Reports & Studies

A Paper Prepared for The Commission on the Social Security 'Notch' Issue

by  James W. Kelley
Former Staff Director, Subcommittee on Social Security, House Ways and Means Committee

and   Joseph R. Humphreys
Former Professional Staff Member, Senate Finance Committee

OVERVIEW

I INTRODUCTION
WHAT IS THE NOTCH?
CONGRESSIONAL INTENT

II THE LEGISLATIVE ROOTS OF THE NOTCH
BENEFIT INCREASES PRIOR TO THE AUTOMATIC COST OF LIVING ADJUSTMENTS
1. Legislative action in 1969
2. Legislative action in 1970
3. Legislative action in 1971
4. Legislative action in 1972

III DEFECTS IN THE AUTOMATICS APPEAR
ACTUARIAL STATUS IN 1972
1973 UNEXPECTED INFLATION
1974 TRUSTEES REPORT INDICATES A PROBLEM
FIRST HSIAO PANEL CONFIRMS A VERY SEVERE FINANCING PROBLEM
1974 ADVISORY COUNCIL RECOMMENDS WAGE INDEXED FORMULA
SECOND HSIAO PANEL RECOMMENDS PRICE INDEXED APPROACH

IV LEGISLATIVE ACTION
FORD ADMINISTRATION PROPOSAL
PRESIDENT CARTER'S RECOMMENDATIONS
DIFFERENCE IN ADMINISTRATION BILLS CONCERNING THE NOTCH
HOUSE ACTION IN 1977
SENATE ACTION IN 1977
FINAL CONGRESSIONAL ACTION
1979 CONGRESSIONAL HEARING ON THE NOTCH

V CONCLUSIONS

OVERVIEW

The 1970's were tumultuous years for the Social Security System. In 1972, legislation was enacted designed to automatically keep benefits up to date with inflation while at the same time assuring adequate financing to support the program into the long-range future. In addition, the 1972 legislation provided a one-time increase of 20% in Social Security benefit levels. Almost immediately economic conditions dramatically worsened with serious consequences for the Social Security program. Rapid inflation caused Congress to add additional ad hoc benefit increases. The unexpectedly high inflation also interacted unfavorably with the new automatic benefit formula, and caused benefit costs to eat into the trust fund reserves. During this same period, it became clear that other factors such as disability experience and long-range fertility trends were developing in ways that would undermine the program's financing. By 1977, the long-range actuarial situation of the Social Security system was enormously out of balance and, in the short-range, trust fund reserves were facing exhaustion within a very few years. At the same time, the 1970's was a recessionary period and the prospect of raising taxes -always unpleasant -- was particularly unappealing.

The 1977 Social Security amendments attempted to restore the program's financial soundness. A major element of those amendments was a change in the formula (i.e. benefit computation method) for Social Security benefits. Individuals born prior to 1917 were left under the old formula. Individuals born after 1916 were required to have their benefits computed under the new rules. Many of those born in 1917 and the years immediately following 1917 believe that the change unfairly discriminates against them as compared with persons born in earlier years and also as compared with those who come well after them.* This alleged discriminatory impact is referred to as the notch issue.

(*Note that individuals born on January 1 of a particular year are considered to have been born in the previous year.)

Social Security is a complex program having many rules which can significantly affect benefit levels. This makes it difficult to compare the benefits payable to individuals whose circumstances are not identical in every respect. Under either the old law or the new law, it is possible to draw examples of individuals who seem in many ways to be similarly situated but who get strikingly different benefits because they are not identically situated. It is also possible to find reasonable examples of a "reverse notch" under which the 1977 amendments have resulted in substantially higher benefits for persons born after 1916 compared to persons born a few days earlier (but just prior to 1917) who have very similar or even identical earnings histories.

Nonetheless, the basic operations of the new and old benefit formulas do tend in many cases to create the results cited as the notch issue. The two sides of the notch -- the comparison with later retirees and the comparison with earlier retirees -- arise from different decisions.

Future beneficiaries do better than those who came on the rolls in the years just after implementation of the 1977 amendments. The 1977 benefit formula is designed to provide stable replacement rates -- benefits as a percentage of preretirement earnings. Since, in most years, wages rise faster than prices, this tends to result in each year's new cohort of beneficiaries receiving benefits which are higher in real terms than the initial benefits for similar retirees in previous years. Both the mechanics of this result and the principle that it gives higher benefits for workers m the future appear to be a conscious policy choice made by Congress. A Congressional panel had recommended an alternative approach of keeping initial benefits level in real terms from year to year. One of the rationales stated for that approach in the panel's report was that it would avoid treating the retirees in future years better than the retirees in earlier years. This alternative approach was considered and specifically rejected by the Congress.

The other, and more widely discussed, side of the notch issue is the differential in benefits between those in the prenotch years (born prior to 1917) and those born in the notch years (the 5 to 10 years after 1916). In general terms, it can be said that the result is consistent with the overall objective that Congress intended to achieve which was to restore the solvency of the program by a combination of increased revenues and constrained benefits -- and in particular by adopting a new benefit formula that was less generous (particularly in times of high inflation) than the old law benefit formula.

In the 1977 Social Security amendments, therefore, Congress acted to restore the solvency of the program by a variety of measures aimed primarily at cutting costs and raising revenues. The most significant cost cutting item was the adoption of a new benefit formula that was intentionally set at a level which represented a rollback from the levels attained by the old formula. In implementing any new system, a decision has to be made as to when and how it will be applied. Congress chose to make the new formula effective for those who would attain retirement age starting about a year after the law was enacted, that is, those born in or after 1917, who would reach age 62 in or after 1979.

The fact that Congressional intent was to implement a less generous benefit formula is, in a general way, consistent with the result that those in the notch years receive lower benefits than those in the prenotch years. However, there is no evidence that Congress directly focused on the question of comparative benefits for the two groups. The legislation did include a transition clause for individuals reaching age 62 in the first five years after implementation, but this provision was aimed at protecting the benefit expectations of those individuals rather than at providing any type of parity with those born in other years. In the hearings on the legislation, one witness did submit with his testimony (before the Senate and House Committees) an addendum which clearly and specifically identified the notch issue, but there is no evidence of any follow up to this testimony.

