Abstract
Objectives:
To describe unclaimed deaths nationwide and to examine contextual factors associated with higher rates of these types of interments.
Methods:
Unclaimed death records from 2009 to 2021 were collected by a combination of direct outreach and obtaining data from publicly available websites. Data were summarized by county-year and linked to American Community Survey demographic variables, as well as markers of social isolation and economic disadvantage.
Results:
The underlying individual-level dataset included 42,942 decedents. Unclaimed deaths in the United States from 2009 to 2021, based on data from more than 248 counties that provided records, are increasing over time. This population appears to be younger at death, more male, and more likely to be of Black race than decedents in the general population. Several county-level markers of social isolation and economic deprivation were correlates of unclaimed deaths.
Conclusions:
This is the first systematic examination of unclaimed deaths at the county level and offers a novel way to understand marginalized populations, such as the unhoused and others living in extreme social or economic deprivation. Unclaimed death rates may also provide an objective measure of social support that can be measured over time.
1. Introduction
Burial sites for persons who are unclaimed at death, commonly known as potter’s fields, have existed since before the founding of the United States (US). Burial or cremation for the unclaimed are funded by local, state, or the federal government when family members are unavailable, unable, or unwilling to pay for a traditional burial. (Unclaimed deaths are separate from missing and unidentified persons because the identity of the person is always known.) These types of interments are largely understudied. Previous work on unclaimed deaths in the US has focused on theoretical analyses of their cultural context (Bowring, 2011; Holleran, 2024; Kenney, 2023; Thompson and Yeung, 2015), been conducted abroad (Allison, 2023; Culler, 2017; Khoo et al., 2023; Steinberg, 2015), or been limited to specific locales, such as Los Angeles (Sohn et al., 2020; Timmermans, 2024) and Indianapolis (Quinet et al., 2016). This is largely because no centralized, nationwide database of unclaimed deaths exists in the US; instead, data are often tracked at the county level, if at all. The definitive number of unclaimed deaths in the US is unknown; however, evidence from Los Angeles (Jordan and Sullivan, 2021; Sohn et al., 2020) and New York City (Heiland et al., 2023) suggests they may represent as many as 1–5% of total deaths each year. There is also some evidence these types of burials may be increasing over time in recent years (Brite et al., 2023). The socio-demographic characteristics and geographic distribution of unclaimed deaths are also under-researched. For example, the two previous epidemiological studies on this topic focused on large metropolitan areas, so the magnitude of unclaimed deaths in rural areas is unknown. However, there is some evidence unclaimed death rates may vary by population density. An unpublished analysis of Hart Island burials in New York City during the wave of the first wave of the COVID-19 pandemic found the least densely populated borough (Staten Island) had lower rates of unclaimed deaths compared to other parts of the city (Heiland et al., 2023). The continuum of urban to rural conditions varies in several aspects that may ultimately lead to an unclaimed death. Social isolation has been found to be greater in urban areas, but populations in more remote rural localities (defined as “non-core”) experience fewer social connections than residents of micropolitan areas (Henning-Smith et al., 2019). Conditions such as extreme poverty, which might influence unclaimed death, may be more prevalent and rising in some types of rural and urban areas compared to suburban places (Thiede et al., 2017).
Moreover, the evidence of racial/ethnic patterning of unclaimed death is limited; the two prior epidemiological studies on this topic found conflicting results in terms of whether Black individuals were at higher risk of becoming an unclaimed death (Quinet et al., 2016; Timmermans, 2024). The role of race/ethnicity may be complex because both Black and Hispanic individuals have greater social ties and kin networks (Jean, 1997), but are also, on average, lower income than White and Asian individuals (Akee et al., 2019). In addition, Black and Hispanic men have the highest rates of incarceration of any demographic group in the United States (Mauer and King, 2007), which can lead them to be both physically and emotionally separated from friends and family for long periods of time.
Despite the scientific gaps described above, emerging evidence suggests unclaimed burials are most common among economically disadvantaged and socially isolated individuals (Brite et al., 2023,; Heiland et al., 2023; Quinet et al., 2016; Sohn et al., 2020), who may be underrepresented in traditional epidemiological or social surveys due to differential attrition by socioeconomic status (Long et al., 2018; Yu et al., 2015). The Indianapolis study is the only unclaimed death study to examine military and housing status; it found 13% of the unclaimed dead were unhoused at the time of death and 16% were veterans (Quinet et al., 2016). Unclaimed deaths may therefore offer a unique measure for understanding mortality patterns in the most vulnerable populations.
This is of public health importance because unclaimed death can be seen as a negative health outcome that may be the result of cumulative disadvantage over the life course (Hawton et al., 2011; Philip et al., 2020; Uchino, 2006). Quantitative research on unclaimed deaths highlights myriad mechanisms through which vulnerability across the life course—poverty, social isolation, veteran’s status, mental health, migration status—places individuals at risk of becoming an unclaimed death (Prickett and Timmermans, 2022). Analysis of health disparities is an important public health objective; however, little is known about disparities in relation to the processes of death. Unclaimed deaths may be seen as a breakdown of social support at both the individual and community level. Social support is made up of tangible support (e.g., financial support for a funeral) and emotional support (e.g., attending a traditional funeral to console friends and family). Social support may be more important in poorer communities as it is a buffer to the effects of poverty (Cohen and McKay, 2020; Takizawa et al., 2006) and lack of social ties has been associated with a range of negative health outcomes and, ultimately, higher mortality (Holt-Lunstad et al., 2010).
More broadly, the study of unclaimed deaths also informs several important public health priorities. For example, in the spring of 2023 Surgeon General Vivek Murthy released an advisory warning of growing loneliness (Jaffe, 2023) and its health effects. In addition, “Deaths of despair,” from economic marginalization and drug and alcohol use, notably have been on the rise, and are distributed unevenly across the US (Case and Deaton, 2015, 2020). Unclaimed deaths may provide an alternative (or supplementary) metric for studying the spatial and temporal patterning of social isolation and economic deprivation. The present study aims to help fill these gaps by providing the first nation-wide analysis of unclaimed deaths. Specifically, we seek to answer: 1. how prevalent is public burial; 2. who is at risk of public burial; and 3. what area-level characteristics are associated with unclaimed deaths? To answer these questions, we conduct descriptive and multivariable analyses of unclaimed deaths over time in a diverse set of counties representing 22% of the U.S. population.
2. Methods and materials
Unclaimed death data were collected in a variety of formats, standardized, and linked with other data sources—notably American Community Survey (ACS) (US Census Bureau) and all-cause mortality data (counts of deaths) in the general population from the Centers for Disease Control (CDC) (Centers for Disease Control and Prevention, 2021)—in order to place unclaimed deaths in context. Ideally, we would have had data for each county, but since that was not available, instead we developed a method to collect data purposively from populous counties and then randomly sample smaller counties for inclusion. A brief overview is provided below. Contextual data—to provide necessary county-level death count denominators and demographic and socio-economic indicators to explain variation in unclaimed deaths—is then described.
2.1. Unclaimed death data collection
Data collection took place between 2021 and 2023. Unclaimed death information was primarily collected via direct outreach to jurisdictions throughout the United States because records are generally, but not always, kept only by county governments. We were unable to contact the more than 3,000 counties in the US due to limited study resources, so we created a sampling frame that included the 200 most populous, as well as a random sample of smaller counties, or when appropriate, states (four states maintain centralized unclaimed death records). We then sent formal requests for unclaimed death records, primarily through the Freedom of Information Act (FOIA); it is notable that most of those requests went unfulfilled. Many counties either did not respond to those requests, required payment for formatting and gathering data, or stipulated they would provide the data at a later date. Additionally, 12 towns and cities provide information on publicly available websites from which we acquired data.
We also obtained data from two other websites: A charity affiliated with Dignity Memorial (which is the largest funeral, cremation, and cemetery services company in the US), which makes public some information on unclaimed veterans (Dignity Memorial). (Because the number of deaths from this source was very small (n = 21), in a sensitivity analysis we removed these observations and find that our results remain unchanged [not shown, but available upon request]). In addition, the U.S. Department of Justice maintains a public registry of deaths associated with missing persons, NamUs, that also contains a smaller section dedicated to the unclaimed (Hanzlick et al., 2011). If data were provided by NamUs and a county, to avoid duplication, only county-reported information was included.
Several states—Ohio, Oregan, and Rhode Island—collect unclaimed death data but do not record county of death or burial and as such as are omitted from this analysis. Although New York City inters unclaimed deaths almost exclusively on Hart Island in Bronx County, the city also provides place of death information, which allows us to assign county (i. e., borough) for each decedent. For all other data, the county of burial was used. In total, we have data from 243 counties (see Appendix 6), from all sources, used in this analysis.
2.2. Analytic population
The current analysis was restricted to 2009–2021 to examine more current predictors of unclaimed death and to better leverage contextual data (see below). One state, Alaska, was excluded due to low number of reported cases (n = 1), and any unclaimed death record with implausible age or death/burial date was also excluded (n = 1,364).
2.3. Contextual data
Population estimates of social isolation (percent separated, living alone, living in group quarters, and never married), economic deprivation (percent below poverty level and unemployed and median income), and various demographic indicators were obtained via the American Community Survey (5-year estimates) by year for each county. In addition, urban-rural continuum was obtained from the National Center for Health Statistics 2013 data, which categorized counties as Large central metro, Large fringe metro, Medium metro, Small metro, Micropolitan, or Non-core (Ingram and Franco, 2014; National Center for Health Statistics, 2022). We used the 5-year ACS data from its inception in 2009 through 2021. Several years of earlier estimates for the 3-year ACS are available, but that series was discontinued in 2013. A 1-year ACS series is available, but it suppresses estimates for smaller counties (population less than 65,000). We have 68 counties where this restriction applies. Consequently, we selected the 5-year ACS over the 1-year series to preserve the maximal number of smaller counties and increase statistical precision overall. While this series was also available for 2022, the mortality data from CDC Wonder were not yet available for 2022.
