Parental incarceration (PI) is both an adverse childhood experience (ACE) and an influencer of pediatric health. Despite evidence that rural America sees the highest incarceration rates and substantial inequities in pediatric health care access and services, it is unclear how the prevalence of PI and associated sociodemographic factors vary across urban, suburban, and rural regions of the United States.
This study used data from the National Survey of Children’s Health (2016–2021; N = 145 281). Based on proximity and population, households were categorized as urban, suburban, or rural. Caregivers reported on household income, race/ethnicity, and living arrangements as well as children’s exposure to ACEs, including PI. Chi-squared and t-tests compared the prevalence of PI across communities and assessed regional differences in ACEs and sociodemographic characteristics in the context of PI.
PI was most common in rural (12%) versus urban (8%) and suburban (6%) areas. ACEs were more prevalent among PI children compared with non-PI peers across regions, with slight differences between PI children across locales. Within all regions, PI was highest for Black, Latinx, Native, and multiracial children; those in poverty; and those in nonparent caregiver placements. However, these prevalences were consistently highest among rural children.
This study points to high rates of adversity and concern racial, economic, and residential disparities for PI children, particularly those in rural communities. Evidence from this study can be used as a foundation for future prevention and intervention pediatric health responses that address inequities and unmet needs for rural populations.
Rural children disproportionately experience parental incarceration, an adverse childhood experience and driver of pediatric health. Inequities in exposure to parental incarceration spark concern given disparities in pediatric health services across geographic regions.
This study uses recent data from children aged 0 to 17 years to examine regional disparities in parental incarceration. It provides new evidence of how parental incarceration is connected to child adversities, poverty, racial disparities, and living arrangements across regions.
The United States is grappling with what the public health implications of 50 years of mass incarceration has meant for individual, family, and community well-being. After a 500% increase in imprisonment rates across the past 5 decades,1–3 an estimated 2 million people were incarcerated in the nation’s jails and prisons at mid-year 2021.4,5 Although prison populations have decreased slightly each year since 2013,4 decarceration has not been the reality for all.6 In fact, jail populations grew 27% in rural counties between 2013 and 2019 (while declining 18% in urban areas), resulting in incarceration rates that were double in rural compared with urban communities.7 Put differently, 3 of 5 people in jail in 2021 were in rural counties.8
There are numerous drivers behind rising rural incarceration rates, some of which are contrary to conventional notions and public perceptions about crime and punishment. For instance, despite higher incarceration rates, rural communities report less crime than urban counties.9 And although criminalization tied to the opioid epidemic in rural communities is likely a contributing factor, this crisis is even more recent than initial mass incarceration booms within rural counties.10 Rather, systemic factors (eg, high caseloads and fewer hearings, longer times in custody before court appearances, a lack of diversion programs, financial incentives to renting jail beds) largely contribute to elevated rates of carceral oversight in rural communities.11
Parental incarceration (PI) rates vary across samples (eg, 47%–57% of people in state and federal prisons nationally, 68% in Minnesota state prisons, 80% of women in jails), though it is estimated that 1 in 14 children experience PI.12–15 PI is a salient driver of pediatric health.16,17 Systemic racism and other structural inequities disproportionately expose Black children, those living in poverty, and, notably, rural children to PI.15 Estimates from the only previous study to use national data are more than a decade old (2011–2012) and before declining incarceration rates (∼2013).4,15 Yet, these estimates suggested that 11% of rural and 6% of urban children experienced PI.15 To our knowledge, no research has extended these findings with recent nationally representative data nor expanded on demographic correlates of PI across regions.
