This month we examine an especially wrenching issue: When is it justified to remove a child from his or her home in the name of safety?

It should be obvious that this question invites two kinds of error: In one type of error, a child remains in a dangerous home, and harm occurs or continues. In the other type of error, a child is removed from a home when no danger exists, or when the danger is so small that it is outweighed by the harm of removal.

And such harm is not negligible: Our lead essayist, family attorney and author Diane Redleaf, describes one such case in her essay, and she offers data and supporting context to argue that in the United States, Child Protective Services can be far too quick to remove children even from untroubled homes.

Yet intervening in family life does seem obviously justified in other cases. Here to talk about the many ways in which the process can both harm and help are Professor James G. Dwyer of the William and Mary School of Law and Cato Institute Senior Fellow Walter Olson. Each will write an essay, and conversation will continue through the end of the month. Comments will be open for one month as well, and readers are invited to join the discussion.

James Dwyer’s response to my essay creates a false dichotomy between two “diametrically opposed” camps within child protection policy circles. He describes my “camp” as the one that criticizes child protection agencies for too “readily separating children” from their abusive or neglectful parents, while his “camp” criticizes child welfare agencies for trying to “make parents happy.” His characterization of my article runs off the rails in its very first sentence, however, when he describes these opposing camps as differing in their respective responses to “child abuse and neglect.”

Walter Olson’s useful reply to both of us puts my discussion of safety plans (child removals under agreements that are labeled as voluntary) into a larger context of negotiations undertaken in the shadow of the law. But this “shadow of the law” can be a more lawless place than Olson’s comments acknowledge. Parent who “agree” to leave their homes or let their children stay with relatives during investigations usually believe they have no choice but to accept such an option as the lesser of two evils. They agree to separation from their children only because a CPS investigator threatened them with taking their children to live with strangers. In this context, safety plans are even more coercive and more troubling than any of the agreements in Olson’s posited analogies. By contrast, safety plan negotiations take place in a vastly under-lawyered arena. Families subject to safety demands typically receive no prior notice of the claims of the CPS agency. They usually are legally unsophisticated, lack access to counsel, and are pressured to make separation agreements on the spot that will have life-altering impacts on their children. When a CPS investigator comes to a parent’s home and says, “Sign this agreement or I’m taking the kids,” a team of corporate lawyers or assigned counsel does not come to the home to review the documentation. And no neutral magistrate looks at the signed agreement to see it is fair or reasonable. The shadows of the law in this area are dark indeed.

Olson’s useful context and analogies don’t respond, however, to the debate that Dwyer wants to have with me, or the one I would prefer to have with Dwyer. Initially, it’s important to clarify that my article wasn’t about appropriate responses to abuse and neglect; it was about appropriate responses to hotline calls that allege abuse and neglect. Dwyer assumes that the allegations in millions of hotline calls are true—or true enough, given that Dwyer, like Judge Posner, also would allow hunches to suffice in the place of evidence because “suspicion can ripen.” Dwyer never engages with the primary question I posed: how child protection services should treat families when addressing allegations of abuse or neglect that have yet to be investigated.

Dwyer also pointedly criticizes me for presenting “merely anecdotal” evidence. But a demand for statistical evidence as to the incidence of a problem, before we agree on the terms of debate, presents a nearly impossible task for anyone, let alone for a civil rights lawyer confronting a “shadow of the law” problem as Olson describes it. Threats used to separate families during CPS investigations are not recorded in central databases so as to allow larger epidemiological studies to be undertaken. In fairness to Illinois and Pennsylvania, however, the two states with the most litigation on this issue, these states have been grappling with problems my article raises, though they have resisted statistical recordkeeping that might show that problem of baseless threats that cause family separations may be even worse than I fear.

Despite these limitations on any study of incidence rates, there are clues that the shadow foster care system has spread. A report from Texas found that nearly two-thirds of the children taken from their homes were placed as a result of safety plans. In the Dupuy litigation (part II), a CPS deputy director testified under oath there were 10,000 safety plan removals every year.

In the Dupuy litigation (part I), the federal court determined that there was a “staggering” rate of error in child abuse findings: CPS teams’ decisions that labeled children abused or neglected were overturned more than 75 percent of the time when those labels were appealed. The Second Circuit Court of Appeals made the same finding as to error rates in Valmonte v. Bane, 18 F. 3d 992 (2nd Cir. 1996) as to New York State CPS system. And a study found that workers are two to six times more likely to wrongly substantiate a case than to wrongly label one unfounded.

Instead of debating whether the problem I discuss is real, I expected to start with the central question of “how do we know when abuse or neglect has occurred?” and then move to a discussion of “what are the essential protections we want to provide to families while we try to determine the truth of a Hotline call?”

Medical analogies can often prove useful. In 2002, the American Board of Internal Medicine Foundation developed the broadly celebrated Choosing Wisely campaign, with a mission of “Advancing a national dialogue around avoiding unnecessary medical tests and treatments.” One concern was that antibiotic prescriptions could lead to antibiotic resistance. The campaign did not deny the value of antibiotics but called into question the judgment of those who prescribed antibiotics for a cold “just in case.” By analogy, CPS policy sets forth a range of choices as to how to use coercive interventions into family life on the assumption that children have an “illness” that their families caused or unreasonably failed to treat. Even when there is, arguendo, evidence to support a concern about a child’s health or wellbeing in their family, I am on the side that favors choosing wisely. Dwyer’s response is analogous to drug companies’ pressure for more “just in case” prescriptions before any doctor has decided the child has cold.

