Many family lawyers wonder about how Child Protective Services agencies make decisions about which calls to investigate, which children to detain, and when it is safe to return a child to the home of a parent. But it is likely that none of us imagined that computers, not social workers, are making most of the decisions.
Today’s Los Angeles Times reports that a computer program is responsible for the vast majority of child detention decisions made by Los Angeles County’s Department of Children and Family Services, which must screen and cope with 3,000 abuse calls each week. The Times notes that social workers are often inexperienced, undertrained and under paid (earning starting salaries of $35,200 per year), and that they leave the department at the rate of 7% each year. Two-thirds have worked for the Department for less than five years. In the absence of staff experience and expertise, Los Angeles County relies heavily on a computer program that employs a program known as Structured Decision Making (SDM). The program was developed by the he National Council on Crime and Delinquency (NCCD) and the Children’s Research Center (CRC).
One factor is the number of prior reports about the same family — although unsubstantiated allegations result in fewer points than do substantiated abuse reports. Developers hoped that the criteria used by the software would prove objective, and that the program would reduce the disproportionate number of black children detained — four times more black children than white children enter the dependency system in Los Angeles County — but that has not proven to be the case.
Moreover, social workers providing reunification services use their knowledge of the software to assist parents in developing a profile that will result in the software recommending return of their children to parental care.
Take a look at the article, and post your comments please.
Wow, what did y’all do to get 7% turnover at children and family services? We’re at 20%+ a year here and have been forever….
On a more serious note, this is one of the many damned if you do, damned if you don’t conundrums of CPS work – when it is humans making decisions based on idiosyncratic factors and “feelings” about how a family is doing it is called unscientific and arbitrary, when a statistical approach is used to risk assessment agency staff are accused of being mechanistic and lacking a “human element.” Really though, it is still people selecting how they categorize information to enter in to the statistical model (i.e. when there are ratings of “social support” the worker’s opinion on Aunt May or Uncle Ben may make a major difference on the “none-to-substantial” scale), and really, how many caseworkers know that there is a X% likelihood of future harm based on the many issues that have been studied for such models (interplay of age and number of children, age and number of parents, significant others, etc., etc.).
Aaron Robb, M.Ed., NCC, LPC-S
Forensic Counseling Services
Keller, TX
This is probably the way of the future, but as Mr. Robb says, we want objectivity but we also want a human factor. It’s one of those difficult situations where the computer can be a tool but a human should make the final decision.