Social Security Beneficiaries Affected by the Windfall Elimination Provision in 2006

Summary

The Windfall Elimination Provision (WEP) is an alternative method of computing benefits for some workers who receive a pension based on work not covered by Social Security. The WEP computation results in a lower benefit than the worker would receive under the regular computation method. This article provides a brief legislative history, describes the WEP computation and applicability, and presents statistical data about beneficiaries affected by the WEP.

The statistical data show that, as of the end of December 2006:

• About 970,000 beneficiaries, mainly retired workers, were affected by the WEP, and the WEP affected the benefits of almost 3 percent of all retired workers.
• Almost half of the retired workers affected by the WEP received a federal pension, and another 36 percent received either a state or local pension.
• Sixty-five percent of both retired and disabled workers affected by the WEP were men.

History of the Windfall Elimination Provision (WEP)

The regular Social Security benefit computation formula is weighted to provide a higher replacement of earnings for workers with low earnings. Most employment and self-employment in the United States is covered by Social Security. Workers and their employers pay taxes up to an annual maximum amount, and the earnings are creditable for Social Security purposes. Before the WEP was enacted, individuals who had worked mainly in employment not covered by Social Security had their benefits computed as if they were long term low-wage earners. The WEP prevents this unintended windfall for workers who receive a pension from a job where they did not pay Social Security taxes, but who benefited from provisions aimed at low lifetime earners. Examples of pensions from noncovered employment are Civil Service Retirement pensions payable to federal employees hired before 1984, state and local government pensions based on noncovered earnings, and certain pensions from earnings in foreign countries.

The Windfall Elimination Provision was one of the many legislative changes included in the Social Security Amendments of 1983 (Public Law 98–21). Major provisions of this legislation included gradually raising the retirement age and making a portion of Social Security benefits received by higher income beneficiaries subject to income taxes. The amendments also provided for mandatory Social Security coverage of newly hired federal employees and current and future employees of nonprofit organizations (Svahn and Ross 1983, 24–27).

Prior to Congressional action, the issue of windfall benefits payable to persons with noncovered employment was considered in two bipartisan national Social Security study commissions. The National Commission on Social Security issued its report on March 12, 1981. One of its recommendations was that "the windfall portion of benefits arising from periods of noncovered government employment in the future (due to the weighted benefit formula) should be eliminated" (National Commission on Social Security 1981, 26).

The WEP was also on the agenda of the later National Commission on Social Security Reform (NCSSR). The commission's report, released in January 1983, recommended "that the method of computing benefits should be revised for persons who first become eligible for pensions from non-covered employment, after 1983, so as to eliminate 'windfall' benefits." The report included two methods of modifying the windfall. One method would make the percentage related to the second bendpoint of the primary insurance amount (PIA) formula applicable to the first bendpoint (32 percent instead of 90 percent) for workers with noncovered pensions. The reduction in benefits would not be larger than the pension from noncovered employment. The second method would apply the current benefit formula to a record that combines both covered and noncovered earnings to determine a replacement rate, which would then be applied to the average earnings based solely on covered employment (NCSSR 1983, 2-9–2-10).

In February 1983, the Social Security Subcommittee of the House Ways and Means Committee completed markup sessions on a draft bill. The subcommittee agreed with the first NCSSR approach to modify the benefit formula on advice that the second method would pose significant administrative problems, and that generally similar results could be achieved by reducing the percentage in the first bendpoint (Svahn and Ross 1983, 10).

After further action in both the House of Representatives and Senate, the conference committee agreement substituted 40 percent for 32 percent as the percentage applicable to the first bendpoint, provided for the 5 year phase-in period, and exempted newly covered employees and those with 30 years of covered work (Committee on Ways and Means 1983, 121). These WEP provisions were included in the legislation signed by President Ronald Reagan on April 21, 1983.

The WEP Computation

Social Security benefits are based on the PIA, which is the monthly benefit payable to the worker upon retirement at full retirement age or upon entitlement to disability benefits. The PIA is derived from the worker's average indexed monthly earnings (AIME). The AIME is based on annual covered earnings that have been indexed to reflect changes in wage levels since the year the earnings were paid. The indexed earnings are then averaged over most of the worker's adult years to determine the AIME.

