The “Racist” Foster Care Myth

Why colorblind removals didn’t just fail, but actually hurt kids instead

There is a prevailing narrative that black kids are vastly overrepresented in foster care because the system itself is structurally racist. To fix this, progressive jurisdictions have championed a new policy called ‘Blind Removals’.

The concept is simple: hide the child’s race and socioeconomic status from the committee deciding whether to remove the child from their home. If you remove the race, you remove the racism.

Yet, the latest data proves that the entire premise of this intervention is false.

The Study

A rigorous 2022 study titled “The Push for Racial Equity in Child Welfare: Can Blind Removals Reduce Disproportionality?” by researchers from Duke, Penn State, and the University of Michigan took a hard look at this exact program. To test if they actually work, researchers had to look at Kent County, Michigan: one of the only places to fully implement the policy.

The credibility of this study lies in its methodological rigor. Critics might try to argue that the system is rigged before the social worker even arrives. That the most biased or strict investigators are sent by their agencies to black families. The reality of how the system works shuts that down entirely. Cases in Michigan are assigned on a strict, next-in-line rotation. When a call comes into the hotline and gets passed to the local office, the case simply goes to whoever happens to be next up in the rotation. Every single investigator gets a perfectly mixed, average batch of cases. It is effectively a random lottery. This made it so that researchers could prove that it was not the case that ‘strict’ investigators were being disproportionately funneled toward black families.

But before they did their micro-analysis, they checked the national data (NCANDS) to see if Kent County was an outlier. That way, we aren’t just looking at a potential local anomaly, but rather, we are looking at a system-wide reality. And what they found completely dismantles the ‘systemic racism’ narrative.

The national data confirmed that across the U.S., conditional on an investigation occurring, the removal gap between black and white kids is virtually zero.

The Reality: A Colorblind System

If the system is racist, we should see investigators removing black kids at much higher rates than white kids when faced with similar circumstances. The data shows this just isn’t the case.

Conditional on an investigation, investigators remove white and black kids similarly.

In Kent County, the difference in removal rates between white and black kids is less than 1%. The researchers explicitly looked for investigator bias and found:

“no clear evidence that racial prejudice on the part of investigators drive racial disparities at the substantiation and removal stages”.

The researchers even developed a ‘stringency’ index to see if individual investigator bias was hidden in the data. They found that an investigator’s personal level of strictness had zero correlation with the race of the child they were assigned. Whether an investigator was naturally ‘tough’ or ‘lenient,’ they applied that standard equally across racial lines.

In fact, the race of the social worker doesn’t matter either, further disproving the effect of race. White and black investigators process cases with near-perfect statistical symmetry, proving the system simply reacts to the reality of the home.

As the researchers noted:

“Black and White investigators both remove Black children at higher rates than White children”.

And as said, this isn’t just a Michigan phenomenon. The researchers pulled nationwide data from the NCANDS and found that in the average U.S. county, investigators remove black kids at a rate “nearly identical” to white kids. The system at the removal stage is effectively race-neutral.

If Not Racism, Then What?

If investigators are colorblind, why are black kids overrepresented? The disparity is entirely driven by the baseline data: black kids are almost 3× more likely to be investigated for maltreatment from the start.

The authors concede this right in their abstract:

We first show that over-representation in most foster care systems is driven by Black children being substantially more likely than White children to be investigated for maltreatment to begin with.

Black children in Michigan are twice as likely as White children to be the subject of a maltreatment investigation, but have nearly identical removal rates conditional on an investigation.

Black children are almost three times more likely than White children to ever be the alleged victim in a child maltreatment investigation.

In Kent County, 17 percent of White children will ever be the alleged victim in a maltreatment investigation, while this share is nearly 50 percent for Black children.

To prove this mathematically, the researchers ran a simulation and found that:

… equalizing the removal probabilities of White and Black children conditional on an initial investigation in Michigan has a negligible effect on overall disproportionality, while equalizing the probability that a child is the subject of an investigation nearly eliminates it …

The math is staggering: the study found that equalizing the initial investigation rates would reduce racial disproportionality by 96%—effectively ending the disparity. In contrast, if you only focused on the removal decision (as Blind Removals do), the disparity would actually increase slightly because white kids currently have a marginally higher removal rate, once an investigation is substantiated.

The system itself is just reacting to the reality on the ground. The racial gap in foster care is not evidence of a prejudiced system. It is simply a statistical mirror reflecting higher baseline rates of family dysfunction and neglect. You cannot blame social workers for the objective reality of the cases they are handed.

The Backfire

Because the system wasn’t actually racist to begin with, the ‘Blind Removals’ program was a spectacular failure.

Did hiding a child’s race from the removal committee reduce racial disparities? No. The researchers found:

“no evidence that blind removals impacted the already small racial disparities in the removal decision”

Additionally, the decline in the removal rate for white kids was similar to that of black kids.

Why did it fail so clearly? Because you cannot engineer away a bias that doesn’t exist. Activists built the ‘Blind Removals’ program on the absolute certainty that social workers were disproportionately tearing black families apart. But the data reveals a devastating irony for the creators of this policy: once abuse or neglect is proven, the system is actually slightly harsher on white families.

According to the study’s data, substantiated cases result in the removal of white kids at a higher rate than black kids. The researchers themselves pointed out this mathematical reality: if investigators are already removing fewer black kids than white kids under similar circumstances, a ‘colorblind’ committee cannot possibly achieve the activist goal of reducing the black foster care population. The policy didn’t fail because it was poorly executed; it failed because it was a ‘solution’ designed to cure a phantom anti-black bias, completely ignoring the reality that the system, if anything, already leans the other way.

But the policy wasn’t just useless; it was actively harmful!

By adding bureaucratic red tape and committee meetings to solve non-existent racism, the program caused a massive bottleneck. The average time for a child to be removed from an unsafe environment surged by 30%!

blind removals ... substantially increased time to removal.

We show that the program substantially increased the amount of time it takes to remove a child following a child welfare investigation: the median time to removal in Kent County increased by roughly 9 days (or 30 percent) relative to counties in the control group.

blind removals increased the median time to removal by roughly 9 days, an increase of roughly 30 percent relative to the pre-blind removal median in control counties.

Think about what that actually means. Children who were deemed to be in ‘imminent harm’ by an investigator were left to sit in potentially abusive, dangerous, or neglectful homes for over a week longer, simply because the bureaucracy demanded an unnecessary, ‘colorblind’ committee review.

Did the wait at least help the kids get better homes? No. The researchers state:

“We find no evidence that this increased time to removal resulted in an increase in the proportion of foster care placements in family settings (as opposed to congregate care)—a measure of placement quality”.

What was even the point?

Conclusion

The perception that ‘systemic racism’ is driving the foster care system is unsupported by the data. Investigators are already making incredibly difficult, colorblind decisions based on the severity of the cases in front of them.

When activists force agencies to adopt policies like Blind Removals to fix disparities that are actually caused by external behavior, not racism, they don’t achieve racial equity. They just further bloat an already bloated bureaucracy and leave abused children in danger for longer.

We must look at the human cost of these delays. In Kent County, the ‘colorblind’ red tape caused the median time to removal to jump from 14 days to 27 days. That is nearly 2 weeks where children remained in ‘imminent risk’ while a committee debated a factor that the data proves was never driving the decision in the first place: race.

You cannot cure a systemic bias that does not exist, but as this data proves, you can certainly cause systemic harm by trying.