Skip to main content
PMC home page
SSM - Population Health logoLink to SSM - Population Health
. 2024 Aug 8;27:101704. doi: 10.1016/j.ssmph.2024.101704

Empty chairs at the dinner table: Black-white disparities in exposure to household member deaths

Angela R Dixon 1
PMCID: PMC11389549  PMID: 39262768

Abstract

As a result of Black–White inequities in life expectancy, recent research has indicated that Black individuals are disproportionately exposed to the deaths of multiple family members compared to White individuals. Black individuals are also more likely to live in coresident households—that is, households that extend beyond the nuclear family. However, it is unclear the degree to which this population may be disproportionately exposed to the loss of deaths marked by the geographic closeness of the household. In this study, I use data from the Panel Study of Income Dynamics to provide the first nationally representative estimates of Black-White disparities in exposure to household member deaths. I find that Black people are significantly more likely than White individuals to have experienced the death of a household member. Based on these findings, I argue the dual inequities of racial disparities in life expectancy and racial disparities in coresidence are an overlooked, salient source of racial disparities in exposure to death. By illuminating a broader range of network sources that contribute to racial inequities in exposure to death, this study sets forth a new conceptual unit of analysis—that of the household—to investigate the intergenerational reproduction of inequality in health and socioeconomic status due to exposure to death.

Highlights

  • •

    Introduces novel measure of exposure to household member death.

  • •

    Shows Black people are 1.6 times more likely to experience the death of at least 1 household member.

  • •

    Finds Black people are 2.5 times more likely to experience the death of at least 2 household members.

  • •

    Identifies dual inequity of higher mortality rates and higher rates of coresidence.

  • •

    Establishes the household as a key source of racial disparities in exposure to death.

There’s a grief that can’t be spoken

There’s a pain goes on and on

Empty chairs at empty tables

– Les Misérables (1985)

1. Introduction

An increasing proportion of Americans are residing in diverse living arrangements that extend beyond the idealized nuclear family model, which consists of two biological parents and their children (Bengtson, 2001). For instance, the percentage of Americans living in multigenerational households increased from 12% in 1980 to 20% in 2016 (Fry, 2019). Yet, this trend of diversifying household living arrangements is not equally distributed by race. Black individuals are more likely to reside in households comprising members beyond the nuclear family model than White people. For example, Black children live with extended family members at a rate almost three times that of White children, with much of this coresidence occurring due to socioeconomic status (SES) constraints (Cross, 2018).

In addition to higher rates of coresident households, Black individuals have also faced lower life expectancy for centuries as a result of historical and contemporary structural racism (Williams et al., 2019). In 1900, Black life expectancy at birth was 33 years, 14.6 years less than the White life expectancy of 47.6 (Hummer & Hamilton, 2019). The gap narrowed by 2019, with Black people living, on average, four years less than White individuals. However, in the last century, Black people had a mortality rate averaging 60 percent higher than White individuals (Rodriguez et al., 2015). These large and persistent Black–White gaps in life expectancy have long been recognized, but researchers have only started to investigate the ways increased exposure to death shapes the lives of Black individuals.

Bereavement, the state of having lost a close relationship due to death, is associated with poor mental and physical health, including an increased risk of mortality among those left behind (Oxford University Press, 2023; Stroebe et al., 2007). Despite the importance of bereavement, the vast majority of work examining Black-White disparities in exposure to death has focused on exposure to the deaths of individuals presumed to be important due to the nature of the relationship, such as primary kin – which include parents, children, and siblings – (Cha et al., 2022, Donnelly et al., 2020, Kidman et al., 2021, Reis and Sprecher, 2009, Thyden et al., 2020, Thyden et al., 2023, Umberson et al., 2019), and to a lesser degree extended family (e.g., cousins, aunts, uncles), and grandparents (Patterson et al., 2020; Verdery et al., 2020). A critical gap in the current literature is the lack of attention to the decedent's place of residence before death, a potential indicator of the social tie's intensity and importance.

In this study, I argue that disparities in coresident households increase Black people's potential exposure to close proximity relationships, thereby serving as another salient source of exposure to death. Household deaths may be especially important because of their relevance to and disruption of daily life routines for surviving household members. Examining this connection is important given two co-occurring inequities: (1) increased rates of coresident households that vary substantially by race and (2) entrenched racial disparities in life expectancy.

In this study, I introduce the household unit as a salient source of exposure to death. This study provides the first nationally representative estimates of Black-White disparities in exposure to household member deaths using data from the Panel Study of Income Dynamics (PSID), the world's longest-running household panel study. This study extends the extant literature on bereavement by examining racialized patterns of exposure to household member deaths. In line with the lifecourse framework, I establish the death of a household member as a pivotal event in the lifecourse that may contribute to cumulative disadvantage and weathering, especially among Black populations (Geronimus, 1992). Through this analysis, I offer novel insights into the ways household member deaths may be a hidden social determinant of health and well-being contributing to racial inequities in health in the U.S.

2. Background

2.1. Racial disparities in coresidence are a racial inequity

American households are increasingly complex. By 2017, 31.9% of adults lived in “a household with at least one adult who is not the household head, the spouse, the unmarried partner of the head, or an 18- to 24-year-old student” (Fry, 2018). Relatedly, between 1996 and 2009, the percentage of children residing in shared households rose from 17.6% to 20.8%, with the greatest increase among children living with grandparents (Pilkauskas & Cross, 2018).

Within these overall trends, there are large differences by race, potentially placing Black individuals at greater risk of exposure to household member deaths. For example, although Cross (2018) found that overall, 35 percent of children ever coresided with extended kin, these rates are much higher for Black (57%) than for White children (20%). Moreover, Black children (15%) are more than twice as likely to live with a grandparent than White children (7%); the disparity in the likelihood of a grandparent being the child's primary caregiver is even greater (8% vs. 3%), potentially resulting in increased risk of exposure to household member deaths given that most deaths occur at older ages (Pew Research Center, 2013).

I contend that the reasons why individuals decide to coreside are of critical concern and central to investigations of racial disparities in exposure to death. Specifically, I argue that higher rates of Black coresidence, in itself, represent an inequity. The difference in prevalence rates of coresidence between Black and White households is not inherently positive or negative. It is possible that Black individuals could simply choose to coreside at higher rates due to personal or cultural preferences. If Black individuals preferred to coreside and then were exposed to the same mortality rates as White people, they would experience higher levels of exposure to household member mortality. In this case, the heightened rate of exposure to death would simply be an outgrowth of a personal/cultural choice that might be balanced by the numerous potential benefits of coresidence. However, this is not the case.

Research suggests there is a nonrandom selection of families into coresident households, with socioeconomic constraints as a primary driver of the formation of these households (Cross, 2018). Coresident households tend to have higher poverty rates and lower levels of education and income (Dunifon et al., 2014). For example, Amorim (2019) found parents in three-generation households earn only about half as much as their two-generation counterparts. Importantly, available evidence suggests coresidence does not cause poverty but is often a response to socioeconomic hardship (Amorim, 2019; Dunifon et al., 2014). Research has also suggested a primary explanation for racial differences in coresidence rates is Black individuals’ lower SES (Cross, 2018). Therefore, given that SES disparities are largely responsible for driving individuals to seek shared households, I argue that racial disparities in coresident households represent and reproduce racial inequities.

2.2. Why household member deaths matter

Black people are not only at greater risk of dying prematurely; research has documented a higher likelihood of experiencing the deaths of multiple family members (Donnelly et al., 2023, Umberson et al., 2017). These losses are significant because growing evidence suggests these racial disparities in exposure to death contribute to racial inequities in health across a range of outcomes, including increased risk of mortality, lower subjective life expectancy, increased risk of cognitive impairment, worse cardiometabolic health, and dementia (Donnelly et al., 2022, Donnelly et al., 2020, Donnelly et al., 2018, Liu et al., 2022, Umberson et al., 2019).

An extensive literature documents the benefits of social relationships for health and well-being (Umberson, 2017, Umberson and Montez, 2010). Although there can be costs associated with poor-quality relationships, research suggests that high-quality close relationships and the quantity of social ties are important resources and potential sources of social, instrumental, emotional, and financial support. Consequently, the loss of such ties can deprive individuals of a key resource (Umberson, 2017, Umberson and Donnelly, 2023). Given Black Americans’ increased exposure to death, Umberson (2017) argues that “bereavement and loss constitute a unique stressor” for Black Americans that impacts social relationships and accumulates over the lifecourse to negatively impact health and well-being.

Although nonresidential family and ties are undeniably important sources of support, losing a household member likely has unique importance. Ruggles (1994:128) argues,

But the very fact that kin do not live together almost inevitably means that they play a relatively small role in one another’s everyday life. According to a recent survey, a majority of elderly say they saw at least one of their children within the previous week. A hundred years ago, however, most elderly saw one of their children at breakfast each morning. However great the interaction of kin who live separately, it is bound to be less than the interaction of kin who live together. Co-residence is not just the best indicator of the intensity of kin interaction over the past century and a half, it is the only consistently available indicator [emphasis added].

Therefore, the death of a household member can deprive the bereaved of intense support that would be hard to replace, even in the presence of other strong nonresidential social ties. Given that coresidence is more common among less resourced individuals and is often driven by the desire to pool resources (Amorim, 2019; Cross, 2018; Dunifon et al., 2014), a household member death may prove particularly detrimental to already disadvantaged households. Furthermore, the death of a household member may increase the likelihood of living alone and social isolation, which increases the risk of poor health, especially for older adults (Donnelly, Lin, & Umberson, 2023; Klinenberg, 2016; Umberson & Donnelly, 2023).

Examining disparities in exposure to household member deaths at different points in the lifecourse is crucial for understanding their impact on trajectories. Prior research suggests Black Americans are less likely than White Americans to have living spouses, kin at older ages, or any living kin due to mortality differentials (Daw et al., 2016; Verdery & Rachel, 2017). The stress accompanying the loss of a loved one or multiple loved ones, along with subsequent stressors, accumulates over the lifecourse to negatively impact well-being (Geronimus, 1992). Studies show that deaths occurring earlier in the lifecourse can have significant and compounding impacts that negatively alter health and socioeconomic trajectories. For instance, the loss of nuclear and extended family members during childhood and adolescence is associated with lower educational attainment (Patterson et al., 2020; Thyden et al., 2023; Thyden et al., 2020) and shorter telomere length, which is a predictor of future poor health (Mitchell et al., 2017). Children living in multigenerational households with grandparents may face a greater risk of exposure to death due to the age gradient of mortality. Household losses later in life can deprive the bereaved of social relationships’ health benefits, which are especially critical in later life (Umberson and Montez, 2010).

2.3. Contributions of this study

Due to the previous focus on the deaths of family members based on the nature of the tie without consideration of coresidence, it is still unclear how exposure to deaths at the household level structures patterns of exposure to death. In this study, I hypothesize that household member deaths are an unexplored source of racial disparities in exposure to death. This disparity in exposure to death arises due to dual inequities in the Black community: higher rates of mortality and higher rates of coresidence. Taken together, if Black individuals are exposed to higher rates of household member mortality, these disparities would represent an avoidable, unjust inequity due to the combination of both higher rates of mortality and coresidence, especially in light of robust evidence that coresidence occurs due to constrained SES resources, not personal preferences.

3. Materials and methods

3.1. Data

For this study, I use data from the PSID, which began in 1968 with over 5000 households (Panel Study of Income Dynamics, 2021). The PSID provides a nationally representative sample of U.S. families, with an oversample of Black individuals. It was administered annually from 1968 to 1997 and biennially since. The PSID has a unique genealogical design that follows the original 1968 Sample Members and their descendants over time as they split off and form their own independent households. PSID Sample Members are defined as those in the original 1968 sample and their descendants who were born after the initial wave (PSID, 2023). The analytic sample is composed of the cross-section of PSID Sample Members who were alive and observed in the 2019 wave of the PSID. The final analytic sample consists of 6817 non-Hispanic Black and 7808 non-Hispanic White respondents.

