Is Minnesota's desire to keep black and American Indian children out of foster care actually putting them in increased danger?

Just last week, a task force convened by Gov. Mark Dayton to tackle problems in the state's child protection system met for the first time. The executive order creating that group called for recommendations on "pre-court child protection protocols, including screening decisions and the family assessment process."

As reported by Brandon Stahl in the Star Tribune (Judge 'sickened' by abuse program," Oct. 12), and by me in the Chronicle of Social Change, serious questions about Minnesota's use of a popular child protection reform called "family assessment" have surfaced.

While family assessment has evolved greatly from its 2000 Minnesota pilot, the basic idea is to offer supports to families accused of abusing their children without forcing them into the formal child protection system.

This is a laudable goal. It is especially important when considering the fates of black and Indian children who enter the system at higher rates than their white or Asian peers.

When family assessment response was launched in Minnesota, child protection professionals' understanding of how race plays into placement in foster care was largely influenced by the congressionally mandated National Incidence Study (NIS), which periodically tracks child abuse and neglect.

The third wave of these widely referenced reports was released in 1993, and it asserted that there was no difference between white and black maltreatment rates. Instead, the authors reasoned, racial disproportionality was driven by "differential attention somewhere during the process of referral, investigation and service allocation" — the implication being that child protection workers rip children from their families more because of racial bias than because of the children's need to be protected.

This is reflected in Minnesota's screening guidelines, which cite the first three NIS studies. "Several national research studies have found that families of color do not abuse or neglect their children at a higher rate than Caucasian families when differential exposure to child maltreatment risk factors are controlled," the guidelines read.

Importantly, Minnesota's guidelines, which came out in 2012, do not cite NIS-4, which came out in 2010 and made an about-face on the prior three studies.

"The NIS-4 found statistically significant differences between black and white rates of child maltreatment, contrary to the findings of the first three NIS cycles," NIS-4 reads.

While Minnesota Department of Human Services (DHS) spokeswoman Katie Bauer said in an e-mail that the DHS intends to "remove references to national studies from the Screening Guidelines in the next revision," the omission of NIS-4 suggests something about the assumptions under which the DHS is operating.

The fact is that black and Indian families are disproportionately exposed to "child maltreatment risk factors" in Minnesota and across the country.

The DHS' 2010 Minnesota Child Welfare Disparities Report provided a comparison of maltreatment risk among racial groups. When the DHS investigates a family, it uses an actuarial risk tool called Structured Decision Making to determine risk of child maltreatment. Citing 2008 data, for example, the DHS determined that Indian children were more likely to be rated as high risk than blacks, whites or Asians, while blacks were more likely to be rated as moderate risk than Indians and whites.

Stahl's recent story cites a Hennepin County judge who fears that too many high-risk cases are being pushed into family assessment, where the likelihood of getting protective services is lower.

When it comes to black and Indian children, that may be by design. One of the goals in DHS' 2015-19 Child and Family Services plan is to "reduce the racial and ethnic disproportionate application of discretionary assignment to investigation response."

What does this mean?

The plan indicates that child protection workers fielding reports of abuse or neglect should assign a higher proportion of black and Indian families to family assessment, when the case is discretionary. By 2019, according to the plan's "interim benchmarks and timetable," the goal is that 65 percent of black and Indian families will get an assessment instead of an investigation, mirroring rates for white families.

When asked about the benchmarks, DHS denied their existence.

Dave Thompson, who recently retired from the DHS and was one of the architects of family assessment response, is adamant that black and Indian families are more exposed to contact with the child protection system and thus receive more calls. But, he said, "When we are setting goals, we don't want to minimize the amount of effort going into the community for child protection for the pure sake of proportionality, when disproportionality of risk is what is driving it."

So the question must be asked: Are goals based on race alone helping or hurting children and families of color?

Daniel Heimpel is the founder of Fostering Media Connections and the publisher of the Chronicle of Social Change. He is a lecturer of journalism and public policy at the Goldman School of Public Policy at the University of California, Berkeley, and the Price School of Public Policy at the University of Southern California.