2.4. Statistical analysis
We estimated OLS regression models and predicted unclaimed death shares using contextual variables as predictors. The unit of analysis in the regression analysis was county-year, where year indicated year of death, if available, or burial year if not. No records in the analytic population were missing county or year. The dependent variable for each model was the Unclaimed Death Share (UDS), defined as the number of unclaimed deaths out of all deaths for that year/county. Multiplied by 100, this represents the share (indicated as a %) of public interment in a given year/county. This allows us to examine unclaimed deaths in the context of all county-level deaths by year. A positive (negative) regression coefficient implies that the covariate is associated with relatively more (fewer) unclaimed deaths in a county-year on average. All models control for state fixed effects to account for (time-invariant) unobserved state-level differences, including differential reporting and unmeasured confounders. They also control for a linear time trend and year fixed effects to account for period effects and whether the unclaimed deaths were recorded exclusively by NamUs rather than directly from the county. (We strongly suspect an undercount in data provided by NamUs because the database is designed for finding missing persons and the unclaimed death portion of it is a consequence of that effort but not its intended purpose, and in a sensitivity analysis [not shown], we found locations where we have NamUs and municipal records, the municipal records always far exceed those of NamUs.) We control for whether data was provided by NamUs in the regression for this reason, but we did not remove NamUs records because they include a much greater geographic distribution, including less populous counties that may not have another way of providing unclaimed death data. The independent variables are primarily generated from Census data from the ACS and are described in Appendix 5. All analyses were conducted in Posit (formerly R Studio).
3. Results
Table 1 describes the number of counties from which we have observations and the completeness of that collection, as well as the percentage of the population coverage and years in which burials are recorded (For brevity, Table 1 is organized by state, but data typically were collected at the county-level.). For example, 127 unclaimed deaths are reported in Alabama counties from 2011 to 2021; these counties represent 12.74% of Alabama’s population (averaging the population over the period in which unclaimed deaths were observed). Owing to the highly decentralized ways in which municipalities collect data (generally not at state-level), it is not possible to fully assess the completeness of the coverage and our estimates are most likely undercounts. For example, in Michigan, data were reported for places that represent more than one-quarter of the state’s population (including Wayne County, where Detroit and its suburbs are located) indicating deaths for the period of 2009–2011 and again 2013–2021. Twelve counties report fully every year and 236 report partially (e.g., counties that did not report consistently from 2009 to 2021 or the death was reported by NamUs). Population-based coverage varies considerably by state.
Table 1.
Coverage of unclaimed death data collection, by state, years, and jurisdictions reported, in this analysis.
| State | Unclaimed deaths | Years1 | Number of counties reporting (% of population with unclaimed death cases reported) | |||
|---|---|---|---|---|---|---|
| Fully reported every year | Partially reported2 | Counties with no data available | % Population coverage3 | |||
| Alabama | 127 | 2011–2021 | 0 | 3 | 64 | 12.74 |
| Alaska | n/a | n/a | 0 | 0 | 32 | n/a |
| Arizona | 903 | 2009–2020 | 0 | 6 | 9 | 31.35 |
| Arkansas | 17 | 2016–2020 | 0 | 3 | 72 | 2.72 |
| California | 17,786 | 2009–2021 | 1 | 16 | 41 | 52.02 |
| Colorado | 5 | 2014, 2016, 2018 | 0 | 2 | 62 | 2.05 |
| Connecticut | 97 | 2015–2019 | 0 | 7 | 1 | 32.96 |
| Delaware | n/a | n/a | 0 | 0 | 3 | n/a |
| District of Columbia | 102 | 2011, 2017, 2019–2021 | 0 | 1 | 0 | 42.23 |
| Florida | 36 | 2011, 2013–2019, 2021 | 0 | 8 | 59 | 2.42 |
| Georgia | 215 | 2009–2021 | 0 | 3 | 156 | 11.14 |
| Hawaii | 346 | 2009–2021 | 0 | 2 | 3 | 76.10 |
| Idaho | 182 | 2009–2021 | 0 | 3 | 41 | 37.71 |
| Illinois | 3035 | 2009–2021 | 1 | 6 | 95 | 51.68 |
| Indiana | 4 | 2015, 2018, 2020 | 0 | 4 | 88 | 0.79 |
| Iowa | 9 | 2010, 2011, 2016–2018, 2020 | 0 | 2 | 97 | 6.30 |
| Kansas | n/a | n/a | 0 | 0 | 105 | n/a |
| Kentucky | 1 | 2012 | 0 | 1 | 119 | 0.09 |
| Louisiana | 13 | 2011–2013, 2015–2018 | 0 | 4 | 60 | 3.93 |
| Maine | n/a | n/a | 0 | 0 | 16 | n/a |
| Maryland | 1 | 2021 | 0 | 1 | 23 | 0.82 |
| Massachusetts | 437 | 2009–2021 | 0 | 12 | 2 | 85.10 |
| Michigan | 296 | 2009–2011, 2013–2021 | 0 | 20 | 63 | 25.94 |
| Minnesota | n/a | n/a | 0 | 0 | 87 | n/a |
| Mississippi | 1 | 2013 | 0 | 1 | 81 | 0.69 |
| Missouri | 7 | 2009–2010, 2016, 2019, 2021 | 0 | 4 | 111 | 4.77 |
| Montana | 74 | 2010–2011, 2014–2021 | 0 | 1 | 55 | 12.53 |
| Nebraska | n/a | n/a | 0 | 0 | 93 | n/a |
| Nevada | 178 | 2013–2021 | 0 | 5 | 12 | 30.82 |
| New Hampshire | 2 | 2009, 2014 | 0 | 2 | 8 | 2.39 |
| New Jersey | 40 | 2010–2013, 2016–2021 | 0 | 13 | 8 | 10.96 |
| New Mexico | 2000 | 2009–2021 | 1 | 4 | 28 | 38.81 |
| New York | 12445 | 2009–2021 | 5 | 4 | 53 | 47.94 |
| North Carolina | 9 | 2017, 2021 | 0 | 3 | 97 | 0.53 |
| North Dakota | n/a | n/a | 0 | 0 | 53 | n/a |
| Ohio | n/a | ** | 0 | 0 | 88 | n/a |
| Oklahoma | 1454 | 2009–2021 | 1 | 21 | 55 | 39.34 |
| Oregon | n/a | ** | 0 | 0 | 36 | n/a |
| Pennsylvania | 360 | 2009–2021 | 2 | 6 | 59 | 14.58 |
| Rhode Island | n/a | ** | 0 | 0 | 5 | n/a |
| South Carolina | 143 | 2009–2021 | 0 | 5 | 41 | 6.75 |
| South Dakota | n/a | n/a | 0 | 0 | 67 | n/a |
| Tennessee | 443 | 2010, 2012–2021 | 0 | 14 | 81 | 14.41 |
| Texas | 469 | 2009–2021 | 0 | 9 | 245 | 23.15 |
| Utah | 19 | 2009, 2016, 2018–2021 | 0 | 1 | 28 | 18.64 |
| Vermont | n/a | n/a | 0 | 0 | 14 | n/a |
| Virginia | n/a | n/a | 0 | 0 | 134 | n/a |
| Washington | 1629 | 2009–2021 | 1 | 14 | 24 | 44.30 |
| West Virginia | 43 | 2009–2015 | 0 | 18 | 37 | 12.56 |
| Wisconsin | 11 | 2010, 2014, 2017–2020 | 0 | 6 | 66 | 3.39 |
| Wyoming | 2 | 2011, 2016 | 0 | 1 | 22 | 1.16 |
Years represent death year if available, otherwise burial year.
Partially reported refers to counties that didn’t report every year or for which data came from NamUS or other source that doesn’t report every unclaimed death.
Population percentage = sum of population in all counties, all years reporting in 2009–2020 divided by sum of total population in all counties, all years 2009–2021.
Data available but only at the state-level. Italics indicate the deaths are observed from NamUs only. Data collection as of January 2024; data collection is on-going.
The individual-level data underlying the county-level analysis included 42,942 unclaimed deaths. Missing information on individuals’ characteristics is common, with 30% of decedents missing information about race, 23% missing gender, and 17% missing age (Table 2). (This does not affect the rate analysis below, which only uses information on how many unclaimed deaths occurred in a year and county.) Among those for whom information is provided, the population of unclaimed deaths appears to be younger and more male than decedents in the general population: the median age of unclaimed death is 65, and there are more than twice as many males than females (72% vs. 28%). Moreover, when examining cases for which race is known, Black individuals make up approximately one-fifth of the reported unclaimed death population (19%), despite the fact they only make up about 13% of the general population the United States.
Table 2.
