The long delayed release of an evaluation of Illinois’ differential response program casts
new doubts on whether one of the country’s most popular child welfare reforms is safe for children and a smart way to spend limited resources dedicated to families on the fringe.
According to the report, children whose parents had benefitted from twice as much social work time, $400 stipends and a philosophy that stresses family strengths were more likely to be reported for child maltreatment and become victims of substantiated abuse or neglect.
This controverts 20 years of evaluations and assurances that children involved with differential response-type (DR) programs across the country were as safe or safer than children who received traditional child protective services.
“I was surprised,” said Tamara Fuller, the lead author of the Illinois evaluation. “We were all surprised. This is the first evaluation to see safety go in the opposite direction.”
Others in the research community had forewarned of the potential shortcomings of DR. While Illinois discontinued its differential response program two years ago because of safety concerns and budget strain, dozens of other states are still pushing ahead.
What is DR?
Over the course of the past year I have read scores of evaluations on differential response, position papers, presentations to public officials and news stories. I have also conducted scores more in-depth interviews with experts ranging from the front-line workers who investigate child abuse to child welfare researchers and elected officials.
What emerges is a picture of a reform movement grounded in the idea that traditional investigations of child abuse are too “police like,” which compromises social workers’ capacity to engage families and prevent children from entering foster care.
To fix this, the designers of DR argued that the child protective system should offer a “differential response” for families that show fewer risk factors when a call of child maltreatment comes into the hotline.
In Illinois, and in other public child welfare administrations, calls that meet the legal threshold to warrant an investigation are broken into two tracks. Those cases deemed less risky fall into the “DR track.” In these cases, workers are instructed to forgo fact-finding investigations for softer, “strengths-based” and “family-centered” assessments. Parents are then offered voluntary services. This is unlike traditional investigations, where families can be compelled to follow a service plan or risk losing their children.
A report published by Casey Family Programs, a charitable foundation that has poured money into DR expansion from California to Illinois, said the following of the practice’s superiority over traditional, forensic investigations. “Due to the non-adversarial approach of DR, its focus on identifying root causes behind parenting difficulties, and provision of family related services, DR has demonstrated improvements in family engagement, child and family outcomes, and some cost savings over time,” the report read.
The perceived benefits of this “non-adversarial approach” are echoed in analyses of social worker surveys found in many of the major evaluations of DR to date. One of the most widely cited of these evaluations is the final report on Minnesota’s DR pilot published in 2004. In it, the Institute for Applied Research, a Missouri-based research firm that has conducted the lion’s share of major evaluations on DR, concludes that:
“Generally, according to workers, AR [another name for DR] builds more positive relationships between families and themselves as well as changes [the] adversarial view of child protection system.
Workers tended to like the fact that AR is not as ‘punitive,’ ‘intrusive,’ ‘threatening,’ ‘labeling,’ ‘blaming,’ ‘shaming or antagonistic’ as the traditional approach, which focuses on investigation, documenting incidents and making determinations.”
Propelled by popularity within the child welfare ranks, glowing evaluations and continuous promotion by some of the most powerful players in the field, DR took off. Since 1993, when differential response was first launched in Missouri and Florida, it has spread to as many as 30 states, sometimes in clusters of counties and, in others, state-wide.
But soaring regard for DR has been tempered of late with criticism from some in the research community.
The September Issue
In September of 2013, Ron Hughes and Judith Rycus of the Ohio-based North American Resource Center for Child Welfare (NARCCW) launched the first major broadside on the evidence base that has been used to speed the adoption of differential response in jurisdictions across the country.
In a special issue dedicated entirely to DR that ran in an academic journal called Research on Social Work Practice, Hughes, Rycus and colleagues claimed that the evaluations consistently cited to sate nerves over child safety and tout DR’s overall effectiveness were at best inconclusive and at worst marketing tools used by “knowledge monopolies and research cartels” bent on proliferating the growth of an unproven practice.
They took aim mostly at Tony Loman and Gary Seigel, the principals behind the Institute for Applied Research, and authors of seminal evaluations of differential response in Missouri, Minnesota, Nevada and Ohio.
The rest of the September issue was filled with responses to Hughes and Rycus’ paper. One peer hailed Hughes and Rycus for writing what “may be the most important article in the child welfare arena in the past 15 years.” A trio of researchers from the National Council on Crime and Delinquency’s Children’s Research Center wrote:
“This is an important, compelling critique of an initiative that has enjoyed near universal support and has been adopted by child welfare agencies nationwide. Points made throughout their review raise serious concerns about nearly every aspect of the DR movement, ranging from conceptual framework to the evaluations that support its effectiveness.”
Loman and Siegel fired back a testy response: “Each of the points made by Hughes et al., we contend, misrepresents our work; together they seem designed to impugn our integrity.”
In an interview conducted shortly after the September issue was published, Tony Loman minimized Hughes’ and Rycus’ assertion that his firm was promoting DR in any way.
“Marketing is a funny term anyway to use with this,” Loman said. “Because it implies that there is some kind of financial gain. And there is not. If you want to make money, you don’t do evaluations of programs like this.”
The Cost of Re-Directing Limited Funds
Beyond disassembling Loman and Siegal’s methodology and interpretation of data, Hughes and Rycus further argued that DR requires a siphoning of limited resources from higher to lower risk cases.
“You only have so much money, so it has got to come from somewhere,” Hughes said in an interview. “Right now, most goes to high risk cases; and they [the proponents of DR] are saying, ‘Take that money and put it into low-risk cases.’ To do that is harmful.’”
While the exact proportions of funding redirected from higher risk cases to lower risk cases on the DR track are hard to fully discern, reports from child welfare monitors in Connecticut and Los Angeles County indicate stress to traditional child protective services when DR is implemented, and potential danger to children.
Connecticut, under a consent degree stemming from a 1989 class action lawsuit lodged against the Department of Children and Family Services, has a court monitor who issues quarterly reports on the child welfare system.
In October of 2013, Court Monitor Raymond Mancuso issued a report stating that:
“Front line staffing levels are inadequate given the complexity of cases that now make up the pool of investigation and ongoing service cases that social workers have on their caseloads since the implementation of the Differential Response System (DRS). DRS results in the diversion of low-risk cases from workers’ caseloads, leaving staff with caseloads made up of only complex cases.”
Three months later, in the following report, Mancuso wrote that the “the situation has worsened.”
In February of 2013, The Los Angles Times published a confidential report that had been written by County Counsel’s Children’s Special Investigative Unit in 2012.
The report detailed the systemic failings that contributed to 14 deaths and one “critical incident” wherein an eight-week-old boy was thrown against a wall and sustained near fatal injuries. “Front-end” services, including under-informed investigations and an over-reliance on L.A.’s differential response experiment called Point of Engagement, contributed to the majority of the deaths, according to the report.
“In recent years, the focus on utilizing voluntary services and safety plans as a means of keeping children at home has clearly conveyed the message that DCFS wants to ‘reduce reliance on out-of-home care,’” the report’s author Amy Shek Naamani wrote. “Like the classic ‘game of telephone’, overtime, the message ‘morphed’ and was understood by the workers and managers as simply ‘do not remove/keep the numbers down.’”
The agency’s primary goals, including shorter paths to permanence and improved child safety, were lost in the shuffle, Naamani wrote. “Individual offices and leadership celebrated as their number of detentions decreased and individual social workers were praised for low detention numbers; all while more children were dying while left in their parent(s) care.”
In Illinois, the Department of Children and Family Services made explicit this tension over limited resources. On Dec. 18, 2013, then-acting director of DCFS, Denise Gonzalez, sent a letter to State Sens. Mattie Hunter and Julie Morrison explaining the rationale behind eliminating the differential response program in 2012. Hunter had introduced the legislation creating DR back in 2009, and Morrison serves on a Senate subcommittee focused on DCFS.
