This week Bill O'Reilly, who recently attacked the Drug Policy Alliance for having the gall to question the war on drugs in a commercial featuring Sting, debated DPA Executive Director Ethan Nadelmann (see video below). O'Reilly once again repeated his favorite drug-related factoid, this time attributing it to the prohibitionist propaganda mill known as the Center on Addiction and Substance Abuse (CASA):
They say 70 percent of abused and neglected children in the USA have alcohol- or drug-involved parents.
O'Reilly has made two crucial changes to this claim since the last time he made it. In his first segment about the DPA ad, he asserted:
Seventy percent of child abuse and neglect in this country is substance abuse driven, most of it narcotics.
O'Reilly is no longer claiming that "most" of these cases involve "narcotics." Since CASA's estimate does not distinguish between alcohol and illegal drugs, he had no basis for saying that to begin with. Just as important, O'Reilly is no longer claiming that 70 percent of child abuse and neglect is "driven" by substance abuse. Since CASA's estimate is based on a correlation, the conclusion that drinking or drug use "cause" people to abuse or neglect their children is not justified. It is not hard to think of alternative explanations. People who are unhappy or have weak impulse control, for instance, may be prone to abuse children as well intoxicants.
The distinction between correlation and causation is important, since O'Reilly, like most prohibitionists, wants to ban certain substances because he believes they take control of people and make them do bad things (an idea I criticize in my book Saying Yes). But it's a distinction that CASA itself frequently ignores. While a 2007 CASA press release said "70 percent of abused and neglected children have alcohol and/or drug abusing parents," a 2001 statement said "at least 70 percent of the cases of abuse and neglect [handled by child welfare systems] stem from alcohol- and drug-abusing parents" (emphasis added). According to a 2004 CASA report (PDF), "Approximately 70 percent of all cases of neglect and abuse are caused or exacerbated by substance abuse and addiction" (emphasis added). CASA cites itself as the source of this figure—specifically, a 1999 report titled No Safe Haven: Children of Substance-Abusing Parents (PDF). But the caveats in that report do not inspire confidence in CASA's estimate (emphasis added):
Reliable national data documenting the prevalence of substance abuse among child welfare cases is not available....
The data that are available suffer from three major methodological problems that make it impossible to confirm the prevalence of substance involvement among child welfare cases. First, study samples may not be large enough to account for sampling errors. Second, samples represent only certain areas of the country. Third, the definitions of substance abuse and addiction vary.
Little methodologically strong data regarding parents involved with the child welfare system exist....Research establishing the prevalence of substance involvement (use, abuse or dependence) generally relies on inconsistent definitions of these terms and of the degree of substance involvement. Moreover, studies are inconsistent in defining whether substance involvement is the primary or causal reason for a parent's involvement with the child welfare system or whether substance involvement is an ancillary or co-occurring problem.
Despite these difficulties, CASA bravely "estimates that substance abuse causes or contributes to about 70 percent of child welfare cases." In doing so, it relies on studies indicating a prevalence of substance abuse of "50 to 78 percent" among parents of abused or neglected children (these are cases where at least one parent is said to be a substance abuser). It settles on a number toward the upper end of that range based on its own survey in which it asked "child welfare professionals" to estimate the share of their cases in which "substance abuse causes or contributes to" abuse or neglect. In short, CASA arrives at its 70 percent figure by combining shaky prevalence numbers that tell us nothing about causality with the impressionistic reports of social workers.
This mushy estimate throws illegal drugs together with alcohol, conflates use ("substance involvement") with abuse, treats a mother who drinks because she's upset about her husband's abuse of their children the same as an angry drunk who hits his kids, and fails to distinguish between substance abusers who abuse their children because the drugs made them do it (O'Reilly's interpretation) and drug abusers who abuse their children because the same personality traits and environmental factors contribute to both kinds of anti-social behavior. And O'Reilly is using this number to argue that people can't be allowed to smoke pot (by far the most popular illegal drug) because it will make them beat their children (or maybe forget to check on them in the tub).
Granted, someone who is drunk or stoned all the time is not likely to be a good parent. But since the vast majority of drinkers and drug users don't mistreat their children, focusing on the intoxicants, rather than the reasons some people use them to excess, does not make much sense.
I reviewed CASA Chairman Joe Califano's book High Society in the April 2008 issue of Reason.