3.2. Measures

3.2.1. Primary dependent variable

The primary dependent variable of interest was the total number of household member deaths an individual experienced. This was operationalized using two binary measures: 1) experiencing at least one household member death versus no deaths, and 2) experiencing at least two household member deaths versus no deaths. The death of a household member was ascertained by using the 2019 PSID Mortality File. I draw on data from 1968 to 2019 to calculate the number of household member deaths each respondent experienced throughout their entire period of observation in the PSID.

I constructed the household member death measure by identifying the last wave in which the decedent was listed as residing in a PSID household. If the recorded date of death from the mortality file was within two years of the last household interview, everyone else in the household at the time of the last interview was considered to have experienced a household member death. For example, if a decedent lived in a household in 2011 and died in 2013, I recorded the decedent as a household death for each person in the 2011 household. If the precise year of death was unknown and the mortality file provided a date range, the death was coded as missing if the interval was greater than three years. If the interval was three years or less, I assigned the date of death as the upper bound of the range. This restriction gave a conservative estimate of the number of household member deaths by reducing the likelihood that the death would qualify for linkage back to a household.

I could not directly determine whether the decedent died while still living in the household as opposed to, for example, having recently moved to a different housing or nursing facility. Therefore, this measure captures the death of a household member who either resided in the household at the time of the death or recently. The PSID does not track short-term coresidents, such as short-term roommates.

3.2.2. Primary independent variable

The primary independent variable of interest was racial identification. Because PSID is a household survey, race is a household-level measure. The first racial identity reported for the reference person was assigned to everyone in the household unless there was a spouse/partner. If a spouse/partner was present, that person's first reported identity was used instead. Only non-Hispanic Black or non-Hispanic White respondents were included in the sample. Other racial groups were not included in the analysis because the PSID was started in 1968 to represent the U.S. population at that time. Although the PSID added additional members from other racial groups with an Immigrant Refresher Sample in 1997, these more recent additions lack the long-term histories that are of most value to this project.

Respondents were categorized into five lifecourse stages based on their age in 2019: early life (age<18),1 transition to adulthood (18–28), midlife (29–49), later life (50–64), and older adults (65+). SES was operationalized using educational attainment and wealth quartiles. Household wealth (quartiles) was defined as the sum of all assets (e.g., home equity, vehicles, investments) minus debts (e.g., credit cards, student loans, personal loans). Wealth was used instead of income because it more accurately captures financial well-being over time and is not as subject to short-term shocks (Killewald et al., 2017). Education was not reported for individuals under 16 years of age. I constructed the following categorical educational attainment variable for those aged 25 and older: less than a high school education, high school degree or GED, some college, and college or greater. The models also included the following variables as covariates: gender (binary) and age (continuous).

3.3. Analytic plan

I first describe characteristics of the study population (Table 1), overall and stratified by race. All percentages and means are survey weighted to be nationally representative. In Table 2, I estimate a series of multilevel mixed-effects Poisson regression models to estimate risk ratios predicting the likelihood of exposure to one or more household deaths compared to no deaths and exposure to two or more household deaths compared to no deaths (Table 2). Multilevel mixed-effect Poisson models are used to account for the multilevel nature of the data with individuals clustered within households. Here, I employ Poisson regressions, as opposed to logistic regression, due to ease of interpretability (Gallis & Turner, 2019; Zou, 2004). All models are adjusted for age and gender. Next, I stratify the main models by the five lifecourse stages to examine racial disparities across various life stages (Table 3). Finally, in Table 4, I replicate the models from Table 2, adjusting for SES to investigate the degree to which Black-White inequities in exposure to household member deaths may be explained by SES. These models are restricted to respondents aged 25+ due to the inclusion of educational attainment. All analyses use the PSID-provided sample weights to adjust for complex sampling design.

Table 1.

Characteristics of the study population from the 2019 PSID.

Race
Total White Black
Household Member Deaths
 At Least 1 Death 16.2% 15.4% 19.7%
 At Least 2 Deaths 2.9% 2.5% 4.5%
Lifecourse Stage
 Early Life (<18) 23.4% 22.1% 28.6%
 Transition to Adulthood (18–28) 15.9% 15.2% 18.8%
 Midlife (29–49) 23.3% 22.8% 25.2%
 Later Life (50–64) 17.3% 17.6% 15.9%
 Older Adults (65+) 20.1% 22.1% 11.6%
Wealth Quartiles
 Q1 (<$5650 (including debt)) 23.8% 19.0% 44.3%
 Q2 ($5651-$76,000) 22.7% 21.2% 29.1%
 Q3 ($76,001-$317,000) 26.0% 27.0% 21.5%
 Q4 ($317,001+) 27.5% 32.7% 5.1%
Education
 Less than High School 8.1% 7.4% 10.9%
 High School 26.7% 26.3% 28.1%
 Some College 18.0% 18.0% 17.9%
 College+ 22.3% 24.7% 12.4%
 Not Reported (age < 16) 25.0% 23.6% 30.8%
Demographics
 Age (mean) 39.87 41.17 34.35
 Female 50.9% 50.1% 54.3%



n 14,625 7808 6817
Open in a new tab

All means and percentages reflect weighted estimates from the sample.

Table 2.

Risk ratios from modified Poisson regressions predicting exposure to household death.

At Least 1 Death
 At Least 2 Deaths
(1) (2)
Race (ref. White)
 Black 1.632∗∗∗ (0.158) 2.518∗∗∗ (0.564)
Age 1.033∗∗∗ (0.001) 1.044∗∗∗ (0.003)
Gender (ref. Male)
 Female 1.145∗ (0.074) 1.210 (0.187)
Constant
 0.028∗∗∗ (0.003)
 0.002∗∗∗ (0.001)
n 14625 14625
Open in a new tab

Exponentiated coefficients; Standard errors in parentheses.

+ p < 0.10, ∗ p < 0.05, ∗∗ p < 0.01, ∗∗∗ p < 0.001.

Table 3.

Risk ratios from modified Poisson regressions predicting exposure to household death by lifecourse stage.

