefit provisions should not be put off until the last possible moment if future beneficiaries and workers are to be able to effectively plan for their retirement.

It was noted earlier in this section that in addition to the measures used in the tests of the overall financial condition of the program, other financial measures are also presented in this report. All of these measures are important factors in arriving at a full understanding of the financial position of the OASDI program.

1. Operations and Status of the Trust Funds During the Period October 1, 1995, to December 31, 2005

These estimates indicate that the assets of the OASI Trust Fund would continue to increase throughout the next 10 years, rapidly under the intermediate and low cost assumptions and moderately under the high cost assumptions. The estimates indicate that the assets of the DI Trust Fund would also continue to increase throughout the next 10 years under the intermediate and low cost assumptions, at a slightly lower rate than for the OASI Trust Fund. Under the high cost assumptions, DI assets would increase for a few years before declining and becoming insufficient to permit the timely payment of benefits by the middle of 2005.

As will be shown later in this subsection, the OASI and DI Trust Funds, both individually and combined, meet the requirements of the Trustees' test of short-range financial adequacy.

a. OASI Trust Fund Operations

The increases in estimated income shown in table II.F1 under each set of assumptions reflect increases in estimated taxable earnings and growth in interest earnings on the invested assets of the trust fund. For each alternative, employment and earnings are assumed to increase in every year through the year 2005 (with the exception that employment is estimated to decline temporarily during the economic recessions assumed under alternative III). The number of persons with taxable earnings would increase on the basis of alternatives I, II, and III from 141 million during calendar year 1995 to about 157 million, 152 million, and 148 million, respectively, in 2005. The total annual amount of taxable earnings is projected to increase from $2,925 billion in 1995 to $5,056 billion, $4,793 billion, and $4,726 billion, in 2005, on the basis of alternatives I, II, and III, respectively. (In 1995 dollars - taking account of assumed increases in the CPI from 1995 to 2005 under each alternative - the estimated amounts of taxable earnings in 2005 are $3,784 billion, $3,368 billion, and $3,004 billion, respectively.) These increases in taxable earnings are due primarily to (1) projected increases in employment levels and average earnings in covered employment, (2) increases in the contribution and benefit base in 1996-2005 under the automatic adjustment provisions, and (3) various provisions enacted in 1983 and later, including extensions of coverage to additional categories of workers.

Growth in interest earnings represents a significant component of the overall increase in trust fund income during this period. Although interest rates payable on trust fund investments are not assumed to change substantially from current levels, the continuing rapid increase in OASI assets will result in a corresponding increase in interest income. By the year 2005, interest income to the OASI Trust Fund is projected to range from 11 to 15 percent of total trust fund income (depending on alternative), as compared to 10 percent in 1995.

Rising expenditures during 1996-2005 reflect automatic benefit increases as well as the upward trend in the numbers of beneficiaries and in the average monthly earnings underlying benefits payable by the program. The growth in the number of beneficiaries in the past and the expected growth in the future result both from the increase in the aged population and from the increase in the proportion of the population which is eligible for benefits. The latter increase is primarily due to various amendments enacted after 1950 which modified eligibility provisions and extended coverage to additional categories of employment.

Growth has also occurred, and will continue to occur, in the proportion of eligible persons who, in fact, receive benefits. This growth is due to several factors, among which are (1) the amendments enacted since 1950 which affect the conditions governing the receipt of benefits and (2) the increasing percentage of eligible persons who are aged 70 and over and who therefore may receive benefits regardless of earnings.

The estimates shown in table II.F1 indicate that income to the OASI Trust Fund would substantially exceed expenditures in every year of the short-range projection period, under each of the three sets of assumptions used in this report. The assets of the OASI Trust Fund at the beginning of 1995 were equal to 139 percent of the fund's expenditures in 1995. As described in the introduction to this section, this ratio is known as the "trust fund ratio;" it provides a useful measure of the relative level of trust fund assets. During 1995, income exceeded disbursements by $45.0 billion. As a result, the trust fund ratio increased to about 148 percent at the beginning of 1996.

Assets are estimated to increase substantially in each year of the short-range projection period, based on each of the three alternative sets of assumptions. The increase in the trust fund ratio from 148 percent at the beginning of 1996 to the range of 166-317 percent at the beginning of the year 2005 is due, in part, to the increases in the OASI tax rate that became effective in 1988 and 1990 (even though much of the increase was reallocated to the DI Trust Fund in 1994). Asset growth is also assisted by growth in taxable earnings that is projected to exceed the growth in benefit payments throughout the short-range projection period (except for certain years under alternative III).

As noted in section II.B, the portion of the OASI Trust Fund that is not needed to meet day-to-day expenditures is used to purchase investments, generally in special public-debt obligations of the U.S. Government. The cash used to make these purchases becomes part of the general fund of the Treasury and is used to meet various Federal outlays. Interest is paid to the trust fund on these securities and, when the securities mature or are redeemed prior to maturity, general fund revenues are used to repay the principal to the trust fund. Thus, the investment operations of the trust fund result in various cash flows between the trust fund and the general fund of the Treasury.

Currently, the excess of tax income to the OASI Trust Fund over the fund's expenditures results in a substantial net cash flow from the trust fund to the general fund. Sometime after the turn of the century, as shown in the following subsection, this cash flow will reverse; as trust fund securities are redeemed to meet benefit payments and other expenditures, revenue from the general fund of the Treasury will be drawn upon to provide the necessary cash. The accumulation and subsequent redemption of substantial trust fund assets has important public policy and economic implications that extend well beyond the operation of the OASDI program itself. Discussion of these broader issues is not within the scope of this report.

