Hispanics' Knowledge of Social Security: New Evidence

Introduction

There is growing concern about the adequacy of retirement planning and saving in the United States. Although many Americans may face economic insecurity in retirement (Government Accountability Office 2015; Rhee and Boivie 2015; Williams and Jackson 2015), studies suggest that certain demographic groups are at particular risk. Hispanics are one such group, facing challenges that include comparatively low-wage jobs, low levels of wealth, limited health insurance coverage, and longer life expectancy (Hopkins 2014). As a result, Hispanics are at greater risk than the general population of having low levels of retirement savings and, therefore, of relying on Social Security benefits as a major source of retirement income (Rabinovich, Peterson, and Smith 2017).

Because Hispanics are likely to rely on Social Security income in retirement, it is important that they be well-informed about program provisions. However, recent research has shown that Hispanics are less knowledgeable than other groups about Social Security (Yoong, Rabinovich, and Wah 2015; Rabinovich, Peterson, and Smith 2017). Increasing Hispanics' understanding of Social Security is important not only for individual Hispanics but also for the Social Security Administration (SSA) and for all organizations supporting the financial and retirement needs of Hispanics. Because Social Security benefits represent a substantial part of Hispanics' retirement income, adequate knowledge of the program and its benefits can support decisions that lead to a secure retirement. Research shows that providing information about Social Security benefits and other financial matters improves financial literacy and that higher levels of financial literacy are correlated with better financial decisions.¹

This article updates and extends the findings of Rabinovich, Peterson, and Smith (2017), which explored these same topics using data from focus groups. By contrast, this article uses data from a large Internet survey panel to explore Hispanics' knowledge of aspects of Social Security both at the broad program level and at the narrower benefit-specific level. Our research makes important contributions to the study of Hispanics and Social Security by identifying and statistically testing differences in Social Security program- and benefits-level knowledge between non-Hispanic whites and Hispanics as well as among Hispanics across ancestry and primary-language groups. We also examine respondents' perceptions of their retirement preparedness and the helpfulness of different SSA information sources and delivery methods. Our findings should be of interest to researchers, financial advisors, and policymakers interested in improving retirement security for U.S. Hispanics. In this article, we use the term “Hispanic” to refer to any “person of Cuban, Mexican, Puerto Rican, South or Central American, or other Spanish culture or origin regardless of race,” as defined by the Office of Management and Budget (Census Bureau 2018).

Characteristics of the U.S. Hispanic Population

According to the Census Bureau's QuickFacts interactive data feature (https://www.census.gov/quickfacts/fact/table/UAS#), Hispanics constitute the nation's largest minority group, at 18 percent of the population as of July 2018. That share is projected to increase to 27 percent by 2050 (Census Bureau 2014b, Table 11). In addition, the Hispanic population aged 65 or older is projected to quintuple from 2012 through 2050. By 2050, the share of Americans aged 65 or older who are Hispanic will exceed 18 percent (Hummer and Hayward 2015, 21). Consequently, the share of Hispanics among Social Security beneficiaries is expected to increase (Rabinovich, Peterson, and Smith 2017).

The Hispanic community is not homogeneous in terms of ancestry or language. According to studies by the Pew Research Center, about 62 percent of Hispanics in America are of Mexican ancestry, almost 10 percent are Puerto Rican, and Cubans and Salvadorans each constitute about 4 percent (Krogstad and Noe-Bustamante 2019).² Among Hispanic Social Security beneficiaries, the three largest ancestry groups are Mexicans, Puerto Ricans, and Cubans, accounting for 52 percent, 14 percent, and 10 percent, respectively (Martin 2007).

In addition to the cultural differences between Latin American places of origin, these U.S. Hispanic subgroups differ in terms of key demographic characteristics, such as age, educational attainment, and income. For example, in 2011, of the three largest subgroups, Mexicans had the youngest median age and Cubans had the oldest; Cubans were most likely to have at least a bachelor's degree, and Mexicans were least likely; and Cubans had the highest median income, and Puerto Ricans had the lowest (Lopez, Gonzalez-Barrera, and Cuddington 2013).

Language preference and proficiency also vary among U.S. Hispanics. According to the Pew Research Center, 36 percent are bilingual, 38 percent speak mainly Spanish, and 25 percent speak mainly English (Krogstad and Gonzalez-Barrera 2015). Further, language preferences vary substantially across first-, second-, and third-generation U.S. Hispanics. Among first-generation families, 61 percent consider Spanish their primary language; that figure falls to 8 percent among second-generation Hispanics and to 1 percent among the third generation (Taylor and others 2012, Chapter IV).

Hispanics tend to be socioeconomically disadvantaged relative to other racial/ethnic groups in the United States. Hispanic adults have the lowest rates of high school and college graduation, are more concentrated in low-wage jobs, and have lower incomes and health insurance coverage rates (Gassoumis, Wilber, and Torres-Gil 2008; Hummer and Hayward 2015). In 2013, the median wealth of a Hispanic family ($14,000) was only one-tenth the median wealth of a non-Hispanic white family ($134,000) (Boshara, Emmons, and Noeth 2015, 7–9).

Despite their socioeconomic disadvantages, Hispanics' life expectancy is greater than that of other population groups. Hispanic men aged 65 in 2014 can expect to live to age 85, versus 83 for non-Hispanic white men; Hispanic women aged 65 in 2014 can expect to live to age 87, versus 86 for non-Hispanic white women (Census Bureau 2014a, Table 2). Higher life expectancy places additional financial burdens on older Hispanics, which in the context of low incomes and lack of savings “predictably lead[s] to aggravated economic problems in old age” (Gassoumis, Wilber, and Torres-Gil 2008, 3), with clear implications for retirement income security.

Hispanics and Retirement Saving

Studies find that Hispanic workers are less likely than other workers to be covered by employer-sponsored retirement plans and Hispanic households are less likely than other households to have dedicated retirement savings (Rhee 2013). Prudential Research (2014) found that only 19 percent of surveyed Hispanics had an individual retirement account (IRA), compared with 39 percent of the general population. In addition, 38 percent of Hispanics participated in a workplace-based retirement plan, such as a 401(k), 403(b), or 457, compared with 51 percent of the general population; and 16 percent of Hispanics had a workplace pension plan, compared with 23 percent of the general population. The National Council of La Raza (2015) found that Hispanics were more likely than other groups to work for an employer that did not offer a retirement plan.³ That study also found that Hispanics who have access to retirement plans at work are less likely to participate in them than other groups are.

Studies on retirement preparedness suggest that Hispanics may place a lower priority on saving for retirement and that they engage less in retirement planning than other demographic groups, often because of competing near-term financial goals such as reducing debt or saving to send their children to college or to buy a home (Prudential Research 2014). Some analysts emphasize the importance of interpreting findings about Hispanics' financial priorities and goals in the contexts of economic realities and core cultural values. They advise observers to recognize that the importance of retirement planning is “a culturally derived concept” and that many Hispanics may “hold on to the value that retirement is a step in life where they will be supported by the children they raised with so much care” (Korzenny 2015).

Importance of Social Security to Hispanics

Social Security benefits constitute a significant proportion of retirement income for Hispanic individuals and households. Among Hispanic beneficiaries aged 65 or older in 2014, 42 percent of married couples and 59 percent of unmarried persons relied on Social Security for 90 percent or more of their income. By comparison, among non-Hispanic white beneficiaries aged 65 or older, 20 percent of married couples and 41 percent of unmarried persons relied on Social Security for 90 percent or more of their retirement income in 2014 (SSA 2016a, Table 9.A3).⁴

Social Security benefits are particularly important for Hispanics in large part because that group is less likely to receive retirement income from other sources such as employer-sponsored retirement plans, as noted earlier. Additionally, because Hispanics tend to have lower average lifetime earnings than do workers overall, they are helped by the progressive formula that determines an individual's Social Security benefit levels because it replaces a larger percentage of preretirement earnings for low earners than it does for high earners.⁵ Further, with their longer life expectancy, Hispanics benefit from guaranteed Social Security income that is adjusted annually for inflation (SSA 2016b). Finally, unlike savings in other retirement plans, Social Security accruals cannot be diverted to other uses. This is important because supporting other members of multigenerational Hispanic families often competes with retirement saving as a financial priority (Prudential Research 2014).

Hispanics' Knowledge About Social Security

Studies such as Greenwald and others (2010) have found that many people do not know enough about the Social Security program to make informed retirement decisions. Previous research that noted disparities in Social Security knowledge across different racial/ethnic groups found Hispanics to be among those at the greatest disadvantage (Yoong, Rabinovich, and Wah 2015). Rabinovich, Peterson, and Smith (2017, Table 1), using Internet panel data, found that Hispanics were less knowledgeable than non-Hispanic whites about Social Security's retirement program and benefits. Using focus groups, that study examined differences in Social Security knowledge across Hispanic ancestry (Mexican, Puerto Rican, and Cuban) and primary-language (English and Spanish) groups. Although no clear pattern of differences emerged across the ancestry groups, English speakers were more knowledgeable than Spanish speakers about Social Security programs and benefits. Regardless of language and ancestry groups, focus-group participants were more knowledgeable about Social Security program-level aspects than about benefit-specific aspects (Rabinovich, Peterson, and Smith 2017). We update that study and contribute to the literature by using Internet panel data to see if differences across ancestry and language groups are statistically significant.

Conceptual Framework and Research Questions

Financial-literacy research shows that providing individuals with information about retirement issues can affect both their knowledge and their behavior. Lusardi and Mitchell (2014) emphasized the link between financial literacy and economic behavior, particularly in making retirement-related decisions. Allen and others (2016) demonstrated that workers who attend employer-provided retirement seminars increase their financial literacy and subsequently change their retirement plans. Smith and Couch (2014) and Smith (2015) showed that workers who received SSA's annual benefits and earnings statement were more knowledgeable about Social Security programs and benefits than workers who did not. To improve Hispanics' knowledge of Social Security, policymakers and providers of financial and retirement advice will have to identify the types of Social Security information that Hispanics most need and the means by which they prefer to receive it.

With data from the 2017 Understanding America Study (UAS) Internet panel, we seek to extend previous research by determining whether differences in Social Security knowledge between Hispanics and non-Hispanic whites, and across Hispanic ancestry and language groups, are statistically significant. We also examine the Internet panel participants' perceptions of their retirement preparedness and the helpfulness of different SSA information sources and delivery methods, and assess the significance of any cross-subgroup differences.

Research Questions

• Are there significant differences in the perceptions of retirement readiness between Hispanics and non-Hispanic whites?
• What do Hispanics know about Social Security?
  • Are there significant differences in Social Security knowledge between Hispanics and non-Hispanic whites? Across Hispanic ancestry and language groups?
  • Are there significant differences between Social Security program-level knowledge and knowledge specific to Social Security benefits?
• Are there significant differences between Hispanics and non-Hispanic whites and across Hispanic ancestry and language subgroups in views of the helpfulness of information from SSA and the means by which SSA provides it?

Method

We examine the Internet panel participants' answers to three sets of questions. The first set addresses the perceived adequacy of the participants' own knowledge of financial issues related to retirement. The second set addresses participants' knowledge of selected aspects of Social Security programs and benefits. The third set addresses participants' perceptions of the helpfulness of SSA information and delivery strategies. We list the questions verbatim below; in the tables and in the discussion of our findings, we paraphrase the wording of some of the questions.

Questions on Perceived Adequacy of Own Knowledge

How knowledgeable do you feel about the following financial issues? Do you feel very knowledgeable, somewhat knowledgeable, not too knowledgeable, or not at all knowledgeable when it comes to…?

1. How inflation will affect your retirement
2. How much you will need to have saved to retire comfortably
3. How the Social Security system works
4. How long you might live in retirement
5. How to invest your retirement money
6. How to manage your spending in retirement

Questions on Social Security Programs and Benefits [with correct answers appended]

Program-level aspects. True or false:

1. Someone who has never worked for pay may still be able to claim benefits if one's spouse qualifies for Social Security. [true]
2. Social Security is paid for by a tax placed on both workers and employers. [true]
3. Workers who pay Social Security taxes are entitled to Social Security disability benefits if they become disabled and are no longer able to work. [true]
4. If a worker who pays Social Security taxes dies, any of his/her children under age 18 may claim Social Security survivor benefits. [true]

Benefit-level aspects. True or false:

1. Social Security benefits are not affected by the age at which someone starts claiming. [false]
2. Social Security benefits are adjusted for inflation. [true]
3. Social security benefits have to be claimed as soon as someone retires. [false]
4. Retired people who continue to earn income from working or investments may have to pay tax on their Social Security benefits. [true]

Questions on Helpfulness of SSA Information

Listed below are some ways the Social Security Administration could provide information to working Americans. Please indicate how helpful each of the following would be to you: very helpful, somewhat helpful, not too helpful, or not at all helpful.

1. Provide written advice and material (via the website or mailed directly) on how to plan for retirement
2. Develop webinars or online video about how to plan for retirement or apply for Social Security benefits
3. Provide worksheets online or by mail to people when they turn age 60 to help them figure out how much they need to be able to afford to stop working or earn less money
4. Provide more information about the financial solvency of the Social Security system and the amount of money in the Social Security trust fund
5. Provide information via the website or statement about how much people can expect to be deducted from their Social Security retirement in order to pay for Medicare premiums
6. Provide a calculator on the Social Security website to help people estimate how much in taxes will be owed on their benefits after they start claiming, based on their expected assets and earnings at the time
7. Provide public-service announcements for television or radio on retirement planning issues

Data

We use data from the UAS, an ongoing Internet panel managed by the Center for Economic and Social Research at the University of Southern California. To establish the pool of respondents, the UAS uses postal codes to draw a nationwide random sample of individuals and invites them to join the study. Invited individuals then choose whether to participate in the study and, if so, whether to answer particular UAS surveys. The panel is recruited by means of address-based sampling. Because the UAS is an Internet panel, respondents answer the surveys wherever they are and whenever they wish to participate, using a computer, tablet, or smart phone. Tablets and broadband Internet are provided to anyone willing to participate but lacking a computer or Internet access. UAS surveys are administered in both English and Spanish.

The full UAS panel comprises adults nationwide aged 18 or older who respond to surveys once or twice a month. The UAS includes a number of different surveys on a variety of topics. Data from all the surveys in the UAS are linked so that a large amount of information is available about panel members. This information includes financial behavior and financial literacy, cognitive ability, personality traits, and knowledge of Social Security.⁶

We use 2017 results of the Social Security–related UAS survey 16 (UAS16), which focused on Social Security program and benefit knowledge and included questions on preferred means of receiving SSA information.⁷ The 5,288 participants who had responded to the 2017 UAS16 survey at the time of our research included 4,245 non-Hispanic whites and 410 Hispanics, including 260 Mexicans, 37 Puerto Ricans, and 13 Cubans. We focus on those three groups because, as noted earlier, they are the three largest ancestry groups among Hispanic Social Security beneficiaries (Martin 2007). Of the Hispanic respondents indicating a primary language, there were 333 English speakers and 70 Spanish speakers. Table 1 provides descriptive statistics for our study sample.

Analysis and Discussion

Table 2 shows that non-Hispanic whites felt more knowledgeable than Hispanics about each of six retirement-related financial issues in 2017. Between 51 percent and 67 percent of non-Hispanic whites felt knowledgeable about these issues; for Hispanics, the percentages ranged from 35 percent to 50 percent. For each issue, the difference between the groups is statistically significant. Both groups felt least knowledgeable about how long they might live in retirement and how to invest their retirement money. Similarly, English-speaking Hispanics perceived themselves as more knowledgeable than Spanish-speaking Hispanics did, and for each issue, the difference in self-assessed knowledge is statistically significant.

Table 3 compares levels of knowledge about Social Security program- and benefit-level aspects between Hispanics and non-Hispanic whites and across Hispanic ancestry and language groups. For each aspect, knowledge is assessed as the percentage of respondents who correctly identify whether a statement is true or false. Large percentages of both non-Hispanic whites and Hispanics (ranging from 75 percent to 91 percent) knew the Social Security program aspects. Hispanics, however, scored much lower than non-Hispanic whites on the benefit-specific aspects. Although 78 percent of Hispanics knew that benefit amounts are affected by claiming age, the percentages with knowledge of the other benefit aspects ranged from 58 percent to 70 percent.