Why did Congress not focus on this issue? The ability to analyze the impact of the 1977 benefit formula was constrained in a number of ways. There were a large number of major issues competing for attention including novel and controversial financing schemes and other difficult benefit rule changes. For most of the period of Congressional consideration, there were multiple contenders for the role of new benefit formula each of which would have affected the comparison in different ways. Also, the complexity of the Social Security system makes it difficult to construct truly typical examples.

Nonetheless, examples could have been constructed to address the question of benefit differentials between those coming under the new law and those remaining under the old law. Had such examples been constructed, they would have shown much smaller differentials than those which actually materialized because the assumptions underlying the 1977 amendments did not foresee the unprecedented inflation of the following several years.

It can only be a matter of conjecture as to how Congress would have reacted if the notch issue had been clearly framed at the time of the 1977 amendments. If Congress had wanted to change the policy, it could have done so. The notch, in fact, would not have occurred under the legislation originally submitted by the Administration in 1976 since it would not have permitted those born before 1917 to use post-1978 wages in an old-law computation. For reasons unrelated to the notch issue, this restraint was dropped when the new Administration resubmitted the legislation in 1977. The notch could also have been lessened by allowing those in the transition years (born 1917-1921) to use post-age-62 earnings, as apparently envisioned by the 1974 Advisory Council or by a transition clause involving a blending of old and new law benefit computations, as recommended by the Hsiao panel. Other approaches also might have been developed for Congressional consideration. If Congress had addressed the issue, the overwhelming concern to restore program solvency would logically have leant support to the approach of limiting benefit growth for those remaining under the old-law, rather than raising benefit levels for those qualifying for benefits under the new benefit formula.

I INTRODUCTION

This report was undertaken at the request of the Commission on the Social Security "Notch" Issue with a view to examining the legislative background of that issue and, insofar as possible, the Congressional intent underlying the legislation involved.

What is the Notch?

In the Social Security Amendments of 1977, Congress changed the formula for computing Social Security benefit amounts. The change in law was made effective for individuals reaching age 62 (the age of first eligibility for Social Security retirement benefits) in 1979 or later. Persons born in 1916 or earlier remained eligible for benefits under the old formula and were not allowed to use the new formula. Persons born in 1917 or later were required to use the new formula. They were not allowed to use the old-law formula, but a special "transitional" formula was provided as an alternative for those reaching age 62 in the first five years of the new law: 1979-1983.

The new formula, in its first year of applicability (1979), was designed to produce benefit levels averaging about 5% lower than those that would have resulted from the old formula.(1) However, under the transition formula, persons reaching eligibility age and retiring in 1979 were guaranteed a benefit at least equal to the benefit the old law formula would have produced.

For persons reaching age 62 after 1978, the new formula generally produced lower benefits than would have been available under the old formula. Both the old and new formulas were automatically adjusted in a way that results in increasing benefit levels from year to year. However, the growth rate in benefit levels under the old formula was higher than the growth rate under the new formula. Consequently, persons who could continue to use the old formula (i.e., persons who were born prior to 1917) and who worked beyond 1978 tended to receive benefits that were higher (in many cases significantly higher) than persons with comparable earnings who were required to use the new formula (i.e. persons born after 1916).(2)

For those born from 1917 through 1921, the transition clause provided a guarantee that their benefits would be no lower than what they would have qualified for under the old law formula as of 1979. However, the old law formula was applied for these persons without adjustment for inflation between 1978 and the year of first eligibility (age 62) and without the use of earnings after age 62. Both of these elements--inflation adjustments and post-62 earnings--tended to heavily influence the growth of benefit levels under the old law formula. Consequently, the transition formula often did little or nothing to lessen the differential between benefit levels for those born in and after 1917(3) compared with those born earlier.

(1)  Conference report on 1977 amendments (House Report No. 95-837) as printed in U.S. Congress, Congressional Record, (Permanent Edition), v. 128, p. 38943. (December 15, 1977).

(2)  It was, however, quite possible for some individuals to receive significantly higher benefits under the new benefit formula than what was payable to persons with similar earnings records who were born prior to 1917 and thus required to use the old law formula. An example would be two women, one born in December 1916 and the other in January 1917. If both had maximum earnings from 1951 through 1965 (age 35 through 49) and then dropped out of the work force to care for an elderly parent, the woman born in 1917 (a notch year) would qualify in January 1979 for a benefit of $241.80 while the pre-notch woman (born a few days earlier but in 1916) would have a benefit of $200.40. In other words, with the same earnings record and a birth date a few days apart, the notch years retiree would get a 20 % higher benefit than the pre-notch retiree.

(3)  There is no universally accepted definition of what constitutes the "notch years''. It clearly begins with those born in 1917 -- the first category permitted and required to use the new wage indexing formula. The birth year ending the notch period is most commonly considered to be either 1921 or 1926.

While most discussions of the notch issue focus particular attention on the noticeably lower benefit level for many of those born in and after 1917 compared with the benefit level for those born earlier, there is also an issue related to comparative benefits for those in the notch years and those who came on the rolls after the notch years. The very term "notch" implies a "V" or "U" shaped cut in which the area cut out is lower than the areas on either side of the cut. Because initial dollar benefit levels do grow from year to year under the new formula, individuals reaching eligibility age well after the 1979 changeover year will receive higher benefits than those who came before them. This is true even if benefit levels are adjusted for inflation.

Thus the notch category - those born in the first few years after 1916 -- did in many cases receive benefits which can be characterized as lower than the benefits for those who came before them and lower than the benefits for those who came after them. Benefits were lower than for those who came before them because the new benefit formula was intended to produce lower benefits than the prior law formula. Benefits were lower than for those who came after them because the new benefit formula is designed to provide each year's cohort of retirees with benefits which, in real terms, are higher than the benefits payable to comparable workers who reached retirement age in earlier years.

