Financial Hardship and Psychological Resilience During COVID-19: Differences by Race/Ethnicity

Abstract

Objectives

Research on the impact of coronavirus disease 2019 (COVID-19) among older adults has primarily focused on relatively acute virus outcomes, but it is likely financial hardships during this time have eroded the adaptive capacity of older adults. It is also possible these impacts vary by race and ethnicity. We examine changes in psychological resilience (PR) among older adults before and during the pandemic to determine whether financial hardships and other stressors have altered this resource for White, Black, and Hispanic older adults.

Method

Using the COVID-19 module released by the Health and Retirement Study (n = 735), we examined changes in PR between 2016 and 2020 related to financial hardships during COVID-19. We tested interactions to determine whether the effects were patterned by race and ethnicity.

Results

Consistent with previous literature, resilience was relatively stable during this time on average. Financial hardship during COVID-19 diminished resilience, but this effect was concentrated primarily among White Americans.

Discussion

The results suggest that PR is a relatively stable resource in later life, even during the pandemic. However, this resource may be affected in the face of specific challenges in later life. Policies related to financial hardship during the pandemic should be seen as supporting the capacity for older adults to adapt to current as well as future challenges.

Coronavirus disease 2019 (COVID-19) has brought numerous hardships for older Americans. In addition to over 41 million cases and 660,000 deaths at the time of this writing (Centers for Disease Control and Prevention, 2021), there have been challenges including loss of family and friends, stressors due to mitigation strategies, job loss, and financial hardship (Pew Research Center, 2021). The impacts of COVID-19 on older adult well-being have been documented (Pearman et al., 2021; Whitehead, 2021). However, the unexpected, uncontrollable, and pervasive nature of the pandemic may also have lasting impacts for perceived resilience to future challenges.

Studies suggest that among stressors experienced during this time, financial loss and related hardships may be among the most enduring in their impacts (Pew Research Center, 2021). Compared to younger adults, older adults are less likely to be in the labor force and have fewer years to accumulate wealth. Therefore, they may have fewer resources to recover from financial setbacks (Morrow-Howell et al., 2020). Although stress related to infection or grief from loss may be more immediately damaging, research suggests financial hardships during the pandemic may undermine a sense of security, personal coping factors, and meaning in life (Coulombe et al., 2020; Pew Research Center, 2021). These internalized resources are key in the ability to face future struggles and challenges often occurring in later life, such as disablement or widowhood (Taylor & Carr, 2020). Over a year into the pandemic, roughly one quarter of households reported difficulty covering expenses such as food, housing, and medical care (Center on Budget and Policy Priorities, 2021). Studies on the Great Recession suggest that financial hardships have lasting effects for older adults and especially for minority individuals (Cohen et al., 2020). Yet, there has been little to no empirical research on the impact of such hardships for older adults during the pandemic.

One way to capture the potential lasting impact of financial hardships during COVID-19 is to examine associations between these experiences and individual levels of psychological resilience (PR), a measure tapping into internalized resources related to the ability to adapt to setbacks and stressors. PR is uniquely relevant to COVID-19, because it is separate from distress outcomes such as depression or anxiety but may better gauge how well older adults are equipped to meet future challenges. Here we define PR as an individual’s internalized ability to adapt effectively to adversities (Manning, 2013). Previous research shows that PR is associated with health, well-being, and engagement in later life (Jason et al., 2017; Taylor & Carr, 2021), suggesting its erosion may predict an array of negative outcomes. PR is generally stable over time among older adults (Taylor et al., 2019), but certain factors can undermine it. Research suggests that financial hardship is one factor that may erode psychological resources associated with adaptive coping in later life (Elder & Liker, 1982). As Levine (2003) states, “even the most resilient soul can be downtrodden, degraded, and ultimately defeated” with the right combination of risk factors and resource deprivation.

However, research also documents that successful responses to adversities in well-resourced and supportive environments may actually promote PR and facilitate a resilient identity (Elder & Liker, 1982; Manning & Bouchard, 2020). Studies of resilience in older adults indicate that PR acts as a multidimensional buffer against challenges and adverse events (Jason et al., 2017; MacLeod et al., 2016). Some people maintain stability or recover after setbacks, and some report personal growth. In summary, hardships during COVID-19 may have variable impacts on later-life PR. Research on the effects of such hardships poses a unique opportunity to understand the current and future extent of pandemic burden for older Americans, but also to understand if such hardships can erode resilience during a time of pervasive and unexpected stressors.

