One of the biggest pieces of news in the policy wonk world this week was Goldman Sach’s announcement that it has agreed to invest $9.6M in an NYC program addressing recidivism as part of a social impact bond.
Social impact bonds work like this: a private actor, such as Goldman, invests money in a program that has the potential to create savings for a government entity; if the program results in a certain amount of savings, the private actor receives its money back from the government entity; if the program exceeds a certain amount of savings, the private actor receives a profit; if the program results in no savings, the government pays nothing to the private actor.
According to this NYTimes article,
The Goldman money will be used to pay MDRC, a social services provider, to design and oversee the program. If the program reduces recidivism by 10 percent, Goldman would be repaid the full $9.6 million; if recidivism drops more, Goldman could make as much as $2.1 million in profit.
This arrangement has been widely celebrated and has inspired several articles and lots of commentary praising the social impact bond model. Social impact bonds are seen as a solution to the budget woes that city and state governments across the country are facing - private resources are being leveraged to further the public good. What’s not to like?
While social impact bonds no doubt have many positive characteristics and, from a practical perspective, cities may need them right now, there are downsides to these financing mechanisms that must be considered.
In the case of Goldman Sach’s bond, public services are essentially being privatized. A nonprofit is being given the responsibility to run a program to enhance the criminal justice system’s success. In addition, it is getting funding for the program from a private actor whose organizational mission is to make profit off of investments. As with any form of privatization, this dynamic poses some skewed incentives. For example, given the profit motive that can be seen to be driving Goldman’s investment, Goldman may pressure the nonprofit to make decisions that help it reach its goal even though the approach taken could result in substantial negative externalities.
The social impact bond model also may have the effect of the profit motive determining which services will be provided rather than the public good. It is reasonable to assume that private actors will only invest in a program or services that they think will realize a reasonably-sized profit - Goldman Sachs didn’t become Goldman Sachs by investing in programs that yield small returns. And, private investment may create a moral hazard situation for governments. With the private sector focusing on an area of public concern, such as recidivism, governments may feel they do not need allocate resources towards addressing this area of concern. Hence, the profit motive could essentially become the main driver of which services are provided even if, as a result of externalities, another approach may be better for the public good.
This last argument foreshadows an even bigger concern that can be applied to any form of privatization. The social impact bond model boils down to an entity with individualistic concerns, the profit motive, being given the responsibility to address public issues. A more community-driven approach may be necessary to ensure all of the public is considered. Sure, we can assert that the contracts involved with privatization align the incentives of the private and the public, but this is a difficult task, especially when one party is extremely desperate as cities often are when they turn to private actors for assistance.
It would be foolish to say that the social impact bond model should be abandoned because of the concerns raised here - the model is a creative, pragmatic solution to an important problem. But, as with all policy approaches, we should make sure we consider the other side.