This is what happened when Goldman Sachs cash paid for teenage inmates on Rikers Island to try a bizarre therapy tinged with Scientology.
In this July 31, 2014 file photo, an inmate at Rikers Island juvenile detention facility in New York City carries a plastic fork behind his back. (AP Photo/Julie Jacobson, File)
In a tale that could only come from America's dystopian justice system, investment banking giant Goldman Sachs funded a failed attempt to reduce recidivism among inmates at New York City's Rikers Island jail by teaching teenagers that their own moral weakness is the source of their problems.
The project's disappointing results were published last year, but the bank's investors weren't exactly going out on a limb with the experiment. The treatment used in the study is known as Moral Reconation Therapy (MRT), and it's listed in a US government database ostensibly designed to spotlight programs backed by scientific data. About a million inmates across America have been exposed to it, according to proponents. Yet the story of MRT at Rikers shows how little thought is typically given to programs intended to help people with addiction and other problems behind bars, and how miserable the evidence base for such treatment really is.
The saga also suggests that even when corporations want to do good, the expert advice they receive is often problematic at best.
Sarah Beller at the Influence (a partner of VICE and an outlet for which I also write) first exposed the extent of MRT's problems in a stunning feature this June, revealing it to be popular with drug courts and widely used in lock-ups in dozens of states. (She also found weird overlap between the program's moral grid and that of Scientology—more on that later.) But despite being introduced decades ago, this supposedly science-based treatment contradicts the modern, medicine-based approach to addiction.
As a 2007 review by the Justice Department explained, "The underlying theory of MRT is that offenders and drug abusers have low moral reasoning." Academics Greg Little and Ken Robinson introduced it in the 1980s and say they based it on theories of moral development devised by psychologist Lawrence Kohlberg, who studied how children develop a sense of right and wrong.
The term "reconation" was coined by MRT's developers to describe the change in moral thinking they seek to inspire. Supporters might say the word "moral" is simply a reference to Kohlberg's work, but it's hard not to see the idea that someone has "low" moral development as at least a bit judgmental. Indeed, much of MRT is about getting prisoners to take complete responsibility for their situation—rather than exploring the role of factors like poverty, unemployment, poor schooling, mental illness, and past trauma.
"In many forms of pop psychology, you are told to blame yourself because you could have everything, you could be on top of the world and you just haven't focused, you haven't worked hard enough," says Barbara Ehrenreich, bestselling author of Bright Sided: How The Relentless Promotion of Positive Thinking Has Undermined America. This, obviously, is in the interest of those who want to keep America's political and financial systems the way they are—if people start to suspect the system is rigged, they might stop beating themselves up and look around for other villains.
MRT, however, doesn't even take the optimistic tone self help usually does. One piece of material, for example, tells readers, "You have probably become aware of the sorry state of your life and the fact that you, alone, are responsible for where you are," and calls those who refuse to start the program "dishonest."
As Ehrenreich explains, this isn't exactly inspirational stuff. "It's not enough to be in prison, which means constant humiliation—and if you are in, it's probably because you're really poor," she says. "On top of that, you have to participate in trashing yourself and that seems to me to be a whole new level." Indeed, research on other programs that confront and shame addicted participants has shown repeatedly that this is counterproductive.
MRT veers even further away from science by apparently borrowing concepts from Scientology. The language it uses to describe its successive phases of moral development are awfully similar to those contained in Scientology's "Life Conditions." According to MRT materials obtained by VICE, we all begin childhood in the moral stage of "disloyalty," which involves dishonesty, "pretense," and "victimizing." Scientology labels a similar, early stage in its moral hierarchy "treason." The idea is that children cry not primarily because they are in need, but because they want to manipulate and control their parents. Under either name, it's a strange way to see babies and toddlers—and is contradicted by research showing children as young as 18 months old will typically try to help others without even being asked, if they know how to do so.
Also, four of MRT's phases have the same name as the comparable Scientology stage. Beller, of the Influence, notes that MRT credits a man named Ron Smothermon with developing and naming the steps of what it calls the "freedom ladder." Smothermon is a self-help author connected to the "human potential movement" of the 1970s and may have been influenced by Scientology.
