Public Knowledge About the Social Security Retirement Program: Differences by Race and Ethnicity
Social Security Bulletin, Vol. 83 No. 3, 2023
We use 2014–2021 survey data from the first three waves of the Understanding America Study to examine public knowledge of the Social Security retirement program. We present descriptive statistics and highlight differences in program knowledge by respondents' race or ethnicity as well as by age, education level, and sex. Social Security retirement benefits are the primary source of income for many people, and program knowledge helps individuals make optimal decisions about saving and the timing of benefit claiming. It is critical to understand any racial-ethnic disparities in retirement program knowledge and to develop solutions to address them. In this article, we find that people of color have significantly lower levels of Social Security retirement program knowledge than non-Hispanic White people. These program knowledge disparities persist across age and education levels and are compounded for women of color. We discuss the implications of these findings and suggest directions for future research.
Richard E. Chard and David Rogofsky are with the Office of Research, Evaluation, and Statistics (ORES), Office of Retirement and Disability Policy (ORDP), Social Security Administration (SSA). When this article was written, Matt Messel was a researcher with ORES, ORDP, SSA.
Acknowledgments: The authors thank Laith Alattar, Sofia Ayala, Robert Weathers, and Mark Sarney for their helpful comments and suggestions.
The findings and conclusions presented in the Bulletin are those of the authors and do not necessarily represent the views of the Social Security Administration.
Introduction
| OASI | Old-Age and Survivors Insurance |
| SSA | Social Security Administration |
| UAS | Understanding America Study |
Social Security benefits provide the majority of retirement income for more than half of Americans aged 65 or older (Dushi, Iams, and Trenkamp 2017). Structural inequities inside and outside the labor market have produced income and wealth disparities (Francis and Weller 2021; Oliver and Shapiro 2013). As a result, people of color tend to have fewer resources than non-Hispanic White people when they reach retirement age. Consequently, Social Security benefits play an even greater role in retirement security for them than for non-Hispanic White people (Hendley and Bilimoria 1999; Rabinovich, Peterson, and Smith 2017).
Knowledge about the Social Security retirement program—that is, the Old-Age and Survivors Insurance (OASI) program—plays an important role in retirement security by helping individuals make optimal decisions about saving and the timing of benefit claiming (Gustman and Steinmeier 1999; Rohwedder and van Soest 2006). Research shows that people tend to have high levels of knowledge about some OASI subject areas, such as the availability of survivor benefits, but less knowledge about others, such as how benefits are calculated (Alattar and others 2019). Previous studies also indicate that Black and Hispanic people are less knowledgeable about Social Security programs (Peterson, Smith, and Guan 2019; Yoong, Rabinovich, and Wah 2015) and inflation's effect on retirement savings (Greenwald and others 2010). It is important for researchers and policymakers to understand potential disparities in the public's knowledge about OASI and to develop solutions for addressing these disparities.
Methods
This section discusses our survey data source, the Understanding America Study (UAS), and our data compilation process.
Data
The UAS is an internet-based panel managed by the University of Southern California. The Social Security Administration (SSA) has funded surveys of Social Security program knowledge (and respondents' preferred channels for receiving program information) as UAS components since 2014. When we conducted this analysis, the UAS panel comprised approximately 9,500 U.S. households who were selected using address-based sampling. The number of households in the UAS panel continues to increase over time. If needed, participants are provided a tablet computer and Internet access. Panel members may respond to multiple surveys covering a wide range of topics, for which they receive nominal compensation. Researchers administer the Social Security program knowledge and information channel surveys on a rolling basis every 2 years to all new panel members and to any current panel member who has not taken that survey for 2 years.1
Sample
In this article, we use data from the first three waves of the Social Security program knowledge survey. If an individual participated in multiple survey waves, we used his or her most recent survey responses. More than 70 percent of the data in our analysis come from surveys completed in 2020 and 2021, and the remaining data come from 2014–2019 survey responses. We weighted results using specially calculated weights supplied by the UAS.2 These survey weights are benchmarked to the Current Population Survey's Annual Social and Economic Supplement to represent the adult noninstitutionalized U.S. population. Our sample includes 10,899 respondents. Of the weighted sample, 63.2 percent are non-Hispanic White, 11.7 percent are non-Hispanic Black, 16.6 percent are Hispanic or Latino, and 5.0 percent are Asian, Hawaiian, or Pacific Islander (Table 1).3 The remaining 3.4 percent of respondents include American Indians, Alaska Natives, and those who identify as multiracial.4 Non-Hispanic White respondents tend to be older than respondents of other racial or ethnic backgrounds. For instance, 73.6 percent of Hispanic and Latino respondents and 66.6 percent of Asian, Hawaiian, and Pacific Islander respondents are younger than 50, compared with 46.2 percent of non-Hispanic White respondents.
