Using data for couples to project the distributional effects of changes in Social Security policy

Under Social Security program rules, the aged receive Social Security benefits either as retired workers, spouses, divorced spouses, or widow(er)s. Retired-worker benefits are paid to workers who have 40 quarters of coverage over their lives. Auxiliary benefits are paid to spouses, divorced spouses, and widow(er)s of retired workers. Spouse benefits are computed using the earnings history of the current spouse for individuals who are married when they apply for benefits. Divorced spouse and widow(er) benefits are computed using the earnings history of the ex-spouse or deceased spouse with the highest PIA. A large number of retired women are entitled to auxiliary benefits. Some women receive only auxiliary benefits, while the majority of women have their retired-worker benefit supplemented by auxiliary benefits. Because the level of Social Security benefits can reflect the relative lifetime earnings of both spouses, as a couple, using individual data to estimate Social Security benefits will tend to underestimate actual benefits, particularly for women. However, detailed data for couples are often difficult to obtain. There is currently no known single data source that includes both marital and earnings history information. As a result, many researchers resort to estimating Social Security benefits using individual data or aggregate data, such as the average earnings of men and women. The Social Security Administration's Office of Research, Evaluation, and Statistics, with substantial assistance from the Brookings Institution, the Urban Institute, and the RAND Corporation, is developing a model that overcomes this problem by using the marital and earnings histories of both marital partners to estimate Social Security benefits. The Modeling Income in the Near Term (MINT) model projects retirement income (Social Security benefits, pension income, asset income, and earnings of working beneficiaries) from 1997 through 2031 for current and future Social Security beneficiaries using a unique data source--the Survey of Income and Program Participation (SIPP)--matched to Social Security Administration records. Using MINT data, this article establishes the importance of using data for couples rather than individuals by examining the impact of changing Social Security benefits to reflect 40 years of lifetime earnings rather than the 35 years required under current law. We compare the effect of this policy change on married women by estimating their benefits with data for couples and with individual data. Results indicate that: Using individual data overestimates the projected reduction in retirement benefits brought about by the policy change and makes the effects on women look more severe than they actually are. Because older birth cohorts are more likely than younger cohorts to receive auxiliary benefits based on their husbands' average lifetime earnings, the bias created by using individual data is projected to be much larger for older cohorts than for younger cohorts. This article emphasizes the importance of using data for couples to estimate Social Security benefits, particularly for women. Although our focus is on married women, using data for couples is just as important for calculating the retirement benefits of divorced and widowed individuals. For individuals who are divorced or widowed at retirement, their Social Security benefits are based on their own earnings history, as well as the earnings histories of each of their previous spouses.     

A profile of social security child beneficiaries and their families: sociodemographic and economic characteristics

Using a rich dataset that links the Census Bureau's Survey of Income and Program Participation calendar-year 2004 file with Social Security benefit records, this article provides a portrait of the sociodemographic and economic characteristics of Social Security child beneficiaries. We find that the incidence ofbenefit receipt in the child population differs substantially across individual and family-level characteristics. Average benefit amounts also vary across subgroups and benefit types. The findings provide a better understanding of the importance of Social Security to families with beneficiary children. Social Security is a major source of family income for many child beneficiaries, particularly among those with low income or family heads with lower education and labor earnings.     

The Social Security Administration's 10-percent sample file of OASDI beneficiaries.

This is another in a series of technical articles describing Social Security Administration sample files and statistical systems. The focus of this article is the 10-percent sample file that provides program and demographic information on Social Security beneficiaries under the Old-Age, Survivors, and Disability Insurance program. The article explains the contents and availability of the 10-percent sample and presents several statistical tabulations as examples of the wide variety of data that can be developed from the file.     

The growth in Social Security benefits among the retirement-age population from increases in the cap on covered earnings

Analysts have proposed raising the maximum level of earnings subject to the Social Security payroll tax (the "tax max") to improve long-term Social Security Trust Fund solvency. This article investigates how raising the tax max leads to the "leakage" of portions of the additional revenue into higher benefit payments. Using Health and Retirement Study data matched to Social Security earnings records, we compare historical payroll tax payments and benefit amounts for Early Boomers (born 1948-1953) with tax and benefit simulations had they been subject to the tax max (adjusted for wage growth) faced by cohorts 12 and 24 years older. We find that 43.2 percent of the additional payroll tax revenue attributable to tax max increases affecting Early Boomers relative to taxes paid by the cohort 12 years older leaked into higher benefits. For Early Boomers relative to those 24 years older, we find 53.5 percent leakage.     