Medical analogies also clarify why discussion of “anecdotal” cases can create new understandings. Every epidemic, from Ebola, to AIDS, to mad cow disease, starts with someone noticing a first case and starting to look for treatments. Luminaries from Pericles to Hippocrates to Confucius have recognized that there is always a tradeoff in any well-intended intervention, and balance is important to achieve.

What may be most needed to start a meaningful discussion is a new and more appropriately tailored theory of the state’s proper role in protecting children. As Thomas Kuhn, author of The Structure of Scientific Revolutions (1970), stated, “New theory, however special its range of application, is seldom or never just an increment to what is already known. Its assimilation requires the reconstruction of prior theory and the reevaluation of prior fact, an intrinsically revolutionary process that is seldom completed by a single [person] and never overnight.” New theory defining appropriate state intervention might start with critique of how the current system gets it wrong when we make CPS workers fact finders, parenting police, and enforcement agents. Such a new theory might also help in setting reasonable limits on the powers that have been conferred on CPS workers to operate in the shadow of the law.

This theory cannot evolve from a blanket denial of family experiences. Dwyer’s suggestion that Hernandezes were not much affected for the worse by the 10-day removal of toddler Jaymz from his home, because Crystelle (his mother) got to stay with him (his father Joshua did not), is chilling. Take it from the Hernandezes: their lives were changed in dozens of ways, large and small, by the experience of having their child coercively removed, even for this relatively short time. Today, with four children born after Jaymz, they worry about CPS intervention every time any one of the kids gets a bump or bruise. What parent, having experienced the seizure of their child without probable cause, wouldn’t experience the same fear?

Dwyer’s lack of “alarm” as to caseworkers who have a “gut” feeling about abuse and who act on their instincts is consistent with Judge Posner declaration in Dupuy that “mere suspicion” should be sufficient to threaten a family with a safety plan separation as long as the state was not lying about the evidence it possessed at the time of the demand. But a parent who can imagine himself on the accused person’s side would immediately have a grave concern. If they knew the right legal questions to ask, they might ask how CPS’s gut-based actions square with their fundamental constitutional right to familial association. Should police be given a similar degree of free range to act on gut instincts and hunches (and if so, what’s the check under our rule of law on police entering homes and telling parents to leave because of such hunch?). What would we think of a medical system in which doctors were allowed to conduct heart transplant surgery based on the hunch that the patient might develop heart disease if they fail to act? Risk of future harm, even when that risk is as serious as child abuse, needs to be supported with “definite and articulable evidence” (in the words of the federal court in Brokaw v. Mercer County, 235 F. 3d 1000 (7th Cir. 2000)). And the intervention needs to be reasonably measured in response to the documented risk.

The challenge for advocates is to enunciate more clearly these standards for intervention in cases of alleged child abuse and neglect. Processes are needed to assure these standards are met. A good gut feeling or unsubstantiated hunch is an abdication of policy, not sound policymaking.

Prof. Dwyer makes a few more observations that also merit a response.

Dwyer argues against the likelihood that innocent parents are unduly coerced and threatened by CPS by looking at caseworker motivation, noting caseworkers are overworked and therefore are not looking for more children for their caseload. Caseworkers never go looking for cases, however. They are assigned to cases when referred by the hotline. This observation about individual caseworker motivation also ignores the systemic policy motivations that allow safety plans as a “safe” and low-cost response for the system, if not for the families. This response Olson describes as typical of systems that allow operation in the “shadow of the law” to flourish.

But Dwyer’s observation about caseworker motivation doesn’t answer the reasons safety plan policies have been allowed to run rampant; it only confirms that they are doing so. It’s easier for a caseworker handling an assigned hotline investigation to separate a family under a safety plan than to mount a case with evidence and take it court. While it may be “easier for the family” (and “make parents happy, ” in keeping with Dwyer’s own pet critique of the CPS system’s predilections) in the cases where there is solid evidence of abuse, this observation fails to reckon with the fact that it’s not necessary to have any evidence at all, according to Judge Posner, in order to threaten a safety plan.

Toward the end of his essay, Dwyer leaves the safety plan terrain altogether, reprising his complaint that the CPS system is “soft” on abusive parents, prone to allowing them to keep custody of their children. At times caseworkers do leave children in genuinely abusive homes. I don’t condone these terrible decisions, though they might be rarer than sensation-seeking media suggest. Rather, these cases are just one side of the same coin that lets errors of all types proliferate. Sometimes, the caseworkers who let children languish in homes after experiencing serious abuse are lazy, untrained, or incompetent. Perhaps the evidence at the time of state’s missed opportunity for intervention wasn’t clear. More probably, these caseworkers are too overworked to get to see the children who are in genuine danger due to the overwhelming caseloads that have resulted from “when in doubt, call the Hotline” policies. A system that allows caseworkers to operate on hunches when it separates children from their parents is the same system in which caseworkers leave children home in the face of danger merely because a parent seems “nice” or the home is spotless.

Despite his calls for reliance on research, Dwyer takes a detour to mischaracterize the research on “differential response.” Differential response is a system, like pretrial diversion, that takes hotline calls out of the investigation track and delivers social services instead. Contrary to his claims that this approach leads to more recurrence of maltreatment, the overwhelming body of research has found differential response to be safe and effective.