At Least 1 Death
 At Least 2 Deaths
(1)
 (2)
 (3)
 (4)
 (5)
 (6)
 (7)
 (8)
 (9)
 (10)
Early Life TAS Midlife Later Life Older Adults Early Life TAS Midlife Later Life Older Adults
Race (ref. White)
 Black 2.088∗ 1.651∗∗ 2.130∗∗∗ 1.698∗∗∗ 1.114 2.625 3.818+ 2.765∗∗ 3.231∗∗∗ 1.924∗
(0.699) (0.283) (0.351) (0.194) (0.144) (3.218) (2.849) (1.027) (1.009) (0.573)
Age 1.211∗∗∗ 1.061+ 1.035∗∗ 1.028∗ 1.034∗∗∗ 1.401∗∗∗ 1.160+ 1.089∗∗ 0.953 1.023
(0.032) (0.032) (0.012) (0.014) (0.005) (0.066) (0.097) (0.034) (0.033) (0.014)
Gender (ref. Male)
 Female 0.896 0.776 0.869 1.089 1.586∗∗∗ 0.121∗ 0.999 0.848 0.814 1.925∗
(0.194) (0.148) (0.090) (0.126) (0.171) (0.102) (0.735) (0.308) (0.236) (0.601)
Constant
 0.003∗∗∗ 0.020∗∗∗ 0.030∗∗∗ 0.041∗∗∗ 0.020∗∗∗ 0.000∗∗∗ 0.000∗∗∗ 0.001∗∗∗ 0.521 0.008∗∗∗
(0.001)
 (0.016)
 (0.014)
 (0.032)
 (0.008)
 (0.000)
 (0.000)
 (0.001)
 (1.066)
 (0.009)
n 4880 2539 3874 1847 1485 4880 2539 3874 1847 1485
Open in a new tab

Exponentiated coefficients; Standard errors in parentheses Note: TAS (Transition to Adulthood).

+ p < 0.10, ∗ p < 0.05, ∗∗ p < 0.01, ∗∗∗ p < 0.001.

Table 4.

Risk ratios from modified Poisson regressions predicting exposure to household death adjusting for SES (age 25+).

At Least 1 Death
 At Least 2 Deaths
(1) (2) (3) (4) (5) (6) (7) (8)
Race (ref. White)
 Black 1.609∗∗∗ 1.471∗∗∗ 1.392∗∗∗ 1.352∗∗ 2.416∗∗∗ 2.104∗∗ 1.705∗ 1.633∗
(0.161) (0.151) (0.133) (0.131) (0.523) (0.479) (0.363) (0.363)
Age 1.028∗∗∗ 1.025∗∗∗ 1.031∗∗∗ 1.028∗∗∗ 1.036∗∗∗ 1.032∗∗∗ 1.043∗∗∗ 1.040∗∗∗
(0.002) (0.002) (0.002) (0.002) (0.005) (0.005) (0.004) (0.004)
Gender (ref. Male)
 Female 1.241∗∗ 1.225∗∗ 1.202∗ 1.204∗ 1.321+ 1.290 1.196 1.201
(0.086) (0.086) (0.086) (0.086) (0.207) (0.207) (0.193) (0.193)
Education (ref. College+)
 Less than High School 2.105∗∗∗ 1.825∗∗∗ 3.207∗∗∗ 2.083∗∗
(0.224) (0.184) (0.905) (0.530)
 High School 1.735∗∗∗ 1.593∗∗∗ 1.953∗∗ 1.517+
(0.118) (0.110) (0.438) (0.316)
 Some College 1.431∗∗∗ 1.340∗∗ 1.776∗ 1.448+
(0.126) (0.116) (0.414) (0.319)
Wealth (ref. Q4)
 Q1 1.727∗∗∗ 1.475∗∗∗ 3.819∗∗∗ 3.215∗∗∗
(0.161) (0.132) (0.835) (0.670)
 Q2 1.591∗∗∗ 1.361∗∗ 3.188∗∗∗ 2.699∗∗∗
(0.169) (0.144) (0.951) (0.751)
 Q3 1.387∗∗∗ 1.249∗∗ 1.628+ 1.457
(0.097) (0.094) (0.410) (0.369)
Constant
 0.037∗∗∗ 0.030∗∗∗ 0.024∗∗∗ 0.023∗∗∗ 0.004∗∗∗ 0.003∗∗∗ 0.001∗∗∗ 0.001∗∗∗
(0.005)
 (0.003)
 (0.003)
 (0.003)
 (0.001)
 (0.001)
 (0.000)
 (0.000)
n 7857 7857 7857 7857 7857 7857 7857 7857
Open in a new tab

Exponentiated coefficients; Standard errors in parentheses Note: Sample restricted to those age 25+.

+ p < 0.10, ∗ p < 0.05, ∗∗ p < 0.01, ∗∗∗ p < 0.001.

4. Results

4.1. Descriptive statistics

Table 1 describes the entire sample, collectively and by race. Overall, 16.2 percent of respondents experienced at least one household member death. The percentage of Black individuals who have experienced at least one household member death (19.7 percent) is greater than White individuals (15.4 percent). Exposure to at least two or more household member deaths is a much less common occurrence (2.9 percent), but Black individuals experienced this at considerably higher rates (Black: 4.5 percent vs. White: 2.5 percent). Additionally, compared to White PSID sample members, Black respondents are, on average, more likely to be in the bottom two wealth quartiles and have lower levels of educational attainment. Black PSID respondents are also younger and more likely to be female.

4.2. Regression results

First, I examine whether racial identification is associated with experiencing the loss of a household member (Table 2). Model 1 presents results for the entire sample. The results show Black people are 1.6 times more likely than White individuals to be exposed to at least one household member death, controlling for age and gender. Furthermore, Black people are 2.5 times more likely than White individuals to be exposed to at least two deaths (Model 2).

