Navigating Child Welfare Data Sources

The Landscape of Child Welfare Data

When analyzing the U.S. child welfare system, we rely on a few key federal data collection systems. These systems provide the raw material for tracking trends, understanding disparities, and evaluating policies. We can group them into two main categories:

1. Administrative Data: These datasets are compiled from the operational records of state child welfare agencies. They aim to be a census of all children interacting with the system.
2. Contextual Data: These are broader demographic and economic datasets, typically from the U.S. Census Bureau. They don't track children in the welfare system specifically, but they provide essential information about the communities where these children live.

Let's dive into the most important datasets in each category.

Core Child Welfare Administrative Datasets

The U.S. Department of Health and Human Services (HHS) manages two primary data collection systems that are indispensable for child welfare analysis: AFCARS and NCANDS. They are the "workhorses" of the field.

To get a detailed overview of these systems, please start by reading the following sections from the "Knowing the Numbers" guide by Child Trends. This document provides a clear breakdown of what each dataset contains, its purpose, and how to access it.

Summarizing AFCARS and NCANDS

As you saw in the reading, AFCARS and NCANDS give us two different, but related, views of the child welfare system.

• NCANDS (National Child Abuse and Neglect Data System): This dataset is your window into the "front door" of the child welfare system. It contains data on reports of child maltreatment made to Child Protective Services (CPS).

  • Key Information: It tells you about the children reported, the types of alleged maltreatment (e.g., neglect, physical abuse), the source of the report (e.g., teacher, doctor), and the outcome of the investigation.
  • Use in Analysis: NCANDS is essential for studying trends in child maltreatment reporting, understanding which populations are most often reported to CPS, and evaluating policies aimed at prevention or altering the initial CPS response.

• AFCARS (Adoption and Foster Care Analysis and Reporting System): If a child is removed from their home and placed in foster care, they will appear in AFCARS. This dataset tracks every child in foster care or adopted through a public agency.

  • Key Information: It includes demographics of the child, reasons for entering care, placement settings (e.g., foster home, group home), and outcomes upon exiting care (e.g., reunification, adoption).
  • Use in Analysis: AFCARS is the primary source for analyzing foster care trends. Your interest in longitudinal analysis of foster care outcomes will rely heavily on this data. You can use it to calculate metrics like time to reunification, placement stability, and rates of re-entry into care.

The image below is a great example of the kind of analysis you can conduct by combining these data sources. It visualizes a key outcome—exiting foster care to a permanent home—and breaks it down by the child's age.

It is important to note that while these are administrative datasets, they are not without limitations. Data quality can vary by state and over time, and definitions for certain variables may differ across jurisdictions. We will address these cleaning and consistency challenges in a future lesson.

Context is Key: The American Community Survey (ACS)

To understand why child welfare outcomes vary, we need to look beyond the child welfare system itself. Are foster care entry rates higher in areas with high poverty? Is access to services different in urban vs. rural counties? To answer these questions, we need data on the social and economic context of the communities where children live.

The single most important source for this contextual data is the American Community Survey (ACS).

To learn more about the ACS, please read the brief description in the "Knowing the Numbers" guide.

As the reading highlights, the ACS is an ongoing survey conducted by the U.S. Census Bureau. Unlike the decennial Census, which is a simple count, the ACS collects detailed information on topics like:

• Poverty status
• Income and employment
• Housing characteristics
• Educational attainment
• Household composition

For your goal of performing geospatial analysis, the ACS is indispensable. Its data is available at multiple geographic levels—including state, county, and even census tract. In a later lesson, you will learn how to merge AFCARS or NCANDS data (aggregated to the county level) with ACS data to explore the relationship between community characteristics and child welfare outcomes. This is a standard and powerful technique in policy analysis.

For a quick, alternative description of these key datasets, you can review the "Data Source Descriptions" section in the following guide.

Other Notable Data Sources

While AFCARS, NCANDS, and the ACS will be our primary focus, it's helpful to be aware of a few other specialized datasets mentioned in the "Knowing the Numbers" guide:

• National Youth in Transition Database (NYTD): This system collects data on services provided to older youth in foster care and tracks outcomes for youth as they transition to adulthood (at ages 17, 19, and 21). It's invaluable for studying the experiences of transition-age youth.
• National Survey of Child and Adolescent Well-Being (NSCAW): This is a longitudinal survey that follows a national sample of children who come into contact with the child welfare system. Unlike administrative data, it includes in-depth interviews and direct assessments of child and family well-being, mental health, and service needs. It's more complex to analyze but offers much richer detail.

These datasets provide deeper insights into specific populations but are generally not used for the same kind of broad, system-wide monitoring as AFCARS and NCANDS.

Conclusion

In this lesson, you've taken the first critical step toward your goal of analyzing child welfare policy. You've learned to identify the key data sources that will serve as the foundation for all the analysis we'll do in this course.

Here are the key takeaways:

• AFCARS (Adoption and Foster Care Analysis and Reporting System) is the primary source for data on children in foster care and adoptions. It's essential for tracking entries, exits, and outcomes.
• NCANDS (National Child Abuse and Neglect Data System) is the main source for data on child maltreatment reports and investigations, representing the "front door" of the system.
• The ACS (American Community Survey) is the workhorse for contextual data, providing social, economic, and demographic information at various geographic levels. It is crucial for geospatial analysis and for controlling for confounding variables in policy evaluations.
• All of these datasets are typically accessed via the National Data Archive on Child Abuse and Neglect (NDACAN) at Cornell University.

Now that you know what data exists, the next logical question is: how do you get it? In our next lesson, we'll move from concepts to code. You'll learn how to programmatically access and download datasets from public sources using APIs or web scraping, putting you on the path to building your first end-to-end analysis pipeline.