Failure to Protect
homelogan marrcaseworker fileschild policydiscussion
interview: richard gelles
photo of richard gelles

Gelles, dean of the University of Pennsylvania's School of Social Work, is a child safety advocate who helped draft the Adoption and Safe Families Act (ASFA) of 1997. In this interview, he tells FRONTLINE that he is comfortable with the widening safety net -- even if it means that children are removed from their homes unnecessarily -- since there is a greater probability that children who are in danger will be protected. He is the author of The Book of David: How Preserving Families Can Cost Children's Lives (1996).

We're going to terminate parental rights in the past we wouldn't have terminated. We're going to remove children from a home that in the past we wouldn't have. And that is the price-tag for child safety.

[What was the Adoption and Safe Families Act?]

ASFA was an attempt to redirect the priorities of the American child welfare system. Now, in 1980, the Adoption Assistance and Child Welfare Act (AACWA) was enacted. Its goal was to prevent children from coming into the foster care system and languishing there. The term was "foster care drift." Kids would come into the system without a particular consistent threshold, and once they got into the child welfare system their only way out often was to age out, to turn 18 or 21. So that bill comes along ... and it says we're going to establish a threshold: ... Before a child can be taken out initially, and before a child can be kept out permanently, the state is obligated to make reasonable efforts. Now the second part of the initial bill also said the state was responsible for having a permanency plan for children once they had been in the system for 18 months. ...

The whole notion of reasonable efforts got lost, and the child welfare system engaged in every possible effort and that came to be a problem in the late 1980s when the caseload changed. They were younger children; they were coming out of multi-problem families; there were lots of drugs in families; there weren't a lot of services available for the caregiver. After the 1980 law, [the placement rate in foster care] had gone up, the length of stays had gone up, and worst of all possible things, the number of permanent placements, the number of adoptions, had dropped to under 20,000.

[It was a crisis?]

There was a crisis in the system and the system had been so used to just swinging its pendulum back and forth on a case-by-case, tragedy-by-tragedy basis [that] it really wasn't able to step back and look and say, "What got us into this mess?" And [U.S. Rep.] George Miller [of California], who was the father of the first law, said in testimony that preceded the drafting of ASFA that [he thought AACWA] ... was a good bill but it [had] not been properly implemented.

So ASFA was designed to properly capture the spirit of the Adoption Assistance and Child Welfare Act by making it very clear what the priority of child welfare was: the safety and well-being of children, not the preservation of families; and two, establishing timelines. You just couldn't open up the front door, have children come in and then make every conceivable effort really waiting for the parents to voluntarily terminate their parental rights. It wasn't fair to the parents, it wasn't fair to the caseworkers, but most of all it wasn't fair to the children. And for all of its problems and all of the criticism of ASFA, within four years the number of adoptions had doubled. Some people will argue that's not good. To me, since we know from research [that] adoption is the most permanent placement a child can have once in the child welfare system, I don't see it as bad. Failed adoptions really are not a very large percentage of cases.

[What about the criticism that ASFA's timelines are unrealistic?]

I don't think that timelines in principle are unrealistic or even that short timelines are unrealistic because you have to balance the family's willingness and capability of turning themselves around with the child's developmental needs. My favorite quote is, "It's almost immoral to hold a child's development hostage while you wait for parents to turn their lives around" -- if in fact they are going to be able to turn their lives around. What if it takes a parent five or six or eight years to turn their life around, with three or four relapses? Is it fair to ask the child to either stay in foster care for eight years with multiple moves? Or for the child to be some sort of guinea pig to test whether the change has been effective and thereby restore the child to the home [and] have the child re-injured, taken out of the home? ... You keep running this revolving door and the only way you know that you've failed is the child is injured or harmed.

Now, those people who look at the timelines as unrealistic, quite frankly, view their client as the parent. When we wrote ASFA and when we at this school train our students, we train them very carefully that the client is the child, that the child comes first. That's why the child welfare system exists. It wasn't created to provide more therapeutic resources for parents. It was created to provide a final ultimate safety net for children whose parents could not meet even the most basic minimal needs that the children had to develop. So I categorically oppose the criticism and I think that the criticism is misplaced and it really ignores children's needs.

[And to those who say that the child's best interests lie with his or her biological parents?]

Well, that's a wonderful ideological position but it doesn't get borne out with actual research. The vast majority of parents are quite capable, loving, wonderful, giving parents. They don't end up in the child welfare system; they don't end up being investigated by the child welfare system. The majority of parents who are investigated by the child welfare system are the most appropriate caregivers for their children; sometimes they're inappropriately investigated, sometimes they just need a little bit of help or a little bit of attention.

But there's a residual population who really don't love their children who are not capable of meeting their needs, who engage in acts of omission or commission, or are quite harmful to their kids. And the final arbitrator of this discussion is if you look at the research as to what optimizes children's development given that they come into the child welfare system. It's not staying with their biological parents. That simply is not the best placement for kids. Once they reach that threshold of removal, in fact, the children who do the best are the ones who have the least number of movements, who come in early and have few movements, who aren't recycled back into the home. ...

I cannot argue this more vehemently: If you look at the population of kids who come into out-of-home care -- 200,000 children a year or 550,000 children total -- the outcome for those children is not enhanced, not maximized, by returning them to unhealed, abusive, and neglectful caregivers.

[But aren't there kids in the system now who don't belong there? Is ASFA's net cast too wide?]

Well, first of all there's no evidence that ASFA has had that effect [of increasing foster care placements]. The only evidence of ASFA's effect is the increased number of adoptions. I don't think that children are being inappropriately adopted. I think that would be a hard argument to make. ... A couple of states have ... swung the pendulum more towards foster placement, but on a national basis that's not the case. ...

[What about the argument that ASFA encourages states to terminate parental rights?]

Terminations don't seem to be rising in any huge numbers. We've just begun a fairly substantial research project because there is the concern that terminations are differentially applied. Now, I've always been concerned that ASFA would be differentially applied to poor, low income, minority families -- and that really is a real risk. But to come back to the basic premise of the question, before you had ASFA you had less than 20,000 adoptions and you had a substantial number of children being re-injured or killed by being left in families that were dangerous. That occurred because the law made it very clear that you were supposed to err on the side of reunification.

When you err on the side of safety and your prime objective -- as AFSA says it is -- is the safety and well-being of children, by the laws of probability theory, you're going to make mistakes in terms of sweeping up children into the system who might not belong in the system and those children are almost always going to be from poor families, from minority families, from Spanish-speaking or non-majority-language families. That's always been the nature of the American residual social welfare, as well as criminal justice system, and there's no way to minimize both risks. You have to decide what mistake you're willing to tolerate.

When we wrote ASFA we knew that the mistake we were willing to tolerate was having the nets be not quite as fine and pulling people in who might not belong there. But the goal was to optimize children's safety and well-being and permanence. I can't say we've optimized safety and well-being, but I can say with some definitive data [that] we've optimized permanence for twice as many children as there were in 1996.

[So it's about choosing what sort of mistake you'd rather make?]

... You can make two types of mistakes. You can make a false negative, in which you see no risk and therefore leave a child in a home or return a child to a home and it turns out that there was risk, and more harm or even a fatality befalls that child. The other mistake you can make is a false positive. You can see risk and danger where there is none and inappropriately substantiate a case, inappropriately take children away from their caregivers, inappropriately underestimate the capacity of the caregivers to change. ...

From 1962 until 1997, essentially the system said, "We'll accept a certain number of false negatives because we don't want to tread upon parents' rights and inappropriately separate children from their parents." Since 1997, we said, "We've done such a poor job of protecting the children at the highest risk, we're going to focus our energies on them. We'll have to accept the fact that we're going to sweep in families who we in the past wouldn't have involved. We're going to terminate parental rights in the past we wouldn't have terminated. We're going to remove children from a home that in the past we wouldn't have. And that is the price-tag for child safety."

The price tag for parental liberty is a residual amount of child endangerment and death. That's it. You can't find a middle ground because from the basic rules of probability theory, there isn't one.