Despite our many disagreements, near the end of his essay Dwyer lands on one point of agreement—a point that, to my surprise, coincides with the enacted policy of the state of Illinois in a series of post-Hernandez lawsuits. He suggests that, if the state had strong evidence of abuse and instead of court action, it proposed course of treatment agreed to by the family, there would be nothing wrong with that from a due process and lawful state intervention perspective.

Dwyer almost has this point right. The alternative to a voluntary agreement in lieu of court action is fair, lawful, and reasonable if the constitutional standard for removing the child has been met first, before a family separation is demanded. In other words, for a safety plan to be lawful, evidence sufficient to impose such a plan involuntarily must be present.

Such a substantive protection against unwarranted coercive threats would be a good start. But even such protection would not suffice to afford fair procedures to assure adherence to the substantive requirements. There must be an available review of the basis for the state’s demand for such agreements, even if the parent agrees to the agreement’s terms. Agreements can too readily be labeled as voluntary if there is no process available to review them on the merits. Parents also need access to counsel to advise them if the agreements and terms under which they must separate from their children, even if the State were to meet a substantive burden of proof, are fair. This area of social control and family life needs to come out of the shadows of the law and into the light.

In law, as in child welfare, there is no alternative to a system that operates under a reasonable burden of proof, even if that burden is mere “probable cause.” Before children get taken from their parents, shouldn’t we be insisting that the state must have more than a subjective hunch that the parents have mistreated their children? And shouldn’t we be discussing how to protect families from terrible harms that we can’t yet adequately measure since we haven’t been paying attention? Shouldn’t we be discussing how to bring light to the shadow foster care system and the CPS practices that allow this system to operate outside the law? These are the debates we should be having.

By Walter Olson
Response Essays

Walter Olson finds that the problems with Child Protective Services are of a kind with problems that affect many other agencies of government. He draws readers to consider these similarities and proposes several ideas that aim at reform. On many of his points, transparency and accountability emerge as key considerations. Bureaucrats seldom want these, but bureaucracy quickly grows abusive without them.

Let me start with the big picture issues, where I think Diane Redleaf and I agree, and then turn to the difficult issues of interim CPS separations concerning which I learned a great deal from her essay, but on which I harbor some differences and reservations.

For as long as I have been around them, libertarians have been trading outrage about the ways Child Protective Services can snatch children away from parents. Such raids can be a life-changing infringement of some of the most prized of personal rights. They can involve an expensive legal fight against state-hired personnel and experts whose contentions a judge might naturally find plausible.

Armed with such power, some CPS agencies combine the raid-first ask-questions-later ethos of a bad law enforcement agency with the nosy paternalism of a bad regulatory agency. Over nearly twenty years, at my website Overlawyered, I’ve gathered CPS horror stories, a genre at which our friend Lenore Skenazy (of Free-Range Kids) so excels: the parents tripped up for letting kids stroll home from the park, walk the family dog alone, stay in the back seat of the car on a mild day while mom dashes into the store for a carton of milk, or simply play outside; the journalist mom’s joke tweet (“3-year-old for sale. $12 or best offer”), the kid held for an unexplained cough, and the criminal suspicions raised by dubious hair-strand testing, the use of perjury and false evidence, the perils of parental consumption of poppy seed bagels, let alone marijuana,living with a boyfriend who has a criminal record, and so forth. It’s not just the United States, either: in the United Kingdom a CPS inquiry might hinge on parental obesity or kids’ having been allowed to play grown-up computer games, and we haven’t even gotten to Scandinavia.

To be fair, CPS employees must work under a set of incentives shaped by headline-grabbing bad outcomes. Policies can lurch back and forth, driven by the latest failure: a child dies in parental custody, so public opinion demands more and earlier raids; then a child dies in a foster care setting, and the cry runs the other way. The next death may occur in grandparent care, or in a group home. When CPS managers vacillate and hold back from any definite final placement or resolution, they may saddle kids with one of the most dangerous risk factors of all: time spent in the system.

Even assuming a strong presumption against intervention in intact families that requires a convincing showing of likelihood of serious injury from not removing a child, we are still left with a lot of unanswered questions about procedure and the role of courts. Which brings us to the distinctive subject matter Diane Redleaf raises in her essay: what sorts of policy response should apply to agencies’ practice of proffering to parents ostensibly voluntary interim placements and “safety plans”? What happens when parents regret—the next month, or the next day—having agreed to those conditions? Can they reopen the concessions they made, and how? Does it matter whether the agency has withheld information from them or menaced them with worst-case scenarios?

I want to start by stepping back and taking note of the way the issues raised here turn up not just in disputes between CPS and parents, but across area after area of law and policy. Generically, they are issues arising from negotiation in the shadow of the law. Some examples:

  • A criminal defendant, half scared to death by a prosecutor’s threat of the long sentence that might be handed down if all charges are proven, agrees to a plea bargain. Since the case ends, no jury ever hears the evidence, which may have holes.
  • A business charged with white-collar crimes, or maybe not even charged yet, accepts a deal proffered by the U.S. Department of Justice for a Deferred Prosecution Agreement (DPA) or Non-Prosecution Agreement (NPA). It agrees to pay the federal treasury hundreds of millions of dollars and submit to years of oversight by a federally appointed monitor, who will review and approve decisions that companies ordinarily get to make for themselves. Since the deal is voluntary, a court never gets into the strength of the evidence.
  • A university, bank, or pharmaceutical manufacturer signs a letter agreeing to adopt practices that are in no way required by federal law, after closed-door meetings in which emissaries from a federal department with oversight authority make known their feeling that these would be good practices to adopt. Pushing back is considered hazardous because a determined regulator can almost certainly find ways to make life unpleasant for one of these entities.
  • Lawyers sue manufacturers claiming that a certain class of vaccine or medical devices cause a side effect. One company, more timid than the others, agrees to a settlement paying tens of millions of dollars. The other defendants go to court and win decisive victories: the law says there was no causation. Too bad for the company that settled; settlements are forever.
  • Idealistic lawyers sue public social service agencies in search of court decrees cleaning up agency bad practices. The case settles and the resulting consent decree requires the agency to get permission from the lawyers before, say, changing certain policies or dropping budgetary commitments. Twenty years later, the private lawyers are still using the bargaining leverage of the consent decree to obtain concessions on unrelated issues.

In short: courts enforce settlements. They often do so even when the settlement is a hard bargain and even when its results stray from the appearance of equal justice. They do so partly out of self-interest (if it were too easy to wriggle out of old settlements, courts would be full of people trying to refight old cases and would have less time for the new). They do so partly out of an ideology that values contract, although settlement is treated as final even in countries with less respect for contract than ours. Yes, of course there are some exceptions—when the proffering party has obtained the settlement by fraud, for example. But the brilliant-but-quirky Judge Richard Posner, and his Seventh Circuit colleague Judge Frank Easterbrook, weren’t just off on some weird contractarian jag when they ruled in the DuPuy case that by and large parents are stuck with the settlement bargains they reach with CPS, even if they were under pressure when they settled, and can’t ask for free-form reconsideration based on all the individualized facts and circumstances.

Yet that is not the end of the story. In recent decades legal scholars, including many of libertarian bent, have stepped up attention to the once neglected issues of bargaining in the shadow of the law. Everyone realizes that most criminal prosecutions end in plea bargain rather than trial, and most civil lawsuits in settlement, and many major regulatory outcomes from the use of guidance documents, informal negotiations, and so forth. In public law the use of agency consent decrees has come under more scrutiny as well.

And if I may oversimplify, these varied lines of scholarly inquiry often wind up echoing much of the critique of CPS practice that Diane Redleaf presents:

Leverage. The side making the demands can deploy fear of extreme or unpredictable outcomes or prolonged legal process to obtain concessions a court probably would not or even could not have imposed.

Opaqueness. Settlements get negotiated away from public view, without the procedural sunlight afforded by open court and written judicial opinions. It is harder for outsiders to know what outcomes are typical, or how wide the range of variance is.

Entanglement over time. If it likes, the side making the demands can sometimes obtain for itself an ongoing, perhaps indefinite supervisory role, even in situations where judicial resolution would tend to afford a once-for-all clarifying ruling, such as a line between guilty and not guilty.

Bypassing court review. Key problems with the demanding side’s case, from quality of evidence to procedural irregularities, may never come before a judge.

Many of the problems Redleaf identifies are versions of these. CPS separations occur “in secret and without fanfare.” Statistics on them are not maintained “in any reliable form.” And so on.

So how do we get a handle on this? My first sense is that attacking the settlement process as such is doomed to frustration: the window for finding duress is awfully narrow. (E.g., “Caseworkers are trained to label safety plans as ‘voluntary agreements.’”)

Instead, obligations to respect rights can be imposed directly on agencies without need to scrutinize the give-and-take of the settlement process. If an agency fails to develop probable cause of abuse, and lacks exigent circumstances or a court order, it must release that child, period.

On the issue of opaqueness, higher-ups such as legislators and governors can impose reporting requirements. Settlements that go beyond the underlying statute in curtailing parents’ rights, entangle a family in obligation for more than a certain length of time, or are considered unusual or extreme for other reasons, can be kicked upstairs to require approval by agency brass.

Safeguarding every family’s rights will, as one of its benefits, shore up families against unwise surrenders of their rights.

By James G. Dwyer
Response Essays

James G. Dwyer says that Child Protective Services may already err in favor of the parents. Child separations are never an easy call, but the incentives at hand, Dwyer argues, are likely already arranged such that in general children are not removed from a home more often than they should be. Indeed it’s quite possible that the opposite is true, Dwyer says.

Debate about child protection policy among scholars and professionals generally reflects two diametrically opposed views about state child protection services (CPS) agencies. Some see the agencies as too readily separating children from parents in overreaction to reports of abuse or neglect, whereas others see the agencies as more concerned about making parents happy than about children’s welfare. Diane Redleaf is clearly in the former category; from her perspective, CPS is “treacherous and error-prone,” “continuously separat[ing] children from their parents,” operating “in secret,” routinely trampling “family rights.”

As is often the case when starkly different views of the same phenomenon persist, there is at least some basis in reality for both views. Redleaf offers anecdotes in support of a claim that CPS caseworkers are systematically and without cause coercing parents into sending their children away to live with relatives following maltreatment reports. Instead of walking away or giving parents a procedurally proper hearing by initiating formal judicial proceedings, caseworkers use threats to induce parents to accept a “safety plan” that entails temporarily placing the child with relatives.

