While a higher level of correlation existed between poverty and maltreatment in older children, disability benefits were still “overwhelmingly” the strongest correlation for all age groups.

From How Can We Identify Kids At Risk? by Naomi Schaefer Riley.  The subheading is Drug abuse and mental illness in the household are better predictors of child abuse than poverty.  

Interesting research.

How do we determine which kids are most at risk of severe abuse or neglect, and what can we do to protect them? Answering these questions should be the primary concern of child protective services. In recent years, more advocates have become interested—reasonably enough—in the question of how to stop abuse and neglect before they happen. In academia and government, it’s about getting “upstream” of the problem.

Several years ago, in Texas, researchers started identifying communities by zip code with high rates of child maltreatment. They included in this measure kids exposed to maltreatment—often siblings who had been in families with substantiated reports of abuse or neglect against another child. Then the researchers tried to figure out the common characteristics of the communities.

This strategy differs from simply assessing the risk factors of a particular family. It would be complicated, as well as controversial, for government to single out a particular family (even if it did so only to offer help) just because it has characteristics that correlate with child maltreatment (such as a former felon living in the home). By focusing on whole communities at risk and then offering various services, the researchers hoped to avoid this difficulty.

Some factors correlated with high-risk communities won’t be surprising: a high percentage of residents with less than a high school diploma and high ratios of hospital-based deliveries to teen mothers or infant emergency-room visits. But the factor with the strongest correlation by far is the percentage of adults aged 35 to 64 receiving Social Security benefits for a qualifying disability. Again, this is not to say that adults getting benefits are more likely to harm a child but that their prevalence tells us something important about the likelihood of child maltreatment in an area.

She then goes on to establish the, to me, non-obvious point that Social Security for a qualifying disability is highly correlated with substance abuse and mental illness.

Dorothy Mandell, a professor of community health at the University of Texas and one of the model’s creators, notes that “if an adult is under age 65 and receiving disability, chances are very high that adult has mental health problems or substance use problems.” The order in which these things occur can vary. As Mandel says, “there is a high association with adults who go on disability to then develop depression. But some go on disability for chronic mental health problems. And we can’t ignore the role of chronic injury and pain medication abuse either.”

The proposition that kids are most likely to suffer abuse in households with a high degree of substance abuse and/or mental illness seems entirely to be expected.  But to have the evidence to support that conclusion is remarkable.

It is hard to quickly see how this might be operationalized to improve the future prospects of children in these environments but it certainly is worth considering.  And it reinforces the general impression that substance abuse and mental illness are upstream from many (most?) social ills including homelessness, child abuse, unemployment, etc.  

As importantly, it directs focus away from the most common, and usually most useless, obsession - poverty.  

Mandel says that advocates regularly ask her about poverty: Isn’t it strongly correlated with child-maltreatment risk in a community? It is not uncommon to hear people say that giving families more material help in the form of housing vouchers or food stamps or just cash would reduce the number of kids reported for neglect. But that is not what the Texas researchers found. While a higher level of correlation existed between poverty and maltreatment in older children, disability benefits were still “overwhelmingly” the strongest correlation for all age groups.

These findings suggest not only that something is wrong with the “neglect is really about poverty” narrative but also that policymakers might better target the crisis. If a community is suffering from elevated substance abuse and mental illness, sending it more cash will not solve things. Opening more rehab facilities may be more effective. Going upstream to fix a problem is fine, but it’s important to look at the actual evidence that we find there.