| Characteristic | All respondents | Non-Hispanic White | Non-Hispanic Black | Hispanic and Latino | Asian, Hawaiian, and Pacific Islander | American Indian, Alaska Native, and multiracial |
|---|---|---|---|---|---|---|
| Number (unweighted) | 10,899 | 7,185 | 887 | 1,707 | 533 | 587 |
| Percent (weighted) | 100.0 | 63.2 | 11.7 | 16.6 | 5.0 | 3.4 |
| Age | ||||||
| 18–29 | 15.4 | 12.0 | 13.5 | 26.0 | 23.1 | 22.1 |
| 30–49 | 38.4 | 34.2 | 43.8 | 47.6 | 43.5 | 45.4 |
| 50–61 | 20.8 | 21.5 | 25.0 | 15.8 | 19.3 | 19.0 |
| 62–69 | 13.4 | 16.5 | 10.3 | 6.9 | 7.7 | 7.1 |
| 70 or older | 12.1 | 15.8 | 7.5 | 3.7 | 6.5 | 6.3 |
| Education level | ||||||
| Less than high school | 8.9 | 8.1 | 10.0 | 12.2 | 6.1 | 9.0 |
| High school diploma | 30.6 | 33.0 | 29.5 | 27.5 | 14.0 | 28.6 |
| Some college | 27.7 | 25.1 | 34.6 | 36.0 | 14.9 | 29.2 |
| Bachelor's degree or higher | 32.8 | 33.8 | 25.9 | 24.4 | 65.0 | 33.4 |
| Sex | ||||||
| Men | 48.7 | 52.8 | 38.0 | 40.7 | 50.8 | 46.2 |
| Women | 51.3 | 47.2 | 62.0 | 59.3 | 49.4 | 53.8 |
| Annual household income | ||||||
| Less than $50,000 | 40.8 | 36.5 | 55.8 | 47.2 | 35.5 | 46.0 |
| $50,000–74,999 | 15.5 | 16.4 | 13.5 | 13.8 | 12.8 | 17.0 |
| $75,000–99,999 | 12.4 | 13.4 | 10.3 | 11.3 | 11.6 | 8.2 |
| $100,000 or more | 31.3 | 33.7 | 20.4 | 27.7 | 40.2 | 28.8 |
| SOURCE: Authors' calculations using UAS survey results for 2014–2021. | ||||||
The UAS Survey of Social Security Program Knowledge
The survey covers respondents' general understanding of the OASI program as well as their specific program knowledge related to benefit claiming ages.5 In this article, we focus on general program knowledge in eight different subject areas as well as knowledge specific to benefit claiming ages in six different subject areas. Box 1 lists the survey questions, arranged by subject area, and gives the possible answer choices and the correct responses. We measure retirement program knowledge as the percentage of these 14 questions a respondent answers correctly and show the results as the average percentage of correct responses for all individuals in a given demographic group.