Projecting retirement income of future retirees with panel data: results from the modeling income in the near-term (MINT) project

Cross-sectional data capture only a point in time and miss individual changes in earnings, labor force participation, marriage, fertility, and health. Because panel data follow individuals over time, they do not have this problem. The problems or concerns with cross-sectional data may be compounded when these data are used to make projections. Iams and Sandell (1997) found that using panel data on earnings explained much more variation in future earnings than using cross-sectional survey data. Panel data are also needed to estimate Social Security benefits, especially for women. Because of auxiliary benefits paid to spouses, ex-spouses, and widow(er)s of entitled workers, an individual's Social Security retirement benefit depends not only on his or her earnings history, but also on his or her marital history and the earnings histories of current and previous spouses. When we compare projected unreduced Social Security benefits with what they would be if we didn't have marital history or earnings history data for men, we find that: Benefits computed using only earnings histories are not very different from benefits computed using both earnings and marital histories. Benefits computed using only current earnings and marital histories underestimate benefits for those in earlier birth cohorts and overestimate benefits for those in the most recent birth cohort. Benefits computed without either marital or earnings histories underestimate benefits for all birth cohorts, but by much more for earlier cohorts than for more recent cohorts. For women we find that benefits computed without marital or earnings histories underestimate benefits in all birth cohorts. The largest differences are for women in earlier birth cohorts. Using both marital and earnings histories to estimate unreduced Social Security benefits, we find that men are projected to continue receiving higher benefits than women, although the gap is expected to narrow as the baby boomers near retirement age. We also look at the composition of projected total income available at retirement for those with incomes in the 45th-55th percentiles of the income distribution and find that: Total income at retirement is projected to be larger for men than for women in every birth cohort. Women are projected to receive the largest share of their total income from Social Security benefits. Men are projected to receive the largest share of their total income from other income sources, although this share declines as the baby boomers near retirement age.     

Mortality reporting in SSA linked data: preliminary results

The earnings and benefit data of the Social Security Administration represent a fairly large and balanced sample of present and past wage earners nationwide. Since the incentives for reporting deaths to the Social Security Administration mean that deaths are reported as a matter of course, this data base serves as an interesting prospect for examining problems of differential mortality. Variables available include age, race, sex, industry, and place of employment. To check on the coverage and content differences between social security and death certificate data, a sample study is being undertaken that links the two sources for 1975 decedents. This article provides a preliminary examination of the differences between the presumably complete frame of death certificates and the social security record data. Related efforts now in progress to improve available information for use in further mortality research are also discussed briefly.     

Noncitizens and the Supplemental Security Income program

The Supplemental Security Income (SSI) program, established by the Social Security Amendments of 1972 (Public Law 92-603), was designed to provide cash assistance to needy aged, blind, and disabled citizens, and noncitizens lawfully admitted for permanent residence or permanently residing under color of law. Since then, this means-tested program has undergone many legislative changes that affect the eligibility status of noncitizens. This article, presented in three parts, discusses the legislative history of noncitizen eligibility, and details relevant laws enacted since the program's inception; provides current data on the trends and changes of the noncitizen population; and describes the larger population of foreign-born SSI recipients, of which the noncitizens are a part. Data on the number of SSI recipients born abroad but who had become citizens before applying for SSI payments were not previously available. Analytical data are from the Supplemental Security Record (SSR) matched to the Social Security Number Identification (Numident) file.     

The role of earnings and financial risk in distributional analyses of Social Security reform measures

The Social Security Trustees project that the Social Security program faces longterm financing difficulties. Several proposals that have been offered to shore‐up the finances of the Social Security program would create individual retirement accounts funded with part of the payroll tax. The authors of many of these proposals claim that future beneficiaries will be better off under their new system than under the current system. This study examines the consequences of differing earnings patterns and year‐to‐year differences in asset returns have for Social Security retired worker benefits in three Social Security reform proposals. Incorporating both actual earnings histories and variation in asset returns shows that none of the three individual account plans can always deliver benefits that are higher than payable current‐law benefits. © 2006 by the Association for Public Policy Analysis and Management     