Demographic attributes of individuals in the unclaimed death population, 2009–2021.
| Characteristic | Count (%) or Median (IQR) | Percent among complete cases only |
|---|---|---|
| Race/Ethnicity | ||
| American Native or Alaskan Native | 107 (0.2%) | 0.4% |
| Asian | 623 (1.5%) | 2.1% |
| Black | 5,661 (13%) | 18.9% |
| Hispanic or Latino | 573 (1.3%) | 1.9% |
| Other | 192 (0.4%) | 0.1% |
| White | 22,773 (53%) | 76.1% |
| Unknown or unreported | 13,013 (30%) | |
| Age (of decedent) | ||
| Median (IQR) | 65 (56, 74) | |
| Mean (SD) | 63.55 (17.3) | |
| Unknown or unreported | 7146 (16.6%) | |
| Sex | ||
| Female | 9,176 (21%) | 27.70% |
| Male | 23,973 (56%) | 72.30% |
| Unknown or unreported | 9,793 (23%) | |
| Year | ||
| 2009 | 1,104 (2.8%) | |
| 2010 | 1,092 (2.8%) | |
| 2011 | 1,547 (3.9%) | |
| 2012 | 2,870 (7.3%) | |
| 2013 | 3,390 (8.6%) | |
| 2014 | 3,463 (8.8%) | |
| 2015 | 2,981 (7.6%) | |
| 2016 | 3,553 (9.0%) | |
| 2017 | 3,885 (9.9%) | |
| 2018 | 4,148 (11%) | |
| 2019 | 4,678 (12%) | |
| 2020 | 6,666 (16%) | |
| 2021 | 3,565 (8.3%) | |
| Total | 42,942 (100%) |
NB: All information was derived from the burial record. Race and ethnicity are combined here. Individuals who were noted on the burial record as either Hispanic Black or Hispanic White are indicated here as Hispanic not Black or White.
The final dataset included 887 county-years. Fig. 1 shows the Unclaimed Death Share (UDS) in % by year for the 13 most populous counties (all have population of 1 million persons or more in 2020) in the data collection. Dotted lines indicate the data were reported by NamUs only. The overarching trend of observed unclaimed death ratios is upward; with a notable increase in 2020, the beginning of the COVID-19 pandemic. There is much variation between counties: among the most populated counties, Bronx, New York, and Los Angeles Counties had the highest UDS, while Cook County had the lowest.
Fig. 1.

Unclaimed death share by 13 most-populous counties from 2009 to 2021*.
NB: Figure includes counties that had a population of 1 million or more persons in 2021 and at least 10 unclaimed deaths. (Counties that reported through NamUS only are indicated in dashed lines; solid lines represent reporting directly from the municipalities.) Not all counties reported during entire time period.
Table 3 presents county-level correlates of UDS. Among county-provided data only, unclaimed deaths made up approximately 1.2% of all deaths throughout the entire study period, but the UDS was not consistent across geography or time. (When including all reported counties, including those in which unclaimed deaths were only reported through NamUs, this rate is much lower likely reflecting an undercount due to the nature of the NamUs collection (Hanzlick et al., 2011).) Models 1 and 2 focus on demographic correlates. Model 3 is a saturated model that includes all demographic variables and two sets of predictors related to two possible mechanisms for unclaimed deaths: social isolation and economic deprivation.
Table 3.
OLS regression coefficients predicting unclaimed death share in the United States (with descriptive statistics) 2009–2021.
| Characteristic | Median or proportions (IQR); or count (%) | Model 1: Basic (Exogenous) Demographics | Model 2: All Demographics | Model 3: Saturated | |||
|---|---|---|---|---|---|---|---|
| Beta | 95% CI | Beta | 95% CI | Beta | 95% CI | ||
| Basic Demographic | |||||||
| Race (mutually-exclusive categories) | |||||||
| Percent Black (single-race) | 5 (2, 13) | −0.002 | −0.006, 0.003 | −0.004 | −0.009, 0.001 | −0.005 | −0.012, 0.003 |
| Percent Asian/Pacific Islander (single-race) | 3.3 (1.4, 6.9) | −0.017 | −0.026, −0.008 | 0.001 | −0.014, 0.015 | 0 | −0.014, 0.015 |
| Percent Native American (single-race) | 0.38 (0.21, 0.97) | 0.01 | −0.018, −0.001 | −0.014 | −0.024, −0.005 | −0.008 | −0.020, 0.003 |
| Percent other race (single-race) | 3.2 (1.3, 8.7) | 0.063 | 0.053, 0.073 | 0.057 | 0.047, 0.067 | 0.053 | 0.041, 0.065 |
| Percent two or more races | 3.18 (2.16, 4.71) | −0.004 | −0.020, 0.012 | −0.01 | −0.026, 0.006 | −0.007 | −0.023, 0.008 |
| Percent Hispanic (any race) | 11 (5, 25) | 0.001 | −0.004, 0.005 | 0.002 | −0.006, 0.009 | 0.002 | −0.005, 0.010 |
| Median age | 37.7 (35.1, 40.3) | 0.033 | 0.012, 0.053 | 0.032 | 0.010, 0.054 | 0.05 | 0.016, 0.085 |
| Percent over 65 | 14.1 (12.1, 16.5) | −0.055 | −0.079, −0.030 | −0.065 | −0.091, −0.039 | −0.08 | −0.108, −0.051 |
| Percent under 18 | 22.9 (21.0, 24.9) | −0.089 | −0.105, −0.073 | −0.053 | −0.079, −0.027 | 0.014 | −0.035, 0.064 |
| Additional Demographic | |||||||
| Percent foreign born | 11 (5, 21) | −0.024 | −0.044, −0.004 | −0.008 | −0.028, 0.013 | ||
| Percent of households with no children under age 18 | 26.0 (23.0, 28.1) | 0.044 | 0.024, 0.064 | 0.079 | 0.025, 0.133 | ||
| Percent moved from a different state | 2.06 (1.39, 3.25) | 0.024 | −0.019, 0.066 | 0.013 | −0.029, 0.056 | ||
| Percent limited English | 7 (3, 14) | 0.024 | −0.001, 0.049 | 0.015 | −0.011, 0.041 | ||
| Percent less than high school diploma | 7.20 (5.60, 9.40) | 0.034 | 0.012, 0.056 | 0.011 | −0.016, 0.039 | ||
| Urban status1 | |||||||
| Large central metro | 241 (27%) | – | – | – | – | ||
| Large fringe metro | 195 (22%) | 0.073 | −0.065, 0.211 | −0.004 | −0.139, 0.131 | ||
| Medium metro | 247 (28%) | −0.012 | −0.153, 0.130 | −0.108 | −0.248, 0.032 | ||
| Small metro | 88 (9.9%) | −0.256 | −0.432, −0.081 | −0.444 | −0.620, −0.268 | ||
| Micropolitan | 82 (9.2%) | 0.038 | −0.143, 0.218 | −0.198 | −0.387, −0.008 | ||
| Non-core | 34 (3.8%) | 0.171 | −0.068, 0.409 | −0.122 | −0.367, 0.122 | ||
| Social Isolation | |||||||
| Percent separated | 1.88 (1.51, 2.50) | 0.004 | −0.106, 0.114 | ||||
| Percent living alone | 13.03 (10.82, 14.86) | 0.019 | −0.029, 0.068 | ||||
| Percent living in group quarters | 2.15 (1.44, 3.13) | 0.096 | 0.071, 0.120 | ||||
| Percent never married | 33 (29, 36) | −0.028 | −0.051, −0.004 | ||||
| Economic Deprivation | |||||||
| Percent below poverty level | 14.3 (10.5, 17.4) | 0.023 | 0.006, 0.040 | ||||
| Median income | 5.59 (4.73, 6.93) | 0.039 | −0.014, 0.092 | ||||
| Percent unemployed | 1.27 (0.93, 1.79) | 0.032 | −0.030, 0.093 | ||||
| Other Control Variables | |||||||
| Year of death (or burial) | n/a | 0.024 | 0.013, 0.036 | 0.037 | 0.024, 0.050 | 0.048 | 0.033, 0.063 |
| Year 2020 (pandemic) | 93 (10%) | 0.086 | −0.019, 0.190 | 0.067 | −0.034, 0.169 | 0.035 | −0.064, 0.134 |
| NamUs observation1 (yes = 1) | 648 (73%) | −0.469 | −0.569, −0.370 | −0.502 | −0.608, −0.396 | −0.509 | −0.612, −0.407 |
| Model fit statistics (R2 adjusted) | 0.608 | 0.635 | 0.663 | ||||
| Mean Unclaimed Death Share | 1.2 (data provided by counties only) | ||||||
NB: The Public Burial Death Ratio is calculated as follows: county-year specific unclaimed deaths/all deaths*100; bold font indicates statistical significance at the 0.05 level. All models are adjusted for state and year fixed effects.
The counts represent county-years.
As indicated by the adjusted R2 adjusted value of Model 1, basic demographics—together with state and year fixed effects—explain 60.8% of the observed differences in the share of unclaimed death across space and time. Looking at individual coefficients, Model 1 shows that the share of unclaimed deaths (UDS) increases over time and is greater in places where a higher percentage of the population identify as Other Race (single-race). Percent Native American (single-race) is associated with lower UDS. Percent Asian (single-race) is also negatively associated with UDS, but this association is no longer statistically significant when controlling for additional demographic factors in Model 2. Higher median age is associated with a greater UDS, while percent over 65 and percent under 18 are associated with lower UDS.
Model 2 adds additional factors related to individual and household demographic characteristics. This further improves model fit (adjusted R2 = 0.635 in Model 2). A decedent is more likely to have an unclaimed death in places with a greater percentage of households without children and those without a high school degree. The latter is consistent with the hypothesis that unclaimed death is more common among socio-economically disadvantaged and vulnerable populations. Unclaimed death is less common in counties with a greater share of foreign-born population, as well as in small metropolitan areas.
