What part of “no evidence” does this child welfare “scholar” not understand

Meet the "scholar" who is turning the concept of "evidence based" upside-down
(Photo by Nick Youngson, pix4free.org)

This post quotes from many tweets. I have not tried to correct the typos in those tweets.

Have you noticed something new about the “child welfare” establishment lately? You know, the wonderful people who created the child welfare surveillance state that tears apart at least 200,000 families a year and subjects more than half of all Black children to traumatic child abuse investigations based on reports that are almost always false.  The tone of their writing seems to be increasingly desperate. 

After 50 years of health terrorism – misrepresenting the nature and scope of a problem, in this case, child abuse, in the name of “raising awareness,” people are catching on.  People are noticing that all that misery inflicted on millions of children in the name of stopping child abuse fatalities has done nothing to stop child abuse fatalities.  And they’re noticing that the misery is inflicted almost exclusively on people who are poor and disproportionately on people who are nonwhite.  I wouldn’t say people aren’t buying the fearmongering anymore, but it’s getting harder to sell. 

Among those sounding increasingly frantic is Richard Barth, former dean of the School of Social Work at the University of Maryland.  Barth, you may recall, is the one who declared that – unlike any other profession in America, child welfare is 100% free of racial bias!   

The evidence, including study after study, says otherwise. 

Perhaps that’s why Barth now is trying to stand the whole concept of evidence on its head. 

This can be seen most recently in his response to an op-ed column in the Hartford Courant by Prof. Kelley Fong, author of the forthcoming book Investigating Families. In that op-ed, she writes: 

Research finds that following high-profile child fatalities, child welfare agencies respond by removing more children from their homes in a “foster care panic.” There’s no evidence, however, that this makes children safer. Instead, such panics leave more children and their parents traumatized by family separation, and spread child welfare workers even thinner. 

Barth responded on his increasingly shrill feed on The Site Formerly Known as Twitter. (No, I will not be linking to it or anything else Barth has written. There’s enough here for anyone to find it if they are so inclined.  But I’ve reprinted the tweet in full, pausing for analysis.)  Let’s have a look, line-by-line. He begins with this: 

Nice to see acknowledgment of the many ways that child welfare services do help support families. 

That’s sort of like saying “what about Officer Friendly?” in response to issues of police brutality. Of course, sometimes individual police officers do good things. That does not justify stop-and-frisk and choking Black people to death.  And the very fact that in order to – maybe – get help, families have to go to what is really a police force adds enormous stress and drives many away from seeking help. 

Misreading We Were Once a Family 

Barth similarly misunderstood We Were Once a Family, Roxanna Asgarian’s brilliant book about Black children who were adopted to death by their white foster/adoptive parents.  Almost every reader understands what that book is all about: the racism that led to these children being taken from extended families who could have raised them and the racism that prompted family police agencies (a more accurate term than “child welfare” agencies) to ignore signs of abuse by the white savior foster / adoptive parents. 

But not Richard Barth. In a commentary for The Imprint about Asgarian’s book, he can’t bring himself even to mention the fact that the children were Black and the murderers were white. And his only solutions are more study and ramping up constant surveillance of all adoptive families. 

Now, back to Barth’s tweet, as he switches gears and discusses foster-care panics: 

 I agree that the standard of removal should not change, markedly, after a child death 

So, how much is “markedly”?  Often there is a 10% increase in a single year.  Is that “markedly”?  Sometimes it’s 20% or 30%.  In the worst example I know of, Florida in 1999, removals skyrocketed 50% in a single year.

More important, where is the evidence that every time there is a high-profile tragedy there should be any increase in the number of children torn from their homes?  It doesn’t exist.  

That does not stop Barth, who continues: 

--these deaths and serious injuries should be factored in to decision making before then. 

So does that mean we all should be tearing apart families at higher rates to begin with?  Again, where is the evidence?  We tear apart vastly more families than we did 50 years ago.  But, as is discussed further below, there is no evidence that child abuse deaths have decreased. 

And now we get to the heart of the matter. Says Barth: 

But not sure what it means that no evidence shows that removals after high profile deaths don't make children safer... 

Really, Prof. Barth?  You don’t know what “no evidence” means?  It means: No. Evidence. 

In fact, no evidence means pretty much what you said yourself in the very next sentence: 

I'd surprised if there is evidence that shows anything about the impact of subsequent removals after failure to protect children. Is there? 

