Since the early 1980s, the federal government has been aware of the concerns of several million older Americans who believe they are being paid less than their fair share of Social Security benefits as a result of Congressional actions taken in the 1970s.

The "Notch" issueCso named because of a v-shaped dip in a graph representing these seniors= benefit levelsChas been the focus of Congressional hearings, study panels, and oversight groups. Volumes of detailed analyses have been written about it,(1) and more than 100 legislative bills addressing it have been introduced in both houses of Congress.(2) Yet despite all the attention and interest, no action has been taken, and much of the proposed legislation involves significant costs.(3)

Some national organizations representing older citizens (the National Committee to Preserve Social Security and Medicare, for instance) have strongly supported legislation to raise benefits for those born in the "Notch" years; other such groups (the American Association of Retired Persons, for example) have argued that those born in the "Notch" years received intended and appropriate benefits, and that no corrective legislation is needed.

In 1992, Congress established the bipartisan Commission on the Social Security "Notch" Issue, charging it with examining the question of whether those born in the "Notch" years had been treated unfairly and recommending, if necessary, remedial legislation and the means to pay for it.(4)

At the outset of its work, the Commission arrived at two basic understandings. First, it realized that, despite the current size of its reserve fund, the Social Security system faces serious long-range fiscal issues.(5) Second, it was keenly aware of the size of the Federal budget deficit. The Commission, therefore, approached its mandate with an explicit understanding that, if it were to recommend remedial action regarding the "Notch" issue, it would not recommend financing it through an invasion of the Social Security trust funds or any use of general revenues. The Commission concluded that it would have to recommend financing any changes with an increase in Social Security taxes or a reduction in some benefitsCand it was fully prepared to do so if a remedy was justified.

The Commission is composed of 12 members who have long experience in evaluating issues of public policy.(6) Many of them have a substantial background of involvement with, and knowledge of, the Social Security system. Its members had the benefit of analysis and advice from many experts on Social Security, both from inside and outside of government. Between April and December 1994, the Commission met seven times (including three public hearings), heard a broad and representative range of testimony, commissioned numerous technical analyses, and spent hundreds of hours exhaustively reviewing the issue. The results of that work follows.

For more than a decade, many older Americans born between 1917 and 1921 have been expressing concern that they have been denied the benefits they deserve as a result of changes Congress made to Social Security in the 1970s. As evidence, they point to the fact that, after taking inflation into account, their benefits are lower than those for persons born both before and after them. Indeed, when displayed on a vertical bar graph, those benefit levels form a kind of v-shaped "notch," dropping sharply from the left then rising again to the right (see Chart 6).

This so-called "Notch issue" has its origins in 1972, when Congress decided to create automatic cost-of-living adjustments to help Social Security benefits keep pace with inflation. Previously, each adjustment had to await legislation, causing beneficiaries' monthly payments to lag behind inflation.

Unfortunately, this new benefit adjustment method was flawed. To function properly, it required that the economy behave in much the same fashion that it had in the 1950s and 1960s, with annual wage increases outpacing prices, and inflation remaining relatively low.

A Revised Computation

That didn't happen. The rapid inflation and high unemployment of the 1970s generated sharp increases in benefits, and millions of seniors born after 1910 began receiving payments far in excess of what Congress had envisioned. Meanwhile, the sluggish economy failed to generate the tax dollars needed to fund those benefit payments, causing experts to express grave concerns about the fiscal stability of Social Security, since the rising benefit levels could not be sustained.

In an effort to end this problem, Congress revised the way that benefits were computed. In making its revisions, Congress decided that it was not proper to reduce benefits for persons already receiving them; it did, however, decide that benefits for all future retirees should be reduced. As a result, those born after January 1, 1917 would, by design, receive benefits that were, in many cases, far less generous.

In an attempt to ease the transition to the new, lower benefit levels, Congress designed a special "transitional computation method" for use by beneficiaries born between 1917 and 1921.

The "Notch" Issue Appears

Some born in the "Notch" years received benefits that were equal to or higher than those paid to beneficiaries born before them, while others received benefits that were higher than those generated by the new (1977) method.(7) Typically, these people retired at 62, which is common under the Social Security program.

However, some of those in the "Notch" yearsCparticularly those who continued to work well beyond age 62Creceived benefits that were significantly lower than they would have been if calculated under the old law. After comparing their benefit checks against the larger checks of their pre-"Notch" colleagues, neighbors, and friends with similar employment records, they began expressing their dissatisfaction to public officialsCand the "Notch" issue was born.

"Notch" year beneficiaries' principal argument is that they have been singled out to receive lower benefits. But the Commission found no evidence to support that position; indeed, the purpose of the 1977 legislation was to reduce benefits for all future beneficiaries, and it has generally done that.

In fact, considering the value of their benefits relative to the Social Security taxes which they paid, those born in the "Notch" years are, in general, receiving a greater return from Social Security than will subsequent generations of beneficiaries. In addition, their "replacement rate"Cthe percentage of pre-retirement earnings replaced by benefit paymentsCis equal to that of retirees who follow them, which was also the intent of the 1977 amendments. In this sense they are "doing well" as beneficiaries of the system, although not as "well" as those who came before them,(8) especially those who worked well beyond 62.

To the extent that disparities in benefit levels do exist, they exist not because those born in the "Notch" years received less than their due; they exist because those born before the "Notch" years (who were "grandfathered" under the old law's more generous computational method) continue to receive substantially inflated benefits. This disparity has created an understandable perception of unfairness.

The Commission's Work

In its work, the Commission has heard from hundreds of people who sincerely believe they are adversely affected by the "Notch" issue, and who understandably believe that they are entitled to legislative relief.

Regretfully for them, the facts suggest otherwise. The Commission's principal conclusion, therefore, is that the "Notch" is a necessary and appropriate result of the 1977 legislation which was designed to substantially reduce the growth of future benefit costs and to restore fiscal balance to Social Security, and no legislative remedy is in order. The Commission does not believe that benefit increases for people born after January 1, 1917 are appropriate, or that they can be justified.