Model 3 adds measures related to social isolation and economic deprivation to the specification with demographic variables. This further improves model fit (adjusted R2 = 0.663). The results point to social isolation and economic deprivation as important contributing factors to unclaimed death. All else equal, in places where a greater share of the population resides in groups quarters or lives in poverty, the share of unclaimed deaths is greater. The estimated coefficient on living in group quarters of 0.096 (95% CI: 0.071, 0.120) implies that a 10-point increase in a county’s percentage living in group quarters is associated with a 0.96 points higher share of unclaimed deaths. Considering data provided by counties only—where the share of unclaimed deaths out of all deaths (UDS) is about 1 percent—this represents a 96% increase. The estimated coefficient on living below the poverty level is 0.023 (95% CI: 0.006, 0.040), which implies that a 10-points greater share living in poverty is associated with a 0.23-points higher UDS.
4. Discussion
The present analysis offered the first examination of unclaimed death across the US. Within the extent of our study counties, our results suggest that the prevalence of unclaimed deaths is increasing over time. We found observed unclaimed death rates are increasing nationwide, particularly since 2016, and not surprisingly highest in 2020 during the height of the COVID-19 pandemic. The prevalence also varies by geographic region. For example, the Bronx, New York (i.e., Manhattan), and Los Angeles counties had relatively high rates of more than 2% of all deaths in most study years (and the Bronx and New York counties are over 4% in 2020). Finally, we found several county-level markers of vulnerability were correlates of unclaimed death. Specifically, a 10-point increase in percent living in group quarters and percent poverty at the county level were associated with a 0.96- and 0.23-point increase in the share of unclaimed deaths respectively. Although these effect sizes are relatively modest, it is important to note that our fully adjusted model, that includes demographics and time trends, explains a substantial percentage of the UDS (adjusted R2 = 0.663).
4.1. Results in context
These results are in line with the emerging literature on unclaimed deaths. For example, similar to this study, an analysis restricted to coroner cases in Indianapolis resulted in approximately one percent unclaimed deaths, and a slightly higher share were found in a study of all Los Angeles deaths (Quinet et al., 2016; Sohn et al., 2020). Both studies also found males were more likely to go unclaimed than females. In Los Angeles, Black individuals were more likely to be unclaimed; in Indianapolis, there was no association between Black race and going unclaimed among coroner cases, but race was not available for cases throughout the second half of the study period. The Indianapolis study found no increases by year in unclaimed deaths but concluded in 2011. In Los Angeles, there was limited evidence of increases over time, but due to limited availability of data at the time of writing only examined every third year from 1976 to 2013.
Only the study of Los Angeles data (Sohn et al., 2020) examined area-level social and economic markers of vulnerability. Unlike the current analysis, that study found the share never married and share separated were positively associated with unclaimed death risk. That study did not examine the share living in group quarters because it utilized data from the Current Population Survey (CPS), which does not include group quarters and institutionalized populations. Group quarters include facilities for both institutionalized (e.g., adult and juvenile correctional facilities, nursing facilities, in-patient hospice care, and care homes for those with disabilities) and non-institutionalized (e.g., college dormitories, military barracks, and emergency shelters) populations (Brault, 2008). Unfortunately the public-use ACS file doesn’t allow for disaggregation by facility type; however, a wide body of literature has shown the detrimental effects of living in certain types of group facilities on social ties, particularly nursing homes and prison facilities (Boamah et al., 2021; Brault, 2008).
In terms of economic predictors, similar to the present analysis, Sohn et al. (2020) also found the share living in poverty was positively associated with unclaimed death risk, but unlike this study, also found a statistically significant positive association for the share unemployed. It is unclear why some markers of social isolation and economic deprivation are significant in that study and not the current nationwide analysis. It may be the phenomenon of unclaimed death is in some way different in Los Angeles or the differences may lie in the fact that study utilized CPS rather than ACS data. However, overall, both the Los Angeles and this study found low area-level socioeconomic status and isolation were correlates of unclaimed death.
More broadly, our analysis of unclaimed death risk fits with the existing body of research on social ties and economic insecurity and how they may determine health outcomes. One of the most important themes in social science research is the role of social integration, defined by the American Psychological Association as “the process by which an individual is assimilated into a group (APA Dictionary of Psychology, 2018).” For example, Durkheim, in his landmark work on suicide (Durkheim, 2005), posited that suicides were tied to individual-level social isolation and could be seen as a marker of broader social disintegration. More recently, Putnam has documented the decline of social club membership and religious attendance as markers of a general social decline (Putnam, 2000). The present work suggests unclaimed deaths may be seen in a similar light, and for many decedents, unclaimed death may be the ultimate outcome of social isolation. For example, lower social integration has been tied to less social support, particularly for health-enhancing behaviors (Harasemiw et al., 2018). Social support is generally defined as assistance and protection given to others and can be broken down into dimensions, such as emotional and tangible support (Langford et al., 1997).
There are several plausible pathways that could lead to lack of social support at the community-level and ultimately resulting in an unclaimed death. Social isolation, while noted to be a hard-to-measure construct (Berkman et al., 2000; Zavaleta et al., 2017), is associated with poor physiological health outcomes: For instance, lack of social support has been shown to produce several negative health effects and premature mortality by altering cardiovascular, neuroendocrine, and immune function (Hawton et al., 2011). The trend of increasing loneliness and isolation in Western societies has been documented by several scholars (Andrew and Meeks, 2018; Carr, 2003; Caswell, 2022; Ogrin et al., 2021), and appears to coincide with increasing unclaimed death rates. Dying without kin is a known indicator of social isolation and loneliness, both of which are strong predictors of negative health outcomes, and untimely, premature death (Hawton et al., 2011; Philip et al., 2020; Uchino, 2006). The number of people without living parents, siblings, spouses, and children is projected to increase rapidly in the coming decades, particularly in older adults (Alburez-Gutierrez et al., 2023). For instance, a recent study found that 7.4% of decedents were kinless at the end of life (Plick et al., 2021). In addition, qualitative news reports suggest drug and alcohol addiction, particularly opioids, may play a role (Jordan and Sullivan, 2021). Finally, the present analysis found large, central metros to have higher rates of unclaimed deaths. With rural population in decline in the US, greater shares of the population will live in urban spaces (Anderson, 2015; Balk et al., 2018); this is meaningful because urban dwellers have been shown to have fewer social ties (Henning-Smith et al., 2019).
For decedents who are not kinless, an unclaimed death is a measure of economic disadvantage, both for the decedent and their social network. Social support may serve as a buffer against the most severe consequences of poverty (Green et al., 2023; Whelan, 1993). However, low socioeconomic status may have detrimental impacts on forming social ties (Weyers et al., 2008). In our ongoing analysis of interments at Hart Island, New York City’s public burial ground, during the COVID-19 pandemic, we found that excess deaths were heavily concentrated in the poorest neighborhoods of the city (Heiland et al., 2023). Although the current study provides a starting point to understanding the complex causal pathways that may lead to a decedent being unclaimed, further research is needed to fully understand this phenomenon. A scientific gap of particular note is the proportion of individuals who are unhoused at the time of death and veterans among the unclaimed dead.
4.2. Strengths and limitations
This analysis has several strengths, one being that it covers more than decade, and includes a large and diverse number of counties covering both large, metropolitan and also rural areas. The data collection includes counties that represent 22% of the U.S. population in 2020 (the year the most counties reported), and, in terms of race/ethnicity classifications and gender, look close to the distributions found in the nation as a whole (See Appendix 7). As the first nation-wide analysis of this topic, it is a first step in strengthening the data infrastructure for studying unclaimed deaths.
Despite the above strengths, several limitations should be noted. Our reported observations may be an underestimate of the phenomenon of unclaimed death. Underreporting, particularly in earlier time periods, is likely as many jurisdictions are generally not mandated to maintain these data. For example, they may start reporting in mid-year in initial data release. Second, there may be delayed reporting. The data analyzed in this study go through 2021, but it’s possible there may be a delay between reporting of burial and death (data collection ended in 2023). Next, when comparing our estimates to local studies, it’s important to note that our case definition only included decedents who received a government-funded interment. However, a broader definition may include indigent burials funded by charities. Lastly, counties may be unable or even hesitant to fully document and report public burial deaths due to resource constraints and public pressure (Prickett and Timmermans, 2022).
Although we present here only county-provided data when available, for several counties (particularly rural ones), only NamUs-provided data were available, which as noted above, are very likely to provide undercounts of deaths. In addition, several large counties, such as Harris County, Texas, and San Diego County, California, did not respond to data requests (though it should be noted these jurisdictions do report at least partial data to NamUs). However, it is unlikely this underreporting would fully explain the patterns seen in the present regression results, as this would suggest more vulnerable counties would be more likely to provide data. Finally, it is important to note that the regression results shown here are at the county level and may not be consistent with associations at the individual level.
4.3. Public health implications
This research has several public health implications. First, the current study suggests unclaimed deaths may make up at least 1.2% of deaths, and conservatively may represent approximately 25,000–30,000 decedents annually. (As discussed above, some counties reported much higher rates and our overall prevalence figure is likely an underestimate.) Very little is known about this population, and the determinants of unclaimed death remain understudied. The present analysis and previous work on this topic suggest that socially and economically vulnerable individuals are at higher risk. Developing novel ways to understand marginalized populations is a public health priority because they are simultaneously at the greatest risk of adverse health outcomes and premature death and are also often the most challenging groups to document in traditional epidemiologic and social surveys. Unclaimed death rates may also provide an objective measure of social support that can be measured over time. No standardized measure of social support exists, and most work on this topic utilizes data from self-administered surveys, such as the widely used the Medical Outcomes Study Modified Social Support Survey (mMOS-SS) (Moser et al., 2012), that asks respondents whether they have someone available for various types of assistance, such as to help with daily chores when they are sick. However, it is not known whether social desirability bias may lead the socially isolated to incorrectly report more social ties than they possess. In addition, those of low socioeconomic status are often underrepresented in traditional epidemiological surveys and more likely to be lost to follow up (Gehlert and Mozersky, 2018; Long et al., 2018; Oris et al., 2016; Rogers and Lange, 2013). For example, an analysis of non-response in the World Trade Center Health Registry, a cohort followed more than 20 years, found non-participants were more likely to be non-White and have lower educational attainment and household income (Yu et al., 2015). Unclaimed death rates can provide an important complement to self-reported data because they are a measure of whether the decedent had sufficient economic and emotional social support in their circle of friends and family to organize a traditional memorial service and burial. Although this line of research examines what happens to an individual after death, it may also provide some insight into the insufficient social processes and networks that were unable to produce a more traditional burial. These disparities at death are important because they most likely are closely related to health disparities across the life course. Our results suggest public health data infrastructure would greatly benefit by expanding surveillance of unclaimed deaths in a systematic and nationwide database.
Unclaimed deaths may also produce negative mental health effects among the living. A wide body of research has established the need for proper, culturally appropriate burial in the grieving process (Doka, 2008; Lobar et al., 2006; Prickett and Timmermans, 2022). Further, because of excess mortality rates in lower-income and minority communities, family members often experience multiple premature bereavements, especially during crises. Thus, the distress imposed on bereaved individuals and communities by unclaimed deaths may be amplified for those communities with the greatest numbers of deaths. For example, mass burials during the first weeks of the COVID-19 pandemic in New York City sparked widespread outrage (Hennigan, 2020). In addition, an in-depth analysis of unclaimed death in Los Angeles found some decedents had a social network willing to hold a traditional burial but no one who officially qualified in the relatively narrow definition of next of kin many jurisdictions require in order to claim the body (Timmermans, 2024). Deaths that are considered “bad deaths,” marked by lack of respect for the decedents, have been found to trigger feelings of anger in survivors that has been linked with a range of other unhealthy outcomes, including elevated risk of heart disease and a tendency to push away those who could provide help and support (Uchino, 2006).
Finally, unclaimed deaths have an economic impact on local or state governments, which generally absorb the cost of body interment. No nationwide estimate exists for the annual cost of unclaimed deaths to local governments, but a report by the National Association of Counties in 2018 found costs were increasing and varied widely by county: For example, Mobile County, Alabama, paid a total of $112,515 in 2017 for unclaimed burials, while Cook County, Illinois paid $77,119 from July 2015 to April 2016. Most counties choose cremation over burial because it is more cost effective, and some counties run out of funding before the end of the fiscal year, particularly in places hit hardest by the opioid epidemic (Barton, 2018).
5. Conclusion
Documenting basic demographic and death-related factors that predict unclaimed death is an important contribution to a topic that has received little systematic study in the scientific literature. With rising levels of economic inequality and increasing attention to systemic racism in the US, it is important to understand trends and patterns in unclaimed deaths as a potential marker of those inequalities. Towards that end, it will be necessary to use a wide range of research approaches—including data on unclaimed deaths as used here but also other untapped quantitative data (such as those from the US Department of Veterans Affairs or more in-depth death certificate or medical examiner data) as well as qualitative studies (Quinet et al., 2016; Sohn et al., 2020; Timmermans, 2024) to capture and understand the experiences of at-risk populations and associated underlying societal conditions.
Acknowledgements
We would like to acknowledge Simon Aytes who wrote the Python code to scrape publicly available websites; Brock Morell, Sascha Garrey, and Allen Mena who aided in jurisdiction outreach; Polina Verkhusha who aided in data cleaning; Emanuel Agu who aided in preparing the Hart Island data; and Jennifer Laird, PhD, who consulted on project design.
Funding sources
CUNY Interdisciplinary Research Grant 80209-01 26, PSC-CUNY Grant 65198-00 53, NIH R03 Grant AI166809-01.
Appendix 1. Correlation coefficient matrix of all predictors

Appendix 2. OLS regression coefficients predicting unclaimed deaths per population in the United States (with descriptive statistics), 2009–2021
| Characteristic | Model 1: Basic (Exogenous) Demographics | Model 2: All Demographics | Model 3: Saturated | |||
|---|---|---|---|---|---|---|
| Beta | 95% CI | Beta | 95% CI | Beta | 95% CI | |
| Basic Demographic | ||||||
| Race (mutually-exclusive categories) | ||||||
| Percent Black (single-race) | −0.000014 | −0.000052, 0.000024 | −3.7E-05 | −0.000085, 0.000011 | −0.000021 | −0.000089, 0.000048 |
| Percent Asian/Pacific Islander (single-race) | −0.000163 | −0.000242, −0.000084 | 0.000001 | −0.000126, 0.000128 | 0.000007 | −0.000120, 0.000135 |
| Percent Native American (single-race) | −0.000106 | −0.000180, −0.000031 | −0.00015 | −0.000228, −0.000062 | −0.000066 | −0.000165, 0.000033 |
| Percent Other Race (single-race) | 0.00044 | 0.000353, 0.000528 | 0.000397 | 0.000310, 0.000483 | 0.00035 | 0.000245, 0.000454 |
| Percent Two or More Races | 0.00006 | −0.000084, 0.000204 | 0.000001 | −0.000139, 0.000142 | 0.000013 | −0.000128, 0.000153 |
| Percent Hispanic (any race) | −0.000002 | −0.000045, 0.000041 | 0.000021 | −0.000045, 0.000087 | 0.000047 | −0.000022, 0.000115 |
| Median Age | 0.000274 | 0.000090, 0.000458 | 0.000249 | 0.000054, 0.000443 | 0.000287 | −0.000020, 0.000595 |
| Percent Over 65 | −0.000343 | −0.000560, −0.000125 | −0.00044 | −0.000671, −0.000211 | −0.000548 | −0.000804, −0.000292 |
| Percent Under 18 | −0.000602 | −0.000744, −0.000461 | −0.00032 | −0.000552, −0.000092 | 0.000157 | −0.000288, 0.000602 |
| Additional Demographic | ||||||
| Percent Foreign Born | −0.00019 | −0.000361, −0.000012 | −0.000051 | −0.000233, 0.000130 | ||
| Percent Of Households with No Children Under Age 18 | 0.000336 | 0.000161, 0.000511 | 0.000631 | 0.000148, 0.001113 | ||
| Percent Moved From a Different State | 0.000118 | −0.000258, 0.000494 | 0.000056 | −0.000325, 0.000438 | ||
| Percent Limited English | 0.00013 | −0.000094, 0.000355 | 0.000054 | −0.000180, 0.000288 | ||
| Percent Less than High School Diploma | 0.000289 | 0.000097, 0.000480 | 0.00013 | −0.000115, 0.000375 | ||
| Urban status1 | ||||||
| Large central metro | – | – | – | – | ||
| Large fringe metro | 0.000104 | −0.001117, 0.001325 | −0.000432 | −0.001640, 0.000777 | ||
| Medium metro | −0.00029 | −0.001545, 0.000956 | −0.001089 | −0.002343, 0.000164 | ||
| Small metro | −0.00248 | −0.004038, −0.000930 | −0.004048 | −0.005623, −0.002473 | ||
| Micropolitan | 0.000158 | −0.001440, 0.001755 | −0.001717 | −0.003412, −0.000022 | ||
| Non-core | 0.002643 | 0.000530, 0.004756 | 0.000451 | −0.001738, 0.002639 | ||
| Social Isolation | ||||||
| Percent Separated | −0.000147 | −0.001132, 0.000839 | ||||
| Percent Living Alone | 0.000188 | −0.000248, 0.000623 | ||||
| Percent Living in Group Quarters | 0.000728 | 0.000508, 0.000947 | ||||
| Percent Never Married | −0.000276 | −0.000488, −0.000063 | ||||
| Economic Deprivation | ||||||
| Percent below poverty level | 0.000063 | −0.000088, 0.000214 | ||||
| Median Income | −0.000082 | −0.000557, 0.000394 | ||||
| Percent Unemployed | 0.000357 | −0.000195, 0.000910 | ||||
| Other Control Variables | ||||||
| Year of death (or burial) | 0.00023 | 0.000126, 0.000334 | 0.00034 | 0.000226, 0.000454 | 0.000487 | 0.000349, 0.000625 |
| Year 2020 (pandemic) | 0.001639 | 0.000710, 0.002567 | 0.001437 | 0.000541, 0.002334 | 0.001093 | 0.000206, 0.001979 |
| NamUS observation1 (yes = 1) | −0.003855 | −0.004744, −0.002967 | −0.00416 | −0.005094, −0.003221 | −0.004314 | −0.005232, −0.003397 |
| Model fit statistics (R2) | 0.530 | 0.567 | 0.591 | |||
NB: Unclaimed deaths per population calculated as follows: county-year specific unclaimed deaths/total population*100; Bold font indicates statistical significance at the 0.05 level. All models are adjusted for state and year fixed effects.
The counts represent county-years.