“The program was cut to allow DCFS to preserve funding for programs that were more critical to child safety,” the letter reads. DCFS analysis showed that “children whose families were diverted to DR were no safer than families that received safety services through traditional child abuse investigation assessments and procedures.”
Further the letter contends: “DR required transferring staff out of frontline child protection positions and into the DR pilot, creating vacancies in investigations that DCFS was unable to fill, and thus driving up caseloads for investigative staff that contributed to high caseloads that put children at risk.”
Gonzalez went on to reference the DR evaluation conducted by Fuller and her team that had, by this time, been completed but not yet published.
“Subsequent analysis by the University of Illinois in a report that will soon be released found that over the 18-month period following the closure of a family’s initial report to the department, families in the DR pilot program were more likely to experience a reported recurrence of maltreatment,” Gonzalez wrote.
The Illinois Evaluation
Fuller, the director of the University of Illinois at Urbana-Champaign’s Center on Child Welfare Research, had been commissioned in December of 2010 to conduct the Illinois DR evaluation as part of a massive federal research grant promising a three-site evaluation of differential response. Of the three state reports submitted, Illinois’ was the first.
While it is dated October 2013, it was not released until January 2014. Fuller and DCFS say that the release was delayed by the high turnover of child welfare directors in Illinois during the past year.
The study zeroed in on 7,584 families who were eligible for DR services from November 2010 through May 2012. Forty-one percent of (3,101) families were randomly assigned to receive differential response services, while the others received a traditional investigation.
Families in the DR track were assigned two social workers, one from DCFS and the other from one of 14 private agencies contracting with the department. Children not on the DR track only had one county-employed social worker assigned to their case.
Caseloads, a critical component in the quality of services given by social workers, were lower for social workers in the DR track.
This was so pronounced that Fuller and her colleagues made note of it in a 2012 report following a site visit the year before. “The lower caseloads and the lower severity of the types of cases being handled by the DCFS DR specialists were highly visible to the DCFS investigators working the traditional track, which often led to resentment,” they wrote.
The disparities did not end there. Families on the DR track had more frequent visits from the private social workers acting on behalf of DCFS and consistently received more services aimed at alleviating the socioeconomic conditions that are thought to contribute to heightened risk of maltreatment. This included cash stipends to pay for car repairs, food and clothing, utilities, furniture, appliances and home repairs.
On most measures it appears that families assigned to the DR track were given more attention, and more resources to mitigate perceived safety risks. Of the 25 percent of DR track families that filled out a survey on their experience with DR, versus the 20 percent who responded to surveys on the traditional track, differential response was the clearly more popular.
“Families like it better,” Fuller said. “We have definitively answered that. In that sense it works fine. If you look at safety you may have a different answer.”
While Fuller and other researchers I interviewed were quick to point out the large size of the sample and how that can make small differences seem more significant, the Illinois evaluation showed that children in the DR track were less safe than their peers in the control group.
Eighteen months after a DR case was closed, 18.8 percent of children received a new report of child maltreatment, compared to 14.7 percent in the traditional track. After the same amount of time, 6.1 percent of children in the DR track were substantiated as victims of abuse or neglect, as compared to 4.7 percent in the traditional track.
And while removals of children from the DR track 18 months after their cases had been closed was only 0.2 percent higher than those who received a traditional investigation, it took, on average, 49 days longer to make the decision to remove.
Of the 3,101 families who ended up on the DR track, only 1,389, or 44.8 percent, would complete the services.
In Illinois, twice as much social work, car payments and a sincere effort to change the face of DCFS did not significantly induce engagement. Instead, less than half of the families made it through the program and the children in the DR track were less safe. All while impeding the traditional system’s ability to respond to higher risk cases.
When asked if Fuller still believes in DR, her answer was fast and earnest.
“Absolutely,” she said. “It’s a promising practice. If you look at totality of the research as a whole; if you look at every evaluation I still think it is a promising practice.”
Daniel Heimpel is the founder of Fostering Media Connections and the publisher of The Chronicle of Social Change.