The Commission did consider the possibility of recommending a reduction in benefitsCor a reduction in the growth in benefitsCfor those born before 1917 in order to reduce the disparity between benefits. Had Congress taken such an action in 1977, the differences in benefits generated by the old law and the new would have been smaller.

However, upon careful review, the Commission decided that reducing benefits, or benefit increases, for beneficiaries now between the ages of 78 and 84Cmost of whom have been retired and drawing benefits for a decade or moreCwould not be an appropriate response to the concerns of those in the "Notch" who feel that their own benefits are too low.

 For more than a decade, a group of older Americans has been petitioning their elected officials to correct what they believe is an injustice which they have encountered regarding their Social Security benefits. Simply put, these beneficiaries - all now between the ages of 73 and 77, all born between 1917 and 1921(9)Care convinced that they are receiving less than their fair share of Social Security benefits as a result of changes Congress made to the system in 1977.

As evidence, they point to the Social Security Administration's own statistics that demonstrate that many of these six million or so retirees are receiving benefit payments that are, on the whole, lower than those of other beneficiaries who share similar work histories but were born either before or after them. And, when those benefit levels are displayed on a vertical bar graph, they do, in fact, drop swiftly, then move upward again, creating a kind of v-shaped "notch" which has come to represent what is now known as "the Notch issue" (see Chart 6).

How did this happen?

The Problem of Ad Hoc Increases

The Social Security program is not static; Congress periodically reviews and revises it. Major changes even took place between the time of the program's inception in 1935 and the day when the first monthly benefit payments were made in 1940.

Originally, Social Security was to provide monthly benefits only to aged individuals who had retired from jobs in commerce and industry. Subsequent amendments added new beneficiary categories, including spouses and children of retired workers; survivors of deceased workers; and disabled workers and their spouses and children.

In addition, coverage has been broadened to include new categories of employees, farmers and other self-employed people, and public and non-profit employees. Periodic amendments have also changed rules of eligibility, such as the amount of time worked that is needed to qualify for benefits, the age of eligibility, and how post-retirement work affects benefits.

Most important to the issue before the Commission, however, are the frequent changes that Congress has made to the rules that determine the amount of monthly benefits. Some of these changes were made to accommodate other modifications in the program;(10) most were made, however, to offset inflation's erosion of the purchasing power of benefit payments.

Before 1972, benefit amounts were determined by using a table(11) that related basic benefit amounts to average "covered" wages.(12) Benefits determined from this table would not change unless Congress amended the law.

But inflation routinely ate into the real value of these static benefits. Therefore, from time to time, Congress would revise upward the benefit amounts in the table.(13) These periodic ad hoc increases both restored lost purchasing power for existing beneficiaries, and provided higher initial benefits for new beneficiaries.

Though not widely noted at the time, these new retirees' benefit levels tended to be significantly higher than those of existing beneficiaries. Why? The answer lies in wage growth.

Since both productivity increases and inflation naturally increase wage levels over time, new beneficiaries reach retirement age with higher average wage histories than workers with similar careers who retired earlier. Since the Social Security Administration bases benefits on a worker's wage history, these new retirees qualified for higher benefits as a result of this wage growth. But they also realized an increase whenever Congress increased benefits for those already on the rolls in order to offset inflation.(14)

A Change in Financing

In the early 1960s, inflation was quite modest, averaging somewhat less than 1.5 percent per year. By the late 1960s, however, the rate had more than tripled, to about five percent per year.

In response, Congress acted more frequently than before to provide ad hoc benefit increases.(15) Some observers were alarmed that those increases outstripped inflation; others, meanwhile, were disturbed that retirees' benefits were being eroded while awaiting action by Congress. These two very different concerns coalesced into support for a law change that would provide automatic annual benefit increases determined by an Ainflation index@ based on the Consumer Price Index, or CPI.

With encouragement and support from the executive branch and major interest groups, Congress passed legislation creating such a system in 1972. The new law adjusted benefits annually if the CPI rose three percent or more since the last benefit increase.

Unfortunately, the new law was flawed in that it simply adjusted benefits using the same method Congress had employed when creating previous ad hoc increases. Again, this meant higher benefits not only for those already on the rolls, but for new beneficiaries, who were also benefiting from the growth in their wages. In sum, the new law would result in raising real benefits over time for successive retirees.

The financing of these increases, however, was quite different from the old ad hoc approach.

Financing previously had relied on the safe but unrealistic assumptions that (1) there would be no further benefit increases, and (2) wage levels would not rise. When wage levels did rise, this produced "unanticipated" tax resources, creating a surplus. Aware of how much of a benefit increase it could "afford," Congress then would provide another ad hoc benefit increase, adjusting upward the amount of wages to be taxed in order to take full advantage of increasing wage levels.

Under the new law, benefit increases would occur automatically, as would increases in maximum taxable wages. This meant Congress had to provide financing for them in advanceCi.e., before knowing how much wages would increase. Congress therefore would have to rely on economic projections to determine whether the increased revenues produced by rising wages would be sufficient to meet the increased cost of rising benefits.

In other words, Social Security could no longer use static (and safe) economic projections; it would have to depend on more risky, dynamic projections of future wage and price levels.(16) It would also have to rely on Congress' willingness to act quickly to raise taxes in order to meet the cost of benefit increases if wage growth proved inadequate.

Congress could scarcely have picked a worse moment to move away from the previous, more cautious financing system. The economy defied all expectations in the 1970s, resulting in serious consequences for the financial viability of Social Security.

Benefits Threaten to Exceed Revenues

How could the economy so drastically affect Social Security?

Three economic factors directly affect the financial stability of the Social Security Trust Funds: inflation, wage growth, and employment. As Chart 1 shows, these three factors performed relatively consistently in the 1950s and 1960s, but changed drastically in the 1970s.

Inflation, which drives benefit costs, increased significantly in the late 1960s. But instead of returning to what was then viewed as the more "normal" level of the 1950s and early 1960s, inflation soared in the 1970s, ultimately reaching double digits and averaging close to eight percent over that period.

