Reductions in Poverty Significantly Greater in the 1990s Than Official Estimates Suggest

Any serious student of poverty understands the many weaknesses of the official poverty measure. In response to these concerns, the National Academy of Sciences issued recommendations to improve the official poverty measure. In addition, the Census Bureau has undertaken a vast amount of research to improve our official poverty series. The methodology used in this analysis incorporates as many of those recommendations as possible to examine the change in poverty from 1979 to 2000. Compared to this methodology, the official poverty measure understates dramatically the reduction in poverty over time.     

Incomes of the aged and nonaged, 1950-82

This article discusses the results of previous studies, both supportive and nonsupportive of the conclusion that the aged are as well-off as the nonaged, and then presents a range of figures from Bureau of the Census reports over the period 1950-82 that measure both the incomes of the aged and nonaged and those of subgroups within these populations. Census figures indicate that the aged and nonaged have about equal levels of average per capita family income and that about the same proportions of these groups have incomes below the poverty line. However, aged unrelated individuals, who account for about a third of all aged persons, have less than three-fifths the income of nonaged unrelated individuals. When the per capita family income of the aged is compared separately with that of families headed by persons aged 25-44 and 45-64, aged persons receive more than those under age 45 but less than those aged 45-64. Trends in the economic status of the aged and nonaged over the period 1950-82 indicate numerous fluctuations rather than a consistent improvement in the income of either group in relation to the other.     

Reconciliation of income and consumption data in poverty measurement

A recent series of papers has renewed interest in the question of whether consumption data are superior to income data for poverty measurement. Although the Census Bureau has provided researchers with an experimental series of variables that can produce a comprehensive income measure, this resource has not been fully exploited in previous analyses. When poverty is measured by a comprehensive income measure, income poverty rates and trends are similar to consumption poverty rates. Arguments that income is measured with more error than consumption at the bottom of the distribution are shown to be based upon inferior income data. © 2008 by the Association for Public Policy Analysis and Management.     

Money incomes of aged and nonaged family units, 1967-84

This article provides a detailed examination of changes in the income of the aged and of other age groups from 1967 to 1984. Levels of income, income inequality, the relative importance of selected types of income, and the poverty rates of various age groups are also analyzed. The data are from the Bureau of the Census Current Population Survey and annual money income before taxes is the measure of income. The relative income gains of the aged, compared with the nonaged, in 1967-84 reversed an earlier pattern in the post-World-War-II period: From 1947 to 1967, the incomes of the nonaged rose at a faster pace than those of the aged. Differences in their income growth were greater in 1979-84 than in 1967-79. Despite the substantial difference in the rates of income growth for the aged and nonaged units in 1967-84, the relationship between age and median income was altered only slightly. In 1984, aged family units continued to have relatively low median income, especially compared with the incomes of those in middle age.     

The increasing labor force participation of older workers and its effect on the income of the aged

The labor force participation rates of men and women aged 62-79 have notably increased since the mid-1990s. The result is a dramatic increase in the share of total money income attributable to earnings. For persons aged 65-69, the earnings share of total income increased from 28 percent in 1980 to 42 percent in 2009. For this age group in the late 1980s and early 1990s, Social Security benefits and earnings were roughly equal shares of total money income (about 30 percent); the earnings share is now more than 12 percentage points larger. When we focus on aged persons who receive Social Security benefits, earnings shares have increased markedly throughout the 62-79 age range since the early 1990s. We show that for aged persons with labor market earnings, those earnings have a large effect on their relative position in the distribution of annual money income of older Americans.     