Appendix 3. OLS regression coefficients predicting unclaimed deaths counts in the United States (with descriptive statistics), 2009–2021
| Characteristic | Model 1: Basic (Exogenous) Demographics | Model 2: All Demographics | Model 3: Saturated | |||
|---|---|---|---|---|---|---|
| Beta | 95% CI | Beta | 95% CI | Beta | 95% CI | |
| Basic Demographic | ||||||
| Race (mutually-exclusive categories) | ||||||
| Percent Black (single-race) | −2.62 | −3.80, −1.43 | −0.141 | −1.61, 1.33 | −2.06 | −4.23, 0.106 |
| Percent Asian/Pacific Islander (single-race) | −6.38 | −8.84, −3.93 | 0.689 | −3.23, 4.61 | 1.68 | −2.33, 5.69 |
| Percent Native American (single-race) | 1.36 | −0.888, 3.60 | 1.38 | −1.16, 3.92 | 0.26 | −2.84, 3.36 |
| Percent Other Race (single-race) | 2.96 | 0.281, 5.64 | 2.48 | −0.208, 5.17 | 2.44 | −0.891, 5.76 |
| Percent Two or More Races | 1.05 | −3.30, 5.40 | 0.462 | −3.86, 4.78 | 0.671 | −3.74, 5.08 |
| Percent Hispanic (any race) | 1.87 | 0.558, 3.18 | 3.78 | 1.74, 5.82 | 3.37 | 1.20, 5.53 |
| Median Age | 3.87 | −1.75, 9.49 | 5.91 | −0.075, 11.9 | 17.4 | 7.66, 27.1 |
| Percent Over 65 | −5.4 | −12.0, 1.24 | −9.37 | −16.4, −2.30 | −14.2 | −22.3, −6.18 |
| Percent Under 18 | −10.7 | −15.1, −6.43 | −11.9 | −18.9, −4.81 | 2.66 | −11.3, 16.6 |
| Additional Demographic | ||||||
| Percent Foreign Born | −6.44 | −11.9, −1.01 | −4.78 | −10.6, 1.02 | ||
| Percent Of Households with No Children Under Age 18 | 3.13 | −2.24, 8.50 | 17.3 | 2.13, 32.5 | ||
| Percent Moved From a Different State | −1.84 | −13.4, 9.70 | −1.78 | −13.8, 10.2 | ||
| Percent Limited English | 1.91 | −4.99, 8.82 | −0.625 | −8.01, 6.76 | ||
| Percent Less than High School Diploma | 2.02 | −3.87, 7.91 | −3.79 | −11.5, 3.90 | ||
| Urban status1 | ||||||
| Large central metro | – | – | – | – | ||
| Large fringe metro | 103 | 64.4, 142 | 104 | 65.1, 143 | ||
| Medium metro | 101 | 61.4, 140 | 96.1 | 55.7, 136 | ||
| Small metro | 102 | 52.4, 151 | 98.5 | 47.5, 149 | ||
| Micropolitan | 118 | 67.7, 169 | 101 | 47.0, 155 | ||
| Non-core | 96 | 30.1, 162 | 78.9 | 9.40, 148 | ||
| Social Isolation | ||||||
| Percent Separated | 22.9 | −8.11, 53.8 | ||||
| Percent Living Alone | −15.8 | −29.5, −2.15 | ||||
| Percent Living in Group Quarters | 5.26 | −1.66, 12.2 | ||||
| Percent Never Married | 5.98 | −0.775, 12.7 | ||||
| Economic Deprivation | ||||||
| Percent below poverty level | 4.58 | −0.189, 9.36 | ||||
| Median Income | 1.45 | −13.7, 16.6 | ||||
| Percent Unemployed | −1.99 | −19.5, 15.5 | ||||
| Other Control Variables | ||||||
| Year of death (or burial) | 4.53 | 1.39, 7.67 | 4.53 | 1.02, 8.03 | 3.43 | −0.908, 7.77 |
| Year 2020 (pandemic) | 23.7 | −4.41, 51.7 | 24.4 | −3.11, 51.9 | 30.2 | 2.30, 58.0 |
| Total Population | 0 | 0.000, 0.000 | 0 | 0.000, 0.000 | 0 | 0.000, 0.000 |
| NamUs Observation1 (yes = 1) | −33.9 | −61.9, −5.85 | −51.4 | −80.6, −22.2 | −47.8 | −77.1, −18.6 |
| Model fit statistics (R2) | 0.595 | 0.615 | 0.619 | |||
NB: Bold font indicates statistical significance at the 0.05 level. All models are adjusted for state and year fixed effects.
The counts represent county-years.
Appendix 4. Conceptual model of unclaimed death

Appendix 5. Description of American Community (ASC) variables
| ACS code | Description | Variable transformation, if applicable |
|---|---|---|
| B01001_001E | Total population | n/a |
| B01002_001E | Median age | n/a |
| B02001_006E | Native Hawaiian and Other Pacific Islander (alone) | Combined with Asian alone (C02003_006E), divided by total population (B01001_001E), multiplied by 100 |
| B02001_008E | Two or more races | Divided by total population (B01001_001E), multiplied by 100 |
| B03001_003E | Hispanic or Latino (any race) | Divided by total population (B01001_001E), multiplied by 100 |
| B05002_013E | Foreign born | Divided by total population (B01001_001E), multiplied by 100 |
| B07003_013E | Moved from different state | Divided by total population (B01001_001E), multiplied by 100 |
| B09016_022E | Living alone | Combined percent male (B09016_022E) and (B09016_025E) female living alone, divided by total population (B01001_001E), multiplied by 100 |
| B11005_011E | Households with no people under 18 years | Divided by total population (B01001_001E), multiplied by 100 |
| B12002_003E | Never married | Combined percent male (B12002_003E) and female (B12001_012E) never married, divided by total population over 15, multiplied by 100 |
| B12002_035E | Separated | Combined percent male (B12002_035E) and female (B12001_016E) separated, divided by total population over 15, multiplied by 100 |
| B14005_005E | Unemployed | Divided by total of those 18–64 years of age in labor force (male: C23002A_004E, female: C23002A_017E), multiplied by 100 |
| B17001_002E | Income in the past 12 months below poverty level | Divided by total population (B01001_001E), multiplied by 100 |
| C02003_003E | White (alone) | Divided by total population (B01001_001E), multiplied by 100 |
| C02003_004E | Black or African American (alone) | Divided by total population (B01001_001E), multiplied by 100 |
| C02003_005E | Native American (alone) | Divided by total population (B01001_001E), multiplied by 100 |
| C02003_006E | Asian (alone) | Combined with Native Hawaiian alone (B02001_006E), divided by total population (B01001_001E), multiplied by 100 |
| C02003_008E | Some other race | Divided by total population (B01001_001E), multiplied by 100 |
| DP02_0060PE | Percent population 25 years and over, 9th to 12th grade, no diploma | n/a |
| DP02_0113PE | Percent speak English less than “very well” | n/a |
| DP03_0063E | Median household income (dollars) | Divided by 10,000 |
| DP05_0021PE | Percent 65 years and over | n/a |
NB: Percent white, black or African American, Asian (combined with Pacific Islander as noted above), Native American, other race, and two or more races equal 100% for each county year. Percent Hispanic is made up of individuals of any race.
