a critical look at the child welfare system

SCREENING OF REPORTS

During the mid-1990s, a major obstacle preventing a direct count of the number of children reported and substantiated for maltreatment was the wide variation among data collection procedures among the states. While all states were asked by the National Committee to Prevent Child Abuse to provide the number of children reported for maltreatment for its 1994 Annual Fifty State Survey, 12 states among those responding were unable to do so.

Thirty six states counted only reports of child abuse and neglect that were actually investigated, while eight other states counted all calls to their hotlines, including calls for help, for services, and informational requests.

States historically have varied greatly in the number of calls they exclude from the investigation process, with 17 states screening out, on average, 47 percent of the calls they received. There is also a tremendous variation in the percentage of reports screened out, with some states screening out as little as 15 percent of calls, to some others screening out as many as 80 percent of them during the mid-1990s.[1]

Forty-eight States reported that about 1,675,000 investigations of alleged abuse or neglect were conducted in 1995, a figure representing roughly one half of the total number of reports at that time. Of those investigations, only about 36 percent resulted in a disposition of either substantiated or indicated child maltreatment of any variety.[2]

“Screen-out rates for Child Protective Service reports nationwide vary greatly and range from 2.8 percent to 75.6 percent of reports received for the 45 states reporting figures to the federal government for 2009. The average screen-out rate for the 45 states combined was 38.1 percent. It is difficult to make comparisons between states due to variations in practice and child maltreatment definitions,” a Wisconsin memorandum of more recent vintage explains.[3]

As of 2010, 45 states reported a total of 2,607,798 abuse or neglect referrals, of which 1,581,882 were passed on for investigations, while 1,025,916 – over one million – were screened out, according to the Children’s Bureau. The national average was 39.3 percent of referrals screened out without being passed on for follow-up.

There were some incredible variations between the states. Forty-five States reported counts of screened-in and screened-out referrals, providing for a nationwide average of 60.7 percent of calls being screened in, with acceptance rates ranging from 25.2 to 98.7 percent. On the other side of the coin, states screened out as few as 1.3 percent, to a maximum of 74.8 percent of their hotline calls in 2010.

Alabama accepted 98.7 percent of its calls, while screening out only 1.3 percent. In contrast, South Dakota screened out fully 74.8 percent of its hotline calls, accepting passing on 25.2 percent ot them for follow-up.

But the acceptance rates at the hotlines only tell a part of the story, while substantiation rates tell another. The Children’s Bureau notes that between 50 states, the District of Columbia, and Puerto Rico, only one-fifth of the reports accepted for investigation were found to involve actual victims of abuse or neglect.

The nationwide substantiation rate – even after screening out the excessive number of inappropriate reports – was only 19.5 percent. As the Children’s Bureau explains it, of the reports screened in and passed along for investigation “four-fifths of the children were found to be nonvictims of maltreatment.”[4]

National screening rate child abuse 2010

2010 CPS case dispositions after investigation.
Source: Children’s Bureau, Child Maltreatment 2010.

A CLOSER LOOK AT THE TRENDS

During the first quarter of Fiscal Year 2012, the Virginia Beach Department of Social Services screened out 57 percent of its reports – a figure that mirrors the regional rate, according to a study issued in July 2012.

During that period, the Virginia Beach Department responded to 366 calls, which works out to an average of 124 calls per month, or 6 per business day. On a per worker basis, this equals approximately 8 new referrals per month, a figure that is “well below the Child Welfare League recommended standard of 12-14 new referrals per month, per worker,” the study explains.[5]

Although lobbyists seeking to promote increased spending on child welfare programs continue to use the raw number of calls, including the duplicated counts, to emphasize the need for increased funding, the reality is that states have historically screened out a significant number of the calls made to their hotlines.[6]

NEW YORK STATE CENTRAL REGISTER: A MODEL FOR THE NATION

“Created by law in 1973, New York’s State Central Register was one of the first of its kind in the country, and the law would serve as a model for other states. Beginning as a makeshift operation with five workers, the hot line is now a far-reaching system that employs 208 people,” explained Albany Times Union reporter Darryl Campagna.

By 1994, the New York State hot line received 486,000 calls. Of these, an estimated 100,000 were follow-up calls to ongoing cases, and 85,000 of these calls calls were determined to be pranks that had to be treated as valid until proven otherwise. Of these calls, only about 26 percent of them were actually being passed on for investigation.[7]

As the number of calls increased, so did the difficulties in keeping up with the ever-growing demand. And the problem was not isolated to New York.

