36

Although most of the increase in cost is expected to occur after the turn of the century (when the effects of the changes in the population projections are fully felt), part of it will already occur within the next few years, thereby producing a marked decline in the near future in the ratio of assets to expenditures in the absence of an immediate increase in income to both the OASI and DI Trust Funds. In the very short run (for the next 5-10 years) a reallocation of the current contributions could cover this problem. The overall OASHDI contribution rate in present law would be enough, if reallocated, to adequately support all three trust funds (OASI, DI and HI) during this period. However, after the next 5-10 years, a tax increase or constraints in the growth of benefits will nonetheless be needed for each of the three programs.

Another important factor that has affected the long-range cost of the program is the recent increase in the number of disabled-worker benefits that are being awarded. As was indicated in last year's report, a significant increase in the disabled-worker benefit awards has been experienced since 1971. The present cost estimates incorporate all the increases that have been experienced through the end of calendar year 1973.

As shown in table 21, other factors affecting the cost of the program are changes in economic assumptions, which include modifications in the projected labor-force participation rates and unemployment rates, short-range assumptions regarding increases in average earnings, and both short-range and long-range assumptions regarding increases in the CPI (with the long-range assumptions changing from 2-3/4% to 3%), as well as the elimination of the 3/8 percent margin used in previous long-range cost estimates. The remaining itemized factor affecting the actuarial balance is a small increase in the rates of retirement among the eligible aged population that has been observed in the last two years.

TABLE 21.-CHANGE IN OLD-AGE, SURVIVORS, AND DISABILITY INSURANCE LONG-RANGE ACTUARIAL BALANCE 1 AS PERCENT OF TAXABLE PAYROLL BY TYPE OF ASSUMPTION

1 Represents the difference over the 75-year period, 1974-2048, between the average tax rate and the average cost. 3 Payroll is adjusted to take into account the lower contribution rate on self-employment income, on tips, and on multiple-employer "excess wages" as compared with the combined employer-employee rate.

Table 22 shows the current-cost of the OASDI System (including the cost of maintaining one year's expenditures on hand) for selected years over the next 75 years, expressed as percent of taxable payroll, in accordance with the dynamic actuarial assumptions.

34-500 O-74-42

37

It may be observed that the annual cost of the Old-Age, Survivors, and Disability Insurance System is projected to increase slowly throughout the remainder of this century and that after the turn of the century it will increase rapidly until leveling at about 17-18 percent of taxable payroll after the year 2025.

According to the present 75-year projections, the cost of the OldAge, Survivors, and Disability Insurance System could be divided into three periods of 25 years each. The first period is projected to be a period of slowly increasing costs. The second period involves fast increases in cost, while the third period is characterized by high but level costs.

TABLE 22.-ESTIMATED "CURRENT-COST" OF OLD-AGE, SURVIVORS, AND DISABILITY INSURANCE SYSTEM AS PERCENT OF TAXABLE PAYROLL UNDER DYNAMIC ASSUMPTIONS, FOR SELECTED YEARS, 1985-2045 [In percent]

1 Represents the cost as percent of taxable payroll of all expenditures in the year, including amounts needed to maintain the funds at about 1 year's expenditures.

* Payroll is adjusted to take into account the lower contribution rate on self-employment income, on tips, and on multipleemployer "excess wages" as compared with the combined employer-employee rate.

* See text for a description of the assumptions.

4 Represents the arithmetic average of the "current-cost" for the 75-year period 1974-2048.

The increasing costs in the second period as well as the high costs in the third period are due principally, but not totally, to the demographic effect of the projected large aged population as compared to the working population. Some of the cost, however, is due to what could be considered anomalies in the automatic benefit adjustment provisions in present law. As is discussed in the appendix and as may be noted from Appendix Table C, the present automatic provisions are projected to result in awarded benefits that would increase faster than average earnings in the future. The differential in trends between average awarded benefits and average covered earnings would be relatively minor during this century (because of the way benefits are calculated under present law), but it is projected to increase substantially thereafter.

The Board of Trustees has suggested to the Department of Health, Education, and Welfare that the present Advisory Council be asked to study this matter, as well as other ways of dealing with the emerging long-range actuarial status of the trust funds.

38

CONCLUSION

The long-range actuarial cost estimates for the old-age, survivors, and disability insurance program prepared in accordance with dynamic assumptions as to both benefits and taxable earnings show an actuarial balance of-2.98 percent of taxable payroll over the valuation period of 75 years, which substantially exceeds the acceptable limit of variation of 5 percent of the cost of the program (0.69 percent of taxable payroll).

The principal reason for the increase in the actuarial imbalance, as compared to that reflected by the cost estimates used last fall by the Congress, is a change in the long-range population projections underlying the cost estimates, which are now based on the results of the 1970 Čensus and on lower future fertility assumptions than were previously used for such projections.

Although the new population and fertility projections will have a major impact after the turn of the century on the long-range cost estimates, they will not have a significant effect in the short run. According to present short-range cost estimates, action to increase the combined income of the OASDI and hospital insurance systems for the next 5-10 years is not necessary right now. Although, when considered separately, the Disability Insurance Trust Fund and, to some extent, the Old-Age and Survivors Insurance Trust Fund decline in terms of both absolute dollar amounts and as a percent of outgo, the Hospital Insurance Trust Fund is increasing more rapidly than previously projected, with the result that it is developing an excess of funds. The Board noted that one of the possible ways that the projected short-range excess of outgo over income in the cash benefit funds can be avoided is a reallocation of the total program income among the three funds (OASI, DI, and HI) by revising the contribution rates scheduled in present law without increasing the total rate. However, in order to maintain the HI Trust Fund in actuarial balance, any reduction in the HI tax rates in the early years would have to be offset by compensatory increases in later years.

The present assumptions as to the rate of increase in the CPI, in both the short-range and the long-range estimates, assume some deceleration from recent rates of increase. If this deceleration does not occur, or occurs more slowly than assumed, the reallocation noted above may not be sufficient over the next 5-10 years to prevent a decline in the funds. And, of course, if such deceleration does not occur and if, as is assumed, recent fertility trends should continue, the additional financing needed over the long-range will be increased.

Although there is of necessity a considerable degree of uncertainty inherent in the long-range demographic and economic assumptions and consequently in the projections that flow from those assumptions, it is certain that additional income to the cash benefits program or some adjustment in the benefit structure will be needed eventually. However, in view of this inherent uncertainty and the fact that the newly appointed Advisory Council on Social Security is studying the long-range financial status of the social security system, the Board is not recommending a specific increase in the combined OASDHI contribution rates scheduled in present law. The Board believes that there is ample time to await the Council's findings and recommendations before making specific proposals.

APPENDIX

STATEMENT OF ASSUMPTIONS, METHODOLOGY, AND DETAILS OF LONGRANGE COST ESTIMATES

(Prepared by Office of the Actuary, Social Security Administration)

The basic assumptions used in the long-range estimates for the Old-Age, Survivors, and Disability Insurance System are described in this Appendix. Also given here are some detailed data in connection with the results of these estimates. Section A of this Appendix provides a description of the demographic aspects of the long-range cost estimates, while Section B discusses the economic aspects. These terms are used in a general sense, since it is not entirely possible to fully separate the effect of these two aspects on the cost estimates. By "demographic aspects" we mean those elements dealing with the population and its characteristics. These include the number, age, sex, marital status, retirement, disability, mortality, fertility, employment, and coverage under the system. By "economic aspects" we mean the monetary elements of the estimates which primarily consist of the projected annual increases in the Consumer Price Index (CPI) and in the average earnings in covered employment.

A. DEMOGRAPHIC ASPECTS

This section of the Appendix discusses the methods used to estimate the demographic elements of the OASDI cost projections and their effect on the cost estimates.

(1) Population

A projection was made of the United States population (including persons overseas covered by the Old-Age, Survivors, and Disability Insurance Program) for future quinquennial years, by 5-year age groups and by sex. The starting point was the population on July 1, 1973, as estimated by the Bureau of the Census from the 1970 Census and from births, deaths, and migration in 1970-73. This population estimate was adjusted for differences in the geographical areas covered by the estimate of the Bureau of the Census and those covered by the Old-Age, Survivors, and Disability Insurance System.

In the population projection it was assumed that mortality rates will decline until the year 2000. Mortality rates for the year 2000 are, on the average, about 10 percent lower than those experienced in 1972.

The total fertility rate was assumed to increase slowly until the year 2005 after which it was assumed to remain constant at a level of 2.1 children per woman, which is close to replacement rate and which would ultimately result in zero population growth. În addition, the projection assumes a small amount of net immigration. Complete details about the population projection will be given in a new actuarial study to be published by the Social Security Administration.

(2) Employment

Assumptions as to the percentage of the population who have covered employment during a year were made for each age group by sex for each quinquennial year. For men, the estimated average percentages for 1968-71 were projected to increase for teenagers, to decrease slightly for those aged 20-59, and to decrease significantly for those aged 60 and over (thus recognizing the trend toward higher retirement rates). For women, the corresponding percentages were assumed to increase, except for those aged 60 and over for whom a decrease was projected. The foregoing projections are consistent with the projected average unemploy ment rate of 5 percent. Although a depression lasting several years could increase the cost, it is assumed that any periods during which unemployment is substantially higher than the assumed level would be of relatively short duration or would be balanced by offsetting periods of low unemployment, and would therefore have virtually no long-range cost effect.

40

(3) Insured Population

The term "insured" is used as meaning fully insured, since the number of persons who are currently-insured only is relatively small and can be disregarded for long-range cost analysis purposes. The percentages of insured persons by age and sex in various future years are estimated from recent experience and from the projected coverage. It is evident that eventually almost all males in the country will be insured for old-age and survivors benefits; the ultimate percentage for aged males is estimated at 97 percent. For females it is estimated that the corresponding proportion will eventually be 71 percent. This is lower than for males because of the lower participation rates of females in the labor force.

The estimated number of persons insured for disability benefits are lower than those insured for old-age and survivor benefits because of the more restrictive insured status provision for disability benefits. These were also estimated on the basis of recent experience and the projected percentage of persons covered. (4) Old-Age and Survivors Insurance Beneficiaries

Old-age beneficiaries were estimated from the aged insured population. The proportions, by age and sex, of the insured population that were receiving benefits at the beginning of 1973 were projected to increase according to past trends after adjustment for changes in the earnings test and in the level of unemployment, thereby reflecting assumed gradual increases in the retirement rates.

Wives aged 62 and over of male old-age beneficiaries were estimated by using census data and mortality projections. These potential wife beneficiaries, after adjustment for eligibility for their own old-age benefits, were assumed to claim benefits as soon as they are eligible, even if this occurred at ages 62-64, when they would have to take reduced benefits. The experience to date indicates that in the vast majority of the cases, such immediate claiming of wife's benefits does occur. Young wives and children of retired workers were estimated by reference to their ratios to male old-age beneficiaries, as derived from recent actual data and projected according to the aforementioned fertility and mortality assumptions. Child-survivor beneficiaries were obtained from estimates of total paternal orphans in the country in future years. The projected child population, by age group, was multiplied by the probability of being a paternal orphan. These probabilities were derived by using distributions of age of fathers at birth of child and death rates consistent with the population projections. The number of paternal orphans was then adjusted to eliminate orphans of uninsured men, to add orphans of insured women and to include the eligible disabled orphans aged 18 and over. For the non-disabled children aged 18–21, a further reduction was made to exclude those not attending school. Mother survivor beneficiaries were estimated by extrapolating the present ratio of mothers to children, after excluding those nondisabled children 18-21 who were attending school, to reflect the lower projected fertility.

To estimate widow beneficiaries the proportions of widows in the female aged population were projected according to mortality assumptions and adjusted for both eligibility for their own old-age benefits and for the insured status of their deceased husbands. These uninsured eligible widows were assumed to claim benefits as soon as available even if this occurred at ages 60 to 64, when they would have to take reduced benefits. For ages 50-59, the disabled widow beneficiaries were estimated from the eligible widows by using disability prevalence

rates.

It can be observed that the assumed wife and widow beneficiaries consist of the uninsured potential beneficiaries. In actual practice, some of the insured potential beneficiaries also receive a residual benefit consisting of the excess of the potential wife's or widow's benefit over their own old-age benefit. These residual benefits, although not giving rise to additional aged beneficiaries, were considered in the cost of the particular type of dependent or survivor benefit concerned.

The minor category of parent beneficiaries was projected by adjusting the previous estimates (which had been projected as a constant proportion of aged persons not eligible for any other benefit) to reflect the experience through the end of calendar year 1973. The insignificant effect of the retirement test as it applies to wife's, widow's and parent's benefits was ignored. No separate estimates were made for benefits to dependent husbands and widowers since their cost is relatively negligible.

Appendix Table A shows the estimated number of beneficiaries in the old-age and survivors insurance program.