Congressional Intent

Congressional intent is sometimes clearly delineated in the legislation itself or in the Committee reports and floor debates accompanying the consideration of a measure. This is not the case with respect to the notch issue.

The legislative record contains some discussion, as noted later in this paper, of the fact that the new benefit formula provided lower real benefits for those reaching retirement age in any given year compared with those who would retire in later years. On the more prominent aspect of the notch issue, however, there is no outright statement in the "official" legislative history to indicate that Congress focused its attention on the question of whether those reaching eligibility in the first several years after enactment would receive substantially different benefits from those coming before them. Consequently, there is also no clear and obvious indication of whether such a result would or would not have been acceptable to the Congress at the time of enactment.

In the absence of any clear statements of intent, this paper attempts to provide an indirect analysis of Congressional intent by examining the context surrounding the 1977 amendments in addition to the official legislative history. What were the main objectives addressed by the legislation and how is the notch result consistent or inconsistent with those objectives? What alternatives did Congress consider and not adopt and how would those alternatives have affected the outcome of this issue? Since the 1977 legislation was designed in large part to remedy problems with the 1972 legislation, what was Congress trying to achieve in 1972 and to achieve better in 1977? What were the purposes of the 1977 transition clause, and how did it achieve, or fail to achieve, those purposes? What were the economic expectations underlying the 1977 amendments, and how did those expectations and the actual economic results affect the notch issue?

II THE LEGISLATIVE ROOTS OF THE 1972 AND 1977 BENEFITS FORMULA

Benefit Increases Prior to the Automatic Cost of Living Adjustments

When the Social Security programs was enacted in the Social Security Act of 1935, it provided for benefit payments only to workers in "commerce and industry" when they retired from employment at age 65 or later. The system then established has been expanded in the intervening years both in coverage and in the types of economic protection afforded to workers and their dependents.

The major subsequent laws which broadened the system included benefits for dependents and survivors of covered workers enacted in 1939; coverage for additional types of workers, largely in the 1950's, 1960's, and 1970's; benefits for disabled workers enacted in 1956; and the creation of the Medicare program, financed in part by employment taxes, enacted in 1965.

The automatic cost of living adjustments were enacted in 1972 following a three-year period of intense Congressional action dealing with proposed changes to a broad range of Social Security Act programs. These included: the enactment of the Supplemental Security Income (SSI) program which federalized the state administered adult public assistance programs; reforms to the Aid to Families with Dependent Children (AFDC) program including the Nixon Administration's proposed Family Assistance Plan, which failed of enactment; numerous Medicare and Medicaid amendments; and major amendments to the unemployment compensation program. During the same period, the House Ways and Means Committee and Senate Finance Committee were also occupied with major trade and tax legislation.

The focus of this section is limited to the history of Social Security benefit increases that preceded the benefit formula provisions of the 1977 amendments, which resulted in the enactment of the "notch." Only those prior amendments that had some relationship to benefit increases will be mentioned.

The automatic cost-of-living benefit adjustments enacted in 1972 were based, in principle, on the same approach used in designing all of the ad hoc benefit increases that were enacted in prior years.

Prior to the 1972 amendments "static" economic assumptions were used to determine the long-range future cost of the Social Security program. The application of these static assumptions was a deliberately conservative approach to financing the program. In brief, the static assumptions held that wages and prices would not increase in the future but would remain at the same level they had attained at the time of the estimate.(5)

(4)  The Social Security Act of 1935 also contained separate titles to create additional programs for the needy and the unemployed which have developed independently. References in this paper to the Social Security program are to the Old-Age, Survivors, and Disability Insurance program (OASDI) which, in the 1935 legislation, was the Old-Age Insurance program.

(5)  See discussion of financing methodology in U.S. House of Representatives, Reports of the 1971 Advisory Council on Social Security, House Document No. 92-80, Washington, U.S. Government Printing Office, pp. 64-66.

The application of static economic assumptions resulted in periodic surpluses in the Social Security trust funds as wages levels increased providing additional Social Security tax income to the program. These surpluses, augmented by legislated changes in Social Security (FICA) tax rates and the amount of annual earnings subject to tax (the wage base), were used to finance ad hoc benefit increases and other program liberalizations prior to the adoption of the automatics in the 1972 amendments. The following table of benefit increases and price increases indicates that, except for the period prior to 1950(6) and in the case of the 1973 Amendments(7), Congress enacted ad hoc benefit increases that exceeded increases in the CPI, prior to the first scheduled operation of the automatics in June 1975.

(6) Social Security benefits were not paid until January 1940

(7) The 1973 amendments provided earlier benefit increases in 1974 that would not have occurred under the automatics until June of 1975.

(1) 1967=100 CPI-W

(2) Average increase of about 77%--from 100% at lowest to 50% at highest level.

(3) Greater of 12.5% or $5

(4) Guarantee of 7% or $3

(5) Guarantee of 7% or $4

(6) 11% increase in benefits effective June 1994, with 7% of this amount payable March 1974. SOURCE: Social Security Administration

Following is a brief summary of legislative actions taken from 1969 through 1972 on ad hoc benefit increases and on the adoption of the automatic provisions

1. Legislative action in 1969

On September 25,1969, President Nixon sent his Social Security recommendations to Congress and they were introduced by Minority Leader Gerald Ford as H.R. 14080.(8)

The bill contained a 10% benefit increase and automatic benefit increases based on future increases in the cost-of-living plus additional amendments relating to benefit categories.

The House Ways and Means Committee started public hearings on September 30, 1969, on H.R. 14080, together with H.R. 14173, President Nixon's proposed welfare reform amendments. The hearings continued until November 13. The Committee went into executive session on November 19 to consider both the Social Security and the welfare reform proposals.