| Subject area | Question and answers |
|---|---|
| General program knowledge | |
| Age adjustment | The amount of Social Security retirement benefits is not affected by the age at which someone starts claiming. ☐ True ☑️ False |
| Benefit calculation | Which of the following best describes how a worker's Social Security benefits are calculated? ☐ They are based on how long you work as well as your pay during the last five years that you are employed. ☑️ They are based on the average of the highest 35 years of your earnings. ☐ They are based on how much Social Security taxes you paid. ☐ They are based on your income tax bracket when you claim benefits. |
| Child survivor benefits | If a worker who pays Social Security taxes dies, any of his/her children under age 18 may claim Social Security survivor benefits. ☑️ True ☐ False |
| Claiming upon retirement | Social Security benefits have to be claimed as soon as someone retires. ☐ True ☑️ False |
| Inflation adjustment | Social Security benefits are adjusted for inflation. ☑️ True ☐ False |
| Payroll tax | Social Security is paid for by a tax placed on both workers and employers. ☑️ True ☐ False |
| Spousal benefits | Someone who has never worked for pay may still be able to claim benefits if his or her spouse qualifies for Social Security. ☑️ True ☐ False |
| Widow(er) benefits | If a worker who pays Social Security taxes dies, his/her spouse may claim Social Security survivor benefits only if they have children. ☐ True ☑️ False |
| Specific benefit claiming age knowledge | |
| Early eligibility age | One of the terms used by Social Security is early eligibility age, or EEA. To the best of your knowledge, what is your personal earliest eligibility age for claiming Social Security retirement benefits? The correct answer is 62. |
| Full retirement age | Another term used by Social Security is full retirement age, or FRA. To the best of your knowledge, what is your personal full retirement age? The correct answer ranges between 65 and 67, depending on the respondent's birth year (see https://www.ssa.gov/policy/docs |
| Relationship between retiring and claiming Social Security benefits | Based on Social Security guidelines, what is the relationship between the age at which you stop working and the age at which you can begin claiming benefits? ☐ Both occur at the same age. ☐ The age at which you stop working should be first. ☐ The Social Security claiming age should be first. ☑️ Any of these combinations are acceptable. |
| Delayed retirement credits (DRCs) | One of the factors that can affect your monthly benefits are the so-called DRCs. Which one of the following statements is correct? ☐ The DRCs are a bonus on Social Security benefits for people who have worked for at least 40 years. ☑️ The DRCs indicate by what percentage monthly benefits increase if one waits until after FRA to claim benefits. ☐ The DRCs are an increase in benefits that comes from earning income by working after age 62. |
| Eligibility age for DRCs | When are/were you first eligible to claim DRCs from the Social Security program? ☐ Early Eligibility Age (EEA) ☑️ Full Retirement Age (FRA) ☐ years old (enter number) |
| Age DRCs stop | At what age would you stop earning DRCs? The correct answer is 70. |
| SOURCE: UAS 16, 94, and 231 questionnaires. | |
| NOTES: Some of the questionnaire's wording has been slightly modified for contextual clarity. | |
| Correct answers are noted or indicated by ☑️. | |
Findings
We present descriptive findings on respondents' knowledge about the Social Security retirement program by race and ethnicity. We further examine these findings by age, educational level, and sex across racial-ethnic groups.
Program Knowledge by Race and Ethnicity
Table 2 presents by race and ethnicity the average percentages of correct responses to all OASI program knowledge questions, general program questions, and specific benefit claiming age questions. The table includes age-adjusted results alongside the unadjusted results because OASI program knowledge tends to be greater among people approaching retirement age than among younger people (Alattar and others 2019).6 Overall, participants answered 50.7 percent of the OASI questions correctly. Respondents were more knowledgeable about general program areas (70.6 percent correct) than about benefit claiming age areas (24.2 percent correct). People of color had significantly lower levels of overall retirement program knowledge than non-Hispanic White respondents, even when adjusting for age. Non-Hispanic White respondents answered 53.9 percent of all questions correctly, compared with 48.8 percent for Asian, Hawaiian, and Pacific Islander respondents, 45.3 percent for non-Hispanic Black respondents, and 43.3 percent for Hispanic and Latino respondents.7 Although each of these differences are statistically significant, the magnitudes of the differences are small, particularly between non-Hispanic White and Asian, Hawaiian, and Pacific Islander respondents. The differences remain significant, however, even when controlling for age differences between racial-ethnic groups.8 People of color also had, on average, lower levels of general program knowledge than non-Hispanic White respondents. However, on questions about Social Security claiming ages, age-adjusted results for Asian, Hawaiian, and Pacific Islander respondents were about the same as those for non-Hispanic White respondents.
| Group | Unadjusted | Age-adjusted |
|---|---|---|
| All questions combined | ||
| All respondents | 50.7 | . . . |
| Non-Hispanic White (reference category) | 53.9 | . . . |
| Non-Hispanic Black | 45.3* | 46.8* |
| Hispanic and Latino | 43.3* | 46.4* |
| Asian, Hawaiian, and Pacific Islander | 48.8* | 51.1* |
| American Indian, Alaska Native, and multiracial | 48.9* | 53.3 |
| General program questions | ||
| All respondents | 70.6 | . . . |
| Non-Hispanic White (reference category) | 73.9 | . . . |
| Non-Hispanic Black | 65.8* | 67.1* |
| Hispanic and Latino | 62.4* | 65.7* |
| Asian, Hawaiian, and Pacific Islander | 68.0* | 69.6* |
| American Indian, Alaska Native, and multiracial | 69.8* | 72.5 |
| Specific benefit claiming age questions | ||
| All respondents | 24.2 | . . . |
| Non-Hispanic White (reference category) | 27.4 | . . . |
| Non-Hispanic Black | 18.0* | 19.7* |
| Hispanic and Latino | 17.2* | 20.5* |
| Asian, Hawaiian, and Pacific Islander | 23.1* | 26.1 |
| American Indian, Alaska Native, and multiracial | 21.1* | 24.3* |
| SOURCE: Authors' calculations using UAS survey results for 2014–2021. | ||
| NOTES: . . . = not applicable.
* = difference from reference category is statistically significant at the 0.05 level.
|
||
Further Differences in Overall Program Knowledge by Age, Education Level, and Sex
Table 3 also presents the percentages of correct responses to all OASI program knowledge questions but provides further subgroup breakdowns by age, education level, and sex.9
| Characteristic | All respondents | Non-Hispanic White | Non-Hispanic Black | Hispanic and Latino | Asian, Hawaiian, and Pacific Islander | American Indian, Alaska Native, and multiracial |
|---|---|---|---|---|---|---|
| Total | 50.7 | 53.9 | 45.3 | 43.3 | 48.8 | 48.9 |
| Age | ||||||
| 18–29 | 39.9 | 41.5 | 39.3 | 36.6 | 42.6 | 39.9 |
| 30–49 | 47.1 | 49.5 | 43.0 | 42.8 | 47.8 | 47.2 |
| 50–61 | 53.7 | 56.5 | 46.2 | 47.5 | 53.5 | 54.4 |
| 62–69 | 60.5 | 62.1 | 54.1 | 53.1 | 56.3 | 63.6 |
| 70 or older | 59.6 | 60.8 | 54.1 | 52.7 | 54.6 | 55.4 |
| Education level a | ||||||
| Less than high school | 41.4 | 44.2 | 38.9 | 36.4 | 38.3 | 38.6 |
| High school diploma | 48.5 | 51.3 | 40.1 | 41.6 | 41.4 | 45.7 |
| Some college | 51.4 | 54.8 | 48.1 | 44.9 | 50.4 | 48.1 |
| Bachelor's degree or higher | 57.5 | 60.6 | 50.9 | 49.2 | 52.5 | 57.2 |
| Sex | ||||||
| Men | 53.3 | 56.0 | 46.3 | 45.5 | 49.8 | 52.2 |
| Women | 48.3 | 51.6 | 44.7 | 41.3 | 47.8 | 46.1 |
| SOURCE: Authors' calculations using UAS survey results for 2014–2021. | ||||||
| a. Excludes respondents younger than 25. | ||||||
Age. Within each racial-ethnic group, retirement program knowledge increases with age. However, racial and ethnic disparities in program knowledge persist across age groups.10 Disparities are larger for people approaching retirement age (50–61) than for younger people. The average difference in correct responses between non-Hispanic White and non-Hispanic Black respondents is 2.2 percentage points for those aged 18–29 and 10.3 percentage points for those aged 50–61. The average difference in correct responses between respondents in the non-Hispanic White group and those in the Hispanic and Latino group is 4.9 percentage points for those aged 18–29 and 9.0 percentage points for those aged 50–61. This could suggest that non-Hispanic White people face fewer barriers or have more opportunities to learn about Social Security programs than do people of other races and ethnicities as they approach retirement age. It could also suggest a cohort effect, in which disparities in retirement program knowledge are smaller for younger cohorts than for older ones. Only longitudinal analysis of retirement program knowledge by race and ethnicity could elucidate the reason for the widening knowledge gap by age.
Education level. Table 3 also shows that within each racial-ethnic group, retirement program knowledge is greater for people with higher levels of education. For instance, non-Hispanic Black respondents with a bachelor's degree answered an average of 50.9 percent of questions correctly, compared with 38.9 percent for those without a high school diploma. At each educational level, however, non-Hispanic White respondents had significantly higher levels of knowledge than respondents in other racial-ethnic groups. For instance, among people with a bachelor's degree or higher, non-Hispanic White respondents scored 11.4 percentage points higher than Hispanic and Latino respondents (60.6 percent versus 49.2 percent correct). Among those without a high school diploma, non-Hispanic White respondents scored 7.8 percentage points higher than Hispanic and Latino respondents (44.2 percent versus 36.4 percent correct). This suggests that program knowledge disparities between non-Hispanic White people and people of color are not explained entirely by differences in educational opportunities.11
Differences in retirement program knowledge by education level also reveal heterogeneity within each racial-ethnic group. The increase in knowledge for respondents with a bachelor's degree versus those with only a high school education is particularly large among Asian, Hawaiian, and Pacific Islander respondents (11.1 percentage points) and non-Hispanic Black respondents (10.8 percentage points). However, the increase across these education levels is somewhat smaller among Hispanic and Latino respondents (7.6 percentage points). Future research could explore educational (and other) factors underlying heterogeneity in retirement program knowledge within racial-ethnic groups.
Sex. Within each racial-ethnic group, men tend to have slightly greater retirement program knowledge than women. This difference is larger among non-Hispanic White respondents (4.4 percentage points) and Hispanic and Latino respondents (4.2 percentage points). By contrast, the difference is smaller among Asian, Hawaiian, and Pacific Islander respondents (2.0 percentage points) and non-Hispanic Black respondents (1.6 percentage points). Women of color tend to have the lowest levels of retirement program knowledge. Program knowledge among Hispanic and Latino women is 14.7 percentage points lower than that of non-Hispanic White men (41.3 percent versus 56.0 percent correct). These intersecting program knowledge disparities by race and ethnicity and sex suggest that women of color are an important target for information interventions.
These UAS data indicate that people of color have significantly lower levels of retirement program knowledge than non-Hispanic White people. These disparities persist at different ages and levels of education, and they are compounded for women of color.
Conclusion
OASI program knowledge helps people make optimal saving and benefit claiming decisions (Gustman and Steinmeier 1999; Rohwedder and van Soest 2006). In this article, we used data from the UAS to measure the public's knowledge about the Social Security retirement program by race and ethnicity. Overall, U.S. adults were able to correctly answer only slightly more than half (50.7 percent) of questions covering various OASI program areas. We found significant racial-ethnic disparities in retirement program knowledge that persist across age groups and education levels and are compounded for women of color. Because these disparities may have real world consequences for the retirement security of people of color, understanding these program knowledge disparities in greater depth is an important research aim. Knowledge of specific topics involving benefit claiming ages was particularly low among non-Hispanic Black and Hispanic or Latino respondents, with both groups answering about 20 percent of questions correctly. Because understanding the optimal age to claim benefits can affect Social Security benefit levels and financial security in retirement, further research could investigate possible causes and policy responses.
Our preliminary findings point toward three potential directions for future research. First, it is important to understand the structural barriers that create disparities in retirement program knowledge by race and ethnicity, including unequal educational opportunities. Because of systemic racial discrimination in the United States, people of color have fewer educational opportunities on average than White people (Noguera, Pierce, and Ahram 2015). Because OASI program knowledge is related to educational attainment, structural barriers to educational opportunity likely play a role in program knowledge disparities by race and ethnicity. Still, program knowledge disparities exist at each educational level. For instance, the difference in program knowledge between respondents in the non-Hispanic White group and those in the Hispanic and Latino group or the non-Hispanic Black group is roughly 10 percentage points at both the high school diploma and bachelor's degree levels (Table 3). This suggests that structural barriers beyond unequal educational opportunities play a role in disparities in retirement program knowledge by race and ethnicity.
Myriad structural barriers, such as unequal access to retirement planning information provided in workplaces (Francis and Weller 2021), language barriers when information is not readily available in a person's primary language (Rabinovich, Peterson, and Smith 2017), or unequal access to program information across social networks that are segregated by race and sex (McDonald and Day 2010), may be driving retirement program knowledge disparities. Further exploration of quantitative data, such as those provided in the UAS, may provide additional insight. Nevertheless, an in-depth exploration of the structural barriers that drive program knowledge disparities may require qualitative research that extends beyond what is available in current surveys. These qualitative studies could focus on specific racial or ethnic groups, identifying relevant factors and potential solutions for addressing barriers to obtaining OASI program knowledge. One example of such qualitative research is Rabinovich, Peterson, and Smith (2017). Their study, based on the UAS sample, found that Hispanic people of different ancestry groups (Mexican, Puerto Rican, and Cuban) are interested in having access to additional Spanish-language information resources about Social Security programs.
Second, it is important to understand how retirement program knowledge disparities develop across the life course. The preliminary finding that program knowledge disparities by race and ethnicity are larger among people approaching retirement age (aged 50–61) is concerning. During these years, many people develop saving and benefit claiming plans that predict their future retirement security. Because our study is cross-sectional, it cannot differentiate whether the larger program knowledge disparities for people aged 50–61 are due to an accumulating disadvantage in learning about Social Security among people of color or to cohort effects (which would suggest that program knowledge disparities are smaller for younger cohorts). Additional waves of the UAS program knowledge survey may provide a longitudinal basis to understand how retirement program knowledge disparities develop over the life course.
Finally, future research could focus on access to retirement program information among women of color, who showed low levels of retirement program knowledge in the UAS survey. A large body of literature demonstrates how women with intersecting identities, such as race and sex, are negatively affected by structural barriers, such as diminished labor market opportunities (Moore and Ghilarducci 2018) and diminished retirement security (Lahey 2018). Little research to date has examined whether and how these disparities affect access to Social Security program information. Some research has focused on access to general retirement planning information among women of color (Angel, Prickett, and Angel 2014; Joo and Pauwels 2002). This research could provide direction for studies on access to OASI program information. Again, efforts to understand how intersecting identities relate to retirement program knowledge disparities may benefit from a qualitative approach that explores how experiences vary for women of different racial and ethnic backgrounds, ancestry groups, income levels, and geographic areas, among other potential factors.
It is also important to research potential solutions addressing these disparities. The Social Security information channels survey in the UAS provides one path to do so. In a forthcoming study, we will investigate the use of and attitudes toward different channels of information on retirement planning in general and Social Security benefit planning in particular across different racial and ethnic backgrounds. These information channels include employers, financial planners, and SSA, among others. By measuring differences in the perceived accessibility, understandability, and accuracy of these information channels, we will explore opportunities to use these channels to reduce both the retirement program knowledge disparities and the barriers to their use that may exist for certain racial or ethnic groups.
Appendix A
| Group | Percent correct |
|---|---|
| All questions combined | |
| All respondents | 50.7 |
| Non-Hispanic White (reference category) | 53.9 |
| Non-Hispanic Black | 45.3* |
| Hispanic and Latino | 43.3* |
| Asian, Hawaiian, and Pacific Islander | 48.8* |
| Asian | 49.1* |
| Hawaiian and Pacific Islander | 36.9* |
| American Indian, Alaska Native, and multiracial | 48.9* |
| American Indian and Alaska Native | 41.5* |
| More than one race | 50.0* |
| General program questions | |
| All respondents | 70.6 |
| Non-Hispanic White (reference category) | 73.9 |
| Non-Hispanic Black | 65.8* |
| Hispanic and Latino | 62.4* |
| Asian, Hawaiian, and Pacific Islander | 68.0* |
| Asian | 68.3* |
| Hawaiian and Pacific Islander | 55.2* |
| American Indian, Alaska Native, and multiracial | 69.8* |
| American Indian and Alaska Native | 62.5* |
| More than one race | 70.8* |
| Specific benefit claiming age questions | |
| All respondents | 24.2 |
| Non-Hispanic White (reference category) | 27.4 |
| Non-Hispanic Black | 18.0* |
| Hispanic and Latino | 17.2* |
| Asian, Hawaiian, and Pacific Islander | 23.1* |
| Asian | 23.4 |
| Hawaiian and Pacific Islander | 12.5* |
| American Indian, Alaska Native, and multiracial | 21.1* |
| American Indian and Alaska Native | 13.4* |
| More than one race | 22.2* |
| SOURCE: Authors' calculations using UAS survey results for 2014–2021. | |
| NOTE: * = difference from reference category is statistically significant at the 0.05 level. | |
Notes
1 Alattar, Messel, and Rogofsky (2018) provide additional information on UAS methodology.
2 Alattar, Messel, and Rogofsky (2018) explain: “Poststratification weights are created using a raking algorithm. The algorithm compares relative frequencies within the target population with relative frequencies in the survey sample by race, sex and age, sex and education, household size and total household income, census region, and urbanicity. When a researcher combines responses from two or more UAS surveys, the UAS team will provide weights unique to the combined data set based on the procedure described above. Alternatively, the UAS team can provide custom poststratification weights using specific raking factors chosen by the researcher.”
3 We chose to combine the Asian population with the Hawaiian and Pacific Islander population to create a sample size large enough to measure knowledge by demographic factors, such as age, education, and sex. Of the 533 respondents in the combined category, 502 identified as Asian and 31 as Hawaiian or Pacific Islander.
4 In these categories, 475 individuals identified as more than one race and 112 individuals identified as American Indian or Alaska Native.
5 Although the survey also explores respondents' knowledge of the disability programs that SSA administers (Disability Insurance and Supplemental Security Income), this study focuses exclusively on OASI program knowledge.
6 For the age adjustments, we reweight each racial-ethnic group to match the age composition of non-Hispanic White respondents.
7 Appendix Table A-1 repeats Table 2 with additional detail for racial-ethnic groups that we collapsed for our analysis: namely, the Asian, Native Hawaiian and Pacific Islander, American Indian and Alaska Native, and multiracial subgroups. With results unadjusted for age, Table A-1 shows that respondents in the Hawaiian and Pacific Islander subgroup and in the American Indian and Alaska Native subgroup have less OASI program knowledge than other groups do, with overall scores of 36.9 percent and 41.5 percent, respectively, compared with 53.9 percent among non-Hispanic White respondents.
8 We do not adjust by factors such as education and sex, which are also associated with OASI program knowledge. Some of the racial-ethnic groups in our sample (such as the non-Hispanic Black group and the Hispanic and Latino group) comprise higher percentages of women and lower percentages of individuals with a bachelor's degree, which are associated with lower levels of retirement program knowledge. Consequently, some of the program knowledge disparities we identify across racial-ethnic groups may also be due to disparities by education or sex. Because this is a descriptive article and structural barriers create program knowledge disparities by race and ethnicity as well as by educational attainment and sex, we leave these additional factors unadjusted. Instead, we provide statistics on program knowledge differences, considering the intersectionality between race and ethnicity, educational attainment, and sex (in Table 3). We believe that these descriptive statistics may be one step toward further investigation of the structural barriers that exist at different intersections of race, ethnicity, educational attainment, and sex.
9 We omit age-adjusted figures from Table 3 because they do not change the statistical significance relative to the unadjusted findings.
10 The exception is that non-Hispanic White respondents and Asian, Hawaiian, and Pacific Islander respondents aged 18–29 have similar levels of knowledge, on average.
11 We found that the knowledge differences at the high school diploma, some college, and bachelor's degree or higher levels were significant, even when accounting for age differences between racial-ethnic groups. The sample size for the population with less than a high school education was not large enough to conduct this analysis.
References
Alattar, Laith, Matt Messel, and David Rogofsky. 2018. “An Introduction to the Understanding America Study Internet Panel.” Social Security Bulletin 78(2): 13–28. https://www.ssa.gov/policy/docs/ssb/v78n2/v78n2p13.html.
Alattar, Laith, Matt Messel, David Rogofsky, and Mark A. Sarney. 2019. “The Use of Longitudinal Data on Social Security Program Knowledge.” Social Security Bulletin 79(4): 1–10. https://www.ssa.gov/policy/docs/ssb/v79n4/v79n4p1.html.
Angel, Jacqueline L., Kate C. Prickett, and Ronald J. Angel. 2014. “Retirement Security for Black, Non-Hispanic White, and Mexican-Origin Women: The Changing Roles of Marriage and Work.” Journal of Women, Politics & Policy 35(3): 222–241.
Dushi, Irena, Howard M. Iams, and Brad Trenkamp. 2017. “The Importance of Social Security Benefits to the Income of the Aged Population.” Social Security Bulletin 77(2): 1–12. https://www.ssa.gov/policy/docs/ssb/v77n2/v77n2p1.html.
Francis, Dania V., and Christian E. Weller. 2021. “Race, Ethnicity, and Retirement Security in the United States.” Oxford Research Encyclopedia of Economics and Finance. https://doi.org/10.1093/acrefore/9780190625979.013.741.
Greenwald, Mathew, Arie Kapteyn, Olivia S. Mitchell, and Lisa Schneider. 2010. “What Do People Know About Social Security?” Pension Research Council Working Paper No. WRM-792-SSA. Philadelphia, PA: Pension Research Council, University of Pennsylvania Wharton School.
Gustman, Alan L., and Thomas L. Steinmeier. 1999. “Effects of Pensions on Savings: Analysis with Data from the Health and Retirement Study.” Carnegie-Rochester Conference Series on Public Policy 50(1): 271–324.
Hendley, Alexa A., and Natasha F. Bilimoria. 1999. “Minorities and Social Security: An Analysis of Ethnic Differences in the Current Program.” Social Security Bulletin 62(2): 59–64. https://www.ssa.gov/policy/docs/ssb/v62n2/v62n2p59.pdf.
Joo, So-Hyun, and Vanda W. Pauwels. 2002. “Factors Affecting Workers' Retirement Confidence: A Gender Perspective.” Journal of Financial Counseling and Planning 13(2): 1–10.
Lahey, Joanna L. 2018. “Understanding Why Black Women Are Not Working Longer.” In Women Working Longer: Increased Employment at Older Ages, edited by Claudia Goldin, and Lawrence F. Katz, 85–110. Chicago, IL: University of Chicago Press.
McDonald, Steve, and Jacob C. Day. 2010. “Race, Gender, and the Invisible Hand of Social Capital.” Sociology Compass 4(7): 532–543.
Moore, Kyle K., and Teresa Ghilarducci. 2018. “Intersectionality and Stratification in the Labor Market.” Generations 42(2): 34–40.
Noguera, Pedro, Jill Pierce, and Roey Ahram, editors. 2015. Race, Equity, and Education: Sixty Years from Brown. Cham, Switzerland: Springer International Publishing.
Oliver, Melvin, and Thomas Shapiro. 2013. Black Wealth/White Wealth: A New Perspective on Racial Inequality. New York, NY: Routledge.
Peterson, Janice, Barbara A. Smith, and Qi Guan. 2019. “Hispanics' Knowledge of Social Security: New Evidence.” Social Security Bulletin 79(4): 11–24. https://www.ssa.gov/policy/docs/ssb/v79n4/v79n4p11.html.
Rabinovich, Lila, Janice Peterson, and Barbara A. Smith. 2017. “Hispanics' Understanding of Social Security and the Implications for Retirement Security: A Qualitative Study.” Social Security Bulletin 77(3): 1–14. https://www.ssa.gov/policy/docs/ssb/v77n3/v77n3p1.html.
Rohwedder, Susan, and Arthur van Soest. 2006. “The Impact of Misperceptions About Social Security on Saving and Well-Being.” MRRC Working Paper No. 2006-118. Ann Arbor, MI: University of Michigan Retirement Research Center.
Yoong, Joanne, Lila Rabinovich, and Saw Htay Wah. 2015. “What Do People Know About Social Security?” CESR-Schaffer Working Paper No. 2015-022. Los Angeles, CA: University of Southern California Center for Economic and Social Research. https://cesr.usc.edu/documents/WP_2015_022.pdf.