Download this column (pdf)
Read the column in your web browser (Scribd)
Social Impact Bonds, or SIBs, are a promising new model for financing social programs that turn traditional government funding structures on their head. Instead of paying contractors or grantees upfront for a set of services, Social Impact Bonds allow government to focus funds on approaches that work—without paying a dime if agreed-upon outcomes aren’t achieved. There is particular interest among state and local governments, foundations, and service providers in using Social Impact Bonds to pay for preventive programs in areas such as reducing prisoner recidivism and homelessness—the kinds of services that often save government money down the road but currently face budget cuts.
How Social Impact Bonds work
Social Impact Bonds are relatively straightforward. In a SIB agreement, a government agency decides on an outcome it wants achieved and contracts an external organization that promises to achieve the outcome. The external organization is free to develop and implement its own strategy. If the external organization succeeds, the government pays them a sum of money that may be based on the expected savings to government from achieving the outcome. If the external organization fails, the government pays nothing.
But Social Impact Bonds also pose significant challenges because they require government agencies to act in unfamiliar ways. Government is used to exerting a great deal of control over social service contractors—a level of control that SIB agreements do not allow.
This means that the agreement itself—the contract signed between the government agency and the external organization that promises to achieve the outcome—is enormously important to a Social Impact Bond’s success. The contract codifies the outcome, payment schedule, and assessment. It also establishes the responsibilities of the government and the external organization, how disputes should be resolved between these two parties, and under which circumstances either party can terminate the agreement.
This column addresses common questions about the roles, responsibilities, and limitations for both government agencies and external organizations in Social Impact Bond arrangements.
Setting the outcome, payment, and assessment method
The contract should clearly establish the outcome the government agency wants achieved. One of the most challenging aspects of a Social Impact Bond is setting an outcome that will not occur absent the external organization’s intervention. But the outcome should not be so difficult to achieve that the external organization has little hope of success. The contract should state what the outcome is, how it will be measured, and the payment schedule for successful outcomes. In some cases the government agency and the external organization may want to appoint an independent assessor to determine whether the external organization has achieved the outcome and received the level of payment due.
If government truly wants the designated outcome achieved, it must fully cooperate and collaborate with the external organization. And the terms and conditions of what “cooperation” entails should be set out in the Social Impact Bond agreement.
At the most basic level, this means the contract should require government to provide the external organization access to the beneficiary population and any information or data about the population that can help the external organization with its work (providing it can be legally shared).
Government should also be open to making changes to its policy or practices to help the external organization succeed. For instance, in a Social Impact Bond agreement aimed at finding jobs for the long-term unemployed, the external organization may ask the government for assistance in getting state IDs issued to program participants to better enable them to apply for employment.
The contract should also place some restrictions on the government. In most SIB agreements this will include clauses prohibiting the government from exerting control over the external organization’s strategy or day-to-day operations. The contract should also prevent the government from intervening in the external organization’s selection of subcontractors and investors, though subcontractors will be held to the same standards as the external organization itself.
The external organization’s responsibilities
The external organization shoulders one overriding responsibility: doing everything reasonable to achieve the outcome.
But the contract will likely set out some limited constraints on the organization’s activities to guarantee the beneficiary population’s safety. For instance, most SIB agreements will include clauses prohibiting the external organization from engaging in activities that they reasonably believe could cause harm. Other clauses may prohibit the external organization from doing things that will harm the government’s reputation or result in increased costs to government or others.
The external organization should be free to modify its strategy and activities, particularly if its original plans are not accomplishing the expected outcome. That’s why the restrictions placed on the external organization in the contract should be broad rather than specific.
Disputes in the Social Impact Bond agreement
Even a cooperative government agency and external organization may disagree with one another. That’s why the contract should include language appointing an arbiter to settle disputes. The arbiter’s key function will be to find ways to deal with disputes quickly and amicably.
In extreme cases, either the external organization or the government agency may wish to terminate a Social Impact Bond agreement and an important part of a SIB contract is the language governing termination and compensation level. An external organization is heavily incentivized to stop providing services and pull out of the arrangement if it begins to think it will not achieve the outcome. The contract should set out an orderly process for termination in these cases.
Similarly, the government may legitimately want to end a SIB arrangement, for example, if it reasonably believes that the external organization is harming the beneficiary population. The government should also be able to terminate a Social Impact Bond agreement “for convenience,” that is, at will. But requiring the government to generously compensate the external organization should make this an unattractive proposition.
Our issue brief describes the mechanisms for terminating and altering the agreement and goes into detail about how a SIB agreement may be formally structured.
This column is by no means the final word on how these unusual arrangements should work. Social Impact Bonds are still developing in the United States and new models will emerge with their own restrictions and permissions as more states and cities explore the concept. But most agreements will need to formalize the responsibilities and expectations of both parties because Social Impact Bonds require openness, trust, and ongoing communication between government agencies and external organizations.
Next in this series, we will explore the possible subject area applications of Social Impact Bonds.
Jitinder Kohli is a Senior Fellow with the Doing What Works project at the Center for American Progress. Douglas J. Besharov is a professor at the University of Maryland School of Public Policy and a Senior Fellow at the Atlantic Council of the United States. Kristina Costa is a Research Assistant at the Center for American Progress.
The Center for American Progress’s work on Social Impact Bonds is supported by the Rockefeller Foundation.
Download this column (pdf)
Read the column in your web browser (Scribd)
Read other pieces in this series: