Exploring Parental Incarceration, US Government Support Programs, and Child Health and Well-Being: A National Cross-Sectional Study

Abstract

Objectives

To describe relationships between parental incarceration and child health and flourishing—a measure of curiosity, resilience, and self-regulation—and to identify government programs that moderate this relationship.

Methods

Using the National Survey of Children’s Health data from 2016 through 2019 for children 6–17 years old, we estimated associations with logistic regression between parental incarceration and overall health and flourishing, adjusting for child, caregiver, and household factors. We secondarily examined physical health (asthma, headaches), mental health (attention deficit disorder/attention deficit hyperactivity disorder, depression), developmental needs (learning disability, special educational plan use), and educational (missing ≥11 school days, repeated grade) outcomes. We performed interaction analyses to determine whether government program participation (eg, free/reduced lunch, cash assistance) moderated relationships between parental incarceration and child outcomes.

Results

Children with parental incarceration accounted for 9.3% of the sample (weighted n = 4 400 000). Black, American Indian/Alaska Native, and multiracial children disproportionately experienced parental incarceration. Parental incarceration was associated with worse health (aOR, 1.31; 95% CI, 1.11–1.55) and higher odds of not flourishing (aOR, 1.66; 95% CI, 1.46–1.89). Physical health, mental health, developmental issues, and educational needs were also associated with parental incarceration. Participation in free and reduced lunch moderated the relationships between parental incarceration and general health and flourishing, and cash assistance moderated the association between parental incarceration and flourishing. For each, parental incarceration had an attenuated association with health among people who participated in government programs.

Conclusions

Parental incarceration is disproportionately experienced by Black and Indigenous children and associated with worse child health and well-being. Government support program participation may mitigate negative associations between parental incarceration and child outcomes.

More than 5 million children in the US have had an incarcerated parent according to the 2011–2012 National Survey of Children’s Health (NSCH).^1,2 Black children are 6 times more likely to have had an incarcerated parent than White children. Indigenous children are also more likely to experience parental incarceration, with estimates for Indigenous children in the Dakotas of 5 times the prevalence among White children.^2–5 Parental incarceration is more common among children living in poverty or whose parents have less than a high school education.^1,2,5 The health impacts of parental incarceration are concentrated in poor communities of color,^2,3,6 prompting discussion of how to reduce incarceration to improve public health through policy change—including decarceration—and community investment.^7,8

Parental incarceration has negative effects on child well-being, including worse general physical and mental health, worse oral health, and more unmet health care needs.^2,4,9–13 Additionally, parental incarceration is associated with more behavioral problems, lower school readiness scores, and lower grade point averages, contributing to negative social and developmental ramifications for children.^12,14,15 Some investigators hypothesize that parental incarceration is associated with poor child outcomes because it is associated with stress, stigma, and family disruption, often in families already experiencing hardship.^4,5,15–18 Additionally, health needs contribute to parenting stress, which in turn contribute to child behavioral and mental health needs, feeding into cycles of child health impacts of parental incarceration.^15,19,20 Another posited explanation for the impact of parental incarceration on child health is that parental incarceration leads to household socioeconomic hardship, including food insecurity and material hardship,^21–23 owing to fewer incoming financial resources and more going out to support incarcerated family members.⁴ These hardships continue after a family member returns home owing to prohibitions on access to monetary assistance, employment, housing, and education for people with felony convictions. For instance, adults with drug-related felony convictions may have limited access to government support programs including the Supplemental Nutrition Assistance Program (SNAP) and Temporary Assistance for Needy Families (TANF), despite the fact that most qualify based on income.²⁴ Although government support program participation improves child health^25,26 and may moderate associations between personal or vicarious incarceration experiences and negative perinatal outcomes,²⁷ to our knowledge, prior studies have not focused on how these programs relate to health outcomes among school-aged children who have experienced parental incarceration.

Our study aims were to update estimates of children who have experienced parental incarceration, describe associations between parental incarceration and child health and well-being, and examine whether structural interventions, specifically government programs, moderate this relationship. We hypothesized that, despite criminal legal system reforms and modest decreases in incarceration,²⁸ parental incarceration would remain associated with worse child health. Additionally, we hypothesized that participation in government programs would be associated with attenuated adverse effects of parental incarceration on child health.