Based on the intermediate (alternative II) assumptions, the assets of the OASI Trust Fund would continue to exceed 100 percent of annual expenditures by a steadily increasing amount through the end of the year 2005. Consequently, the OASI Trust Fund satisfies the test of short-range financial adequacy by a wide margin. The estimates in table II.F1 also indicate that the short-range test would be satisfied even under the high cost assumptions.

In interpreting the trust fund ratios in table II.F1, it should be noted that at the beginning of any month there must be sufficient assets on hand to meet the benefit payments that are payable at the beginning of that month. The specific minimum amount of assets required for this purpose depends on a number of factors and varies somewhat from month to month. Assets of roughly 8 to 9 percent of annual expenditures are normally sufficient for this purpose. If the assets of either the OASI or DI Trust Fund at the end of a month fall below the minimum amount needed to meet the benefits payable at the beginning of the next month, section 201(a) of the Social Security Act provides for an advance transfer to the trust fund of all the taxes that are expected to be received by the fund in the next month. Thus, the difference between (1) the sum of the estimated trust fund ratios shown in table II.F1 and the advance tax transfers for January expressed as a percentage of total expenditures in the year and (2) the minimum required level of about 8-9 percent, represents the reserve available to handle adverse contingencies.

b. DI Trust Fund Operations

Expenditures are estimated to increase because of automatic benefit increases and projected increases in the amounts of average monthly earnings on which benefits are based. In addition, on the basis of all three sets of assumptions, the number of DI beneficiaries is projected to continue increasing throughout the short-range projection period. The projected growth in the number of DI beneficiaries is attributable to several factors, including (1) gradual increases in the number of persons estimated to be insured for disability benefits and (2) an assumption that the number of insured workers who apply for and are awarded disability benefits will continue to substantially exceed the number of disabled worker beneficiaries whose benefits terminate each year as a result of death, recovery, or attainment of normal retirement age.

The proportion of insured workers who apply for and are awarded disability benefits in a given year is referred to as the "disability incidence rate." This rate has fluctuated substantially in past years and the causes for the variation have not been precisely determined. Incidence rates increased during 1970-75, declined during 1976-82, increased again during 1983-85, and remained steady in 1986-89. During 1990-92 the incidence rate resumed increasing, with unusually rapid increases (on a relative basis) of 8, 12, and 17 percent in those 3 years. In 1993-95, the observed incidence rate declined slightly from the 1992 level. There remains, however, a backlog of pending disability applications awaiting final adjudication that is relatively large compared to historical levels. This suggests that the recent declines in the incidence rate may, in part, represent a delay in awards from 1993-95 to later years.

The rapid increases in disability benefit applications and awards during 1990-92 appear to be attributable, in part, to the rise in unemployment associated with the 1990-91 economic recession (although the evidence is not conclusive). Other explanatory factors may include changes to the conditions governing receipt of disability benefits, as introduced through recent legislation, regulations, and court decisions, and increased awareness of the DI program by the public.

These and other factors were discussed at some length in a report issued December 1992, entitled "The Social Security Disability Insurance Program: An Analysis" prepared by the Department of Health and Human Services at the request of the Board of Trustees. Subsequent to that report, the Social Security Administration, together with the Office of the Assistant Secretary for Planning and Evaluation in the Department of Health and Human Services, commissioned a series of studies attempting to quantify some of the reasons for the rapid growth in the DI program on the early 1990s. Reference should be made to these studies for further details on the possible factors contributing to the increases in disability incidence rates observed in the period 1990-92, and the subsequent decline observed since 1992.

Due to the substantial variation exhibited by incidence rates in the past and the difficulty in determining reliable explanatory factors for this variation, any projection of future incidence rates necessarily will be uncertain. The 1995 disability incidence rate (calculated on an age-sex-adjusted basis) was 5.25 awards per 1,000 insured workers. This figure was about 15 percent higher than the average incidence rate of 4.6 per thousand that was experienced during 1975 through 1995. Under the intermediate assumptions, incidence rates are assumed to increase by another 1 percent in 1996-97 and then to decline gradually for the remainder of the short-range projection period, to about 3 percent below the level experienced in 1995. A small portion of this decline is attributable to the prohibition on future awards of benefits to persons disabled by drug addiction or alcoholism enacted in Public Law 104-121. Under the low cost alternative, incidence rates decline by about 18 percent during 1996-2005, dipping slightly under the 1975-95 average at the end of the period. The high cost alternative assumes that incidence rates increase by another 18 percent over the next 7 years (nearing briefly the highest levels experienced during the 1970s) and then decline slightly over the remaining 3 years of the short-range period.

The proportion of DI beneficiaries whose benefits terminate in a given year has also fluctuated significantly in the past. Over the last 20 years, the rates of benefit termination due to death or conversion to retirement benefits (at attainment of normal retirement age) have declined very gradually. This trend is attributable, in part, to the lower average age of new beneficiaries. The termination rate due to recovery has been much more volatile. Currently, the proportion of disabled beneficiaries whose benefits cease because of their recovery from disability is very low in comparison to past levels.

