LEXIS-NEXIS® Congressional Universe-Document

Search Terms: Government Pension Offset, House or Senate or Joint

Document 10 of 26.

View Related Topics

June 27, 2000, Tuesday

SECTION: PREPARED TESTIMONY

LENGTH: 3216 words

HEADLINE: PREPARED TESTIMONY OF PAUL R. CULLINAN CHIEF HUMAN RESOURCES COST ESTIMATES UNIT CONGRESSIONAL BUDGET OFFICE

BEFORE THE HOUSE COMMITTEE ON WAYS AND MEANS SUBCOMMITTEE ON SOCIAL SECURITY

SUBJECT - SOCIAL SECURITY "GOVERNMENT PENSION OFFSET"

BODY:
Mr. Chairman and Members of the Subcommittee, I am pleased to be here today to discuss an aspect of the current benefit structure of the Social Security program: the government pension offset, or GPO. The provision applies to people with pensions from government employment that was not covered by Social Security, and it curtails the benefits from certain features of the program that those people would otherwise enjoy.

My testimony today focuses on three major questions: What is the offset, and how does it work? Who is affected by the offset? What are the alternatives to current law and their costs?

WHAT IS THE GPO, AND HOW DOES IT WORK?

The GPO was enacted in 1977 and modified somewhat in the Social Security Amendments of 1983. The provision affects benefits for people who meet both of the following criteria: They receive pensions based on employment that was not covered by Social Security (termed noncovered employment)---for example, federal workers who participate in the Civil Service Retirement System (CSRS) and a significant number of state and local government workers; and They are entitled to Social Security benefits because they are the spouse or survivor of an individual entitled to such benefits. In general, in the absence of any offset provision, the spouse of a retired or disabled worker may receive up to 50 percent of that worker's Social Security benefit, and a surviving spouse may receive up to 100 percent. However, if the spouse has an earnings record and receives his or her own worker's benefit, then that benefit reduces spousal benefits dollar for dollar. This offset, known as the dual- entitlement provision, prevents the higher earner in a couple from receiving any auxiliary benefits and limits the spousal benefits paid to the lower earner.

These benefits for wives and widows---often referred to as dependents' benefits--were enacted in the amendments to the 1939 Social Security Act, and they essentially presumed that a married woman would have little or no employment history of her own and would rely heavily on her husband's earnings. (More limited coverage for husbands and widowers was enacted in 1950.) Even in 1939, the assumption about a wife's financial reliance on her husband was not always correct, but the exceptions probably were not glaring.

Developments since 1939 have undermined that assumption. For example, over many decades, women's participation in the labor force has grown, and government employment that was not covered by Social Security expanded significantly during World War II and the postwar period. Thus, a gap in a woman's wage history under Social Security might no longer signal that her husband was the couple's only earner. Moreover, a man who earned a civil service pension was now more likely to be married to a woman who worked enough to become eligible for Social Security benefits in her own right. Still, the Congress did not address those facts until the Supreme Court held--in the 1977 Califano v. Goldfarb case--that the Social Security Act discriminated against men as beneficiaries and women as taxpayers. Women who qualified as wives or widows could get benefits automatically, but men who sought benefits as husbands or widowers had to show that they were previously dependent on their wives for at least one-half of their support. The Goldfarb case opened the door to a flood of claims by men who had civil service pensions or pensions from other noncovered employment and who could not have met the previous dependency test.

The Congress addressed that problem in a gender-neutral fashion by enacting the government pension offset provision. Because the provision was gender neutral, women with pensions from noncovered employment who previously would have encountered no reduction in spousal benefits paid under Social Security now faced an offset to their benefits. The provision reduces (offsets) the spouse's or surviving spouse's Social Security benefit by an amount that is not permitted to exceed two-thirds of his or her noncovered pension. The one-third of the pension excluded from the offset acknowledges that part of the pension is akin to the private pension that many employers provide as a supplement to Social Security and that is not subject to an offset. Effectively, the GPO serves as an alternative to the dual entitlement provision in cases in which one spouse has a significant period of noncovered employment on which a pension is based. (Table 1 shows examples of couples' benefits under the dual-entitlement provision and the GPO.)