[The choice is clear to you?]

The choice is clear to me, I think, because I've been in the system for 35 years. The system was set up as a residual system to make sure that ... children wouldn't suffer at the hands of inadequate parents. And that was always the purpose of the system. The system was never designed to be a default welfare system, a default parental training system, a default drug abuse and alcohol abuse treatment system. That's not what it was set up for. It was set up to watch out for vulnerable children. That should be its prime objective. I guess I've become comfortable with the consequences of having a child sent to system. ...

[And again, what about the parents' liberty to raise their children?]

The constitution says really clearly that parents have the liberty interest to raise their children without government interference. The constitution in the form of case law also says that that liberty interest is not unlimited, that there are conditions in which government has an interest to step in when children's well-being and safety has been significantly compromised. That's etched into American law. Where we argue is where that line is to be drawn. There's no bright line legal standard for where parent behavior drops to the point where it begs for and demands government intervention.

[Are there enough adoptive homes? And is foster care sufficient?]

I'm satisfied that there could be enough adoptive homes and that there could be enough foster homes because ASFA should be prompting or motivating child welfare systems to make the decisions earlier. We get into trouble when we have 8-, 9-, and 10-year-olds who are still floating around in the system whose decision on their well-being should have been made when they were two. Secondly, and I think if you interview foster families they will tell you this over and over again, if we treated adoptive families better, if we treated and gave agency to foster families and listened to their concerns about the children in their care, we would not have a shrinking number. We drive people out of the foster care system by giving them nothing other than a check and "Thank you very much, but by the way, if you complain, we're taking the kid out." And we shrink the population of adoptive families because we say to them, "The child's not free [for adoption]." There's 100,000 children freed for adoption in the child welfare system, but they were freed so late that the interest of adoptive parents declined. ...

[Your portrait is somewhat at odds with what we've observed, with these systems in chaos and children subsequently getting hurt.]

... I think public deaths such as [those in] Maine, New York, Chicago, Florida, that get an awful lot of media attention short-circuit the normal way the agency works [and] swings the pendulum dramatically to the safety and well-being side. [It] maximizes what I call the false positive errors because the caseworkers will tell you, "I'm not going to wake up and read my name in the front page of the newspaper. I'm not going to be pilloried. I'm not going to be thrown to the wolves. Therefore I'm not going to make any more false negative mistakes." Well, it guarantees they're going to make false positive mistakes. ...

Fundamental, this is about better and good decision-making, and when you don't make [good] decisions, you simply knee-jerk the pendulum from one side to the other. Now there are states where if a child dies in foster care the pendulum will hard swing to the other side, say, "We got to keep these kids at home. The mother ... had a minor heroin problem. And the child was not being fed. We took the child out and six months later the child is dead. This was not a good thing to do. We really should have worked hard on mom." So they'll swing the pendulum hard back, freeze it on that side, until the mom they thought they could help turns out to be homicidal or her boyfriend turns out to be homicidal. This is a 50-year trend, and somewhere along the line somebody has to stand up and say, "Look, we need a better decision-making template because this knee-jerk short-circuit way of running 51 child welfare systems is dysfunctional at best."

[Talk about the effects of media coverage and public pressure.]

The departments have to restore the public confidence and the political confidence in what they do. ... The departments of child welfare unquestionably have the worst public relations units of any public agency in the world, including the defense department, and that's because they're reactive. Metaphorically they let themselves get up to their ass in alligators. They become satisfied that being a good alligator fighter is a qualification to be a good administrator and they lose sight that the job was to drain the swamp. Eventually, when the alligator bites their head off, they're replaced by a new alligator fighter and another new alligator fighter. But nobody ever steps back, or they rarely step back, and say, "You know, we really have to drain the swamp. And we're going to go public and tell people we're going to drain the swamp. And we're going to go public and tell people these are the mistakes we're going to make, and when we make those mistakes we're going to say, 'Yes, we understood that that was a risk, that was the price we were going to pay for doing what we do.'" ...

[What is mandatory reporting?]