Despite some research with rural incarcerated parents,18,19 limited work has focused on rural children impacted by PI.16,20,21 One study of 1400 rural youth in North Carolina found that 24% experienced PI, an adversity linked with poor mental and behavioral health.22 However, this study was limited in its ability to compare across geographic regions. Another study of 113 000 youth across Minnesota detailed PI prevalence rates that grew in magnitude along urban-rural lines (cities, 16%; suburbs, 14%; towns, 20%; rural, 22%), with stark racial disparities.16 This is salient given that, relative to peers, both children with incarcerated parents as well as rural youth report higher rates of other adversities, known contributors to poor health.23–26
Social-ecological models consider the complex interplay between individual, relational, and societal factors as drivers of pediatric health.27 Such theoretical models posit that by addressing factors at these levels concurrently, prevention and intervention efforts can more holistically and sustainably impact child health.27 Thus, the inclusion of individual exposure to adversity, relational aspects such as family structure and economic resources, and manifestations of systemic racism (ie, disproportionate PI) are paramount in studies of pediatric health equity in contexts of PI. Yet, our knowledge of these factors among households impacted by PI across the urban-rural continuum is underexplored. For instance, despite PI being linked with more diverse living arrangements for children,28 no known research has examined regional heterogeneity. Similarly, although children living in poverty are 3 times more likely to experience PI,15 no known work has identified differences by region. More studies examining social-ecological factors of urban, suburban, and rural households impacted by PI are needed to develop locally tailored pediatric health responses that address disparities and unmet needs.
The current study extends and intentionally builds on previous work, including 1 study using data from 2011 to 2012 and another using a state-based sample of adolescents.15,23 The current study uses recent, nationally representative data of children (aged 0–17 years) across urban, suburban, and rural areas to describe rates of PI, examine racial disparities, and detail associated adversities. Additionally, it provides new insights regarding children’s experiences of poverty and living arrangements across this urban-rural continuum in instances of PI. Findings inform pediatric health services that meet the diverse needs of families impacted by PI across regions of the United States.
Methods
Data
The current study used data from the 2016–2021 cohorts of the National Survey of Children’s Health (NSCH). The NSCH is a survey of a cross-sectional weighted probability sample of noninstitutionalized US children ranging in age from 0 to 17 years. The survey is funded by Health Resources and Services Administration’s Maternal and Child Health Bureau and conducted by the US Census Bureau. The sample was taken from the Census Bureau’s Master Address File that contains a complete listing of all known residences in the United States and includes an administrative flag to identify households most likely to have children.29 A screener was used to confirm the presence of any minor children in the household; in cases of more than 1 child, 1 was selected as the subject of the survey (termed “focal child”), oversampling children aged 0 to 5 years and those with special health care needs. These data are well-suited to the present inquiry because they are nationally representative, contain information on geographic regions and other family demographics, and assess children’s exposure to a variety of adversities, including PI. Given the variability in children’s ages (0–17 years), primary caregivers (predominantly parents) are the sole respondents to NSCH surveys, providing information about themselves and the focal child.
In total, 225 443 questionnaires were completed from 2016 to 2021 for all focal children (2016, N = 50 212; 2017, N = 21 599; 2018, N = 30 530; 2019, N = 29 433; 2020, N = 42 777; 2021, N = 50 892). However, given our focus on geographic regions and in line with prior NSCH research examining regional disparities,24,30 the current study was restricted to families residing in 1 of the 31 states or the District of Columbia for which geographic indicators were available in the public use dataset (Supplemental Table 3). This resulted in a final sample of 145 281 children (2016, N = 31 719; 2017, N = 13 375; 2018, N = 19 459; 2019, N = 18 522; 2020, N = 28 340; 2021, N = 33 866).
Measures
Geographic Region
The NSCH public use data do not include a precoded regional variable. Based on prior research, we constructed the following categories24,30 : rural (including noncore rural [towns with <10 000 persons] and micropolitan rural [towns with 10 000–50 000 residents]), suburban (urbanized population >50 000 but not including the principal city), and urban (the urban core, which includes the principal city in a metropolitan area). These were derived from the following items: mpc_yn (Metropolitan Principal City [MPC] status), metro_yn (Metropolitan Statistical Area [MSA] status), and cbsafp_yn (Core Based Statistical Area [CBSA] status). Cases were coded as urban if people resided in a location noted as an MPC. Cases were coded as suburban if they resided outside the MPC, but inside the MSA. Finally, cases were coded as rural if they were outside the MSA, regardless of whether they were inside the CBSA (micropolitan rural) or outside the CBSA (noncore rural).
Parental Incarceration
Caregivers answered questions pertaining to events that may have happened during the child’s life, including whether, to the best of their knowledge, the child had ever experienced a parent or guardian serving time in jail. Response options included Yes (1) or No (2) and were recoded to assign experiences of PI a value of 1 and non-PI assigned 0.