Actually, there is some evidence about this.  It is, indeed, limited, but it does not support Barth’s position. 

In one sense the relative lack of evidence is a good thing.  Though each is the worst imaginable tragedy, child abuse deaths are as rare as they are horrifying.  So in all but the largest jurisdictions, they may rise or fall due to random chance.  And even in those large jurisdictions, there are other confounding variables, including bias in determining if a given death was due to, say, neglect, or an accident. 

But in those very large jurisdictions, I have never seen a foster-care panic followed by a decline in child abuse deaths.  Several times, there have been increases. 

Correlation is not causation.  But the best people like Barth can say is there is no evidence one way or the other.  Yet this is how he makes his case to maintain the status quo:  There’s no evidence foster care panics make children safer, but you can't prove they don’t! So apparently, we shouldn’t worry too much about them.  In another tweet, responding to Prof. Fong, who cited family defense attorneys who used statistics from government agencies, Barth doubled down: 

Hmm. Public defenders arguing from single trends don't give me any confidence. It's time for public defenders, and attorneys, and CW commentators, in general, to team up with scientists before drawing conclusions. I see that lawyers have now tweeted out your claims that there is no evidence. Sigh. 

Darn those pesky lawyers!  If Barth wants to convince people that it’s wrong to claim there’s no evidence all he has to do is provide the evidence.  But he can’t.  Because there’s no evidence! 

A failed defense of mandatory reporting 

In an opinion column he co-authored for JAMA Pediatrics Barth again turned the concept of “evidence-based” on its head to justify mandatory child abuse reporting laws. 

Those laws were deployed more than 50 years ago with no evidence base at all. There were no studies beforehand to see if they would work.  Now that research finally has been done, those studies show mandatory reporting backfires, scaring families away from seeking help and overloading workers so they have less time to find the relatively few children in real danger. 

So what does Barth say?  Keep mandatory reporting – but “Research is needed to find optimal approaches to certain circumstances…” 

So to review: Here is the traditional definition of “evidence-based”: First you come up with an idea.  Then you test the idea with rigorous studies.  If the studies show it works, then you deploy. 

Here’s the Richard Barth approach when it involves something he likes: Deploy mandatory reporting with no testing.  After 50+ years, when zero evidence says it works and the studies done so far say it’s harmful – then do more studies. 

Notably, even child welfare establishment types are becoming more reluctant to buy what Barth keeps trying to sell. 

That JAMA Pediatrics opinion column was followed by a comment from none other than Dr. Richard Krugman. Krugman is the very personification of the child welfare establishment.  He used to be director of the C. Henry Kempe National Center for the Prevention and Treatment of Child Abuse and Neglect. Here is his comment, in full: 

This is a thoughtful commentary stressing the need for balance in our views of Child Protective Services (CPS) agencies. As one who has worked in this field for the past 40 years, I agree with Dubowitz and Barth that just ablating CPS agencies is the wrong approach. BUT, we now have 40 years of experience with this approach and have made no progress in reducing the mortality from physical abuse of children (decades with 1500-2500 children dying annually). What is missing from CPS are data on the outcomes for children and families of the "services" being provided to them. What is needed are carefully designed studies testing alternative approaches to community child protection efforts that can tell us in the next decade whether those approaches lead to better outcomes for children and families that what we are doing now. Doing the same thing for 40 years that doesn't seem (or can't be shown) to be working was someone's definition of insanity. [Emphasis in original.] 

Alas, like Barth, Krugman can’t bring himself to actually call for not doing things that are not evidence-based.  But for someone like Richard Krugman to offer even this genteel critique, with its implication that Barth’s approach may be, um,  lacking in sanity, means that extremists like Barth are becoming increasingly isolated. 

Barth isn’t just trying to redefine evidence-based.  When it suits him, he tries to redefine foster care. 

Consider this bizarre 2021 exchange on Twitter between Barth and Prof. Alan Dettlaff: 

Dettlaff tweets:

 In April 2020, Cornelius Fredericks threw a sandwich at another resident of the institution where he was placed. He was then killed by staff members who restrained him until he could no longer breathe. This is what happens to children in foster care. 

Barth then replies: 

This is tragic but did not occur in foster care. (This was a residential center.) So it can't fit your sloganeering about what happens to children in foster care. Lazy thinking amd imprecise language wont help us serve children better. 