After the AIME has been determined, the PIA is computed by applying a formula to the AIME. The formula applies three percentages to three brackets of the AIME. The formula is weighted to provide a higher PIA to the AIME ratio for workers with relatively low earnings by applying declining percentage rates to the three AIME brackets. The dollar amounts defining the AIME brackets are referred to as bendpoints and are updated each year in proportion to increases in the national average wage level. For workers who first became eligible for benefits—those who reach age 62 or become disabled—in 2006, the PIA is equal to the sum of:

90 percent of the first $656 of AIME, plus

32 percent of the next $3,299 of AIME, plus

15 percent of the AIME over $3,955.

The PIA is the monthly amount payable at full retirement age (FRA)—age 66—for workers who attained age 62 in 2006. Retirement benefits are reduced for each month of benefit receipt before the FRA. Disabled workers may receive 100 percent of the PIA, unless they receive a reduced retirement benefit for months before disability entitlement (SSA 2007a, 14–17).

The WEP computation for the PIA generally applies 40 percent to the first bendpoint instead of the 90 percent used to compute the regular PIA. The maximum amount of PIA reduction is half of the amount of the first bendpoint applicable to the year of first eligibility. The following example shows the maximum PIA reduction due to the WEP for a worker who attains age 62 or becomes disabled in 2006:

Regular PIA, based on AIME of $800

$656 x .90 = $590.40

$144 x .32 = $46.08

Regular PIA = $636.40

Windfall Elimination Provision PIA, based on AIME of $800

$656 x .40 = $262.40

$144 x .32 = $46.08

WEP PIA = $308.40

After the calculation, the PIA is rounded down to the nearest $0.10. The maximum PIA reduction due to the WEP is $328, one-half of the first bendpoint of $656.

WEP Applicability

The WEP computation reduces the PIA for some retired and disabled workers and their spouses and children. It is the PIA of record. However, the WEP computation does not apply to survivor benefits. If a worker dies, benefits for widow(er)s and children are based on the regular PIA.

The WEP may apply to workers who attained age 62 or became eligible for disability benefits after 1985 and became eligible after 1985 for a pension based in whole or in part on earnings in employment not covered by Social Security. Workers who have 30 or more years of substantial earnings covered under Social Security are exempt from the WEP. The annual amount of substantial earnings is based on a formula and is updated each year based on the increase in the national average wage level. The formula for substantial earnings is 25 percent of the old law's contribution and benefit base, the amount that would have determined maximum taxable earnings for benefit computation purposes had the 1977 Social Security Amendments (which included ad hoc increases in the maximum taxable earnings) not been enacted. Amounts of substantial covered earnings were $11,625 for 1996 and $17,475 for 2006 (SSA 2007c, 2). These amounts are 25 percent of the old law's base amounts of $46,500 and $69,900, respectively.

Benefits for workers with 21–29 years of substantial covered earnings are not fully impacted by the WEP. The 40 percent applicable to the first bendpoint percentage is increased by 5 percentage points for each year of substantial covered earnings beginning with the 21st year and through the 30th year, at which point the WEP no longer applies. The first bendpoint percentage ranges from 45 percent for workers with 21 years of substantial covered earnings to 85 percent for workers with 29 years (SSA 2007a, 16).

There is a guarantee provision for workers with relatively low pensions based on noncovered employment. The reduction in the Social Security benefit due to the WEP cannot exceed one-half of the amount of the pension based on noncovered earnings after 1956 (SSA 2007c, 2).

The WEP was phased in for workers first eligible for retirement or disability insurance in 1986 through 1989. For those first eligible in 1986 and subject to the WEP, the applicable first bendpoint percentage was 80 percent. The percentage decreased by 10 percentage points each year for those first eligible in 1987–1989 and reached 40 percent for those first eligible in 1990 or later (SSA 2007a, Table 2.A11.1).

The WEP does not apply to two groups of employees whose Social Security coverage was mandated in 1984—federal employees who were first hired after December 31, 1983, and certain employees of nonprofit organizations who started employment on December 31, 1983. Other WEP exceptions include pensions based on railroad employment, pensions based solely on noncovered employment before 1957, and certain pensions paid as a result of totalization agreements between the United States and foreign countries (SSA 2007c, 2).

Beneficiaries Affected by the WEP

The following sections present statistical data on beneficiaries affected by the WEP who were in current-payment status as of December 2006. The beneficiary statistical data were derived from the Social Security Administration's Master Beneficiary Record (MBR). The MBR is the principal file used in the administration of the Social Security benefit program. The data show demographic and program characteristics of those affected, information on pensions and years of coverage, and effects of the WEP on the PIA and monthly benefits.

Demographic and Program Characteristics

At the end of 2006, 971,310 beneficiaries were affected by the WEP. Ninety percent were retired workers. Another 2 percent were disabled workers, and the remaining 8 percent were spouses and children of affected workers (Table 1). From 1999 to 2006 the number of affected beneficiaries more than doubled. Although small numbers of disabled workers are affected by the WEP, the percentage increase was slightly higher for these workers (137 percent) than for retired workers (110 percent) during this period.

About 65 percent of both retired and disabled workers affected by the WEP were men. The WEP affected the benefits of 3.6 percent of all men receiving retired-worker benefits, compared with 2.0 percent of all women (Table 2).

The WEP affected the benefits of about 4 percent of retired workers under age 75, but only 2.3 percent of those aged 75–79 and 0.3 percent of those aged 80 or older. The WEP did not apply to persons first eligible for benefits prior to 1986.

Also, those first eligible after 1985 were affected only if they were entitled to a noncovered pension that began after 1985. Thus, fewer older beneficiaries were affected by the WEP.

Tables 3 and 4 present state-by-state breakdowns of beneficiaries affected by the WEP. In some states, the percentage of retired workers affected by the WEP was substantially higher than the national figure of 2.8 percent. More than one-tenth of the retired workers in Alaska and the District of Columbia were affected as well as about 5 percent of those in Colorado, Maine, Maryland, Nevada, and Ohio. The percentages are higher in these states because they had either many federal employees or low percentages of state and local employees who were covered by Social Security. In both 1991 and 2001, less than 50 percent of state and local workers in Alaska, California, Colorado, Maine, Louisiana, Massachusetts, Nevada, and Ohio were covered by Social Security. (Committee on Ways and Means 1996, 10–11 and 2004, 1-43–1-44).

Because of the small number of disabled workers affected by the WEP, percentages affected in each state are not presented.

Pensions and Years of Coverage

Table 5 shows the sources of pensions for retired and disabled workers affected by the WEP. Almost half of the retired workers and over two-fifths of the disabled workers received federal pensions. The percentages of retired and disabled workers receiving state and local pensions were similar, 36 percent and 38 percent, respectively.

Almost three-fourths of workers affected by the WEP had fewer than 21 years of substantial covered earnings (Table 6). Thus, almost three-quarters of the workers were subject to the maximum PIA reduction—50 percent of the first bendpoint, unless they were covered by the WEP guarantee. Another 10 percent had 21–24 years, and about 8 percent had 25–29 years. Retired and disabled workers had similar distributions of years of covered earnings, but women tended to have fewer years than men.

Table 7 shows distributions of the monthly noncovered pension amount for workers affected by the WEP who became entitled to Social Security benefits in 2004–2006. These years were selected to provide fairly current data on pensions. For beneficiaries entitled in these years, higher percentages of disabled workers and women received lower pensions. About two-thirds of disabled workers received monthly noncovered pensions of less than $2,000, compared with 45 percent of retired workers. About two-thirds of women, compared with one-third of men received less than $2,000 per month.

Table 8 relates the amount of the monthly noncovered pension to the primary insurance amount for recently entitled retired workers who were affected by the WEP. Since persons affected by the WEP had many years in noncovered employment, PIAs tended to be low. The years of zero Social Security earnings were included in the computation of the AIME, resulting in a lower PIA, which is further reduced by the WEP.

Overall, about 33 percent of the men and 44 percent of the women in this group had PIAs of less than $300 per month. However, only 11 percent of the men and 32 percent of the women with noncovered pensions of less than $1,000 had PIAs under $300. Some workers with the lowest noncovered pension had fewer years of noncovered employment and more years of covered employment and thus were able to earn higher PIAs.

Effect of the WEP on the PIA

For about 87 percent of the affected workers, there was enough information to estimate the PIA before application of the WEP. The application of the WEP substantially reduced the PIAs of both retired and disabled workers. For retired workers, the reduction in the PIA averaged $246 or 35 percent of the PIA before application of the WEP (Table 9). Among disabled workers, the PIA reduction averaged $262 or 30 percent of the PIA before the WEP (Table 10).

The maximum reduction in the PIA due to the WEP depends on the year of first eligibility for benefits and was $328 for workers first eligible in 2006. Among retired workers aged 62–64, the estimated reduction in the PIA was $273. As noted in Table 6, almost three-quarters of retired workers affected by the WEP had fewer than 21 years of substantial covered earnings and could be subject to the maximum PIA reduction. However, the PIA reduction averaged only $119 or 18 percent for retired workers aged 80 or older (Table 9). These individuals were first eligible for benefits prior to 1989 and, thus, were affected by the gradual phase in of the WEP reduction (as discussed above).