Across the Hispanic ancestry and primary-language groups, there is greater knowledge of Social Security program-level aspects than of benefit-specific aspects. At least 75 percent of the respondents in each ancestry and language group knew that Social Security offers disability benefits and that children may qualify for survivor benefits, and the percentages who knew that the program is funded by payroll taxes, and that individuals who never worked can receive benefits if their spouses qualify, were nearly as high. The percentages of Mexicans and Puerto Ricans who knew each of the program aspects were almost identical. The percentages of English-speaking Hispanics who knew the program aspects equaled or exceeded those of Spanish speakers. Spanish speakers were least knowledgeable about whether individuals who never worked could get benefits and whether benefits are funded by a payroll tax, with correct-response rates of 60 percent and 67 percent, respectively.

The percentages of correct responses were much lower for the benefit aspects, in many cases 70 percent or less. Puerto Ricans were generally more knowledgeable of benefit aspects than Mexicans. The statement about whether benefits are adjusted for inflation had the lowest correct-response rate among all program and benefit aspects (58 percent for Mexicans and 68 percent for Puerto Ricans). Low percentages of both English and Spanish speakers knew that benefits are adjusted for inflation (57 percent and 58 percent, respectively). Less than half of Spanish speakers (42 percent) knew that benefits do not have to be claimed at retirement from work.

Table 4 isolates the differences between pairs of ethnicity, ancestry, and language groups in the percentages shown in Table 3. Compared with Hispanics, significantly greater shares of non-Hispanic whites knew the various aspects of Social Security, particularly items relating to benefits; but nearly all of the differences across Hispanic ancestry and language groups were not significant. Notable exceptions were that English speakers were significantly more likely to know that individuals who never worked can get benefits as a spouse or survivor and that benefits do not need to be claimed at retirement.

We conducted sensitivity tests assuming that our samples of Puerto Ricans (37) and Cubans (13) were the same size as our sample of Mexicans (260) and that the sample of Spanish-speaking Hispanics (70) was the same size as our sample of English-speaking Hispanics (333). Assuming equal sample sizes, the differences between the Hispanic ancestry and language groups in knowledge of benefit aspects would all be significant. Assuming equal sample sizes also resulted in significant differences in program knowledge between the paired groups except for the pairing of Mexicans and Puerto Ricans. As the focus-group analysis found in Rabinovich, Peterson, and Smith (2017), no ancestry group was consistently higher or lower than the others in knowledge of the various program and benefit aspects. English-speaking Hispanics, however, were more knowledgeable about both program- and benefit-level aspects than were Spanish-speaking Hispanics.

All ethnicity and Hispanic ancestry and primary-language groups were more knowledgeable about program aspects than about benefit aspects (Table 5). The average percentages of correct responses were higher for the statements about the programs than for the statements about benefits, and the differences between those percentages were statistically significant for four of the seven groups we studied: non-Hispanic whites, Hispanics, Mexicans, and English-speaking Hispanics. The groups for which we found no statistically significant difference—Puerto Ricans, Cubans, and Spanish-speaking Hispanics—were the groups with the smallest number of respondents. If we assumed equal sample sizes for Spanish-speaking and English-speaking Hispanics, then the difference between program knowledge and benefits knowledge would also be significant for Spanish speakers. Assuming the same sample sizes for Cubans and Puerto Ricans as that of Mexicans did not affect the significance level.

Perceived Helpfulness of SSA Information Materials

Table 6 shows participants' views on the helpfulness of SSA information and the means by which they would prefer to receive it. There were no significant differences between Hispanics and non-Hispanic whites in perceived usefulness of written material on planning for retirement, worksheets for estimating their retirement income needs, an online calculator for estimating tax on Social Security benefits, and information on Social Security trust fund solvency. Hispanics were more interested than non-Hispanic whites in having SSA offer webinars or online videos about retirement planning and applying for Social Security benefits and in having SSA provide public-service announcements for television or radio on retirement planning issues. Non-Hispanic whites were more interested than Hispanics in having SSA provide information on how much will be deducted from their Social Security retirement benefits to pay their Medicare premiums. In all three cases, the differences are statistically significant. With few exceptions, 80 percent or more of the participants found these suggested ways of providing SSA information helpful. There were no significant differences between English-speaking and Spanish-speaking Hispanics in the perceived helpfulness and preferred means of receiving SSA information. Note that these questions were not asked of the full study sample; only respondents who were already retired or had exhibited knowledge of the program's purpose in a separate survey question (not covered in our analysis) were queried.

