Reports & Studies

Research Note #14:
Key Data From Annual Trust Fund Reports

Understanding The Trust Fund Reports
The business of predicting the future is always difficult. How much more so when we are asked to predict the future in precise numerical detail! But that is the task facing the Social Security actuaries. They are asked to make actuarial estimates that summarize the future income and expenditures to the Social Security system for many decades into the future. The actuaries must consider all sorts of demographic, economic and legislative factors in making their estimates. Such things as birth rates, death rates, future wage levels, future benefit levels, future inflation rates, etc. etc.

Given all these complex factors, and the inherent ambiguities of the future, it should be obvious that actuarial estimates, despite their seeming numerical precision, contain large amounts of uncertainty. Because of this inescapable uncertainty, some unique concepts are deployed in making the long-range estimates.

"Close Actuarial Balance"
Because of the inherent imprecision and uncertainties in long-range actuarial forecasts, early on the Trustees deployed a concept known as "close actuarial balance." The idea being that even if the forecasts showed a small deficit in the long-range balance, such small deficits would not prevent a judgment that the system was adequately funded. One way to understand this is that deficits of small magnitude could be considered within the range of error in such long range estimates. This concept was first deployed in the 1957 Report with the addition of the new Disability Trust Fund. Each Trust Fund (OASI and DI) was assigned a range in which deficits were judged to be small enough to not raise questions regarding the system's overall solvency.

Percent of Payroll & Level Premium Cost
The Social Security Trust Fund involve billions of dollars over decades of time. Trying to project Trust Fund performance decades into the future only in dollar terms is a complicated and sometime unwieldy matter. For this reason, the actuaries early on developed an alternative way of assessing the Funds--in terms of "percent of payroll." The "percent of payroll" is the same thing as the tax rate for covered earnings. So if the tax rate is 3% on the employee and the employer, we could express the income to the program as being 6% of payroll. We could then express the future expenditures in the same way, and by comparing the two figures we would have an easy way to see the status of the system. So if the long-range cost of the system is, say, 7% of payroll and the existing tax schedule is 6% of payroll, then the system would have a deficit of 1% of payroll. One potential advantage of this type of assessment is that it tells policymakers precisely how much payroll taxes would have to be raised to close any deficit--if raising payroll taxes were an option under consideration.

A closely-related idea is that of the "level premium cost" of the system. The "level premium cost" is the tax rate that would have to be put in force at the start of the estimation period in order to fully meet the projected obligations of the system. This is actually the basis for the long-range comparisons--the level premium cost of the system versus the scheduled tax rates for the system. Although the idea of the "level premium cost" may be a bit strange to non-actuaries, the simplest way to think about it is as a comparison of the future income and outgo of the system expressed as a percentage of payroll--that's the essence of the point here.

Estimation Period
Currently, the Trust Fund estimation period is 75 years. This period has become the de facto standard for evaluating the long-range solvency of the Social Security program. In itself, this is a remarkable standard for assessing a fund's solvency. Few private pension systems attempt anything like this long assessment period. It is interesting to note, however, that this idea of a 75-year evaluation period has not always been used. Over the years, evaluation periods as short as 35 years and as long as 80 years have been used to assess Trust Fund solvency. The 75-year evaluation period has been in use only since 1965.

Intermediate Cost Estimates
Starting with the 1951 Report, the actuaries introduced the concept of an Intermediate cost estimate, which initially was the average of the High and Low cost estimates. This change was mandated by the 1950 Amendments. Although the actuaries frequently issued disclaimers to the effect that the Intermediate estimate was not to be considered the "most likely" outcome, the disclaimers were mostly ignored. Probably frustrated by the uncertain conclusion in the early Reports, Congress began mandating the use of a single estimate as the basis for the conclusion as to long-range solvency, and the Intermediate estimate was adopted for this purpose.

Generally speaking, there usually are three (sometimes more) sets of estimates which we can somewhat loosely describe as Optimistic, Pessimistic and Intermediate. Since the introduction of the Intermediate estimates most commentary on Trust Fund solvency is usually based on the Intermediate estimates.