In this report, termination rates due to attainment of normal retirement age are estimated to continue their downward trend through 2002. This rate would drop in 2003 and remain at a depressed level for 5 more years as a result of the increase in the normal retirement age) which begins in that year. Age-specific death rates for disabled beneficiaries are assumed to remain at about their current level. Terminations due to recovery are assumed to increase from their current levels in response to the additional funding for continuing disability reviews authorized under Public Law 104-121. In addition, the prohibition placed by Public Law 104-121 on benefits payable to individuals disabled by drug addiction and alcoholism, is expected to result in a one-time increase in terminations during 1997. Ignoring this one-time effect, the overall termination rate (reflecting all causes) is projected under all three alternatives to continue declining gradually during 1996-98 and level off during 1999-2002. The overall rate then declines in 2003 due largely to the increase in the normal retirement age cited above.

The continuing spread of Acquired Immunodeficiency Syndrome (AIDS) has contributed to the recent increases in DI awards. Due to the extremely high mortality rates of affected individuals, the total number of disabled workers currently receiving benefits has not increased greatly as a result of AIDS. Although many aspects of the AIDS epidemic are well understood, there remains considerable uncertainty regarding future medical advances and future incidence of HIV infection. To reflect this uncertainty, the projected numbers of benefit awards to AIDS patients are varied by alternative. Under the intermediate assumptions, benefit awards to persons with AIDS are projected to increase slightly through 2000 before beginning to decline. Under the low cost assumptions, the number of new awards declines gradually throughout the projection period, while the number projected under the high cost assumptions increases at a rapid rate through 2001 before beginning to decline.

At the beginning of calendar year 1995, the assets of the DI Trust Fund represented 55 percent of annual expenditures. During 1995, DI income exceeded DI expenditures by $14.6 billion, with the result that the trust fund ratio for the beginning of 1996 increased to about 83 percent. Under the intermediate and low cost sets of assumptions, income is estimated to exceed expenditures in each year of the short-range projection period. The increase in the trust fund ratio from 55 percent at the beginning of 1995 to 83 percent at the beginning of 1996, and the further increase to 136 percent at the beginning of 2002 on the basis of the intermediate assumptions, are largely due to the tax rate reallocation enacted in 1994. The decline in the trust fund ratio to 127 percent at the beginning of 2005 is an early warning of trouble for the DI Trust Fund soon after the short-range period.

Under the low cost assumptions, the trust fund ratio would increase rapidly to 264 percent at the beginning of 2005. Under the high cost assumptions, the assets of the DI Trust Fund would increase through 1999, decline steadily thereafter, and would be exhausted in 2005.

Because DI assets reach the level of 1 year's expenditures at the beginning of 1997 under the intermediate assumptions and would remain above that level in 1998 and later, the DI Trust Fund satisfies the Trustees' short-range test of financial adequacy. However, as indicated above, under the high cost assumptions not only does DI fail to meet the short-range test of financial adequacy, but the DI Trust Fund is exhausted near the end of the short-range projection period.

c. Combined OASI and DI Trust Fund Operations

At the beginning of 1995, the trust fund ratio for the OASI and DI Trust Funds combined was 128 percent, as shown in table II.F3. During 1995, total income to the two trust funds was $59.7 billion higher than total expenditures. As a result of this increase, combined OASDI assets at the beginning of 1996 represented about 140 percent of estimated combined expenditures for the year. Based on the intermediate assumptions, the trust fund ratio for the combined funds is projected to increase substantially, to 221 percent by 2005. The ratio would grow at an even faster rate under the low cost assumptions, reaching 309 percent at the beginning of the year 2005. Under the high cost assumptions, assets would grow more slowly, reach a maximum of 159 percent in 1999-2000, and decline to 139 percent at the beginning of 2005.

Under the intermediate assumptions, the total assets of the OASI and DI Trust Funds would remain above 100 percent of annual OASDI expenditures throughout the short-range projection period. Therefore, the combined trust funds meet the requirements of the short-range test of financial adequacy. Under the high cost assumptions, the fund ratio for OASI and DI combined would still remain above 100 percent through 2005 (although, as indicated in the section on long-range projections, the ratio would fall below this level shortly thereafter). Thus, even under adverse conditions the combined funds would satisfy the short-range test of financial adequacy, although by a narrower margin.

Section 215(i) of the Social Security Act includes a provision to stabilize automatic benefit increases in the event of high inflation at a time when the combined assets of the OASI and DI Trust Funds are at very low levels (see section II.E of this report). Under all three alternatives, the level of OASDI assets during 1996-2005 would substantially exceed the applicable threshold. Thus, the stabilizer provision would not be triggered during the short-range projection period under any of the sets of assumptions used in this report.

Figure II.F1 presents the estimated total assets of the OASI and DI Trust Funds at the end of each year 1996-2005, based on the three sets of assumptions (together with actual assets at the end of each year 1985-95). Figure II.F2 illustrates the pattern of actual past and estimated future OASDI trust fund ratios under the three alternatives. Trust fund ratios for selected years prior to 1996, and estimates for 1996-2005 under the three alternatives, are shown in table II.F4 for OASI, DI, and both funds combined. In evaluating the ratios shown in figure II.F2 and table II.F4, it should be recalled that a minimum of roughly 8 to 9 percent is generally needed to meet monthly cash-flow requirements. The shaded area in figure II.F2 depicts this requirement.

