Skip Navigation

Mr. Chairman and Members of the Committee, I am delighted to be here today to discuss the implications for women of introducing personal retirement accounts and other changes into the Social Security program.

The Social Security program faces long-term pressures as the baby-boom generation ages into retirement. The current pay-as-you-go system, which relies on payroll taxes from current workers to pay benefits to current beneficiaries, cannot be sustained as the number of retirees per worker climbs in the coming decades. Financial pressures will require significant policy change. The longer reform is delayed, the less time people will have to adapt to the changes, and the bigger the changes may need to be.

Although the potential for increasing economic growth is not generally considered when assessing Social Security reform proposals, it is an important consideration. The size of the economy ultimately determines the nation's ability to support a growing elderly population with fewer workers per retiree. Policies that encourage economic growth could enhance the future living standards of everyone--the elderly, workers, and their children.

The Committee has asked the Congressional Budget Office (CBO) to address the impact of Social Security reform on women. Neither the current program nor proposed reforms include sex-specific rules. Nonetheless, Social Security reform could affect women more than men, on average, because women are expected to earn less and live longer than men. For an individual woman, however, the consequences of Social Security reform will depend on many factors: the specific rules enacted for individual accounts; the specific rules determining defined benefits for workers and their families; the woman's own lifetime earnings; her marital history; and the lifetime earnings of her husband(s), if she has married.

In this testimony, I restrict my focus to the potential impact of Social Security reform on women's income in old age. The impact of reform on Social Security benefits for the disabled and young survivors, though a vitally important consideration, is beyond the scope of this testimony.
 

WOMEN AND THE CURRENT SOCIAL SECURITY PROGRAM

Social Security is a central part of the social safety net for older women today. The progressive schedule for benefits, the auxiliary benefits available for spouses and survivors, and the inflation-protected lifetime annuity payouts all favor women more than men. That is because, on average, women live longer and earn less than men. Women also have lower incomes than men from pensions and other sources. Social Security benefits represented more than half of family income for a majority of women ages 65 and older in 1997 and an even greater share for widows and other unmarried women. For a majority of older women with family income below the poverty line, Social Security represented at least 90 percent of family income.

The changing position of women in the economy must be considered in any discussion of the impact of Social Security reform on women in the future, of course. The movement of women into the paid labor force has been dramatic (see Figure 1). Women born since 1950, in particular, are working in much greater numbers than women born before that year. The participation rate has grown most for women in their childbearing years.
 

FIGURE 1. LABOR FORCE PARTICIPATION RATES OF WOMEN, BY AGE IN FIVE-YEAR BIRTH COHORTS

SOURCE: Congressional Budget Office based on data from the Department of Labor, Bureau of Labor Statistics, Handbook of Labor Statistics, Bulletin 2340 (August 1989) for data prior to 1960, Current Population Survey, 1948-1987, Bulletin 2307 (August 1988) for data between 1960 and 1985, and Employment and Earnings (January 1991 and January 1996) for 1990 and 1995 data.

Two more patterns emerge from Figure 1, however. First, a large minority of women still have little or no attachment to the labor market, despite the major influx of women into the workforce. One out of four women ages 18 to 64 did no work for pay in 1997, and 42 percent of working women that age worked less than full time year-round. Second, increases in women's labor force participation between birth cohorts have tapered off significantly, suggesting that the long-term upward trend in women's participation could be ending. Overall, it appears that women's diverse attachments to the labor market could persist far into the future and that women's attachments will continue to lag behind those of men, on average (see Figure 2).
 

FIGURE 2. LABOR FORCE PARTICIPATION RATES OF MEN, BY AGE IN FIVE-YEAR BIRTH COHORTS

SOURCE: Congressional Budget Office based on data from the Department of Labor, Bureau of Labor Statistics, Handbook of Labor Statistics, Bulletin 2340 (August 1989) for data prior to 1960, Current Population Survey, 1948-1987, Bulletin 2307 (August 1988) for data between 1960 and 1985, and Employment and Earnings (January 1991 and January 1996) for 1990 and 1995 data.