Methods

Study Design

We performed a cross-sectional analysis including children 6–17 years old with valid responses for the independent variable (parental incarceration) in the 2016–2019 NSCH, a nationally representative survey. The survey is administered by the US Census Bureau and sponsored by the Maternal and Child Health Bureau of the Health Resources and Services Administration. Each year, households are mailed instructions for completing the survey or a screening questionnaire asking whether a child aged 0–17 is in the household. One child from each eligible household is selected as the subject of the questionnaire, which is completed on paper or online, and includes caregiver-reported measures of child physical and mental health, family, school, and social context.²⁹ Between 2016 and 2019, there were 131 774 survey respondents, 51.7% of whom were male and 48.3% of whom were female, with a mean child age of 9.4 years.

Variables of Interest

Independent Variable.

Our exposure variable was the experience of ever having an incarcerated parent, which we defined as responding affirmatively to the question, “To the best of your knowledge, has this child EVER experienced any of the following? Parent or guardian served time in jail.”^13,16,30,31

Dependent Variables.

We had 2 primary dependent variables: measures of a child’s general health and their well-being. We dichotomized the caregiver’s response to the question, “In general, how would you describe this child’s health?” into excellent or very good vs poor, fair, or good, as has been done in other pediatric literature.³² To define a child’s well-being, we used a validated measure of social-emotional flourishing—a measure of a child’s curiosity, resilience, and self-regulation—that includes three questions for children ages 6–17 years old^33,34: “How often does this child: show interest and curiosity in learning new things, work to finish tasks he or she starts, and stay calm and in control when faced with a challenge?” A child met each flourishing measure with an “always” or “usually” response and did not if the response was “sometimes” or “never.” The flourishing measure was dichotomized into whether a child met all 3 vs 0–2 flourishing measures.^30,33,34

Secondary outcomes included dichotomous (yes/no) caregiver-reported of ever having physical health diagnoses (asthma, headaches), behavioral health diagnoses (attention deficit disorder/attention deficit hyperactivity disorder, depression), developmental needs (learning disability, special educational plan use), and educational problems (missing ≥11 days of school, repeating a grade). Child well-being is best examined through multiple lenses, including physical health, mental health, social/developmental status, and educational attainment.^30,34–38 With the aim of incorporating this multifaceted perspective on defining child well-being into study design, we selected these variables based on previous research to examine parental incarceration and child well-being across multiple domains, while additionally providing updated population estimates of these important child health outcomes.^{4,16,31,39,40}

Covariates.

We examined child age, race and ethnicity, sex (male, female), birth weight (very low birth weight, low birth weight, not low birth weight), and nativity (born in the US, born outside of the US). Race and ethnicity were included in the following self-reported categories: Asian, non-Hispanic Black, Hispanic, Native American/Alaska Native, Native Hawaiian/Pacific Islander, non-Hispanic White, and other race (as reported by the respondent) or multiracial. We conceptualize race and ethnicity as a social construct that results in a differential exposure to parental incarceration owing to structural racism and include it in analyses to highlight the disproportionate impact of mass incarceration on Black and Indigenous families in particular. Additionally, we examined household characteristics, including food insecurity (food secure, the ability to afford nutritious food; mild food insecurity, the ability to afford food, but not always healthy food; and moderate-to-severe food insecurity, the inability to consistently afford food); presence of a biological or adoptive parent; whether English was the primary language; and whether the caregiver completing the survey had fair or poor physical or mental health. Finally, we categorized household income by a percentage of the federal poverty level (FPL) (≤100% FPL, 101%–200% FPL, and ≥201% FPL) using an imputed value of the FPL. These income strata were selected to account for how income-based program eligibility varies from state to state; most families with an income of ≤100% FPL are eligible for all the programs, whereas families with an income of 101%–200% FPL would face more varied eligibility status based on their state of residence, and as income increased beyond 201% of the FPL, eligibility for government programs is less common.

Interaction Terms.

We tested whether participation in state and federal benefit programs moderates the association between child health and parental incarceration. Participants were asked whether during the past 12 months a child’s household had received food stamps or SNAP benefits (yes/no); benefits from the Women, Infants, and Children (WIC) program (yes/no); free or reduced cost breakfasts or lunches at school (yes/no); or cash assistance from government welfare program (yes/no).

Statistical Analysis

We used percentages and χ² tests to characterize the population of children who had experienced parental incarceration compared with those who had not. Using multivariable logistic regression analyses, we estimated the associations between parental incarceration and children’s general health and flourishing. We performed multivariable logistic regression analyses for each secondary outcome. All regression models controlled for variables chosen a priori as potential confounders, including age, sex, race and ethnicity, birth weight, nativity, caregiver physical and mental health, residential adoptive or biological parent presence, household size, income, and primary language.

We performed an interaction analysis between government support program participation and parental incarceration.²⁷ For all statistically significant interactions, we calculated the average marginal effects to compare differences in the primary outcomes between children with parental incarceration and those without.

All analyses were performed in STATA version 16.1 (Stata-Corp, College Station, TX) using the svy suite of commands to produce nationally representative population estimates accounting for missing data, survey oversampling, and other factors affecting population estimates.⁴¹ This study was deemed exempt by our institution’s Human Investigation Committee and followed the STROBE reporting guidelines.

Results

Sample Characteristics

Between 2016 and 2019, there were 94 369 survey responses for children ages 6–17 years old, and 91 431 survey responses for children with valid responses to the parental incarceration question, representing 47 700 000 US children ages 6–17 years old annually (mean age, 11.5 years). Of the included children, 48.9% were female and 51.1% were male. Additionally, 50.9% were White, 25.3% were Hispanic, 13.5% were Black, 5.3% were multiracial or identified as another race, 4.5% were Asian, 0.4% were Native American or Alaska Native, and 0.15% were Native Hawaiian or Pacific Islander (Table I).

Table I.

Characteristics of US children age 6–17 years with history of parental incarceration in 2016–2019

Characteristics	Parental incarceration		No parental incarceration		Total (survey estimate*)	P value
	No. (survey estimate*)	%	No. (survey estimate*)	%
Child demographics
Sex						.26
Male	3465 (2 200 000)	50	43 676 (22 200 000)	51	47 141 (24 400 000)
Female	3398 (2 200 000)	50	40 892 (21 100 000)	49	44 290 (23 300 000)
Race						<.001
White, Non-Hispanic	4095 (1 900 000)	43	60 000 (22 400 000)	52	64 095 (24,300 000)
Black, Non-Hispanic	843 (1 100 000)	24	4879 (5 400 000)	12	5722 (6 400 000)
Hispanic	1008 (1 000 000)	23	9367 (11 000 000)	26	10 375 (12,100 000)
American Indian/Alaska Native, Non-Hispanic	124 (40 000)	0.8	417 (140 000)	0.3	541 (180 000)
Asian, Non-Hispanic	52 (30 000)	0.6	4528 (2 100 000)	5	4580 (2 100 000)
Native Hawaiian/Pacific Islander, Non-Hispanic	26 (10 000)	0.1	215 (70 000)	0.2	241 (70 000)
Other races, or ≥2 races, non-Hispanic	715 (400 000)	9	5162 (2 100 000)	5	5877 (2 500 000)
Child born in the US	6774 (4 300 000)	98	80 902 (40 800 000)	95	87 676 (45,100 000)	<.001
Child’s birthweight						<.001
Very low (<1500 g)	122 (100 000)	3	991 (580 000)	1	1113 (680 000)
Low (<2500 g)	582 (380 000)	10	5555 (3 100 000)	8	6137 (3 500 000)
Not low	5577 (3 500 000)	88	74 179 (37 200 000)	91	79 756 (40,700 000)
Household characteristics
Lives with ≥1 biologic or adopted parent	5155 (3 300 000)	75	82 269 (41 400 000)	96	87 424 (44,700 000)	<.001
English as primary home language	6633 (4 200 000)	95	78 863 (36 500 000)	85	85 496 (40,700 000)	<.001
Primary caregiver with fair or poor physical health	1071 (810 000)	18	4688 (3 100 000)	7	5759 (3 900 000)	<.001
Primary caregiver with fair or poor mental health	743 (540 000)	12	3235 (1 800 000)	4	3978 (2 300 000)	<.001

^*Population estimates rounded to the nearest 100 000 and averaged over the study period based on recommendations from the NSCH, with the exception of populations of <1 million, which are rounded to the nearest 10 000. Column percentages presented.