The factors underlying the changes in the intermediate estimates for the OASI Trust Fund, from last year's annual report to this year's, are analyzed in table II.F5. In the 1995 Annual Report, the trust fund ratio for OASI was estimated to reach 246 percent at the beginning of the year 2004 - the tenth projection year from that report. The corresponding ratio shown in this report for the tenth projection year (2005) is 239 percent. As indicated in table II.F5, the net effect of the provisions in Public Law 104-121 affecting OASI (liberalization of the retirement earnings test for persons aged 65-69 and restrictions on benefits payable to stepchildren) was to reduce the tenth-year ratio by 2 percentage points. If there had been no changes to the projections other than those necessitated by this legislation, then the estimated ratio at the beginning of 2005 would have been 12 percentage points higher than at the beginning of 2004. There were changes, however, to reflect the latest actual data, as well as adjustments to the assumptions for future years. The cumulative net effects of changes in economic assumptions (including re-estimates of future tax revenue consistent with recent revisions to historical data) resulted in a decrease in the trust fund ratio of 13 percentage points by the beginning of 2005. Finally, the tenth year trust ratio was reduced an additional 4 percentage points due to the net effect of revised assumptions regarding future average benefit levels and projected numbers of survivor beneficiaries.

Corresponding estimates of the factors underlying the changes in the financial projections for the DI Trust Fund, and for the OASI and DI Trust Funds combined, are also shown in table II.F5. As was the case for OASI, the key factors affecting the new estimates for the DI Trust Fund were legislation, and the cumulative effects of changes in assumptions related to economic performance and terminations of disability benefits.

For the DI Trust Fund during 1996-2005, the estimated operations in this report under all three alternatives show a slight worsening since the 1995 Annual Report, primarily due to the downward revisions in projections of tax revenue. As for benefit payments from the DI Trust Fund, the number of new disability awards to insured workers in 1995 was less than anticipated in last year's report, but (as noted earlier) the backlog of pending disability claims continued to remain at a high level. The assumed disability incidence rates for the 1996 Annual Report are similar to the corresponding rates from the 1995 report, with a slight reduction in the number of awards in the next few years.

The overall disability termination rate experienced in 1995 was only slightly higher than assumed under the intermediate assumptions of the 1995 Annual Report (9.6 percent versus 9.5 percent). Consequently, the termination rate assumptions for this report were not changed significantly as compared to the 1995 Annual Report, except for the effects of Public Law 104-121 on continuing disability reviews and benefits payable to individuals disabled by drug addiction and alcoholism.

Table II.F6 shows that total expenditures in calendar year 1995 from the OASI and DI Trust Funds increased to 11.64 percent of taxable payroll for the year - 0.95 percentage point less than the income rate of 12.59 percent. This increase in the total cost rate for OASDI is primarily attributable to the re-estimate of the OASDI taxable payroll, as described previously. Under the intermediate assumptions, the OASDI cost rate would increase gradually during the short-range projection period, to 12.07 percent in 2005. Based on the low cost assumptions, the cost rate is estimated to decline steadily, reaching 10.58 percent in 2005. The high cost alternative indicates a significant increase, to 13.72 percent in 2005.

These cost rate projections are shown in table II.F6 for both trust funds, separately and combined. Table II.F6 also shows a comparison of the cost rates with the corresponding income rates. As explained previously, the income rate represents the sum of the combined employee-employer payroll tax rate and the income derived from the Federal income taxation of OASDI benefits, expressed as a percentage of taxable payroll. The difference between the income rate and the cost rate for a year is referred to as the "balance" for that year.

Estimates of the operations of the trust funds during calendar years 1996-2005 have been presented in the preceding tables on the basis of three different sets of economic assumptions, because of the uncertainty of future economic and demographic developments. Under the provisions of the Social Security Act, estimates of the expected operations and status of the trust funds during the next 5 fiscal years are required to be shown in this report. Accordingly, detailed estimates of the expected operations and status of the trust funds during fiscal years 1996-2000 are shown in the remaining tables of this section for the intermediate assumptions (alternative II) only. Similar detailed estimates are also shown for 5 additional fiscal years (2001-05) and on a calendar-year basis for 1996-2005.

Data on the actual operations of the OASI Trust Fund for selected years during 1940-95, and estimates of the expected operations of the trust fund during 1996-2005 on the basis of the intermediate assumptions, are shown in tables II.F7 and II.F8 on a fiscal- and calendar-year basis, respectively. Corresponding figures on the operations of the DI Trust Fund are shown in tables II.F9 and II.F10. Operations of both trust funds combined are shown in tables II.F11 and II.F12. (Data relating to the operations of the two trust funds for years not shown in tables II.F7 - II.F12 are contained in past annual reports.) The figures shown in tables II.F8, II.F10, and II.F12 for 1987, 1988, 1992, and 1993 are adjusted to reflect 12 months of benefit payments in each year. The amounts estimated for 1998 and 1999 are similarly adjusted.

2. Long-Range Actuarial Status of the Trust Funds

Beginning with the 1991 Annual Report, actuarial balances have also been computed based on the intermediate (alternative II) assumptions for valuation periods that are 10 years, 11 years, and continuing through 75 years in length. This series of actuarial balances provides the basis for the test of long-range close actuarial balance, described earlier in this section.

In addition to these actuarial balances, other indicators of the financial condition of the program are shown in this report. One is the series of projected annual balances (that is, the differences between the projected annual income rates and annual cost rates), with particular attention being paid to the level of the annual balances at the end of the long-range period and the time at which the annual balances may change from positive to negative values. Another is the series of projected trust fund ratios, with particular attention being paid to the amount and year of maximum fund ratio accumulation and to the year of exhaustion of the funds. These additional indicators are defined in the introduction to this section.