Hey Prof. Barth, have you told the federal government?  All these years, they’ve been under the misimpression that residential treatment is foster care!  For some reason, the federal database known as the Adoption and Foster Care Analysis and Reporting System has a whole category for “institutions” Similarly, this excellent database from ChildTrends includes “Group Home or Institution” under “Placement settings and stability for children in foster care” [Emphasis added.]  

And apparently, the entire federal government got its regulations wrong, too.  Because Federal regulations define foster care as 

24 hour substitute care for all children placed away from their parents or guardians and for whom the State agency has placement and care responsibility.  

So, Prof. Barth, tell us again about “lazy thinking.”

UPDATE, AUGUST 28

Sometime today, Prof. Barth responded to having his tweets criticized:

The tattletale factor in “child welfare”

It’s why all those plans for kinder, gentler family policing won’t do much good – and might even backfire.  Connecticut is a case in point. 

It seems like a week doesn’t go by without some “child welfare” agency announcing an initiative that supposedly will make family policing kinder and gentler.  

There’s the longstanding “differential response” in which, in cases deemed less serious, caseworkers are sent out to do a supposedly family-friendly “assessment” and offer voluntary help, instead of an investigation with the accompanying risk of onerous surveillance.  On the one hand, there are a lot of reasons to be concerned that this leads to needless net-widening. On the other hand, another in a long line of studies suggests it may reduce foster care entries. 

Then there are “family resource centers” – drop-in centers, in which families can stop by and get help, including the concrete help families really need. 

But for these and similar interventions there is one huge catch:  Call it the tattletale factor.  

All of these alternatives are either run by the family police themselves or contracted out to agencies whose workers are mandatory reporters of “child abuse” and “child neglect.” At worst, this can wind up widening the net of needless intervention into families.  At best, this creates a deterrent to families coming forward to seek help because they never know if the “helper” will feel compelled to call the family police.  Prof. Kelley Fong, who’s done extensive research on this issue, summed it up in an article for Family Integrity and Justice Quarterly: 

“Seeking help is no easy decision when those in a position to assist are also potential tattletales.” 

The tattletale factor persists because no state or locality creating these plans has taken the one step necessary to make them work: surrendering some of their vast, untrammeled power to intervene in the lives of families. 

And that brings me to something included in Connecticut’s plan for how to use funds under the federal Family First Act.  Here’s how The Imprint describes the plan: 

The state is planning to support a community-based organization that will help steer some reports away from its surveillance-oriented hotline, known as the Careline. 

Funding for this venture — described wonkily in the Family First plan as a Care Management Entity (CME) — has already been included in the state’s budget negotiations this year, and department leadership are confident it will secure about $1.5 million for this venture. 

“We want to go further upstream and serve families without having them become known to us … and really do prevention in a more robust way other than cases accepted,” said DCF Deputy Commissioner of Operations Michael Williams. 

The idea is that reports of concern that do not rise to the level of abuse or neglect would come to the CME from schools or child care centers, law enforcement officials or the community. Parents could also come directly in search of help. 

The effort is probably sincere.  It may have been inspired by Prof. Fong’s research, some of which was done in Connecticut.  And when it comes to reducing needless foster care, over the past decade or so, the Connecticut Department of Children and Families has shown some real leadership and done better than other states (although using the Orwellian term “Careline” to describe the number you call to reach the family police does not inspire confidence). 

But the closer one looks at the fine print, the less there is to Connecticut’s plan. 

For starters, the claim that this entity will “serve families without having them become known to us”
isn’t quite accurate.  Under the plan, mandated reporters will keep right on calling the policing line (no way I’m calling it “Careline”).  Then, if the family police decide the report doesn’t rise to a level that needs investigation, the reporter will be told to call the CME instead. 

So this means that a second entity will, in fact, go out and knock on a family’s door in situations where that does not happen now, thus reinforcing fears that Family First opens the door to making needless intervention into families worse. 

And while it’s true that, in theory, this new entity is only there to help, there’s another catch: According to the plan “CME staff are mandated reporters and if they hear something that meets the statutory criteria of abuse or neglect, they will refer to the [family police].”  Why, then, should families be less afraid of the CME than they are of any other mandated reporter? 

Prof. Fong describes parents forced constantly to weigh whether to reach out for help and, if they do, what they dare not say to “helping” professionals: 

In communities highly exposed to [child protective services], the specter of CPS accompanies families to doctors’ visits, to parent-teacher conferences, to homeless shelters, to therapy

Appointments … Parents in need face a no-win situation: close off opportunities for support or open themselves up to the risk of state scrutiny and family separation. 

