Differences in Retirement Knowledge and Financial Preparedness between Hispanic and Non-Hispanic White Older Adults in the United States

Abstract

Background and Objectives:

We assess differences in retirement knowledge between older Hispanic and non-Hispanic White adults, and the extent to which individual, household, and neighborhood characteristics account for these differences. We also evaluate whether the relationships of retirement knowledge with financial outcomes (i.e., wealth) and health differ between Hispanic and non-Hispanic White older adults.

Research Design and Methods:

We administered the Retirement Knowledge Scale (RKS) in the 2020 Health and Retirement Study (N=1,350). We use a regression approach with a Blinder-Oaxaca decomposition analysis.

Results:

The average RKS was significantly lower among Hispanic than among White older adults. The top three factors explaining differences in retirement knowledge between Hispanic and White older adults were educational attainment, financial literacy, and neighborhood socioeconomic characteristics. RKS was significantly (p <.05) and positively associated with financial account balances and the likelihood of reporting very good or excellent health. We find that the association of RKS with wealth is of smaller magnitude for Hispanic than for White older adults. However, the relationship between retirement knowledge and health did not differ across racial/ethnic groups. Finally, differences between Hispanic and White older adults on the RKS were larger for men than for women.

Discussion and Implications:

The results suggest that reducing gaps in financial literacy and education, and acknowledging the role of neighborhood characteristics, can be helpful channels for reducing differences in retirement preparedness among Hispanic and White older adults. It is also significant that retirement knowledge is more strongly linked to wealth among White older adults.

I. Background and Objectives

Understanding what individuals know about financial planning for retirement is essential for promoting retirement preparedness and wealth accumulation. Hispanic adults lag behind non-Hispanic White adults in retirement knowledge, planning, and savings (Lusardi & Mitchell, 2011; Tamborini & Kim, 2020; Viceisza et al., 2023).¹ A systematic review of nationally representative surveys in the United States indicated that White adults scored 14 percentage points higher on objective financial knowledge than Hispanic adults (Blanco et al. 2024). Among non-retired adults, 68% of White adults and 35% of Hispanic adults have a tax-preferred retirement account, and 23% versus 16% have defined benefit pensions (Federal Reserve, 2024). In addition, White adults are more likely than Hispanic adults (40% vs 21%) to feel their retirement savings are on track (Federal Reserve, 2023). Indeed, the median retirement wealth of Hispanic households in 2022 was estimated to be 47% of White households (Weller et al., 2024).

The lack of access to an employer-sponsored retirement account, the complexity of information about retirement savings plans, and language barriers prevent many people, especially Hispanic immigrants, from financially preparing for retirement in the United States. Several studies have found that a lack of knowledge about retirement leads to mistakes in retirement decision-making (Clark et al., 2012; Angrisani & Casanova, 2021; Angrisani et al., 2023). Additionally, retirement knowledge is positively related to wealth accumulation among older adults (Lusardi et al., 2017; Blanco & Hays, 2024). Thus, given the key role financial knowledge plays in explaining wealth inequality, it is crucial to better understand how different racial/ethnic groups acquire and interact with information related to financial planning for retirement.

Wealth accumulation and financial preparation tend to be lower among Hispanic households than for White households due to barriers related to immigration status and socioeconomic status (SES). Undocumented immigrants contribute to Social Security through payroll taxes but are not eligible for retirement benefits (SSA, 2025). The 2022 median 12-month household income of Hispanic households represented 81% of that of White households ($65,882 versus $80,404; OMH, 2025). Interestingly, when it comes to expectations about age-related functional decline, Hispanic adults expect less decline in comparison to other racial groups (Menkin et al., 2017). Expectations about age-related functional decline is likely to influence how Hispanic adults save and plan for retirement. In addition, Hispanic older adults face significant barriers related to retirement preparedness due to less access to financial information from family and peer networks (Blanco et al. 2017). Thus, financial education targeted at retirement preparedness could help overcome barriers faced by Hispanic adults in the United States.

We examine factors related to differences in retirement knowledge and financial preparedness between Hispanic and White older adults in the United States, using the Retirement Knowledge Scale (RKS). The RKS was designed for low- and moderate-income (LMI) individuals and was included in the Health and Retirement Study (HRS) experimental module for the 2020 wave (Blanco & Hays, 2024). This experimental module also collected data on the “Big Three” financial literacy questions, which gauge knowledge about inflation, interest rates, and risk (Lusardi & Mitchell, 2011).

Our study has three main objectives. First, we describe the differences in retirement knowledge and financial preparedness between older Hispanic and White adults. Second, we examine how individual, household, and neighborhood characteristics contribute to differences in retirement knowledge and preparedness among older Hispanic and White adults. Third, we investigate the relationship between retirement knowledge and retirement savings, and whether this relationship differs between older Hispanic and White adults. We also examine the relationship between retirement knowledge and overall health, as well as whether this relationship differs across our target populations. Our analysis also contributes to the existing literature by examining the intersection of race and ethnicity with gender.

Financial Knowledge in the Context of Retirement Preparedness

Measurements

We assess retirement knowledge in terms of the financial knowledge individuals acquire and apply when making decisions related to financial preparedness for retirement (Blanco & Hays, 2024). Gallego-Losada et al. (2022) conducted a comprehensive bibliometric analysis of the literature on retirement planning and financial literacy, spanning the search period from 1990 to 2019. They found an increase in publications in this area, especially since 2011. They also note that almost half of the most cited authors and publications come from the United States. There is a unique centrality of financial investing in the United States, particularly regarding retirement. The proportion of workers under a Defined Benefits (DB) plan decreased from 84% to 25% between 1980 and 2015 (Li et al., 2019). As a result, there has been an increasing interest in improving retirement planning and financial knowledge due to the increase in defined-contribution (DC) plans in the United States.

Researchers have commonly used the “Big Three” financial literacy questions to assess levels of understanding of different financial topics relevant to everyday life developed by Lusardi & Mitchell (2023). The Lusardi & Mitchell (2007) measure of financial knowledge was first fielded through an experimental model in the 2004 wave of the HRS. They found that individuals with a greater propensity to plan for retirement have higher wealth and financial literacy than those who do not plan. Analyses examining the relationship between objective measures of financial knowledge and financial outcomes related to retirement preparedness often utilize a financial literacy indicator composed of these questions (Blanco et al., 2024).

Other objective measures assess financial knowledge related to retirement planning. For example, Knoll & Houts (2012) developed a financial knowledge scale incorporating specific questions about annuities, 401(k) plans, and Investment Retirement Accounts (IRAs). Lusardi et al. (2014) analyzed financial sophistication, going beyond the big three questions and collecting information on financial knowledge specific to risk diversification, asset valuation, portfolio choice, and investment fees among older adults in the HRS. Lusardi et al. (2014) show that most older adults lack the financial sophistication necessary to make financial decisions related to retirement security.

