Preventive Benefits of U.S. Childcare Subsidies in Supervisory Child Neglect

Abstract

Using data from age 3 of the Fragile Families and Child Wellbeing Study, the current study explores the complex relationships between U.S. child care subsidies and neglect. Specifically, the study examines two research questions: (1) Are U.S. child care subsidies associated with self-reported neglect among low-income mothers? (2) What individual types of self-reported neglect are significantly reduced by receipt of child care subsidy? Using negative binomial regression examining the relationships among mothers who were income-eligible for child care subsidy, we found that child care subsidy was associated with lower levels of supervisory neglect, indicating an important role of subsidy in the lives of low-income families.

Introduction

Despite the many rewarding aspects of parenting, caregiving can be a stressful task especially in the context of low-income families. Low-income parents face multiple demands including meeting the expectations of their employer, finding appropriate care and education for their children, and meeting the needs of their children during nonworking hours. High-quality child care settings have the potential to reduce neglect, because it in theory increases supervision and safety of children. Government-sponsored child care subsidies are intended to be a work support and provide lower-income working mothers with the opportunity to have access to child care for their children (Healy & Dunifon, 2014). The current study investigates the relationships between child care subsidies and neglect.

Literature Review

Child neglect.

Child maltreatment is a significant and persistent social problem impacting many children in the United States. In federal fiscal year 2015, 3.4 million children were the subject of a child maltreatment report, with a total of 683,000 unique child victims of abuse and neglect (U.S. Department of Health & Human Services [USDHHS], 2017). Of these children, over one-third were age 3 and under (USDHHS, 2017). Over the first 18 years of life, an estimated 1 in 8 children in the United States will become a victim of child maltreatment (Wildeman et al., 2014). The effects of child maltreatment are far reaching and long lasting. Children who are maltreated are at risk for a variety of negative outcomes including physical and mental health problems (Felitti et al., 1998), violent and delinquent behavior problems (Fang & Corso, 2007; Widom & Maxfield, 2001), and economic hardships into adulthood (Currie & Widom, 2010). Child neglect is the most common form of child maltreatment in the United States (USDHHS, 2017).

Child neglect remains an understudied area in child maltreatment research (Stoltenborgh, Bakermans-Kranenburg, & van IJzendoorn, 2013). The Federal Child Abuse Prevention and Treatment Act (CAPTA) (42 U.S.C.A. §5106g), as amended by the CAPTA Reauthorization Act of 2010, defines child maltreatment as “any act or failure to act on the part of a parent or caretaker which results in death, serious physical or emotional harm, sexual abuse, or exploitation; or an act or failure to act which presents an imminent risk of serious harm.” Child neglect typically falls into the “failure to act” portion of the definition, and includes a wide range of behaviors including failing to meet a child’s basic physical (food, shelter, medical), emotional, and educational needs, providing improper or inadequate supervision, and leaving a child with an inappropriate caregiver.

Poverty and neglect are strongly related (Hussey, Chang, & Kotch, 2006; Schumacher, Slep, & Heyman, 2001). Although parents may not have the intention of putting their children in harm’s way, those living in and near poverty may have insufficient resources to meet their children’s basic needs (Drake & Pandey, 1996; Sedlak et al., 2010; Jonson-Reid, Drake, & Zhou, 2013). Failure to provide adequate food, shelter, and clothing is referred to as “basic needs” neglect (Coohey, 2003). Failure to provide adequate protection from harmful people and situations is considered supervisory neglect (Coohey, 2003). Economically disadvantaged parents may be unable to afford child care, and thus rely on informal networks to provide care for their children (Capizzano, Adams, & Sonenstein, 2000; Huston, Chang, & Gennetian, 2002; Roditti, 2000; Sonenstein, Gates, Schmidt, & Bolshun, 2002). Informal child care may not provide the same level of quality and reliability compared to formal child care arrangements (Fuller, Kagan, Loeb, & Chang, 2004).

Supervisory neglect is one of the most prevalent forms of neglect and has the highest fatality rate among maltreatment forms (Coohey; Hussey, Change, & Kotch, 2006). This form of neglect includes failing to provide adequate protection from harmful people and situations (Coohey, 2003). Despite its prevalence, studies on supervisory neglect are rare (exceptions include Coohey, 1998 and Jonson-Reid, et al., 2013).

Child care and neglect.

There are many different aspects (that are not universally agreed upon) of child care that contribute to whether a center is considered to be high quality, including class size, child-to-teacher ratio, and training of teachers (Howes, Phillips, & Whitebook, 1992). Child care services are increasingly highlighted as important sources of support for vulnerable families that enhance children’s school readiness, enable parents to work, and have the potential to promote positive parenting (Magnuson & Waldfogel, 2005). These services may even help prevent child maltreatment by socializing parents in prosocial parenting norms and offering respite from the sometimes overwhelming demands of caregiving (Horton, 2003).

A small body of work has examined the role of child care in neglect. Mothers who report child care concerns are likely to report supervisory neglect (Yang & Maguire-Jack, 2016). Further, lacking a “back-up” person to step-in for emergency child care needs is associated with overall levels of neglect (Ha et al., 2015). Klein (2011) found that preschool attendance is related to lower rates of maltreatment referrals and substantiations from child protective services. As such, having consistent and reliable child care may be protective against child neglect. When parents do not have to make difficult decisions about whom to leave their children with when they go to work (because of consistent quality care from a child care provider), child supervisory neglect may be especially impacted.

Child care subsidies and neglect.

The Child Care Development Fund (CCDF) was established in the 1990s during welfare reform, a time when work involvement among welfare recipients was emphasized. Consequently, the CCDF was created to provide subsidized child care as a work support for low-income families (Healy & Dunifon, 2014). Child care subsidies provide vouchers/certificates for care provided directly to the parent or reimbursements provided directly to the child care provider. If child care subsidies increase access to reliable child care, subsidies may be related to child neglect in that they: (a) decrease the economic burden associated with child care; (b) provide appropriate supervision for children; and (c) reduce irregularities in the availability of care (Healy & Dunifon, 2014). Given the connection between poverty and neglect, reducing this economic strain on families may enhance a family’s ability to meet the child’s basic needs.

Current Study

The current study examines the following research questions: (1) Are U.S. child care subsidies associated with neglect among eligible low-income mothers? (2) What individual types of neglect are significantly reduced by receipt of child care subsidy among eligible low-income mothers? For the first research question, we hypothesize that child care subsidies will be associated with lower neglectful behaviors in the sample for two reasons. First, we hypothesized that receipt of a child care subsidy would mean increased economic well-being for the parent, which would, in turn, reduce neglect because of the link between poverty and neglect (Hussey et al., 2006; Schumacher et al, 2001). Second, we hypothesized that increased ability to use child care services would increase the amount of time for which the child is being properly supervised. For the second research question, we expect that child care subsidies will be associated with decreased neglect related to basic needs (inability to provide food or go to the doctor for needed medical care) and supervision. We do not expect a relationship between child care subsidies and neglect due to mental health or substance use problems. Although Ha and colleagues (2015) examined the relation between child care burden and child maltreatment, the current study expands that work by specifically investigating the role of child care subsidies and examining neglectful behaviors.

Methods

Source of Data

We used data from the longitudinal birth cohort study—Fragile Families and Child Wellbeing (FFCW)—to examine the relationships between child care subsidies and neglect. The FFCW study began following a birth cohort of children born during 1998–2000 primarily to unmarried parents. The children and their parents have been followed in subsequent waves when the child was 1, 3, 5 and 9 years of age. At each wave, data were collected regarding a host of social circumstances including the children’s and families’ socioeconomic situation, parenting and child behaviors, and interactions with social support networks and the greater community (for further review of FFCW methods, see Reichman, Teitler, Garfinkel, & McLanahan, 2001). The FFCW dataset is ideal for examining relationships between child care subsidies and neglect for several reasons. First, the includes an oversample of low-income parents, who are more likely to be eligible for child care subsidies. Second, the study examines young children, who are likely to be enrolled in child care. Third, the dataset includes self-report information on child neglect.

The current study used a sample of mothers who participated in the third wave of the FFCW study, when the focal child was approximately 3 years old. Cases were selected if the mother was eligible for child care subsidy (N = 2,250) and had full information on the study variables (N = 1,179). We used the Child Care and Development Fund Reports to Congress to determine income eligibility for the subsidy by state (United States Department of Health and Human Services, 2003a, 2003b). Specifically, we created a dichotomous eligibility variable (1 = yes, 0 = no), which compared the state of the participant to the income eligibility for the subsidy of that specific state. If the mother’s income met the eligibility requirement, she was coded as “1” for that variable.

Measures

Predictor variables.

Child care subsidy receipt was measured as a single binary item indicating whether or not the mother received government assistance for child care. This measure came from two sources: mothers and their child care providers were both asked whether the care was paid for by government assistance. If either the mother or the provider indicated that the child care was paid for by government assistance, we considered that as the mother receiving child care subsidy.

Outcome variables.

Neglect was assessed by five maternal self-report items originating from the Parent-Child Conflict Tactics Scales (CTS-PC) (Straus, Hamby, Finkelor, Moore, & Runyan, 1998). Items include: being “so caught up in own problems that they (parents) did not show love to child”, “left the child home alone but thought adult supervision was needed”, “not able to make sure the child got the food he/she needed”, “were not able to take the child to the doctor or hospital when needed”, and “were so drunk/high that they had a problem taking care of child” (Cronbach’s alpha=0.57). Though the reliability coefficient is lower than the typical cut-off for reliability, this result is common for the CTS-PC scale because neglect behaviors do not typically cluster together (Straus et al., 1998). The response options included “never,” “not in the past year, but it happened,” “1 time,” “2 times,” “3–5 times,” “6–10 times,” “10–20 times,” and “more than 20 times.” As recommended by the scale developers (Straus et al., 1998), we used the frequency of the behaviors, which were assessed as the midpoint within the ranges provided (i.e. 8 times if the individual selected 6–10 times). We assigned 0 if the respondent chose “never” or “not in the past year, but it happened.”

Covariates.

We controlled for a number of variables related to child neglect (Stith et al., 2009). Economic hardship was measured by a scale consisting of eight dichotomous items that originated from the Survey of Income and Program Participation (SIPP; Bauman, 1998) and Social Indicators Survey (SIS; Social Indicators Survey Center, Columbia University School of Social Work, 1999) measuring participants’ perceptions of economic hardship. Item examples include: whether or not in the past year the mother was unable to pay rent/mortgage on time, and whether or not the mother was unable to get her child to a medical provider due to being unable to pay. Higher economic hardship indicates that families are experiencing multiple forms of hardship. Maternal education was coded as a series of dummy variables indicating less than high school education (1=yes, 0=no), high school education (1=yes, 0=no), and more than high school education (1=yes, 0=no, comparison group). Maternal marital status was dichotomously coded to indicate whether the mother was married to the father of the focal child or a new partner, versus not married. Maternal depression was coded dichotomously to indicate where or not the mother met the conservative criteria for the Composite International Diagnostic Interview (CIDI) Short Form (Kessler, Andrews, Mroczek, Ustun, & Wittchen, 1998). Child sex was measured dichotomously to indicate whether the child was male. Number of children living in the home was assessed continuously. Finally, we controlled for child externalizing behaviors, measured by parents’ reports of child behavior using the Child Behavior Checklist (Achenbach & Rescorla, 2001). FFCW included 24 items from the attention problems and aggressive subscales of externalizing behaviors included in the Child Behavior Checklist (Achenbach & Rescorla, 2001). Each item was scored (0 = not true, 1 = somewhat or sometimes true, 2 = very true or often true). We summed these scores for a total possible score ranging from 0–48 with a higher score indicating more externalizing behavior problems.