The Windfall Elimination Provision

The GPO is not the only provision in the Social Security Act that reduces benefits because someone receives a pension from noncovered employment. The other provision--the windfall elimination provision, or WEP--applies to a retired or disabled worker's own benefit rather than to his or her benefit as a spouse or survivor. Only a few pension annuitants are affected by both provisions, but many annuitants find them confusing.

The WEP was enacted because many government workers have blended careers that qualify them for both a government pension and Social Security in their own right. Today, a 62-year-old needs just 40 quarters, or 10 years, of covered earnings of about $3,000 a year to qualify for a small retired-worker benefit under Social Security. The first step in computing a Social Security benefit is to index the worker's past earnings to today's dollars, pick the top 35 years of earnings, and average them. That step cannot distinguish, however, between a person who toiled for 35 years at low wages and someone who spent 25 years in the federal civil service and 10 years in a second career covered under Social Security.

TABLE 1. HYPOTHETICAL EXAMPLES OF SOCIAL SECURITY BENEFITS WITH AND WITHOUT THE GOVERNMENT PENSION OFFSET (In dollars) Spouse 1 Spouse 2

Case 1: Couple with One Earner in Covered Employment

Social Security benefit based on own earnings 1,000 0 Social Security benefit based on spouse's earnings 0 500 Pension based on covered employment 500 0 Pension based on noncovered employment 0 0

Total Social Security/pension income 1,500 500

Case 2: Couple with Two Earners in Covered Employment

Social Security benefit based on own earnings 1,000 500 Social Security benefit based on spouse's earnings/a 0 0 Pension based on covered employment 500 200 Pension based on noncovered employment 0 0

Total Social Security/pension income 1,500 700

Case 3: Two-Earner Couple with Higher Earner in Noncovered Employment (No GPO in effect)

Social Security benefit based on own earnings 0 500 Social Security benefit based on spouse's earnings 250 0 Pension based on covered employment 0 200 Pension based on noncovered employment 1,500 0

Total Social Security/pension income 1,750 700

Case 4: Two-Earner Couple with Higher Earner in Noncovered Employment (GPO in effect)

Social Security benefit based on own earnings 0 500 Social Security benefit based on spouse's earnings 0 0 Pension based on covered employment 0 200 Pension based on noncovered employment 1,500 0

Total Social Security/pension income 1,500 700

SOURCE: Congressional Budget Office.

a. No spousal benefits are payable to either spouse because the dual- entitlement provision totally offsets the benefit.Specifically, for a worker who turns 62 in 2000, the primary insurance amount (PIA), on which all Social Security benefits are based, is computed using a three-bracket formula: 90 percent of the first $531 of average indexed monthly earnings; plus 32 percent of such earnings between $531 and $3,202; plus 15 percent of earnings above $3,202.

If our hypothetical retiree spent 30 years in the federal civil service and 10 years in covered employment at an average indexed salary of $42,000, his or her covered earnings--averaged over 35 years--would be $12,000 a year, or $1,000 a month. Under the formula, his PIA would be $628. And that would be on top of what could be a substantial civil service pension.

The windfall elimination provision scales back the 90 percent factor in the first bracket of the benefit formula for workers who have pensions based on noncovered employment. If a worker spent 20 years or less in covered work, the 90 percent factor is cut to 40 percent. For 21 through 29 years of covered work, that percentage (40 percent) rises by 5 points per year. Finally, the recipient of a noncovered pension who nevertheless spent 30 or more years in covered work is entitled to the regular 90 percent factor and is thus exempt from the WEP. For the purposes of the provision, a year of covered work in 2000 requires earnings of at least $14,000. In the case of our hypothetical retiree, the WEP would reduce his or her PIA by $265. Roughly one-half million Social Security recipients are currently affected by the WEP, and that number is growing by about 60,000 annually.

Features of the Social Security Program That Give Rise to the GPO and the WEP

The perceived need for provisions such as the GPO and the WEP arises from three characteristics of the Social Security program: Not all workers are covered under Social Security, and the majority of noncovered workers are employees of federal, state, or local governments. Social Security provides benefits to spouses and survivors of retired and disabled workers without reducing those workers' own benefits. The benefit formula is weighted to replace a greater percentage of earnings for beneficiaries with low lifetime earnings.

Coverage. According to the Social Security Administration (SSA), about 96 percent of the workforce is employed in jobs covered by Social Security. The other 4 percent is mostly employed by state and local governments (which account for about one-half of those remaining workers, or over 3 million employees) or by the federal government (which employs about 15 percent, or roughly 1 million people). Those figures represent a single year, and they understate the percentage of workers who move between covered and noncovered employment during their lifetime. Therefore, many workers who receive pensions from noncovered employment also work long enough in covered employment to gain insured status under Social Security. Indeed, a 1997 study published in the Social Security Bulletin indicated that over two- thirds of federal civil service retirees ages 65 to 69 were also entitled to Social Security benefits on the basis of their work history.

Auxiliary Benefits. Social Security, unlike most private and public pensions, pays additional amounts to the families of insured workers without reducing the workers' own benefits. As noted earlier, the spouse of a retired or disabled worker is eligible to receive a benefit equal to one-half of the basic Social Security benefit paid to that worker, and a survivor can receive the entire benefit. But with a typical private or public pension, a worker must elect a "joint and survivor" annuity to ensure benefits for a widow or widower and must accept, in mm, a smaller benefit over his or her own lifetime.

Weighting of the Benefit Formula. As described earlier, Social Security benefits are calculated by using a three-bracket formula that translates average indexed monthly earnings over a 35-year period into a basic benefit amount, or PlA. That formula is progressive, and the weighting is designed to help low-wage workers. However, the formula also provides an advantage to the annuitant who moves into covered employment and qualifies for retirement benefits under Social Security on the basis of a fraction of his or her working life.

Each of the above features of the Social Security system contributes to the argument that noncovered government pensions should be factored into the calculation of Social Security benefits. For example, if Social Security coverage were universal, the existing dual entitlement provision would serve the same purpose as the GPO. If there were no benefits for spouses or surviving spouses, the marital status of the person with the pension from noncovered employment would be irrelevant in figuring Social Security benefits. And because the benefit formula is weighted rather than strictly proportional to earnings, workers with low average earnings have an advantage, regardless of whether the average reflects a lifetime of low-wage employment or a short period of covered work.

WHO IS AFFECTED BY THE GPO?

Because Social Security coverage is so extensive, the government pension offset does not affect a significant proportion of the program's roughly 45 million recipients. But for those who are affected, the provision makes a huge difference. The Congressional Budget Office's (CBO's) tabulations of data supplied by SSA indicate that at the end of 1999, the GPO was reducing Social Security benefits for about 305,000 people (see Table 2). In about 225,000 cases, or three-quarters of that total, the GPO caused potential benefits to be completely withheld; the average loss was $353 per month. The remaining 80,000 recipients continued to be paid some benefits even after the GPO was applied, with the reduction averaging $336 per month.

TABLE 2. SOCIAL SECURITY CASES AFFECTED BY THE GOVERNMENT PENSION OFFSET, DECEMBER 1999 (NOTE: Table not transmittable)

SOURCE: Congressional Budget Office based on the One-Percent Monthly Old-Age, Survivors, and Disability Insurance Sample file, December 1999.

What the data do not reveal are those annuitants with pensions from public-sector jobs who do not apply for Social Security because they are aware that the GPO would wipe out any spousal benefits to which they would otherwise have been entitled. Therefore, any counts of people affected by the GPO and any estimates of the costs of altering it are likely to understate the potential effects of such changes. Moreover, that understatement could be substantial for major changes such as a complete repeal of the GPO.

CBO's analysis of the SSA data revealed some differences in how the GPO affected men and women. Seventy percent of the affected beneficiaries reflected in the data were women, and women were more likely than men to experience a reduction (rather than a complete withholding) of their benefit. Benefits were totally withheld for almost two-thirds of the women affected by the GPO; the comparable figure for men was 98 percent. Partial offsets were much more likely to occur for widows or widowers (41 percent) than for spouses (17 percent), simply because Social Security provides larger benefits to survivors.