Hispanics, Social Security, and Supplemental Security Income

The U.S. Census Bureau reports that Hispanics are the country's largest and fastest growing minority, representing about 14.4 percent of the population in 2005 (Census Bureau 2006b). By 2050, Hispanics will account for an estimated 24.4 percent of the population--or 1 in every 4 persons in the United States (Census Bureau 2004, Table 1 a). The Hispanic population tends to be younger than the overall population and currently represents a relatively small but growing fraction of the Social Security beneficiary population. The representation of Hispanics in the Supplemental Security Income (SSI) program, however, approximates that of their representation in the overall population. This article compares the Hispanic population with the overall population along several dimensions, with a particular focus on the Social Security beneficiary and SSI recipient populations. Data are drawn mainly from the 2005 Public Use Microdata Sample of the American Community Survey (ACS PUMS), a relatively new data source with a rich set of economic and demographic variables. Fully implemented nationwide for the first time in 2005, the ACS became the largest household survey in the United States with a sample of almost 3 million addresses. The analysis using the ACS finds that the Hispanic population is significantly different from the general population, particularly in the areas of age distribution, educational attainment, and economic well-being. Compared with the general population, the Hispanic segment is younger and is characterized by lower levels of educational attainment and a higher rate of poverty. The Hispanic Social Security beneficiary population also differs significantly from the general beneficiary population in the same areas. In contrast, the Hispanic and general SSI populations are more comparable with regard to age and economic status and differ significantly only with regard to education.     

The earnings test and the short-run work response to its elimination

Eliminating the earnings test will have different effects on the work effort of persons aged 65-69, depending on whether or not they are currently working or currently receiving Social Security benefits. This article reviews the development of the earnings test and examines the theoretical implications on work effort of removing the test for members of this age group. It looks at the Current Population Survey (CPS) data to determine how many persons aged 65-69 have characteristics that can be identified with groups that would theoretically increase, decrease, or not change their work effort should they no longer be subject to the earnings test. This analysis suggests that at least 80 percent, and perhaps more than 90 percent, of the 9.7 million persons aged 65-69 will not change their level of work effort if the earnings test is eliminated. Individuals who would modify their hours worked and earnings are fairly evenly split among those who would increase, decrease, or have an undetermined direction of change in their work effort.     

Economic security, 1935-85

The economic well-being of both working and retired persons has improved significantly since the Social Security Act was passed in 1935. More people are employed now than at any time since then, despite declining employment among the aged and more years of school attendance among the young. The ratio of non-workers to workers--a broad measure of dependency--is lower now than at any time since the 1930's. Social security has grown and matured to become a strong foundation of retirement income, and other work-related employee benefits have grown in tandem with social security. Employer contributions for social insurance and related employee benefits have grown from being about a 1-percent supplement to aggregate wages and salaries in 1929 to nearly 20 percent today. Social security and Medicare account for just over a fourth of employer contributions, while other public and private retirement systems represent just over another fourth. The balance of benefits for active workers includes group health and life insurance, unemployment insurance, workers' compensation, temporary disability insurance, and related benefits. Pay for holidays, vacations, and sick leave is estimated to have increased from less than 1 percent of aggregate pay in 1929 to about 10 percent today. The improved economic status of the aged has been documented by a series of surveys beginning in 1941-42 and carried out from time to time until 1972 and biennially since 1976. The earlier surveys were supplemented with estimates from record data and tables from the Bureau of the Census. The income of the aged as a whole has grown by about 75 percent over the past 2 decades after taking inflation into account. The income of the aged as a whole grew faster than that of the nonaged in the 1970's and early 1980's when real social security benefits increased faster than inflation and wages lagged behind it. New beneficiaries in 1982 were in better health and were more likely to retire because they wanted to than was true of their counterparts in the early 1940's. Not only have benefits continued to be the main component of income of the aged as total incomes have grown, but also benefits have become much larger in relation to average earnings than used to be the case. Retired workers are much more likely now than in the early 1940's to have other pensions or income from assets to supplement benefits.(ABSTRACT TRUNCATED AT 400 WORDS)     

Welcome Mats and On‐Ramps for Older Adults: The Impact of the Affordable Care Act's Medicaid Expansions on Dual Enrollment in Medicare and Medicaid