Children who have ever experienced parental incarceration accounted for 9.3% of the sample (weighted n = 4 400 000). Black, American Indian/Alaska Native, and multiracial children made up 14.0%, 0.4%, and 5.0% of the population overall, yet experienced parental incarceration at disproportionately high prevalence (24.0%, 0.8%, and 9.0%, respectively; P < .001) compared with their share of the population. Children who have experienced parental incarceration had a lower prevalence of living with a biological or adoptive parent (75% vs 96%; P < .001) and a higher prevalence of living with a primary caregiver who reported fair or poor physical health (18% vs 7%; P < .001) or mental health (12% vs 4%; P < .001) compared with children who have never experienced parental incarceration (Table I).

Parental Incarceration and Child Health Status

Fewer children who had ever experienced parental incarceration had very good or excellent general health status (80% vs 90%; P < .001) or met all flourishing measures (67% vs 82%; P < .001) compared with those who have not had parental incarceration. Children who had experienced parental incarceration had higher rates of poverty, food insecurity, and participation in government support programs compared with children without parental incarceration (Table II).

Table II.

Health and well-being measures for children age 6–17 years with history of parental incarceration in 2016–2019

Measures	Parental incarceration Observations (%), survey estimate		No parental incarceration Observations (%), survey estimate		Total (survey estimate*)	P value
	No. (survey estimate*)	%	No. (survey estimate*)	%
Primary outcomes
Very good or excellent general health	5609 (870 000)	80	77 328 (38 600 000)	90	82 937 (42 100 000)	<.001
Flourishing						<.001
Met 0–1 items	1089 (700 000)	16	5188 (2 900 000)	7	6277 (3 600 000)
Met 2 items	1155 (710 000)	16	8778 (4 800 000)	11	9933 (5 500 000)
Met 3 items	4508 (2 900 000)	67	68 976 (34 200 000)	82	73 484 (37 100 000)
Secondary outcomes
Asthma	1351 (890 000)	20	12 401 (6 200 000)	14	13 752 (7 100 000)	<.001
Headache	703 (420 000)	9	4928 (2 300 000)	5	5631 (2 700 000)	<.001
Diagnosis of depression	1142 (590 000)	13	5006 (2 000 000)	5	6148 (2 600 000)	<.001
Diagnosis of ADD/ADHD	1794 (990 000)	23	10 279 (4 500 000)	11	12 073 (5 500 000)	<.001
Special education plan	1739 (1 100 000)	25	13 724 (6 300 000)	15	15 463 (7 400 000)	<.001
Diagnosis of learning disability	1158 (720 000)	16	6913 (3 300 000)	8	8071 (4 100 000)	<.001
Missed school						<.001
Missed ≤10 days	6201 (4 000 000)	93	80 113 (41 000 000)	96	86 314 (45 000 000)
Missed ≥11 days in last year	540 (320 000)	7	3324 (1 600 000)	4	3864 (1 900 000)
Repeated a grade	827 (610 000)	14	3989 (2 400 000)	6	4816 (3 000 000)	<.001
Social determinants of health
Income, % of FPL						<.001
≤100%	1777 (1 600 000)	37	8081 (7 700 000)	18	9858 (9 400 000)
101%–200%	1925 (1 400 000)	31	12 542 (9 100 000)	21	14 467 (10 500 000)
≥201%	3161 (1 400 000)	32	63 945 (26 400 000)	61	67 106 (27 800 000)
Food insecurity						<.001
Food secure	3195 (1 800 000)	41	63 378 (29 800 000)	69	66 573 (31 600 000)
Mildly food insecure	2698 (1 800 000)	41	17 920 (11 100 000)	26	20 618 (12 900 000)
Moderate to severely food insecure	928 (770 000)	18	2875 (2 200 000)	5	3803 (2 900 000)
Household member received food stamps in the last year	2172 (2 000 000)	45	6580 (6 500 000)	15	8752 (8 500 000)	<.001
Household member received cash assistance in the last year	609 (560 000)	13	1363 (1 300 000)	3	1972 (1 900 000)	<.001
Household member participated in free or reduced lunch program in the last year	3994 (3 100 000)	72	16 704 (14 500 000)	34	20 698 (17 700 000)	<.001
Household member received WIC food assistance in the last year	453 (550 000)	13	2029 (2 800 000)	7	2482 (3 400 000)	<.001

ADD, attention deficit disorder; ADHD, attention deficit hyperactivity disorder.