As of 1997, of the 355,579 calls made to the Illinois Department of Children and Family Services hotlines, only 19.2 percent were actually passed on for investigation. According to the Department’s own data, over a ten year period the percentage of reports accepted for follow-up had steadily declined, even as the sheer number of reports increased.[8]

Illinois DCFS reports graph

By 2012, Illinois’ DCFS reported facing “challenges with long wait times for calls coming into its hotline,” looking at Indiana’s new centralized Hotline system as a possible solution to the problem.[9]

In the state of Georgia, during calendar year 1993 21,182 of the 73,701 calls made to the hotlines – nearly 29 percent – were screened out. Of the remaining calls actually passed on for investigation, 33 percent were determined to be unfounded, while another 33 percent were closed as “unconfirmed.”

Approximately 23 percent of Georgia’s reports were determined to be founded, however among those findings, 46 percent were for one or another variety of neglect or other maltreatment, while 6 percent were for “emotional abuse.”[10]

Georgia child abuse dispositions

Just as there are many variations among the states, there are wide variations
in how reports are handled from one county to another. This has always been the case.

During the mid-1990s, 32 percent of all reports in California were screened out statewide. But the percentage of reports screened out varied substantially from one county to another. In 1994, for example, Los Angeles County screened out about 19 percent of the reports its hotline received, compared to 55 percent in Contra Costa County. According to the California Office of the Legislative Analyst: “Some of the variation among counties may be due to a lack of specificity in the state guideline, thereby allowing counties to adopt different screening policies.”[11]

California child abuse screening statistics

“In Wisconsin, the range for screened-out CPS Reports for 2009 was 7 percent to 91 percent. The statewide average was 55 percent,” the state’s Division of Safety and Permanence explains.

Reviewers sought to identify the source of the wide county-to-county variations, and conducted a review that examined 476 reports across 31 counties.

The reviewers for the most part agreed with the screening assessments, however they thought that a greater percentage of reports met the definition of “possible impending danger threats”
and should have been screened in for follow-up. Based on the reviewers more liberal interpretation of what may consitute such a threat, they concluded that 76 percent of the reports examined did not contain an allegation of impending danger to a child.[12]

Wisconsin child abuse reports graph


Source: Wisconsin Division of Safety and Permanence
Child Protective Services Access Screening Review Summary
April 2011

The variations between counties are somewhat smaller in West Virginia. In 2009, the state’s hotline took in 35,468 calls, of which 23,515 were accepted for further investigation making
for a statewide acceptance rate of 66.30 percent.

With its acceptance rate of 87.80 percent, Tyler County had the highest rate of calls passed along for follow-up, with Gilmer County coming in as a close second with 87.14 percent. At the lower end of the spectrum, Barbour County accepted 40 percent of the calls that it recieved, with Preston being the runner-up, having accepted 44.07 percent of its calls for follow-up.

While the county-to-county variations in West Virginia are not as extreme as they are in some other states, the counties at the lower end of the spectrum nevertheless took in less than half of of the calls on a percentage basis than did their counterparts at the higher end of the spectrum.[13]

Florida’s 2010 Home Visiting Needs Assessment provides some potentially valuable insights into the “broad variations” in reporting between counties and local communities. The Assessment explains:

The reporting of child abuse/neglect typically occurs as tips from neighbors, teachers, service providers, or police officers transmitted via the Florida Abuse Hotline. Cultural norms as to what actions and behaviors, or lack thereof, constitute “abuse or neglect” influence which tips are formally reported. Additionally, cultural norms dictate whether incidents are handled internally, rather than reported to the hotline. Thus, reporting can vary considerably from community to community throughout the state, making one region appear to have a higher volume and rate of cases and another region a lower volume and rate of cases.

In many states, the percentage of reports screened out has apparently increased, even as the raw number of reports increased. The state of Massachusetts, for example, screened out about one third of the calls made to its hotlines in 1991. As of 2010, the percentage of calls screened out had increased to 44.7 percent.[14]

Similarly, the state of Washington screened out approximately 30 percent of the hotline calls that it received, according to a 1996 performance audit. By 2010, Washington state was screening out 54.2 percent of its referrals.[15]

CASTING WIDER NETS

“In 2010, Minnesota’s child protection agencies screened out 68 percent of recorded referrals statewide. A decade earlier, in 2000, the state’s child protection agencies reportedly screened out 38 percent of child maltreatment referrals. The apparent increase in the percentage of child maltreatment referrals that agencies were not investigating or assessing has concerned some people,” explains Minnesota’s Office of the Legislative Auditor.

Concerned by this discrepancy, the auditors conducted a number of interviews, arriving at the conclusion that: “Increases in the statewide screen-out rate between 2000 and 2010 may reflect child protection agency data recording practices rather than changes in agencies screening decisions.”

As for the wide variations between counties, the audit notes that “one screener said his agency tends to err on the side of caution and screen in referrals other agencies would screen out. In contrast, staff at another agency stated that they do not accept most of the referrals they receive. They said, technically, according to statutes they could screen in practically every call. However, they draw the line at serious maltreatment and focus on those allegations. Staff from a third agency told us the agency prefers to use a non-child-protection response unless children are in imminent danger.”

The audit explains also that “a major contributor to agency variation in screening is vague statutes defining abuse and neglect.” Keeping statutes vague “may reflect a policy choice to preserve the ability of child protection agencies to practice in ways that accommodate the variation among communities they serve,” the report explains.

Significantly, the audit notes that the 2011 Minnesota Legislature enacted the “Vulnerable Children and Adults Act,” which specifies that DHS will distribute state and federal funds based on a formula that considers the number of children who were the subject of screened-in maltreatment referrals. “Thus, agencies that cast wider nets for screening in child protection referrals could receive more funding, even if their caseloads reflect less-serious child protection cases than agencies that more selectively screen their referrals.”[16]

THE RANDOM ELEMENT

By the mid-1990s, the deluge of reports had introduced a random element into reporting – the patience of the individual reporter. In California, for example, the Orange County Grand Jury reported that one out of four callers to the County Child Abuse Registry never got through. 11,591 of the 46,313 callers to the County’s Child Abuse Registry during an 11-month period in 1993 hung up the phone after lengthy periods of time spent on hold or after getting constant busy signals. One grand juror noted that it was impossible to know what prompted the calls that went unanswered, but that the threat of missing a true emergency call was genuine.[17]

The flood of inappropriate reports into the New York state registry was having a similar impact, with 10 percent of its calls lost to hang-ups. Says Lori Beer, adult services coordinator with the Saratoga Center for the Family, a private, non-profit organization that provides services for abused or neglected children and their families: “I make my hot line calls at 11 o’clock at night. I’ll only wait 20 minutes.”

According to agency director Kristine Hoagland, her staff sometimes has to wait an hour to get through to the hot line.

“We don’t like to say calls are ‘lost,'” said Terrance McGrath, a spokesman
for state Social Services. “We do know that about 10 percent of the calls hang
up before they get through, over the course of a year or a day. We assume that
most of them will call back later.”[18]

As of 1997, callers in Prince William County, Virginia, were waiting 30 minutes to get through to an operator.[19]

The problem of lost calls continues to this day, and efforts to keep up with the volume of calls often serves to frustrate hotline staffers, as the Virginia Beach study referenced above explains:

Intake is currently staffed by four social workers. This is a recent change in response to complaints of “dropped” calls and long waiting periods for callers. Staff expressed concerns that no coverage schedule for Intake means staff cannot take lunch or breaks unless they can secure their own coverage. A problem with the current system is no way of tracking how many calls are waiting to be answered. Staff shared the DSS phone system is the same as the police use yet the feature of a screen showing calls waiting are not available to DSS.[20]

The New Jersey experience is hardly any different. As of June 2011, the hotline answered 16,325 calls, of which 716 were “abandoned calls.” Among these calls, 5,592 calls were screened out, while 1,232 resulted in referrals for follow-up.[21]

“A majority of callers to Illinois’ child abuse hotline — a front line in protecting battered and neglected children — don’t initially get through to someone who could dispatch an investigator,” the Chicago Tribune reports. The Department of Children and Family Services took messages for the majority of its more than 236,000 calls logged over an 11-month period ending May 31, 2012.

“The percentage of callers who reach a specialist on the crucial first attempt has plummeted over the last 11 years. It’s now less than 40 percent, compared with nearly 70 percent in 2001,” the newspaper found. According to hotline workers and mandated reporters, including the police, teachers, and doctors, it may take several hours during peak periods to get a response.

A DCFS spokesman acknowledged the problem, saying that the alternative would be that of more abandoned calls. Less than 3 percent of the hotline calls over the course of the 11-month period being reviewed resulted in a hang-up. The newspaper reported that this reflected “the best rate chronicled in more than a decade.”

As for the national rate of calls lost to hang-ups, the current average is about 5 percent.[22]

THE NEW AND IMPROVED NATIONAL MODEL

Seeking “to bring consistency to the way abuse and neglect calls were managed across the state,” Indiana established a centralized Child Abuse and Neglect Hotline that was phased in over the course of some time during 2010, a process that was completed on August 30, 2010. Clearly, something needed to be done, as the Indiana Department of Child Services explains:

Prior to Indiana launching the hotline, reports of child abuse and neglect were directed to more than 300 different telephone numbers across 92 counties, with some calls rolling over to law enforcement offices, or being answered inadvertently by overnight cleaning staff. Call record keeping also differed from county to county, making it difficult to determine how many calls in total did not meet the legal definition of abuse or neglect.[23]

This was no ordinary centralized Hotline, as Indiana’s web site explains: “The Hotline is staffed with 62 specially-trained Family Case Managers, know as Intake Specialists, who are professionally trained to take reports of abuse and neglect. These Intake Specialists gather information from callers, determine whether the information provided meets statutory criteria for DCS to conduct an assessment, and if appropriate, route reports directly to DCS local offices for response and assessment.”