Argument by anecdote is a dangerous approach to advocacy. An opponent typically can easily pick apart the anecdotes or offer up contrary anecdotes. And many people are susceptible to being persuaded by such argument even though it is illogical to draw inferences about systemic problems from isolated examples. What one wants in this field is data presenting a picture of the system as a whole, so that any policy changes respond to sufficiently widespread phenomena rather than to anomalous occurrences. Are there a large number of cases in which CPS separates children from parents without sufficient reason?

Focusing on a few specific cases actually creates an impression that the answer is “no.” Particularly so when one omits facts that might spoil the impression one is trying to make. In Redleaf’s case in chief, Hernandez, the agency’s basis for removing from parental custody a child who suffered an arm fracture was not just the mother’s inconsistent statements about whether the child was walking yet (a fact that a wrist fracture would not change), but also the facts that “Jaymz had an unexplained, older bruise above his left eyelid,” “the parents gave conflicting reports about who was home at the time of the incident,” and “the parents denied that anything was in the crib, but [the caseworker] observed objects in the crib.” Hernandez ex rel. Hernandez v. Foster, 657 F.3d 463, 477 (7th Cir. 2011). The courts ultimately concluded that the initial removal was reasonable, in light of those facts, but that the child should have been returned to parental custody after medical professionals concluded that the injury was consistent with an accidental fall from a crib. Id. at 478-83. The case does stand as an example of CPS wrongfully coercing parents into accepting a safety plan that entails having a child stay with relatives. It was wrongful because the agency threatened something it was incapable of doing in that case (placement of the child in formal foster care with strangers). A fuller picture of the facts, though, makes the agency’s actions more understandable. There was insufficient evidence to bring the case to court, as is often true even when child abuse has occurred, but there were several red flags that made the agency wary about closing the case.

Absent statistics about how often agencies are coercing parents with empty threats of non-kin foster care, it would be helpful to have a clear picture of how caseworkers and other players in the child welfare system operate, what incentives they have, and what precise legal rules constrain their actions and decisions. That picture, along with some anecdotes, might support an impression that it is happening systematically, as Redleaf speculates. But even if so, the proper response might not be simple or clear, because any alternative rules or practices might raise other concerns—in particular, leaving too many children in danger or forcing agencies to take more drastic action.

We know CPS agencies are generally overburdened, so caseworkers are not roaming about hoping to find more children to add to their caseload. We know the actual rate of maltreatment exceeds the rate at which CPS receives reports of maltreatment, and it greatly exceeds the rate at which CPS is able to substantiate reported maltreatment. So it is implausible to suppose that many CPS intrusions into family life are entirely unwarranted. Softer responses to maltreatment reports have developed in part to minimize CPS workload and in part to minimize the trauma for children and parents following a report. One of those softer alternatives is a safety plan parents can choose instead of going to court for formal hearings and adjudication of abuse or neglect charges against them.

As I discuss in my new book Liberal Child Welfare Policy and Its Destruction of Black Lives (Routledge 2018), there is ample reason to fear CPS caseworkers will too often choose soft responses in cases where they could reach a finding of maltreatment, seek an adjudication in court, and take stronger measures to protect a child. They might do this to avoid upsetting parents and to spare themselves from having to prepare for and attend court hearings, find and monitor foster parents, and so on. And the result might be to leave children in danger (e.g., because relatives with whom a child is placed will not adequately protect the child from parents or are not themselves adequate caregivers). Objective studies of “Differential Response,” the latest policy innovation in child protection practice, whose basic thrust is to offer parents voluntary services rather than to do formal investigations, show that CPS agencies in many states have diverted a shockingly high percentage of maltreatment cases onto the soft “assessment track” rather than the more formal and coercive investigative track, with the result that more children suffer a recurrence of abuse or neglect.

Redleaf’s concern is the opposite—that caseworkers are inclined toward a heavy-handed, intrusive approach, always seeking to prolong oversight of parents and separation of children from parents, using devious informal means when unable to initiate formal proceedings. She does not suggest what motivation caseworkers or their supervisors would have for routinely doing this. One motivation might be the noble one of protecting a child when their gut rather than hard evidence tells them the child is in danger. I am not certain we should be alarmed about caseworkers acting on their instincts in such situations, though their instincts are certainly fallible. Another motivation might be more self-serving— avoiding lawsuits should they return a child to parental custody and the child is subsequently seriously maltreated. But such lawsuits are far less likely than suits by parents of the sort Redleaf recounts, charging CPS with violating parents’ rights. Whereas parents are generally able to ask a lawyer to sue on their behalf when they feel wronged, children generally are not. The lawsuits parents have brought have clarified the law for agency directors and caseworkers, so that wrongful continuation of state custody and coercion of parents into “voluntary” safety plans should no longer occur in the jurisdictions where courts have issued decisions like that in Hernandez.

In sum, though Redleaf might be correct in suggesting there is today, despite rulings like Hermandez, a widespread problem of illicit coercion of parents, she does not document that or make it seem plausible. Nor does she make clear when, in her view, coercion is illicit rather than appropriate. If an agency could make a formal finding of maltreatment, secure a court removal order, and place a child in foster care with non-relatives, conditioning reunification on parental compliance with a plan of treatment and services, is it wrong to offer the parents a more informal approach predicated on their acceptance of temporary placement with relatives? Redleaf fails to address what the alternative would be to the practice of giving parents such a choice. Would it be to require in all cases formal investigation, court proceedings, and a permanent record of findings—in other words, elimination of softer responses? If so, would that be better for parents on the whole?

