UP AND UP: The consultant was asked by DCFS to investigate the reasons behind this unprecedented spike in foster care numbers. HORNBY ZELLER ASSOCIATES

Twenty-two percent of Arkansas children who were placed in foster care in 2015 and early 2016 should have potentially been left with their families, representing a nearly 30 percent increase — amounting to at least 300 additional children — in “questionable removals” by the state’s child protection authorities, according to a consultant of the Arkansas Department of Human Services. That’s one of the findings in an unreleased, internal report prepared for DHS’s Division of Children and Family Services (DCFS) and obtained by the Arkansas Times through a Freedom of Information Act request.

Arkansas’s foster care system has seen record numbers of children entering state custody for well over a year, and in March DCFS requested that an external consultancy group, Hornby Zeller Associates, Inc., investigate the reasons behind the increase. This June, HZA delivered the results of its research. (HZA has completed a number of reports for DCFS in the past as well.) 


As of late March, there had been a 25 percent increase in the foster care population since January 2015; by May 31 that had grown to a 30 percent increase, totaling nearly 5,000 children in state care. In the several years before this spike, from 2011 through 2014, the state had maintained a steady range of around 3,700 to 4,000 children in care at any one time. But in January 2015, the numbers began to climb, and haven’t stopped since. While we’ve reported on pockets of this “foster care crisis” in the past — Sebastian County has long been an outlier — HZA found that the increase has happened across the state, with 18 percent more children coming into foster care and seven percent fewer leaving it.

Why that is the case is a complicated question, but it does not seem to be due to an increase in the rate of child maltreatment. The only thing that hasn’t increased in the 15-month period HZA studied (from Jan. 2015 to March 2016) was the percentage of investigations into child abuse or neglect that were substantiated, or found “true.” Instead, the report declares, “the increase in foster care is due largely to two factors: DCFS removing more children [from their homes] immediately upon investigation and the courts ordering removals against the recommendations of the agency.” That is, HZA attributes the increase to changes in “decision-making” at the agency and within the court system.


DCFS said it has “concerns” about the “conclusions drawn and recommendations made” in HZA’s report. In a letter that DCFS Director Mischa Martin wrote on Tuesday (and which the agency appended to the copy of the report it sent to the Arkansas Times), she pointed to methodological issues that could have impacted HZA’s findings, such as the overrepresentation of cases from Sebastian and Pulaski counties, which have long had higher numbers of children in care. (However, she conceded that the methodology was statistically valid.) She also criticized HZA’s “limited” focus on “subjective factors.” 

“The two broad issues related to decision making led to an oversimplification of the problems within the system as well as the potential solutions,” Martin wrote. DHS spokesperson Amy Webb said DCFS is currently working on its own, broader report investigating the surge in foster care numbers; Webb anticipates that report could be ready for release in the next few days.


The two main factors HZA identified — divisions within DCFS and the behavior of the courts — are combining to disastrous effect for families navigating the system. They are also helping lead to one of DCFS’s more nuanced and intermediate means of intervention “being phased out of DCFS practice,” the report says. One alternative to removing a child from a home and placing them in foster care is to instead institute a protection plan, in which the child remains with the family while DCFS works with the parents to address problems. But in part because of a 2015 law giving increased power to the courts in child welfare cases, the use of protection plans has declined.

(It should be noted that some conservative state lawmakers — particularly Sen. Alan Clark (R-Hot Springs) — have been criticizing juvenile judges for what they perceive as overzealous removals of children from their homes.)

To put HZA’s study in larger context, over roughly the same 15-month period that HZA was studying, DCFS has come under nearly constant criticism, sometimes from seemingly diametrically opposed directions. The agency has been faulted for failing to intervene in cases that led to severe child abuse or death; for mishandling a backlog of prospective foster parents, including relative caregivers, in the midst of a severe shortage of foster homes; for taking in too many children over non-abuse issues like drugs, unstable housing and incarceration; and, for allegedly removing some children from their homes without proper cause. Faced with this onslaught of criticism, DCFS has tended to explain problems by noting — accurately — that they’re woefully understaffed and that caseworkers are stretched too thin to meet all the needs of the state’s children.

National foster care reform advocates say these problems aren’t contradictory, but related: the same system that takes in more children than it can adequately care for is liable to overlook the cases where children really do need to be removed. And HZA’s report seems to suggest a similar conclusion. “If all of the removals are necessary, the only thing DCFS can do is increase the resources it needs to handle the additional workload,” HZA wrote. On the other hand, it continued, “If a substantial number of the decisions are inappropriate, a different set of solutions is required.”


Looking at 400 randomly selected DCFS cases, and speaking to several dozen agency staff and five juvenile court judges, HZA undertook to determine what had changed in the decision-making process to lead to an increase in “questionable removals.” The most significant reasons HZA identified for the increase fall into two broad categories: one, a faction of DCFS workers who view their role more as police than social workers, and two, local courts both overruling DCFS and (because of the threat of court intervention) prompting inexperienced workers to make recommendations against DCFS best practices and their own judgment.

In many areas around the state, DCFS has created standalone, specialized investigative units that, HZA found, saw “themselves as law enforcement officials rather than social workers,” adopting an adversarial role with the families they’re investigating and often requesting police backup on their visits — both of which escalate tensions with families and reliably lead to a greater number of children being removed.

These investigative units, which tend to be siloed off from the rest of the agency — even reporting to a different chain of command — seem to proceed from a different understanding than DCFS’s mission to be a support system for struggling families. Regular caseworkers told HZA that investigative units had a “low threshold” for removing children, whose parents often could have been dealt with instead by monitoring them in their homes with protection plans. As one caseworker told HZA, “We get all these kids assigned to us and we wonder why they were removed to begin with.” 
Regular caseworkers suspected that investigators sometimes chose to remove children because it’s easier than providing in-home services, and because the task of finding foster homes for children is passed off to regular caseworkers. The different units rarely met or shared information, leading to “substantial misunderstanding between the two units,” HZA found. The disconnect between the different parts of the agency was so substantial that, in one service area, the area director required that full-time investigators arrange for the first foster care placement of the children they removed, in an apparent attempt to prompt them to assess each situation more carefully. For their part, investigators expressed mistrust of caseworkers’ abilities to effectively monitor children left in homes. 

But other times, the pressure came from the outside, notably from courts demanding more removals. In July 2015, a new state law went into effect requiring court oversight of DCFS protection plans — the agency’s more moderate “in home” care plans. This also had the effect of further empowering judges to remove children from their families.

“The legislation changed the role of the court from one of ensuring that DCFS did not overstep its bounds in removing children to one in which the court became an independent actor,” HZA wrote. DCFS staff reported that this law has led directly to an increase in foster care numbers, with one area director telling HZA, “Judges hate protection plans. If we submit one to the court the judge will order that kid into foster care immediately.” Since the law went into effect, there’s been a sharp decline in protection plans across the state, including a 74 percent drop in one county.

Many judges, HZA noted, are already more likely to remove children than DCFS, whether because they have a zero-tolerance approach to drug use or because some are notoriously hostile to placing children with the relatives of parents who have lost custody. Some judges have also seemingly made decisions about children’s placement based less on facts than on their apparent annoyance with DCFS’s handling of a case.

“From workers’ perspectives, these types of decisions send a message to DCFS never to challenge the court’s judgment, and they serve their intended purpose of intimidating DCFS staff into doing what they perceive the court to prefer instead of making decisions rooted in their training,” HZA found. “More importantly, it unnecessarily punishes the families who are caught at the crossroads between the two entities.”

One result is substantial inconsistency around the state when it comes to DCFS practices. As one area director told HZA, “I might approve a removal for one county and then ten minutes later not approve a removal for the same [conditions] in another county because of who the judge is.”
This problem has been exacerbated by the influx of new and inexperienced DCFS hires in the agency — itself the result of a longstanding staffing shortage that, in recent years, has amounted to what DCFS Director Martin called an “exodus of experienced field staff.” Today, the average DCFS caseworker has only about one-and-a-half years of experience, compared with nearly three in 2013. Rookie caseworkers are less likely to feel confident in defending their decisions against a skeptical or hostile judge, and are more likely to begin to shape their recommendations based on what their local courts prefer. (The very fact of the newness of DCFS’s staff might be another driver of the foster care increase, as HZA noted that just six months after the biggest drop in caseworker experience, the foster care population began to rise.)

Another factor in the increase is the responsiveness of DCFS caseworkers, and the agency as a whole, to the prospect of bad press. A majority of the staff HZA spoke to said that their decisions were effected by negative media coverage, with one supervisor reporting that she tells her staff, “Never leave a kid in a vulnerable situation if you’re afraid you’d be on the front page.” Mapping the number of children brought into care over the last two years against high-profile child abuse scandals or deaths — from the “rehoming” of two adopted children by Rep. Justin Harris to the hot car death of the son of Judge Wade Naramore — there’s a clear spike associated with each news event, whether or not it involved DCFS.

DCFS hastened to point out their concerns about the report after it was released to the Times. In addition to questioning HZA’s methodology, Martin’s letter also argued that the report failed to look closely enough at the reasons fewer children are leaving foster care and at the impact of DCFS staffing cuts from 2011-2013. But in fact the report does address that. It notes that even the decrease in the number of children exiting foster care — whether to be reunited with their families or to be adopted — is likely tied to the increased number of children coming in, as caseworkers struggling to meet the needs of an additional 1,100 children statewide don’t have time to resolve their other cases. The growing foster care population, one caseworker told HZA, causes “‘the entire system to slow down.’”

DHS spokesperson Amy Webb said in May the agency began holding “foster care war room meetings” with both DCFS staff and national experts. Those sessions, Webb said, approached the problem of increased foster care population from the “macro level” of “what barriers exist to keeping more kids safely in their homes,” and will lead to the DCFS report that Webb anticipates will be ready for release soon. 

In its recommendations, HZA called for better communication between both DCFS’s different units and also between the agency and the court system. In other states, the report noted, child protective services agencies have worked with the Court Improvement Program to come to a better consensus on how to address substance abuse in child welfare cases. It also called for lessening restrictions on placing children with relatives.

But some of its strongest words were for the investigators that it says are taking an overly punitive approach to their cases: “The existence of ‘probable cause’ means that the agency can remove a child, not that it should remove the child. In other words, investigators must exhibit the courage to exercise restraint. After all, removals are justified only when reasonable efforts to prevent the removal have been made.”

Benjamin Hardy contributed to this report.