Analytic Strategy

We used Stata version 15 (StataCorp, 2017) for all analyses. Due to the count nature and overdispersed nature of the outcomes, we conducted a series of negative binomial regressions to estimate our models. Negative binomial regression does not require the mean to be equal to the variance, as required by Poisson models, and are therefore a best fit for the distribution.

Results

Descriptive Statistics

The descriptive statistics of the sample are provided in Table 1. In terms of the key independent variable, 30.87% of mothers reported receiving child care subsidy. The majority of the mothers were unmarried (83.5%) and approximately one-third (36.3%) attained less than a high school education. In the sample, 23.24% of the mothers met clinical criterion for depression and the average score of economic hardship experienced was about 2.00 (M = 1.98, SD = 1.68), suggesting mothers experienced an average of two forms of economic hardship. The average level of externalizing problem behaviors that children showed was 12.14 of a total possible of 48 (though the range within the sample was 0–25; SD = 5.79) indicating low levels of problem behaviors. There were about 2.5 children per family and approximately 53% of the focal children were male.

Table 1.

Descriptive Statistics for all study variables (N=1,179)

Variable	Mean (SD) or %	Range
Dependent Variables
Neglect	1.00(5.33)	0-125
Left child home	.11(1.26)	0-25
Unable to show love	.50(2.34)	0-25
Unable to provide food	.21(1.85)	0-25
Unable to take to doctor	.10(1.15)	0-25
Too intoxicated to care for child	.08(1.19)	0-25
Independent Variables
Child care subsidy	30.87%
Covariates
Married	16.54%
Economic hardship	1.98(1.68)	0-9
Maternal Depression	23.24%
Child Externalizing	12.10(5.79)	0-25
Mother education: Less than high school	36.47%
Mother education: High school	27.0%
Number of children in home	2.46(1.38)	0-9
Male child	53.01%

In terms of neglect, on average parents in the sample reported about one neglectful act in the past year. The plausible range was from 0–125, so this level is low. Parents reported less than one neglectful act on all of the individual neglect items; .10 for leaving the child home alone, .50 for being so caught up in the parent’s own problems that s/he couldn’t show love to the child, .21 for not providing food to the child, .10 for not taking the child to the doctor when it was needed, and .08 for being too “drunk or high” to care for the child.

Child Care Subsidy and Neglect

Table 2 shows the results of the first regression model, examining the association between receipt of child care subsidy and overall child neglect score. Receipt of child care subsidy was not associated with self-reported neglect. However, in the next set of models (Table 3), each neglect behavior was examined separately to understand the types of neglect to which child care subsidy was related. Of the individual neglect items, receipt of child care subsidy was associated with a lower rate of supervisory neglect, assessed as leaving a child home alone when the mother thought the child should be with an adult. In terms of the other individual items, while the coefficients were in the same direction as supervisory neglect, they were not significant.

Table 2.

Negative binomial regression predicting overall neglect score (N=1,179)

Variable	Coefficient	Standard Error
Child care subsidy	−.42	.34
Married	1.03	.40
Economic hardship	.06	.11
Depression	.12	.37
Externalizing	.10	.03
Less than high school	.69	.40
High school	.40	.40
Number of children	−.05	.11
Male child	−.26	.31

^*p<0.05

Table 3.

