Testimony for Hearing before the House Ways and Means Committee Subcommittee on Social Security
on Government Pension Offset and Windfall Elimination Provision (WEP-GPO)
Jane L. Ross, Deputy Commissioner for Policy Testified
June 27, 2000

Mr. Chairman, Members of the Subcommittee: Thank you for giving me the opportunity to appear before you today to discuss the government pension offset provision, commonly referred to as the GPO. Today I will describe for you how this provision works and the reasons Congress enacted it. I will then discuss H.R. 1217, a bill that is before this Subcommittee that would change the way this provision is applied.

The GPO affects Social Security beneficiaries who receive pensions based on work not covered under Social Security. These beneficiaries are retired workers who were employed by Federal, State, and local government entities. The GPO reduces or eliminates the Social Security benefit payable to a person as the spouse or surviving spouse of a worker.

In enacting the GPO, Congress intended to remove an unfair advantage available to government workers whose employment was not covered under the Social Security program. The goal of the GPO is to assure that Social Security dependent's benefits will not be paid to persons not dependent on the worker.

Description of GPO

The GPO affects government retirees who are eligible for two retirement benefits:

  • A pension based on their own work in a Federal, State, or local government job that was not covered by Social Security, and
  • A Social Security dependent's benefit based on their husband's or wife's work in covered employment.

For such a retiree, the GPO requires that the Social Security spouse's benefit be reduced--or "offset"--by two-thirds of the amount of the noncovered government pension.

To simplify the discussion I will generally talk about how the GPO affects Social Security benefits paid to wives. Actually, the provision also similarly affects benefits for other dependents--husbands, widows, and widowers. Also, when I speak of government

employment, unless I note otherwise, I am referring to employment that was not covered by Social Security on the last day of the worker's employment. This is the test used under the current law offset provision.

Rationale for the GPO

Let me now turn to a discussion of the rationale for the GPO. The GPO provision, enacted by Congress in 1977, is designed to replicate the dual-entitlement provision of the Social Security program for workers receiving a pension from noncovered government employment. The dual-entitlement provision, which has applied since 1940, requires that Social Security benefits payable to a person as a spouse or surviving spouse be reduced by the amount of that person's own Social Security worker's benefit. Thus, a person who works in a job that is covered under Social Security and receives a Social Security worker's benefit cannot also receive a full Social Security spouse's benefit. The dual-entitlement provision was intended to restrict the payment of benefits to those family members who were actually dependent on the worker.

Under the dual entitlement provision, if a person is entitled to a larger Social Security benefit as a worker than as a spouse, no spouse's benefit is payable because the person is not considered dependent on the other spouse. Similarly, if the benefit payable as a spouse exceeds the worker's benefit for that person on their own record, then the spouse's benefit is offset by the amount of the worker's benefit. As a result of the dual entitlement provision, nearly 6 million beneficiaries receive reduced benefits as spouse--which is to say that they receive the equivalent of the worker's benefit or the spouse's benefit, whichever is higher.

The GPO acts as a surrogate for the dual entitlement provision for workers receiving a government pension based on work not covered under Social Security because if the work had been covered, any spouse's or surviving spouse's benefit would have been reduced by the person's own Social Security worker's benefit. Government pensions are, to a large extent, a substitute for Social Security benefits. The result of enactment of the GPO is that spouses and surviving spouses are treated similarly, regardless of whether their jobs are covered under Social Security or not. Thus, the GPO helps ensure that those who receive Social security benefits as dependents were, in fact, dependent to some extent on the worker for financial support.

Visual Presentation

To better illustrate the dual entitlement provision and how the GPO helps replicate dual-entitlement, I've brought along several illustrative charts.

The first chart shows Mary, a woman worker whose monthly Social Security retirement benefit is $600. Mary's husband has a Social Security worker's benefit of $1,400, entitling Mary to a Social Security spouse's benefit of $700. However, because of the dual-entitlement provision, Mary cannot receive both benefits in full. Her $700 benefit as a spouse is offset dollar for dollar by her $600 worker's benefit. Consequently, she receives her worker's benefit of $600 plus the remaining spouse's benefit of $100 for a total monthly benefit of $700.

The second chart shows Mary's sister Nancy. Nancy worked in a government job not covered by Social Security. While Nancy does not receive a Social Security worker's benefit, she does receive a monthly government pension of $600. Like Mary's husband, Nancy's husband also receives a monthly Social Security benefit of $1,400 so Nancy is eligible for a Social Security spouse's benefit of $700. Since Nancy does not receive a Social Security worker's benefit, the dual entitlement provision does not apply. If the GPO provision did not exist, Nancy would receive her monthly spouse's benefit of $700 plus her government pension of $600 for a total of $1,300 in monthly benefits.

The third chart reflects present law and shows how the GPO provision affects Nancy's Social Security benefits. Under the GPO, Nancy's $700 spouse's benefit is reduced by $400, two-thirds of the amount of her monthly government pension. Thus, under GPO, Nancy receives her monthly pension of $600 plus a monthly spouse's benefit of $300 for a total of $900. As a result of the GPO, Nancy, who worked in a noncovered job, is treated like Mary, who is affected by the dual entitlement provision.

The fourth chart shows a comparison of the three examples. I believe that this chart demonstrates how the GPO was designed to replicate the dual entitlement provision and remove the favorable treatment under Social Security that previously existed for government workers who were not covered under Social Security.

Why the Offset is Less Than 100 Percent

Although the GPO provision is intended to accomplish the same purpose as the dual entitlement provision, the amount of the reduction under the GPO is different:

Under the dual entitlement provision, there is a dollar-for-dollar reduction: if a woman gets a monthly Social Security benefit of $300 based on her own work, then $300 is subtracted from any Social Security benefit she would get as a wife.

Under GPO, there is a two-thirds reduction. If a woman gets a monthly pension of $300 based on her own work in government, then two-thirds of it ($200) is subtracted from any Social Security benefit she would get as a wife.

The GPO replicates the Social Security dual-entitlement rule by assuming that two-thirds of the government pension is approximately equivalent to a Social Security retirement benefit the worker would receive if his/her job had been covered by Social Security. Therefore, only two-thirds of the government pension is used to offset Social Security benefits.

The other third of the government pension is considered as the equivalent of a private pension and is not used to offset Social Security benefits. Here again, the GPO affects government workers in the way that dual entitlement affects non-government workers: both groups of workers can receive a private pension without having it offset against their Social Security spouse's benefit.

While Congress settled on two-thirds of the government pension as the amount to be considered equivalent to a Social Security worker's benefit, actually there is no single offset rate that would be a precise match in every case. This is because Social Security and government pensions have different benefit structures and serve different purposes.

A 1990 Congressional Research Service study of the effects of the GPO on employees under the federal Civil Service Retirement System concluded, in part, that the net effect of these differences between Social Security and government pensions is that low earners probably have less of a reduction in their spouse's benefits under the GPO than they would under dual entitlement. Thus, low earners subject to the GPO generally do somewhat better than similarly-situated low earners subject to the dual entitlement provision and high earners do somewhat worse under the GPO than similarly-situated high-earners subject to dual entitlement.

Impact of GPO

As of December 1999, 305,000 beneficiaries had their benefits fully or partially offset due to the GPO. The following table shows some important distinctions in its effects on men and women:

  Men Women
Affected by the GPO 96,000 209,000
Benefits Fully Offset 98% 5%
Average Monthly Offset $276 $391

The difference in the average pension received by those affected by the GPO varies significantly by gender. For women, the average monthly pension amount is currently about $1,150; for men it is $1,950. As expected, the pension amount is larger for those who are fully offset. For women whose benefits are fully offset, the average pension is $1,400, but only $540 for women whose benefits are partially offset under the GPO provision.

Arguments for Eliminating the GPO

Critics of the GPO say that:

  • The provision is not well understood and many people are unprepared for a smaller Social Security benefit than they had assumed in making retirement plans.
  • Reducing everyone's spousal benefit by two-thirds of their government pension is an imprecise way to estimate what the spousal benefit would be had the government job been covered by Social Security. Ideally, the way to compute the dual entitlement rule would be to apply the Social Security benefit formula to an individual's total earnings, including the noncovered portion, and reduce the resulting Social Security benefit by the proportion of total earnings attributable to noncovered earnings. However, this is not possible from an administrative standpoint because SSA does not have information on a person's noncovered earnings history.
  • The GPO unfairly singles-out workers with government pensions, compared to those with private pensions.

Arguments for Retaining the GPO

Defenders of the GPO maintain that:

  • It is an effective method to cut back what otherwise would be an unfair advantage for government workers.
  • It is appropriate to provide different treatment for workers with government pensions if they did not pay Social Security tax on the employment that generated the pension.
  • There has been ample time for people to adjust their retirement plans since the provision has been in the law for nearly 23 years.
  • Had these workers been covered by Social Security, in many cases Social Security's dual entitlement rule would actually produce a greater reduction of spouse's benefits than does the GPO.

While not always perfect, because administrative considerations precluded applying the Social Security benefit computation rules to government employment, the GPO is defended as a practical way to prevent undue Social Security benefits from going to government annuitants.