Next, Table 3 presents results for the five lifecourse stages, first for at least one death (Models 1–5) and then for at least two deaths (Models 6–10). Black children in early life (ages <18) are 2.1 times more likely than White children to be exposed to at least one household member death (Model 1). On the other hand, exposure to at least two household member deaths is a rare event in early life, and there is no statistically significant difference between Black and White children (Model 6). Black people in the transition to adulthood lifecourse stage are 1.7 times more likely than White individuals to have been exposed to at least one death up to that point in their lifetime (Model 2). In terms of exposure to at least two deaths, the coefficient is large in the transition to adulthood model, but only marginally significant (Model 7, Relative Risk [RR]: 3.818, p < 0.10). For midlife and later life, a sizeable Black–White disparity also exists. Black people are 2.1 and 1.7 times more likely to be exposed to at least one death by midlife and by later life respectively (Models 3–4). For at least two deaths, they are 2.8 and 3.2 times more likely (Models 8–9). The risk of exposure does not significantly differ by race for at least one death (Model 5), but is significant for at least two deaths (Model 10, RR: 1.924, p < 0.05) among older adults.

Finally, in Table 4, I return to the original models presented in Table 2 but adjust for SES. The sample is restricted to adults aged 25 and older due to the inclusion of educational attainment. Model 1 shows that among those 25 and older, Black adults are 1.609 times more likely than White adults to be exposed to at least one household member death, compared to 1.632 in Table 2. Adjusting independently for educational attainment (Model 2) and wealth (Model 3) leads to a slight attenuation in the association between race and experiencing at least one household member death. In Model 4, Black adults are still 1.4 times more likely to experience at least one household member death when both education and wealth are included in the same model.

Models 2–4 show that both measures of SES are significantly associated with experiencing at least one death. In Model 4, compared to those with a college degree or more, individuals with less than a high school degree (RR = 1.825, p < 0.001), high school degree (RR = 1.593, p < 0.001), or some college (RR = 1.340, p < 0.01) are more likely to have experienced at least one death. Lower levels of wealth are also associated with an increased risk of experiencing at least one death. Compared to being in the highest wealth quartile, those in Q1 (RR = 1.475, p < 0.001), Q2 (RR = 1.361, p < 0.01), and Q3 (RR = 1.249, p < 0.01) are more likely to have experienced at least one death.

Models 5–8 demonstrate that the association between race and experiencing at least two household member deaths is greatly attenuated by the inclusion of SES. Among those aged 25 and older, Black adults are 2.4 times more likely than White adults to have experienced at least two deaths (compared to 2.5 in Table 2). This association is reduced considerably but remains significant after the inclusion of both educational attainment and wealth in Model 8 (RR = 1.633, p < 0.05). Model 8 shows that compared to those with a college degree or more, individuals with less than a high school degree (RR = 2.083, p < 0.001), high school degree (RR = 1.517 p < 0.1), or some college (RR = 1.448, p < 0.1) are also more likely to have experienced at least two household member deaths. Those in the first wealth quartile are 3.2 times more likely than those in the top wealth quartile to experience at least two household member deaths, while those in the second quartile are 2.7 times more likely than those in the top quartile. There is no statistically significant difference between the third and top quartile.

5. Discussion

Although research has demonstrated that Black individuals are significantly more likely to experience the deaths of multiple family members than White people (Umberson et al., 2017), little work has considered the role of residential proximity in exposure to death. This study leveraged the longest-running household panel study in the world to examine racial disparities in exposure to household member death.

The first key finding was that there are marked racial disparities in experiencing household member deaths. Almost one in five Black people have experienced at least one death, compared to about one in seven White individuals. While experiencing the death of multiple household members was relatively rare, it was much more prevalent for Black people (4.5 percent versus 2.5 percent). Even after controlling for relevant factors, I find Black people are over 1.6 times more likely to have experienced at least one household member death and over 2.5 times more likely to have experienced at least two household member deaths. The disproportionate risk of exposure to the death of a household member for Black individuals may therefore not only mirror well-documented racial disparities in mortality but also racial differences in household structure (Pilkauskas et al., 2014).

It is essential to note that coresidence is not by itself inherently detrimental; instead, I argue the disparity arises due to the dual inequity of higher mortality rates and higher rates of coresidence. In fact, research suggests coresidence can be associated with numerous benefits for household members, especially Black children (Mollborn et al., 2011, 2012). Rather, the combination of higher rates of coresidence and differential mortality rates, which reflect long-standing historical and structural racism against Black people (Williams et al., 2019), appears to put Black people at heightened risk of exposure to deaths in their households. Although coresidence has many beneficial aspects, such as social and instrumental support, when coupled with lower life expectancy, it may perpetuate cumulative disadvantage through the stressor of bereavement. The results from this study thus demonstrate the need to not only examine racial disparities in exposure to death but also consider the proximity of these deaths.

The second key finding was that racial disparities in the likelihood of experiencing the loss of household members were significant for most lifecourse stages. In early life through later life, Black individuals were more likely to be exposed to at least one household member death. These results are consistent with past work showing large disparities earlier in the lifecourse when considering Black-White disparities in exposure to family member death (Donnelly et al., 2023, Umberson et al., 2017). The elevated exposure to both close family members and household member deaths may reflect that despite the rising prevalence of coresidence households, the majority of Americans do not coreside (Cross, 2018). This study therefore extends prior work by capturing individuals outside of the idealized nuclear family model, which is important given the higher rate of coresidence among Black households compared to White households.

Disparities in exposure to at least one household member death early in the lifecourse are consequential because the death of key figures early in the lifecourse is particularly detrimental (Berman et al., 2015; Stroebe et al., 2007) and may be an important source of disadvantage that accumulates over the lifecourse (Donnelly et al., 2020; Umberson, 2017). Much of previous work has focused on the impact of early parental death, showing a negative association with outcomes such as household income (Mitchell et al., 2017) and educational attainment (Patterson et al., 2020, Thyden et al., 2020, Thyden et al., 2023). Disparities in exposure to close proximity deaths during periods when educational and career trajectories are being established, such as during the transition to adulthood and midlife, represent a previously unidentified potential driver of broader Black–White inequities. This is especially the case given the unique support that household members are more readily able to provide due to proximity (Daw et al., 2019; Ruggles, 1994). My finding suggests the existence of an additional, unstudied source of loss with potential long-term consequences. Further research is necessary to investigate how these household member deaths may uniquely contribute to cumulative disadvantage.