The estimates are sensitive to changes in the underlying economic and demographic assumptions. The degree of sensitivity, however, varies considerably among the various assumptions. For example, variations in assumed fertility rates have little effect on the estimates for the early years, because almost all of the covered workers and beneficiaries projected for the early years were born prior to the start of the projection period. However, lower fertility rates have large impacts on the actuarial balance in the later years. Variations in economic factors, such as interest rates and increases in wages and prices, have significant effects on the estimates for the short term, as well as for the long term. In general, the degree of confidence that can be placed in the assumptions and estimates is greater for the earlier years than for the later years. Nonetheless, even for the earlier years, the estimates are only an indication of the expected trend and general range of future program experience. Section II.G contains a more detailed discussion of the effects on the estimates of varying certain economic and demographic assumptions.

a. Annual Income Rates, Cost Rates, and Balances

The projections for OASDI under the intermediate alternative II assumptions show income rates that increase slowly and steadily due to the combination of the flat payroll tax rate and the gradually increasing effect of the taxation of benefits. The pattern followed by the cost rates is much different. Costs as a percent of taxable payroll are projected to rise slowly for the next 15 years and then to increase rather rapidly for about the next 20 years (through 2030) as the baby-boom generation reaches retirement age. Cost rates continue rising slowly through 2036 and then decline slightly for the next 6 years as the baby-boom generation ages and the relatively small birth cohorts of the late 1970s reach retirement age. Thereafter, cost rates rise steadily, but slowly, reflecting projected increases in life expectancy. The cost rates during the third 25-year subperiod rise to a level of nearly 19 percent of taxable payroll under the intermediate alternative II assumptions. The income rate during the third 25-year subperiod is just over 13 percent of taxable payroll under alternative II.

Projected income rates under the low cost and high cost sets of assumptions (alternatives I and III, respectively) are very similar to those projected for alternative II as they are largely a reflection of the tax rates specified in the law. OASDI combined cost rates for alternatives I and III differ significantly in size from those projected for alternative II, but follow generally similar patterns. For the low cost alternative I, cost rates decline somewhat for about the first 12 years, and then rise, reaching the current level around 2015 and a peak of 14.26 percent of payroll in 2032. Thereafter, cost rates decline gradually, reaching a level of 13.12 percent of payroll in 2070. For the high cost alternative III, cost rates rise throughout the 75-year period, but at a relatively faster pace during the next 5 years due to the assumed economic recessions, and between 2010 and 2030 because of the aging of the baby-boom generation. During the third 25-year subperiod, the projected cost rate continues rising and reaches 28.02 percent of payroll in 2070.

The projected pattern of the OASDI annual balances (that is, the difference between the income rates and the cost rates) is important in the analysis of the financial condition of the program. Under the alternative II assumptions the annual balances are positive for 16 years (through 2011) and are negative thereafter. This annual deficit rises rapidly reaching 2 percent of taxable payroll by 2020 and continues rising thereafter, to a level of 5.51 percent of taxable payroll for 2070.

Under alternative I, projected OASDI annual balances are positive for 25 years (through 2020), and thereafter are negative. Deficits under alternative I rise to a peak of 1.27 percent of taxable payroll in 2032, but diminish thereafter, as the effect of the baby-boom generation diminishes and the assumed higher fertility rates increase the work force. Deficits under alternative I diminish to 0.08 percent of payroll by 2070. Under the more pessimistic alternative III, however, the OASDI actuarial balance is projected to be positive for only 4 years (through 1999) and to be negative thereafter, reaching deficits of 4 percent of payroll by 2020, nearly 10 percent by 2050, and over 14 percent of payroll by 2070.

Also of interest are the long-range financial conditions of the separate OASI and DI programs. Annual balances under alternative II remain positive through 2013 for the OASI program, but only through 2002 for the DI program.

Figure II.F3 shows in graphical form the patterns of the OASDI annual income rates and cost rates. The income rates are shown only for alternative II in order to simplify the graphical presentation and because, as shown in table II.F13, the variation in the income rates by alternative is very small. The OASDI long-range summarized income rates for alternatives I and III, for the 75-year valuation period, differ by only about 0.3 percent of taxable payroll. By 2070, the annual income rates under alternatives I and III differ by less than 0.8 percent of taxable payroll. Only small fluctuations are projected in the income rate, as the rate of income from taxation of benefits varies only slightly, for each alternative, reflecting changes in the cost rate and the fact that benefit-taxation threshold amounts are not indexed.

The patterns of the annual balances are indicated in figure II.F3. For each alternative, the magnitude of each of the positive balances in the early years, as a percent of taxable payroll, is represented by the distance between the appropriate cost-rate curve and the income-rate curve above it. The magnitude of each of the deficits in subsequent years is represented by the distance between the appropriate cost-rate curve and the income-rate curve below it.

In the future, the cost of the OASDI program, as a percent of taxable payroll, will not necessarily be within the range encompassed by alternatives I and III. Nonetheless, because alternatives I and III define a reasonably wide range of economic and demographic conditions, the resulting estimates delineate a reasonable range for future program costs.

b. Summarized Income Rates, Cost Rates, and Balances

Table II.F14 shows rates on a present-value basis summarized for each of the 25-year subperiods, excluding both the funds on hand at the beginning of the period and the cost of accumulating a target trust fund balance by the end of the period. These rates are useful for comparing the cash flows of tax income and expenditures, as an indicator of the degree to which tax income during the period is sufficient to meet the outgo estimated for the period.