Mandatory reporting was designed to require a specific number of professionals who had frequent contact with children to report suspected cases of abuse or neglect or sexual abuse to a designated agency, which became the American child protective service system, so that a more detailed, thorough investigation could be done. ... Some states decided that well, kids come into contact with all kinds of people so we'll make every adult a mandated reporter. Other states made it a more confined list of people. ...

[What has happened?]

Frankenstein. It's become a Frankenstein monster. ...

Because of mandatory reporting, we've developed a very cautious system that essentially says, "We will investigate the vast majority, if not nearly all, of the reports that come to us unless it's clearly outside of the law." And sometimes even if it's marginal to law, it'll be investigated.

So the front end of the system, the American child welfare system, is staffed and trained to do 3 million investigations a year, that yield 1 million substantiated cases of child abuse and neglect, that yield 500,000 cases that actually receive services. So 2.5 million contacts between the child welfare system and families are nothing more than investigations. But it's extraordinarily time-consuming; it's extraordinarily costly; it requires lots of technology, lots of training. Since resources are finite, it leaves very little energy, very little skill, very little money for the 500,000 cases that are left.

I am one of the outlaws. I believe, with a couple of other people, [that] the gate should not be so wide, [that] we were a little bit too aggressive in defining what is child abuse and neglect. Some things we define as abuse and neglect, quite frankly, we don't have services for. Even if we substantiate it, it's not like we have anything we can offer. And then we have this big wide gate and then we seal it or weld it open, and we let everything through it. No triage. We'll let triage happen after we invest 30 days, 45 days, 90 days in doing an investigation with all the collateral contacts and all of the medical contacts. The files get to be 24 inches wide, [and] for what? So that the case would be unsubstantiated, not receive services?

That was not the child welfare system that 35 years ago I envisioned and that 40 years ago Henry Kemp envisioned. We had a much narrower view of children and families that needed services. Mandatory reporting was designed not as the Child Welfare Worker Full Employment Act, but as a way of breaking down the reluctance of reporters to bring these kids to public attention. It really did become the Child Welfare Worker Full Employment Act because people aren't stupid. Social services agencies understood that the bigger the problem, the more staff they'd have. ...

There were 6,000 reports of child abuse [and] neglect in 1965. There were 3 million last year. ... Child protective agencies [used to be] 9-to-5 operations. They went out of business Saturday, Sunday, Memorial Day weekend, Labor Day weekend. Now they're 24-hour operations; they have cell phones, they have computers, they've got pagers, they have 800-numbers. The carrying capacity of the system expanded as the number of cases came in. Again, you could go to the legislature and say, "We got 100 percent increase in reports last year, so we need 100 percent more workers." Well the legislature would say, "We're not going to give you 100 more workers, but we'll give you 35." Well, that increased their carrying capacity, so they were able to take in more reports and more reports and more reports. ...

[Is it a culture of "ratting" on your neighbors? Is that how people are brought into the system?]

Well, what I've observed is that the more heterogeneous the community, the more reporting you get. Communities that are all alike, that all have the same privilege or the same oppression, are less likely to report on their neighbors. But you should pardon the stereotype, a white elementary school teacher is going to be more likely to report out an African-American child who comes into his or her classroom poorly dressed, not having eaten, with bilateral abrasions on both sides of the forehead. An African-American daycare worker in a community church that's next door to the home is probably going to be less likely to do that.

[What proportion of child welfare cases are classified as neglect vs. abuse?]

The reality is that half the cases that come in are neglect of one form or another. The kind of case that gets the most attention is sexual abuse. They constitute about 3 percent of the cases, and the significant physical abuse constitutes maybe 22-25 percent of the cases. The media cases are the rarer occurrences, the ones that people worry about.

The typical case is a neglect case where that bright line doesn't exist. ... Is it OK to run out and go to work and leave your 8-year-old at home to watch your 4-year-old? How much vigilance can a mother be expected to provide over her boyfriend who lives with her that keeps her from being on welfare? That's the typical case. The typical case is a soft case. I think the public thinks a typical case is a handprint on the side of the face with the college ring embedded into the cheeks so that you know immediately who the perpetrator was. That's really unusual. ...