Negative binomial regressions predicting individual neglect items (N=1,179)

Variable	Left Home Coeff(SE)	Show love Coeff(SE)	Food Coeff(SE)	Doctor Coeff(SE)	Intoxicated Coeff(SE)
Child care subsidy	−3.51(1.38)*	−.36(.39)	−.07(.96)	−.89(.75)	−.81(1.05)
Married	1.94(1.09)	1.00(.47)*	.86(.97)	1.42(.88)	3.70(1.29)*
Economic hardship	−.01(.36)	.11(.13)	−.02(.31)	.06(.23)	−.26(.33)
Depression	.10(1.40)	.41(.43)	−.45(.85)	.34(.85)	1.99(1.21)
Externalizing	.08(.11)	.12(.03)*	.18(.08)*	.05(.06)	.26(.11)*
Less than high school	1.50(1.43)	1.01(.47)*	−.64(1.01)	2.56(.97)*	3.14(1.62)
High school	1.90(1.51)	.93(.45)*	−1.55(1.15)	2.07(.96)*	2.07(1.62)
Number of children	.38(.30)	.00(.12)	−.41(.32)	.12(.20)	−.50(.45)
Male child	.85(.92)	.08(.36)	−.79(.76)	−.40(.61)	−1.21(.99)

^*p< 0.05

Discussion

The current study sought to examine the relationship between child care subsidy and child neglect. While child care subsidy may only be used for child care, it is not surprising that receiving the subsidy decreases supervisory neglect. By virtue of increasing access to care for the child, supervisory neglect should be directly, inversely related to subsidy. Because of the high cost of child care (statewide averages ranged from $402 per month in Mississippi to $1,886 per month in Washington, DC; Economic Policy Institute, 2017), low-income families may be unable to gain access to formal child care. The lack of reliable child care may make it more difficult for low-income parents to consistently find safe and appropriate supervision for their child. The relationship between receipt of subsidy and supervisory neglect suggests that the subsidies may be effectively increasing access to adequate supervision for low-income children and is supported by findings from related studies (Ha et al., 2015).

Contrary to our hypothesis, our data did not find support for a relationship child care subsidy and basic needs neglect. It may be possible that the amount of funding received for child care subsidy is insufficient to raise the family above an income level that is sufficient to allow the family to meet all of their child’s basic needs. Additionally, this may be due to the low incidence of neglect within our sample.

Consistent with our hypotheses, we did not find a relationship between receipt of child care subsidy and neglect due to mental health issues or substance use issues. We thought that these two types of neglect were potentially more complex than basic needs and supervisory neglect, which can be easily tied directly to receipt of child care subsidy.

Limitations

There are limitations of the current study. First, the FFCW study purposefully oversampled families with unmarried parents, which resulted in an oversample of low-income and minority children. The study was also conducted in large, urban cities. Consequently, the results may not generalize to other populations, particularly to subsidy-eligible families in non-urban areas. However, given the focus on child care subsidy, which is targeted to low-income families, the sample restrictions may not be a disadvantage. Second, the models were conducted cross-sectionally, thus limiting our ability to make causal interpretations of the results. Third, child neglect was measured using a parental survey, which is susceptible to social desirability bias. Additionally, in our subsequent models, we investigated individual neglect questions, which were part of an overall neglect scale. As a result, the reliability of the CTS-PC from prior studies may not apply. Finally, although we attempted to control for potentially confounding variables (economic hardship, marital status, maternal depression, child behavior problems, maternal education, family size, and child sex), there may be additional variables that are not included that may be driving the associations we found.

Implications and Conclusion

Child care subsidies were implemented during the era of welfare reform as a policy to reduce barriers to work. Despite the policy intentions focusing on increasing the work of low-income parents, the current study suggests that these subsidies are an important support for families above and beyond the impact of work. Specifically, receipt of child care subsidy is associated with lower levels of self-reported supervisory neglect. Therefore, increasing the availability of child care subsidies may further decrease supervisory neglect in the United States.

Future work should include studies that are able to administratively examine whether parents received child care subsidies and engaged in supervisory neglect in order to avoid the problem of potential social desirability bias. Additionally, understanding whether the mediators of the relationship between child care subsidies and supervisory neglect is an important next step to disentangle the reason that subsidies seem to have a protective effect.