On December 8, the Ways and Means Committee unanimously reported a separate bill (H.R. 15095) which was introduced by Chairman Wilbur D. Mills and ranking Minority member John Byrnes on December 4. The bill provided a 15% benefit increase effective for January 1970 and numerous modifications to the President's proposals. The bill did not contain any financing amendments since the OASDI trust funds had a favorable actuarial balance of 1.16% of payroll.(9) Following floor debate, the House approved the bill by a 398 to 0 vote on December 15.

Meanwhile H.R. 13270, the-proposed Tax Reform Act of 1969, was being considered in the Senate. The Senate attached several Social Security amendments to the bill, including a 15% benefit increase effective for January 1970, and an increase in the minimum benefit from $55 to $100/month, and an increase in the taxable earnings base to $12,000 beginning in 1973. H.R. 13270 passed the Senate by a vote of 69 to 22 and it was sent to a Senate-House conference committee on December 11,1969, which considered the differences between the two bills. The conference committee agreed to the 15% benefit increase, approved by both Houses, omitted the Senate financing provisions and settled the differences on the other provisions of the bill. The conference committee bill was agreed to by both the House and the Senate on December 22 and the President signed the bill as Public Law 91-172 on December 30,1969.

2. Legislative action in 1970

The Ways and Means Committee carried on extensive executive sessions in the early months of 1970 and further considered the Administration's Social Security, Medicare, and Medicaid proposals. On May 11, 1970, Chairman Mills and Congressman Byrnes introduced H.R. 17550, containing the Committee's decisions. The bill contained an additional 5% benefit increase, effective for January 1971 but did not include the Administration's automatic benefit increase proposal.

H.R. 17550 also contained many additional amendments to the OASDI program and to the Medicare and Medicaid programs, along with financing provisions to maintain the fiscal integrity of the payroll-financed OASDI and Medicare (Hl) programs.

(8) Weekly Compilation of Presidential Documents, vol. 5, no. 39, pp. 1319-1324.

(9) The long-range solvency of the Social Security program is usually stated in terms of percentages of taxable payroll. Roughly speaking, a long-range deficit of l% of taxable payroll means that the actuarial projections of the income and outgo of the trust funds over the next 75 years show that outgo will exceed income and that income could be made to equal outgo by increasing the combined employer/employee tax rate by l percentage point (or by making other changes of an equivalent magnitude to the income or outgo of the program). Similarly a surplus of 1% of taxable payroll could be used to decrease the tax rates by a combined 1 percentage point or to make program changes with comparable costs.

While being debated in the House, H.R. 17550 was ordered recommitted with directions to report the bill back to the House with the Administration's recommendation for automatic adjustment of benefits, the wage base, and the retirement test. As so amended, the bill passed the House on May 21, 1970, by a vote of 344 to 42.

The Senate Committee on Finance began public hearings on H.R. 17550 and on H.R. 16311, the proposed welfare reform amendments on June 17 and went into executive session to consider the two bills on September 29. The Finance Committee bill was reported on December 9 with major modifications.

In place of the 5% benefit increase in the House passed bill, the Finance Committee bill provided a 10% increase. The minimum benefit would be increased to $100, rather than $67.20 which would have been payable under the 5% increase passed by the House. The Finance Committee bill also differed from the House bill with regard to financing the automatics. The House bill assumed that the program would be adequately financed from the additional revenues generated by automatic increases in the wage baseCthe maximum amount of annual earnings subject to Social Security tax. The Senate Finance committee bill would have required an actuarial evaluation of program finances each time a benefit increase was triggered. On the basis of that evaluation, program financing would be increased to the full extent necessary to cover the additional costs. One-half of those costs would be financed from increases in FICA tax rates and one-half from increases in the wage base.

Numerous other OASDI and HI amendments in the Finance Committee bill either were the same as modifications or deletions of, or additions to the provisions of the House passed bill.

The Finance Committee bill also would have created a new Federal program of catastrophic health insurance for all persons under age 65 who were insured for or entitled to Social Security, their spouses, and dependent children. The health services would have been the same as those provided under the Medicare program and would have been available after family health care expenses exceeded defined limits. Also included in the bill were provisions related to international trade and veterans benefits.

With regard to welfare changes, the Finance Committee considered H.R. 16311 the Administration's welfare reform proposal and, with major modifications, included its provisions in H.R. 17550. These modifications would have eliminated the House passed provisions to federalize the adult public assistance programs and establish a federal Family Assistance Plan in place of the State-run AFDC program, but the bill included a number of other modifications to the adult assistance programs and to the AFDC program.

During floor debate the Senate voted to recommit the bill to delete the catastrophic health insurance program and certain other provisions of the bill. As so amended, H.R. 17550 was passed by a vote of 81 to 0 on December 29.

The Senate requested a conference and appointed conferees. The House declined the request since Congress had set January 2, 1971, as final adjournment day. Chairman Mills declared his intention to make Social Security legislation the first order of business for the Ways and Means Committee in the about-to-convene 92nd Congress.

3. Legislative action in 1971

When the 92nd Congress convened on January 21, 1971, Chairman Mills and Representative Byrnes jointly introduced H.R. 1, the Social Security Amendments of 1971. The bill=s provisions were, for the most part, the same as those contained in H.R. 17550 as passed by the House in 1970 (but they did not include the automatic benefit increase provisions). The bill also included the welfare reform proposals passed by the House in H.R. 16311 in 1970.

The Committee on Ways and Means held executive sessions from February through May 1971 to consider H.R. 1.

During this same time, action was required on a bill to increase the public debt limit. When the public debt bill (H.R. 4690) was before the Senate, several Social Security amendments were adopted during floor debate. The principal amendment was a 10% Social Security benefit increase, including family maximum benefits, effective for January 1971. The conference committee on the debt limit bill deleted two of the Senate amendments (one providing for a $100 minimum and one providing for an increase in the retirement test (10) annual exempt amount to $1200), but accepted the remaining Senate amendments, including the 10% increase in benefits and family maximum benefits.(11)

The President signed H.R. 4690 into law (Public Law 92-5) on March 17, 1971.