The results show that for at least one death, there was not a significant racial disparity among older adults. Similarly, Umberson et al. (2017) found racial disparities in exposure to primary kin deaths narrow or disappear in later life stages. A reason for these disparities narrowing may be selection regarding who survives to these older ages. Given Black individuals’ lower life expectancy, they are less likely to live to age 65. Relatedly, those exposed to more deaths may be less likely to be observed, especially given that exposure to the death of close family is associated with increased risk of mortality (Donnelly et al., 2020; Stroebe et al., 2007). On the other hand, Black older adults were almost two times more likely than White older adults to experience multiple household member deaths. Future work should investigate the mechanisms undergirding why this population is at an elevated risk of experiencing multiple household member deaths.

A third key finding was that socioeconomic status attenuates, but does not fully explain Black-White inequities in experiencing household member deaths. It is crucial to recognize that the vast Black-White SES disparities are themselves a product of historical and contemporary racism and discrimination (Williams et al., 2019). Notably, there are considerable Black-White gaps in educational attainment; furthermore, only 5 percent of Black individuals were in the top category of wealth, compared to 33 percent of White individuals, reflecting the well-documented Black–White gap in wealth (Killewald et al., 2017). Both lower levels of educational attainment and wealth were associated with increased risk of experiencing a household member death. Given that coresident households are often formed to pool SES resources, those with higher SES may have the resources needed to help individuals who would have otherwise become coresiders remain in their own homes, thereby lessening the risk of exposure to a household member death. Taken together, the results suggest that to mitigate racial disparities in experiencing household member deaths, disparities in both life expectancy and SES will need to be addressed.

5.1. Limitations

This study has several limitations. First, because PSID does not track all household members, especially short-term ones, individuals were only observed in the household if they were living in the household at the time of the survey. Therefore, the results are likely a conservative estimate of the number of household member deaths. Individuals who are temporary household residents are likely not captured in the PSID. For example, an individual who moved into a household in 2016 between the 2015 and 2017 waves and died before the 2017 survey is not likely to have been recorded. However, the duration of exposure is important given that the deaths of individuals with whom the focal person has resided for longer periods (i.e., more than two years) may be more meaningful due to emotional and instrumental bonds. Relatedly, given the sample was from 2019, individuals over 51 years old would not have been observed in the PSID throughout their entire life because the PSID did not begin until 1968. Therefore, the estimates of the number of deaths individuals aged 51 and over experienced is a conservative measure of lifetime exposure.

Another limitation was the issue of selection. Although participants were followed retrospectively from the 2019 wave, individuals who died before the 2019 wave were omitted. Given that Black people have higher rates of mortality, they were more likely to have died before the 2019 wave and not be included in the analysis. However, the PSID survey weights account for attrition due to mortality and other causes of loss to follow-up. Lastly, while this study estimates racial disparities in experiencing a household member death by a certain lifecourse stage, it does not pinpoint exactly when in the lifecourse the death was experienced. Future research should investigate the exact timing of household member deaths and how this impacts the socioeconomic and health trajectories of the bereaved.

6. Conclusion

Despite stark racial differences in coresidence, in most studies on the relationship between racial disparities in exposure to death, researchers have not assessed the role of household member death. By introducing the role of the household to this literature, I provide a new key conceptual framework to understand the sources of racial disparities in exposure to death. Such a reframe provides a critical and urgent lens to this emerging body of research to understand not only the magnitude of racial disparities in exposure to death, but also the location of these exposures. I demonstrate how two interlocking contributors to racial disparities in exposure to death are themselves rooted in longstanding racial inequities. It is imperative that researchers incorporate the household as a critical source of exposure to death. If exposure to death is an important, understudied source of Black disadvantage throughout the lifecourse, we must understand the prevalence of all meaningful and, in particular, proximal sources of death.

Funding sources

Research reported in this publication was supported in part by the National Institute On Minority Health And Health Disparities of the National Institutes of Health under Award Number K01MD016170. The content is solely the responsibility of the author and does not necessarily represent the official views of the National Institutes of Health. This research was also supported in part by a grant from the Russell Sage Foundation. The research reported in the article was made possible in part by a grant from the Spencer Foundation (#202100298). The views expressed are those of the author and do not necessarily reflect the views of the Spencer Foundation. The collection of data used in this study was partly supported by the National Institutes of Health under grant number R01 HD069609 and R01 AG040213, and the National Science Foundation under award numbers SES 1157698 and 1623684. This research was also supported in part by an NICHD center grant to the Population Studies Center at the University of Michigan (P2CHD041028).

CRediT authorship contribution statement

Angela R. Dixon: Writing – review & editing, Writing – original draft, Funding acquisition, Formal analysis, Conceptualization.

Declaration of competing interest

The author declares to have no conflicting interests relevant to the research presented in this article.

Footnotes

1

The early life stage includes only children aged 2–17. Children under 2 were not present in the prior wave (i.e., 2017) and thus could not have been assigned a household member death based on how the measure was constructed.

Data availability

All data are available through the Inter-university Consortium for Political and Social Research (ICPSR).

