There was an interesting op-ed in US News this week, in which two researchers reported on their study showing that increasing annual spending on anti-poverty initiatives by 13% person in all 50 states and D.C. “might be associated with approximately 181,000 fewer children reported for maltreatment, 28,500 fewer victims, 4,100 fewer children entering foster care and 130 fewer children dying – every year.”
I found the study interesting not because it’s new information — there are numerous studies showing that maltreatment reporting and poverty go hand-in-hand — but rather because of a principle I learned freshman year in college: correlation is not causation.
The study’s authors — Puls, Hall, and others — compared state spending on poverty (e.g., TANF, Medicaid, food stamps, etc) with state maltreatment rates (reports, substantiations, foster care entries, and abusive deaths). Where they may go wrong, however, is in assuming that there is a relationship between actual maltreatment and the number of reports, substantiations, and entries into foster care. Our own research here in Georgia suggests there’s little relationship between actual maltreatment and foster care rates. Rather, review of the federal child welfare data over the past 20 years more likely suggests a correlation between the foster care population and media attention on child abuse deaths and serious injuries.
From 2000-2005, for example, following a series of widely-publicized child deaths, the foster care population in Georgia swelled to almost 15,000. Under Governor Perdue (who sent a strong message that poverty is not maltreatment and not grounds for removing a child from his home), the numbers fell to half that. We’ve had another flow in more recent years (up to 15,000 again in early 2018) and ebb (down to 10,000 in 2021). One conclusion is that the number of children who come into foster care is mostly dependent on the State agency’s policies, practices, and attitudes toward removing children from their homes.1
Substantiations, likewise, can be dependent not upon actual maltreatment but rather upon policies and practices. When Georgia instituted a child abuse registry (CAR) in 2016, DFCS substantiations of neglect or abuse dropped dramatically, from around 23% down to 15%. The reason? For the first time in recent years, a case manager’s determination that there was “likely” neglect or abuse had far-reaching impact, placing a parent on a registry in a way that could impact his or her ability to do certain kinds of work. Case managers didn’t want to do further damage to parents who were already struggling. Once the legislature abolished the CAR in 2020, substantiation rates returned to their pre-CAR levels.
The study also doesn’t take into account the recent voices who have raised concerns that the child welfare system is really a “family regulatory system” that unfairly targets the poor and minorities. Under their contention, the best way to reduce maltreatment reports and entries into foster care might be to ban mandated reporting and require that parents in CPS cases be informed of their rights.
These diverging points of view bring into focus an important, if often overlooked, set of facts. First, we don’t know the actual rate of child maltreatment in the US. Are our systems capturing it all? Are we “over-policing” families? Second, the system is a reactive one, addressing reports of child abuse rather than actual child abuse — and after the fact, at that. We have a long way to go to create a system that proactively helps families safely care for their children.
h/t Andy Barclay, Christopher Church.