^*Population estimates rounded to the nearest 100 000 and averaged over the study period based on recommendations from the NSCH, with the exception of populations of <1 million, which are rounded to the nearest 10 000. Column percentages presented.

Experiencing parental incarceration was associated with a higher unadjusted odds of worse general health (OR; 2.12; 95% CI, 1.84–2.44) and of not flourishing (OR, 2.17; 95% CI, 1.94–2.41) compared with children without parental incarceration. In multivariable logistic regression models, children who have ever experienced parental incarceration were more likely to have worse health (aOR, 1.31; 95% CI, 1.11–1.55) and less likely to meet flourishing goals (aOR, 1.66; 95% CI, 1.46–1.89) compared with children who never had a parent incarcerated (Table III). Children who had experienced parental incarceration had a higher odds of having frequent/severe headaches (aOR, 1.40; 95% CI, 1.14–1.73), attention deficit disorder/attention deficit hyperactivity disorder (aOR, 1.78; 95% CI, 1.55–2.04), and depression (aOR, 2.00; 95% CI-1.67, 2.39), but did not have a higher odds of asthma diagnosis (aOR, 1.16; 95% CI, 0.997–1.34) compared with children without parental incarceration. They were more likely to have had a special education plan (aOR, 1.41; 95% CI, 1.23–1.62), have a learning disability (aOR, 1.42; 95% CI, 1.21–1.68), miss ≥11 days of school in the last year (aOR, 1.39; 95% CI, 1.09–1.76), or repeat a grade (aOR, 1.81; 95% CI, 1.49–2.20) compared with children without parental incarceration.

Table III.

The unadjusted and adjusted associations between ever experiencing parental incarceration and measures of health and well-being, United States children age 6–17 years in 2016–2019

Outcomes	Unadjusted OR (95% CI)	P value	Adjusted* OR (95% CI)	P value
Primary outcomes
General health status
Fair, poor, or good health	2.12 (1.84–2.44)	<.001	1.31 (1.11–1.55)	<.001
Flourishing status
Meets 0–2 flourishing measures	2.17 (1.94–2.41)	<.001	1.66 (1.46–1.89)	<.001
Secondary outcomes
Physical health
Asthma diagnosis	1.51 (1.33–1.72)	<.001	1.16 (0.997–1.34)	.055
Frequent/severe headaches diagnosis	1.87 (1.56–2.23)	<.001	1.40 (1.14–1.73)	.002
Behavioral health
Diagnosis of ADD or ADHD	2.52 (2.25–2.82)	<.001	1.78 (1.55–2.04)	<.001
Diagnosis of depression	3.16 (2.74–3.65)	<.001	2.00 (1.67–2.39)	<.001
Social/developmental well-being
Special education plan	1.96 (1.73–2.23)	<.001	1.41 (1.23–1.62)	<.001
Learning disability diagnosis	2.32 (2.02–2.67)	<.001	1.42 (1.21 1.68)	<.001
Educational well-being
Missed ≥11 days of school in last 12 months	2.07 (1.66–2.58)	<.001	1.39 (1.09–1.76)	.007
Repeated grade	2.74 (2.33–3.22)	<.001	1.81 (1.49–2.20)	<.001

ADD, attention deficit disorder; ADHD, attention deficit hyperactivity disorder.

^*Model controls for age, sex, race and ethnicity, income, primary caregiver physical and mental health, presence of a biological/adoptive parent in the household, household size, household primary language, child nativity, and child birth weight.

Interaction Analyses

General Health.

Participation in the free or reduced lunch program moderates the association between parental incarceration and children’s general health. Among children with ever parental incarceration, not receiving free or reduced lunch was associated with a higher predicted probability having poor, fair, or good health (4.1% higher predicted probability; 95% CI, 2.0%–6.2%; P < .001) compared with children without parental incarceration. In contrast, among people who participated in free or reduced lunch, there was no negative association between parental incarceration and general health (Figure).

Figure.