Women's earnings have also grown sharply over time. But on average, women still earn substantially less than men at every age. For full-time wage-and-salary workers ages 16 and older, median weekly earnings of women equaled three-quarters of men's in the first quarter of 1999. The difference between men's and women's year-round earnings has been even wider.

For married women, the division of earnings between two spouses is most important for determining Social Security benefits under current rules. Although wives' contributions to family earnings have also grown significantly over time, most wives still earn much less than their husbands. In 1997, earnings of wives ages 18 to 64 accounted for 31 percent of a couple's earnings, on average.

Because of their lower earnings and greater longevity, women of the future could still rely more heavily than men on their Social Security income, on average. The program could also continue to favor women more than men because of its progressive schedule for benefits, auxiliary benefits for spouses and survivors, and inflation-protected lifetime annuity payouts, if those provisions remain in effect.

Some people consider the current Social Security program to have certain shortcomings for women. From the start, the program has had two conflicting objectives: to provide retirement income linked to workers' payroll taxes, like an individual pension plan, and to provide an adequate level of income for retired workers and their dependent family members who survive them, like a social safety net. More fully serving one of those goals comes at the cost of serving the other less successfully.

The current program--with its individual-based payroll taxes and family-based benefits--provides different amounts of benefits to different workers who have paid exactly the same lifetime payroll taxes, depending on their marital history and their spouse's earnings. A worker who has married may collect either his or her retired-worker benefit or an auxiliary benefit based on his or her spouse's earnings history, whichever is higher. For the spouse of a living retired worker, the full auxiliary benefit is one-half the spouse's benefit; for a survivor, it is the deceased spouse's benefit.

Under those rules, a couple with two earners might actually receive lower monthly benefits than a couple with one earner in return for the same or higher Social Security payroll taxes. For the survivors of such couples, the contrast in benefits can even be greater. If the two couples pay exactly the same total payroll taxes, the surviving spouse of a couple with equal-earning spouses can receive a much smaller benefit than the survivor of a one-earner couple. Because most wives have lower lifetime earnings than their husbands, and because they outlive their husbands and collect survivor benefits, the current program is generally viewed as favoring wives who have worked less over wives who have worked more.

Despite several features of the current program that favor women with low lifetime earnings, many older women also have very low income. The poverty rate for women ages 65 and older was 13.1 percent in 1997--about the same as the rate for the entire U.S. population, but nearly twice the poverty rate for older men. Widowed and other unmarried women fared significantly worse than others.
 

POSSIBLE EFFECTS OF PROPOSED REFORMS

The net impact of Social Security reform on the income of women depends on the returns from private accounts and on how much defined benefit levels change. Many of the proposals to reform Social Security would keep the defined benefit part of the program, but they would require or encourage workers to have personal retirement accounts. Many would also revise the defined benefit program.

Personal Retirement Accounts

Any person's income from a personal retirement account program would depend partly on the program's rules for contributions, investments, and distributions. But some program features could be particularly important for certain women.

Account balances might be distributed as mandatory single-life annuities, for example, so that a worker would receive a steady monthly payment based on his or her account balance until death. Because such a payout arrangement could leave a worker's survivors in poor financial shape, however, a plan might instead require all married workers to purchase joint-survivor annuities. A joint-survivor annuity would pay lower monthly benefits to a couple than a single-life annuity based on the same account balance while both spouses were alive, but the joint-survivor benefits would continue as long as either spouse was alive. A third alternative would allow married workers to buy single-life annuities only if their spouse consented in writing.

Annuities could be entirely optional, of course, which would undoubtedly push up annuity prices for buyers. Because workers who expected to live a long life would be more likely than others to buy annuities, the average payout period would be longer than the average payout in a mandatory annuity plan, and the average monthly payment for each dollar annuitized would be lower accordingly. Administrative costs per annuity could also be higher in an optional plan than in a mandatory plan because fewer accounts would share the fixed costs of administration.

Such decisions about annuities would affect women more than men because women have greater longevity. Moreover, any annuity plan might charge women higher prices than men for single-life annuities for the same reason.