[What problems do neglect cases pose?]

Neglect by definition [is an act] of omission. The difficulty is, is someone not caring for a child because they don't have the resources? Or because they have some constraint that's remediable -- impacted wisdom teeth, a health care problem that's not all that difficult to deal with? ... Or is this a deliberate act of omission?

There are many caregivers who have left their children home alone and isolated, wealthy communities who never come to the attention of the child welfare system, and they'll argue, "Well, I just ran down to get a pack of cigarettes." Do that in another setting, that act of omission becomes an act of neglect and it's investigated. Now, it's got a 55 percent chance of going nowhere, but that assumes the investigation itself is benign, and they're simply not. ...

[What do you mean investigations aren't benign?]

Investigations are what I call the third lie. ... The third lie is, "I'm from the government and I'm here to help you." Investigations are not attempts by friendly visitors that come to the home and meet the caregivers' needs. They are coercive activities designed to detect or uncover an act that could be criminal or ... could result in the removal of a child from the home. They bring out with them a scarlet letter "A," or a scarlet letter "N" that says, "You have been charged with abuse" or "You have been charged with neglect."

Now, frankly, you can call me a lot of names and you can say I do a lot of bad things, but when you call me a bad parent, you've gone to the core of who I am. And even if you don't substantiate my case, you're going to change my self-image, you're going to change my self-esteem, you're going to change my relationship with the neighbors who may or may not have seen your car pull up. ...

[Why shouldn't they investigate, though?]

Because they're also supposed to help. They're wearing two hats. They're bringing the velvet glove and the Billy club, and there's enough role conflict in the human service industry as it stands without having to add cop to it. So a number of us in this field have slowly come to the realization that if you get a police officer when you abuse your wife, if you get a police officer when neglect your 85-year-old mother, ... then you might as well get a police officer when the suspicion is that you've abused and neglected your child. They're going to be better at evidence-gathering, they're going to be trained in applying the legal parameters, and they're not there with any ambiguous hat that says, "Oh, by the way, I'm here to help you, but if you don't want to be helped I'm here to punish you."

[So how does that dual-role impact child welfare agencies?]

How can you accept that the caseworker is there to help you when they wave the stick around that if you don't accept [their] help [they're] taking your children away? You've got one week, two weeks, three weeks, four weeks to get your act cleaned up or this is what [they're] going to do.

I know that in the principles that we teach our students in this school of social work -- which is to meet the client where they are, to use values, to understand that the change process is slow -- to then give them that level of coercion is going to undermine their capacity and ability to establish a relationship with the family such that the family will accept and engage in the service or helping activity. How can they not look at this worker as anything other than an agent of social control who is eye-balling them and about to hammer them in court and say, "I've watched this mother up close and she can't do X and Y and Z, and she refuses to admit that such-and-such happened to her child. Therefore, your honor, I recommend that the state take custody of the child."

It's simple. I mean, you can walk through any street in Philadelphia that has a high proportion of child welfare cases and ask them, "What does the Department of Human Services do in the city of Philadelphia?" And their answer is, "They take children away." It doesn't matter whether you're in Maine or Philadelphia or Washington or California. The vision of child protective services is [that] they take children away. Nobody is going to volunteer, "Oh, they're really nice. They come and help me control my anger and they got me off of drugs and they taught me some really valuable parenting skills." Nobody says that.

[What about those types of services -- anger management, parenting skills, etc. -- that child welfare agencies can provide to families?]

The toolbox that the child protective service system has is pretty limited. Parenting classes being one of them, homemaking services, advocacy. But the toolbox is essentially set up to preserve families. ... You give a little boy a hammer, the whole world becomes a nail. Every problem, no matter how diverse or different, gets the same one-size-fits-all solution. Sixteen weeks intensive family preservation services, parenting classes, home visits, insight therapy, referral to drug treatment. It's a pretty standard package and it's given whether you're fighting with the worker and completely resisting the label that you've done anything wrong, or whether you say, "You're absolutely right. I did this." ... Both clients get the same services. Now, which one do you think the service might work for? ...