Also during March 1971 the Advisory Council on Social Security(12), which had been appointed in 1969, issued its report. Its recommendations included most of the major changes in the OASDI program that were contained in H.R. 1. It also recommended changing from static to dynamic economic assumptions in estimating the costs of the OASDI program, and it endorsed the principle of current cost financing by which trust fund revenues and expenditures would generally be kept in balance with sufficient reserves to maintain funds to assure annual benefit payments.

The Ways and Means Committee reported H.R. 1 to the House on May 26, 1971. The reported bill provided a 5% Social Security benefit increase effective for June 1972. The OASDI, Medicare, and Medicaid provisions remained essentially patterned after H.R. 17550 but with additional modifications along the lines of Senate amendments to H.R. 17550 in 1970.

H.R. 1 as reported in the House also contained welfare reform provisions which differed considerably from those contained in H.R. 16311 of the previous Congress.

H.R. 1 passed the House by a vote of 288 to 132 on June 22 and was sent to the Senate. The Senate Finance Committee held public hearings in July and August of 1971 but took no further action on H.R. 1 during the remainder of 1971.

(10) The retirement test or earnings limit is the provision of law under which Social Security benefits are reduced when an individual to whom the test applies has earned income above a certain amount, referred to as the exempt amount. The test now generally applies to beneficiaries under age 70.

(11) It is of interest to note that in dealing with the l0% increase in the family maximum the amendment eliminated "notches" that had accrued for families affected by the family maximum under prior benefit increases. Such earlier benefit increases went only to families on the rolls on the effective date of a benefit increase, but not to families coming on the rolls in the future. This resulted because family maximum benefits were previously determined as a percentage of a worker's average monthly wage, which remained constant regardless of any benefit changes. The family maximum was, in effect, redefined by being stated as a percentage of a worker's primary insurance amount (PIA) which is the basic figure that is used to determine all individual benefits. As thus redefined, benefit increases were made applicable to family maximum benefits regardless of when a family became subject to the maximum. Since then, all benefit increases have applied to current and future families limited by the family maximum.

(12) Section 706 of the Social Security Act provided for the appointment every four years by the Secretary of Health Education and Welfare of a 13-member Advisory Council whose "members, shall to the extent possible, represent organizations of employers and employees in equal numbers and represent self-employed persons and the public." Legislation enacted in 1994 establishes a permanent advisory board and eliminates the requirement for quadrennial advisory councils.

4. Legislative action 1972

The Senate Finance Committee held additional public hearings on H.R. 1 in January and February of 1972 followed by executive sessions which lasted through June.

On February 23, 1972, while hearings on H.R. 1 were being conducted by the Senate Finance Committee, Chairman Mills introduced a bill (H.R. 13320) to provide a 20% Social Security benefit increase, effective for June 1972. The bill's financing provisions provided for increases in the wage base to $10,200 in 1972 and to $12,000 in 1973, and automatic increases in the wage base thereafter, and for reductions in the tax rates in the early years according to the following table showing rates for current law, H.R. 1 as passed by the House and for H.R. 13320.

(1) $9,000 contribution and benefit base for 1972 and after.

(2) $10,200 contribution and benefit base for 1972 with automatic adjustments to increases in earnings levels thereafter.

(3) $10,200 contribution and benefit base for 1972 and $12,000 for 1973 with automatic adjustments to increases in earnings levels thereafter.

In introducing H.R. 13320, Chairman Mills explained that it was possible to finance a 20% benefit increase and the provisions in H.R. 1 and at the same time lower tax rates during the early years by applying the dynamic economic assumptions and the current-cost financing that had been recommended by the Advisory Council on Social Security and endorsed by the boards of trustees of the Social Security trust funds and by the Nixon Administration.(13)

Mr. Mills' bill met with an uncharacteristic questioning of its soundness. Employer interests and the conservative press and Congressional critics expressed concerns relative to the risks the system would assume if the finely-tuned economic assumptions on which financing of the Mills' bill was based failed to occur.(14)

Action on the 20% benefit increase and the automatics followed an unusual and unpredictable path in the Senate. This legislative course bypassed consideration by the legislative committees in both Houses of the 20% increase and the shift from static to dynamic assumptions.

(13)  U.S. Congress, Congressional Record, (Permanent edition), pp. 5269-5272, Feb. 23, 1972.

(14)  See statement of Congressman Conable at U.S. Congress, Congressional Record, (Daily edition), p. H1429, Feb. 24, 1972.

When the Senate was debating a bill (H.R. 15390) to extend the public debt limit, Senator Frank Church on June 28, 1972, offered a floor amendment to the bill that was essentially the same as H.R. 13320 but with a later effective date for the 20% benefit increase (September rather than June) and a one year delay in the wage base increases (to $10,800 in 1973 rather than 1972, and to $12,000 in 1974 rather than 1973).

H.R. 15390, with these and other technical amendments, was passed by both the Senate and the House on June 30 and signed by the President on July 1, the date by which final action on the debt limit was mandatory. (Public Law 92-336, 92nd Cong.)

Such was the highly charged legislative environment in which the automatic (COLA) provisions were enacted.

The Senate Finance Committee resumed consideration of H.R. 1 in September 1972 and reported the bill on September 26 with numerous amendments to the Social Security, Medicare/Medicaid, and welfare programs.

The Senate debated H.R. 1 from September 26 to October 6, again adding additional amendments, and passed the bill by a vote of 68 to 5.

The conference committee met on the bill on October 10 through October 14. The conference report was adopted by the House on October 17 by a vote of 305 to one and on the same day by the Senate by a vote of 61 to zero. On October 30, 1972, the President signed the bill into law as Public Law 92-603.

III DEFECTS IN THE AUTOMATICS APPEAR

Actuarial Status in 1972

In 1972, when the automatic benefit increase provisions were adopted, inflation rates seemed to be declining from the unusually high levels (averaging 5% annually) of the past five years. For 1972, the inflation rate was 3.2%.(15) Near the end of 1972, the Finance Committee reported that the Social Security program of Old-age, Survivors, and Disability Insurance had a positive actuarial balance. The future cost of the system assumed an average annual inflation rate of 2.75% and a wage growth rate of 5%. For the first 37 years (through 2010), a safety factor of 0.375% was added so that the assumed real wage differential was 1.875% (5% wage growth minus 2.75% inflation. minus 0.375% safety factor). The first two automatic benefit increases were projected to take place in 1975 (5.1%) and 1977 (5.5%). (16)

1973 - Unexpected Inflation

By early 1973, it was clear that inflation would be much higher than had been expected in 1972. Instead of waiting for the first automatic increase to take effect in January of 1975, Congress passed Public Law 93-66 in July 1973 providing an ad hoc increase of 5.9% to become effective in June of 1974. This was a "down payment" on the January 1975 automatic increase which was then expected to be about 8%.

As inflation continued to mount, Congress again passed legislation in the closing days of the session to provide a two-step increase - an 11% increase effective for June 1974, 7% of which was payable in April. The previously adopted 5.9% increase was absorbed into this increase.

(15) Economic Report of the President, U.S. Government Printing Of lice, Washington, 1992., p. 361.

(16) Committee on Finance, U.S. Senate, Social Security Amendments of 1972 (Senate Report 92-1230 to accompany H.R. 1), U.S. Government Printing Of lice, Washington, 1972, pp. 341-344. Under the 1972 law, inflation would have to increase by at least 3 % since the last increase in order to trigger an automatic benefit adjustment. Under subsequent amendments, there is now an annual increase without regard to any such minimum.

Action on this legislation began with the introduction of H.R. 11333. The bill was favorably reported by the House Ways and Means Committee on November 9, 1973 (House Report No. 93-627). The bill, as it was reported and as it passed the House on November 15, provided for a two-step increase. A "flat" 7% payable for March 1974 and another increase payable for June 1974 which, together with the 7% increase, would equal a combined 11% increase.

The Senate acted on this increase as an amendment to H.R. 3153, a lengthy technical amendments bill to eliminate minor drafting errors in the Social Security law. The 1974 two-step benefit increase in the Senate amendment differed from the House provision in H.R. 11333 in three respects: first, the interim 7% increase would be for the month in which the bill was enacted rather than for March 1974 as in H.R. 11333; second, additional financing was provided in the form of increases in Social Security tax rates and in the taxable wage base; third, the interim 7% increase would be a "refined" increase rather than a "flat" increase as in H.R. 11333.(17)

A conference committee began to meet on H.R. 3153 in late December but no conference report was issued. Instead, by unanimous consent, floor action was taken in the House on December 20 and in the Senate on December 21. Both houses approved H.R. 11333 with the two-step 11% increase along the lines of the Senate amendment to H.R. 3153. The bill was signed on December 31, 1973, as Public Law 92-233.

In its report on the legislation which would become Public Law 92-233, the Senate Finance Committee indicated that it would leave the Social Security program with a long-range actuarial deficit of 0.56% of taxable payroll and a short-range situation in which end-of-year balances would grow from $44 billion in 1973 to $52 billion in 1978. (Although the Committee report did not point this out, the size of the short range fund balances relative to annual outgo were projected to decline over that period.) (18)

1974 Trustees Report Indicates a Problem

While the Finance Committee report in 1973 showed no great concern over the financial status of the program, the situation changed dramatically in 1974. Inflation, which, in the years since 1950, had never reached as much as 5% came in at 6.2 % for 1973 and continued to accelerate to nearly twice that rate in 1974. When the Social Security Board of Trustees(19) issued its annual report on the financial status of the program at the end of May, it revealed a startling actuarial imbalance of 2.98 % of taxable payroll. For the program to meet its benefit obligations over the 75-year estimating period, tax rates would have to be increased by some 27 percent. (20)

(17) Briefly, a "flat" increase involves applying the percentage increase to actual benefit amounts paid to beneficiaries rather than to the primary insurance amount (PIA) on which benefits are normally based and on which a "refined" benefit increase is based. The use of the former gives rise to more difficult administrative problems than the latter.

(18) Committee on Finance, U.S. Senate, Social Security Amendments of 1973 (Senate Report 93-553 to accompany H.R. 3153), U.S. Government Printing Office, Washington, 1972, pp. 10-16

(19) By law, the OASI and Dl trust funds each have a board who are required to make an annual report on the status of the funds. In 1974, the Board of Trustees was composed of the Secretary of Treasury, the Secretary of Health, Education, and Welfare, and the Secretary of Labor, with the Commissioner of Social Security acting as secretary to the Board.

(20) U.S. House of Representatives, 1974 Annual Report of the Board of Trustees of the Federal Old-Age and Survivors Insurance and Disability Insurance Trust Funds, House Document 93-313, Washington, U.S. Government Printing Office June 3, 1974, p. 36. The primary reason assigned by the trustees for the worsening of the financial condition of the funds was a change in population projections.

First Hsiao Panel Confirms a Very Severe Financing Problem

The Senate Committee on Finance responded to the Trustees' report by recommending a Senate Resolution calling for the appointment of a panel of actuaries and economists to provide a second opinion or, in the words of the resolution, "an expert independent analysis of the actuarial status of the Social Security system." The resolution was adopted by the Senate In of 1974, and a panel, headed by William C. L. Hsiao was convened.(21)

The Hsiao panel issued its report in January of 1975. The panel not only confirmed the Trustees' findings that the program was badly out of actuarial balance but also concluded that the deficit was roughly twice as bad as indicated in the Trustees' report. Where the trustees had indicated that tax rate increases of 27% would be needed for the program to meet its benefit obligations over the long-range estimating period, the Hsiao panel found that restoring solvency would require a 55% increase in tax rates (or other changes to benefits and financing with an equivalent impact).(22)