Table II.F15 shows summarized rates including the funds on hand at the start of the period and the cost of accumulating a target trust fund balance equal to 100 percent of annual expenditures by the end of the period, for valuation periods of the first 25 years, the first 50 years, and the entire 75-year period. Therefore, the actuarial balance for each of these three valuation periods is equal to the difference between the summarized income rate and cost rate for the corresponding period. A balance of zero for any period on this basis would indicate that estimated outgo for the period could be met, on the average, with a remaining trust fund balance at the end of the period equal to 100 percent of the following year's outgo.

The values in table II.F15 show that the combined OASDI program is expected to operate with a positive balance over shorter valuation periods under alternatives I and II. For the first 25-year valuation period the summarized values indicate balances of 1.75 percent of taxable payroll under alternative I, 0.36 percent under alternative II, and -1.18 percent under alternative III. Thus, the program is more than adequately financed for the next 25-year valuation period under all but the high cost alternative III projections. Over the 50-year valuation period, 1996-2045, the OASDI program would have a positive balance of 0.65 percent under alternative I, but would have deficits of 1.39 percent under alternative II and 3.82 percent under alternative III. Thus, the program is more than adequately financed for the 50-year valuation period under only the low cost set of assumptions, alternative I.

For the entire 75-year valuation period, the combined OASDI program would again have actuarial deficits except for the low cost set of assumptions, alternative I. The actuarial balance for this long-range valuation period is projected to be 0.46 percent of taxable payroll under alternative I, -2.19 percent under alternative II, and -5.67 percent under alternative III.

As may be concluded from tables II.F14 and II.F15, the financial condition of the DI program is somewhat poorer than that of the OASI program for the first 25 years. Summarized over the full 75-year period, however, long-range deficits for the OASI and DI programs under intermediate assumptions are about the same relative to program costs.

c. Test of Long-Range Close Actuarial Balance

Table II.F16 presents a comparison of the estimated actuarial balances with the minimum allowable balance (or maximum allowable deficit) under the long-range test, each expressed as a percentage of the summarized cost rate, based on the intermediate alternative II estimates. Values are shown for only 14 of the valuation periods: those of length 10 years, 15 years, and continuing in 5-year increments through 75 years. However, each of the 66 periods - those of length 10 years, 11 years, and continuing in 1-year increments through 75 years - is considered for the test. These minimum allowable balances are calculated to show the limit for each valuation period resulting from the graduated tolerance scale. The patterns in the estimated balances as a percentage of the summarized cost rates, as well as that for the minimum allowable balance, are presented graphically in figure II.F4, for the OASI, DI and combined OASDI programs. Values shown for the 25-year, 50-year, and 75-year valuation periods correspond to those presented in table II.F15.

As discussed earlier, a program is found not to be in long-range close actuarial balance if, for any of the valuation periods ending with the 10th through 75th years of the projection period, the estimated actuarial balance is less than the minimum allowable balance. The minimum allowable balance as a percentage of the summarized cost rate is -5.0 percent for the full 75-year long-range period and is reduced uniformly for shorter valuation periods, reaching zero for the 10-year valuation period.

For the OASI program, the estimated actuarial balance as a percentage of the summarized cost rate exceeds the minimum allowable for valuation periods of length 10 years through 33 years, under the intermediate alternative II estimates. For valuation periods of length greater than 33 years, the estimated actuarial balance is less than the minimum allowable. For the full 75-year long-range period the estimated actuarial balance reaches -13.90 percent of the summarized cost rate, for a shortfall of nearly 9 percent, from the minimum allowable balance of -5.0 percent of the summarized cost rate. Thus, although the OASI program satisfies the short-range test of financial adequacy (as discussed earlier in this section), it is not in long-range close actuarial balance.

For the DI program, the estimated actuarial balance as a percentage of the summarized cost rate exceeds the minimum allowable balance for valuation periods of length 10 through 12 years under the intermediate alternative II estimates. For valuation periods of length greater than 12 years, the estimated actuarial balance is less than the minimum allowable. The shortfall from the minimum allowable balance rises to a level of 15.54 percent of the summarized cost rate for the full long-range period, for a shortfall of over 10 percent, from the minimum allowable balance of -5.0 percent of the summarized cost rate. Thus, although the DI program satisfies the short-range test of financial adequacy (as discussed earlier in this section), it is also not in long-range close actuarial balance.

As indicated above, financing for the DI program is less adequate than for the OASI program during the first 25 years even though long-range actuarial deficits are comparable over the entire 75-year period. This occurs because the cost of the OASI program rises much faster during the long-range period. As a result, tax rates that are relatively more adequate for the OASI program during the first 25 years, become relatively less adequate thereafter.

For the combined OASDI program, the estimated actuarial balance as a percentage of the summarized cost rate exceeds the minimum allowable balance for valuation periods of length 10 years through 31 years. For valuation periods of length greater than 31 years, the estimated actuarial balance is below the minimum allowable balance. The size of the shortfall from the minimum allowable balance rises gradually reaching 9.13 percent of the summarized cost rate for the full 75-year long-range valuation period. Thus, although the OASDI program satisfies the short-range test of financial adequacy (as discussed earlier in this section), it is out of long-range close actuarial balance.

The OASI and DI programs, both separate and combined, were also found to be out of close actuarial balance in last year's report. The estimated deficits for the OASI, DI, and combined OASDI programs in this report are similar to those shown in last year's report.

d. Income and Cost Rates by Component

Summarized values for the annual income and cost rates, along with their components, are presented in table II.F18 for 25-year, 50-year, and 75-year valuation periods. Summarized income rates include the starting trust fund balance in addition to the components included in the annual income rates. The summarized cost rates include the cost of reaching and maintaining an ending trust fund target of 100 percent of annual expenditures by the end of the period in addition to the expenditures included in the annual cost rates. Thus, the total summarized rates shown in table II.F18 are the same as the summarized income and cost rates shown in table II.F15 for the 25-year, 50-year, and 75-year valuation periods.