More importantly, which regime is best for children? Redleaf devotes little attention to the child welfare impact of the practice she describes. Was Jaymz Hernandez harmed by being placed in the custody of relatives for ten days, for eight of which his mother was able to stay with them? Is there evidence of other children incurring harm of any sort as a result of coerced safety plans? As with others who criticize CPS for being overly intrusive, Redleaf manifests a parent-focused view of the child welfare system. And though interests of children and parents are intertwined and overlapping to some extent, they are not entirely unified in any family. Simplistic assertions that children belong with their parents, or suffer when separated from parents, are not helpful to adjudication of individual cases nor to policy making. Direct, sophisticated, research-informed attention to children’s developmental needs and experience is essential.

By Diane Redleaf
Lead Essay

Diane Redleaf describes how Child Protective Services operates, using one particularly troubling case as a focus. Yet millions of children are named as suspected victims of abuse every year, and many of these end up being removed from their families, whether temporarily or permanently. Redleaf describes how parents may be subject to strong tactics to secure removal—and how these tactics can easily affect the innocent.

They Took the Kids Last Night is the title and opening line of my just-published book (Praeger, October 31, 2018), drawn from over 30 years of helping families navigate a treacherous and error-prone Child Protective Services system (CPS).

CPS caseworkers continuously separate children from their parents at a monthly rate 300 times greater than the number of the separations at the Mexican border that took place in May 2018. These separations occur day in and day out. They commonly happen in secret and without fanfare. CPS caseworkers take children from their homes, their schools, and from hospitals if the children have been taken for medical care following an injury or medical condition. Children are also separated from parents, and sometimes from their siblings too, at CPS offices and in juvenile/child protection/dependency courthouses.

How It Happens

In 2015 alone, over 7.4 million children were named as suspected victims of abuse or neglect in Hotline calls. Massive bureaucratic triage systems screen out 40% of these calls and send the remaining allegations to field offices for investigation or alternative community responses. Then, CPS investigative caseworkers make findings of abuse or neglect in approximately one-fifth of the screened-in calls, meaning that some 650,000 children a year are labeled abused or neglected at the conclusion of a Hotline investigation. These labels then get registered in a state-administered child abuse register, based on sometimes minimal levels of evidence (“credible evidence” in many states). This is determined by a caseworker and almost never by a judge.

During or as a result of this investigative process, in 2016 alone, 273,539 children were placed into foster care. But much less attention has been paid to the untold number of children who are the subject of coerced “voluntary” separations of families are a common result of CPS Hotline investigations in as many as 37 states, and such separations may constitute an equally large number among those that were initiated directly by CPS. Unfortunately, statistics on these informal separations, never ratified in a court of law, are not maintained by state or federal agencies in any reliable form. One of the policy issues my book discusses is why safety plans have come to be so pervasive and so tolerated by a country that claims to treat family rights as sacred and the rights of families as fundamental.

It is time to reexamine CPS family separation policies. The public has recently been served a large dose of the visible trauma that children and families experience when they are separated by force. Yet vocal expressions of public sympathies have not reached the parents of children taken into foster care, nor much to those children themselves. But even parents who never have been charged crime and whose sole fault is bad luck can and do lose their children to CPS. While the child protection system operates disproportionately to separate families of color, no one is immune. Indeed, the primary story the book tells, in unfolding day-by-day narrative detail, involves an unexplained fracture that suddenly was found one night in the leg of newborn grandson of a federal judge.

If CPS caseworkers take children from their parents into state or local CPS agency custody, a formal legal process is required, both to authorize the children’s removal from their homes and to transfer legal custody to CPS. Once it obtains custody, CPS places children into substitute care (usually a foster home, including with a relative, but also into group homes and residential facilities). Timely court orders—no later than five days after an emergency removal—are constitutionally mandated. To get these court orders, every state has a process that triggers judicial oversight, ongoing formal assigned casework responsibilities for the CPS agency, and federal funding if the child qualifies for federal aid.

But CPS systems have improvised clever tactics to avoid ever presenting a legal case to a judge when they remove a child. There is a long history in many states of using so-called voluntary placement agreements that explicitly aim to avoid the requirement of court orders. In more than thirty states, policies and practices authorize even more informal separations, which are agreements in name only. Innocuously called “safety plans,” these devices are utilized to circumvent the duty to afford due process to the families from whom children are taken.[1] Safety plans are usually oral directives, issued to parents following a CPS Hotline call, that require parents to give their children over to relatives under threat of forced removal and placement with strangers. These separations are labelled as “voluntary” even when parents are led to believe that entrusting their children to a relative is the only way they can see their kids at all. This informal separation system, operating under the unfettered authority of CPS caseworkers to fashion onerous demands and restrictions upon parents and other family members, creates what I call a “shadow foster care system.”

Litigation, news reports, and research into these practices has shown them to be prevalent and pervasive, especially in Illinois, Pennsylvania, and Texas (where one-third of the children in the child welfare system were separated from their parents without legal process and under a form of so-called “voluntary” agreement). CPS spokespeople (including CPS attorneys) have defened safety plans as a lesser intrusion than formal separations and placements into foster care, but have generally avoided answering the question of whether there is a legal basis to demand family separation in the first place.

A Troubling Case and an Important Precedent

Multiple attempts to challenge the exercise of coercive state authority through safety plan demands have been made in Pennsylvania and Illinois, with mixed but growing success. The case of Crystelle and Joshua Hernandez, which established the constitutional standard for the lawful taking of children from their parents in the Seventh Circuit in 2011, is one of the six stories my book narrates in detail. What happened to the Hernandez family could have happened to any family in America. In 2009, when Jaymz Hernandez was 15 months old, his parents Crystelle and Joshua had just moved back to Illinois from Texas. Joshua was an Iraq war veteran and Crystelle was a stay-at-home mom. On September 8, 2008, after living in Illinois for all of three weeks, Crystelle heard, through a baby monitor, that Jaymz had started crying after his nap.

Crystelle ran to Jaymz’s room and she saw him on the floor of his room, holding his arm. She naturally concluded he must have fallen out of his crib. The Hernandezes had no idea Jaymz could climb out already. They quickly lowered the mattress to its lowest possible position so that he couldn’t get out again.

New to the area, Crystelle took Jaymz to an unfamiliar doctor’s office to have his arm checked. Jaymz turned out to have a torus fracture of his right arm. During their visit, the nurse noted Crystelle had said Jaymz couldn’t “run or climb.” If Crystelle had said such a thing, what she had meant was that Jaymz couldn’t run or climb since the fall. Unaware of nurse’s note until much later, Crystelle had no chance to clarify the point at the time the note was made. A minor scratch above Jaymz’s eye and confusing responses as to whether Joshua was home at the time of the fall led the doctor’s office called the CPS hotline, viewing the statement that Jaymz couldn’t run or climb as a “red flag” for possible child abuse.

A CPS caseworker came to the Hernandez home that afternoon. She saw Jaymz walking, playing with toys, and interacting with his mother as if nothing were wrong. While she reported to her supervisor that Jaymz looked fine and that she didn’t know why she had been sent there, the CPS supervisor directed her, over the phone and without ever seeing Jaymz, to take the toddler into state protective custody.

Fortunately, the CPS caseworker placed Jaymz with his great grandmother, though the toddler barely knew her. And the CPS caseworker forbade Crystelle and Joshua from seeing their son while he remained in state protective custody. Luckily, the next day Crystelle was allowed to attend Jaymz’s orthopedist appointment to cast Jaymz’s arm. The doctor told the CPS caseworker that it “didn’t look like abuse.” Indeed, the doctor said that abuse could not account for the type of fracture Jaymz had; the break could only have occurred from a fall.

Later that same day, the local State’s Attorney decided that there wasn’t enough evidence to file a petition in the local juvenile court to start a case to adjudicate Jaymz as an abused child or to take Jaymz into the formal foster care system. Under Illinois state law, the CPS agency’s legal authority to keep Jaymz in protective custody, and away from his parents, lapsed when the State’s Attorney decided not to bring the case to a judge. The power to hold a child away from his parents is an emergency police power. Nevertheless, CPS continued to hold Jaymz away from his parents. The CPS caseworker told Crystelle and Joshua that, despite that State’s Attorney’s rejecting the case, CPS “still had more investigating to do,” and they now had to sign a “safety plan.” Jaymz still could not come home and the Hernandezes could not see him unless they signed the safety plan agreement. The CPS caseworker also told them that they had no parental rights.

Naturally, Crystelle and Joshua believed the caseworker and signed the demanded agreement. The agreement allowed them to see their son, and even to stay overnight with him at the relative’s home in which he had been placed, as long as a relative supervised all of their contact with Jaymz.

New medical opinions continued to confirm the lack of child abuse. On September 16, 2008, Crystelle and Joshua sought legal help to bring their son home. After their lawyer implored the caseworker to end the restrictions, and as medical evidence in the family’s favor continued to mount, the CPS caseworkers ended the safety plan on September 18. Without such attorney intervention, plans often last 60 days or more while investigations continue, even when evidence of parental innocence mounts. The CPS investigation stayed open for another two months, however, closing finally on November 7, 2008, with a finding that the allegations of abuse against the Hernandezes were “unfounded.”

Terrible and frightening as it experience was, the experience

presents a clear and compelling set of facts for challenging the authority of the state to hold the child in custody without probable cause or exigent circumstances. The coerciveness of the CPS demand, coupled with the representation that the Hernandezes had no legal rights to see their son (and hence were compelled to agree) were glaring too.

Despite the strong evidence that the safety plan the Hernandez parents signed was involuntary, their federal civil rights lawsuit was initially dismissed in federal district court. The 2006 case Dupuy v. Samuels, with an opinion by Judge Richard Posner, held that a “mere suspicion” of abuse by a parent, without any evidence yet gathered to support the allegation made to a Hotline, sufficed for CPS caseworkers to make a safety plan demand that required parents to put their children into a relative’s care. Suspicion could “ripen,” and the hands of CPS workers were not to be tied by the need for evidence before they were allowed to “offer” a safety plan. Viewing safety plan choices as “options,” the Seventh Circuit declared that parents who were offered such safety plans could simply “call the bluff “of the caseworkers. The opinion pointedly compares parents’ options (the choice between having one’s children go live with grandma or be taken into foster care) to those of cocktail party guests first offered a martini or a Manhattan, but then told that the only choice was a Manhattan, concluding that even such limited “offers” made them “no worse off.” The opinion concluded that safety plans were voluntary. Indeed, using the cocktail party analogy and its language of “offers,” the opinion concluded that Illinois’ pervasive safety plan practices were so plainly innocuous that they did not reach the threshold showing of a deprivation of liberty sufficient to allow the plaintiffs to state any colorable claim for a constitutional violation impairing family liberty.

The Search for Better Options

No good host forces his guests to drink. Yet CPS caseworkers never offered parents the choice of “no thank you” when they were given the forced choice between putting a child with a relative or having them go live with strangers in a foster home. The Dupuy opinion’s cocktail party analogy was thus plainly flawed, as the massive factual record before the district court in the 22-day trial confirms. The evidence—never referenced in the Posner opinion—showed that not one parent ever called a CPS caseworker’s bluff. Every time that a parent was threatened with foster care if they did not agree to separate from their child by enlisting a relative to care for the child during the investigation, the parent “agreed” to the safety plan. The federal trial court had found express coercion was rampant in the Illinois safety plan regime.

Far from a voluntary choice of options, the choice between having a child live with a relative during an investigation or go into foster care (with the possibility of never seeing their child again) is akin to being offered a choice of “your money or your life.” Labeling such choices “offers” of “options” is Orwellian. Yet, in the regime of CPS, safety plans have been largely accepted as though they were voluntary options for the families under investigation. Caseworkers are trained to label safety plans as “voluntary agreements.” In the wake of the Dupuy decision, families face an uphill battle in arguing for any entitlement to due process as they lose custodial rights to their child.

Fortunately for the Hernandez family and the plaintiffs who followed them, there was a loophole in the Dupuy decision. As plaintiffs’ lawyers working to create a constitutional child welfare system for families, we mounted a series of cases that aimed to turn that loophole into a set of constitutional policies that would, if implemented as required, move the CPS system into a direction of affording due process to families under CPS investigation.

The Posner opinion had posited that families who refused safety plans were “no worse off” for having been offered the “choice” between taking their child into foster care or having the child live with a relative. The loophole arose because when the CPS caseworker demanded the Hernandezes sign the plan, their child had already been taken into protective custody. That made the so-called “options” more plainly coercive. Contrary to the assumption in the Dupuy opinion that the parents could “call the State’s bluff” and not be worse off, if the Hernandezes had exercised that “option,” it would mean their child would stay separated from them, and they would not be able to see him at all.

Ultimately, the Hernandezes won reversal of the Dupuy-based dismissal of their federal civil rights suit and secured an opinion that was more than a modest corrective to Dupuy. First, it clarified that the state bears an immediate duty to release a child from state protective custody as soon as it is clear there was no probable cause to believe the parents had abused the child. And while the Seventh Circuit refused to award damages to the Hernandezes arising from initial seizure of their children, it declared that, henceforth, taking children from their parents without probable cause, without exigent circumstances, and without a court order first, violates the Fourth Amendment. Several other circuits have articulated the same legal standard.

The Seventh Circuit also labeled the Hernandezes safety plan coercive, knocking a gigantic hole into the Dupuy opinions conclusory assumption that all safety plans, including those issued without any investigation and upon “mere suspicion,” are uniformly voluntary and thus work no deprivation of liberty.

Despite this landmark ruling, however, practices on the ground in Illinois have not changed significantly. Five subsequent lawsuits in Illinois successfully extended the holding of Hernandez v. Foster to the nonconsensual transfer of custody to another parent and the holding of children at hospitals. Still, policies and practices intended to clarify the limited circumstances in which a state can demand a safety plan are yet to be fully implemented. Safety plans continue to be demanded of parents in the absence of evidence showing any serious likelihood of abuse or neglect to the alleged child victim and absent emergency justification for separating the family without a court order.

Juvenile court intervention provides a necessary but insufficient correction to the abuses of power committed in the name of child protection. The Hernandezes story was the shortest of the separation stories told in They Took the Kids Last Night. One ordeal lasted 3½ years, and another required a trip to the appellate court before an unfair presumption of guilt for alleged Shaken Baby Syndrome was overturned. The Hernandezes never had to respond to a formal legal action, since the request to file a case against them in the juvenile court was rejected by the prosecutor. Nevertheless, the juvenile court practices my book examines in detail add to the challenges families face when CPS separates them from their children. Parental innocence of wrongdoing provides little protection in the face of pervasive policies that support taking children first and then shift the burden to the families to prove their innocence.

New federal amendments to the Social Security Act pursuant to the recently enacted Family First Prevention Services Act could make the problem of family separation without due process even worse. Federal matching funding soon will become available to support relatives caring for children involuntarily taken from their parents, without requiring any formal judicial review of the validity of the separation in the first place. While advocates like me hope for some stronger federal guidance to reinforce the principle that children cannot be taken from their parents under coercive threats, there is substantial concern about an ever-widening child protection reach into family life. While the Families First Act was intended to support families’ ability to raise their children and avoid foster care placements, it may have the unintended effect of encouraging more safety plan separations into relative care without due process.

Wider public attention and scrutiny to these operation of CPS, both in the formal foster care system and in the shadow foster care system that operates under the radar, is long overdue. If awareness of the trauma inflicted on families by ICE practices at the border leads to a re-examination of that agency and its actions, then perhaps a re-examination is also in order when American resident children, from Alaska and Hawaii to Florida and Maine are taken away too.

Coming Up

Conversation through the end of the month.

Related at Cato

Cato Unbound: Children’s Safety and Liberty,” February 2013

Latest Issues