The Hsiao panel attributed the long-range financing problems of the Social Security program about equally to two main causes: demographics (a lower fertility rate than had previously been projected so that there would be fewer workers to pay taxes to support the program) and the nature of the benefit formula. The panel also examined the various factors which have an impact on program solvency and found that those factors with the highest impact (fertility, wage patterns, and inflation) were also the factors that are hardest to predict accurately.(23)

With respect to the benefit formula, the panel found that the formula adopted in 1972 "responds irrationally to changes in the rate of inflation, and can produce patterns of replacement ratios inconsistent with the generally understood purpose of the Social Security system." The panel recommended "That the benefit structure be changed to eliminate its irrational response to changes in the rate of inflation. This is essential to achieve financial soundness. The first step should be a prompt thorough study of several possible changes in the benefit structure."(24)

1974 Advisory Council Recommends Wage Indexed Formula

A new Social Security advisory council had been appointed to meet in 1974 under the chairmanship of W. Allen Wallis, Chancellor of the University of Rochester. The advisory council devoted its primary attention to the need to correct problems with the benefit formula adopted in 1972. Its report issued in March 1975 noted that "the method used for automatic cost-of-living adjustments has the side-effect of making replacement ratios (benefits as a proportion of earnings just before retirement) subject to unpredictable variations caused by changes in wage and price levels, an effect that presumably was not intended."(25)

(21) Committee on Finance, U. S. Senate, Report of the Panel on Social Security Financing pursuant to S. Res. 350, 93rd Congress. Committee Print, U.S. Government Printing Office, Washington, 1975, p. 1 (cited hereafter as Hsiao I).

(22) Hsiao I., p. 2.

(23) Hsiao I., pp. 2-6.

(24) Hsiao I., pp. 3-4.

(25) U.S. Congress, Reports of the Quadrennial Advisory Council on Social Security, House Document 94-75, U.S. Government Printing Office, Washington, 1975, p. l 30. Hereafter cited as Advisory council (74). Note: This quotation and some of the others in this paper come from an appendix to the advisory council report which was prepared by a panel of consultants to the council. Since the Advisory Council chose to incorporate their finding in its report and indicated no disagreement with them, this paper treats that appendix as part of the report.

The Hsiao panel had characterized the problem with the 1972 automatic benefit adjustment mechanism as one of "overindexing."(26) The advisory council report described it as a "coupled" system.(27) Both of these terms were widely used in discussions and documents at that time. They referred to the fact that initial benefit levels increased from year to year as a result of two separate indexing mechanisms. Initial benefit levels were computed by applying a formula to the retiree's average wages under Social Security. Under the 1972 legislation, the factors in the formula were periodically increased by the percentage increase in the Consumer Price Index.(28) The average wages to which this formula was applied also tended to increase as wage levels in the economy grew from year to year (and the 1972 legislation facilitated this impact by indexing for wage growth the maximum amount of annual earnings that could be counted in determining the average wages to which the benefit formula was applied). Thus, initial benefit levels were automatically increased by a mechanism which "coupled" the impact of price growth through an explicit indexing of the benefit formula and the impact of wage growth through the use of average wages. This "coupling" of the two factors made the increase in year to year benefit levels extremely difficult to predict since the increase would be based on both the absolute values and the interrelationship of inflation and wage growth. The result, under the revised economic assumptions of the mid-1970's was that benefit levels were increasing more rapidly than the financing of the system could sustain. Consequently, the "coupled" system adopted in 1972 was resulting in "overindexing" of benefits.(29)

The advisory council recommended that the situation be corrected by replacing the "coupled" mechanism for increasing initial benefit levels with a "decoupled" system which would rely entirely on wage indexing. Under the advisory council approach, the 1972 system in which the percentages in the benefit formula were indexed to the CPI each year would be dropped. A new formula would be adopted in which the percentage factors would not change from year to year. Instead of indexing the formula for price inflation, the new mechanism would index the wages to which the formula was applied. A retiree's creditable wages for each year would be adjusted to reflect wage growth in the economy between the year in which they were earned and the year of retirement. The benefit formula would then be applied to the average of those indexed wages.(30) Once individuals had their initial benefit levels computed at retirement, those benefits would be kept up to date through price indexing.

Although differing in some details, the wage indexing approach recommended by the advisory council, was essentially the approach ultimately enacted in the 1977 amendments.(31)

(26) Hsiao I., p. 17

(27) Advisory council (74), pp. 130-131.

(28) Strictly speaking there was no actual formula. Rather the law contained a benefit table which prescribed the full-rate benefit payable for workers at each level of average earnings. When benefits increased by a given percentage, all the benefit amounts in the table were increased by that percentage and, if the law also increased taxable wages, the table was

extended to prescribe benefit amounts at the resulting higher average wage levels. For all practical purposes, however, the result was the same as if there had been a benefit formula and all the factors in the formula were increased by the specified percentage.

(29) There was no "official" definition of the term "coupling," and it was somewhat confusingly used in two different meanings. As described here, it referred to the fact that the mechanism for adjusting initial benefits combined elements of price indexing with elements of wage indexing. It was also sometimes used in reference to the fact that the 1972 law used

the same mechanism--price indexed benefit formula factors--to increase initial benefit levels and to adjust benefits for those already on the benefit rolls. (A proposal described as "simple decoupling" would have eliminated price indexing for the factors in the benefit formula as it related to initial benefit levels but continued price indexing as it applied once an individual was on the benefit rolls.)

(30) The Social Security program weights the benefit formula to the advantage of lower income individuals. The benefit formula provides a higher percentage return on the lower part of an individual's average earnings than on the higher part. The "bend points" -- the levels of earnings at which the percentage in the formula changes -- were also to be indexed for wage growth in the economy under the advisory council proposal. Advisory council (74), p. 17.

(31) The advisory council proposed formula was 100 % of the first "A" dollars of average indexed monthly earnings (AIME) plus "B" % of AIME above "A" dollars. Instead of this 2-step formula, the actual 1977 legislation created a 3-step formula starting with 90% of the first "A" dollars. Advisory council (74), p. 17.