Appendix 6. Counties included in regression analysis
| Ada County, Idaho | Adams County, Illinois | Adams County, Oklahoma |
| Adams County, Pennsylvania | Adams County, Washington | Alachua County, Florida |
| Alachua County, Florida | Alameda County, California | Albany County, New York |
| Alpena County, Michigan | Anderson County, South Carolina | Anne Arundel County, Maryland |
| Anson County, North Carolina | Atlantic County, New Jersey | Barbour County, West Virginia |
| Barnstable County, Massachusetts | Barron County, Wisconsin | Berkeley County, West Virginia |
| Berks County, Pennsylvania | Berkshire County, Massachusetts | Bernalillo County, New Mexico |
| Blaine County, Idaho | Blount County, Tennessee | Boone County, Missouri |
| Bristol County, Massachusetts | Bronx County, New York | Brown County, Wisconsin |
| Bucks County, Pennsylvania | Burlington County, New Jersey | Cabell County, West Virginia |
| Calcasieu Parish, Louisiana | Calhoun County, Michigan | Camden County, New Jersey |
| Cameron County, Texas | Campbell County, Tennessee | Canadian County, Oklahoma |
| Canyon County, Idaho | Carson City, Nevada | Charleston County, South Carolina |
| Cherokee County, Oklahoma | Chester County, Pennsylvania | Clare County, Michigan |
| Clark County, Nevada | Clark County, Washington | Cleveland County, Oklahoma |
| Cocke County, Tennessee | Cole County, Missouri | Columbia County, Florida |
| Comanche County, Oklahoma | Cook County, Illinois | Craig County, Oklahoma |
| Crawford County, Michigan | Creek County, Oklahoma | Crittenden County, Arkansas |
| Cumberland County, Tennessee | Dallas County, Texas | Dane County, Wisconsin |
| Darlington County, South Carolina | Delaware County, Oklahoma | Denver County, Colorado |
| District of Columbia, District of Columbia | Dona Ana County, New Mexico | Douglas County, Nevada |
| DuPage County, Illinois | Duval County, Florida | El Paso County, Texas |
| Emmet County, Michigan | Essex County, Massachusetts | Essex County, New Jersey |
| Fairfield County, Connecticut | Forsyth County, North Carolina | Franklin County, Massachusetts |
| Fremont County, Wyoming | Fresno County, California | Fulton County, Georgia |
| Galveston County, Texas | Gila County, Arizona | Gloucester County, New Jersey |
| Grafton County, New Hampshire | Grand Traverse County, Michigan | Grant County, New Mexico |
| Grays Harbor County, Washington | Greenville County, South Carolina | Grundy County, Illinois |
| Hamblen County, Tennessee | Hamilton County, Tennessee | Hampden County, Massachusetts |
| Hampshire County, Massachusetts | Hampshire County, West Virginia | Harris County, Texas |
| Harrison County, West Virginia | Hartford County, Connecticut | Hawaii County, Hawaii |
| Hendry County, Florida | Hidalgo County, Texas | Hinds County, Mississippi |
| Honolulu County, Hawaii | Howard County, Indiana | Hudson County, New Jersey |
| Hudspeth County, Texas | Humboldt County, California | Imperial County, California |
| Ingham County, Michigan | Ionia County, Michigan | Island County, Washington |
| Jackson County, West Virginia | Jefferson County, Alabama | Jefferson County, Missouri |
| Jefferson County, West Virginia | Jefferson Parish, Louisiana | Juneau County, Wisconsin |
| Kalamazoo County, Michigan | Kanawha County, West Virginia | King County, Washington |
| Kings County, New York | Kitsap County, Washington | Kittitas County, Washington |
| Knox County, Tennessee | Lake County, Indiana | Lake County, Michigan |
| LaSalle County, Illinois | Lawrence County, Arkansas | Le Flore County, Oklahoma |
| Lee County, Florida | Levy County, Florida | Lewis County, West Virginia |
| Lincoln County, Oklahoma | Litchfield County, Connecticut | Los Angeles County, California |
| Loudon County, Tennessee | Macomb County, Michigan | Maricopa County, Arizona |
| Marin County, California | Marion County, West Virginia | Mayes County, Oklahoma |
| McDowell County, West Virginia | McKinley County, New Mexico | Mecosta County, Michigan |
| Merced County, California | Mercer County, New Jersey | Middlesex County, Connecticut |
| Middlesex County, Massachusetts | Middlesex County, New Jersey | Missaukee County, Michigan |
| Mohave County, Arizona | Monmouth County, New Jersey | Monongalia County, West Virginia |
| Monroe County, Florida | Monroe County, Tennessee | Montcalm County, Michigan |
| Montgomery County, Pennsylvania | Montgomery County, Texas | Morgan County, Colorado |
| Morgan County, West Virginia | Muskegon County, Michigan | Muskogee County, Oklahoma |
| Natchitoches Parish, Louisiana | New Haven County, Connecticut | New London County, Connecticut |
| New York County, New York | Newton County, Indiana | Norfolk County, Massachusetts |
| Oakland County, Michigan | Ocean County, New Jersey | Oceana County, Michigan |
| Ohio County, West Virginia | Okfuskee County, Oklahoma | Oklahoma County, Oklahoma |
| Okmulgee County, Oklahoma | Oneida County, Wisconsin | Orange County, California |
| Osage County, Oklahoma | Oscoda County, Michigan | Ottawa County, Oklahoma |
| Pacific County, Washington | Passaic County, New Jersey | Pershing County, Nevada |
| Philadelphia County, Pennsylvania | Pierce County, Washington | Pima County, Arizona |
| Pinal County, Arizona | Pittsburg County, Oklahoma | Plymouth County, Massachusetts |
| Pocahontas County, West Virginia | Polk County, Iowa | Polk County, Tennessee |
| Pottawattamie County, Iowa | Queens County, New York | Raleigh County, West Virginia |
| Randolph County, West Virginia | Rapides Parish, Louisiana | Richmond County, New York |
| Riverside County, California | Rockingham County, New Hampshire | Rogers County, Oklahoma |
| Salem County, New Jersey | Salt Lake County, Utah | San Bernardino County, California |
| San Diego County, California | San Francisco County, California | San Joaquin County, California |
| Santa Barbara County, California | Santa Cruz County, Arizona | Schuylkill County, Pennsylvania |
| Scott County, Kentucky | Sequoyah County, Oklahoma | Sevier County, Tennessee |
| Shelby County, Indiana | Shelby County, Tennessee | Siskiyou County, California |
| Skagit County, Washington | Snohomish County, Washington | Sonoma County, California |
| Spokane County, Washington | St. Clair County, Alabama | St. Louis County, Missouri |
| Stanislaus County, California | Suffolk County, Massachusetts | Suffolk County, New York |
| Sullivan County, New York | Tulsa County, Oklahoma | Union County, New Jersey |
| Union County, New Mexico | Union County, North Carolina | Union County, Tennessee |
| Victoria County, Texas | Wagoner County, Oklahoma | Walworth County, Wisconsin |
| Washington County, Arkansas | Washington County, Oklahoma | Washington County, Pennsylvania |
| Washington County, Tennessee | Washoe County, Nevada | Wayne County, Michigan |
| Westchester County, New York | Wexford County, Michigan | Whatcom County, Washington |
| Whitman County, Washington | Will County, Illinois | Windham County, Connecticut |
| Winnebago County, Illinois | Wood County, West Virginia | Worcester County, Massachusetts |
| Yakima County, Washington | Yellowstone County, Montana |
Appendix 7. Comparisons to the study counties and all US, 2021
| Characteristic | Percent in US general population | Percent in general population in 243 counties represented in unclaimed burial data collection |
|---|---|---|
| Race/Ethnicity | ||
| Race (mutually-exclusive categories) | ||
| White (single-race) | 61.6% | 58.5% |
| Black (single-race) | 12.4% | 14.6% |
| Asian/Pacific Islander (single-race) | 6.2% | 9.0% |
| American Native or Alaskan Native (single-race) | 2.9% | 0.8% |
| Other (single-race) | 8.4% | 8.4% |
| Two or more races | 10.2% | 8.6% |
| Hispanic (any race) | 18.9% | 25.6% |
| Sex | ||
| Female | 51.1% | 50.8% |
| Male | 48.9% | 49.2% |
Footnotes
An earlier version was presented at the Population Association of American Annual Meeting, 2023.
CRediT authorship contribution statement
Jennifer Brite: Writing – original draft, Supervision, Methodology, Funding acquisition, Formal analysis, Data curation, Conceptualization. Frank W. Heiland: Writing – review & editing, Methodology, Formal analysis. Deborah Balk: Writing – review & editing, Supervision, Methodology, Conceptualization.
- The work described has not been published previously except in the form of a preprint, an abstract, a published lecture, academic thesis or registered report. See our policy on multiple, redundant or con-current publication.
- The article is not under consideration for publication elsewhere.
- The article’s publication is approved by all authors and tacitly or explicitly by the responsible authorities where the work was carried out.
- If accepted, the article will not be published elsewhere in the same form, in English or in any other language, including electronically without the written consent of the copyright-holder.
Data availability
Data will be made available on request.
References
- Akee R, Jones MR, Porter SR, 2019. Race matters: income shares, income inequality, and income mobility for all US races. Demography 56, 999–1021. [DOI] [PubMed] [Google Scholar]
- Alburez-Gutierrez D, Williams I, Caswell H, 2023. Projections of human kinship for all countries. In: Proceedings of the National Academy of Sciences, vol. 120, e2315722120. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Allison A, 2023. Being Dead Otherwise. Duke University Press. [Google Scholar]
- Anderson MJ, 2015. The American Census: A Social History. Yale University Press. [Google Scholar]
- Andrew N, Meeks S, 2018. Fulfilled preferences, perceived control, life satisfaction, and loneliness in elderly long-term care residents. Aging Ment. Health 22, 183–189. [DOI] [PubMed] [Google Scholar]
- APA Dictionary of Psychology. (2018). American Psychology Association. [Google Scholar]
- Balk D, Leyk S, Jones B, Montgomery MR, Clark A, 2018. Understanding urbanization: a study of census and satellite-derived urban classes in the United States, 1990–2010. PLoS One 13, e0208487. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Barton MA, 2018. Undertakers of Last Resort: Indigent Burials on the Rise, Denting County Budgets. National Association of Counties (NACOO). [Google Scholar]
- Berkman LF, Glass T, Brissette I, Seeman TE, 2000. From social integration to health: Durkheim in the new millennium. Soc. Sci. Med 51, 843–857. [DOI] [PubMed] [Google Scholar]
- Boamah SA, Weldrick R, Lee T-SJ, Taylor N, 2021. Social isolation among older adults in long-term care: a scoping review. J. Aging Health 33, 618–632. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Bowring J, 2011. Containing Marginal Memories: the Melancholy Landscapes of Hart Island (New York), Cockatoo Island (Sydney), and Ripapa Island (Christchurch).
- Brault M, 2008. Disability Status and the Characteristics of People in Group Quarters: A Brief Analysis of Disability Prevalence Among the Civilian Noninstitutionalized and Total Populations in the American Community Survey. US Census Bureau. [Google Scholar]
- Brite Jennifer, Laird Jennifer, Heiland Frank, Balk Deborah, April 13, 2023. Life and death among the most vulnerable Americans: An examination of public burials. Flash Session presented at: Population Association of America. New Orleans. https://submissions.mirasmart.com/PAA2023/Itinerary/EventDetail.aspx?evt=191. [Google Scholar]
- Carr D, 2003. A “good death” for whom? Quality of spouse’s death and psychological distress among older widowed persons. J. Health Soc. Behav 215–232. [PubMed] [Google Scholar]
- Case A, Deaton A, 2015. Rising morbidity and mortality in midlife among white non-Hispanic Americans in the 21st century. Proc. Natl. Acad. Sci. USA 112, 15078–15083. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Case A, Deaton A, 2020. Deaths of Despair and the Future of Capitalism. Princeton University Press. [Google Scholar]
- Caswell G, 2022. How Can We Know and Think about Dying Alone? Dying Alone: Challenging Assumptions. Springer, pp. 19–43. [Google Scholar]
- Centers for Disease Control and Prevention, N.C.f.H.S, 2021. National vital statistics system, mortality 2018–2021 on CDC WONDER online database, released in 2021. In: Data Are from the Multiple Cause of Death Files, 2018–2021, as Compiled from Data provided by the 57 Vital Statistics Jurisdictions through the Vital Statistics Cooperative Program. [Google Scholar]
- Cohen S, McKay G, 2020. Social support, stress and the buffering hypothesis: a theoretical analysis. In: Handbook of Psychology and Health, vol. IV. Routledge, pp. 253–267. [Google Scholar]
- Culler DH, 2017. The invisible dead, a silent epidemic: violating the right of sepulcher through scientific experimentation and mass disposal of unclaimed human remains. Wake Forest JL Pol. 7, 295. [Google Scholar]
- Dignity Memorial. https://www.dignitymemorial.com/veterans/21-homeless-vet-burials.