Connecticut’s plan doesn’t change this. 

A far better approach would be a plan that said: “We will seek legislation exempting CME staff from mandatory reporting requirements.”  That doesn’t mean they’d be prohibited from reporting, only that they’d be free to exercise their professional judgment without fear.  But that would require the family policing agency to relinquish some of its power. 

A good plan also would have included a provision allowing “mandated reporters” to fulfill their legal obligations by calling this new entity instead of calling the family police.  But that, too, would require the family policing agency to give up some of its power. 

The fact that this proposal is part of Connecticut’s plan to implement the so-called Family First Act creates additional problems.  Family First will only help states pay for a service if the service is provided for families where children are, using the law’s own Orwellian term, “candidates” for foster care.  That creates even more entanglement with the family police.  As the proposal itself says: 

…the CME will be expected to conduct an assessment protocol for all families coming through the CME. This will include 1) an evaluation of safety; if there is a safety concern the CME will make a call to the Careline for further evaluation, and 2) an assessment of risks, strengths, and needs to inform case planning, and service matching… 

So how can this be purely voluntary, if the family must allow all that “risk assessment”? Will a family that says: “no, thank you, we’re not interested” automatically become higher risk?  And how does all this constitute serving families “without having them become known to us”? 

At best, this seems no different from differential response, at worst is it simply subcontracting child abuse investigations by another name? 

Why entangle this with Family First at all?  Surely the richest state in America could come up with the $1.5 million needed for the CME without federal reimbursement. [UPDATE: I now understand the state is, in fact, going to come up with that $1.5 million without Family First.] As for Family First eligible services, with the strings that attach to them, savings from reduced foster care would cover the costs of any services the CME might arrange. 

A better approach would be to take the money the state would spend on the CMEs and give it to community-based community-run service providers, no strings attached, and let them figure out how to reach out to families and help them – in other words, what happened in New York City, not by design but by way of COVID-19.  But that would require the family police agency to relinquish some of its power. 

Short of that, there is one thing that would make this plan far more reassuring: If the mandate for the CME included immediately linking the family to high-quality interdisciplinary legal representation. 

But that would require the family police agency to – well, you know.

New study is a damning indictment of the child welfare surveillance state. It’s also excellent journalism

Study shows how even when everyone wants the system to work, it leaves families worse off.

 There’s a new study out by Dr. Kelly Fong, now a professor of sociology at the Georgia Institute of Technology. Dr. Fong’s graduate work at Harvard included a study of mothers in Rhode Island who are scared away from seeking help due to mandatory child abuse reporting laws, something I discuss here.

 Dr. Fong has a new, even more extensive study out. For this one, she effectively “embedded” with the Connecticut Department of Children and Families, doing more than 100 interviews with caseworkers, family members and mandated reporters.  The result is an indictment of our entire overpoliced, surveillance-state approach to child welfare.

 

People will probably be tempted to read only the plain-language summary published by the Council on Contemporary Families. But I hope readers won’t stop there.  The middle portion of the full study, published in the American Sociological Review, where we read about some of the families caught up in the system, is great journalism.  It starts on Page 11 with the section called  “Gaby’s Investigation.”

 That case turns out to be the perfect illustration of why the child welfare system, in its current form, is fundamentally unfixable.

 As you read that part of the study, consider:

 ● This study was done when Connecticut was a rapidly improving system with good leadership.

 ● The mandatory reporter who turned in Gaby meant well.

 ● The investigation was conducted by a caseworker who was experienced, smart and empathetic.

 ● The amount of intrusion into the family was the bare minimum possible when agencies such as Connecticut DCF intervene.

And yet, the family still was left worse off.

 That is the theme throughout the study.  You can’t fix child welfare just by getting rid of the kind of caseworker who thinks it’s funny to wear a T-shirt emblazoned with the words “Professional kidnapper” – as happened in Arizona last month.

 In fact, the biggest lesson from the study is simply this: The notion that people working for an agency with the power to take away children can ever really “take off the investigator hats, and put on our services and support hats,” as one child welfare administrator in Michigan put it, is a self-serving myth.

 Child protective services is a police force.  Its primary job is policing poor people, especially poor people of color.

 The helping function in child welfare needs to be separated from the policing function.  And the helpers must be exempt from our failed system of mandatory reporting.