(Little insisted in an email to VICE, "I know nothing at all about Scientology, except what I saw on a single episode of South Park.")
But let's leave the program's apparent debt to Scientology's rhetoric aside. At least six published evaluation studies of MRT are listed in PubMed—the official US database of scientific studies—as well as dozens of papers that are unpublished or remain outside the peer-reviewed literature. A 2013 meta-analysis of data from 33 studies, which included more than 30,000 participants, found an average recidivism reduction of around one third. That sounds impressive, but 50 percent of the studies showed no effect at all, and the overall "effect size" was calculated to be 0.16
Effect size is a measure of how much of a difference something makes— researchers typically rate them as "small," "medium," or "large," and obviously for something like reducing recidivism, the bigger the effect, the better. The typical cut-off for being seen as having a enough of an effect to make a real world difference is .20, though the authors of the MRT meta-analysis note that many widely used prison programs have similarly small effects.
That's another reason it seems especially odd a program with such small and often null results would be listed on an official US website highlighting treatments that are evidence-based and effective. Nonetheless, MRT can be found in the National Registry of Evidence Based Programs and Practices (NREPP), run by the federal government's Substance Abuse and Mental Health Services Administration (SAMHSA). (It is listed as a "legacy" program.)
I reached out to the feds to ask why MRT is included and what that designation means. In an email statement, a SAMHSA spokesperson explained that the registry is undergoing a major transition: In July 2015, it changed its mission from being a database of programs that work to one that lists any program about which there is any evidence, positive or negative.
"Now evaluations are published whether a program is harmful or ineffective, as well as if it is promising or effective," the spokesman said. "This changes the entire nature of the registry from a list of evidence-based programs to a list of all programs that demonstrates not only what is effective, but also what is harmful or ineffective so communities can better invest scarce resources."
That sounds like an excellent idea. But it should probably be accompanied by a change in the name of the database, which as of now implies that it includes only programs backed by evidence. SAMHSA is aware of the contradiction, its spokesperson adding, "As NREPP transitions into its more mature phase as a registry, the appropriateness of the name will be considered."
In the meantime, a disclaimer on the website would help. After all, programs like MRT use NREPP as something like a Good Housekeeping seal of approval; on the treatment's website, its inclusion in NREPP is cited as "recognition" that it is "an evidence based program." In the statement, SAMHSA's spokesperson added they are currently reviewing the data on MRT and other "legacy" programs that were reviewed before 2015, but this will not be completed for several years.
For his part, Little thinks his program was implemented on Rikers in a way that made it unlikely to work. "MRT is generally designed for implementation in more than 30 sessions," he wrote VICE. "As I understand it, the 'clients' at Rikers came to an average of less than 7 sessions, the median number was three meetings as I heard it. To me, it was designed to fail at some level."
If there's any good news here, it is that the poor outcome cost a rich bank $1.2 million, rather than taxpayers. That's because the research was conducted as one of the first US experiments in what are known as "social impact bonds." The idea is to get companies to invest in programs that might cut government costs by improving outcomes. If it works, the investors make money and the government wins, too, by gaining data and early implementation of an effective program. If the program fails, however, only the investors lose money. (Through a spokesperson, the bank declined to comment other than to say, "Goldman Sachs funded a social impact bond to help reduce recidivism on Rikers—the funding was not for research on the therapy.")
The issues run deeper, though. Study participants experienced an outdated and humiliating treatment that basically taught them that their problems lie only in themselves—not in a society that locks up kids even when data shows that this, in and of itself, increases recidivism compared to noncustodial placements. No one involved even seemed to consider what it would be like to get beaten by guards at night—as has been widely documented at Rikers—while getting told each day that their own disobedience is the real problem.
If we want a more just society, maybe it's not prisoners who need to take a moral inventory, but leaders and corporations who accept this corrupt, unequal, and biased justice system as it is. Politicians across the spectrum keep claiming to believe that addiction is a disease. But our courts and jails and prisons—and the programs we choose to study in them—keep saying otherwise.
Follow Maia Szalavitz on Twitter.