Poverty-based Supplemental Security Income

Supplemental Security Income (SSI) is a poverty-based social assistance program in which eligibility is based on income and assets. Since it was created in 1972, SSI has provided monthly cash benefits to poor people who are age sixty-five and older, blind or disabled. Women live longer than men and are more likely to be poor, thus nearly two-thirds of older SSI recipients are women (Social Security Administration, 2010). SSI benefits have been indexed to cost of living increases since the implementation of the program, and have risen slowly each year. These benefits do not raise many older people above the poverty line, however, because maximum federal benefits are equal to about 77 per cent of the federal poverty line for single beneficiaries and 87 per cent of the federal poverty line for older couples (Social Security Administration, 2010). Benefits for those who are not living independently, such as an older woman living with her adult daughter, are reduced by one-third under the assumption that other household members will provide in-kind benefits, and that two can live almost as cheaply as one.

Like most poverty-based programs in the US, SSI is underused. Only about one-third of poor older people, and only one-half of those who are SSI eligible, actually receive benefits. SSI use among those age sixty-five and older has dropped from 10 per cent in 1975 to a current all-time low of 3 per cent (Social Security Administration, 2010). Some of the decline is due to expansion of Social Security and the decrease in old age poverty over this period of time. But some of the decline is due to features of the SSI program itself (Harrington Meyer and Herd, Reference Harrington Meyer and Herd2007). Some older people who are eligible for SSI do not apply because they are unaware of the benefits, overwhelmed by the cumbersome eligibility forms or discouraged by assets and earnings tests that have not been updated in decades. SSI asset tests are quite strict. Assets, excluding a house, car and burial funds under certain conditions, must be below $2,000 for individuals and $3,000 for couples. SSI's asset maximums have been frozen in place since 1989 (Social Security Administration, 2010). Neither the income disregard nor the asset test has been linked to cost of living and neither can be raised except by congressional legislation.

Unlike Social Security, which has been easing the earnings test for older people, SSI has maintained a strict earnings test that considers all earned and unearned income in calculating benefits. Under federal guidelines the first $65 in earned income, along with an additional $20, is disregarded each month. These modest income disregards have been frozen in place since 1981 (Social Security Administration, 2010). Any additional earned income, beyond the first $65, decreases benefits by $1 for every $2 earned. Unearned income decreases benefits by $1 for every $1 received. Unearned income includes Social Security and gifts or payments from family and friends. Thus a grandmother who is being paid to babysit her own grandchildren is likely to have her benefits reduced by $1 for every $1 she receives for her work.

Both the asset and earnings tests could be indexed to inflation. When poverty based programs are left to languish in this way, older people who rely on these programs face even more difficult financial circumstances and have little choice but to turn to family members for assistance. Means-tested welfare programs are politically unpopular and, therefore, politically vulnerable. Moreover, they are often underused, rarely indexed to inflation and rarely effective at pulling people out of poverty (Korpi and Palme, Reference Korpi and Palme1998; Korpi, Reference Korpi2000; Quadagno, Reference Quadagno2005; Harrington Meyer and Herd, Reference Harrington Meyer and Herd2007).

Employer-based private pensions

Historically, employer-based private pensions have been mainstays of US social welfare provision. Private pensions are the third largest income stream, accounting for 18 per cent of all old age income (Social Security Administration, 2010). Most Americans think of these pensions as employer expenses, but in fact, they are funded in part by forgone taxes totaling over $100 billion a year. The receipt of a private pension often reduces the likelihood of poverty for older families. But pensions have become less reliable and pension coverage has shrunk dramatically. Increasingly, private pension income for many older individuals ranges from little to none (Selden and Gray, Reference Selden and Gray2006; Harrington Meyer and Herd, Reference Harrington Meyer and Herd2007; Butrica et al., Reference Butrica, Johnson and Smith2011).