Adverse Childhood Experiences
In line with previous NSCH research,24,31,32 the following 8 lifetime reports of adverse childhood experiences (ACEs) were assessed: (1) economic hardship; (2) parental divorce or separation; (3) parental death; (4) domestic violence; (5) neighborhood violence; (6) household mental health concern; (7) household substance use; and (8) racial discrimination (see Table 1 for full item labels). Although NSCH’s measurement of ACEs has demonstrated strong internal validity and good psychometric properties,33 some ACEs (parent death, racial/ethnic discrimination) were not included in Felitti and colleagues’ original instrument.26
Adverse Childhood Experiences Prevalence Across Regions by Parental Incarceration Status, NSCH 2016–2021 (N = 145 281)
| . | Parental Incarceration . | . | |||||
|---|---|---|---|---|---|---|---|
| ACEs . | Urban (n = 44 182) . | Suburban (n = 77 585) . | Rural (n = 23 515) . | . | |||
| . | No (92.25%) . | Yes (7.75%) . | No (93.92%) . | Yes (6.08%) . | No (88.36%) . | Yes (11.64%) . | Significance . |
| Parent or guardian divorced or separated | 20.52% | 70.97% | 18.49% | 72.91% | 22.31% | 73.29% | A |
| Parent or guardian died | 2.85% | 11.95% | 2.44% | 11.86% | 2.83% | 9.07% | B |
| Saw or heard parents or adults slap, hit, kick, punch one another in the home | 3.34% | 33.80% | 2.65% | 36.06% | 3.66% | 37.64% | A, B |
| Was a victim of violence or witnessed violence in the neighborhood | 3.46% | 26.01% | 1.88% | 20.18% | 2.26% | 23.09% | A |
| Lived with anyone who was mentally ill, suicidal, or severely depressed | 5.94% | 25.42% | 6.29% | 27.61% | 7.46% | 30.24% | |
| Lived with anyone who had a problem with alcohol or drugs | 5.22% | 41.69% | 5.20% | 48.09% | 6.57% | 53.02% | A |
| Treated or judged unfairly because of his or her race or ethnic group | 4.60% | 15.49% | 3.77% | 12.59% | 2.54% | 7.53% | |
| Hard to cover basics like food or housing | 17.33% | 38.99% | 14.50% | 38.44% | 18.67% | 37.44% | B |
| . | Parental Incarceration . | . | |||||
|---|---|---|---|---|---|---|---|
| ACEs . | Urban (n = 44 182) . | Suburban (n = 77 585) . | Rural (n = 23 515) . | . | |||
| . | No (92.25%) . | Yes (7.75%) . | No (93.92%) . | Yes (6.08%) . | No (88.36%) . | Yes (11.64%) . | Significance . |
| Parent or guardian divorced or separated | 20.52% | 70.97% | 18.49% | 72.91% | 22.31% | 73.29% | A |
| Parent or guardian died | 2.85% | 11.95% | 2.44% | 11.86% | 2.83% | 9.07% | B |
| Saw or heard parents or adults slap, hit, kick, punch one another in the home | 3.34% | 33.80% | 2.65% | 36.06% | 3.66% | 37.64% | A, B |
| Was a victim of violence or witnessed violence in the neighborhood | 3.46% | 26.01% | 1.88% | 20.18% | 2.26% | 23.09% | A |
| Lived with anyone who was mentally ill, suicidal, or severely depressed | 5.94% | 25.42% | 6.29% | 27.61% | 7.46% | 30.24% | |
| Lived with anyone who had a problem with alcohol or drugs | 5.22% | 41.69% | 5.20% | 48.09% | 6.57% | 53.02% | A |
| Treated or judged unfairly because of his or her race or ethnic group | 4.60% | 15.49% | 3.77% | 12.59% | 2.54% | 7.53% | |
| Hard to cover basics like food or housing | 17.33% | 38.99% | 14.50% | 38.44% | 18.67% | 37.44% | B |
All 2-tailed, .05 level. A, significant difference between urban and suburban; ACEs, adverse childhood experiences; B, significant difference between suburban and rural; C, significant difference between urban and rural.