For many low‐income Medicare beneficiaries, Medicaid provides important supplemental insurance that covers out‐of‐pocket costs and additional benefits. We examine whether Medicaid participation by low‐income adults age 65 and up increased as a result of Medicaid expansions to working‐age adults under the Affordable Care Act (ACA). Previous literature documents so‐called “welcome mat” effects in other populations but has not explicitly studied older persons dually eligible for Medicare and Medicaid. We extend this literature by estimating models of Medicaid participation among persons age 65 and up using American Community Survey data from 2010 to 2017 and state variation in ACA Medicaid expansions. We find that Medicaid expansions to working‐age adults increased Medicaid participation among low‐income older adults by 1.8 percentage points (4.4 percent). We also find evidence of an “on‐ramp” effect; that is, low‐income Medicare beneficiaries residing in expansion states who were young enough to gain coverage under the 2014 ACA Medicaid expansions before aging into Medicare were 4 percentage points (9.5 percent) more likely to have dual Medicaid coverage relative to similar individuals who either turned 65 before the 2014 expansions or resided in non‐expansion states. This on‐ramp effect is an important mechanism behind welcome mat effects among some older adults.     

A note on sampling variance estimates for Social Security program participants from the Survey of Income and Program Participation

The Census Bureau's Survey of Income and Program Participation (SIPP) provides data that can be used to study the characteristics of Old-Age, Survivors, and Disability Insurance (OASDI) and Supplemental Security Income (SSI) program participants. It is important that estimates of sampling errors accompany such studies because the estimates may have large sampling errors due to the small number of sample cases available for specific analyses. The generalized sampling variances provided by the Census Bureau did not identify separately either program's participants and, therefore, do not pertain directly to analyses of these groups. This article describes an approach to the direct computation of sampling variances for OASDI and SSI program participants. The approach uses the pseudo stratum and half-sample codes available in SIPP public use data files. A table of generalized standard errors is constructed for participants of both programs aged 18 or older. Generalized standard errors could not be computed for child beneficiaries under age 18 because of a wide variation of design effects across subpopulation estimates.     

Mandatory retirement and labor-force participation of respondents in the Retirement History Study

The Age Discrimination in Employment Act (ADEA) was amended in 1978 to prohibit mandatory retirement before age 70 in most occupations. The impact of this legislation on the probability of older persons remaining in the labor force is the primary concern of this article. Specifically, questions concerning which older workers are affected by mandatory-retirement provisions and the extent to which they are forced to retire and leave the labor force are examined. Tabular analysis of data from the Retirement History Study on persons aged 62-63 in 1969 shows significant variation in mandatory-retirement coverage between the public and private sectors and across industries, occupations, and demographic groups. Until age 65, the labor-force participation rate of those facing compulsory retirement is higher than or equal to that of those not covered but it drops significantly below the noncovered rate after 65. Logit analysis of the labor-force participation of persons before and after age 65 indicates that mandatory retirement at that age reduces the probability of retirement by approximately 16.7 percentage points for white men wage earners. This results in a decline in the labor-force participation rate of all men aged 66-67 of approximately 4 percentage points.     

How have people responded to changes in the retirement earnings test in 2000?

This article describes responses to removing the retirement earnings test in 2000 for persons at the full retirement age or older. We examine annual earnings and retirement benefit claims from Social Security administrative data that cover the 4 years before and after the change. Three findings emerge from the study. First, the effect on earnings of removing the earnings test is uneven across people with different earnings levels. We find little effect on earnings at lower levels, but the effect on earnings in the mid to upper levels (50th to 80th percentiles) is large and significant. Such a finding indicates that the removal most affects people with earnings levels above the earnings test threshold. The largest increases in earnings are found at the 70th percentile for persons who have attained ages 65-69 and at the 60th percentile for those turning 65. Second, there is no clear evidence of the effect of the test's removal on the overall rate of labor force participation. A small rise in work participation among individuals aged 65-69 may be at least partially attributable to the trend already under way. Increases in work participation that do occur are mostly attributable to retaining older workers rather than inducing older workers back into the workforce. The effect appears to increase over time, suggesting that the removal has long-lasting effects on work participation. Third, the removal of the earnings test accelerated applications for benefits by 2 to 5 percentage points among individuals aged 65-69 and by 3 to 7 percentage points among those reaching age 65.     