The “double jeopardy” of age and racial minority status has been a substantial topic of COVID-19 research (Chatters et al., 2020; Macias Gil et al., 2020). Black and Hispanic/Latinx individuals experience higher infection and death rates (CDC, 2020). Older minority individuals are more likely to experience job and income loss and economic hardship during this time (Cohen et al., 2020). Literature suggests that because minority older adults disproportionately experience structural racism, life course inequality, and COVID-19 hardships, they may experience reduced well-being and greater stress-related “weathering” processes (Garcia et al., 2021; Greer, 2011; Williams et al., 2019).

However, research on psychosocial resources and flourishing among minority groups also highlights relatively high observed levels of many resources related to PR such as self-esteem, mastery, social support, and religious involvement (Erving et al., 2020; Ryff et al., 2003; Twenge & Crocker, 2002). Although high-effort coping strategies tied to pervasive inequalities may be linked to worse physical health outcomes (James, 1994), it is possible that older minority individuals will better maintain PR in the face of stressors during the pandemic. Our primary research questions in this report are (a) How do financial hardships during COVID-19 impact change in PR between 2016 and 2020, net of other stressors? (b) How do the impacts of financial hardships vary for White, Black, and Hispanic older adults?

Method

Data and Study Sample

Our study utilized the Health and Retirement Study (HRS), a National Institute on Aging-funded nationally representative longitudinal study with the biennial collection since 1992 and new cohorts added every 6 years (HRS, 2016). Our sample focuses on a subset of the 2016 HRS cohort participating in a 2020 COVID-19 module on experiences during the pandemic. An alternating randomly drawn half of HRS respondents complete the battery of psychological functioning measures as part of the Leave-Behind Questionnaire (LBQ) in each biennial wave. The sample from the COVID-19 module was linked to measures in both 2016 and the summer of 2020, allowing the evaluation of longitudinal changes in psychosocial outcomes related to COVID-19 and other experiences.

Our sample was derived from the 3,266 HRS respondents completing the June 2020 base study and COVID-19 module. The sample was restricted to the randomly sampled 1,146 respondents completing the 2020 LBQ, because the HRS deploys the LBQ to a random half of study participants each wave. We further restricted the sample to respondents who also completed the 2016 LBQ. We removed individuals with a baseline age less than 51 (n = 59) and individuals reporting “other” race (n = 54). This resulted in a sample of 735 individuals with data on PR in both 2016 and 2020 who completed the COVID-19 module.

Measures

Resilience

Our primary interest is in changes in PR between 2016 and 2020. PR was measured with the Simplified Resilience Score, developed from items in the HRS (Manning et al., 2016). This measure draws on LBQ measures, has high reliability (α = 0.84), and is strongly predictive of health and well-being (Taylor & Carr, 2020). See Supplementary Appendix 1 for further description, measure items, and coding. All items were standardized (0–1) because items varied in their original ranges. The standardized items were summed with observed values ranging from 2.8 to 12. The outcome measure was then standardized.

Financial hardships and race/ethnicity

Financial hardships were measured as a count, experienced since the start of the pandemic. Hardships include missing any regular payment on (a) rent or mortgage, (b) credit cards or other debt, (c) utilities or insurance, or if respondents report they, (d) could not pay medical bills, or (e) did not have enough money to buy food, or (f) were otherwise unable to buy food. Observed values ranged from 0 to 5. Because we are interested in the differential impact of financial hardships during the COVID-19 pandemic on PR for White non-Hispanic, Black non-Hispanic, and Hispanic individuals, we restricted the sample to these mutually exclusive groups, including binary markers (Black = 1, Hispanic = 1, White = referent) in all models.

Other COVID-19-related hardships and controls

We control for other COVID-19-related stressors including binary measures of (permanent or temporary) job loss during the pandemic, whether the respondent was infected or knew others who were infected with COVID-19, and/or had any friends/family who died of COVID-19. Controls for previous financial strain and health-related stressors were measured with two binary indicators taken from the 2016 LBQ capturing chronic stress related to ongoing financial strain and ongoing health problems. Respondents were asked on a scale of 1–4 to indicate whether these were current and ongoing problems lasting at least 12 months, along with how upsetting they were. We also controlled for 2016 self-rated health using a binary indicator capturing fair or poor health and respondent’s count of depressive symptoms ranging from 0 to 8. Respondent’s binary gender (female = 1), education (in years), and age (in years at 2016) were also included.

Analysis

Ordinary least squares (OLS) regression with robust standard errors was used to examine the level of PR in 2020 controlling on the respondent’s level in 2016. Therefore, models can be interpreted as the effect of covariates on standard deviation increases/decreases in 2020 PR net of 2016 PR. We present nested models to (a) establish the main effects of financial hardships and race/ethnicity, (b) introduce interaction terms between race/ethnicity and financial hardship, and (c) adjust for other COVID-19-related stressors, previous hardships, and controls.

Results

Descriptive statistics are presented in Table 1. Consistent with prior work (Taylor et al., 2019), change in resilience between 2016 and 2020 was not significant on average across race/ethnicity. Black respondents had significantly lower 2020 PR compared to Whites, but Hispanic respondents were not significantly different. Both Black and Hispanic respondents reported greater financial hardships during COVID-19, job loss, and infection rates. Black respondents also experienced greater COVID-19 loss and financial strain in 2016. Both Black and Hispanic respondents were more disadvantaged than their White counterparts, reporting lower levels of education and physical and mental health.

Table 1.

Descriptive Characteristics of Study Sample and by Race Groups

	All (n = 735)				White (n = 509)		Black (n = 140)		Hispanic (n = 86)
	Mean	SD	Min	Max	Mean	SD	Mean	SD	Mean	SD
Key variables of interest
Resiliency (post-COVID-19)	9.630	1.719	2.810	12	9.728	1.692	9.327*	1.830	9.542	1.645
Resiliency (baseline)	9.573	1.715	2.786	12	9.665	1.660	9.384	1.677	9.333	2.041
Total financial hardships	0.181	0.623	0	5	0.079	0.000	0.407***	0.988	0.419***	0.887
Race
White	0.693		0	1
Black	0.190		0	1
Hispanic	0.117		0	1
COVID-19-related measures
COVID-19 exposure
No exposure	0.460		0	1	0.642		0.407***		0.500*
Self, friend, or family infected	0.122		0	1	0.332		0.493***		0.465*
Friend or family died	2.079		0	1	0.026		0.100***		0.035
Lost job due to COVID-19	0.122		0	1	0.098		0.164*		0.198**
Ongoing stressors (baseline)
Ongoing health problems	2.079	0.936	1	4	2.029	0.918	2.171	0.921	2.221	0.975
Ongoing financial strain	1.688	0.907	1	4	1.640	0.866	1.871**	0.958	1.674	0.874
Demographic factors
Female	0.562		0	1	0.538		0.643*		0.570
Years of education	13.578	2.836	0	17	14.084	2.324	13.400**	2.521	10.872***	4.220
Age	64.550	7.338	51	1	65.442	7.881	62.657***	5.832	62.349***	4.750
Fair/poor health	0.199	0.399	0	1	0.138	0.345	0.279***	0.450	0.430***	0.498
Depressive symptoms	1.068	1.710	0	8	0.890	1.551	1.443***	1.909	1.512**	2.074

Notes: COVID-19 = coronavirus disease 2019. Statistical significance indicates differences relative to Whites.

***p < .001, **p < .01, *p < .05.

We present results from OLS regressions with robust standard errors in Table 2. Model 1 suggests there were no significant race/ethnicity differences in PR change between 2016 and 2020, and that financial hardships were not significant on average. In Model 2, the main effect of COVID-19 financial hardships became significant and negative, and the interaction effect became significant and positive for Black respondents. This suggests that Black older adults better maintained PR in the face of financial hardships compared to their White counterparts. These impacts were not significantly different between White and Hispanic respondents. Model 3 introduces other COVID-19-related stressors, baseline financial and health strain, and other controls. Figure 1 provides a visual interpretation of the estimated marginal effects for Model 3 with other measures held at the mean. Among those reporting no financial hardships during COVID-19, White, Black, and Hispanic respondents experienced a negligible change in PR. Black respondents showed a slight nonsignificant increase in PR change with additional financial hardship, but White and Hispanic respondents experienced reductions in PR. At high levels of hardship (4.5 out of 5), White respondents saw a 1 SD reduction in PR change between 2016 and 2020, the greatest of any group. Model 3 also suggests there were no significant effects for other COVID-19 stressors. Alternate interaction terms testing race differences for other COVID-19 stressors (not shown) were not significant. Together, Model 3 suggests that financial hardships during COVID-19 were the only stressor measured here to significantly affect PR, and this effect was only observed significantly among White older adults.