Among retired and disabled workers affected by the WEP, average PIAs before the WEP reduction were lower than average PIAs for all workers in these groups. For affected retired workers, the average PIA before application of the WEP was $706, compared with $1,072 for all retired workers receiving benefits at the end of 2006 (Table 9) (SSA 2008, Table 5.B7). Comparable figures for disabled workers were $882 and $979, respectively (Table 10) (SSA 2008, Table 5.E1).

Benefit Amounts for Retired Workers

Retired workers receive the full PIA if they begin receiving benefits at the FRA. Monthly benefits are permanently reduced for each month of entitlement before the FRA. The maximum reduction for workers who attained age 62 before 2000 was 20 percent and is gradually increasing along with the FRA (SSA 2007a, Table 2.A17.1). The maximum reduction for workers who attained age 62 in 2006 is 25 percent, based on a FRA of 66.

The monthly benefit for retired workers is increased if the worker is dually entitled to a higher benefit as a spouse or widow(er). This worker receives the amount of the worker benefit plus the difference between that amount and the benefit he or she would receive as a spouse or widow(er). A beneficiary cannot receive both the full worker benefit and the full spouse or widow(er) benefit.

Table 11 shows average PIAs and monthly benefits for all retired workers affected by the WEP. The PIA is the amount after application of the WEP. The average benefit for those affected by the WEP—$431 (Table 11)—was much lower than average benefits for all retired workers receiving benefits—$1,044 (SSA 2008, Table 5A.1.1).

For men affected by the WEP, monthly benefits averaged $443 and PIAs averaged $500. Average benefits were lower than average PIAs for all age groups, reflecting a number of early retirements. Since the benefits for all men aged 62–64 were reduced for early retirement, the average benefit of $364 was only 77 percent of the average PIA of $473.

The average benefit for all women affected by the WEP ($408) was higher than the average PIA ($391). For women aged 75 or older, the average benefit was about 26 percent higher than the average PIA. About 51,000 women, one-sixth of those affected by the WEP, were dually entitled to a spouse or widow benefit. In contrast, only 2,400 men were dually entitled (Table 12). Benefits averaged $664 for dually entitled women and $616 for dually entitled men affected by the WEP.

For some of the dually entitled beneficiaries affected by the WEP, the amount of the spouse or widow(er) portion of the benefit may have been reduced due to the Government Pension Offset (GPO) provision. The GPO provision also affects the benefits of persons who work in noncovered employment. The WEP affects the worker benefit, and the GPO affects the spouse and widow(er) benefits of persons who receive a pension based on their own work in noncovered government employment. Thus, persons who work in noncovered employment can have their worker benefit reduced due to the WEP and their spousal or widow(er) benefit offset by the GPO. Unless certain exceptions apply, the spousal or widow(er) benefit is generally reduced by two-thirds of the noncovered pension (SSA 2007b). The GPO could completely offset the spousal or widow(er) portion of the benefit, and, thus, the retired worker would not be counted as dually entitled in Table 12.

Conclusion

The WEP was enacted to prevent a windfall for workers who receive a pension from a job where they did not pay Social Security taxes, but would benefit from provisions aimed at low earners. The impact of the WEP is as intended: it helps to ensure that workers with pensions from noncovered employment do not receive the advantage of the weighted benefit formula that is intended for career-long low earners.

The number of beneficiaries affected by the WEP has been increasing and should continue to increase as the baby-boom generation retires. Certain advocacy groups and individuals affected by the WEP have raised concerns about the WEP, arguing that it unfairly targets public employees who are low earners or have careers split between covered and noncovered employment. In response, during the past several Congressional sessions, bills were introduced to eliminate or modify the WEP:

• The Social Security Fairness Act of 2007 (H. R. 82 and S. 206) would repeal the Windfall Elimination Provision and also the Government Pension Offset Provision for monthly benefits payable after December 2007.
• The Windfall Elimination Provision Relief Act of 2007 (H. R. 726) would combine amounts of the monthly noncovered pension and the worker's monthly PIA and apply less stringent WEP provisions for smaller combined income amounts.¹
• The Public Servant Retirement Protection Act of 2007 (H. R. 2772 and S. 1647) would repeal the current WEP provisions and substitute a new PIA formula based on the proportion of earnings in Social Security covered employment to the total earnings in both covered and noncovered employment.²