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.     

Collecting information on disability in the 2000 Census: an example of interagency cooperation

The United States Constitution requires that an enumeration (or census) of the population be conducted every 10 years to apportion seats in the House of Representatives. Census information is also used to allocate funds and to plan and manage programs. Census 2000 occurs on April 1, 2000, when one-sixth of all American households will be mailed the "long form," containing disability, demographic, economic, and housing questions. Although no short set of commonly accepted questions on disability existed, one was developed for Census 2000 by a collaborative, federal interagency work group on disability, convened by the Office of Management and Budget. The work group consisted of staff from the Social Security Administration (SSA), the Department of Health and Human Services, the U.S. Census Bureau, and other agencies. They reviewed questions initially proposed by the Census Bureau, developed an alternative proposal, tested both versions in the Census Bureau's cognitive questionnaire lab, and on the basis of testing, derived a consensus version for Census 2000. In many ways, the six questions now contained on Census 2000 are an improvement over previous efforts. Disability is ascertained for children as well as for adults, and information will be collected separately for several domains of disability (for example, sensory, mental, physical). The need for a brief set of disability measures goes beyond Census 2000. If such data were collected regularly on national surveys, critical policy and program concerns across agencies could be addressed because better information could be gathered on changes in disability prevalence and on the characteristics of persons with disabilities. Other similar efforts include the former Disability Evaluation Study, now known as the National Study of Health and Activity--a national sample survey on working-age disability to be conducted by SSA--and the President's Task Force on the Employment of Adults with Disabilities (Executive Order 13078).     

The Characteristics and Unmet Housing Program Needs of Disabled HUD-Assisted Households

The mismatch between the housing needs of persons with a disability and the housing programs designed to accommodate those needs is an important housing policy concern. The U.S. Department of Housing and Urban Development (HUD) sponsors several programs designed to improve the housing conditions of persons with a disability, but we know little about the characteristics of persons with a disability, among those receiving federal housing assistance, or the degree to which persons with a disability are served by HUD-sponsored housing programs that are designed to meet the needs of persons with a disability. Our study relies on administrative data from HUD and the U.S. Census Bureau to address this research gap. We find that many persons with a disability are served by HUD-sponsored programs that are not designated for persons with a disability, even when disability accommodations have been requested, and a similarly large share of persons with a disability live in potentially eligible low-income households that do not receive HUD assistance.     

Benefit adequacy among elderly Social Security retired-worker beneficiaries and the SSI federal benefit rate

Both target effectiveness and administrative simplicity are desirable properties in the design of minimum benefit packages for public retirement programs. The federal benefit rate (FBR) of the Supplemental Security Income (SSI) program has been proposed by some analysts as a potentially attractive basis of establishing a new minimum benefit for Social Security on both of these grounds. This type of proposal is related to a broader array of minimum benefit proposals that would establish a Social Security benefit floor based on the poverty rate. In contrast to Social Security, the SSI program is means tested, including both an income and asset screen and also a categorical eligibility screen (the requirement to qualify as aged or disabled). The SSI FBR provides an inflation-adjusted, guaranteed income floor for aged and disabled people with low assets. The FBR has been perceived by proponents as a minimal measure of Social Security benefit adequacy because it represents a subpoverty income level for a family of one or two depending on marital status. For this same reason it has been seen as a target-effective tool of designing a minimum Social Security benefit. An FBR-based minimum benefit has also been viewed as administratively simple to implement; the benefit can be calculated from Social Security administrative records using a completely automated electronic process. Therefore-in contrast to the SSI program itself-an FBR-based minimum benefit would incur virtually no ongoing administrative costs, would not require a separate application for a means-tested program, and would avoid the perception of welfare stigma. While these ideas have been discussed in the literature and among policymakers in the United States over the years, and similar proposals have been considered or implemented in several foreign countries, there have been no previous analyses measuring the size of the potentially affected beneficiary population. Nor has there been any systematic assessment of the FBR as a measure of benefit adequacy or the tradeoffs between potential target effectiveness and administrative simplicity. Based on a series of simulations, we assess the FBR as a potential foundation for minimum Social Security benefits and we examine the tradeoffs between administrative simplicity and target effectiveness using microdata from the 1996 panel of the Survey of Income and Program Participation (SIPP). Our empirical analysis is limited to Social Security retired-worker beneficiaries aged 65 or older. We start with the assessment of the FBR as a measure of benefit adequacy. We are particularly concerned about two types of error: (1) incorrectly identifying some Social Security beneficiaries as "economically vulnerable," and (2) incorrectly identifying others as "not economically vulnerable." Operationally we measure economic vulnerability by two alternative standards. One of our measures considers beneficiaries with family income below the official poverty threshold as vulnerable. Our second measure is more restrictive; it uses a family income threshold equal to 75 percent of the official poverty threshold. We find that a substantial minority of retired workers have Social Security benefits below the FBR. The results also show that the FBR-based measure of Social Security benefit adequacy is very imprecise in terms of identifying economically vulnerable people. We estimate that the vast majority of beneficiaries with Social Security benefits below the FBR are not economically vulnerable. Conversely, an FBR-level Social Security benefit threshold fails to identify some beneficiaries who are economically vulnerable. Thus an FBR-level minimum benefit would be poorly targeted in terms of both types of errors we are concerned about. An FBR-level minimum benefit would provide minimum Social Security benefits to many people who are clearly not poor. Conversely, an FBR-level minimum benefit would not provide any income relief to some who are poor. The administrative simplicity behind these screening errors also results in additional program cost that may be perceived as substantial. We estimate that an FBR-level minimum benefit would increase aggregate program cost for retired workers aged 65 or older by roughly 2 percent. There are two fundamental reasons for these findings. First, the concept of an FBR-level minimum benefit looks at the individual or married couple in artificial isolation; however, the family is the main consumption unit in our society. The income of an unmarried partner or family members other than a married spouse is ignored. Second, individuals and couples may also have income from sources other than Social Security or SSI, which is also ignored by a simple FBR-based minimum benefit concept. The substantial empirical magnitude of measurement error arising from these conceptual simplifications naturally leads to the assessment of the tradeoff between target effectiveness and administrative simplicity. To facilitate this analysis, we simulate the potential effect of alternative screening methods designed to increase target effectiveness; while reducing program cost, such alternatives also may increase administrative complexity. For example, considering the combined Social Security benefit of a married couple (rather than looking at the husband and wife in isolation) might substantially increase target effectiveness with a relatively small increase in administrative complexity. Adding a family income screen might increase administrative complexity to a greater degree, but also would increase target effectiveness dramatically. The results also suggest that at some point adding new screens-such as a comprehensive asset test-may drastically increase administrative complexity with diminishing returns in terms of increased target effectiveness and reduced program cost. Whether a broad-based minimum benefit concept that is not tied to previous work experience is perceived by policymakers as desirable or not may depend on several factors not addressed in this article. However, to the extent that this type of minimum benefit design is regarded as potentially desirable, the tradeoffs between administrative simplicity and target effectiveness need to be considered.     