Table 2.

OLS Regression Models Predicting Change in Resilience (n = 735)

	Model 1	Model 2	Model 3
	Beta (SE)	Beta (SE)	Beta (SE)
Resiliency (baseline)	0.671*** (−0.032)	0.671*** (−0.032)	0.607*** (−0.037)
Total financial hardships	−0.047 (−0.063)	−0.240* (−0.114)	−0.178 (−0.118)
Race
Black	0.006 (−0.076)	−0.065 (−0.079)	−0.055 (−0.077)
Hispanic	−0.107 (−0.097)	−0.071 (−0.105)	0.046 (−0.106)
Black × Total financial hardships		0.360** (−0.129)	0.284* (−0.133)
Hispanic × Total financial hardships		0.061 (−0.19)	0.026 (−0.18)
COVID-19 exposure
Self, friend, or family infected			0.059 (−0.056)
Friend or family died			−0.059 (−0.146)
Lost job due to COVID-19			0.006 (−0.087)
Ongoing stressors (baseline)
Ongoing health problems			−0.088* (−0.036)
Ongoing financial strain			0.043 (−0.036)
Demographic factors
Female			0.09 (−0.055)
Years of education			0.028* (−0.011)
Age			−0.009* (−0.004)
Fair/poor self-rated health			−0.178† (−0.095)
Depressive symptoms			−0.018 (−0.02)
Constant	0.019 (−0.032)	0.032 (−0.032)	0.319 (−0.333)
R  2	0.457	0.465	0.496
−2 log-likelihood	−818.1	−812.8	−790.5

Notes: OLS = ordinary least squares; COVID-19 = coronavirus disease 2019. Standard errors are robust.

Statistical significance indicates: ***p < .001, **p < .01, *p < .05, ^†p < .10.

Figure 1.

The figure is based on Model 3 with covariates held at the mean. Statistical significance indicates statistical difference at each number of financial hardships relative to Whites: **p < .01; *p < .05. Estimates for Hispanic older adults are not significantly different from Whites. The slope for White older adults is significantly different from 0 at p < .05, indicating a decline in psychological resistance associated with increased financial hardship. The slopes for Black and Hispanic older adults are not significantly different from 0.

Discussion

Our findings suggest there were no substantial reductions in PR between 2016 and 2020 on average. This is consistent with the stability of this personal resource, but also a testament to the ability of older adults to maintain resilience in the face of the pandemic. White older adults, and to some degree Hispanic/Latinx older adults, did experience erosion in PR related to financial hardships. The resilience of Black older adults remained stable. These findings are consistent with research suggesting higher observed psychosocial resources and coping factors among African Americans (Erving et al., 2020; Ryff et al., 2003). Research also suggests financial stability matters less for well-being among those who derive self-worth from social rather than status resources (Richards, 2016).

Overall, these findings suggest that policies related to financial hardship during the pandemic, such as tax credits and unemployment benefits, will be important in bolstering PR for the future. Research suggests that adaptive coping is fostered after setbacks when individuals are not faced with chronic economic or social deprivation (Elder & Liker, 1982). We note that although our findings highlight the erosion of PR among White respondents, and to a lesser degree Hispanic respondents, Black respondents reported significant ongoing financial strain in 2016 coupled with significantly poorer health and depressive symptoms. It is likely that different outcome variables would better capture the impacts of financial strain for these older adults.

We acknowledge other limitations here. Due to measurement wording and wave spacing, we cannot know if financial hardship was caused by the pandemic and could only control for previous hardships in 2016. We did not have measures on other pandemic stressors (e.g., social isolation) and coping (e.g., engaging social networks) that are tied to PR. It is also possible that factors shaping PR for Black and Hispanic respondents during 2020 were not captured in our data, such as political climate or media coverage of violence. We encourage future research to reexamine conceptualizations of resilience and alternate measures of stressors, hardships, coping, and well-being to understand adaptive capacity among diverse older adult populations.

Funding

The third author (to K. J.) reports funding from Advanced Respiratory Research for Equity (AIRE) Programs to Increase Diversity Among Individuals Engaged in Health-Related Research (PRIDE) grant number 2R25- HL126140.

Conflict of Interest

None declared.