Static vs. Dynamic Assumptions
Among the factors in the actuarial estimates, two key factors are the level of benefits paid and the level of earnings subject to payroll taxation. Prior to the 1972 Amendments, there were no provisions in the law that would automatically increase these two factors. An increase in either benefit levels or the taxable wages required a legislative act. Or to say it another way, the law presumed the program to be static as to these two factors. As a result, the actuarial estimates assumed these two factors to be static as well. This was sometimes referred to as the "level-wage" and "level-benefit" assumptions.

However, the facts in the economy were that wage levels tended to increase over time and prices tended to increase over time. (Wages tended to increase a little more than prices throughout the post-War period, until the early 1970s when the phenomenon economists came to call "stagflation" first appeared.)

Since in the real economy both of these factors tended to rise over time, the program was frequently amended to adjust to these increases. One interesting consequence of the use of the static assumptions was that the cost of the program tended to be overestimated. That is, since wages generally increased more than prices, the potential income to the system would tend to increase faster than the potential outgo to the program. So that when a legislative adjustment was made to reflect the actual increases in wage and prices, lawmakers would often find themselves in the happy circumstance of having more income to the system that had been contemplated in the actuarial estimates. This often allowed benefit increases to be made without any attendant increase in the contribution rates. This kind of "windfall" was in fact used many times as a justification for benefit increases in the period prior to the shift to dynamic estimates.

The Trustees had always maintained the position that unless provisions were added to the law to provide for automatic increases in these factors, the actuarial estimates should remain static. In 1972 the law was amended to provide for automatic benefit increases (COLAs) and automatic increases in the taxable wage base to keep pace with the increase in average wages. As a result of these changes in the law, the Trustees adopted the automatic estimating procedures starting with the 1973 Report. These dynamic estimates have been used in all Reports since that time.

An Overview of the Results
During the 61 years in which Trust Fund Reports have been issued, the Funds have been out of long-range balance almost as often as they have been in balance. In 26 of the 61 years, the Funds have been in long-range balance, and in 24 years they have not. In 11 Annual Reports the status of the Funds is uncertain based on the data presented. Of the 26 years in which the Funds were said to be in balance, in 15 of these years the Funds actually showed a small deficit but were judged to be "in close actuarial balance." In only 11 years since Reports were first issued have the Funds been in simple positive balance--the last time being shortly after the passage of the 1983 Amendments.

The first financing shortfall occurred in 1973 and continued throughout the 1970s. Legislation enacted in 1977 was intended to address Social Security's financing, but with limited success. Long-range imbalances continued until passage of the 1983 Amendments restored the system to long-term solvency--but only temporarily. The system again went out-of-balance in 1988 and has been out-of-balance ever since. The peak year of imbalance occurred in 1997. The Trust Funds have been steadily improving since the 1997 Report (as can be seen by examining the date-of-exhaustion and percent-of-payroll values), but the Funds are still in substantial imbalance.

Footnotes:

{1} Long-range balance was sometimes given in dollar terms and sometimes as a percent of payroll. Figures in (brackets) indicate negative values. In some periods, the numerical figures are not available and the narrative conclusion is the only available information. In other periods, actuarial balance was found despite the presence of negative values in the Trust Fund balance.

{2} In the 1941, 1942, 1943, 1944 and 1945 Reports two sets of estimates are given: one indicating the Trust Fund is in long-range balance and one indicating it is not. No basis is given for choosing between the two alternatives.

{3} Although no definitive conclusion is given as to the long-range actuarial status of the program, the Trustees issued a warning that the refusal of Congress to allow scheduled tax rates to rise as contemplated in the 1939 law was potentially placing the program in financial jeopardy. The Trustees urged an immediate tax rate increase to 2% each on employers and employees (from the existing 1% level).

{4} In this Report the Trustees again complain about the failure to allow tax rates to rise on the schedule contemplated in the 1939 law. They point out that in the 1943 Deficiency Act a provision was added permitting General Revenue contributions to Social Security to make-up any financing shortfall. They suggest this is an indication that the program is not adequately financed, but no such explicit conclusion is taken by the Trustees..

{5} No long-range estimates provided in Report.

{6} The Report shows four alternative scenarios, two high-cost and two low-cost. Trust Fund in balance in two low-cost alternatives, but not in high-cost. No most likely scenario identified.

{7} The Report shows four alternative scenarios, two high-cost and two low-cost. Trust Fund in balance in three of the four scenarios. No most likely scenario identified. Report presents first chart indicating that the value of benefits not keeping pace with increases in cost-of-living.