The Hotline “streamlines the Agency’s approach to taking reports, improves the Intake Specialists’ ability to gather information from callers, and expedites the process of preparing comprehensive reports and disseminating those reports to local offices for assessment,” the site explains. It also “allows Family Case Managers in the local offices to spend more time partnering with children and families because they are no longer responsible for handling intake functions. It also allows DCS Intake Specialists to ask more probing questions to obtain comprehensive information about factors that may impact worker safety.”

It features a “computerized call system” that is said to provide DCS with quicker data entry, as well as the ability to track the number of calls received and the timeliness and quality of responses to callers. This “state of the art system” also allows staff the opportunity to listen to individual calls, such that if specific concerns regarding a call are raised, staff can “utilize these features to review the call and address any concerns.”

Clearly, this is no ordinary hotline. This is the Cadillac of hotlines; so much so that several states were looking at it as a model on which to improve the operations of their own Hotlines.

Hotlines are a “best-practice trend spreading across the country,” Indiana explains, and Arizona, Florida, Illinois, Missouri, Ohio, Texas and West Virginia “are just a sampling of states that have adopted centralized hotlines. Recently Oklahoma and Louisiana also launched their own centralized call systems.” Representatives from Arizona, Illinois, West Virginia, Iowa and Michigan have explored Indiana’s new Hotline “as an example of how to implement or improve their own operations in addressing reports of child abuse or neglect,” the Department explains.

“If you find yourself on hold please be patient and do not hang up. Your call will be answered by the next available Intake Specialist,” the Hotline’s web site instructs would-be reporters. “If you are calling regarding an emergency situation and/or if you believe the victim is in imminent danger please dial 911 immediately.”

What happens if someone actually has a genuine emergency to report, and calls 911 or the police department directly? Law enforcement agents are instructed to “call the Child Abuse and Neglect Hotline especially in cases of emergency,” the Hotline’s explains. “To ensure law enforcement officials have the highest priority, they have an access code which routes their calls to the front of the call sequence. On average, law enforcement officials utilizing the access code are connected to the hotline within 46 seconds.”[24]

THE NEW MODEL’s CRITICS

Even as Indiana officials continued to tout the benefits of their “model” program, criticisms continued to mount, with local officials familiar with the unit painting a very different picture, describing it as a “constant problem,” “very frustrating,” and “inefficient.”

“They do not seem to understand the issues that are actually going on in the field,” a Warrick County sheriff’s detective says. A Knox City police detective was more direct: “Children are not getting the help they need.”

Those comments are contained in responses to an informal survey conducted by Sen. Brent Steele (R-Bedford) which he shared exclusively with The Indianapolis Star

“Critics have complained that the system wrests control from local officials with local knowledge and that too many cases are now deemed unworthy of investigation,” journalist Alex Campbell explains. Others who were surveyed echoed these complaints while raising many others, including “critical delays in response, a lack of follow-up, miscommunication, incompetence and, in some cases, an unwillingness to acknowledge and address problems.”[25]

Indiana Governor Mitch Daniels chimed in, saying that his administration would continue to review the Department of Child Services’ abuse Hotline, adding that: “You can always get better.”[26]

On September 5, 2012, the Hotline came under renewed criticism; this time from social workers, members of the public, as well as some current and former hotline workers during the second round of legislative hearings concerning the Hotline, held at the Indiana Statehouse.

“Some described a hostile work environment. Others decried lack of follow-up and bad advice that posed a danger to vulnerable children,” reports The Indiana Star.

The first to testify during the three-hour session was an ex-hotline supervisor. Amber Turientine left the agency in October 2011 due to what she described as a work environment in which supervisors were “encouraged to bully and target” their subordinates. She also submitted written testimony from four former hotline workers, as well as three current ones, all of whom raised similar issues. All but one former worker submitted their testimony anonymously.

“I am very afraid that a child fatality will be directly the result of failure of this hotline,” one current employee wrote. “Things are that serious.”

Changes stemming from a decision to allow local law enforcement authorities direct access to local DCS caseworkers when reporting suspected abuse were welcomed by Bedford City Police Detective Robert Herr, who testified in favor of the recent change in policy.[27]

Further investigations into the operations of the Hotline are planned to continue.

Why do states screen out so many of these reports? Is there a specific criteria used in determining which calls to accept and pass along for investigation?