The reliability of private pensions changed dramatically with the shift from defined benefit to defined contribution plans. Historically, most employees were covered by defined benefit plans, and benefits increased with hours, salaries and years of service. Between 1979 and 2004, the share of workers covered only by defined benefit plans dropped from 62 per cent to 10 per cent and the share covered only by defined contribution plans rose from 16 per cent to 63 per cent (Employee Benefit Research Institute, 2005). Now most workers are covered by a defined contribution plan and the value of benefits will be determined by subsequent contributions, investment decisions and a fair amount of luck.

Defined contribution plans shift the responsibility risks, and most costs associated with private pension accumulation from employers to employees. Over time, employers are contributing less to these plans, making employees even more responsible for their own pension accumulations (Munnell and Sundén, Reference Munnell and Sundén2004; Shuey and O'Rand, Reference Shuey and O'Rand2006). Moreover, because they are not as compulsory or as automatic as defined benefit plans, employees with defined contribution plans are more readily able to opt out, reduce their contributions or withdraw money before retirement to cover education, health expenses or family obligations. Lower income workers are more likely to do all three and thus to be less economically secure in old age (Munnell and Sundén, Reference Munnell and Sundén2004; Shuey and O'Rand, Reference Shuey and O'Rand2006).

Access to private pensions, which has long been unequal by income, gender and race, is becoming more unequal as employers tighten the reins on this employee benefit. Today 70 per cent of workers in the upper one-third of earnings work for an employer who offers a private pension, compared to only 30 per cent of workers in the lowest one-third of earners (Karamcheva and Sanzenbacher, Reference Karamcheva and Sanzenbacher2010). Take-up rates also vary by earnings bracket. Nearly all high earners participate, but only 69 per cent of lower earners did in 2008. The overall impact is that households of older individuals in the top one-third of the income distribution receive 31 per cent of their non-earned income from private pensions, whereas households in the bottom one-third receive just 3 per cent (Karamcheva and Sanzenbacher, Reference Karamcheva and Sanzenbacher2010).

Women have always had smaller pensions and the inequality may be growing (Munnell and Sundén, Reference Munnell and Sundén2004; Shuey and O'Rand, Reference Shuey and O'Rand2006). Across the life course, women are less likely than men to be employed, and when employed they work fewer hours at lower pay. Thus women are only 60 per cent as likely as men to receive private pensions and their pensions are just one-half of the value of men's. The size of women's defined contribution plans lags well behind the size of their defined benefit plans. Among those born between 1931 and 1941, women's average defined benefit savings were 77 per cent of men's while their defined contribution plan account sizes were just 37 per cent of men's (Johnson, Reference Johnson1999). One benefit simulation estimates that women's pension wealth in defined contribution plans will be 25 per cent lower than in traditional defined benefit plans, and 50 per cent lower in 401(k) plans than in traditional defined benefit plans (Even and Macpherson, Reference Even and Macpherson2004). Thus gender inequality in private pension income is expected to increase for future cohorts.

Blacks and Hispanics have always had lower pensions and pension coverage as well. Across the life course, they are more likely than whites to be unemployed, underemployed, incarcerated, chronically ill and in lower paying jobs (Shuey and O'Rand, Reference Shuey and O'Rand2006). The race gap in private pensions had been closing under defined benefit plans but has widened again under defined contribution plans (Harrington Meyer and Herd, Reference Harrington Meyer and Herd2007; AARP, 2011). Among those aged fifty and older in 2009, 35 per cent of whites and blacks and 22 per cent of Hispanics were offered defined benefit plans, and 34 per cent of whites and blacks and 20 per cent of Hispanics participated. By contrast, 54 per cent of whites compared to 47 per cent of blacks and 37 per cent of Hispanics were offered defined contribution plans, and 44 per cent of whites compared to 33 per cent of blacks and 26 per cent of Hispanics participated (AARP, 2011). Thus, with the shift to defined contribution plans, pension coverage and size is down more for blacks and Hispanics than whites.

Moreover, private pensions have become much more meager, and much less secure, with the onset of the 2007−09 depression, the worst economic downturn in the US since the Great Depression. Butrica et al. (Reference Butrica, Johnson and Smith2011) estimate that for those aged fifty-five to fifty-nine in 2008, the Great Recession will reduce average income when they are seventy by 5 per cent. Looking ahead, estimates of the percentage of households at risk of being unable to sustain their pre-retirement lifestyles during retirement rose from 43 per cent in 2004 to 51 per cent in 2009, largely due to the recession (Munnell et al., Reference Munnell, Webb and Gulab-Sass2009). Thus employer-based benefits tend to help only those with sufficiently strong links to the labour force to receive the benefits even during economic downturns (Harrington Meyer and Pavalko, Reference Harrington Meyer and Pavalko1996; Korpi and Palme, Reference Korpi and Palme1998). Far from reducing inequality in old age, private pensions tend to exaggerate that inequality.

Nearly universal Social Security

Social Security is a nearly universal, contribution-based, social insurance program. Unlike targeted programs that emphasise selective acceptance, universal programs boast broad eligibility, redistributive benefit formulas and widespread approval by voters (Estes and Associates, Reference Estes and Associates2001; AARP, 2005; Public Agenda, 2005; Quadagno, Reference Quadagno2005). Universal programs are more effective than targeted programs at reducing poverty and inequality, particularly when the benefit formula is redistributive. Social Security replaces 78 per cent of income for the lowest earners, compared to about 28 per cent of income for the highest earners (Center on Budget and Policy Priorities, 2010). Indeed, Social Security is credited with reducing poverty among older adults from over 45 per cent to 9 per cent today (Center on Budget and Policy Priorities, 2010; Administration on Aging, 2011). Social Security provides monthly benefits to millions of retirees, their spouses, and people who are permanently disabled. The program initially excluded workers in many occupations, including agriculture and domestic service, as well as self-employed workers and those employed by religious, charitable and educational organisations (Quadagno, Reference Quadagno1984). These exclusions eliminated from coverage nearly one-half of all workers and nearly all women and blacks. Over time Social Security incorporated most of these workers and by 2009 over 94 per cent of workers were covered and 91 per cent of older people were receiving benefits (Social Security Administration, 2010).

Although initially designed as an income supplement, Social Security is the primary source of income for many older Americans and a particularly important source of income for women, blacks, and Hispanics. Social Security comprises about 40 per cent of income for all older people, 60 per cent of income for all older women and 100 per cent of income for one in five older women (Social Security Administration, 2010). Because of reduced access to private pensions and private savings, one-half of older Hispanics and African Americans rely on Social Security for 90 per cent or more of their income (Torres-Gil et al., Reference Torres-Gil, Greenstein and Kamin2005; Wu, Reference Wu2005). Social Security comprises almost 80 per cent of the income of black and Hispanic women. Social Security also provides an important actuarial advantage for women. In the US, women live an average five years longer than men, and their benefits are not reduced to offset these additional years of benefits (Harrington Meyer and Herd, Reference Harrington Meyer and Herd2007). Thus, Social Security has transformed from a system that once excluded women and blacks, to one that provides a majority of their income in old age.

Retired worker benefits

Retirees may be eligible for Social Security retired worker benefits or spouse and widow benefits. To be eligible for worker benefits, workers must have worked in covered employment for a total of ten years, or forty quarters, over their life course (Social Security Administration, 2010). The quarters need not be consecutive, so eligibility is not compromised when workers disrupt waged labour to rear children or perform other unpaid care work. Such disruptions do, however, affect benefit size. Maximum benefits, $2,513 a month in 2012, are available only to those with lengthy and continuous labour force participation in higher paying jobs. Those with interrupted participation and lower paying jobs receive much less. Blacks and Hispanics, who are more likely to struggle with higher unemployment and incarceration rates, lower education levels, and lower wages, receive smaller retired worker benefits than whites on average (Harrington Meyer and Herd, Reference Harrington Meyer and Herd2007). Women, who are more likely to juggle the conflicting demands of waged and unwaged labour and persistent gender segregation and discrimination in the labour force, receive significantly lower benefits than men (Harrington Meyer and Herd, Reference Harrington Meyer and Herd2007; Social Security Administration, 2010). Even though the Social Security benefit formula is designed to redistribute from higher to lower paid workers, average retired worker Social Security benefits for black men are 83 per cent of those for white men. Average retired worker benefits for white women are 76 per cent, and average benefits for black women are just 69 per cent, of those for white men (Social Security Administration, 2010).

Spouse and widow benefits

Eligibility for spouse and widow benefits is determined by marital rather than work history. Although the rules are gender neutral, nearly 97 per cent of spousal beneficiaries are women (Social Security Administration, 2010). Spouse and widow benefits are not available to the growing numbers of both homosexual and heterosexual couples that cohabitate without marriage. Currently about 60 per cent of older women receive spouse or widow, rather than retired worker, benefits. Despite advances in the labour force, the majority of women are expected to continue to rely on spouse and widow benefits for the next several cohorts as they reach old age (Levine et al., Reference Levine, Mitchell and Phillips2000; Social Security Administration, 2010). Spouse beneficiaries must be married at the time of eligibility, or, if divorced, have been married at least ten years to a covered worker. Women receiving spouse benefits receive an amount equal to 50 per cent of their husbands’ benefits, those receiving widow benefits receive an amount equal to 100 per cent of their husbands’ benefits. Race differences in men's wages translate into race differences in women's spouse and widow benefits. Black women's average spouse benefits are 82 per cent, and their widow benefits just 78 per cent, of white women's (Social Security Administration, 2010).

Many women are dually entitled, that is they are entitled as workers and as spouses or widows. The Social Security Administration automatically gives them whatever is the largest benefit. Historically, spouse and widow benefits have provided much-needed retirement income for women with low wages or interrupted labour. But access to these noncontributory benefits is linked to marital status, and currently there is both a retreat from marriage and a growing race gap in marriage. By the time women born in the 1960s reach old age, 80 per cent of white and Hispanic women, compared to only 50 per cent of black women, will be eligible to receive benefits based on marriage, widowhood and divorce (Harrington Meyer et al., Reference Harrington Meyer, Wolf and Himes2006; Iams and Tamborini, Reference Iams and Tamborini2012). This means that older black women, who already have fewer economic resources and higher poverty rates, may be especially vulnerable in old age in future cohorts.

Fiscal considerations

Social Security benefits are funded through a regressive payroll tax system. In 2010, workers paid a flat tax of 6.2 per cent on earnings, matched by employer, up to the Social Security tax cap of $106,800 (Social Security Administration, 2010). Earnings above the ceiling are not subject to Social Security taxes, thus lower income earners contribute a higher proportion of their total earnings to Social Security than higher income earners. Because of race differences in annual family income, a higher percentage of blacks and Hispanics pay the tax on all of their income. In 2006, 27 per cent of white and 27 per cent of Asian, compared to 13 per cent of black and 11 per cent of Hispanic families, had income above $100,000 that was, therefore, untaxed (US Census Bureau, 2008).

Social Security faces a fiscal shortfall in the next few decades. It is a pay-as-you-go system in which workers pay in through the FICA tax and the money is paid out to older beneficiaries. Aware that cohorts come in different sizes, and that aging baby boomers were going to draw sizable benefits, the Social Security Administration purposely began accumulating a trust fund surplus decades ago. In 2012, that surplus is $2.7 trillion and it is expected to reach $3.1 trillion by 2020 (Ruffing and Van de Water, Reference Ruffing and Van de Water2009; Reno et al., Reference Reno, Walker and Bethell2012). As more baby boomers make claims, the surplus will begin to dwindle. In 2017, the predicted ratio of workers to those collecting Social Security benefits will be 2.7 to 1, down from 3.3 to 1 in 2007 (Board of Trustees, 2009). By 2033, it is estimated that the surplus will be gone and the amount coming in from workers each month will cover only 75 per cent of the payouts (Center of Budget and Policy Priorities, 2010; Reno et al., Reference Reno, Walker and Bethell2012). The funds in the surplus have been borrowed by the US government for expenditures unrelated to Social Security, however, and those loans will need to be repaid with interest to the Social Security trust funds.

The Social Security shortfall can be readily addressed without reducing benefits to recipients. Congress could raise the Social Security payroll tax by 1 per cent, matched by a similar 1 per cent increase for employers (AARP, 2005; Harrington Meyer and Herd, Reference Harrington Meyer and Herd2007; Social Security Administration, 2010; Center on Budget and Policy Priorities, 2010). Congress could also raise the cap on taxable earnings to at least $180,000 and index it to inflation. Currently the ceiling on payroll taxes, $106,000, covers just 84 per cent of earnings. The proposed higher cap would cover 90 per cent of earnings. Finally, Congress could broaden the eligibility base to include the 30 per cent of state and local government employees who are currently not participating in Social Security. This is a temporary fix in that these employees would contribute now but eventually they would also draw benefits (AARP, 2005; Harrington Meyer and Herd, Reference Harrington Meyer and Herd2007; Social Security Administration, 2010). These proposals would provide fiscal solvency for Social Security as aging baby boomers draw benefits without reducing benefits to older recipients.

Concerns about the fiscal shortfall, coupled with the recognition that life expectancy has risen dramatically since the inception of the program, have already led policy makers to implement changes aimed at encouraging people to delay taking Social Security benefits. Beginning in 2003, the age of eligibility for full retirement benefits has been gradually increasing from age sixty-five to age sixty-seven, where it will be in 2027. The age for early retirement remains sixty-two, but the penalty for taking early benefits, which most men and nearly all women take, has increased from 20 per cent to 30 per cent (Social Security Administration, 2010). Hardest hit are those who are unable to continue working due to poor health or unemployment. Women and blacks and Hispanics are more likely to be affected by both (Moon and Herd, Reference Moon and Herd2002; AARP, 2011). Whereas those with good health who are in jobs that are not physically demanding may be able to continue working until sixty-seven, those that are not able to do so find that they have to make do with fewer economic resources.

Another policy that attempts to cut costs by encouraging people to delay taking Social Security benefits is that the Social Security earnings test penalises wages among early retirees. Aware that many older persons need or want to continue working, policy makers have relaxed the earnings test over the decades. Currently, there is no earnings test for those aged sixty-five and older, but those aged sixty-two to sixty-five lose $1 in benefits for every $2 in earnings after earning $14,000 for the year (Social Security Administration, 2010). The penalty on wages for early retirees tends to affect women more as they tend to take their Social Security benefits earlier than men, often timing their receipt of benefits to coincide with their slightly older husbands.

Privatisation of social security?

Recent debates about Social Security have focused on privatising the program. Indeed, after he was re-elected in 2004, President George Bush made the privatisation of Social Security his top domestic priority (Schulz and Binstock, Reference Schulz and Binstock2006). The plan most often discussed would have allowed those under age fifty-five to divert as much as 4 per cent of their 6.2 per cent Social Security contribution to private accounts (President's Commission to Strengthen Social Security, 2001; Herd and Kingson, Reference Herd, Kingson and Hudson2005). Supporters of privatisation argued that the existing program would soon be fiscally unsustainable and proposed individual accounts that would maximise each person's choice and responsibility. Opponents pointed out that the kind of privatisation being proposed would make the fiscal situation worse, while increasing program costs and exposure to risk for older people and their families. Moreover, many of the features that make the existing program advantageous for lower income older people, particularly women, such as the spouse and widow benefits, the redistributive benefit formula and the actuarial advantage would most likely disappear or be reduced under privatisation (Harrington Meyer and Herd, Reference Harrington Meyer and Herd2007). The proposal was discussed continuously but never introduced in Congress, in part because downward trends in the stock market made voters very nervous about considering such investments as their primary source of old age income.

The payroll tax cut

At the end of 2010, a time when proponents of Social Security were asking for a 1 per cent increase in the Social Security payroll tax to finance benefits for the baby boomers, President Obama and the liberal and conservative wings of Congress instead passed a payroll tax decrease. Defined as a payroll tax holiday, the law reduced the individual payroll tax from 6.2 to 4.2 per cent for the year 2011, but kept the employer contribution at 6.2 per cent. For a worker earning $50,000 a year, this tax cut would result in a $1,000 increase in take-home pay. The impetus was intended to jump start the economy and create more jobs by putting $120 billion back in consumer's pockets (Bethell, Reference Bethell2010). In subsequent legislation, the payroll tax holiday was then renewed for 2012. The government has promised that this will not affect Social Security's fiscal solvency because Congress will pay the funds back through a transfer of general tax revenues. However, some fear that Congress will not actually transfer those funds (Reno et al., Reference Reno, Walker and Bethell2012). Moreover, many supporters of Social Security fear the payroll tax holiday will become a permanent reduction. Nancy Altman, co-director of Social Security Works, suggests that if the payroll tax cut does indeed become permanent, ‘it's good-bye, Social Security’ (Bethell, Reference Bethell2010). Social Security already faces significant fiscal shortfalls in the decades ahead. A reduction in the payroll tax could expedite the arrival of the shortfall. Any retrenchment in Social Security will increase inequality and insecurity among the elderly in general, and among the older women, blacks, Hispanics and single persons who rely on it disproportionately, in particular.

A new minimum benefit?

In its first half century, Social Security was expanded numerous times: benefits were extended to wives and widows, formulas were made gender neutral and strict earnings tests were relaxed. Supporters of the program have called for additional expansions that would help families juggle work and children (see Harrington Meyer and Herd, Reference Harrington Meyer and Herd2007). Many scholars suggest that Social Security does not effectively compensate women for time spent out of the labour force caring for family members (Quadagno, Reference Quadagno1984; Harrington Meyer and Herd, Reference Harrington Meyer and Herd2007). All workers are allowed to drop the five lowest years of earnings, but many women spend more than five years out of the labour force and in lower paying jobs while they are performing unpaid care work (Harrington Meyer and Herd, Reference Harrington Meyer and Herd2007; Social Security Administration, 2009). In contrast to many EU countries, the US has not seriously entertained policy changes that would help families juggle paid and unpaid work, such as care credits or a generous minimum benefit (Harrington Meyer and Herd, Reference Harrington Meyer and Herd2007; Social Security Administration, 2010).

One solution would be a new minimum benefit. The US had a minimum benefit previously, but it was terminated in the early 1980s. Today, Social Security has a special minimum benefit, but few qualify and the benefits are meager. Some policy analysts favour a plan that functions like the Earned Income Tax Credit (EITC), so that each year when older people with lower incomes file their tax returns, they receive additional Social Security benefits. Others prefer a plan in which the minimum benefit slides up with the number of years of contributing to the system, and, in the European tradition, in which a care credit offsets time out of the labour market (Favreault and Steurle, Reference Favrerault and Steurle2007). The simplest plan, however, provides a minimum benefit to all who are eligible for Social Security (Harrington Meyer and Herd, Reference Harrington Meyer and Herd2007). If that minimum was set near the poverty line, it could nearly eliminate poverty among older people. It would also reduce the need for spouse benefits or Supplemental Security Income (SSI), and such offsets would help pay for the benefit expansions. A minimum benefit would also reduce reliance on marital history to determine eligibility for benefits, reducing the current inequalities linked to the growing race gap in marital rates and the federal government's failure to recognise gay marriage.

Attention to these policies would make Social Security more responsive to changing American families by broadening the program's capacity to shore up economic security in old age and by reducing inequalities linked to gender, race, class and marital status. But critics have argued that with the aging of the baby boom generation, the country can no longer afford such universal and redistributive polices and that the time has come to shift greater amounts of risk and responsibility away from the welfare state and markets and to older people and their families (Yergin and Stanislaw, Reference Yergin and Stanislaw1998; Harrington Meyer and Herd, Reference Harrington Meyer and Herd2007).