Modeling SSI financial eligibility and simulating the effect of policy options

This article simulates eligibility for Supplemental Security Income (SSI) among the elderly, analyzes factors affecting participation, and looks at the potential effects of various options to modify financial eligibility standards for the federal SSI program. We find that in the estimated noninstitutional elderly population of 30.2 million in the United States in 1991, approximately 2 million individuals aged 65 or older were eligible for SSI (a 6.6 percent rate of eligibility). Our overall estimate of the rate of participation among eligible elderly is approximately 63 percent, suggesting that more than a third of those who are eligible do not participate in the program. The results of our analysis of factors affecting participation among the eligible elderly show that expected SSI benefits and a number of demographic and socioeconomic variables are associated with the probability of participation. We also simulate the effects of various policy options on the poverty rate, poverty gap, annual program cost, the number of participants, and the average estimated benefits among participants. The simulations consider the potential effects of five policy alternatives: Increase the general income exclusion (GIE) from $20 to $80. Increase the earned income exclusion (EIE) from $65 to $260. Increase the federal benefit rate (FBR) by $50 for individuals and $75 for couples and eliminate the GIE. Increase the asset threshold to $3,000 for individuals and $4,500 for couples. Increase the asset threshold to $6,000 for individuals and $9,000 for couples. Using 1991 microdata from the Survey of Income and Program Participation (SIPP) matched to Social Security Administration administrative records and making adjustments reflecting aggregate program statistics, we present the results of our simulations for December 1999. The results show substantial variation in the simulated effects of the five policy alternatives along the various outcome dimensions considered. The simulated effects on the poverty gap of the elderly population range from a 7.9 percent reduction ("Increase the GIE from $20 to $80") to a 0.1 percent reduction ("Increase the EIE from $65 to $260"). All simulated interventions are expected to increase the rate of SSI participation among the elderly from a high of 20.3 percent ("Increase the GIE from $20 to $80") to a low of 0.5 percent ("Increase the EIE from $65 to $260"). We also find that the interventions that have greater estimated effects in terms of increased participation and reduced poverty tend to cost more. At the high end, we estimate that increasing the GIE from $20 to $80 could raise annual federal SSI cash benefit outlays by about 46 percent, compared with only 0.9 percent for increasing the EIE from $65 to $260. Similar to the EIE intervention, raising the resource thresholds by 50 percent would reduce the overall poverty gap of the elderly by only 0.2 percent, would increase SSI participation only modestly (by 1.3 percent), but would entail slightly higher program costs (by 1.4 percent). Increasing the asset threshold by 200 percent would have higher estimated effects on all three outcomes, but it would still be associated with relatively low increases in both costs and benefits. Finally, the simulated effects on the three key outcomes of increasing the FBR by $50 for individuals and $75 for couples, combined with eliminating the GIE, are relatively large but are clearly less substantial than increasing the GIE from $20 to $80. This work relies on data from the SIPP matched to administrative data on federal SSI benefits that provide a more accurate picture of SSI participation than has been feasible for previous studies. We simulate eligibility for federal SSI benefits by applying the program rules to detailed information on the characteristics of individuals and couples based on the rich array of demographic and socioeconomic data in the SIPP, particularly the comprehensive information SIPP provides on assets and monthly income. A probit model is estimated to analyze factors affecting participation among the eligible elderly. Finally, we conduct the policy simulations using altered program rules represented by the policy alternatives and predicted participation probabilities to estimate outcomes under simulated program rules. We compare those simulated outcomes to observed outcomes under current program rules. The results of our simulations are conditional on the characteristics of participants and eligibles in 1991, but they also reflect aggregate adjustments capturing substantial changes in overall participation and program benefit levels between 1991 and 1999.     

Leveraging Technologies in Public Agencies: The Case of the U.S. Census Bureau and the 2010 Census

Emerging technologies are transforming government agencies and the nature of governance and pose new challenges for public managers. Under the leadership of Steven J. Jost, associate director for communications, the U.S. Census Bureau leveraged emerging technologies during the 2010 Census to complete the project under budget while also engaging citizens through the design of viable participatory platforms. The 2010 Census campaign focused on increasing response rates and encouraging citizen participation through innovations in the communication process with citizens and through the infusion of technology. The Census Bureau also effectively managed risks associated with the use of emerging technologies. The authors examine the innovations, the risks, and the effort to manage those risks under Jost’s leadership.     