Income growth and future poverty rates of the aged

This article estimates the effects of 50 years of steady growth in incomes on poverty rates among the elderly. It assumes that the poverty threshold continues to be adjusted for inflation but not for increases in real incomes. Simulations with the March 1998 Current Population Survey indicate that if the benefit rules for Social Security and Supplemental Security Income (SSI) are not changed and if earnings and other sources of income in an otherwise unchanging population grow by 1 percent per year (the intermediate assumption about earnings growth used in the Social Security Trustees' Report), poverty among the elderly will decrease from 10.5 percent in 1997 to about 7.2 percent in 2020 and to 4.1 percent in 2047. These projected poverty rates are quite sensitive to both the assumption about earnings growth and the assumption that benefits are not further reduced to maintain solvency. This article quantifies the sensitivity of the results to these assumptions and discusses several other aspects that might affect future poverty rates--changes in other income components like SSI, earnings, and pensions; changes in longevity and marital patterns; and changes in the distribution of earnings.     

Characteristics of individuals with integrated pensions

Employer pensions that integrate benefits with Social Security have been the focus of relatively little research. Since changes in Social Security benefit levels and other program characteristics can affect the benefit levels and other features of integrated pension plans, it is important to know who is covered by these plans. This article examines the characteristics of workers covered by integrated pension plans, compared to those with nonintegrated plans and those with no pension coverage. Integrated pension plans are those that explicitly adjust their benefit structure to help compensate for the employer's contributions to the Social Security program. There are two basic integration methods used by defined benefit (DB) plans. The offset method causes a reduction in employer pension benefits by up to half of the Social Security retirement benefit; the excess rate method is characterized by an accrual rate that is lower for earnings below the Social Security taxable maximum than above it. Defined contribution (DC) pension plans can be integrated along the lines of the excess rate method. To date, research on integrated pensions has focused on plan characteristics, as reported to the Bureau of Labor Statistics (BLS) through its Employee Benefits Survey (EBS). This research has examined the prevalence of integration among full-time, private sector workers by industry, firm size, and broad occupational categories. However, because the EBS provides virtually no data on worker characteristics, analyses of the effects of pension integration on retirement benefits have used hypothetical workers, varying according to assumed levels of earnings and job tenure. This kind of analysis is not particularly helpful in examining the potential effects of changes in the Social Security program on workers' pension benefits. However, data on pension integration at the individual level are available, most recently from the Health and Retirement Study (HRS), a nationally representative survey of individuals aged 51-61 in 1992. This dataset provides the basis for the analysis presented here. The following are some of the major findings from this analysis. The incidence of pension integration in the HRS sample is 32 percent of all workers with a pension (14 percent of all workers). The HRS can also identify integrated DC plans, a statistic that is not available from BLS data. The rate of integration for workers with only DC plans is 8 percent. After controlling for other variables, several socio-demographic characteristics are significantly related to the incidence of integration. The probability of having an integrated pension is 4.6 percentage points less for men compared to women. Non-Hispanic blacks are 6.4 percentage points less likely than non-Hispanic whites to have integrated pensions. Union members are 14 percentage points less likely to have integrated pensions, while workers with less than a graduate level education are at least 15 percentage points more likely to have a pension that is integrated. Some earnings and pension characteristics are also significantly correlated with pension integration. Earnings are positively related, with the probability of having an integrated pension increasing by 2 percentage points for an increase of $1,000 in annual pay. An even larger effect comes from earning at or above the Social Security taxable maximum. Workers at or above this income level are 10 percentage points more likely to have an integrated plan, but for those with more than one plan the probability of pension integration goes up by 13 percentage points.     

Comparing replacement rates under private and federal retirement systems

One measure of the adequacy of retirement income is replacement rate - the percentage of pre-retirement salary that is available to a worker in retirement. This article compares salary replacement rates for private-sector employees of medium and large private establishments with those for federal employees under the Civil Service Retirement System and the Federal Employees Retirement System. Because there is no standard benefit formula to represent the variety of formulas available in the private sector, a composite defined benefit formula was developed using the characteristics of plans summarized in the Bureau of Labor Statistics Medium and Large Employer Plan Survey. The resulting "typical" private-sector defined benefit plan, with an accompanying defined contribution plan, was then compared with the two federal systems. The Civil Service Retirement System (CSRS) is a stand-alone defined benefit plan whose participants are not covered by Social Security. Until passage of the 1983 Amendments to Social Security Act, it was the only retirement plan for most federal civilian employees. Provisions of the 1983 Amendments were designed to restore long-term financial stability to the Social Security trust funds. One provision created the Federal Employees Retirement System (FERS), which covers federal employees hired after 1983. It was one of the provisions designed to restore long-term financial stability to the Social Security trust funds. FERS employees contribute to and are covered by Social Security. FERS, which is a defined benefit plan, also includes a basic benefit and a 401(k)-type plan known as the Thrift Savings Plan (TSP). To compare how retirees would fare under the three different retirement systems, benefits of employees retiring at age 65 with 35 years of service were calculated using hypothetical workers with steady earnings. Workers were classified according to a percentage of the average wage in the economy: low earners (45 percent), average earners (100 percent) high earners (160 percent), and maximum earners (earnings at the taxable maximum amount). Overall, this analysis found that: Excluding Social Security benefits and TSP and defined contribution annuities, CSRS retirees have a higher pre-retirement salary replacement rate than either FERS or private-sector retirees. Private-sector retirees, however, have higher replacement rate than their FERS counterparts. Including Social Security benefits but not TSP and defined contribution plan annuities, CSRS retirees who are maximum earners have a higher pre-retirement salary replacement rate (despite receiving no Social Security benefits) than FERS retirees with the same earnings. Private-sector retirees in all earnings categories have a higher replacement rate than federal retirees with the same earnings. Including Social Security and TSP and defined contribution plan annuities, private-sector retirees in all earnings categories have a higher replacement rate than federal retirees, but their rate is close to that of FERS retirees. The rate is higher for FERS retirees than for CSRS retirees in all earnings categories. This analysis shows that replacement creates could exceed 100 percent for FERS employees who contribute who contribute 6 percent of earnings to the TSP over full working career. Private-sector replacement rates were quite similar for those with both a defined benefit and a defined contribution pension plan. Social Security replacement rates make up the highest proportion of benefits for th private sector's lowest income quartile group. The replacement rate for 401(k) plans and the TSP account for a higher proportion of benefits than does Social Security for all other income groups, assuming the absence of a defined benefit plan.     

Effect on benefits of earnings at ages 65 or older, 1995.

A major policy issue for the Social Security program is the treatment of earnings of persons who have attained retirement age. This article discusses the retirement test and recomputation of benefit provisions, and provides statistical data for 1995. In 1995, about 806,000 persons aged 65-70 had significant earnings resulting in the withholding of benefits by the retirement test. About 1,659,000 persons aged 65 or older realized an increase in their benefit amount because of their earnings.     

The impact of changes in couples' earnings on married women's Social Security benefits

Women's labor force participation and earnings dramatically increased after World War II. Those changes have important implications for women's Social Security benefits. This article uses the Social Security Administration's Modeling Income in the Near Term (version 6) to examine Social Security benefits for current and future beneficiary wives. The projections show that fewer wives in more recent birth cohorts will be eligible for auxiliary benefits as spouses because their earnings are too high. If their husbands die, however, most wives will still be eligible for survivor benefits because, despite the increase in their earnings over time, they still typically have lower earnings than their husbands. Even so, the share of wives who would be ineligible for widow benefits is projected to double between cohorts.     

Evaluating the initial impact of eliminating the retirement earnings test

Under the retirement earnings test, Social Security benefits are reduced if earnings exceed specified amounts, although the benefit reduction is partly offset by future benefit increases. By imposing a tax on the earnings of beneficiaries, the earnings test provided a disincentive for them to supplement retirement income by working. The Senior Citizens Freedom to Work Act of 2000 eliminated the earnings test for Social Security beneficiaries who have reached the full retirement age. This article presents the first study of labor force activity (earnings and employment) among individuals aged 65-69 before and immediately after this sudden rule change. Drawing on Social Security administrative data, the author examines three widely expected reactions: increased return to work, increased hours worked, and accelerated applications for old-age benefits. The analysis finds that removing the retirement earnings test: Encouraged some workers to increase their earnings. The increases in earnings are large and significant among higher earners but are not statistically significant among lower earners. Had little effect on employment. Removing the earnings test appears to have had no immediate, significant effect on the employment rate of older workers. Employment of older people may be affected in the longer run, however. Slightly increased the pace of applications for benefits. Applications rose about 2 percent in the 65-69 age group in 2000. The overall acceleration will probably be small, however, because most individuals in this age group apply for benefits before reaching the full retirement age. Although the current analysis captures the effects of retaining older workers in the labor force, these initial results may not capture all the effects of eliminating the retirement earnings test, however, for two reasons. First, the analysis covers only a single year--the year the earnings test was eliminated. Since eliminating the earnings test may have had little effect on people who had already retired, its full effect may not be apparent for several years. Second, the analysis applies only to workers aged 65-69. Eliminating the earnings test for people above the full retirement age may also encourage younger workers to delay retirement and therefore increase their labor supply. Further analysis will therefore be required to determine the longer-run impact of eliminating the retirement earnings test.     

Projecting immigrant earnings: the significance of country of origin

This article asks whether information about immigrants beyond their age, education, and years since migration can be productively used to project their earnings. Although many factors could affect immigrant earnings, what is most useful for Social Security modelling purposes is relevant information that is readily available on a continuous basis. Country of origin is a good candidate, as it is regularly and readily available from several administrative and survey data sources. In this article, microdata samples from the 1960-90 censuses are used to examine the relationship between country of origin and the earnings of immigrants. By following cohorts of immigrants over 10-year intervals, we learn how country of origin affects the initial earnings of immigrants and how the relationship between country of origin and immigrant earnings changes as immigrants continue to live in the United States. The article also presents theoretical insights and empirical evidence about the underlying causes of the link between country of origin and immigrant earnings.     

Incorporating immigrant flows into microsimulation models

Building on the research on immigrant earnings reviewed in the first article of this series, "Research on Immigrant Earnings," the preceding article, "Adding Immigrants to Microsimulation Models," linked research results to various issues essential for incorporating immigrant earnings into microsimulation models. The discussions of that article were in terms of a closed system. That is, it examined a system in which immigrant earnings and emigration are forecast for a given population represented in the base sample in the microsimulation model. This article, the last in the series, addresses immigrant earnings projections for open systems--microsimulation models that include projections of future immigration. The article suggests a simple method to project future immigrants and their earnings. Including the future flow of immigrants in microsimulation models can dramatically affect the projected Social Security benefits of some groups.     

Development of diagnostic data in the 10-percent sample of disabled SSI recipients

The Social Security Administration has created a 10-percent sample data base of blind and disabled recipients from the Supplemental Security Income (SSI) program. Codes showing the primary medical diagnosis were obtained for the sample by matching several files, and by imputing codes to sample cases where no diagnostic codes were found. The data base is updated each year by repeating this matching operation and by bringing forward the diagnostic codes from the previous year's file. This article describes the sources of diagnostic information in the administrative record system; the methodology used in the development of the 10-percent disability data base, and the technique chosen to compensate for missing values.     

Lifetime distributional effects of Social Security retirement benefits

This article presents three measures of the distribution of actual and projected net benefits (benefits minus payroll taxes) from Social Security's Old-Age and Survivors Insurance (OASI) for people born between 1931 and 1960. The results are based on simulations with the Social Security Administration's Model of Income in the Near Term (MINT), which projects retirement income through 2020. The base sample for MINT is the U.S. Census Bureau's Survey of Income and Program Participation panels for 1990 to 1993, matched with Social Security administrative records. The study population is grouped into 5-year birth cohorts and then ranked by economic status in three ways. First, the population is divided into five groups on the basis of individual lifetime covered earnings, and their lifetime present values of OASI benefits received and payroll taxes paid are calculated. By this measure, OASI provides much higher benefits to the lowest quintile of earners than to other groups, but it becomes less redistributive toward lower earners in more recent birth cohorts. Second, people are ranked by shared lifetime covered earnings, and the values of shared benefits received and payroll taxes paid are computed. Individuals are assumed to split covered earnings, benefits, and payroll taxes with their spouses in the years they are married. By the shared covered earnings measure, OASI is still much more favorable to persons in the lower income quintiles, although to a lesser degree than when people are ranked by individual covered earnings. OASI becomes more progressive among recent cohorts, even as net lifetime benefits decline for the entire population. Finally, individuals are ranked on the basis of their shared permanent income from age 62, when they become eligible for early retirement benefits, until death. Their annual Social Security benefits are compared with the benefits they would have received if they had saved their payroll taxes in individual accounts and used the proceeds to buy either of two annuities that provide level payments from age 62 until death: a unisex annuity that is based on the average life expectancy of the birth cohort or an age-adjusted annuity that is based on the worker's own life expectancy. On the permanent income measure, OASI is generally more favorable to people in higher income quintiles. Moreover, it is particularly unfavorable to those in the lowest quintile. Because people in the lowest quintile have a shorter life expectancy, they receive OASI benefits for a shorter period. This group would receive greater benefits in retirement if they invested their payroll taxes in the age-adjusted annuity. OASI is more favorable to them than the unisex annuity, however, OASI is becoming more progressive in that the net benefits it provides drop more rapidly among higher income quintiles than lower ones. This article also examines how OASI affects individuals by educational attainment, race, and sex. On both the lifetime covered earnings and the permanent income measures, OASI is more favorable to workers with less education and more favorable to women. The results by race and ethnicity are mixed. When people are ranked by the present value of their shared lifetime covered earnings, OASI appears more favorable to non-Hispanic blacks and Hispanics than to non-Hispanic whites. When people are ranked by shared permanent income in retirement, however, OASI produces negative returns for both non-Hispanic blacks and non-Hispanic whites in the most recent birth cohorts, with non-Hispanic blacks faring relatively worse. The changes across cohorts occur partly because of changes in tax rates and benefits, but more importantly because of changing demographics and earnings patterns of the workforce. Of particular importance is the increasing share of beneficiaries who receive worker benefits instead of auxiliary benefits as wives or widows. OASI benefits are based on the lifetime covered earnings of current or former married couples, as well as on earned retirement benefits of individuals. The reduced importance of auxiliary benefits (due to the higher lifetime covered earnings of women) and the increased proportion of divorced retirees make OASI more progressive--even as net benefits decline--for current and future cohorts than for cohorts who retired in the 1990s. Analysis of these findings suggests that simulations of policy changes in Social Security must take into account the decreasing importance of auxiliary benefits across birth cohorts and the complex changes in individuals' marital histories.     

Social Security: a financial appraisal for the median voter

Several explanations have been proposed for why voters continue to support unfunded social security systems. Browning (1975) suggests that the extremely large unfunded pension systems of most democracies depend on the existence of a voting majority composed of middle-aged and older people who fail to fully internalize the cost of financing the system. In fact, when voting, economically rational workers consider only their current and future contributions to the system and their expected pension benefits--not their past contributions, which they regard as sunk costs. If, for a majority of voters, the expected continuation return from social security exceeds the return from alternative assets, an unfunded social security system is politically sustainable. This article explores the validity of Browning's proposition by quantifying the returns that U.S. voters in presidential elections from 1964 to 1996 have obtained, or expect to obtain, from Social Security. Did "investments" in Social Security outperform alternative forms of investment, such as mutual funds or pension funds, for a majority of the voters? What can be expected for the future? The U.S. Social Security system redistributes income within age cohorts on the basis of sex, income, and marital status. To account for some of these features, the median voter is represented by a family unit whose members--a husband who accounts for 70 percent of household earnings and a wife who accounts for 30 percent--make joint economic and voting decisions. Thus, retirement and survival benefits paid out to the spouse of an insured worker can be included in the calculation of Social Security returns. Interval estimates of voters' family incomes from the U.S. Census Bureau were used to obtain the median voter's household earnings. The median voter's age is derived from the ages of those who voted in presidential elections, not from the ages of the entire electorate. The median voter's contributions to Social Security are the product of the joint employer/employee Old-Age and Survivors Insurance (OASI) tax rate and employee earnings. Data on actual contributions are available for median voters in the 1964 to 1976 elections; Social Security Administration (SSA) estimates are used for future tax rates and average wage growth rates. Data on actual old-age, retirement, and survivor benefits, as well as estimates of future benefits, are also available from SSA. Analysis of ex-post returns from "investing" in Social Security and from a buy-and-hold strategy applied to three alternative assets--the Standard & Poor's Composite Index (S&P), the Dow Jones Industrial Average (DJIA), and U.S. government bonds--shows surprising results. In 1964 and 1968, Social Security largely outperformed the other three assets. In 1972, Social Security and the stock market performed almost equally. In 1976, however, the median voter would have been better off in the stock market. The expected returns for median voters in later elections cannot be directly compared with realized returns from alternative assets. However, estimates range from 5.7 percent in 1984 to 7.0 percent in 1996 and thus compare favorably with average returns of 5.6 percent for S&P, 5.3 percent for DJIA, and 2.1 percent for government bonds over the 1964-1996 period. Although these findings must be taken with caution since they compare ex-post returns, they show that, despite a continuous reduction in profitability, Social Security still represents a safe, high-return asset for a majority of families.