Hopkins & Pearce (2019) developed a comprehensive retirement income literacy quiz to assess financial knowledge in the context of retirement. They found that only 26% of a sample of retirement-age adults (aged 60 to 75 years) in the United States passed the quiz. Those who passed the quiz were more likely to have made better decisions related to retirement preparedness, such as having a retirement plan or estate plan, feeling more confident about managing their investments in retirement, and having a long-term care plan.

Other recently developed retirement knowledge measures include those by Seidl et al. (2021) and Ghafoori et al. (2023). The Retirement Planning Scale (Seidl et al., 2020) was developed in New Zealand and later translated into Spanish, where it was fielded among 1,827 Spanish employees. This 24-item measure gathers information about retirement planning in 4 dimensions: representations, goals, decisions, and behaviors.

Based on a literature review and consultation with a panel of experts, Ghafoori et al. (2023) identified four key dimensions of retirement confidence: Financial Awareness and Skills, Health and Well-being, Social Factors and Retirement, and Awareness and Planning. They developed the Retirement Confidence Index and evaluated its reliability in a sample of 321 participants in Australia, yielding a composite score of how confident people are moving towards and throughout the retirement years.

Differences between Hispanic and White adults

There is extensive evidence of racial/ethnic differences in retirement preparedness, where Hispanic adults lag behind White adults. Data show that Hispanic older adults tend to struggle more in retirement than White older adults due to lower levels of wealth and a higher propensity to live in poverty (Butrica et al., 2009).

Two factors that explain the low levels of wealth in retirement among the Hispanic population in the United States are that the proportion of Hispanic adults with access to employer-sponsored retirement plans and who own a home is significantly lower than that of White adults (Francis et al., 2021). Data from 2022 show that while 83% of White households worked for an employer that offered a retirement plan, only 57% of Hispanic households had access to a retirement plan through their employer (Weller et al. 2024).

There is evidence that Hispanics made progress in retirement income between 1989 and 2007, but there was a setback in 2007–2009 related to the Great Recession (Wolff, 2023). Wolff (2022) provides evidence that the net worth of Hispanic families decreased in absolute terms between 2007 and 2016. Hispanic adults have less retirement wealth than White adults, and thus rely more on Social Security income in retirement. Social security benefits for Hispanics and Whites represent 30% and 11%, respectively, of expected retirement income (Wolff, 2023). Previous work has shown that Hispanic adults have less knowledge than White adults about Social Security benefits (Peterson et al., 2019). It is estimated that 30–40% of retirement wealth inequality is attributed to financial knowledge (Lusardi et al., 2017).

Studies led by Lusardi & Mitchell (2007, 2011a, 2014), using HRS data, have shown that Hispanic adults lag behind White adults in financial knowledge, and these differences in financial knowledge have implications for retirement planning and preparedness. A comparison of wealth between Baby Boomers in 2004 and individuals in the same age group in 1992 shows that Baby Boomers rely more on housing equity (Lusardi & Mitchell, 2007). Interestingly, this study reveals that individuals who plan for retirement have higher levels of wealth and financial literacy compared to those who do not plan. Lusardi & Mitchell (2011) found that minorities are more financially illiterate than non-minorities; however, those with higher levels of financial literacy are more likely to plan for retirement and achieve successful retirement outcomes. Individuals acquire financial knowledge related to retirement preparedness by using retirement calculators, attending retirement seminars, and talking with financial experts Lusardi & Mitchell (2011). Furthermore, Lusardi & Mitchell (2014) collected data specific to more complex questions related to retirement finances. They found that non-White older adults are less financially savvy than their White counterparts.

Other studies looked at racial/ethnic differences in the context of financial knowledge and retirement preparedness using HRS data from a different perspective. This work provides evidence that workers are often unfamiliar with retirement benefits, which can lead to less optimal decisions when preparing for retirement (Ekerdt & Hackney, 2002). Li et al. (2019) conducted an interesting study on pension plan types and financial literacy in later life, revealing that Hispanic older adults had lower financial literacy compared to White older adults. Their analysis revealed that individuals with access to DC pension plans exhibit higher financial knowledge and skills compared to those without a DC pension plan, and that Hispanic adults are less likely to have access to DC pension plans than White adults. Furthermore, a recent study measuring how older adults respond to objective information on longevity found that Hispanic participants were more likely to regret not saving enough for retirement and not purchasing long-term care than White participants (Hurwitz & Mitchell, 2025).

Other studies focused on racial/ethnic differences in financial knowledge and found relevant implications of these differences for retirement preparedness. Lusardi & Mitchell’s (2011b) analysis of the National Financial Capability Survey data showed that Hispanic adults score lower than White adults in financial literacy concepts. Individuals who score higher on financial knowledge are more likely to plan for retirement and have higher levels of wealth. Using data from the NLSY79, White et al. (2022) provide evidence that Black and Hispanic adults are less likely to engage in retirement planning and have lower retirement assets compared to White adults. In their analysis, factors that explain these racial/ethnic differences in retirement preparedness are having an employer-sponsored retirement plan and higher education.

Kim et al. (2021) used data from the 2016 Survey of Consumer Finances to evaluate the factors that explain differences in retirement saving motives among White, Black, and Hispanic adults. Decomposition analysis revealed that individual characteristics, including homeownership, objective financial knowledge, planning horizons, and age, explain some of the racial/ethnic differences in retirement savings motives.

Our Contribution

We contribute to the literature by comprehensively analyzing the factors related to differences in retirement knowledge between Hispanic and White older adults in the United States. First, we use data from the RKS and the Big Three financial literacy questions to provide evidence of the differences between Hispanic and White adults. The RKS was developed to measure retirement preparedness (Blanco & Hays, 2024). This measure was piloted in the HRS 2020 wave to provide a comparable measure of retirement knowledge among retired and non-retired adults, as well as among English- and Spanish-speaking individuals. This is particularly noteworthy when considering the differences in retirement preparedness between Hispanic and White older adults in the United States.

Using data from the RKS allows us to study racial/ethnic differences in retirement knowledge and preparedness from three dimensions, such as whether people have reflected on what age they want to retire and how much money they will need in retirement (reflection dimension), how have they engaged with different sources of information related to financial planning for retirement such as print, web/online, family/friends, financial professional (engagement dimension), and their self-evaluation of knowledge about different types of retirement accounts, how retirement accounts work, and how well prepared they feel they are for retirement (self-evaluation dimension). Blanco & Hays (2024) found that Hispanic adults had lower RKS scores (i.e., less knowledge) than White adults.

Our analysis also considers the intersectionality of race and ethnicity, as well as gender. Additionally, we examine the characteristics of individuals, households, and neighborhoods that likely explain the differences in retirement knowledge between Hispanic and White older adults. We leverage the richness of the HRS dataset to analyze relevant attributes at both the individual and household levels as potential factors explaining differences across groups. To study how neighborhood socioeconomic and demographic characteristics explain differences in retirement knowledge among our racial/ethnic groups of interest, we work with the restricted version of the HRS, which provides access to the geocodes of the census tracts where survey participants reside. We hypothesize similarities with previous work that explains racial/ethnic differences in financial knowledge, where higher levels of Socioeconomic Status (SES) at the individual, household, and neighborhood levels are associated with higher levels of financial knowledge (Angrisani et al., 2021).²

We expand on previous work evaluating how retirement knowledge, as measured by the RKS, relates to retirement saving and wealth accumulation among older adults, and whether the relationship between retirement knowledge and wealth differs among Hispanic and White older adults. We also evaluate whether neighborhood characteristics influence this relationship similarly or differently for racial/ethnic subgroups. Finally, we aim to investigate the relationship between retirement knowledge and overall health, as well as whether this relationship varies across different racial/ethnic groups.