The SSA data do not reveal the source of the government pension paid to people who are subject to the offset. But by matching records with the Office of Personnel Management for June 1996, SSA researchers found that the GPO reduced Social Security benefits for 116,000 former federal employees who received civil service pensions. That figure appeared to represent a little less than half of the people who were affected by the GPO in 1996.

The number of people subject to the GPO has climbed by about 15,000 (or about 6 percent) per year. Over time, that growth will fade since all federal civil servants hired after 1983 (and some who were hired earlier and opted to switch) are covered by Social Security in combination with the Federal Employees' Retirement System. But the slackening of growth in the number of affected people will take many decades to play out.

WHAT ARE THE ALTERNATIVES TO CURRENT LAW AND THEIR COSTS?

The government pension offset is a relatively simple way to trim spousal benefits paid to some two-career couples, but it has many detractors. Some critics believe that it reneges on the implied promise of Social Security, under which benefits accrue from working and paying taxes and those benefits are an "earned right." Others contend that it is unfair to treat pensions from noncovered employment differently from other pensions or even from dividends, interest, and royalties. Still others argue that although it might be desirable to limit spousal benefits for people with relatively large government pensions, the GPO distinguishes the percentage reduction neither by the size of the pension nor by the total income of the annuitant.

Approaches other than the GPO may be feasible for addressing the underlying problems it was meant to solve, but such approaches cost money and often introduce new complexities or inequities. One obvious alternative is to eliminate the GPO altogether. The drawbacks to that option are its cost and, to many analysts, the inappropriateness of paying full benefits to people who worked only a fraction of their careers in covered employment (see Table 3). Eliminating the GPO would cost at least $21 billion (probably substantially more) over the next 10 years, CBO estimates, and would increase the 75-year shortfall in the Social Security trust funds by 0.03 percent of taxable payroll. Although that increment is small in comparison with the overall imbalance in the program (1.89 percent of taxable payroll), repealing the GPO would make it more difficult for the country to address the looming fiscal burden posed by the aging of the baby boomers.

Another approach might seek to shield lower-income beneficiaries from the full brunt of the offset. H.R. 1217, as introduced by Representative William J. Jefferson, is an example of that strategy. The bill would base the reduction in benefits on the combined amount of Social Security and the noncovered pension; because SSA now collects that same information to calculate the offsets, no new data collection system would have to be developed. Other income, such as interest and dividends, would be irrelevant in calculating benefits-- just as it is in such calculations for other beneficiaries.

TABLE 3. ESTIMATED COSTS OF VARIOUS ALTERNATIVES TO THE CURRENT GOVERNMENT PENSION OFFSET (By fiscal year, in billions of dollars) (NOTE: Table not transmittable)

SOURCE: Congressional Budget Office.

NOTES: The proposals would take effect in January 2001.

These estimates all understate the true costs of the options since the estimates do not take into account the options' effects on people who do not apply for benefits because of the GPO. That understatement is likely to be substantial for the repeal option.

H.R. 1217 would limit reductions under the GPO to beneficiaries whose total unreduced Social Security benefit and government pension exceeded $1,200 a month. Under that provision, about 25 percent of those affected under current law would be exempt from the GPO, and another 20 percent would see a boost in their benefits. CBO estimates that H.R. 1217 would increase Social Security outlays by at least $4.9 billion over the 2001-2010 period; the Social Security actuaries estimate that, over the next 75 years, the costs would be negligible (less than 0.005 percent of taxable payroll).

Lowering the maximum offset to one-half rather than two-thirds of the noncovered pension is another alternative. That change would increase benefits for about two-fifths of those affected under current law and add at least $2 billion in costs to the program over the next decade. The long-term impact would be insignificant.

END

LOAD-DATE: June 28, 2000

Document 10 of 26.


Search Terms: Government Pension Offset, House or Senate or Joint
To narrow your search, please enter a word or phrase: