The views here are my own. They do not represent my employer or any organization I am affiliated with.

Last week, the Wall Street Journal published an investigation looking at five large autism therapy chains owned by investment firms. Investment firms are large corporations that buy smaller companies to grow them bigger and then sell them for a profit. According to the article, in 2024, these autism therapy chains billed Medicaid for 29% more therapy hours per week than other companies (Mathews et al., 2026). Old employees from these autism therapy chains say they felt pressure to maximize hours, which the companies dispute. With some colleagues we have had concerns and long conversations on this for the last 10-15 years as things have been changing for the behavior analytic practice with insurance coverage and the emergence of large autism therapy chains.

I remember in 2006 when I started working in Houston, ABA services were delivered at university clinics, homes, or by small, independent practices. Most ABA centers had one owner, most often was the BCBA, and as employees some therapists. The larger centers had more BCBAs or BCaBAs. These were the good old days before the birth of the behavioral technician certification. Ugh, we will talk about the controversy around the behavior technician credential in another blog.

As large chains started opening autism centers in the last 10-15 years, parts of the ABA service started to become standardized, and the quality of care may not always be a priority. Financial incentives do create conflicts of interest, and the article is asking how it influences treatment hours. A researcher told the Journal: “If it costs you money to do 10 hours a week, but you profit at 40 hours a week, even if you’re ethical, how do you balance those incentives?” (Mathews et al., 2026).

The million dollar question: does every child need 40 hours of therapy every week? A full-time schedule for a 2- or 3-year-old looks like an adult workday for a child who still needs naps, playtime, and time with family. More hours don’t automatically mean better outcomes. These important decisions need empirical evidence.

The 30 to 40 hour model is based on research from the 1980s, including a study of 19 children who started full-time therapy by age 4 (Lovaas, 1987; Mathews et al., 2026). Reviews of early intensive behavioral intervention suggest that it helps many children, but the evidence comes mostly from small studies, and those studies compare intensive programs with usual care (Reichow et al., 2018). They don’t test whether 40 hours work better than 20. Some studies have compared different amounts of therapy. One found that children who received about 30 hours a week gained more than children who received about 12 (Reed et al., 2007), and the Council of Autism Service Providers has said that children who receive 30 or more hours of ABA a week are twice as likely to close cognitive development gaps than children who receive fewer than 12 (Council of Autism Service Providers, 2026). The study behind that claim is a 2026 meta-analysis of individual participant data from fifteen studies, which found that treatment intensity contributed to change across adaptive behavior, intellectual functioning and autism severity (Eldevik et al., 2026). Its authors also report that every study they included carried a serious risk of bias, because none randomly assigned children to groups. In these comparisons, families were not randomly assigned, so the groups may have differed in more than hours. The one randomized trial that compared hours directly found no significant difference between 15 and 25 hours a week for toddlers (Rogers et al., 2021). To my knowledge, a randomized trial comparing 20 hours of ABA with 40 has not been done yet. One large analysis found more goals mastered with more hours, in children averaging about 15 hours a week (Linstead et al., 2017). A 2026 study of 725 children found that more hours predicted more goals met, but slower progress in adaptive behavior and communication. Dangerous behavior decreased whatever the number of hours, and a child’s communication skills at the start predicted progress better than hours did (Samelson et al., 2026). Recent federal Medicaid guidance calls 40 hours a week for young children “not a best practice” and notes that children need time for napping and eating (Centers for Medicare & Medicaid Services, 2026; Mathews et al., 2026). We did not participate in a study, but I remember twelve years ago, we provided ABA services for young children for 15 hours per week and had strong, wonderful outcomes.

An important note from reading the WSJ article: billing data can show a pattern, but they don’t suggest a conclusion. A gap in hours can’t tell us whether treatment was necessary, effective or inappropriate. For that we need each child’s assessment results that show needs, goals, progress and quality of care. The Journal also notes that several companies said their patients have more severe needs, and that Learn Behavioral billed fewer hours than its local competitors, so it’s important to individualize (Mathews et al., 2026). Also, looking at children who moved between providers is a way to compare the same children (Mathews et al., 2026). But hours alone can’t tell us whether more intensive services were clinically justified. Outcome data though could.

Some stories are scary. The accounts of a child woken from a nap, or of families asked to delay kindergarten, are serious (Mathews et al., 2026). The large autism companies dispute them, and individual cases can’t tell us how common these practices are. As a practitioner, I have heard stories like this.

Behavior analysts and families have described the same pattern for years: with long waiting lists, some ABA centers accept only children who will attend the full 40 hour time. Former employees quoted by the Journal said centers turned away children who needed fewer hours, and one family was told their son could keep his place only if he kept his full-time schedule (Mathews et al., 2026). For a family that has waited months, that is not a free choice. Hours should follow the child’s needs, not the provider’s schedule.

Who delivers the therapy hours? Most therapy hours are delivered by behavior technicians, who can begin work with a high school diploma and a 40-hour training course. Technicians are often paid about $20 an hour (or less), while Medicaid pays companies about $60 (Mathews et al., 2026). Most are paid only for the hours they work with a child, so when a child is sick or a session is cancelled, they lose that pay. Although their income changes from month to month, their rent and bills don’t. So, only few people can build a career on that, so many leave the ABA centers after a short time, and children lose familiar instructors often. The least experienced staff often work with the children who have the most complex needs. Quality care needs a stable workforce, and a stable workforce needs salaries, like the schools provide. The more hours a model requires, the more it depends on staff who are paid the least and leave the soonest.

Insurance companies decide what is authorized and billable, and their limits often push ABA centers to cut corners. Assessments are often given too few hours to understand a child’s skills and learning style well. Reviewing the research is given no time at all, yet that is what the science of behavior analysis is, reviewing the literature and choosing the interventions that work best for each child. When the time for assessment and research disappears, plans start to look copied and pasted from one child to the next. If we want individualized care, the system we use must allow pay for the work that makes it possible.

What would help is outcome data on communication, daily living skills, independence, and participation in school and community, plus more clear evidence on how clinicians decided on hours and how those decisions compared with each child’s assessed needs.

Some thoughts, the 30 to 40 hours per week model, may not be best for young children, it is not supported by the evidence showing individualized decisions based on each child’s needs. Intensive programs are right for some children, based on needs, and progress should be measured and hours reduced as goals are met. Clinical decisions must put the child first. Hours should be individualized, children need time to sleep, play, attend school and be with family. ABA centers should be transparent about outcomes and about how they set hours, and families should be able to choose kindergarten without losing their place in therapy.

Here are some questions I wish more families would ask

Billing and business practices need a fresh review. Intensive ABA is right for some children, and billing records alone can’t tell us otherwise. But the investigation raises a legitimate question about financial incentives.

Behavior analysis is a wonderful science. As a field, we have allowed profit-driven companies to manipulate and cut corners, and our clients and our profession are paying the price. We need to do better.