We summarize here our theoretical framework for differences in retirement knowledge between Hispanic and White older adults. We have the following hypotheses:

• H1: There are significant differences in retirement knowledge and financial literacy between groups.

• H2: Financial literacy, individual and household, and neighborhood SES explain differences in retirement knowledge between groups.

• H3: The retirement knowledge–wealth relationship is the same across groups.

• H4: The retirement knowledge–health relationship is the same across groups.

II. Research Design and Methods

Data Source and Study Sample

We used data from the 2020 HRS wave.³ Since 1992, the HRS has used a complex sample design to obtain a representative sample of U.S. adults 50 and older. The HRS collects data related to health, labor force participation, economic status, family structure, and retirement. The RKS data come from an HRS experimental module, where a 10% random subsample is invited to participate in a 2–3-minute survey after completing the core survey. The RKS experimental module also included the Big Three questions on financial literacy, developed by Lusardi & Mitchell (2011a), which have been used to construct the Financial Literacy Scale (FLS). A total of 1350 individuals answered the RKS experimental module. For the analysis, we restrict our sample to Hispanic and White adults 50 and older.⁴

We use the publicly available version of the HRS dataset to construct most of the independent and dependent variables. To incorporate neighborhood socioeconomic characteristics into the analysis, we used the restricted version of the HRS, which provides access to the geocodes of the census tracts where the survey participants reside. We constructed the Neighborhood Socioeconomic Score (NSES) using data from the 2016 American Community Survey (ACS, US Census Bureau, 2016), which was the latest year available in the HRS-restricted data enclave at the time of our analysis.

Dependent Variables

We used dependent variables to assess retirement knowledge, financial, and health outcomes. As mentioned, the RKS was designed to measure retirement knowledge among retired and non-retired adults with low to moderate income levels, who speak English or Spanish. The RKS is composed of three subscales: Reflection, Engagement, and Evaluation. Please refer to Blanco & Hays (2024) for a full analysis of the reliability and validity of the RKS as a measure of retirement knowledge among older adults in the United States. We constructed this variable in accordance with Blanco & Hays' (2024) analysis. Please refer to Table A1 (online supplemental material) for how the RKS was constructed.

To examine the effects of one’s financial preparedness for retirement on financial outcomes, we worked with three dependent variables: 1) individual retirement account (IRA) or Keogh account balance, 2) stock account balance, and 3) probability of leaving a bequest of $10,000 or more (binary indicator variable).

We also analyze the relationship between retirement knowledge and self-reported health. We used two dependent variables related to health 1) a dummy variable equal to one if excellent or very good health (equal to zero if good, fair, or poor health), and 2) the Patient-Reported Outcomes Measurement and Information System (PROMIS^®) global physical health score estimated as in Hays et al. (2015).

Independent Variables

In all regressions, we control for individual sociodemographic and economic characteristics such as age, gender, educational attainment, marital status, household income (in $1000s), and whether the respondent works for pay, was born in the United States, is retired, receives social security, and is the financial respondent in the household for the HRS survey.

Moreover, we control for financial literacy using the FLS from the Big Three financial literacy questions. We also control for the financial planning horizon for family savings and spending, as well as other financial outcomes, such as IRA or Keogh account balances, stocks/mutual funds accounts, checking account balances, and Certificate of Deposit (CD) and Bonds account balances.

We created a NSES for the census tracts with data from the ACS, similar to the methodology of Diez-Roux et al. (2001) using the following variables: the median household income, the median value of housing units, dividend or net rental income, education (percentage of adults 25 and older who completed high school and percentage of adults who completed college), and employment (percentage of those 16 and older employed in executive, managerial, or professional specialty occupations).⁵

Please refer to Table A1 (online supplemental material) for a comprehensive description of HRS and ACS variables and the construction of the variables used in our analysis. Table 1 shows the summary statistics of the variables used in our analysis. We also provide a summary statistics table by racial/ethnic and gender groups in Table A2 (online supplemental material).

Table 1.

Summary Statistics for Demographic and Socio-Economic Variables

Variables	N	Mean	SD	Min	Max
Retirement Knowledge Scale	964	38.86	24.56	0	100
Financial Literacy Scale	950	68.81	30.75	0	100
Race/Ethnicity
White	964	79%	40.72	0	100
Hispanic	964	21%	40.72	0	100
Women	964	57%	49.48	0	100
coupled	964	62%	48.56	0	100
US born	964	85%	35.66	0	100
Age	964	69.07	10.27	50	99
Education
GED or less than high sch.	964	20%	39.64	0	100
High school grad	964	28%	44.83	0	100
Some college	964	26%	43.74	0	100
College and above	964	27%	44.40	0	100
Working for pay	954	34%	47.5	0	100
Retired	954	54%	49.9	0	100
Household income (thousands)	964	77.10	97.31	0	1134.66
Receives social security	964	75%	43.38	0	100
Financial respondent	964	70%	45.75	0	100
IRA/Keogh account (thousands)	964	131.27	324.92	0	2422.75
Stocks (thousands)	964	89.27	411.20	0	4708.14
Checking acct. (thousands)	964	35.88	83.94	0	900
CD/Bonds (thousands)	964	8.21	66.31	0	1700
Prob. of leaving a bequest of $10,000	940	69.16%	38.29	0	100
Financial planning horizon
next few months	958	14%	34.81	0	100
next year	958	12%	32.15	0	100
next few years	958	28%	45.11	0	100
next 5–10 years	958	32%	46.77	0	100
longer than 10 year	958	14%	34.26	0	100
Excellent/very good health	963	39%	48.86	0	100
PROMIS health Score	964	47.33	8.40	8	62.45
Neighborhood characteristics
NSES	954	0.445	2.88
Household Income (Thousands)	960	651.39	292.18
Med. house value (Thousands)	954	242.56	178.15
% with high school degree	960	57	13
% with college degree	960	30	17
% employed in Mgmt/Prof occ.	960	36	14

Notes: Data HRS 2020 and 2016 ACS. The sample includes older Hispanic and White adults 50 years old and older. See Table A1 in Appendix for a full description of variables construction. See Table A2 (online supplemental material) for summary statistics disaggregated by racial/ethnic and gender groups.

GED = General Education Degree; IRA = Individual Retirement Account; CD = Certificate of Deposit; Prob. = Probability; NSES: Neighborhood Socioeconomic Score; Mgmt=management; Prof = professional; occ.=occupation; PROMIS=Patient-Reported Outcomes Measurement Information System).

