Over the next 30 years, the share of the U.S. population age 65 and older will increase from about 15 percent to almost 22 percent, spurring growing interest in understanding whether people will have adequate income in retirement. As reflected in an extensive body of literature on the topic, researchers have developed diverse approaches for quantifying the adequacy of retirement income, focusing on different groups of retirees and employing different definitions of income and adequacy. For example, some researchers have computed the fraction of current retired workers whose income is below the poverty threshold and found it to be less than one-tenth of retirees. In contrast, others who have examined how income changes upon retirement project that a much larger fraction of current workers would experience a substantial decline in income as they retire. This report explains the various measures and approaches, providing a framework for further analysis of retirement income.
What Does Having an Adequate Retirement Income Mean?
Researchers have defined the adequacy of retirement income in two main ways: whether it satisfies basic needs and whether it allows retirees to maintain the standard of living they experienced before retirement. If the goal of the analysis is to determine how many retirees would be able to afford essential living expenses, then researchers would apply the first definition, which generally considers retirement income to be adequate if it exceeds a poverty threshold. Variations of that threshold are often used to determine eligibility for government programs targeted to low-income groups, such as the Supplemental Nutrition Assistance Program (SNAP) and parts of Medicaid.
In contrast, if the goal of the analysis is to examine whether retirees’ income would allow them to enjoy the same standard of living they experienced during their working years, then the second definition would apply. Economists and financial advisers generally use that second definition because large drops in consumption in retirement are typically considered undesirable. (However, maintaining a preretirement living standard does not necessarily mean that people would be able to avoid significant material hardship in retirement. For example, low-income workers who maintain their living standard in retirement could still have income that is below a poverty threshold.)
What Measures of Adequacy Have Researchers Used?
Gauges of retirement income can focus on a single year of retirement, typically the first year, or they can evaluate the adequacy of income over many years of retirement.
Single-Year Analysis. For simplicity, some studies analyze retirement income in a single year of retirement, providing a snapshot of retirement adequacy at a given point in time. Measures that evaluate retirement income for a single year can be based on either of the two definitions of adequacy.
Adequacy measures that are based on the basic needs definition include the official federal poverty thresholds (commonly referred to as the poverty thresholds) and the supplemental poverty measure (SPM) thresholds, both compiled by the Census Bureau, and elderly-specific thresholds, which particularly reflect the living expenses of retirees. (The various thresholds are described in greater detail in the section titled “Measures of Adequacy Used in Single-Year Analyses.”) For all three types of thresholds, researchers calculate retirement income for the population of interest in a selected year, and if that income is at least as high as the chosen basic needs threshold, then it is considered to be adequate.
The most widely used measure of the adequacy of retirement income—a threshold based on the standard-of-living definition—is known as the target replacement rate. That rate is broadly defined as the amount of income in retirement, expressed as a percentage of income before retirement, that enables retirees to maintain the standard of living they enjoyed while working. Although a common rule of thumb is that replacing at least 70 percent of gross preretirement income would avoid a marked decline in retirees’ standard of living, that specific goal is not appropriate for all people. To better capture the diversity of people’s circumstances, researchers have developed a range of target rates that vary with individual characteristics, such as marital status, lifetime income, and homeownership.
Multiyear Analysis. Although many researchers analyze retirement income in a single year, financial security is dynamic and may change as retirees’ spending patterns evolve. For a more comprehensive measure of the adequacy of retirement income, basic needs and replacement rate thresholds can be applied to resources over multiple years, ultimately encompassing the full duration of the retirement period. Unlike single-year analyses, studies that look at multiple years can show whether or not retirees are able to maintain their standard of living as their needs change over time.
Multiyear analyses fall into two broad groups: limited multiyear analysis and a comprehensive simulation-based approach. Limited multiyear analysis applies basic needs thresholds or target replacement rates to resources in several different years of retirement, examining how the adequacy of retirement income changes between discrete points in time. The simulation-based approach, however, evaluates the adequacy in every year of retirement until the end of life, fully capturing changes in retirement income over time. That approach generally requires complex projections and more detailed data. One advantage of the simulation-based approach is its ability to incorporate increases in health care expenses that typically occur over the course of retirement as well as potentially large long-term care costs.
What Do the Measures Reveal About the Adequacy of Retirement Income?
Current measures of the adequacy of retirement income provide diverse answers about the state of retirement income security in the United States. That diversity stems from a number of sources. One source is which definition of adequacy is used (meeting basic needs or maintaining the preretirement standard of living). Other sources include which cohorts are analyzed and precisely how income and wealth are counted.
What proportion of retirees will have an adequate income? The answer varies depending on the threshold for “adequate.” In general, researchers conclude that fewer than one-tenth of retirees will have income below the lowest basic needs threshold—the poverty threshold—while a much larger fraction of people are projected to have income that will fall short of maintaining their preretirement standard of living. For example, in one study, the authors projected that only about 6 percent of workers in their 40s and 50s would have future retirement income below the poverty threshold. In contrast, recent analysis using the National Retirement Risk Index (NRRI), a measure developed by the Center for Retirement Research at Boston College, indicated that about half of working-age U.S. households were “at risk” of not being able to maintain their living standard in retirement.
Even when the thresholds are based on the same broad definition of adequate income, researchers have made different analytical choices when constructing the adequacy measures, resulting in different findings. For example, in contrast with basic needs studies that rely on the poverty threshold, a recent analysis using the Elder Economic Security Standard Index, or the Elder Index—which is a basic needs threshold based on the larger expenses that are common among households with elderly members—found that about a quarter of two-person households whose members were age 65 or over and more than half of the elderly people in single-person households would have inadequate income.
Similarly, according to a recent literature review by the Government Accountability Office (GAO), studies that evaluated the extent to which workers were able to maintain their preretirement living standards reached a broad range of conclusions. Overall, across the studies, the share of current workers who were at risk of having inadequate income ranged from about one-third to two-thirds.
Using more than one measure at a time can provide additional insights. For example, analyses pegged to the poverty threshold indicate how many people are experiencing significant material hardship; however, they do not reveal how those people’s standard of living changed upon retirement. Similarly, analyses based on replacement rates generally show the fraction of retirees who are judged unable to maintain their preretirement standard of living, but they might not convey whether retirees are experiencing poverty or are near poverty. Comparing the results of such analyses sheds additional light, enabling better understanding of the financial security of retirees.
Although a number of factors cause conclusions about the adequacy of retirement income to diverge, recent research delves further into the role of data quality, examining differences between publicly available survey data and restricted-use administrative data. (Administrative data are collected for administrative purposes by governmental agencies, such as the Internal Revenue Service or the Social Security Administration.) The vast majority of studies on the adequacy of retirement income use self-reported data because that information is widely available. However, income is known to be underreported in surveys, particularly income from employer-sponsored pensions and other retirement accounts. Ongoing research is seeking to improve current understanding of the extent and nature of underreporting and to clarify how underreporting could affect conclusions about retirement adequacy.