About this project

The Measuring Up project seeks to clarify the relationship between relevant federal and state policies and well-being among children from prenatally through age 2, assessed at the population level. Specifically, the goal of this project is to study how variation in state-level policies across states is associated with positive child outcomes to increase our understanding of how to improve children’s early outcomes. This information can be used to inform state and national policy agendas for supporting young children and their families. It can also inform state and federal policymakers who design and implement child maltreatment data systems.

As a first step, the Child Trends research team[1] conducted a scan of outcome measures for infants and toddlers. Child outcomes that were considered include cognitive and language development, developmental delay, maltreatment, and health outcomes such as infant mortality, preterm birth, and low birthweight. We concluded that exploring child maltreatment rates has inherent value (as an important measure of child well-being from birth through age 2) and data at the state level could be examined over time. With input from the Measuring Up Collaborative (the Collaborative)[2] and both internal and external subject matter and methodological experts, we then completed a scan of relevant policies to identify policies with a theory- and research-based justification for reducing child maltreatment rates among children through age 2 (see summary below).

This brief documents the various policies that were examined, explains the criteria used to select policies that warrant new analysis (called “focal policies”), and briefly describes the characteristics of seven focal policies. The piece concludes with recommendations for national and state policymakers to close the gaps in implementation and knowledge about policies that may reduce rates of child maltreatment.

Child maltreatment overview

Although federal law has some requirements for the definition of child maltreatment, states have flexibility in how they define it. Generally, child maltreatment definitions recognize four types—physical abuse, sexual abuse, neglect, and emotional maltreatment—and can be defined as “any recent 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 which presents an imminent risk of serious harm.” It is important to note that there is no consensus on a more specific definition of “child maltreatment,” which results in variation across state measures of maltreatment. There is no available measure that can consistently and accurately assess the actual incidence of maltreatment. As a proxy, we examine maltreatment reports and substantiated reports from child welfare administrative data. A key limitation of these measures is that we don’t know how closely these correspond with the actual incidence of maltreatment because many children who are reported have not been maltreated, and there are likely many children who have been maltreated who are never reported. Further, substantiation decisions can be incorrect. Despite its limitations as a proxy for maltreatment, maltreatment reports and substantiated reports have inherent value as an outcome measure because involvement in the child welfare system is often traumatic for children and families. For this brief, discussions of state maltreatment rates pertain to maltreatment report rates and substantiated report rates.

In 2020, there were 3,145,000 children (42.9 per 1,000) subject to maltreatment reports. Of these, the child welfare system identified 618,000 (18.4 per 1,000) as victims of neglect.[3] Children younger than age 1 had more than twice the rate of maltreatment than children at age 1—25 versus 11 per 1,000 children, respectively. Black and American Indian/Alaska Native children experience higher rates of reported maltreatment and this has been linked to a history of separating families of color, economic disparities, and bias in reporting. Child maltreatment risk factors at the individual and family level include but are not limited to substance use issues, lower socioeconomic status, lower education or income levels, and acceptance of corporal punishment for discipline. Protective factors include parental resilience; social connections; access to basic needs (e.g., food, clothing, housing, transportation, access to services); and nurturing environments. There are also community risk factors such as lack of community connection, community violence, lack of access to stable housing, and food insecurity. Community protective factors include support for families, safe environments, and access to high-quality child-care options. Our work identifies policy factors hypothesized to affect trends in child maltreatment reports and substantiated reports.

---

Policy scan methods

The policy scan focused on state policies that might be associated with reductions in child maltreatment over time and drew on preventive strategies and approaches described in the Centers for Disease Control (CDC) framework for preventing child abuse and neglect. Policies and practices which improve family access to services and programs can promote child maltreatment protective factors and/or reduce risk factors. Examples of policies and programs found to be associated with reductions in maltreatment reports and substantiated reports include the federal Earned Income Tax Credit (EITC)  policy, which can help to provide financial supports, and home visiting programs which provide information and supports and link families to a variety of services.