Methods

We provide a descriptive analysis of the differences in retirement knowledge and preparedness among older Hispanic and White adults. We compare the means of different subscale items and the RKS and FLS total scales among older Hispanic and White adults, disaggregating these groups by gender. We conduct between-group t-tests to determine whether the mean differences are statistically significant. We also evaluate the magnitude of these differences by calculating the effect size, which is obtained by dividing the difference by the standard deviation calculated for the entire sample. We use the rule of thumb that 0.20–0.49 is considered small, 0.50–0.79 is considered medium, and 0.80 and above is considered large (Cohen, 1998).

We then evaluate which individual, household, and neighborhood factors contribute the most to the differences in retirement knowledge and financial preparedness among Hispanic and White older adults. To evaluate the factors that contribute most to explaining the differences in retirement knowledge and financial preparedness between Hispanic and White older adults, we employed a regression approach with a Blinder-Oaxaca Decomposition (BOD) analysis.

In our BOD analysis, we include independent variables associated with retirement knowledge. Previous work has shown that women, minorities, immigrants, and individuals who are younger, working, receiving social security benefits, and have lower levels of education and income, in comparison to their counterparts, are likely to have lower retirement knowledge and be less financially prepared for retirement (Lusardi & Mitchell 2011a, Lusardi & Mitchell 2011b). We expect a positive association between retirement knowledge and preparedness with wealth-related variables, such as having a bank account and an IRA, stocks, or CDs, and the balance of such accounts (Lusardi et al., 2017; Lusardi & Mitchell, 2007). We also expect that individuals who are financial respondents of the household and who have higher levels of financial literacy, and a longer financial planning horizon are likely to show higher levels of retirement knowledge and preparedness than their counterparts (Angrisani et al., 2023; Lusardi & Mitchell 2023). In relation to neighborhood characteristics, previous work has shown that lower NSES can be associated with a lower likelihood of owning a bank account among older adults and lower levels of financial literacy (Angrisani et al, 2021; Blanco et al., 2019; La Chance, 2014). Thus, we hypothesize that individuals residing in neighborhoods with higher NSES will likely exhibit higher retirement knowledge and preparedness than those living in lower NSES neighborhoods.

Using ordinary least squares (OLS) regression, we also examine how retirement knowledge is related to financial outcomes associated with wealth accumulation in older ages. We will evaluate whether there are differences across racial/ethnic groups in these relationships using interaction terms. We estimate the following OLS regressions for three outcome variables: 1) IRA/Keogh account balance, 2) Stock account balance, and 3) probability of leaving a bequest of $10,000 or more:

$$\mathit{\text{Financial}}\mathit{\text{outcome}}{s}_i={\beta }_0+{\beta }_1\mathit{\text{RK}}{S}_i+{\beta }_2\mathit{\text{RK}}{S}_i*\mathit{\text{Hispani}}{c}_i+{\beta }_3\mathit{\text{Hispani}}{c}_i+{\beta }_4{X}_i+{\beta }_5\mathit{\text{NSES}}+{\epsilon }_i$$ (1)

X_i is a vector of individual-level variables (age, gender, education attainment, marital status, being US born, working status, retirement status, social security recipient, household income, financial respondent, and financial planning horizon). The primary coefficient of interest is β₂, which indicates the association between retirement preparedness and financial outcomes and how it differs between Hispanic and White older adults. Additionally, we estimate Equation 1 by splitting the sample into two groups based on whether the individual lived in a census tract with an NSES above or below the median. This approach helps to evaluate whether the retirement knowledge-wealth relationship varies depending on neighborhood characteristics.⁶

To evaluate the relationship between retirement knowledge and health, and whether it differs for older Hispanic and White adults, we estimate Equation 1 for two self-reported health outcome variables: reporting excellent or very good health (versus good, fair, or poor) and the estimated PROMIS global physical health score.⁷

III. Results

Sample Characteristics

Table 1 describes the individual, household, and neighborhood variables we use in our analysis. Our sample contains 21% Hispanic and 79% White older adults. The average age is 69, and 27% of the sample have a college degree or beyond. Sixty-two percent of the respondents are in coupled households, 85% receive social security, 54% are retired, and 34% work for pay. Thirty-nine percent reported having very good or excellent health. The average RKS was 39, whereas the FLS was 69. The average NSES in our sample was 0.445.

There are differences in sample characteristics between Hispanic and White older adults (Table A2). White older adults are more likely to born in the US (96% versus 65%), older (average age 70 versus 65 years old), more educated (13% versus 43% have a GED or less than high school degree), have almost double the average income ($84,595 versus $48,811) and less likely to work for pay (33% versus 42%), more likely to be retired (58% versus 33%) and receive Social Security benefits (77% versus 65%). We also observe that older White adults have higher levels of RKS, FLS, and PROMIS global physical health scores, as well as a higher probability of having excellent or very good health. Summary statistics for older Hispanic and White adults disaggregated by gender are also provided in Table A2.

Racial/Ethnic Differences in Retirement and Financial Knowledge

In Table 2, we further disentangle the differences between the RKS and the FLS among Hispanic and White older adults (Panel A) and by gender within each group separately (Panel B). We observe in Table 2, Panel A, that the average for both RKS and FLS and its subscales was significantly higher among White than Hispanic older adults. Except for one item of the RKS (engagement with retirement information online), all differences between Hispanic and White older adults were statistically significant at the p < 0.05 level. Within RKS, the largest difference between Hispanic and White older adults was in the evaluation subscale (difference of 22.19, or 0.79 SD), especially in financial preparedness (difference of 26.31, or 0.76 SD). Following Cohen’s rule of thumb, these differences are of medium magnitude and close to the 0.80 threshold, representing a large effect size.

Table 2.

Mean Differences on Retirement Knowledge Scale (RKS) Between Hispanic and White older adults, Hispanic men and White men, Hispanic women and White women

PANEL A: OVERALL SAMPLE								PANEL B: BY GENDER GROUPS
	Overall			White	Hispanic	Diff.	Diff. Mag.	Men				Women
	N	Mean	SD	Mean	Mean	W-H		White Mean	Hispanic Mean	Diff. W-H	Diff. Mag.	White Mean	Hispanic Mean	Diff. W-H	Diff. Mag.
Retirement Knowledge Scale Items
Reflection: what age retire	937	60.97	41.53	64.50	47.81	16.69***	0.40	69.09	46.82	22.27***	0.54	61.06	48.62	12.43***	0.30
Reflection: how much money	951	60.25	42.85	64.63	43.72	20.91***	0.49	70.96	42.15	28.82***	0.67	59.98	44.94	15.04***	0.35
Engagement: Print source	961	37.60	40.19	39.92	28.88	11.04***	0.27	43.75	23.22	20.53***	0.51	37.13	33.33	3.80	0.09
Engagement: Web	960	26.04	38.12	27.27	21.39	5.88*	0.15	31.45	22.47	8.98*	0.24	24.24	20.54	3.71	0.10
Engagement: Family/friend	958	34.97	39.48	37.03	27.17	9.86**	0.25	38.26	31.44	6.82	0.17	36.14	23.81	12.33***	0.31
Engagement: Finance prof.	960	27.05	37.76	31.13	11.72	19.42***	0.51	33.75	9.74	24.01***	0.64	29.23	13.27	15.96***	0.42
Engagement: Class	960	12.50	24.33	13.76	7.76	6.01**	0.25	15.88	4.49	11.39***	0.47	12.22	10.32	1.90	0.08
Evaluation: How IRAs work	956	40.27	33.10	44.74	23.60	21.14***	0.64	52.98	25.47	27.51***	0.83	38.70	22.12	16.57***	0.50
Evaluation: Types of IRAs	958	37.89	31.29	41.84	23.10	18.74***	0.60	49.69	23.60	26.09***	0.83	36.09	22.71	13.37***	0.43
Evaluation: Finc. preparedness	954	52.48	34.74	57.97	31.66	26.31***	0.76	63.33	36.33	27.0***	0.78	54.02	27.88	26.14***	0.75
Retirement Knowledge Subscales and Total Scale
Reflections scale	959	60.41	37.41	64.40	45.36	19.05***	0.51	70.26	44.38	25.88***	0.69	60.12	46.13	13.99***	0.37
Engagement Scale	963	27.60	26.95	29.79	19.35	10.44***	0.39	32.57	18.20	14.37***	0.53	27.77	20.26	7.51******	0.28
Evaluation Scale	962	43.52	28.08	48.18	25.99	22.19***	0.79	55.35	28.46	26.88***	0.96	42.94	24.04	18.90***	0.67
Total Scale	964	38.86	24.56	42.13	26.51	15.62***	0.64	46.98	26.50	20.48***	0.83	38.58	26.52	12.07***	0.49
Financial Literacy Scale Items and Total Scale
Inflation item	935	74.22	43.76	78.13	59.80	18.33***	0.42	83.54	64.04	19.50	0.45	74.05	56.36	17.68***	0.40
Interest rate item	937	66.60	47.19	70.23	53.03	17.20***	0.36	76.97	56.32	20.65***	0.47	65.17	50.45	14.72***	0.34
Risk item	931	65.52	47.56	70.08	48.74	21.34***	0.45	76.51	52.27	24.24***	0.51	65.23	45.95	19.28***	0.41
Total scale	950	68.81	30.75	72.81	53.90	18.91**	0.62	79.10	57.68	21.42***	0.45	68.14	50.89	17.25***	0.36

Notes: Difference (denoted as Diff. W-H) calculated subtracting mean for older Hispanic adults from older White adults. Difference magnitude (denoted as Diff. Mag.) calculated by dividing difference by standard deviation.

^***p<0.01,

^**p<0.05,

^*p<0.1.

Sample includes older Hispanic and White adults 50 years old and older. There are 739 Whites (316 men, 424 women) and 196 Hispanics (88 men,108 women). IRA = Individual Retirement Account.

Differences between Hispanic and White older adults for the RKS and FLS total scales were of medium magnitude (0.64 SD and 0.62 SD, respectively). For the reflection subscale of the RKS, we also observe a medium magnitude difference (0.51 SD). Interestingly, the difference we see for the engagement subscale of the RKS is small (0.39 SD). Nonetheless, for the item of engagement subscale related to engagement with a financial professional, the difference between Hispanic and White older adults is of medium magnitude (0.51 SD).

Racial/Ethnic and Gender intersectionality in Retirement and Financial Knowledge

When we split the sample to consider the intersectionality of race/ethnicity and gender to compare the average RKS and FLS total scales, their subscales, and items, we continue to see higher scores and significant differences in most cases for White older adults compared to Hispanic older adults, regardless of the gender (Table 2, Panel B).

Some interesting patterns emerge here. The differences between Hispanic and White older adults for the RKS total scale are large for men but small for women (0.83 SD versus 0.49 SD). For the FLS total scale, the difference between Hispanic and White older adults is of small magnitude, regardless of the gender group considered (0.45 SD and 0.36 SD for men and women, respectively).

With the RKS subscales, differences between Hispanic and White older men are of medium magnitude for the RKS reflection and engagement subscales (0.69 and 0.53, respectively). There is a large difference between Hispanic and White older men on the RKS evaluation subscale (0.96 SD).

Differences in the RKS reflection and engagement subscales between Hispanic and White older women are of small magnitude (0.37 SD and 0.28 SD, respectively). The difference on the RKS evaluation scale between Hispanic and White older women is smaller than between men and is of medium magnitude (0.67 SD).

For the items of the RKS reflection subscale, we observe that differences between Hispanic and White older men are of medium magnitude (0.54 SD and 0.67 SD), but for women, they are of small magnitude (0.30 SD and 0.35 SD).

For the items of the RKS subscale specific to engagement with information about retirement, we see some differences across racial/ethnic and gender groups. The difference between Hispanic and White older men is of medium magnitude for the RKS item specific to engagement with retirement-related information from a printed source. In contrast, the difference between Hispanic and White older women is not statistically significant. For engagement with retirement-related information online, the difference between the two gender groups is not statistically significant at the 5% level. While older White women are more likely to talk to a family or friend about retirement preparedness than older Hispanic women, there is no statistically significant difference in this behavior between Hispanic and White older men.

The difference in the RKS item related to engagement with a financial professional is larger for men than women (0.64 SD versus 0.42 SD). Older White men are more likely to attend a class related to retirement preparedness than older Hispanic men, and this difference is of small magnitude (0.47 SD). We do not observe a significant difference between Hispanic and White older women attending a retirement preparedness class.

The differences in knowledge on how retirement saving account works and different types of retirement accounts is large between Hispanic and White older men (0.83 SD and 0.83 SD), but medium and small, respectively, between Hispanic and White older women (0.50 SD and 0.43 SD). Regarding self-evaluation of retirement preparedness, the differences between Hispanic and White older men and Hispanic and White older women are of medium magnitude (0.78 SD and 0.75 SD, respectively).

Factors Explaining Differences in RKS among Hispanic and White Older Adults

Table 3 presents results from the BOD to analyze what individual and neighborhood characteristics explain differences in financial knowledge for retirement among older Hispanic and White adults (Column 1). We also include results from the BOD to study factors explaining differences between older Hispanic and White men (Column 2) and between Hispanic and White older women (Column 3). Please refer to Figure A1 (online supplemental material), which shows the differences in the RKS associated with racial/ethnic and gender groups from the BO decomposition analysis shown in Table 3.

Table 3.

Blinder-Oaxaca decomposition of RKS total score for Hispanic and White older adults, Hispanic men and White men, Hispanic women and White women

	White (G1) vs Hispanic (G2) (1)	White men (G1) vs Hispanic men (G2) (2)	White women (G1) vs Hispanic women(G2) (3)
Gap	15.250***	20.025***	11.699***
% of gap explained by diff. in variables	71.601***	62.438***	75.467***
% of the quantity effect explained by…
Financial literacy scale	21.192***	24.698***	22.352***
NSES	17.718***	1.412	25.483***
Women	−0.321
Coupled	−1.357	−0.222	−1.008
US born	10.310	7.874	6.210
Age	−17.80***	−9.656	−27.248***
Education	28.348***	18.322**	39.870***
Employed	−1.354	5.379	−7.401
Retired	12.524**	0.577	25.217***
Household income	5.283*	10.333	2.483
Social security	−0.944	−1.772	2.093
Financial respondent	1.317	2.641	0.373
IRA	13.05***	19.22***	10.521**
Stock	3.350**	5.86***	1.792
Checking account	0.329	3.074	−2.889
CD	0.051	0.276	−1.375
Financial planning horizon	8.303**	11.984**	3.526
Observations (Total)	937	403	534
Observations (G1)	744	317	427
Observations (G2)	193	86	107

Notes:

^***p<0.01,

^**p<0.05,

^*p<0.1.

HRS 2020 data. Sample includes older Hispanic and White adults 50 years old and older.

NSES = Neighborhood Socioeconomic Score; IRA = Individual Retirement Account; CD = Certificate of Deposit; G1 = Group 1; G2 = Group 2.

The results indicate the percentage of the difference explained by our model and how much each independent variable contributes to explaining this difference between the two groups in percentages. Based on Blanco & Hays (2024) analysis, we hypothesized that the following factors would be positively associated with the RKS: financial literacy, being in a couple household, US-born, educational attainment, being employed, having a retirement status, household income, being a financial respondent, owning an IRA, stock, checking account, and/or a CD, and the financial planning horizon. The factors negatively associated with the RKS in Blanco & Hays (2024) were being a woman, age, and receiving social security. We expect NSES to be positively correlated with the RKS.

As shown in column 1 of Table 3, the model explains 72% of the White-Hispanic gap in the RKS total score, where education (28%), financial literacy scale (21%), NSES (18%), age (−18%), IRA account balance (13%), financial planning horizon (8%), and Stock account balance (3%) contributed to the gap.⁸

As expected, the top two factors explaining differences in retirement knowledge between older Hispanic and White adults are educational attainment and financial literacy (28% and 21%, respectively). Interestingly, the following two factors contributed to explaining the White-Hispanic gap in the RKS total score: NSES and age, where both explain the gap by 18% but have opposite signs. NSES contribution to the gap is denoted as 18%, and age contribution to the gap is −18%. The 18% associated with the NSES indicates that neighborhood characteristics appear to be relevant, as Hispanics are more likely to reside in areas with lower NSES. In our regression estimates, age has a negative coefficient, indicating that Hispanic older adults would have even lower RKS scores if they were the same age as White older adults. Hispanic older adults are younger than White older adults in our sample, and age is negatively associated with the RKS (see Table A2, mean of 65 versus 70 years old).

Factors Explaining Differences in RKS among Hispanic and White Older Men and Women

A similar pattern holds when we compare Hispanic men with White men and Hispanic women with White women in columns 2 and 3 of Table 3. Our model explains about 62% of the unconditional gap in financial preparedness between Hispanic and White men and 75% of the unconditional gap in financial preparedness between Hispanic and White women. For Hispanic and White men, the contributing factors to the gap (listed in order of importance based on the percentage contribution) were financial literacy scale (25%), IRA account balance (19%), education (18%), financial planning horizon (12%), and stock account balance (6%). For Hispanic and White women, education (40%), age (-27%), NSES (25%), being retired (25%), financial literacy scale (22%), and IRA account balance (11%) contributed to the racial/ethnic gap (also listed in order of importance). In this estimation, age is denoted as contributing to explaining the gap with a negative sign (-27%), where Hispanic women would have lower RKS scores if they were the same age as White women.

Characteristics that explain differences between older Hispanic and White adults by gender are interesting. Financial literacy and educational attainment are the top factors explaining differences across gender groups. NSES and being retired are significant factors explaining differences between Hispanic and White older women, but not relevant factors explaining differences between Hispanic and White older men. Education accounts for a larger share of the difference between Hispanic and White older women than between Hispanic and White older men (40% versus 18%). The IRA balance explains a greater portion of the difference between men and women (19% versus 11%). Interestingly, while financial outcomes-related variables such as stock account balance and financial planning horizon explain differences between Hispanic and White older men, these factors do not explain differences between older Hispanic and White women. Age is not statistically significant in explaining the difference between Hispanic and White older men, but it is negatively associated with differences between Hispanic and White older women.

The Relationship between RKS and Financial Outcomes

Table 4 shows the association between RKS and different financial outcomes, including balances in IRA accounts and stock accounts and the probability of leaving a bequest of $10,000 or more. In Table 4, we present the regressions for the overall sample and for those with NSES above and below the median, using an interaction term of the RKS variable with a dummy that accounts for being a Hispanic older adult.

Table 4.

Regression of relationship retirement knowledge and financial preparedness outcomes for Hispanic and White older adults

	Overall sample			Above median NSES sample			Below median NSES sample
	IRA/Keogh Acct. Bal.	Stock Acct. Bal.	Prob. of bequest	IRA/Keogh Acct. Bal.	Stock Acct. Bal.	Prob. of bequest	IRA/Keogh Acct. Bal.	Stock Acct. Bal.	Prob. of beques
RKS	2,997.24***	2,436.21***	0.31***	3,773.80***	3,683.41***	0.21***	1,840.00***	878.63	0.39***
	(484.80)	(804.03)	(0.05)	(735.07)	(1336.17)	(0.07)	(521.14)	(544.39)	(0.09)
Hispanic	76,838.45**	66,066.97*	−7.7478	−22,843.56	81,68.00	−34.42***	50,338.28*	7,585.52	0.08
	(31481.97)	(35578.78)	(6.15)	(78591.28)	(80943.82)	(9.71)	(26007.95)	(19899.52)	(7.32)
RKS*Hispanic	−2,799.09**	−2,283.44**	0.1023	−1,491.61	−3,552.17*	0.53***	−2,093.91***	−340.85	−0.06
	(1326.86)	(1006.21)	(0.13)	(3392.01)	(1966.09)	(0.17)	(598.82)	(631.28)	(0.18)
N	954	954	930	477	477	462	477	477	468
R-squared	0.21	0.11	0.26	0.20	0.11	0.25	0.15	0.07	0.24

Notes:

^***p<0.01,

^**p<0.05,

^*p<0.1.

HRS 2020 data. The sample includes older Hispanic and White adults 50 years old and older. Regressions control for gender, age, education attainment, citizenship status, marital status, retirement status, financial respondent, whether the respondent works for pay, household income (1000s), and social security recipient. We control for NSES when we estimate regressions with the overall sample. We estimate regressions separately for individuals who live in census tracts with NSES above median and those who live in census tracts with NSES below the median separately. The IRA/Keogh Acct. Bal. dependent variable refers to the net value of Individual Retirement Accounts (IRA) and/or Keogh accounts in the household (self, husband/wife/partner, continuous variable). The Stock Acct. Bal. dependent variable refers to the net value of stocks, mutual funds, and investment trusts in the household (self, husband/wife/partner, continuous variable). Prob. of bequest dependent variable refers to the probability of leaving a bequest of $10,000 or more (continuous variable with probabilities ranging from 0 to 100).

NSES = Neighborhood Socioeconomic Score; IRA = Individual Retirement Account; RKS = Retirement Knowledge Scale.

The regression estimates for the overall sample indicate that higher levels of the RKS are associated with higher balances in IRA and stock accounts, and a higher probability of leaving a bequest of $10,000 or more. Interestingly, being a Hispanic older adult is significantly positively associated with IRA account balances.

RKS is significantly associated with $2,997 more in the IRA account and $2,436 more in the stock account balance. However, the magnitude of the association is weaker for Hispanic older adults than for White older adults. The total effect of RKS on financial outcomes among Hispanic older adults is a $198 IRA account balance, $153 stock account balance, and 0.31 percent probability of leaving a bequest, and among White older adults, is a $2,999 IRA account balance, $2,436 stock account balance, and 0.31 percent probability of leaving a bequest. Thus, these estimates indicate that the association between RKS and financial outcomes is of smaller magnitude for Hispanic older adults compared to White older adults.

To provide some context on the magnitude of the effect of the RKS on financial outcomes for Hispanics, Appendix Table A3 presents the estimates for our financial outcome dependent variables associated with the overall sample mean RKS score, as well as half a standard deviation below and above the mean among Hispanics.

Table 4 also presents the results when we estimate the model separately for those above and below the sample median NSES. Here, we observe that the RKS was positively and significantly associated with IRA account balance, Stock account balance, and the probability of leaving a bequest of $10,000 for those living in neighborhoods above the sample median NSES. RKS was positively associated with IRA account balance and the probability of leaving a bequest for those living in neighborhoods below the sample median.

Some interesting results emerge when we examine whether the retirement knowledge-wealth relationship differs among Hispanic and White older adults when we separate the groups by NSES. For those in neighborhoods with NSES above the median, the association between RKS and financial outcomes related to IRA and stock account balances appears to be the same for Hispanic and White older adults, as the interaction of race/ethnicity with RKS is not statistically significant. We observe that for Hispanic older adults, the association between RKS and the probability of leaving a bequest of $10,000 is higher than that of White older adults (a 0.53 percentage point increase).

When examining the sample among those living in neighborhoods with an NSES below the median, we find that the total association of RKS with IRA account balance is weaker for Hispanic older adults (—$254) than for White older adults ($1,840). Please note that the magnitude of the relationship between RKS and financial outcomes seems small in the context of the standard deviations of our financial outcomes.

The Relationship between RKS and Health Outcomes

Table 5 provides estimates of relationships between retirement knowledge and health outcomes. The RKS was significantly and positively related to reporting excellent or very good health and to the estimated PROMIS global physical health score for the entire sample. We also find no significant differences in the relationship between retirement knowledge and health outcomes among Hispanic and White older adults, given that the interaction terms are not significant in all cases. When we split the sample by NSES above and below the median, the RKS positive relationship with health outcomes only holds for individuals residing in the NSES below the median.

Table 5.

Regression of relationship retirement knowledge and health for Hispanic and White older adults

	Overall sample		Above sample median NSES sample		Below sample median NSES sample
	Excellent/very good health	PROMIS score	Excellent/very good health	PROMIS score	Excellent/very good health	PROMIS score
RKS	0.0016**	0.0461***	0.0012	0.0309*	0.0025**	0.0709***
	(0.00)	(0.01)	(0.00)	(0.02)	(0.00)	(0.02)
Hispanic	−0.0663	−0.2355	−0.2263*	−3.9488*	−0.0073	1.0216
	(0.07)	(1.38)	(0.12)	(2.23)	(0.08)	(1.72)
RKS*Hispanic	0.0004	−0.0085	0.0042*	0.0747	−0.0019	−0.056
	(0.00)	(0.03)	(0.00)	(0.05)	(0.00)	(0.04)
N	953	954	476	477	477	477
R-squared	0.11	0.17	0.09	0.12	0.11	0.19

Notes:

^***p<0.01,

^**p<0.05,

^*p<0.1.

HRS 2020 data. The sample includes older Hispanic and White adults 50 years old and older. Regressions control for gender, age, education attainment, citizenship status, marital status, retirement status, financial respondent, whether the respondent works for pay, household income (1000s), and social security recipient. We control for NSES when we estimate regressions with overall sample. We estimate regressions separately for individuals who live in census tracts with NSES above median and those who live in census tracts with NSES below median separately. Excellent/very good health is a dependent variable that refers to a dummy variable equal to one if health is excellent or very good, equal to zero otherwise. The PROMIS score dependent variable refers to the estimated PROMIS global health T-score based on the five response categories: Excellent (62.45), Very good (54.42), Good (46.71), Fair (38.04), and Poor (29.48). NSES = Neighborhood Socioeconomic Score; RKS = Retirement Knowledge Scale; PROMIS=Patient-Reported Outcomes Measurement Information System).

We find that for the entire sample, RKS is associated with a 0.0016 percentage point (4%) increase in the likelihood of reporting very good or excellent health. This association was more pronounced among those living in neighborhoods with NSES below the sample median, where a one-point increase in RKS is associated with a 0.0025 percentage point (0.6%) increase in the probability of reporting excellent or very good health. Although the magnitude is small (in relation to the standard deviation of the health variables), being prepared for retirement through retirement knowledge acquisition, measured with the RKS, is associated with better health.

IV. Discussion and Implications

Limitations

There are limitations in our study. First, our analysis is based on a randomly selected subsample of the HRS, which includes approximately 10% of HRS participants. It is possible that our sample may not fully capture the exact characteristics of the full HRS sample, and there may be some selection bias. We compared key sociodemographic variables common in both the experimental module and the entire HRS 2020 wave (public data). We found statistically significant differences at the 5% level for two variables: being born in the US and completing the survey in Spanish. In the experimental module, we have a higher share of White older adults (79% in the experimental module, 77% in the overall sample) than Hispanic older adults (21% in the experimental module, 23% in the overall sample), compared to the full sample. We also have a higher share of US-born respondents (85% in the experimental module, 82% in the overall sample) and a smaller share of respondents who completed the survey in Spanish (13% in the overall sample, 11% in the experimental module) in the experimental module compared to the full sample. Thus, we suggest that future research evaluate racial/ethnic differences in retirement knowledge in the context of a nationally representative sample.

Second, we cannot capture important differences in our sample concerning language and other aspects of immigration. While the RKS was fielded in English and Spanish, only Hispanic older adults answered the survey in Spanish. We also control for whether the individual was born in the US in our analysis, but we are unable to capture other relevant variables, such as years in the United States or immigration status in our main analysis that compares Hispanic and White older adults.

We conducted an analysis to evaluate whether there are statistically significant differences in the RKS among Hispanic older adults with different characteristics related to language and immigration status, using the publicly available data from the HRS. We evaluate whether there are statistically significant differences between the following groups among older Hispanic adults: 1) Spanish Versus English survey, 2) Mexican origin versus other country of origin, 3) permission to provide SSN in previous wave (as a proxy of documentation status), and 4) immigrated to US before 1982 versus in 1982 or after (to account for amnesty given to those who immigrated before January 1^st of 1982). Table A4 presents the mean for the RKS, the difference between groups, and whether the difference is statistically significant (two-sided t-test). We found a statistically significant difference only between older Hispanic adults who completed the survey in Spanish versus those who completed it in English. Hispanic older adults who answered the survey in Spanish have a lower RKS compared to those who answered in English.