Household Federal Poverty Level
The Department of Health and Human Services’ federal poverty guidelines are used to reflect household income as a percentage of the federal poverty level (FPL; 100% representing the poverty line). In line with previous NSCH research,34 this includes the following 4 categories: FPL <100% (ie, below the poverty line), FPL 100% to 199%, FPL 200% to 399%, and FPL ≥400%.
Children’s Race/Ethnicity
Given that systemic racism has contributed to marked racial disparities in the criminal legal system, we examine racial and ethnic disproportionalities of children exposed to PI. In the NSCH, these categories included: white, Black, Asian/Pacific Islander, Latinx, Multiracial, and Native American/Other Race.
Children’s Living Arrangements
To capture children’s living arrangements, we followed prior NSCH research that uses available data to construct the following 6 categories28 : (1) 2 parent caregivers, married; (2) 2 parent caregivers, unmarried; (3) 1 parent caregiver, no other; (4) 1 parent caregiver, nonparent other; (5) 2 nonparent caregivers (eg, 2 grandparents); and (6) 1 nonparent caregiver, no other (eg, 1 aunt). More detail on this construction is included in Supplemental Table 4.
Analysis Plan
To address missing data in multivariate analyses, we present multiply imputed results calculated in Stata 17.1 using the MI commands (chained equations; 20 imputations).35 Results were invariant to the number of imputations used (eg, 5, 10) and substantively similar to that of listwise deletion. The results begin with an examination of the prevalence of all 8 ACE indicators for children with and without PI exposure across urban, suburban, and rural regions. Specifically, this analysis examines the differential in the prevalence of a specific ACE between PI and no PI children in a given region (eg, suburban) versus another region (eg, urban). Significant differences in these analyses provide evidence that the association between PI and a given ACE is stronger or weaker in 1 region versus another. Next, we calculated the prevalence of PI for each of the FPL, race/ethnicity, and living arrangements categories across regions. Chi-square and t tests assessed significant differences between groups across strata.
Results
Across the sample, 11.6% of rural, 7.8% of urban, and 6.1% of suburban children nationwide experienced PI (differences by state and region are presented in Supplemental Tables 5 and 6, respectively). Although there is a higher prevalence of each ACE among PI-exposed children regardless of region, some significant differences across locales emerged (Table 1). In general, PI was more strongly linked to 6 of the 8 ACEs in suburban regions relative to urban or rural regions. For instance, PI was more strongly associated with parent separation or divorce, neighborhood violence, and household substance use in suburban versus urban regions. Meanwhile, PI was more strongly associated with parent death and economic hardship in suburban versus rural regions. In suburban versus urban and rural regions, PI was more strongly associated with domestic violence. Critically, household mental illness and racial discrimination were not differentially associated with PI across regions.
Table 2 displays the percentage of children who experienced PI by region and household FPL, race/ethnicity, and living arrangements. Overall, within regions, findings indicate that the prevalence of PI is highest among (1) children in families below the FPL (versus other income levels), 2) Black, multiracial, and Native American/other race children (versus white children), and (3) children in 2 nonparent caregiver or 1 nonparent caregiver (no other) living arrangements (versus other living arrangements). Still, there were some noteworthy patterns when examining PI among each of these demographic categories by region.
Percentage of Children Who Have Experienced Parental Incarceration by Region and Household FPL, Child Race/Ethnicity, and Children’s Living Arrangements, NSCH 2016–2021 (N = 145 281)
| . | % Children Ages 0–17 Who Experienced Parental Incarceration . | ||
|---|---|---|---|
| . | Urban . | Suburban . | Rural . |
| Poverty | |||
| FPL <100% | 14.47% | 12.78% | 21.54% |
| FPL 100%–199% | 10.29% | 9.46% | 13.59% |
| FPL 200%–399% | 5.68% | 5.17% | 7.15% |
| FPL ≥400% | 1.89% | 2.20% | 5.46% |
| Race/ethnicity | |||
| White | 5.88% | 5.37% | 10.30% |
| Black | 14.73% | 12.39% | 18.05% |
| Asian/Pacific Islander | 0.94% | 1.07% | 2.48% |
| Latinx | 6.60% | 5.43% | 13.06% |
| Multiracial | 11.97% | 9.73% | 21.82% |
| Native American/Other race | 12.36% | 11.80% | 16.30% |
| Living arrangements | |||
| 2 parent caregivers, married | 1.47% | 1.07% | 2.88% |
| 2 parent caregivers, unmarried | 6.92% | 5.72% | 10.94% |
| 1 parent caregiver, no other | 15.66% | 13.13% | 19.74% |
| 1 parent caregiver, nonparent other | 17.68% | 15.49% | 23.05% |
| 2 nonparent caregivers | 27.85% | 33.41% | 42.33% |
| 1 nonparent caregiver, no other | 26.88% | 35.14% | 42.13% |
| . | % Children Ages 0–17 Who Experienced Parental Incarceration . | ||
|---|---|---|---|
| . | Urban . | Suburban . | Rural . |
| Poverty | |||
| FPL <100% | 14.47% | 12.78% | 21.54% |
| FPL 100%–199% | 10.29% | 9.46% | 13.59% |
| FPL 200%–399% | 5.68% | 5.17% | 7.15% |
| FPL ≥400% | 1.89% | 2.20% | 5.46% |
| Race/ethnicity | |||
| White | 5.88% | 5.37% | 10.30% |
| Black | 14.73% | 12.39% | 18.05% |
| Asian/Pacific Islander | 0.94% | 1.07% | 2.48% |
| Latinx | 6.60% | 5.43% | 13.06% |
| Multiracial | 11.97% | 9.73% | 21.82% |
| Native American/Other race | 12.36% | 11.80% | 16.30% |
| Living arrangements | |||
| 2 parent caregivers, married | 1.47% | 1.07% | 2.88% |
| 2 parent caregivers, unmarried | 6.92% | 5.72% | 10.94% |
| 1 parent caregiver, no other | 15.66% | 13.13% | 19.74% |
| 1 parent caregiver, nonparent other | 17.68% | 15.49% | 23.05% |
| 2 nonparent caregivers | 27.85% | 33.41% | 42.33% |
| 1 nonparent caregiver, no other | 26.88% | 35.14% | 42.13% |
FPL, federal poverty level; NSCH, National Survey of Children’s Health.
First, the highest prevalence of PI was consistently among rural children, with high rates of rural PI for (1) children living with 2 nonparent caregivers (∼42%) or 1 nonparent caregiver, no other (∼42%); (2) multiracial (∼22%) and Black (∼18%) children; and (3) those below the FPL (∼22%). Although these demographics were also correlated with PI in other regions, the prevalence of PI remains lower (eg, ∼12% of multiracial urban youth experience PI and ∼28% of urban youth living with 2 nonparent caregivers experience PI).
Overall, the lowest prevalence of PI was among children residing in households at the highest income level (>400% FPL; range, 1.89%–5.46%), Asian/Pacific Islander children (range, 0.94%–2.48%), and children living with 2 married parents (range, 1.47%–2.88%). Even so, among these less affected demographic groups, rural children experienced PI at 2.64 to 2.89 times greater odds than urban or suburban children.
Discussion
PI is a notable influencer of child health, calling for the expansion of pediatric health services for affected children across the country. Given evidence of regional disparities in incarceration, the current study used nationally representative data to expand our understanding of rates of PI, racial disproportionalities, and associated adversities across urban, suburban, and rural communities. It also provided new evidence related to poverty and living arrangements across regions in instances of PI. Results suggest that PI is most prevalent in rural communities where associated racial, economic, and residential disparities in PI are notably stark.
Rural America reports the highest incarceration rates,11 yet few studies have examined regional heterogeneity in PI. One prior NSCH (2011–2012) study reported PI for 11% of rural and 6% of urban children.15 We extended this work with recent NSCH data and nuanced categories to detail similar PI patterns: 12% of rural, 8% of urban, and 6% of suburban children. These studies assess PI exposure for children aged 0 to 17 years but there may be greater differences for adolescents. For instance, a recent statewide sample of adolescents found that 22% of rural youth experienced PI.23 This body of work raises concern for rural children disproportionately disadvantaged by PI, particularly given inequities in pediatric health services by region.36
Notably, exposure to PI is not equally distributed. Black, Latinx, Native, and Multiracial children across urban, suburban, and rural communities were consistently more likely than white youth to experience PI. Although a higher proportion of white youth in rural areas experienced PI relative to urban and suburban white peers, these prevalence rates were lower than that of rural multiracial, Black, Native, and Latinx children. Underlying these disparities are discriminatory practices across the criminal legal system that disadvantage communities of color.37 Systemwide policy reform is needed at the macro level to address these disparities while culturally responsive primary care services are expanded at the micro level to meet the needs of racially, ethnically, and regionally diverse children.