Conclusion

Using the UAS Internet panel, we extend and test the findings of previous research on what Hispanics know about Social Security and what kinds of information they most need to ensure their retirement security. Our research is one of the few attempts to look at differences in Social Security knowledge between Hispanics and non-Hispanic whites and among Hispanics across ancestry and primary-language groups; and ours is the first, as far as we are aware, to assess whether these differences are statistically significant. Such information is important for policymakers and program providers, as well as financial counselors, planners, and educators because of its implications for outreach to Hispanics.

We corroborate earlier research showing that Hispanics are less knowledgeable about Social Security than non-Hispanic whites. This suggests that useful outreach efforts, in addition to providing informational publications in Spanish as well as English, might include partnering with community-based language programs—both English-as-a-second-language and Spanish-language—to better inform Spanish-speaking Hispanics about Social Security benefits and programs. Other outreach efforts could identify minority-focused retirement planners and counselors and provide them with Social Security educational materials in Spanish.

Although the differences we observe across Hispanic ancestry and language groups are not statistically significant, Hispanics have significantly greater knowledge of Social Security program aspects than of Social Security benefit aspects. An individual who does not know certain facts about benefits—for example, that benefits are inflation-indexed and increase with delayed claiming—could make suboptimal benefit-claiming decisions, with consequences for retirement security. Earlier research noted a correlation between the type of information SSA provides in its publications and respondents' levels of benefit- and program-specific knowledge (Smith and Couch 2014). That study found that SSA provides program-specific information in most of its publications but is less likely to provide benefit-specific information, especially regarding inflation indexing. Our findings suggest that SSA outreach efforts should provide more detail on benefits and should not overemphasize program knowledge.

Our findings also suggest that financial educators can play an important role in improving Hispanics' knowledge of Social Security by providing details on Social Security benefits to complement agency-provided program information. Research on financial literacy suggests that providing Hispanics with information on Social Security benefits will affect their benefit-claiming decisions in particular and their retirement planning more generally.

We find that Hispanics' self-assessed knowledge of retirement-related financial issues is significantly lower than that of non-Hispanic whites. Hispanics feel less knowledgeable about how inflation affects retirement, how much to save for retirement, their longevity in retirement, how to invest their retirement money, and how to manage their spending in retirement. Financial counselors, planners, and educators are well suited to address these concerns. Spanish-speaking Hispanics feel even less knowledgeable than English-speaking Hispanics about retirement-related financial issues, which points to the importance of providing information in Spanish.

Our research on the perceived usefulness of Social Security information and how it is provided finds that respondents rated material and tools found on the SSA website or mailed by the agency as more useful than webinars and online videos or public-service announcements. This was true for both Hispanics and non-Hispanic whites and for English- and Spanish-speaking Hispanics. Hispanics rated information provided by public-service announcements and by webinars or online videos significantly more useful than did non-Hispanic whites. Non-Hispanic whites rated the information provided on the Medicare premium significantly more useful than did Hispanics. Otherwise, there were no significant differences across Hispanics and non-Hispanic whites or across language groups. Taking these perceptions into consideration in the development of materials and outreach strategies might be useful in identifying when targeted or more general information will be most helpful and how best to provide this information.

Limitations and Future Directions

This study is limited by the small sample sizes for Cubans, Puerto Ricans, and Spanish-speaking Hispanics. Our sensitivity analyses suggested that larger sample sizes would increase the likelihood of finding significant differences in levels of program- and benefit-specific knowledge across the ancestry and language groups. Another limitation is that we did not use population weights. UAS panel members are randomly selected; however, once selected, members choose whether to participate in the individual surveys. Larger samples, and weighting to account for differential response rates across population groups, would benefit future work.

Although we did not find statistically significant differences in Social Security knowledge across Hispanic ancestry and language groups, factors such as socioeconomic status, education level, and immigrant/citizenship status might play a role in explaining these differences between non-Hispanic whites and Hispanics. These differences merit further exploration. For example, are low retirement savings and limited knowledge of Social Security characteristic of Hispanics at all socioeconomic levels? Given the predicted growth of the U.S. Hispanic population, further research should explore how best to make relevant information easily accessible to improve their retirement outcomes.