The question of how the change from the old to the new system would affect benefit levels in the years following the changeover was addressed in the advisory council report to some extent:

THE PHASING-IN PROCESS

The consultants have no intention of reducing benefits when the new formula produces a lower result than the old one. It is clearly important that the new beneficiary of year y actually receive the greater of the new-formula PIA(32) and the old-system PIA. For those becoming beneficiaries after year y the same comparison is to be made and the greater benefit granted; with the understanding, however, that the old-system benefit table will not be updated for CPI changes after year y. As time goes on, the new-formula result (dynamic with average wages) will be greater than the PIA from the old system (static at the year y level) for a larger and larger percentage of the new beneficiaries, and the old system will be slowly phased out. The young-age death and disability cases will likely be the last to be payable on the new formula.(33)

In connection with the notch issue, there are several important points to be noted about the above quotation from the 1974 advisory council report.

The phase in described by the advisory council report matches the transition provision actually adopted in the 1977 legislation with 2 important exceptions. The actually enacted transition provision prohibits the use of wages earned in or after the year of reaching age 62 in computing a benefit under the old law formula. The benefit level differentials which give rise to the notch issue would have been less striking if the transition clause had permitted the use of post-age-61 wages.(34)

While the advisory council phase-in clearly anticipated that there would be no differential between old-law and new-law benefits for individuals retiring in the year the change became effective ("year y@), the above quotation shows that the report did recognize that the old-law benefits would for some time be higher for some beneficiaries than what would be payable under the new system. This finding, however, relates solely to a comparison of what the same individual would qualify for under the two approaches. There is no indication that the advisory council was aware of or addressed in any way the actual notch issue; that is, a differential in actual benefit levels payable to individuals subject to the new system and individuals who continued to be eligible under the prior system.

A second notable difference is that the Advisory Council Transition would have been available to all future retirees rather than only to those reaching retirement age in the first five years of the new formula. As the Council pointed out, however, the new formula would increasingly provide higher benefits than the static transition formula.

(32) The term "PIA" is an abbreviation for "Primary Insurance Amount". This is the "basic" benefit amount for an insured worker. For a retiree, it represents the amount that would be payable if the individual began receiving benefits at age 65. The actual benefit payable may differ, for example, because of reduction for early retirement.

(33) Advisory council (74), p. 130.

(34) For persons born in 1917, allowing post-6l wages to be used in an old-law computation would have eliminated any notch effect since such individuals were not affected by the other part of the notch, the rule that cost-of-living increases would not apply after 1978 and before the year of attaining age 62.

Second Hsiao Panel Recommends Price Indexed Approach

In April of 1975, the Congressional Research Service of the Library of Congress, on the basis of a request from the Senate Finance Committee and the House Ways and Means Committee, appointed a consultant panel of actuaries and economists to "examine the various ways in which the benefit structure could be revised to correct the problem of any overreaction to changes in price levels."(35) William Hsiao once again was selected as project director for this study. The panel submitted its report to the Congressional Research Service in April of 1976. The report was subsequently printed as a joint committee print in August, 1976. By the time the report was printed, the Ford administration had submitted legislation proposing a switch to a wage indexed mechanism along the lines of the recommendations made by the 1974 advisory council.

The Hsiao II panel looked at 5 alternatives to the 1972 benefit adjustment mechanism: a flat benefit formula, a money purchase plan, a "high-5" plan, a wage-indexed formula, and a price-indexed formula.(36) For the most part, however, the report concentrated on the price and wage indexing alternatives. The panel recommended a price indexing approach. This would have operated in the same way as the wage indexing proposals except that wages would have been adjusted for price inflation between the year of earnings and retirement instead of being adjusted for wage growth.

The price indexing approach recommended by the Hsiao II panel was expected to result in lower program costs than wage indexing.(37) The panel argued that price indexing would maintain the real, inflation-adjusted value of benefit levels while preserving a greater degree of control and flexibility for Congress to determine whether benefit increases in excess of inflation were appropriate. A second argument was that the price indexing approach involved a substantially lower long-range cost. The panel's third argument was that the price indexing approach would allow Congress to provide real increases in benefits for those already on the rolls:

• In contrast to this, the wage-indexing method provides a sharp tilt in favor of workers retiring in the future. The increases in benefits for workers already retired are limited to increases in the rise in the Consumer Price Index. Yet workers who retire five years later will receive increments due to both price changes and increases in real wages. The difference in retirement benefits can be substantial.(38)

This last argument corresponds to that part of the notch issue which is based on the contention that those in the notch years received benefits which were lower than those paid to later retirees.

The Hsiao II panel also dealt with the issue of transition. They recommended that persons born in 1917 or earlier remain subject to the old law regardless of when they retired. For those born during the 4 transition years 1918-1921, the panel recommended the payment of a blended benefit comprised of an increasing percentage (20,40,60, and 80 percent) of the new law benefit and a decreasing percentage of the old-law benefit. Like the advisory council, the Hsiao II panel did not directly address the question of discrepancy between benefits for those born before 1918 and those born after 1917.

(35) Committee on Finance, U.S. Senate, and Committee on Ways and Means, U.S. House of Representatives, Report of the Consultant Panel on Social Security to the Congressional Research Service. Joint Committee Print, U.S. Government Printing Office, Washington, 1976, p. 1. (cited hereafter as Hsiao II)

(36) Hsiao II, p. 15.

(37) As it turned out, during the first few years during which the 1977 amendments were effective, price inflation outpaced wage growth so that costs of price indexing during that period would have been greater than the costs of the wage indexing system which was enacted. Over the long run, however, the price indexing approach would have cost significantly less.

(38) Hsiao II, p. 9.

IV LEGISLATIVE ACTION

As described under the preceding heading, during the period between the enactment of the 1973 ad hoc increase and the 1977 amendments, concerns over the financial integrity of the Social Security system mounted. Although no legislation to deal frontally