Now I'd like to discuss H.R. 1217, the bill introduced by Representative Jefferson to modify the GPO. Under H.R. 1217, the GPO would be eliminated for individuals with combined monthly Social Security spouse's/surviving spouse's benefits and non-covered government pensions of $1,200 or less. If the combined amount exceeds $1,200, there would be a dollar-for-dollar offset for the excess above $1,200. The bill would also guarantee that the offset could not exceed two-thirds of the pension, as guaranteed under present law. The $1,200 amount would be indexed by annual cost-of-living adjustments (COLAs).

Effects of H.R. 1217

We've prepared another chart to show the effects of H.R. 1217. You'll recall, Nancy, our government worker, receives a monthly pension of $600 based on her noncovered government employment. In addition, her monthly spouse's benefit of $700 is reduced to $300 because of the current law GPO provision. The total that she would receive would be reduced from $1,300 to $900. Under H.R. 1217, no reduction would apply to the first $1,200 in combined monthly benefits. The combined amount of her government pension and Social Security benefits--$1,300-would exceed the $1,200 threshold by $100. Consequently, Nancy's Social Security spouse's benefit would be reduced under the bill by this $100 excess. Nancy would then receive her full pension of $600 and $600 per month in spouse's benefits (after the offset).

The final chart shows a comparison of all four examples. It shows how H.R. 1217 compares to present law and how it compares to the law that was in effect prior to enactment of the GPO. It also shows the effect H.R. 1217 would have on noncovered government workers compared to similarly-situated persons who worked in covered employment.

By linking the application of the GPO to a dollar threshold for the combined amount of the monthly government pension and the Social Security benefit, H.R. 1217 would be a significant departure from the earned-right nature of the program. The strong public support that Social Security has enjoyed since its inception is based on the concept that the Social Security taxes workers pay establishes a right to benefits that is not dependent on a showing of presumed financial need. In this regard, H.R. 1217 does nothing for the millions of similarly-situated spouses with low benefits who have worked and paid contributions into the system but whose benefits are affected by the dual-entitlement provision.

SSA's Office of Policy prepared an analysis of the effects of H.R. 1217. Based on this study our analysis shows that:

  • 50 percent of the beneficiaries now affected by the GPO would have their benefits increased under the bill, including 29 percent who would come out of offset.
  • Of the women with increased benefits, 64 percent would be wives and 36 percent would be widows. (This parallels the distribution of women now in offset.)
  • 90 percent of those who would benefit from H.R. 1217 have income above the poverty level and over 80 percent have family income over 150 percent of poverty.

I would like to highlight this last point in particular. It demonstrates that H.R. 1217 is not well targeted at low-income elderly beneficiaries, especially widows. There are legitimate issues regarding the need to alleviate poverty among low-income elderly, including those impacted by the GPO. H.R. 1217, however, is not well targeted to address those issues.

Cost of H.R. 1217

SSA's Office of the Actuary estimates that H.R. 1217 would cost approximately $2.3 billion over the first five years and $5.9 billion over the first ten years. The long-range cost is estimated to be negligible (i.e., less than 0.005 percent of taxable payroll.)

Informing the Public

While the GPO has been part of the Social Security law for many years, SSA continues to actively work to inform government employees about it so that they can properly plan their retirement. The annual Social Security Statement advises workers that Social Security benefits may be affected by a pension based on noncovered work and provides the title of the SSA publication, Government Pension Offset, that explains the GPO. This publication, as well as information about the GPO, is also available on SSA's website.

For Federal government employees, SSA works closely with the Office of Personnel Management to provide background materials on the provision including fact sheets and pamphlets. These materials are made available at OPM seminars open to all Federal employees who are within 5 years of eligibility for retirement. SSA also provides the same background materials to State and local government agencies and to unions, such as the American Federation of State, County, and Municipal Employees, for use in informing State and local employees.


Congress had good reason to enact the GPO. It was established to prevent workers who spent a portion of their careers in employment not covered by Social Security from receiving more favorable treatment under Social Security than comparable workers who had worked a lifetime in covered employment.

H.R. 1217 would provide higher Social Security benefits for many career government workers whose pensions from noncovered employment, in combination with their Social Security benefits, are below certain levels. Thus, this bill focuses on providing higher Social Security benefits to public sector retirees, who were not covered by Social Security during their years in government work, simply because their combined public pension and Social Security benefits are deemed to be "too low". But as previously indicated, 90 percent are not poor and 80 percent have family income that is at least 150 percent of poverty.

Finally, I think that at this time, any change in the GPO should only be considered as part of the broader context of long-term reform and extending the solvency of the Social Security program. This is because a change in the GPO would inherently involve changing the relative level of benefits for one group of beneficiaries over another.

I want to again thank the Chairman and the Subcommittee for giving me this opportunity to discuss the GPO and to share SSA's analysis on the legislation before this Subcommittee. As always, I am more than happy to provide assistance to the Members and will be glad to work with you to provide any additional information you request. I would be glad to answer any questions you might have concerning the GPO provision.