Suppose, hypothetically, after an extreme case of police brutality made headlines in New York City, the Mayor ordered the city’s Department of Investigation to investigate community-police relations. But lets suppose the method he chose was to have DOI single out the nine most horrendous instances of alleged police abuse and see what lessons there are to be learned about the police department as a whole.
No doubt, the resulting report would be filled with broad-brush condemnations of how police “often” attacked innocent civilians, “regularly” trampled on civil liberties and “routinely” abused their power. After all, case after case – among those nine cases – revealed those problems. Indeed, DOI might well suggest that the NYPD hire, say, 100 social workers to accompany cops and aid them in learning how to properly deal with the communities they serve.
And everyone immediately would realize why the whole report was absurd.
But of course it doesn’t work that way with child welfare.
So when DOI, on orders from the Mayor, examined nine cases in which children “known to the system” were killed and filled the resulting report with “oftens” “regularlys” and “routinelys” every news organization accepted it at face value. And everyone seized upon DOI’s recommendation that the city’s Administration for Children’s Services hire 100 more people with law enforcement backgrounds, typically former police detectives, to assist its caseworkers.
Of course, it didn’t help that ACS meekly accepted the report and its recommendations.
One can only imagine what DOI would have recommended had it gotten a very different set of marching orders. Suppose the Mayor had said: Canvass the organizations providing defense counsel in child abuse cases, and those helping birth parents advocate for themselves in such cases, and ask them for the nine worst cases of taking children from innocent families they’ve seen in the past year. Then examine each case and generalize from it.
The resulting report would be all about how ACS often tears apart families for no reason, regularly confuses poverty with neglect and routinely distorts information presented to the court. It would, no doubt include a series of recommendations to curb the power of ACS, not add to it. But, of course, no political leader is ever likely to commission a report like that.
A clear example of how the skewed sample biased DOI’s findings concerns its contention that ACS workers “often” labeled cases as unfounded when they should have been substantiated. Perhaps they do. I’m not aware of any recent objective data on that. But older data suggest the opposite. First, the one time a national study second-guessed these decisions it found that workers were two to six times more likely to wrongly substantiate a case than they were to wrongly label a guilty parent innocent. And in New York, statewide, a class-action lawsuit in the early 1990s revealed that 75 percent of the time, when parents appeal a finding of substantiation, that finding is overturned – even though it’s an administrative hearing that applies the same standard of proof as the caseworker from the child welfare agency. (NCCPR’s Vice President was co-counsel for plaintiffs in that lawsuit.)
The only way to know what a child welfare agency does often, regularly, and routinely is to examine a random sample of cases – and have the examination done by experts with a track record for objectivity, or experts whose biases balance each other.
That’s what the first head of New Jersey’s Office of Child Advocate, Kevin Ryan, did. He recognized that making recommendations based solely on reviewing fatalities inherently distorts the process and gives an erroneous impression of what kinds of mistakes are routine, regular and happen often. So, at NCCPR’s suggestion, he audited a random sample of cases involving families under the supervision of the state child welfare agency. He found that caseworkers routinely, regularly and often made mistakes – in all directions. He also found a series of problems beyond those workers’ control. (Now, Ryan has the unenviable task of trying to solve the problem; he runs the state child welfare agency).
Similarly, Leroy Pelton, professor of social work at the University of Nevada – Las Vegas, and one of the most distinguished scholars in the field, recently completed a case reading of a random sample of cases in metropolitan Las Vegas.
Among his findings:
“case plans ordinarily consisted merely of sets of demands: ‘Father will complete drug treatment program…parents will obtain housing, attend parenting classes…’; ‘Mother will apply for housing; maintain adequate income’; ‘Father will complete parenting classes’; ‘(Mother) will obtain stable housing’; ‘Mother will complete a six week parenting class…will obtain and maintain stable housing and employment’; ‘Parents will obtain and maintain stable and adequate housing’; ‘Mother will maintain stable and appropriate housing.’ All of the foregoing quotations are from case plans in cases in which the parents were homeless.”
The study found that DFS workers seemed to have little concept of the enormous emotional trauma inflicted upon children when they are torn from everyone they know and love, institutionalized, and then moved from placement to placement.
Prof. Pelton found that some common practices by child welfare authorities are “breathtaking in their insensitivity to the safety and well-being of children” and wind up “creating a child protection problem” where none existed before. The study also documented errors in the other direction: two cases where children were left in dangerous homes when they should have been removed.
Like DOI, Prof. Pelton also recommended some new hiring. But instead of former police officers, he recommended hiring housing counselors whose one task would be to help families at risk of losing their children because of housing problems to get the housing they need.
It’s amazing what you find out when you look at cases at random.
In fact, having experienced law enforcement officers as a resource for child protective caseworkers makes sense – and not only for the obvious reasons. Having people around who understand concepts like “evidence” can help exonerate innocent families.
Even where the entire job has been turned over to law enforcement, the results haven’t been anything like what backers of a take-the-child-and-run approach had expected. In one state where investigations in several counties are handled by law enforcement there was little change in rates of removal, or quality of investigations. The investigator in charge of child abuse cases in one county law enforcement agency even called up NCCPR to describe how angry he was at the lack of discretion given his officers by the state’s child abuse hotline. Whereas his officers could ignore obviously false leads, crank calls and efforts to harass people in criminal cases, the child abuse unit had to pursue everything sent from the state hotline, no matter how absurd.
But it does not follow that hiring more ex-cops should be any child welfare agency’s top priority. ACS already has 18 former police detectives available to help its caseworkers (two more positions are vacant). Hiring 100 more would cost several million dollars. There are lots of things ACS could do with that money that would do far more to keep children safe – like pouring those funds into rent subsidies, day care, and other concrete help.
But the larger problem with the report is simply that it is one more document leaving the false impression that the errors go only one way. That – plus ACS’ dutiful response – makes it one more message to caseworkers that they can take away every child in sight and suffer no penalty. The only time they put themselves at risk is when they leave a child in her or his home and something goes wrong.
It’s one more reason to expect that just as the surge in removals of children might be slowing down, it’s going to start up again.
And that means more devastated families and a higher load for caseworkers – and overloaded caseworkers are the root cause of most preventable child abuse tragedies. Not always, of course, but often, regularly, and routinely.