Fees for investment transactions and account maintenance could also affect workers' rates of return. A flat fee for each transaction would leave a smaller percentage of each dollar contribution for actual investment, for example, and the impact of such fees would increase as the fee rose and the size of the contribution declined. Such fees would affect women's rates of return more than men's, on average, because of women's lower average earnings. Conversely, women would benefit more than men, on average, from a fee schedule that was proportional to a worker's contributions or earnings.

Account contributions might also be partly voluntary. Catch-up contributions might be allowed for workers with long spells outside the labor force, for example, or for workers with periods of low earnings. Such provisions would benefit women more than men, on average, because of greater variation in women's labor force participation and earnings over their lifetime.

Different women could respond quite differently to the same investment options, of course, depending on their circumstances outside the program. The same holds for men. Spouses' account decisions could also be linked. Because private account returns are uncertain, CBO cannot conclude whether reform would increase or decrease any one person's total income from Social Security.

Defined Benefits

Proposed changes in the level and distribution of defined benefits could significantly affect the well-being of many women--with or without the introduction of personal retirement accounts. Such changes include direct changes in the rules for women's own retired-worker benefits and for their auxiliary benefits as spouses and survivors. The changes also include indirect changes in the amount of auxiliary benefits that women would receive as spouses and survivors because of changes in the size of their husband's retired-worker benefits.

Some proposed changes would almost surely reduce women's retired-worker benefits by more than men's benefits, on average. For example, basing the retired-worker benefit on the average of a worker's highest 38 years of earnings, rather than the current 35 years, would cut benefit amounts for most workers because more years with low or no earnings would pull down average earnings. But because women's earnings and labor force participation vary more than men's, basing benefits on more years' earnings would reduce women's retired-worker benefits more than men's, on average.

Other proposals would lower women's benefits by less than they would lower men's. For example, proposals to scale back the current retired-worker benefit schedule progressively--so that workers with high lifetime earnings faced larger benefit reductions than workers with low lifetime earnings--would tend to lower women's benefits by less than men's, on average.

Of all proposed changes in defined benefits, the one that could affect older women's well-being the most is a new minimum benefit for surviving spouses equal to 75 percent of the couple benefit that the deceased spouse and survivor would have received. Under that proposal, a surviving spouse could receive the 75 percent survivor benefit, the deceased spouse's retired-worker benefit, or the survivor's own retired-worker benefit, whichever was highest.

Additional proposals for auxiliary benefits would generally reduce benefit amounts. Such changes include a cut in the spousal benefit from one-half to one-third of a retired-worker benefit and increases in the eligibility ages for full and reduced auxiliary benefits. Auxiliary benefit amounts could also be based on lower retired-worker benefit amounts.

All of those changes in the rules for defined benefits could significantly alter the lifetime benefits of many women relative to their lifetime benefits under current rules. But the changes could raise lifetime benefits for some women and lower them for others, and differences in benefits could be large or small, depending on the rules and the earnings histories of the women and their spouses.

On balance, most proposed changes in defined benefits would favor women with a greater lifetime attachment to the labor force relative to women with a weaker attachment. A woman with a stronger attachment to the labor force would probably benefit more from a 75 percent benefit for surviving spouses, for example. She would also be less likely to collect spousal benefits while her husband was alive--even without a cut in spousal benefits.
 

CONCLUSION

In sum, no easy answers are forthcoming to the question: What Social Security reforms are best for women? Women's economic roles vary greatly. Many women have strong attachments to the labor market and high lifetime earnings. But many others have low lifetime earnings. If Social Security benefits declined or became less certain in the future, women--and men--might adjust their work and saving behavior to partly offset any direct effects of reform on their income in old age. Just like today, however, Social Security will still be just one of several factors that influence people's decisions.

Reforms now being considered could shift the balance between Social Security's two competing roles as an individual pension plan and a social safety net. Incorporating individual worker accounts, in particular, could make Social Security more like an individual pension plan. Such a shift could significantly affect the well-being of many women. But the consequences for any particular woman would depend on her own lifetime earnings, marital history, and other individual characteristics--and on the full set of Social Security rules.