Typically, high inflation is accompanied by low levels of unemployment. But the 1970s proved to be a decade of "stagflation" during which both inflation and unemployment were high.

This peculiar economic circumstance undermined the financing of Social Security in three important ways: (1) It slowed the flow of taxes into the Trust Funds; (2) it produced rapid increases in benefits, both for existing beneficiaries and for newly retiring workers; and (3) it further increased the amount of benefits paid as older unemployed workers took early retirements.(17)

What happened with wage growth? In general, higher average wages are good for Social Security, because more taxes flow into the Trust Funds. But it is not the absolute level of wage growth that is important; it is the relationship of wage growth to inflation that is critical.

As Chart 1 shows, wage growth in the 1950s was more than double that of inflation. In the 1960s, wage growth again outstripped inflation by a wide margin. In the 1970s, however, it actually fell short of inflation.

As inflation drives up benefit costs, wages (and the taxes thereon) must grow to pay for benefits. The adequacy of financing for the 1972 law relied on the assumption that wage growth would continue at a rate roughly twice that of inflation. When wage growth failed even to match inflation in the 1970s, it resulted in a badly under-financed Social Security program.

Unfortunately, this was not a temporary aberration. While inflation was expected to decline, actuaries were no longer willing to make long-range projections based on the assumption that wage growth would outstrip inflation as it had in the 1950s and 1960s. Under these conditions, the method of calculating initial benefits based on a combination of wage growth and inflation inevitably resulted in benefits growing faster than revenues.

For those already on the rolls, the new law worked as expected: Changes in benefit levels reflected inflation. For new beneficiaries, however, initial benefit levels grew faster than wage levels. As a result, the percentage of wages replaced by Social Security benefits would increase from year to year, eventually rising to the point that many new beneficiaries would receive benefits higher than their earnings before retirement - a condition that is as unreasonable as it is unsustainable.(18)

The Problem is Addressed

The bad economic news was quick to arrive. At the time that the 1972 amendments were passed, Congress intended the automatic benefit increase mechanism to be triggered in 1975. But high inflation caused Congress to enact an additional ad hoc increase during 1973, thereby making the need for change evident even sooner. By early 1974, the Board of Trustees of the Social Security Trust Funds reported that the program had a very serious long-range deficit.

Appointed in 1974, the statutory Advisory Council on Social Security devoted its attention primarily to these financial problems. Also, two panels of actuarial and economic experts were convened by the House Ways and Means Committee and Senate Finance Committee to assess the magnitude of the problem and to propose solutions.

While opinions differed regarding the best solution, there was widespread agreement that any solution must include a major revision of the method for determining initial benefits. In addition, it was agreed that measures would have to be found to address other factors such as declining fertility rates, increasing longevity, and an increasing incidence of disability that were so negatively affecting the long-term integrity of the system.

How serious was the situation? By the mid-1970s, Social Security's long-range costs were projected to exceed income by more than 50 percent. Without corrective action, Social Security's funds would be exhausted, and the system would be unable to meet its statutory obligations by the early 1980s.

The Solution: Wage Indexing

By 1976, the need for substantial revisions in the Social Security program and its financing was overwhelmingly clear.(19) In June, the Ford Administration submitted to Congress a proposal to create a new method for determining initial benefits(20) based on an approach called "wage indexing," a method which adjusts a worker's wages to reflect economy-wide changes in wages over his or her lifetime.'' Hearings were held on this proposal, but Congress adjourned for the upcoming Presidential elections before completing a full review.

The new administration of President Carter sent its proposals to Congress in May 1977. Its package included the same "wage indexing" solution proposed by the Ford Administration as well as many new tax and financing proposals.

Despite the two Administrations' support for wage indexing, Congress examined numerous alternative proposals!! in a lengthy series of hearings. Ultimately, both the House and the Senate adopted legislation replacing the flawed 1972 method with a wage indexing method, and President Carter signed these new Social Security amendments into law on December 20, 1977.

These new amendments preserved the way that benefits were adjusted for inflation for those already on the rollsCin other words, existing beneficiaries continued to receive annual increases (COLAs) based on the percentage increase in the CPI. The way initial benefits were calculated. however, was completely revised.

Under the old law, a person's initial benefit was determined by averaging the actual wages he or she earned (in "covered" jobs) over a period roughly equivalent to a working lifetime. A benefit table was then used to determine the basic amount payable.

But since earnings levels in the economy tend to increase each year, initial benefits tended to creep up as the worker's average earnings rose. In addition, these benefits were also "price-indexed" -adjusted for inflation - since the figures in the table rose by the percentage increase in the CPI.

Fixed Formula Introduces Wage Indexing

Thus the old law generated, under some economic conditions, inflated initial benefits by linking, or "coupling," the effect of both wage and price increases. The 1977 legislation "de-coupled"(23) those two elements, substituting a fixed formula for determining initial benefits:(24)

• 90 percent of the lowest range of average indexed monthly earnings, plus
• 32 percent of the mid range of such earnings, plus
• 15 percent of the highest range of such earnings (up to a maximum based on amount of earnings on which taxes are paid).

Like the old approach, this new approach used average earnings over a "working lifetime." But those earnings would now be adjusted ("indexed") to reflect the growth of wages in the economy Cin other words, past wages would be translated into their equivalent in current wage levels.(25)

By adopting this new method, Congress purposely lowered initial benefits to offset the unintended increases that would have occurred as a result of the flawed 1972 method. However, it protected anyone who reached eligibility age prior to 1979Cthat is, anyone born before January 2, 1917 Cby "grandfathering" them under the old law. This protected people already on the benefit rolls as well as those who could have retired in 1978 or earlier but continued working. For those who continued working, the initial benefit calculations resulting from this grandfathering proved especially generous.

Thus a worker retiring under the new law would generally receive lower benefits than a worker I retiring under the old law, which was the intent of Congress. To minimize the abruptness of this change, however, Congress created a special five-year "transitional method" for people who would become eligible for benefits beginning in 1979.(26) In other words, those born between 1917-21 would I be the first to have their benefits calculated under the new law. This "transitional method" was designed to ease their transition to the new, lower level of benefits.

The transitional method was identical to the old method except (1) earnings after age 61 could not be used in figuring benefits, and (2) after 1978, no inflation adjustments would be made until age 62.(27) Individuals eligible for the transitional method would have their benefits computed under the new law method if it produced higher benefits.

The transitional method did not alterCnor was it intended to alterCthe fact that people born after January 1, 1917 would receive, with few exceptions, lower benefits than those born prior to that year. That was the purpose of the 1977 law.

Replacement Rates Are Stabilized

By moving to a system in which initial benefits were tied only to wage growth, Congress made Social Security less sensitive to unforeseen economic changes. It also created a system in which benefits would replace about the same percentage of pre-retirement earnings for future retirees.

This stabilization of "replacement rates" (benefits in relation to earnings in the year prior to retirement) was sorely needed: For a worker who had average earnings and retired at 65, the old law increased replacement rates from 34 to 42 percent between 1970 and 1974.(29) Had the 1977 amendments not passed, that rate would have reached 46 percent by 1979, 60 percent by 2020, and 67 percent by 2050. Replacement rates for low-income workers would have reached more than 100 percent.

As it turned out, those long-range projections substantially underestimated the growth in replacement rates. Under 1994 assumptions, the 1972 method (had it remained in force) would have produced a 70 percent replacement rate for average-wage workers retiring at 65 by the end of this decade.

The objective of the 1977 amendments was to stabilize replacement rates for all future workers with similar earnings patterns, with the goal of replacing 42 percent of an average worker's wages upon retirement at 65, and 35 percent upon retirement at 62Cas always, replacing a higher proportion of wages for workers with below average earnings, and a lower portion of wages for those with above-average earnings.

Charts 2 and 3 show these replacement rates based on earnings in the year prior to retirement (the charts assume individuals have earnings in each year equal to the average wage, and that they retire either at age 62 or 65).

As these charts demonstrate, the new law does, in fact, correct the unintended upsurge in replacement rates created by the 1972 legislation. The transitional method then eases the movement to the desired replacement rate of about 42 percent (for retirement at 65) or 35 percent (for retirement at 62).(30) In Chart 2, the differences in replacement rates between those born in 1917 and those who follow are much less than those presented in Chart 3.

The 42 percent goal represented a reduction of roughly 10 percent from what the rate was expected to reach by 1979,(31) if the old law had applied. The size of the difference between replacement rates generated by the old law and replacement rates generated by the new law would increase over time.

The replacement rate goal of 42 percent has been closely approximated in actual experience as noted in Chart 4 (which also presents the anticipated versus actual growth in Social Security replacement ratios projected to the year 2000).

The new method stabilized the relationship between initial benefits and pre-retirement earnings. Over the long run, these changes made the financing of Social Security more stable as well, since both revenues (taxes) and expenses (initial benefit levels) would be driven by the same economic factor: wage growth. Further, once an initial benefit was calculated, its purchasing power would be kept current through price indexing.(32)

Thus, Social Security was intentionally redesigned to assure that (1) each year's cohort of new beneficiaries began at the same replacement rate relative to their pre-retirement earnings, as beneficiaries in prior years, and (2) their benefits would keep up with inflation.(33)

For some individuals, the new law resulted in benefit increases.(34) But for most beneficiaries who reached 62 after 1978, the new method produced lower benefits than would have been received under the old law (in calculating those benefits, the Social Security Administration always uses the methodCnew law or transitionalCthat yields the highest amount).

For some born between January 2, 1917 and January 1, 1922, the impact of this deliberate move to lower benefits was softened somewhat by the transitional method.(35) As intended, however, that method quickly became less and less effective in limiting the differences in benefit levels generated by the old versus new laws.(35)

Since Congress made the changeover to the new law effective on the basis of birth date and "grandfathered" workers born before 1917 under the old law, two individuals with similar work histories, but slightly different birth dates, could receive substantially different benefits.(37)

 For individuals retiring at age 62, the differences tended to be modest. For those retiring at later ages, however, the differences could be quite large.(38) It is important also to note that for individuals born before 1910, the differences, whether they retire at 62 or 65, are substantial, both in relation to those born between 1910 and 1917 and those born after 1917.

Charts 5 and 6 show benefit levels (in 1994 dollars) for workers who had average earnings every year and who retired at ages 62 and 65, respectively. The drop in benefit levels for those retiring at 65 and born in 1917, coupled with the continuing drops over the next three years, reflect the impact of the new law and the declining impact of the transitional provision.(39) Thereafter, benefit levels begin to climbCprimarily a consequence of the fact that, given a wage-indexed method for computing initial benefit levels and ensuring stable replacement rates, all future beneficiaries will receive a higher benefit base than those born before them, reflecting the gradual growth of earnings in the economy.(40)

Over the course of the last nine months, the Commission has conducted a careful and thorough review of the "Notch" issue. During its deliberations, it has had the opportunity to interact with hundreds of people who consider themselves adversely affected. They appeared as witnesses, attended hearings, or otherwise communicated with the Commission. The Commission appreciates the depth of feeling and the sincerity of those who seek relief from Congress for what they truly perceive as an unfairness.

The conviction of those born in the "Notch" years that they are entitled to legislative relief is based on their belief that they have been singled out for unfavorable treatment in the way their Social Security benefits are computed. They reach this conclusion from statistical evidence showing that their benefits are lower than benefits for those born in earlier years and somewhat lower than benefits for those born in later years. For many of them, this statistical evidence is bolstered by personal knowledge of large inconsistencies between their benefit amounts and those reported by friends, neighbors, or relatives with similar wage histories.

The Commission does not disputeCindeed, it confirmsCthe factual allegations that benefits for many in the "Notch" years tend to be lower than benefits for many of those born in the years just before the "Notch" period and somewhat lower than for some born later. In light of these findings, the perception of those in the "Notch" years that they have been unfairly treated is entirely understandable. Nonetheless, the Commission finds that such is not the case.

Benefit Differences are Appropriate

In order to support a recommendation for remedial action, the Commission would have to conclude that the differences in benefits between those in the "Notch" years and those with earlier or later birth dates were not only significant, but also inappropriate. We do not find that to be the case.

There is a significant difference in the benefit levels between those in the "Notch" years and those born in earlier yearsCparticularly those born in the few years just preceding the "Notch" period, and especially for those who worked well beyond age 62. This is the result of a conscious decision by Congress to amend the law in 1977 to reduce initial benefit levels for future retirees starting with those who would reach age 62 in 1979.

Congress made this decision because of a very serious flaw in the benefit computation method that had been a part of the 1972 lawCa flaw that was causing initial benefit levels to increase very rapidly. Without such action, the very viability of the Social Security program would have been undermined.

The Commission therefore finds that the reduction in benefit levels instituted by the 1977 amendments was a necessary and, indeed, unavoidable response to a problem. The Commission finds, moreover, that the extent of the benefit reduction was appropriate. Benefits for future retirees were not reduced by more than was then assumed necessary to restore the fiscal viability of the program. In fact, had Congress correctly projected the continuing adverse economic conditions of the years after 1977, it could have reduced benefits more, raised taxes more, or done both.

Why Benefit Differences Exist

The Commission notes that the perception of uneven treatment between those in the "Notch" years and those just before them was magnified by the decision of Congress to fully grandfather under the old law individuals born prior to 1917. Congress could haveCand in retrospect probably should haveClimited the growth in the benefits received by pre-"Notch" individuals, especially those who worked well beyond age 62 after 1978.(41) In other words, the prospective growth in those benefits might have been slowed, thereby making benefit levels somewhat nearer the levels of those in the "Notch" who followed.

The Commission believes that it would not be appropriate at this time to take action reducing benefits, or their growth, for those individuals, all age 78 and older, who benefited from the flawed 1972 benefit method. Such action would not result in higher benefits for those in the "Notch" years.

Any difference in benefits between those in the "Notch" years and those in the post-"Notch" period arises from the operation of the new method for computing initial benefits adopted in 1977. The Commission finds that this difference also represents a reasonable policy outcome.

Under the new method,(42) initial benefits for each new cohort of beneficiaries represent approximately the same percentage of their pre-retirement income as was true for earlier cohorts, and will be true for later cohorts (Charts 2 and 3).(43) In dollar terms, initial benefits increase each year (as shown in Charts 5 and 6).

The important point is that the difference in benefit levels for those born in the "Notch" years and those born in later years arises not because they are treated differently but rather because they are treated in the same way.

It is true that a substantial increase in the rate of wage growth in the mid-1980s created a steep increase in benefits for those born in the years following 1920. However, the fact is that the same general benefit computation method was applied by the 1977 amendments to those born in the "Notch" years and to those born in all subsequent years.

Those in the "Notch" Treated Fairly

If anything, those born in the "Notch" years are more favorably treated, since they have available the special transitional method of determining benefits. Many of those in the "Notch" years get higher benefits as a result of that transitional method than the benefits payable under the general benefit method established under the 1977 legislation, which applies to all later beneficiaries.

The Commission also notes, that, despite the reduction in benefits under the 1977 amendments, those in the "Notch" years receive benefits which represent a generous return relative to the Social Security taxes which they paid in during their working years. This is, of course, also true of those born in the pre-"Notch" years and of those born in the several years after the "Notch" period. However, as the system matures it will become less and less true for future beneficiaries.

Congress' Authority to Change System

The Commission wishes to note that Congress must retain the ability to make changes in the Social Security program that are necessary to ensure its continued viability, even if those changes sometimes adversely affect some beneficiaries. For example, Charts 2 and 3 show that, for persons born after 1937, replacement rates will decline below current levels (and below the levels payable to those in the "Notch" years). This will occur as a result of changes made by the 1983 amendments to Social Security that increase the age at which full benefits are payable.

The Commission finds that Congress' actions in 1977 to reduce benefits were necessary, appropriate, and intentional. Because of unanticipated economic developments, the earlier benefit computation method, adopted in 1972, had disrupted the fiscal balance of the Social Security system by providing unintentionally high benefits.

The reduction in benefits was necessary to stabilize benefit levels and save the Social Security program. The amount of the reduction was appropriate. It eliminated part of the unexpected and under-financed growth in benefits which had been generated by the flawed 1972 benefit computation method.

In its three public hearings, the Commission heard four common misunderstandings about how the 1977 amendments disadvantaged those born in the "Notch" years (1917-21). What follows is a statement of each misunderstanding and a brief explanation of how the law works.

1. People born in the ANotch" years are not allowed to use all of their earnings in calculating their initial benefits. Therefore, they receive lower benefits than they should under the new law.

• For those born in the "Notch" years, initial benefit levels are calculated in one of two ways: via the "new law" (1977) method, which applies to everyone born after January 1, 1917, or the special "transitional" method, whichever guarantees the higher benefit. The new law method considers all earnings; the transitional method does not. The transitional method and the accompanying exclusion of earnings is only used when it results in the highest possible initial benefit.

2. Those born in the "Notch" years are denied cost of living adjustments (COLAs).

• All Social Security retireesCwhether born before, during or after the "Notch" yearsCget automatic annual cost of living benefit increases (COLAs) beginning at age 62 and continuing every year thereafter. Both the new law method and the special transitional method ended the practice of increasing the amount of benefits by price inflation before age 62 as well. This was the flaw that Congress intended to correct. Following the implementation of the 1977 law, benefits were adjusted for price inflation only after age 62.

3. Retirees who follow the Notch yearsCi.e., those born after 1921Care treated better than those in the Notch years.

• The same benefit computation method is used to determine initial benefit amounts for all beneficiaries following the implementation of the 1977 amendmentsCboth those born in the "Notch" years as well as those born after. The only exception is that some born in the "Notch" years receive a more favorable benefit computation method than those who follow as a result of the transitional method. In general, those who follow the "Notch" years receive larger benefits, not as a result of any difference in computation methods, but as a result of the fact that wages increase over time and initial benefit levels are linked to these wage levels.

4. Those born in the "Notch" years were to have been given the choice between using the new law method or the special transitional methodCwhichever yielded the higher benefit. This choice was not offered.

• The Social Security Administration automatically determines which method to use based on which will yield the higher benefit. In determining initial benefit levels for those in the "Notch" years, the Social Security Administration always computes benefits under the new law method as well as the special transitional method. Individuals are automatically paid whichever benefit is higher.

 (1) The two principal reports constituting the body of work previously done on the "Notch" issue are: National Academy of Social Insurance. 1988. Me Social Security "Notch". A Study. Washington, D.C.: National Academy of Social Insurance, and General Accounting Office. 1988. Social Security: The "Notch" Issue. Washington, D.C.: Government Printing Office.

(2) Neither of the two principal committees with jurisdiction over Social Security (the House Ways and Means and the Senate Finance committees) have recommended changing the law. Apart from one procedural vote, no action on "Notch" legislation has ever been taken, either in the House or Senate.

(3) The most recent piece of "Notch" legislation, HR 1883/S 173, carries with it a cost of $42 billion in the first 10 years. For a discussion of these and other costs related to the Notch issue, see Koitz, D., Legislative Proposals to Address Me "Notch" Issue. Statement for the Social Security Notch Commission, June 1994.

(4) As mandated under Public Law 102-393, the Commission was to submit to Congress, no later than December 31, 1993, a detailed report of its findings, conclusions and recommendations. However, due to a delay in completing the appointments (the Presidential appointments were not made until March 1994), the original reporting date was extended by Congress to December 31, 1994 [Public Law 103-256].

(5) The current reserve is equal to approximately one year of benefit payments. Due to the rapid growth of spending with the anticipated retirement of the "baby boom" generation, current analyses indicate that reserves will be drawn down beginning in 2019 (allowing for both tax revenues and interest earned on the assets in the Trust Funds) and depleted in 2029, unless structural changes are made to the benefits system or taxes are increased.

As in any social insurance program, Social Security must keep track of the relationship between its projected income (taxes) and outgo (benefit payments and administrative expenses), not only in current terms, but in relation to the future. Careful actuarial analysis on an ongoing basis is conducted by a staff of professional actuaries within the Social Security Administration. Overall responsibility for the trust fund is now vested by law in a Board of Trustees, including the Secretary of the Treasury, the Commissioner of Social Security, the Secretary of Labor and the Secretary of Health and Human Services, and two members of the public, who are appointed by the President and confirmed by the Senate. The trustees are required to provide an annual report on the actuarial soundness of the program. This report examines the actuarial balance of income and outgo over a 75-year long range period, and a 10-year short range period.

(6) The Commission consists of four appointees each by the House leadership, the Senate leadership, and the President, with the chair named by the President. Following the completion of the appointments process in March 1994, the Commission began its analysis and deliberations.

(7) Insured workers born in 1917-21 had their benefits calculated using whichever calculation method - the new or the transitional - produced the higher benefit.

(8) To date, the vast majority of Social Security beneficiaries receive more value than they paid in to the system in taxes. For example, a one-earner, low-wage couple, both of whom turn 65 in 1995, can expect to receive $94,100 more in benefit dollars (in constant 1993 dollars) than was paid into the trust fund in tax dollars. A similar high wage earning couple can expect $140,700 more in benefits than was paid in Old-Age and Survivors Insurance taxes. For a single, low wage female, the figure is $45,600, while a single, high wage female receives $29,000. Similar categories for males are $28,500 and a negative $10,400, respectively. These figures represent the difference between (I) the value of past employee-employer OAS taxes, accumulated with interest at a real rate of 2 percent, and (ii) the present value of future retirement and survivor benefits, discounted at a real interest rate of 2 percent and reflecting average future life expectancy. The figures are not adjusted to reflect the possibility of death prior to reaching age 65 and thus are related to individuals who are known to have survived. Source: Steuerle, C. Eugene; and J. Bakija. 1994. Retooling Social Security for the 21st Century: Right and Wrong Approaches to Reform. Washington, D.C.: The Urban Institute Press.

(9) While the Commission adopted this five-year time period as a working definition of the "Notch," the investigations, findings and recommendations embodied in this report apply equally to the ten-year period of 1917-26 that some consider to be "Notch" years.

(10) For example, the substantial broadening of coverage in the 1950s brought into the program people who would have been severely disadvantaged under the preexisting benefit formula that looked at covered wages over a period starting with 1937.

(11) Each time Congress enacted legislation to increase benefits, the law it passed would include a new benefit table to take the place of the prior table. The 1968 Social Security amendments, for example, placed a new table in the law that was approximately two pages in length. For workers who, over their working lifetime, had average monthly wages subject to Social Security tax of $74 or less, the table specified a base monthly benefit (called a "Primary Insurance Amount") of $55. For average wages from $75 to $76, the base benefit would be $55.40 and so forth up to the maximum possible average wage, which was then $650, yielding a base monthly benefit of $218. In December 1969, a new set of amendments substituted a new table which showed 15 percent higher benefits for each range of monthly earnings. This 15 percent increase was figured into each wage increment in the existing table and added on to the new highest wage increment as the ceiling of taxable wages increased. Examples taken from the two benefit tables (for average monthly wages around $400) are:

(12) Covered wages are those wages to which the Social Security taxes apply. Also referred to as "covered taxable earnings," these taxable wages are taxed only to an allowable limit. Wage income beyond that limit is exempt from further Social Security tax.

(13) Generally, Congress did not try to match the benefit increases exactly to the level of inflation; in some cases, benefit increases fell short of inflation. More often, Congress provided increases that offset inflation and also provided some additional increases in benefit levels to make up for losses during the years when benefits were not adjusted for inflation.

(14) Although now sometimes referred to as "double indexing," this dual adjustment for wages and prices worked appropriately under the economic conditions prevailing in the 1950s and 1960s. The net result was that benefit levels tended to rise from year to year in such a way that the program continued to provide initial benefits which represented a fairly constant replacement of preretirement earnings. This would no longer be true under the significantly different economic conditions of the 1970s.