[Can those types of services work when they're offered coercively? When they're required in order to get your kids back?]

Well, change is difficult to accomplish for any behavior. We could take any behavior off the rack -- our weight, our height, our exercise, our diet. You don't just wake up one morning and change. ... Secondly, you don't bring about change by layering on more and more and more and more. One of the least disseminated findings from evaluation research in the area of child welfare services is the more services you provide, the less likely they are to be effective. ... At a certain point, you so paralyze the family with [all of these services that] they're not capable of having any of those things be meaningful. ...

The second issue is when the change is not forthcoming, the worker immediately goes to the hammer and says, "Now, keep in mind that if you don't do what I want you to do, I'm taking your children away." ... The best you can hope for in that circumstance is ... compliance. "OK, I'll jump through your hoops for you, I'll go to your damn parenting class, I'll sit there for 16 weeks." But it's not going to change anybody. Compliance isn't change. Lots of families have become very good at complying with the wishes of welfare like agencies. But that doesn't mean they've changed and the proof is in the recidivism rate. The recidivism rate for children returned to homes that have complied ... is appallingly high. [It] can run to 35-40 percent.

[Do you support those rehabilitation efforts?]

... I've gone as far as to say -- and it's an arcane answer to the question -- [let's] make Title IV-E, which is the largest federal allocation of funds to the states, a block grant. Rather than say you can only use it for the ambulance service at the bottom of the cliff -- foster care -- you can use it for any number of preventive services, secondary prevention services, or legal guardianship. ... The funding is so categorical that we don't have a diversity of services that we can offer. ... We got two things in [the] toolbox: either you take my ameliorative services or I put your child in out of home care. That's all I got. There's nothing else in there. And you're surprised that it doesn't work?

[Talk a little about the average caseworker.]

... They are not social workers. They have not gone through an accredited professional degree program in social work. That means that by and large they have a liberal arts degree; some even have just an associate degree. They have 20-40 hours of training and then they get a caseload. They may or may not know anything about child development. They may or may not know anything about child abuse and neglect. They may or may not know anything about human violent behavior. ... They are paid no more than $35,000 by the best of jurisdictions and as little as $21,000 by the most frugal. And then they're sent out to make some of the most important decisions that government workers could possibly make. ... I've just read a deposition of a supervisor whose training was, "I went to cosmetology school and then I worked for 11 years as a secretary to the director. And when the director of our county office left, I became the director." ...

I mean, they really try to help. But ... they are young, they're inexperienced. ... And they often become amateur psychologists. ... You hear them using the language in a nonsensical fashion. ... It's just an awful system. Who in their right mind would have life and death in critical decisions made by well-meaning individuals with this level of training, this level of resources? ... Really terrible way to run a system. ...

[What about the complaint that caseworkers take their cases to supervisors, who then create a narrative about a certain family's situation and act on it without even meeting the client?]

Supervision is a standard component of the system. ... The federal guidelines are that a worker should see a family in care once a month. We know from research that that's achieved only 50 percent of the time. So let's say the average worker has contact for an hour or so with a family, with a child, every six weeks. They take that slice of life back to a supervisor who hasn't seen a real case in years -- because the beauty of being a supervisor is you don't have to leave your office -- and they form that narrative. They shape it and form it and ... they'll fit it into a box. And one box will be, "This doesn't seem all that dangerous. Let's make sure that we get reunification." [Or], "This seems really horrible. Let's go for compelling reasons and a termination of parental rights." And, "This is sort of a gray area, but cases like this tend to go down this road. Therefore, this is how we'll treat it." ...

I wouldn't dare make decisions like that, that effect human beings, with that narrow and stereotypical slice of life. ...

home | introduction | the taking of logan marr | a national dialogue
caseworker files | child policy | producer chat
join the discussion | interview with the producers
readings, links & resources | tapes & transcripts | press reaction
credits | privacy policy | FRONTLINE home | wgbh | pbsi

web site copyright 1995-2014 WGBH educational foundation.