In The Politics of Autism, I write:
As long as government funds so much research, politics will shape the questions that scientists ask and determine the kinds of research that receive funding. Politics will even influence which scientists the policymakers will believe and which findings will guide public policy. In the end, science cannot tell us what kinds of outcomes we should want. ABA “works” in the sense that it helps some autistic people become more like their typically developing peers. Most parents regard such an outcome as desirable, but not all people on the spectrum agree.
Whether ABA is helpful or harmful has become a highly contentious topic—such a flashpoint that few people who aren't already advocates are willing to speak about it publicly. Many who were asked to be interviewed for this article declined, saying they anticipate negative feedback no matter which side they are on. One woman who blogs with her daughter who has autism says she had to shut down comments on a post that was critical of their experience with an intensive ABA program because the volume of comments—many from ABA therapists defending the therapy—was so high. Shannon Des Roches Rosa, co-founder of the influential advocacy group Thinking Person’s Guide to Autism, says that when she posts about ABA on the group’s Facebook page, she must set aside days to moderate comments.
Given the diversity of treatments, it’s hard to get a handle on the evidence base of ABA. There is no one study that proves it works. It’s difficult to enroll children with autism in a study to test a new therapy, and especially to enroll them in control groups. Most parents are eager to begin treating their children with the therapy that is the standard of care.
There is a large body of research on ABA, but few studies meet the gold standard of the randomized trial. In fact, the first randomized trial of any version of ABA after Lovaas’ 1987 paper wasn’t published until 2010. It found that toddlers who received ESDM therapy for 20 hours a week over a two-year period made significant gains over those who got the usual care available in the community.
That year, a report from the U.S. Department of Education’s What Works Clearinghouse, a source of scientific evidence for education practices, found that of 58 studies on Lovaas’ ABA model, only one met its standards, and another met them only with reservations.