To identify federal policies potentially linked to preventing child maltreatment, the Collaborative used a multi-pronged approach. In the first step of the policy scan, we used the CDC framework to categorize an array of federal policies related to child and family well-being (e.g., Temporary Assistance to Needy Families [TANF], minimum wage, home visiting). The CDC framework strategies include strengthening economic supports for families, changing social norms to support parents and positive parenting, quality early education in early life, enhancing parenting skills to promote healthy child development, and intervening to lessen the likelihood of maltreatment. We also added two more constructs: family planning and children with disabilities. Family planning policies (including access to contraception) were considered important as unintended pregnancies may add stress to a family, potentially resulting in child maltreatment. Similarly, given the financial, physical, and emotional challenges faced by families that include children with disabilities, we examined whether federal policy, which governs services for children with disabilities, provided concrete supports to prevent child maltreatment. Given that information on these two topics is scarce, we searched widely for state-level data to no avail.

Next, we mapped 23 different policies or nationally embraced strategies (e.g., home visiting) being implemented by states onto the identified approaches. To check the validity of our approach, we conducted multiple convenings with internal and external experts in child maltreatment prevention, service systems, policy, and data. These subject matter and methodological experts provided feedback on the selected policies and provided additional recommendations on our planned analytic model. Advisors’ concern about variation across states in how maltreatment is defined and reported were ameliorated upon learning that difference in difference analyses are planned. Difference in difference analyses will examine changes over time in one state’s policies alongside the changes over time in that state’s rates of maltreatment reports and substantiated reports to determine whether there is any association.

After the convenings, we used the following criteria to create a final list of potential focal policies to be examined:

• Is there sufficient variation in state policy to complete analyses?;
• Is state policy data available for coding policy variation over time?;
• Were similar studies conducted in the past? (noting that we seek to fill gaps, not replicate past work); and
• Might we expect a change in rates of maltreatment given the features of the policy and the size of the population served?

Constructs that have been examined in related studies are not highlighted here as focal policies, but these constructs will be included in a forthcoming analyses as control variables if data are available. The next section discusses the focal policies assessed by the research team in more detail.

---

Policy scan results

Table 1 summarizes the results of the scan. Most of the policies identified are in the “Strengthen Economic Supports for Families” section (14 policies), and less than half met the above criteria to be identified as focal policies in relation to infant and toddler maltreatment for three major reasons, described below.

First, for some policies, a change in maltreatment would not be expected given the scope of the policy and the population served. For instance, some state prevention policies mandate the use of public awareness campaigns. These efforts tend to be broad and short-term, and they are often targeted locally—at the county level. Still, the effects of these campaigns are not well-studied. We also considered whether access to early care and education might influence family and child well-being. Relevant early childcare and education programs include public pre-kindergarten (pre-K) funding and enrollment (e.g., targeted vs. universal pre-K) and childcare subsidy through the Child Care and Development Block Grant (CCDBG). In both cases, variations across states in family eligibility requirements for these programs would have restricted the population of interest and available data could not be considered representative.

Second, some policies lacked state-level data to identify and measure variation over time. This includes policies around implementing home visiting programs, where there is a variety of programs across states but there are no publicly available measures of variation in implementation across time. Similarly, the Individuals with Disabilities Education Act program for infants and toddlers (IDEA Part C) only has programmatic state-level data and national funding data available; it does not include measures of implementation differences across states.

This lack of consistent data over time for a majority of the states was most evident for family planning policies. For this specific topic, we completed an in-depth scan of 23 potential measures, including number of clinics providing publicly supported contraceptive services, reported public expenditures for family planning client services, family planning funding restrictions, and number of claims and average payments for Medicaid services for family planning. As part of our scan, we also had personal conversations with the Guttmacher Institute and the Centers for Medicare & Medicaid Services. Our results yielded the unfortunate discovery that there are no consistent measures of family planning policies during the last 20 years for all states. From our work reviewing and coding state legislation passed in the last 15 years, and related to contraceptive access policies, we only found one consistent measure: a policy allowing pharmacists to directly dispense emergency contraception without a prescription, which states began implementing in 2016.