{8} The Report shows four alternative scenarios, two high-cost and two low-cost. Trust Fund in balance in only one of the four scenarios. No most likely scenario identified. Report presents chart and narrative arguing that the value of benefits have not kept pace with increases in cost-of-living.

{9} The Report shows four alternative scenarios, two high-cost and two low-cost. Trust Fund in balance in only one of the four scenarios. No most likely scenario identified.

{10} The Report introduces the idea of a high-cost, low-cost and intermediate-cost set of scenarios. Congress mandated selection of a single set of estimates to assess long-range actuarial balance. Intermediate-cost scenario used as the basis for this assessment.

{11} Long-range estimates same as those in the 1951 Report.

{12} In this Report, there are six sets of estimates provided: two Low, two High and two Intermediate estimates. Under one of the Intermediate estimates, a positive balance of $56 billion exists at the end of the estimation period. Under the other, the fund is exhausted in 1995 (five years before the end of the estimation period). The Report identifies no clear set of estimates to be used in assessing actuarial balance. Furthermore, the Report also represents long-range balance by comparing the "level cost" of the system to the tax rate schedule in the law. Under both Intermediate assumptions, the level-cost of the system is slightly higher than the tax rates in the law. Thus, the system is not fully self-supporting under this metric--using either of the Intermediate estimates. Given these results, and the studied ambiguity of the Report's narrative, it is impossible to come to a firm assessment as to whether the program was in long-range balance in the 1954 Report.

{13} Starting with this Report, the new Disability Trust Fund is added.

{14} This Report is the first to use the 75-year estimation period for purposes of assessing actuarial balance--following a recommendation of the Advisory Council.

{15} Starting with the 1966 Report, the Medicare Insurance Trust Funds were also included in the annual reports--although Medicare is not included in this chart.

{16} This Report used, for the first time, an additional set of estimates based on so-called dynamic actuarial assumptions, as recommended by the 1971 Advisory Council. However, since dynamic provisions were not yet part of the law, this set of estimates was for illustration purposes only.

{17} This Report, and all future Reports, now make use of the dynamic actuarial assumptions since automatic provisions were added to the law in the 1972 Amendments.

{18} This breathtaking acceleration in the date-of-exhaustion was an expression of a short-term financing crisis facing the program. If the short-term crisis could be overcome, the long-range outlook for the program was much more favorable, with the subsequent date of exhaustion of 2035 being very close to that reported in the 1979 Report.

{19} The 1981 Report introduced the idea of two Intermediate estimates, designated II-A and II-B. These two estimates used the same demographic assumptions, but the II-A estimate used more favorable economic assumptions. Over time, the II-B assumptions came to be the indicators preferred for most uses. For ease of presentation, we use only the II-B figures in this chart. The reader should keep in mind that in every case, the II-A assumptions would be more favorable. Like the 1980 Report, the dates of Trust Fund exhaustion reflect a short-term financing crisis, with the financing improving in the medium and long terms. The long-range date of exhaustion under II-A was 2040 and under II-B was 2025.

{20} Same short-term financing issues as in the two prior Reports. Date of depletion of the OASI Trust Fund was in fact pegged at July 1983. Long-range date of exhaustion under II-B was 2030.

{21} Although the actuarial deficit in this report was small enough for the program to still be considered "in close actuarial balance," this was the first post-1983 report showing a trust fund depletion date within the 75-year estimating period. By that measure--trust fund depletion date--the program's long range financing was already inadequate. Also, the actuaries changed their estimating methodology in this report. Previously, they had assessed actuarial balance on an "average cost" basis throughout the estimating period. Starting with this report, they shifted to a "level financing" basis. Under the prior methodology, the actuarial deficit in 1988 would have been 0.87% of payroll rather than the 0.58% reported under the level financing method. Under the old measure the system would already have been judged out of long range actuarial balance.

{22} Starting with the 1991 Report, the two sets of Intermediate estimates known as II-A and II-B were dropped in favor of a single Intermediate estimate (along with a Highcost and a Lowcost set of estimates). From this point on in the table, all values shown are from the Intermediate set of estimates.

{23} Starting with the 2003 Report, the actuaries began including--in addition to the traditional 75-year estimates--a set of estimates designed to project the financing of the program to "infinity." This was not a replacement to the standard 75-year estimates, but rather, an additional measure of financial adequacy.