An Analysis of Racial/Ethnic Disproportionality and Disparity at the National, State, and County Levels

Print More

Published by the Casey-CSSP Alliance for Racial Equity in Child Welfare
Available at

This report makes several important new contributions to the study of disproportionate minority representation in child welfare, largely by incorporating a variety of communities not typically examined in such studies: Native Americans, Asians/Pacific Islanders and Hispanics.

Measures of disproportionality compare children within a racial or ethnic group. Disparity ratios compare information among such groups to understand the relationship of populations to each other.

“To better understand disproportionality, you have to go below the national level,” said Robert Hill, the study’s author and a senior researcher for the Race Matters Consortium at the Maryland-based research firm Westat.

For example, because Hispanics are usually underrepresented at the national level, “people say it’s not a problem,” Hill said. “But they are overrepresented when you look at different states and counties … because they’re concentrated in localities.”

Using nationally representative data from the National Child Abuse and Neglect Data System (NCANDS) and the Adoption and Foster Care Analysis and Reporting System (AFCARS), Hill examined the presence of children by race and ethnic group in child protection investigations, substantiated investigations and placement into foster care.

Hill analyzed county-level data for five counties identified by the Casey-CSSP Alliance for Racial Equity in Child Welfare as having implemented promising practices to promote racial equity in their child welfare systems: Bexar County, Texas; Guilford and Wake Counties, N.C.; King County, Wash.; and Ramsey County, Minn. In addition, he examined state-level data for the four states in which those counties are located.

At the national level, Hill found that:

• Both black children and Native American children are overrepresented in the child welfare system at all three decision-making points: investigation, substantiation and placement.

• At investigation and substantiation, the percentages of both black children and Native American children are double their percentages in the general population. At placement, their percentages are two to three times their percentages in the general population.

• Asian/Pacific Islander, Hispanic and white children are all represented at each of the decision stages at lower rates than they are in the general population.

• Children of all racial and ethnic groups except white show increasing percentages of representation as they progress through the child welfare system toward placement. The proportion of white children decreases as they advance through the decision stages.

• American Indian children have higher disproportionality rates and disparity ratios than blacks at all three stages of decision-making.

At the state level, the rates among races in Washington and Minnesota are similar to the national rates. In Washington, however, not only are Hispanics overrepresented at all three stages, they are twice as likely as whites to be investigated, substantiated or placed in foster care.

In Minnesota, Asians and Pacific Islanders were surprisingly more likely than whites to undergo investigation and substantiation. Disproportionality trends in North Carolina and Texas were similar to each other, but slightly out of sync with national trends.

The study also provides detailed analyses of racial/ethnic trends in the five counties. Hill hopes that the ground-level information can be put to good use. “We want localities to focus on early stages of intake,” Hill said.

“All of the findings show [disproportionality] is a cumulative effect as you go deeper into the system. If you do some things to intervene at the earlier points, you may be able to reduce disproportionality at the later stages.”