As noted, in some cases, previous research has linked infant and toddler maltreatment with state level variation in the policies; therefore, these constructs have been defined as control variables rather than focal policies. For instance, researchers found that states’ minimum wage laws predict lower levels of maltreatment, a decline concentrated among young children (ages 0–5). Similarly, researchers have found that the Medicaid expansion was associated with a reduction in the reported child neglect rate for children younger than 6 years of age.

---

Click here to download Table 1: Policies or nationally embraced strategies grounded in the CDC framework for preventing child abuse and neglect.

---

Seven policies met the criteria for inclusion as focal policies in our study, some of which have multiple measures and some of which lack data (see Table 2); other policies and contextual variables will be included in analyses as control variables. These seven focal policies have predominantly been identified as potentially reducing maltreatment by providing economic support for families. State policies such as EITC, CTC, TANF, and SNAP target low-income families, provide supplemental financial supports, and reduce stresses linked to poverty. In addition, paid family leave is a focal policy; studies comparing the effect of paid family leave in California to states that have not implemented paid family leave found reduced infant hospitalizations due to head injuries, reduced infant mortality within the first year (although it is not clear what percentage of these is due to maltreatment), and improved maternal and child health outcomes.

---

Click here to download Table 2. Focal policies and their characteristics.

---

Call to Action and Next Steps

One strategy for reducing infant and toddler maltreatment is to strengthen existing policies or enact new policies based on existing evidence of what drives reductions in child maltreatment. For example, federal minimum wage requirements are low, and increasing the minimum wage at the state level has been found to positively impact economic well-being for families and children, suggesting the value of higher minimum wages. Paid sick leave and paid family leave are two other policy examples that have found positively associated with family well-being; however, only a limited number of states have enacted these policies.

Another strategy to support effective policy design and implementation is to establish or improve requirements around data collection, review, and publication for accountability. For example, data are needed on family planning services provided by states over time. Also, there are insufficient data on state variation in implementation to include awareness campaigns and CARA, a provision of CAPTA . While WIC does have data on participation numbers and grant amounts, it does not have data on policies aimed at addressing barriers to benefit access such as eligibility misconceptions, language and cultural barriers, or satisfaction with services. Similarly, the data available on IDEA Part C are only at the national level, which makes it impossible to examine state-level variation. Finally, policies and programs such as home visiting programs and housing assistance programs enacted at the state or local levels do not have sufficient data to understand variations that can be accurately attributed to each state’s policy implementation strategy.

Finally, research strategies that focus on policy implementation will give the public a better understanding of why there might be a connection between certain policy characteristics and positive child outcomes, and how policies might be implemented to enhance their effectiveness. For instance, analyses have found that variation in how long a family is able to receive welfare services and supports (e.g., SNAP, WIC, TANF) is associated with reduced maltreatment reporting. Additionally, there is an opportunity for future studies to examine more or different components of these policies, or to focus more on the vulnerable target population of children prenatally through toddlerhood. For example, existing TANF data do not allow for disaggregating benefit amounts received by families with children of different ages.

If implemented, these strategies will provide researchers and policymakers with better-quality information to assess the benefits and effects of widely utilized policies. At a minimum, these recommendations can help policymakers and advocacy groups verify state and national returns on substantial public investments. More importantly, such efforts will promote high-quality, evidence-informed policy making and programming, which can lead to meaningful differences in the health and well-being of very young children by preventing or reducing instances of child maltreatment.

In the final phase of this project, we will conduct a multivariate analysis using the policy characteristics with trends over time in state data (i.e., difference in difference analyses) on child maltreatment reports and substantiated reports. In this analysis, we will use the overall maltreatment reporting rates, which are based on the total number of reports (per year) made to Child Protective Services (CPS) within each state for children ages from birth through 2, divided by the total number of children ages from birth through 2 in each state. Findings will be used to develop a toolkit for policymakers, which will go into more detail about policy characteristics with the most promise for preventing child maltreatment. Future project developments and products will be posted on our webpage https://www.childtrends.org/project/measuring-up.