References

  1. Amorim M. Are grandparents a blessing or a burden? Multigenerational coresidence and child-related spending. Social Science Research. 2019;80:132–144. doi: 10.1016/j.ssresearch.2019.02.002. [DOI] [PubMed] [Google Scholar]
  2. Bengtson V.L. Beyond the nuclear family: The increasing importance of multigenerational bonds. Journal of Marriage and Family. 2001;63(1):1–16. doi: 10.1111/j.1741-3737.2001.00001.x. [DOI] [Google Scholar]
  3. Berman L.R., Snow R.C., Moorman J.D., Policicchio D., Geronimus A.T., Padilla M.B. Parental loss and residential instability: The impact on young women from low-income households in detroit. Journal of Child and Family Studies. 2015;24(2):416–426. doi: 10.1007/s10826-013-9852-9. [DOI] [Google Scholar]
  4. Cha H., Thomas P.A., Umberson D. Sibling deaths, racial/ethnic disadvantage, and dementia in later life. The Journals of Gerontology: Serie Bibliographique. 2022;77(8):1539–1549. doi: 10.1093/geronb/gbab202. [DOI] [PMC free article] [PubMed] [Google Scholar]
  5. Cross C.J. Extended family households among children in the United States: Differences by race/ethnicity and socio-economic status. Population Studies. 2018;72(2):235–251. doi: 10.1080/00324728.2018.1468476. [DOI] [PubMed] [Google Scholar]
  6. Daw J., Verdery A.M., Margolis R. Kin count(s): Educational and racial differences in extended kinship in the United States: Kin count(s) Population and Development Review. 2016;42(3):491–517. doi: 10.1111/j.1728-4457.2016.00150.x. [DOI] [PMC free article] [PubMed] [Google Scholar]
  7. Daw J., Verdery A.M., Patterson S.E. Beyond household walls: The spatial structure of American extended kinship networks. Mathematical Population Studies. 2019;26(4):208–237. doi: 10.1080/08898480.2019.1592637. [DOI] [PMC free article] [PubMed] [Google Scholar]
  8. Donnelly R., Cha H., Umberson D. Multiple family member deaths and cardiometabolic health among black and white older adults. Journal of Health and Social Behavior. 2022;63(4):610–625. doi: 10.1177/00221465221114485. [DOI] [PMC free article] [PubMed] [Google Scholar]
  9. Donnelly R., Garcia M.A., Cha H., Hummer R.A., Umberson D. Exposure to family member deaths across the life course for hispanic individuals. Demography. 2023;10604036 doi: 10.1215/00703370-10604036. [DOI] [PMC free article] [PubMed] [Google Scholar]
  10. Donnelly R., Lin Z., Umberson D. Parental death across the life course, social isolation, and health in later life: Racial/ethnic disadvantage in the U.S. Social Forces. 2023;102(2):586–608. doi: 10.1093/sf/soad027. [DOI] [PMC free article] [PubMed] [Google Scholar]
  11. Donnelly R., Umberson D., Hummer R.A., Garcia M.A. Race, death of a child, and mortality risk among aging parents in the United States. Social Science & Medicine. 2020;249 doi: 10.1016/j.socscimed.2020.112853. [DOI] [PMC free article] [PubMed] [Google Scholar]
  12. Donnelly R., Umberson D., Pudrovska T. Family member death and subjective life expectancy among black and white older adults. Journal of Aging and Health. 2018 doi: 10.1177/0898264318809798. [DOI] [PMC free article] [PubMed] [Google Scholar]
  13. Dunifon R.E., Ziol-Guest K.M., Kopko K. Grandparent coresidence and family well-being: Implications for research and policy. The Annals of the American Academy of Political and Social Science. 2014;654(1):110–126. doi: 10.1177/0002716214526530. [DOI] [Google Scholar]
  14. Fry R. Driven in part by parents living with their adult children. Pew Research Center; 2018. More adults now share their living space.https://www.pewresearch.org/fact-tank/2018/01/31/more-adults-now-share-their-living-space-driven-in-part-by-parents-living-with-their-adult-children/ [Google Scholar]
  15. Fry R. Pew Research Center; 2019. The number of people in the average U.S. Household is going up for the first time in over 160 years.https://www.pewresearch.org/fact-tank/2019/10/01/the-number-of-people-in-the-average-u-s-household-is-going-up-for-the-first-time-in-over-160-years/ Retrieved June 2, 2022. [Google Scholar]
  16. Gallis J.A., Turner E.L. Relative measures of association for binary outcomes: Challenges and recommendations for the global health researcher. Annals of Global Health. 2019;85(1):137. doi: 10.5334/aogh.2581. [DOI] [PMC free article] [PubMed] [Google Scholar]
  17. Geronimus A.T. The weathering hypothesis and the health of African-American women and infants: Evidence and speculations. Ethnicity & Disease. 1992;2(3):207–221. [PubMed] [Google Scholar]
  18. Hummer R.A., Hamilton E.R. Vol. 5. Univ of California Press; 2019. (Population health in America). [Google Scholar]
  19. Kidman R., Margolis R., Smith-Greenaway E., Verdery A.M. Estimates and projections of COVID-19 and parental death in the US. JAMA Pediatrics. 2021;175(7):745–746. doi: 10.1001/jamapediatrics.2021.0161. [DOI] [PMC free article] [PubMed] [Google Scholar]
  20. Killewald A., Pfeffer F.T., Schachner J.N. Wealth inequality and accumulation. Annual Review of Sociology. 2017;43(1):379–404. doi: 10.1146/annurev-soc-060116-053331. [DOI] [PMC free article] [PubMed] [Google Scholar]
  21. Klinenberg E. Social isolation, loneliness, and living alone: Identifying the risks for public health. American Journal of Public Health. 2016;106(5):786. doi: 10.2105/AJPH.2016.303166. [DOI] [PMC free article] [PubMed] [Google Scholar]
  22. Liu H., Lin Z., Umberson D. In: Kelley J., editor. Vol. 77. 2022. Parental death and cognitive impairment: An examination by gender and race–ethnicity; pp. 1164–1176. (The journals of gerontology: Series B). 6. [DOI] [PMC free article] [PubMed] [Google Scholar]
  23. Mitchell C., McLanahan S., Schneper L., Garfinkel I., Brooks-Gunn J., Notterman D. Father loss and child telomere length. Pediatrics. 2017;140(2) doi: 10.1542/peds.2016-3245. [DOI] [PMC free article] [PubMed] [Google Scholar]
  24. Mollborn S., Fomby P., Dennis J.A. Who matters for children's early development? Race/Ethnicity and extended household structures in the United States. Child Indicators Research. 2011;4(3):389–411. doi: 10.1007/s12187-010-9090-2. [DOI] [PMC free article] [PubMed] [Google Scholar]
  25. Mollborn S., Fomby P., Dennis J.A. Extended household transitions, race/ethnicity, and early childhood cognitive outcomes. Social Science Research. 2012;41(5):1152–1165. doi: 10.1016/j.ssresearch.2012.04.002. [DOI] [PMC free article] [PubMed] [Google Scholar]
  26. Oxford University Press . Oxford Learner’s Dictionaries; 2023. Bereavement.https://www.oxfordlearnersdictionaries.com/us/definition/english/bereavement [Google Scholar]
  27. Panel Study of Income Dynamics . Produced and distributed by the Survey. Research Center, Institute for Social Research, University of Michigan; Ann Arbor, MI: 2021. restricted use dataset. [Google Scholar]
  28. Patterson S.E., Verdery A.M., Daw J. Linked lives and childhood experience of family death on educational attainment. Socius. 2020;6 doi: 10.1177/2378023120975594. [DOI] [PMC free article] [PubMed] [Google Scholar]
  29. Pew Research Center . Pew Research Center’s Social & Demographic Trends Project; 2013. Children living with or being cared for by a grandparent.https://www.pewsocialtrends.org/2013/09/04/children-living-with-or-being-cared-for-by-a-grandparent/ [Google Scholar]
  30. Pilkauskas N.V., Cross C. Beyond the nuclear family: Trends in children living in shared households. Demography. 2018;55(6):2283–2297. doi: 10.1007/s13524-018-0719-y. [DOI] [PubMed] [Google Scholar]
  31. Pilkauskas N.V., Garfinkel I., McLanahan S.S. The prevalence and economic value of doubling up. Demography. 2014;51(5):1667–1676. doi: 10.1007/s13524-014-0327-4. [DOI] [PMC free article] [PubMed] [Google Scholar]
  32. PSID PSID terminology. 2023. https://psidonline.isr.umich.edu/Guide/FAQ.aspx?Type=5
  33. Reis H.T., Sprecher S. Encyclopedia of human relationships. Sage Publications; Thousand Oaks, CA: 2009. Kin relationships; pp. 951–954. [Google Scholar]
  34. Rodriguez J.M., Geronimus A.T., Bound J., Dorling D. Black lives matter: Differential mortality and the racial composition of the U.S. Electorate, 1970–2004. Social Science & Medicine. 2015;136–137:193–199. doi: 10.1016/j.socscimed.2015.04.014. [DOI] [PMC free article] [PubMed] [Google Scholar]
  35. Ruggles S. The transformation of American family structure. The American Historical Review. 1994;99(1):103. doi: 10.2307/2166164. [DOI] [Google Scholar]
  36. Stroebe M., Schut H., Stroebe W. Health outcomes of bereavement. The Lancet. 2007;370(9603):1960–1973. doi: 10.1016/S0140-6736(07)61816-9. [DOI] [PubMed] [Google Scholar]
  37. Thyden N.H., Schmidt N.M., Osypuk T.L. The unequal distribution of sibling and parent deaths by race and its effect on attaining a college degree. Annals of Epidemiology. 2020 doi: 10.1016/j.annepidem.2020.03.002. [DOI] [PMC free article] [PubMed] [Google Scholar]
  38. Thyden N.H., Slaughter-Acey J., Widome R., Warren J.R., Osypuk T.L. Family deaths in the early life course and their association with later educational attainment in a longitudinal cohort study. Social Science & Medicine. 2023;333 doi: 10.1016/j.socscimed.2023.116161. [DOI] [PMC free article] [PubMed] [Google Scholar]
  39. Umberson D. Black deaths matter: Race, relationship loss, and effects on survivors. Journal of Health and Social Behavior. 2017;58(4):405–420. doi: 10.1177/0022146517739317. [DOI] [PMC free article] [PubMed] [Google Scholar]
  40. Umberson D., Donnelly R. Social isolation: An unequally distributed health hazard. Annual Review of Sociology. 2023;49(1):379–399. doi: 10.1146/annurev-soc-031021-012001. [DOI] [PMC free article] [PubMed] [Google Scholar]
  41. Umberson D., Donnelly R., Xu M., Farina M., Garcia M.A. In: The journals of gerontology: Series B gbz154. Carr D., editor. 2019. Death of a child prior to midlife, dementia risk, and racial disparities. [DOI] [PMC free article] [PubMed] [Google Scholar]
  42. Umberson D., Montez J.K. Social relationships and health: A flashpoint for health policy. Journal of Health and Social Behavior. 2010;51(1_suppl):S54–S66. doi: 10.1177/0022146510383501. [DOI] [PMC free article] [PubMed] [Google Scholar]
  43. Umberson D., Olson J.S., Crosnoe R., Liu H., Pudrovska T., Donnelly R. Death of family members as an overlooked source of racial disadvantage in the United States. Proceedings of the National Academy of Sciences. 2017;114(5):915–920. doi: 10.1073/pnas.1605599114. [DOI] [PMC free article] [PubMed] [Google Scholar]
  44. Verdery A.M., Rachel M. Projections of white and black older adults without living kin in the United States, 2015 to 2060. Proceedings of the National Academy of Sciences. 2017;114(42):11109–11114. doi: 10.1073/pnas.1710341114. [DOI] [PMC free article] [PubMed] [Google Scholar]
  45. Verdery A.M., Smith-Greenaway E., Margolis R., Daw J. Tracking the reach of COVID-19 kin loss with a bereavement multiplier applied to the United States. Proceedings of the National Academy of Sciences. 2020;117(30):17695–17701. doi: 10.1073/pnas.2007476117. [DOI] [PMC free article] [PubMed] [Google Scholar]
  46. Williams D.R., Lawrence J.A., Davis B.A. Racism and health: Evidence and needed research. Annual Review of Public Health. 2019;40(1) doi: 10.1146/annurev-publhealth-040218-043750. [DOI] [PMC free article] [PubMed] [Google Scholar]
  47. Zou G. A modified Poisson regression approach to prospective studies with binary data. American Journal of Epidemiology. 2004;159(7):702–706. doi: 10.1093/aje/kwh090. [DOI] [PubMed] [Google Scholar]

Associated Data

This section collects any data citations, data availability statements, or supplementary materials included in this article.

Data Availability Statement

All data are available through the Inter-university Consortium for Political and Social Research (ICPSR).

Articles from SSM - Population Health are provided here courtesy of Elsevier

RESOURCES