(15) In 1965-71, Congress enacted four benefit increases, raising benefits by 53 percent as against a 37 percent increase in price inflation.

(16) The change to this new dynamic methodology produced what appeared to be a significant actuarial surplus in 1972. Congress used this surplus to provide a one-time 20 percent benefit increase in the same legislation that set up the automatic benefit increase system.

(17) High levels of unemployment were also believed to have translated into more applications for disability benefits (in addition to replacing a portion of the earnings lost due to retirement, Social Security provides monthly benefits to individuals who are unable to continue working due to disability, and to survivors of deceased workers).

(18) The size of the retirement benefit is related to "covered" taxable earnings - that is, the amount of wages on which taxes are paid. But the relationship is not direct. Benefit levels are calculated in a weighted fashion, giving a greater return (in relation to wage taxes paid) to lower wage earners. Covered workers are thus "insured" by the system and "entitled," as a result of their tax contributions, to retirement benefit payments. It is important to underscore that Congress expected these payments would replace only a portion of a retiree's wage income. Additional retirement income would have to be provided through a worker's savings and/or pension plan. The original benefits formula, in fact, was designed to ultimately replace, on average, only about 40 percent of a recipient's prior monthly wage.

(19) While this report focuses on those benefit computation changes and their role in creating the "Notch" issue, it is important to understand that those changes were related to, and circumscribed by, other elements of the overall amendment package. Adopting a "tighter" benefit formula, for instance, would reduce the need to raise taxes; a "looser" benefit formula would mean higher taxes or other benefit cutbacks.

(20) The COLA mechanism for all beneficiaries, however, remained essentially unchanged from the 1972 law to the 1977 law.

(21) For example, the wages which a worker earned in 1951, very low by today's standards (because of the combined effects of productivity growth and inflation since 1951), would be adjusted upward to reflect wage levels in the economy at the time of his or her retirement. In addition, the benefit formula (and the amount of wages taxed) would be adjusted for wage growth. These changes would affect initial benefits only. Benefits for people in the rolls would continue to be adjusted annually to reflect price inflation.

(22) Those proposals included the use of price indexing for both initial benefit determinations and for adjusting the benefits of those already retired.

(23) The indexing formula reflects wage increases only, up through the year that a worker reaches age 60, and price increases only from age 62 on, thereby separating the two components completely.

(24) Lower wage earners receive a higher rate of return on their contributions than do higher wage earners. Because Social Security benefits were intended to serve as a hedge against the threat of poverty in old age, such a redistribution from the prosperous to the less prosperous was part of the original design of the program. For example, for a typical worker with low average earnings, born in 1921 and retiring at 65, the replacement rate (before reflecting the reduction for early retirement) is 60 percent. For a similar worker with maximum earnings, the replacement rate is 23 percent. The earnings levels at which the three factors apply are increased for persons becoming 62 (or becoming disabled or dying before age 62) from year to year to keep them current with wage levels in the economy. For the 1994 cohort, for example, the 90 percent factor applied to the first $422 of average indexed monthly earnings; the 32 percent factor applied to such earnings between $422 and $2,545; and the 15 percent factor applied to such earnings above $2,545 up to the covered taxable earning limit.

(25) As noted, the wage adjustments would be made up to the year that the worker turned 60; thus, during this period, only the growth of wages would be used to determine increases in initial benefits. Price changes would affect benefit levels only starting with the year the individual turned 62 (regardless of the year in which benefits began). This procedure resulted in a stable relationship between benefits and pre-retirement wages for workers retiring in all future years.

(26) While the law was signed by President Carter in 1977, the new benefit computation method did not take effect until 1979.

(27) As a practical matter, no one could receive lower retirement benefits under the new law until at least 1980, the third year after enactment. After that point, the two restrictions would increasingly limit the extent to which persons qualifying for the transitional method would receive higher benefits similar to those grandfathered under the old (1972) law.

(28) The opportunity to benefit from the transitional formula was not available to people born after 1921.

(29) This change primarily reflected the impact of the ad hoc 20 percent benefit increase included in the 1972 amendments.

(30) Some have pointed out that another, smaller "Notch" appears on this replacement rate graph. Some workers retiring at later ages - particularly, those born in 1920-23 - received replacement rates of about 41 percent at age 65. This anomaly is the result of the fact that, as the computation method interacts with the annual wage and price fluctuations in the economy, it can, in fact, produce some small variation in the replacement rate, especially for those who work past age 62. The Commission has found, however, that this is solely a consequence of the dynamic relationship between the benefit computation method and these larger forces of the economy, as expressed in annual wage and price fluctuations. Similar differentials are likely to occur again in the future. The Commission has concluded that this degree of variation is both minor and acceptable.

(31) The corresponding reduction at age 62 was five percent. In practice, the actual differences were significantly greater than originally expected. This occurred because the benefit level under the old law continued to increase more rapidly than expected, while the level if computed under the new law was stable at the intended level.

(32) In fact, with automatic annual increases in benefits for those already retired based on price inflation, it is possible that high inflation and low wage growth could produce an imbalance between revenues and benefits - but is less likely to do so than a system that uses both price inflation and wage inflation in determining initial benefits.

(33) At the time of the 1977 amendments, it was estimated that the additional financing, benefit adjustments, and stabilized benefit computation method provided in the law would restore the fiscal integrity of the program for many years. Unfortunately, the economy continued to defy the expectations of estimators. In adopting the amendments, Congress relied on projections that the rate of inflation and the rate of unemployment would both decline over the next five years, and that wage growth would be about 1.75 percentage points higher than inflation. In fact, there was negative real wage growth, inflation hit double-digit levels, and unemployment climbed. By the early 1980s, Social Security was again in serious financial difficulty. But, the new benefit computation method did perform as expected, stabilizing replacement rates and reducing the long-term cost of the program. Without these changes, the financial problems of the 1980s would have been much more severe.

(34) The transitional method and the new indexed method allowed approximately 20 percent of the "Notch" beneficiaries to receive benefits equal to or higher than what they would have received under the old law. About 14 percent of them received benefits that, while lower than those that would have been paid under the 1972 law, were higher than would have been paid under the new method alone. This effect was strongest for those born in the earliest Notch years; the effect gradually lessened over the later years. For example, about 40 percent of retirees born in 1917 had their benefits determined under the transitional method rather than the new, wage-indexed method; by contrast, only 2 percent of those born in 1921 benefited from the transitional method.

(35) No retiree in 1979 could get a lower benefit under the new law than under the old law since, in that year, the transitional method fully duplicated the old-law benefit method. Starting in 1980, however, the transitional method gave only partial protection. It would not provide cost-of-living increases for years after 1978 and prior to the year of reaching age 62. Nor could earnings after age 61 be involved in computing benefits under the transition.

(36) In 1980, for example, retired workers born in 1917 received the full benefit of the old law formula when they used the transitional method with the exception that they could not count any earnings after 1978. Beneficiaries born in 1918 and retiring in 1980 also received the full benefit of the old law method with the exception that they would receive no cost-of- living increase for 1979 and they could not count any earnings after 1979. It is important to note however, that such earnings after age 61 are used in the new law method if that computation provides beneficiaries with a higher monthly benefit payment than the transitional method.

By 1986, a worker born in 1917 would have wages earned in 1979-86 excluded from the transitional method; a worker born in 1921 would not be able to use the cost-of-living increases for 1979-82 and would also be unable to use earnings in 1983-86. Importantly, however, any worker who would receive a higher benefit using the new method (counting all earnings and receiving all their COLAs) would have their benefits calculated using the new method. The great majority of such workers would receive a higher benefit under the new wage-indexed method.

Allowing for these options the majority of workers - particularly those retiring after the first few years of the transition period - received benefits that were significantly lower than they would have received under the old 1972 method (and intentionally so).

(37) It should be pointed out that differences could occur even if both received benefits computed under the same method, However, the extent of the difference was magnified when the benefits were computed under the two different methods, of which one was known to be over-indexing initial benefit amounts, while the other was designed to prevent such over-indexing

(38) For example, consider two workers who both retire at the same age after having average earnings throughout their careers. One was born on December 31, 1916 and the other on January 2, 1917. If both retire in 1979 at age 62, then the initial difference in their benefits is only about $7 per month (and arises for reasons unrelated to the "notch" issue). If they retire in 1982 at age 65, however, the initial difference is $110 per month. Similarly, the difference for retirement in 1987 at age 70 is $199. (See memorandum of August 31, 1994 in the Commission's files, "Comparison of Monthly Benefits for Persons Who Attain Age 62 in December 1978 and In January 1979 Under Present Law and Under a Modified Benefit Formula").

(39) Without the transitional method, there would have been a severe drop in benefits for the 1917 cohort who retired at age 62.

(40) The decrease for those born after 1937 is due to the legislated increase in the Normal Retirement Age, resulting in reduced benefits for retirement at age 65 and larger reductions for retirement at age 62 than is now the case.

(41) Under the economic projections used by Congress when it adopted the 1977 amendments, the gap between the benefits for those "grandfathered" and for those who came under the new law would have been much smaller than the gap actually turned out to beCas a result of the double digit inflation of the late 1970s.

(42) This was not the only policy choice available to the Congress, and it is beyond the scope of this Commission's mandate to consider the merits of the alternative methods considered at the time. However, the benefit method adopted was one which Congress considered the most appropriate approach and one which was designed to treat all future beneficiaries fairly.

(43) As noted previously, replacement rates are much more stable under the new law than under the Old, but there can be minor differences from year to year. Consequently Charts 2 and 3 show a slight lowering of replacement rates for those born in the years 1920-1924 for retirement at age 65. This is a minor variability resulting from the new benefit method and its interaction with certain economic conditions. It is not a part of the "Notch" issue and is likely to occur again as economic circumstances change from year to year. Moreover, this effect applies only to a limited number of individuals. Other analyses done for the Commission show that most persons born in those years did not experience this dip in replacement rates (see the memorandum of December 5, 1994 in the Commission's files, "Comparison of Actual Replacement Ratios for Retired Workers Born in 1917-26 Versus Illustrative Replacement Ratios for Theoretical Steady Workers").

American Enterprise Institute. 1985. Proposals to Deal with the Social Security Notch Problem. Washington, D.C.: American Enterprise Institute.

Congressional Budget Office. 1977. Financing Social Security: Issues For the Short and Long Term. Washington, D.C.: U.S. Government Printing Office.

Duggan, James E., Robert Gillingham, and John G. Greenless. 1994. Distributional Effects of Social Security: The Notch Issue Revisited. Washington, D.C. U.S. Treasury Department.

General Accounting Office. 1988. Social Security: The Notch Issue. Washington, D.C.: Government Accounting Office.

Kingson, Eric R., and Edward D. Berkowitz. 1993. Social Security and Medicare: A Policy Primer. Westport, CT: Auburn House.

Krueger, Alan B., Jorn-Steffen Pischke. "The Effects of Social Security on Labor Supply: A Cohort Analysis of the Notch Generation." Journal of Labor Economics. 10 (1992): 412- 437.

Myers, Robert J. 1993. Social Security. Philadelphia, PA: University of Pennsylvania Press.

National Academy of Social Insurance. 1988. The Social Security Benefit Notch: A Study. Washington, D.C.: National Academy of Social Insurance.

National Commission on Social Security Reform. 1983. Report of the National Commission on Social Security Reform. Washington, D.C.: U.S. Government Printing Office.

Sandell, Steven H. 1993. Are Retirement Benefits Too High and Will They Be Too Low?: An Examination of Wage Indexation in Social Security. Washington D.C.: Population Studies Center.

Stererle, C. Eugene, Jon M. Bakija. 1994. Retooling Social Security for the 21st Century: Right and Wrong Approaches to Reform. Washington, D.C.: The Urban Institute Press.