It may be noted that the payroll tax income expressed as a percentage of taxable payroll is slightly smaller than the actual tax rates in effect for each period. This results from the fact that all OASDI income and outgo amounts presented in this report are computed on a cash basis, i.e., amounts are attributed to the year in which they are actually received by, or expended from, the fund, while taxable payroll is allocated to the year in which earnings are paid. Because earnings are paid to workers before the corresponding payroll taxes are credited to the funds, payroll tax income for a particular year reflects a combination of the taxable payrolls from that year and from prior years, when payroll was smaller. Dividing payroll tax income by taxable payroll for a particular year, or period of years, will thus generally result in an income rate that is slightly less than the applicable tax rate for the period.

e. Comparison of Workers to Beneficiaries

Table II.F19 shows that the number of covered workers per beneficiary, which was about 3.3 in 1995, is estimated to decline in the future. Based on alternative I, for which high fertility rates and small reductions in death rates are assumed, the ratio declines to a level of 2.3 by 2030, and increases slowly thereafter. Based on alternative III, for which low fertility rates and substantial reductions in death rates are assumed, the decline is much greater, reaching 1.3 workers per beneficiary by 2065. Based on alternative II, the ratio declines to 1.8 workers per beneficiary by 2065.

The impact of the demographic shifts under the three alternatives on the OASDI cost rates is better understood by considering the projected number of beneficiaries per 100 workers. As compared to the 1995 level of 31 beneficiaries per 100 covered workers, this ratio is estimated to rise by the year 2070 to significantly higher levels, which are 41 under alternative I, 56 under alternative II, and 78 under alternative III. The significance of these numbers can be seen by comparing figure II.F3 to figure II.F5.

For each alternative, the shape of the curve in figure II.F5, which shows beneficiaries per 100 covered workers, is strikingly similar to that of the corresponding cost-rate curve in figure II.F3, thereby emphasizing the extent to which the cost of the OASDI program is determined by the age patterns of the population. Because the cost rate is basically the product of the number of beneficiaries and their average benefit, divided by the product of the number of covered workers and their average taxable earnings (and because average benefits rise at about the same rate as average earnings), it is to be expected that the pattern of the annual cost rates is similar to that of the annual ratios of beneficiaries to workers. A graphical presentation of covered workers per beneficiary is shown in section I.H of the Overview.

f. Trust Fund Ratios

Based on alternative II, the OASI trust fund ratio rises steadily from 148 percent at the beginning of 1996, reaching a peak of 284 percent at the beginning of 2012. This increase in the OASI trust fund ratio results from the fact that the annual income rate (excluding interest) exceeds annual outgo for several years (see table II.F13). Thereafter, the OASI ratio declines steadily, with the OASI Trust Fund becoming exhausted in 2031. The DI trust fund ratio follows a similar pattern, except that it unfolds more rapidly. The DI trust fund ratio is estimated to rise from 83 percent at the beginning of 1995 to a peak of 136 percent in 2002, and to decline thereafter until becoming exhausted in 2015.

The trust fund ratio for the hypothetical combined OASI and DI Trust Funds rises from 140 percent for 1996 to a peak of 245 percent at the beginning of 2011. Thereafter, the ratio declines, with the combined funds becoming exhausted in 2029. Based on the intermediate estimates in last year's report, the peak fund ratio for the combined funds was estimated to be 269 percent and the year of exhaustion was estimated to be 2030.

The trust fund ratio for the combined OASDI program begins to decline in 2012, the same year annual expenditures begin to exceed noninterest income. Although the dollar amount of assets will continue to rise through the beginning of 2018, because interest income more than offsets the shortfall in noninterest income, revenue from the general fund of the Treasury will be needed in increasingly large amounts, beginning in 2012, to redeem the trust funds' public-debt obligations due to the cash-flow shortfall. This will differ from the experience of recent years when the trust funds have been net lenders to the general fund. The change in the cash flow between the trust funds and the general fund is expected to have important public policy and economic implications that go well beyond the operation of the OASDI program itself. Discussion of these issues is outside the scope of this report.

Based on the low cost alternative I assumptions, the trust fund ratio for the DI program increases throughout the long-range projection period, reaching an extremely high level by 2070, of 1,390 percent. For the OASI program, the trust fund ratio rises to a peak of 487 percent in 2017, dropping thereafter to a stable level around 340 percent by 2045. For the combined OASDI program, trust fund ratios follow a pattern similar to that for OASI, peaking at 479 percent in 2018, and then falling, until around 2040, but increasing thereafter, to a level of 471 percent for 2070.

In contrast, under alternative III, the OASI trust fund ratio is estimated to peak at 172 percent in 2001, thereafter declining to fund exhaustion by the end of 2020. The DI Trust Fund is estimated to begin declining in 2000, becoming depleted in 2005. The combined trust fund ratio is estimated to rise to a peak of 159 percent in 2000, declining thereafter to fund exhaustion by the end of 2016.

The fact that the financing for the DI program is relatively more adequate compared to the financing for the OASI program under low cost assumptions, but relatively less adequate under high cost assumptions is due to the tax rate reallocation enacted in 1994. This reallocation roughly equalized the size of the long-range actuarial deficits of the OASI and DI programs in relation to the summarized cost rates under intermediate assumptions. A smaller reallocation would have been needed to equalize the deficits in this manner under low cost alternative I assumptions, while a larger reallocation would have been needed under high cost alternative III assumptions.

Thus, because of the high ultimate cost rates that are projected under all but the most optimistic assumptions, income will eventually need to be increased and/or program costs will need to be reduced in order to prevent the trust funds from becoming exhausted.

Even under the high cost assumptions, however, the combined OASI and DI funds on hand plus their estimated future income would be able to cover their combined expenditures for 20 years into the future (until 2016). Under the alternative II assumptions the combined starting funds plus estimated future income would be able to cover expenditures for about 33 years into the future (until 2029). The program would be able to cover expenditures for the indefinite future under the more optimistic assumptions in alternative I. In the 1995 report, the combined trust funds were projected to be exhausted in 2016 under alternative III and in 2030 under alternative II.

A graphic illustration of the trust fund ratios for the combined trust funds is shown in figure II.F6 for each of the alternative sets of assumptions.

g. Reasons for Change in Actuarial Balance From Last Report

Legislation enacted since last year's report (see section II.A) is estimated to increase the long-range OASDI actuarial balance. The Senior Citizens' Right to Work Act of 1996 (Title I of Public Law 104-121) made several changes affecting OASDI benefits, including an increase in the retirement earnings test exempt amount for persons between their normal retirement age and age 70. Three provisions of this Act resulted in significant changes in the OASDI actuarial balance. Disability benefits will no longer be payable to persons found unable to work solely because of drug addiction or alcohol dependency. Additional funding will be provided through 2002 for periodic continuing disability reviews of disabled beneficiaries who were earlier found to have impairments that are not permanent. Benefits for a stepchild will be terminated if the stepparent, upon whose earnings record benefits are based, divorces the natural parent.

In changing from the valuation period of last year's report, which was 1995-2069, to the valuation period of this report, 1996-2070, the relatively large negative annual balance for the year 2070 is included. This results in a decrease in the long-range actuarial balance. (Note that the positive balance for 1995 is, in effect, retained because the funds accumulated during the year are included in the income rate and the actuarial balance for this year's report.)

Several demographic assumptions were modified: (1) the starting population was updated to reflect revised postcensal estimates (1990 through 1994) by the Bureau of the Census, which showed fewer people at working ages than did earlier estimates; (2) projected mortality rates were decreased, reflecting the latest data, which were, on balance, lower than expected for 1992 through 1994; (3) the ultimate rates of improvement in mortality for persons under age 65 were increased relative to the rates of improvement for older persons, consistent with long-term historical data; (4) projected fertility rates were decreased slightly through 2005, consistent with recent data that shows lower birth rates than did earlier estimates; (5) projected marriage and divorce rates were lowered, based on recent data; and (6) net legal immigration was decreased by 50,000 per year and net other-than-legal immigration was increased by 50,000 per year based on recent analysis by the Immigration and Naturalization Service. These modifications result in a net decrease in the long-range actuarial balance.

Ultimate economic assumptions were not changed for this report. However, revised short-range economic assumptions, including reduced revenue from payroll taxes based on revised earnings data for recent years, resulted in a reduction in the long-range actuarial balance.

Disability incidence rates were lowered throughout the 75-year projection period to reflect the exclusion of persons unable to work due to drug addiction or alcohol dependence, and recovery termination rates were increased within the first 25-year subperiod to reflect the increased funding provided for periodic continuing disability reviews; the effects of these changes are reflected under legislation, as they are based on provisions of Public Law 104-121. Additional changes in disability rates include (1) slightly lower ultimate incidence rates for males and higher incidence rates for females, consistent with recent data; and (2) lower death termination rates for disabled workers, in relation to general population death rates, consistent with recent data. These modifications result in a reduction in the long-range actuarial balance for the DI program.

Several significant improvements and updates were made in the methods used to project the cost and income of the OASDI program. The method used for projecting the number of dually entitled beneficiaries (those receiving both a retired or disabled worker benefit, and a residual spouse or survivor benefit) was changed to better reflect the projected population by marital status. This change significantly increased the long-range actuarial balance for the OASI program. The rate of change in benefits for female retired workers, after first becoming entitled, was increased, reflecting recent data. Benefits tend to increase after entitlement faster than suggested by the cost-of-living adjustment because of increments based on benefit withholding for earnings, and because individuals with higher benefits tend to have slightly lower death rates. This change resulted in a small reduction in the OASDI actuarial balance. Updated sample data for benefits awarded in 1994 were used as the starting point for projecting the level of average benefits for future beneficiaries. The increase in average benefit levels from the previous sample, for 1993 awards, to the new sample was slightly higher than had been previously projected, resulting in a small reduction in the OASDI actuarial balance. An error that resulted in an understatement of the estimated amount of revenue from taxation of benefits in the 1995 report was corrected, resulting in a small increase in the actuarial balance.

The cost of the OASDI program has been discussed in this section in relation to taxable payroll, which is a program-related concept that is very useful in analyzing the financial status of the OASDI program. The cost can also be discussed in relation to broader economic concepts, such as the gross domestic product (GDP). OASDI outlays generally rise from a little less than 5 percent of GDP currently to about 6.6 percent of GDP by the end of the 75-year projection period under alternative II. Discussion of both the cost and the taxable payroll of the OASDI program in relation to GDP is presented in appendix C.