Economic disadvantage is well-documented among households impacted by PI.15,20,38–40 We shed light on the differential risk and prevalence of household poverty by PI across regions. For instance, among those below the FPL, a notably higher proportion of rural children experienced PI than low-income urban and suburban children. Poverty can be particularly acute among rural households, contributing to increased reliance on public health insurance and reduced utilization of pediatric health services.36 Given that PI is, itself, associated with foregone health care,41 health risks may be compounded for rural children who experience poverty and PI.
Regional heterogeneity in children’s household structures in contexts of PI has received little empirical attention. Speaking broadly, PI is linked with residential instability, guardianship changes, and diverse living arrangements for children.26,42–44 We found that there may be potentially differential risk for this across regions. For instance, among children living with 2 nonparent caregivers in rural areas, more than 2 in 5 experienced PI, rates notably higher than urban and suburban peers. This reinforces the notion that PI destabilizes households and likely shifts caregiving responsibilities, processes that may be more prevalent for rural families.
PI is a salient ACE that puts impacted children across regions at greater risk for additional challenges.15,23,25 We point to factors that may present the most additive risk within regions. For example, in instances of PI in rural contexts, 1 in 2 children lived with someone who had substance use problems. That said, these challenges are not siloed within regions; instead, PI was associated with cooccurring adversities across regions. Given that combating household challenges may look different in urban, suburban, and rural settings, there is a need for community-driven solutions that expand services to support family fulfillment and success.
Implications
This study has implications for pediatric health services for families experiencing PI in rural contexts. Improving rural pediatric healthcare requires a multifaceted approach that responds to structural inequities, including: (1) expanding telehealth services to improve access to care and mitigate transportation issues; (2) promoting recruitment and retention of pediatric providers and other specialists (ie, mental health providers); (3) investing in preventive care; and (4) collaborating with community agencies (eg, schools, churches). Providing health services in the context of PI calls for attention to the inequities identified here. For instance, even with insurance, PI households experiencing poverty may struggle to afford necessary pediatric care. Offering low-cost or sliding-scale services, as well as transportation (a notable barrier for rural families), can make these services more accessible. By addressing needs and offering high-quality care, rural pediatric providers can better intervene against health consequences of PI.
Limitations
This study has many strengths, including its recent and representative data, nuanced regional indicator, and sample of children aged 0 to 17 years. However, there are also limitations to consider. First, the dichotomous PI measure prevents examination of frequency, duration, recency of exposure, and which parent was incarcerated (eg, mother or father). Next, the NSCH uses an address-based sample that may miss those who are mobile, and certain states suppress residence data altogether, which may limit generalizability. Third, the survey is cross-sectional, meaning no causal relationships can be assessed. Finally, as with surveys that assess sensitive topics, there is risk for social desirability and recall bias that may impact caregivers’ willingness or ability to report on children’s exposure to PI and other adversities.
Conclusions
The United States is grappling with the consequences of 50 years of mass incarceration, including the disproportionate impact for rural families. Findings from this study extends prior work that outlines the heightened risk of PI for rural households, linking it with cooccurring ACEs and marked racial, economic, and residential disparities. Given that PI can adversely influence pediatric health, mitigating these disparities is crucial to support child well-being.
Dr Muentner conceptualized and designed the study, drafted the initial manuscript, and critically reviewed and revised the manuscript; Dr Shlafer conceptualized and designed the study and critically reviewed and revised the manuscript; Dr Heard-Garris critically reviewed and revised the manuscript; Dr Jackson conceptualized and designed the study, carried out the initial analyses, drafted the initial manuscript, and critically reviewed and revised the manuscript; and all authors approve the final manuscript as submitted and agree to be accountable for all aspects of the work.
FUNDING: No external funding.
CONFLICT OF INTEREST DISCLOSURES: The authors have indicated they have no potential conflicts of interest to disclose.
Comments