- Doka KJ, 2008. Disenfranchised Grief in Historical and Cultural Perspective.
- Durkheim E, 2005. Suicide: A Study in Sociology. Routledge. [Google Scholar]
- Gehlert S, Mozersky J, 2018. Seeing beyond the margins: challenges to informed inclusion of vulnerable populations in research. J. Law Med. Ethics 46, 30–43. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Green KM, Doherty EE, Bugbee BA, 2023. Can early disadvantage be overcome? A life course approach to understanding how disadvantage, education, and social integration impact mortality into middle adulthood among a Black American cohort. Prev. Sci 24, 829–840. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Hanzlick R, Clark S, Lothridge K, 2011. History of the national missing and unidentified persons system (NamUs). Acad. Forensic Pathol 1, 310–321. [Google Scholar]
- Harasemiw O, Newall N, Shooshtari S, Mackenzie C, Menec V, 2018. From social integration to social isolation: the relationship between social network types and perceived availability of social support in a national sample of older Canadians. Res. Aging 40, 715–739. [DOI] [PubMed] [Google Scholar]
- Hawton A, Green C, Dickens AP, Richards SH, Taylor RS, Edwards R, et al. , 2011. The impact of social isolation on the health status and health-related quality of life of older people. Qual. Life Res 20, 57–67. [DOI] [PubMed] [Google Scholar]
- Heiland F, Brite J, Balk D, Agú E, 2023. Spatial Dynamics of Mortality during the COVID Pandemic in New York City: Evidence from Hart Island Mass Burials. Population Association of America, New Orleans. [Google Scholar]
- Hennigan WJ, 2020. Lost in the Pandemic: Inside New York City’s Mass Graveyard on Hart Island. Time. [Google Scholar]
- Henning-Smith C, Moscovice I, Kozhimannil K, 2019. Differences in social isolation and its relationship to health by rurality. J. Rural Health 35, 540–549. [DOI] [PubMed] [Google Scholar]
- Holleran S, 2024. Dying apart, buried together: COVID-19, cemeteries and fears of collective burial. In: Death’s Social and Material Meaning beyond the Human. Bristol University Press, pp. 100–113. [Google Scholar]
- Holt-Lunstad J, Smith TB, Layton JB, 2010. Social relationships and mortality risk: a meta-analytic review. PLoS Med. 7, e1000316. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Ingram DD, Franco SJ, 2014. 2013 NCHS Urban-Rural Classification Scheme for Counties. US Department of Health and Human Services, Centers for Disease Control and … [PubMed] [Google Scholar]
- Jaffe S, 2023. US Surgeon General: loneliness is a public health crisis. Lancet 401, 1560. [DOI] [PubMed] [Google Scholar]
- Jean C, 1997. No more kin: exploring race, class and gender in family networks. Family Matters. [Google Scholar]
- Jordan M, Sullivan K, 2021. Alone in death. Wash. Post [Google Scholar]
- Kenney FL, 2023. ‘A prison for the dead’: Hart Island and spatial histories of marginalization. Difficult Death, Dying and the Dead in Media and Culture. Springer, pp. 149–159. [Google Scholar]
- Khoo LS, Lai PS, Siew SF, Ibrahim MA, 2023. Management of unidentified and unclaimed bodies: a comparison of model from four countries in the Asia Pacific Region. Forensic Sci. Med. Pathol 1–16. [DOI] [PubMed] [Google Scholar]
- Langford CPH, Bowsher J, Maloney JP, Lillis PP, 1997. Social support: a conceptual analysis. J. Adv. Nurs 25, 95–100. [DOI] [PubMed] [Google Scholar]
- Lobar SL, Youngblut JM, Brooten D, 2006. Cross-cultural beliefs, ceremonies, and rituals surrounding death of a loved one. Pediatr. Nurs 32, 44–50. [PubMed] [Google Scholar]
- Long DL, Howard G, Long DM, Judd S, Manly JJ, McClure LA, et al. , 2018. An investigation of selection bias in estimating racial disparity in stroke risk factors: the REGARDS study. Am. J. Epidemiol 188, 587–597. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Mauer M, King RS, 2007. Uneven Justice: State Rates of Incarceration by Race and Ethnicity.
- Moser A, Stuck AE, Silliman RA, Ganz PA, Clough-Gorr KM, 2012. The eight-item modified Medical Outcomes Study Social Support Survey: psychometric evaluation showed excellent performance. J. Clin. Epidemiol 65, 1107–1116. [DOI] [PMC free article] [PubMed] [Google Scholar]
- National Center for Health Statistics, 2022. Urban rural classification scheme for counties. 2019. URL: https://www.cdc.gov/nchs/data_access/urban_rural.htm.
- Ogrin R, Cyarto EV, Harrington KD, Haslam C, Lim MH, Golenko X, et al. , 2021. Loneliness in older age: what is it, why is it happening and what should we do about it in Australia? Australas. J. Ageing 40, 202–207. [DOI] [PubMed] [Google Scholar]
- Oris M, Roberts C, Joye D, Ernst Stähli M, 2016. Introduction: Surveying Human Vulnerabilities across the Life Course: Balancing Substantive and Methodological Challenges.
- Philip KEJ, Polkey MI, Hopkinson NS, Steptoe A, Fancourt D, 2020. Social isolation, loneliness and physical performance in older-adults: fixed effects analyses of a cohort study. Sci. Rep 10, 13908. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Plick NP, Ankuda CK, Mair CA, Husain M, Ornstein KA, 2021. A national profile of kinlessness at the end of life among older adults: findings from the Health and Retirement Study. J. Am. Geriatr. Soc 69, 2143–2151. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Prickett PJ, Timmermans S, 2022. “If no one grieves, no one will remember”: cultural palimpsests and the creation of social ties through rituals. Br. J. Sociol 73, 244–258. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Putnam RD, 2000. Bowling Alone: the Collapse and Revival of American Community. Simon and schuster. [Google Scholar]
- Quinet K, Nunn S, Ballew A, 2016. Who are the unclaimed dead? J. Forensic Sci 61, S131–S139. [DOI] [PubMed] [Google Scholar]
- Rogers W, Lange MM, 2013. Rethinking the vulnerability of minority populations in research. Am. J. Publ. Health 103, 2141–2146. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Sohn H, Timmermans S, Prickett PJ, 2020. Loneliness in life and in death? Social and demographic patterns of unclaimed deaths. PLoS One 15, e0238348. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Steinberg J, 2015. The social life of death on Delhi’s streets: unclaimed souls, pollutive bodies, dead kin and the kinless dead. Ethnos 80, 248–271. [Google Scholar]
- Takizawa T, Kondo T, Sakihara S, Ariizumi M, Watanabe N, Oyama H, 2006. Stress buffering effects of social support on depressive symptoms in middle age: reciprocity and community mental health. Psychiatr. Clin. Neurosci 60, 652–661. [DOI] [PubMed] [Google Scholar]
- Thiede BC, Kim H, Valasik M, 2017. Concentrated Poverty Increased in Both Rural and Urban Areas since 2000, Reversing Declines in the 1990s.
- Thompson P, Yeung P, 2015. ‘Is a funeral a right?’ Exploring indigent funerals from social work perspectives. Aotearoa N. Z. Soc. Work 27, 73–86. [Google Scholar]
- Timmermans P.P.a.S., 2024. The Unclaimed: Abandonment and Hope in the City of Angels. Penguin Random House. [Google Scholar]
- Uchino BN, 2006. Social support and health: a review of physiological processes potentially underlying links to disease outcomes. J. Behav. Med 29, 377–387. [DOI] [PubMed] [Google Scholar]
- US Census Bureau. Selected Housing Characteristics, 2009–2021 American Community Survey 5-year Estimates, n.d.
- Weyers S, Dragano N, Möbus S, Beck E-M, Stang A, Möhlenkamp S, et al. , 2008. Low socio-economic position is associated with poor social networks and social support: results from the Heinz Nixdorf Recall Study. Int. J. Equity Health 7, 1–7. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Whelan CT, 1993. The role of social support in mediating the psychological consequences of economic stress. Sociol. Health Illness 15, 86–101. [Google Scholar]
- Yu S, Brackbill RM, Stellman SD, Ghuman S, Farfel MR, 2015. Evaluation of non-response bias in a cohort study of World Trade Center terrorist attack survivors. BMC Res. Notes 8, 1–11. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Zavaleta D, Samuel K, Mills CT, 2017. Measures of social isolation. Soc. Indicat. Res 131, 367–391. [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.
Data Availability Statement
Data will be made available on request.