The economic status of the aged.

This article discusses what is known about the economic status of the aged. Numerous complexities involved in the assessment of their status are discussed. Compared with most other recent assessments, this study finds a less favorable status for the aged relative to other age groups. The focus is on an examination of detailed age groups, rather than summary aged and nonaged groups--thus providing a more complete picture of age differences. More than most other assessments, this study stresses uncertainty about the relative status of the aged and emphasizes what we do not know. It stresses that better adjustments for differences in needs among age groups and other subgroups of the population are necessary. It emphasizes that consistency between the definition of resources and the specification of needs is essential. Also discussed is the vulnerability of the aged to economic risks. Major findings include: Median cash income is highest for middle-aged family units and lowest for the oldest and youngest units. The poverty rate for aged persons is above the rates for other adult age groups, but below the rate for children. When noncash income is considered in addition to cash income, the income of the aged tends to improve relative to that of the nonaged, but serious measurement problems exist. When wealth is considered in addition to cash income, the economic status of the aged improves relative to that of the nonaged.     

Trends in the economic status of the elderly, 1976-2000

The economic well-being of elderly Americans (aged 65 or older) improved between 1976 and 2000. Overall, poverty rates fell during this period, median real income rose, and median income relative to the working-age population was relatively stable. Most population subgroups shared in the reduced poverty rates; however, the economic status of elderly Hispanics did not improve. This article attempts to explain those economic trends by identifying changes in five sources of income for the elderly and analyzing the changes in the context of demographic changes in the elderly populations over the past 25 years. As a result of increased longevity, for example, larger proportions of elderly men and women are now 80 or older, and smaller proportions are 65 to 69. Hispanics and Asian Americans make up a larger share of the elderly population and whites a smaller share. The fraction of women who are married has increased, the fraction who are widowed has fallen, and the fraction who are divorced has grown. Such demographic changes can greatly affect the economic status of subgroups as well as the overall elderly population. Of the five sources of income for the elderly, Social Security remains the most prevalent and important. While both the rate of receipt and the share of aggregate income from Social Security benefits stayed relatively steady over the past 25 years, the average real Social Security benefit increased because of rising wages. Income from assets, the second most important source of income for the elderly, fluctuated. Because the elderly are more likely to hold interest-bearing assets such as bonds rather than stocks, their asset income is responsive to changes in nominal interest rates and bond yields. Receipt of pension income increased during this period, although it leveled off during the 1990s. Factors contributing to this pattern include enactment of the Employee Retirement Income Security Act of 1974, which increased protections of pension benefits for spouses, and improved labor market opportunities for blacks and women. In recent years, defined contribution pension plans have become more prevalent than defined benefit plans, but the full effect of this change on pension income may not yet be apparent. After decades of decline, labor force participation rates of older men leveled out in the mid-1980s and then increased. For older women, the trend before the mid-1980s was flat, but since then rates have risen substantially. The increased use of part-time jobs or self-employment to ease the transition into retirement, the economic expansion of the 1990s, and the liberalization of the Social Security earnings test may all have contributed to those trends. Although the percentage of elderly people with earnings has increased only modestly in the past few years, the share of income from earnings has grown substantially--from 16 percent of income in 1984 to 23 percent in 2000. Finally, Supplemental Security Income (SSI) benefits are indexed for inflation but not for growth in real wages. As real incomes of the elderly rose, therefore, fewer elderly persons were eligible to receive SSI or, for those receiving SSI, were eligible for smaller benefits. The proportion of elderly persons receiving public assistance, primarily SSI, declined from 11 percent in 1976 to 5 percent in 2000.