Effect modification of government support programs on health and well-being for children ages 6–17 with parental incarceration.^a,b ○ Average marginal effect on odds of endorsing poor, fair, or good general health. • Average marginal effect on odds of not meeting all flourishing measures. ^aAnalyses incorporating interactions between each individual federal assistance programs and parental incarceration into fully adjusted logistic regression model (controlling for age, sex, race and ethnicity, income, primary caregiver physical and mental heath, presence of a biological/adoptive parent in the household, household size, household primary language, child nativity, child birth weight). ^bAverage marginal effect compares the difference in likelihood of endorsing the outcome of interest among children with parental incarceration compared with children without parental incarceration.

The association between parental incarceration and child general health was not moderated by participation in SNAP (P = .35), WIC (P = .45), or cash assistance (P = .46).

Flourishing.

Cash assistance and free and reduced lunch participation moderated the association between ever experiencing parental incarceration and flourishing. Among children whose households did not receive cash assistance, parental incarceration was associated with a 9.7% higher predicted probability (95% CI, 7.1%–12.2%; P < .001) of not flourishing, whereas among children whose households received cash assistance, there was no association between parental incarceration and flourishing. Parental incarceration was associated with a higher probability of not flourishing, with a higher predicted probability (10.7%; 95% CI, 7.0%–14.4%; P < .001) among people not participating in the free and reduced lunch program compared with people participating in the program (7.2%; 95% CI, 4.0%–10.4%; P < .001) (Figure).

Participation in SNAP (P = .06) and WIC (P = .68) did not moderate the association between parental incarceration and child flourishing status.

Discussion

In this nationally representative study, we found that nearly 1 in 10 US children ages 6–17 years have had a parent incarcerated, with the experience more common in Black and Indigenous families than White families. Ever having a parent incarcerated was associated with worse general health and a lower likelihood of flourishing. Children who have had a parent incarcerated were more likely to have physical health, mental health, developmental, and educational problems than those without parental incarceration. Participation in free and reduced lunch and cash assistance each moderated the relationship between parental incarceration and child health and flourishing, suggesting that these programs may be protective. Our study uses recent NSCH data to extend prior work on the negative impact of parental incarceration on children.^{2,4,12,18,30,42,43} We highlight the continued high prevalence, the disproportionate experience of parental incarceration among racial and ethnically minoritized groups, its pervasive impact across multiple domains of child health, and the potential benefits of government support programs for children who experience it.⁴⁴

We found that participation in certain government support programs, but not others, was associated with attenuated health-harming effects of having had a parent incarcerated. Specifically, participating in the free and reduced lunch program attenuated the relationship between parental incarceration and both general health and flourishing. Income supports like cash assistance also attenuated the association between parental incarceration and child flourishing. Differences in the programs that moderated general health and flourishing may be due to differences in program administration and perception. For example, the free and reduced lunch program is a nutrition program with more strict guidelines placed on the available food, potentially meaning that program participants regularly have access to healthier food. It also generally implies contact with schools, where physical and developmental monitoring take place, potentially contributing to better health outcomes.⁴⁵ In contrast with food assistance programs, TANF cash assistance places resources directly in the hands of caregivers, which can potentially reduce the stressors of poverty, thus leading to improved social-emotional well-being, as demonstrated by flourishing, but not necessarily health. In contrast with free and reduced lunch programs and cash assistance, we observed a lack of moderation among families participating in SNAP and WIC, which may be due to differences in program administration (including eligibility, target populations, and accessing benefits) that affect how they are experienced by families impacted by incarceration. Future studies that delve into qualitative perspectives of how these families perceive and interact with government support programs may elucidate causal pathways and options for optimizing program implementation for this population. Even so, these findings suggest that government program participation may mitigate the intergenerational collateral health consequences of experiencing parental incarceration.

Our findings are important in the context of laws like the Personal Responsibility and Work Opportunity Reconciliation Act of 1996 (PRWORA). PRWORA placed a lifetime ban on the receipt of SNAP and TANF for people convicted of drug felonies. Impacted households would then receive fewer benefits that would then be divided among more people, diluting the program’s ability to meet a family’s needs.^46,47 PRWORA allowed states to opt out or modify the ban; as of 2022, one state—South Carolina—has a full lifetime SNAP ban for people with drug felony convictions, and 21 states have modified SNAP bans.⁴⁷ Modified SNAP bans use restrictions like drug testing, postrelease ineligibility periods, or parole compliance to restrict food stamps eligibility.^48,49 Our findings highlight that policymakers should consider the collateral consequences of policies that limit eligibility for government support programs,⁴⁸ and should instead consider lifting these bans—including modified bans—because they may improve household financial well-being.

TANF restrictions have consistently been stricter than SNAP restrictions—as of 2022, 7 states had a full TANF ban.⁴⁷ In April 2021, the Biden-Harris administration recommended expanding SNAP eligibility to formerly incarcerated individuals, but made no recommendations regarding TANF.⁵⁰ Our study suggests that, as policymakers consider government program alterations, expanding access to the social safety net may improve child health and flourishing for children who are exposed to parental incarceration.

Government support program access has been a salient topic in recent years as policymakers increased benefits for families and attempted to reduce administrative barriers to receiving these benefits.⁵¹ For example, many states expanded free and reduced lunch program access to meet children’s needs during remote learning.^52,53 In October 2021, SNAP benefits increased by 25% of prepandemic levels,⁵⁴ the largest single change since the program started. Additionally, Child Tax Credit expansions were anticipated to decrease child poverty by 40% through regular direct payments; although this program expired, studies indicated it decreased food insufficiency among participating families.^55,56 Increasing government program benefits may benefit children experiencing parental incarceration, who experience poverty and participate in government support programs at higher rates,²³ but their positive effects may be diminished by policies that limit access for their eligible caregivers.⁴⁹

Our study highlights the negative impact that parental incarceration has on children and that it is disproportionately experienced by Black and Indigenous children. Structurally racist policies have created the mass incarceration of Black and Indigenous people, contributing to inequities in parental incarceration. Therefore, intentional policy changes are necessary to mitigate and reverse these inequities. As discussions on how to address health inequities across racial and socioeconomic lines persist, our study highlights a significant population-level association between parental incarceration on child health, which is an important area for future interventions.

Beyond government support programs, there are ongoing efforts around the country—with laws passed in states such as Massachusetts and Tennessee—to decrease parental incarceration itself via primary caretaker laws. These laws work by considering caregiver status before sentencing (ie allowing judges to consider sentences that minimize family separation), investing in community-based alternatives to incarceration, and promoting increased funding for community or health-based organizations, such as drug treatment programs.⁵⁷ Decreasing lengthy sentences, reducing reliance on re-incarceration as a response to probation or parole violations, addressing policies that contribute to racialized disparities, and prioritizing treatment-based alternatives to confinement can help to reduce incarceration at the federal, state, and local levels.¹¹ Along with government support program access, policies that reduce incarceration and promote decarceration through family and community investments are necessary to reduce the impact of parental incarceration on child health and well-being.

Our study has limitations. First, our exposure variable identifies whether a child has ever had a parent incarcerated in jail, but does not identify key characteristics about the parents’ incarceration, including whether the incarcerated parent lived in the child’s household before incarceration, whether parents have only been incarcerated in jails (generally used for pretrial detention or shorter postconviction confinement), in prisons (generally used for longer postconviction sentences), or both, nor does it allow for the description of frequency or duration of incarceration. As written, the question allows for broad characterization of a nationally representative population at high risk for worse health, but other data sources remain relevant for understanding the nuances between each of these variables. Second, owing to the cross-sectional nature of the survey, we are unable to identify causal relationships. Additionally, the NSCH includes only noninstitutionalized children; because children who have experienced parental incarceration are at a higher risk of being involved in the criminal legal system themselves,⁴ it is possible that the impact of parental incarceration on US children is underestimated in this study sample. Although we cannot determine causality in this study, future longitudinal studies may elucidate the mechanisms between parental incarceration, government support program participation, and child health and well-being.

Overall, parental incarceration is associated with worse child health and well-being. This relationship may be mitigated by participation in government programs like TANF and free and reduced lunches. Considering the intergenerational health harms of incarceration and racial inequities in parental incarceration, expansion of government program access, support for families and communities, and decarceration are necessary to improve health equity.

Glossary

FPL

Federal poverty level

NSCH

National Survey of Children’s Health

PRWORA

Personal Responsibility and Work Opportunity Reconciliation Act of 1996

SNAP

Supplemental Nutrition Assistance Program

TANF

Temporary Assistance for Needy Families

WIC

Women, Infants, and Children

Data Statement

Data sharing statement available at www.jpeds.com.