In 2012, the Adolescent Behavioral Learning Experience (ABLE) program was launched at Rikers Island, financed by the nation’s first Social Impact Bond, an innovative way to fund new programs at no cost to taxpayers. An evidence-based cognitive behavioral therapy program for 16- to 18-year-olds detained at New York City’s Rikers Island jail, ABLE’s goal was to reduce the high recidivism rate for this population.

In a Social Impact Bond (SIB), private investors fund the intervention through a nonprofit contractor and the government repays the funder only if the program meets certain goals. In this particular initiative, Goldman Sachs provided program financing, Bloomberg Philanthropies provided grant support to guarantee a portion of Goldman’s investment, and MDRC oversaw project implementation. Working with the New York City Mayor’s Office and the Department of Correction, MDRC contracted with The Osborne Association, which, in collaboration with Friends of Island Academy, ran the program. The Vera Institute of Justice conducted an independent evaluation.

Today, the Vera Institute released an overview of findings from its one-year assessment of ABLE’s effect on recidivism. It found that the program did not lead to reductions in recidivism for program participants and the program did not meet the target required for the City to pay the investors. The ABLE program will be discontinued as of August 31.

While it is disappointing that the program did not meet its goals, the Social Impact Bond financing arrangement worked as it was supposed to. With the SIB, New York City was able to provide a promising intervention to more than 4,000 young people over the three years the program operated, using financing from the private sector. The City agreed to pay for this program only if it achieved a certain magnitude of impact on reducing recidivism. Because the program did not meet the impact requirements, the City is not paying for a program that did not produce results — a positive outcome for the City and taxpayers.

In addition, the Rikers SIB showed that it was possible for investors and government officials to consider and agree upon specific numerical metrics about what defines success. It is often difficult for governments to quantify program success or failure in this way. Particularly since it was one of the first SIBs attempted, all of the partners knew that this was uncharted territory — with the risk that the program might not meet its goals. The arrangement of this SIB, with private financing backed in part by philanthropy, reflected this understanding of risk.

The experience of implementing the ABLE program at Rikers offers lessons about the challenges associated with scaling and replicating evidence-based programs. Cognitive Behavioral Therapy (CBT) — and the particular intervention provided at Rikers, Moral Reconation Therapy (MRT) — had considerable evidence of prior success. This particular brand of CBT was chosen because it offered the best fit with the Rikers context and had shown positive results even for people who had not completed the entire program. However, the prior research was not definitive, and an MRT program had never been implemented at scale in as challenging an environment as Rikers. As described in MDRC’s 2013 report, the implementation of ABLE faced a number of challenges, including reaching adolescents who did not attend the Rikers school and the indeterminate lengths of stay for many of the young people. However, the program was implemented with overall fidelity to the model. And, according to the Vera Institute, 87 percent of the 16- to 18-year-olds who entered the jail during 2013 and were held for more than six days attended at least one ABLE session, and 44 percent reached a programmatic milestone found in prior studies to be associated with positive outcomes.

Even though the program did not reduce recidivism, the experience of implementing it may leave a positive legacy at Rikers. This program attempted to change part of the culture inside Rikers Island by introducing an intervention for a very high-needs population for whom little to no programming was previously offered. Promoting a culture of sound decision-making and nonviolence inside Rikers was a worthwhile attempt at contributing to positive change. In addition, this project was able to generate significant buy-in from Rikers management, non-uniformed staff and uniformed staff, and school personnel. They worked very closely with the nonprofit service providers to make the program operate on a daily basis.

The partnership among government, nonprofit service deliverers, the financial sector, and philanthropy in the ABLE SIB was forged with a determination to tackle one of New York City’s most challenging problems — recidivism among adolescents incarcerated at Rikers Island. All of the parties understood that it was a high-risk undertaking. While the program did not have the hoped-for effects on recidivism, the experience offers valuable lessons for the constructive role social impact bonds can play in encouraging and supporting innovation in government.

Additional resources:

• Impact Evaluation of the Adolescent Behavioral Learning Experience (ABLE) Program at Rikers Island: Summary of Findings, Vera Institute
• Financing Promising Evidence-Based Programs: Early Lessons from the New York City Social Impact Bond, Timothy Rudd, Elisa Nicoletti, Kristin Misner, and Janae Bonsu, MDRC
• “What We Learned From the Nation’s First Social Impact Bond,” James Anderson and Andrea Phillips, Huffington Post
• “No Success Like Failure: N.Y. Sees Social Impact Bond Pluses,” by Paul Burton, The Bond Buyer