We conducted additional estimations based on the models in Tables 4 and 5, where we classified participants into three groups: White (all English speakers), Spanish-speaking Hispanic, and English-speaking Hispanic older adults.⁹ We interacted the racial/ethnic/language dummies with the RKS and included them in the Appendix (Tables A5 and A6). We note that when we include this interaction that accounts for language in our analysis on the relationship between the RKS and financial outcomes for different racial/ethnic groups, results are similar to what we found before but seem to be driven by Spanish-speaking Hispanic older adults (here we compare Table 4 with Table A5). When examining the relationship between the RKS and health outcomes, we observe similar results; however, there is no difference between Hispanic older adults who speak English and those who speak Spanish (as shown in Table 5 compared to Table A6). Due to the small sample size, we exercise caution regarding these additional estimations and suggest conducting this analysis in the future with a larger sample of Hispanic adults who speak both English and Spanish.

Third, while our measure of RKS captures whether individuals have discussed retirement preparedness issues with family and friends, we did not assess other factors associated with knowledge, such as family reliance, informal saving strategies, and the role of remittances. Future work is needed to examine these other aspects of retirement knowledge.

Fourth, our analysis cannot provide any implications related to causality. We can only provide empirical evidence of associations of factors explaining differences in retirement knowledge and how retirement knowledge relates to financial and health outcomes among Hispanic and White older adults. Educational interventions designed with a randomized-controlled trial approach and adequate sample sizes in the context of retirement preparedness are warranted.

Implications of Racial/Ethnic Differences in Retirement Knowledge

Our analysis contributes to the literature by quantifying the differences in retirement knowledge and financial preparedness between older Hispanic and White adults, and by evaluating the factors that explain these differences. We also assess the intersectionality of race and ethnicity with gender when quantifying differences and evaluating the factors that explain these differences.

Our analysis reveals that older Hispanics reported substantially less retirement knowledge and financial literacy than older Whites, with a substantial difference in magnitude (medium effect size). Differences in all items but one of the RKS and FLS were statistically significant across these two groups. The gap was of a larger magnitude for the evaluation dimension than for the reflection and engagement dimensions.

Implications of Factors Explaining Racial/Ethnic Differences in Retirement Knowledge

When analyzing the individual, household, and neighborhood factors that explain differences in retirement knowledge and preparedness among older Hispanic and White adults, we find that financial literacy, educational attainment, and neighborhood characteristics appear to be key factors in explaining these differences. Our findings align with previous research, which has shown that higher levels of SES at both the individual and household levels are associated with greater financial knowledge (Blanco et al., 2024).

Our findings, which relate to the importance of neighborhood characteristics in explaining racial/ethnic differences in retirement knowledge, are similar to those of previous work (Angrisani et al., 2021; Blanco et al., 2019; La Chance, 2014). The Social Determinants of Financial Wellbeing (SDFW) framework is relevant, as neighborhood characteristics relate to structural and institutional barriers that hinder financial knowledge acquisition. Our findings are consistent with the Personal Financial Ecosystem (NEFE, 2022), which suggests that geographic factors related to systematic inequality at the local level are relevant to financial knowledge and financial well-being. Further research is needed to better understand whether these geographic factors are related to access to financial institutions or peer networks.

Implications of Racial/Ethnic Differences and the Intersectionality with Gender

Differences between the racial/ethnic subgroups for most scales and items were similar for men and women. Still, differences between Hispanic and White older men were larger for most items than those between Hispanic and White older women. Previous work has documented that men tend to have higher levels of financial knowledge than women, as measured by both objective and subjective measures (Blanco et al., 2024).

Sample characteristics are relevant to some of our findings (Table A2, Panel A). Hispanics have lower levels of SES than White older adults. Interestingly, SES characteristics of Hispanic older women and older men are similar (Table A2, Panels B and C). For example, 43% of Hispanic older men and women have an educational attainment of GED or less than high school, while for White older men and women, only 13% fall into that category. Hispanic older men have an income level lower than that of Hispanic women by around $1000, while the income of White men exceeds that of women by around $22,000.

Educational attainment appears to explain the difference to a greater degree among Hispanic and White women than among men, and neighborhood characteristics are only relevant in explaining differences among women. One plausible explanation is that there are higher barriers to retirement knowledge acquisition in the context of retirement financial planning. Women with higher levels of education might be able to navigate these barriers to a greater degree. It is also possible that peer networks have a greater impact on women than on men. Further research is warranted to understand the underlying mechanisms that lead to these differences between men and women.

Implications of Racial/Ethnic Differences in the Retirement Knowledge-Wealth Relationship

We found that the association between retirement knowledge and financial outcomes related to wealth differed in some instances for older Hispanic and White adults. Hispanic older adults seemed to be at a disadvantage with respect to the relationship of retirement knowledge with wealth compared to White older adults, and this disadvantage seemed to be more prevalent among those individuals with an NSES below the median. Thus, further research is warranted to explore the channels explaining these differences in the retirement knowledge-wealth relationship by race/ethnicity and neighborhood characteristics.

Examining the SDFW framework is crucial in this context, as structural and institutional barriers can help explain these differences. Further research is warranted to understand whether these differences can be attributed to the supply or demand of financial products essential for asset building, such as retirement and investment accounts.

Implications of Racial/Ethnic Differences in the Retirement Knowledge-Health Relationship

On a positive note, the association between retirement knowledge and health variables was not different between older Hispanic and White adults. This finding is particularly relevant, as retirement knowledge is linked to financial planning for retirement, which in turn is associated with improved health outcomes. While we cannot establish causality in our analysis, this association suggests that financial education may lead to improved health outcomes and reduce health disparities between Hispanic and White older adults.

Our analysis finds health outcome differences between Hispanic and White older adults (Table A2, Panel A). We observe that while 24% of Hispanic older adults reported excellent or very good health, 43% of White older adults fall into this category. Thus, based on the associations shown in our analysis, further research is also warranted on how we can leverage the retirement knowledge-health relationship to address health disparities between older Hispanic and White adults. Here, it will be crucial to investigate the mechanisms by which retirement knowledge is linked to health outcomes, specifically whether this association is due to physical health, mental health, or both.

Data Availability

We utilized data from the Health and Retirement Survey (HRS) 2020 Core and Experimental Module 1: Retirement Knowledge Scale, as well as the American Community Survey (ACS) data available in the restricted-use HRS data enclave. We merged the HRS dataset with the ACS data at the census tract level to construct a neighborhood socio-economic score. All data analyses were conducted in the restricted-use HRS data enclave. Stata code for the analyses is available upon request. The HRS is sponsored by the National Institute on Aging (grant number NIA U01AG009740) and is conducted by the University of Michigan. The study was not preregistered.