Some recent articles in Time Magazine and Huffington Post about the promise of Pay for Success financing have brought the topic forward again and I’ve been asked what happened to the Minnesota pilot to try pay for performance based funding. Good question.
To refresh your memory, the Minnesota Pay for Performance law was signed into law in 2011, ushering in a new experiment in public, private and nonprofit partnerships. It was the first state to pass legislation authorizing social impact financing through a state bond. However, as other pay for performance initiatives have moved forward, Minnesota has failed to close an agreement to issue the appropriations bonds authorized in the Act. What happened, and why, after 4 years, has this stalled out?
To find out, Propel Nonprofits (formerly Nonprofits Assistance Fund) engaged a research team from the Humphrey School of Public Affairs last year to dig into the question and identify some lessons learned. The resulting report, Pay for Performance Financing to Expand Cost-Effective Social Services: Lessons Learned in Minnesota, draws on documents and interviews and concludes with seven lessons learned. We hope that the report will be valuable for stakeholders in public, private, and nonprofit sectors to examine and determine whether and how to move forward.
At the heart of both the lessons learned and the recommendations are critical structural and model distinctions between the Minnesota Pay for Performance bond pilot and other versions of Social Impact Bond (SIB) or Pay for Success transactions that have been implemented elsewhere.
From my perspective, the key issue and obstacle for the Minnesota pilot has been the assignment of risk. If the services that are delivered do not achieve the intended outcomes, who will bear the risk and absorb any losses? In the Minnesota case, because there are investors who purchase a bond, it appears that the investors are taking on risk that the state can avoid. In reality, though, the risk in the structure approved by the legislature transfers the risk to the nonprofit service providers. This is why Propel Nonprofits has been involved in this policy area. During the four years of exploration and development there were attempts to shift or mitigate that risk by creating intermediaries and a working capital pool, but none of these solved this central problem.
Let’s talk about what this isn’t.
Who takes the risk in SIBs? The idea behind Pay for Success based financing is to expand effective social programs by creating a new financial model that captures the long term savings to the public. Because outcomes almost always take time to be achieved, the risk of whether or not they are achieved is shifted away from the public entity (city or state, for example) to others. In the Social Impact Bond (SIB) model, the risk is transferred to private investors with a financial structure that factors in the risks, costs, and rewards. Whether or not this is the actual practice is explored in The Payoff of Pay for Success published in Stanford Social Innovation Review this summer. The service provider (often a nonprofit) doesn’t bear the risk – they are paid for delivering services that are designed to achieve the outcomes. In these structures, the public entity has little or no financial/budget risk because they will only pay on the contract if the outcomes can be documented. This was demonstrated in the Riker’s Island project. The private investors (Goldman Sachs, with a guarantee from Bloomberg Philanthropies) took a loss when the outcomes were not achieved. They were investors who took a financial risk on a program designed for social impact.
How is the Minnesota pilot structure different?
The pilot project adopted in Minnesota in 2011 is structured very differently, and here is the deficiency. In this structure, the State is directed to develop contracts with service providers that require the State to make payments only if and when the provider can demonstrate outcomes, even if that takes several years to achieve. In the meantime, the service providers (likely to be nonprofits) are on their own to pay the staff and other costs required to deliver the services to their clients that are designed to achieve the outcomes. Even in the best case scenario (that all of the intended outcomes are achieved) the service providers will need to have available working capital for what could be multiple years before payments are made under the contract. If the contracted outcome targets are not reached, the nonprofit might not recoup their costs.
Private Investor Risk is Low
The only private investors involved are those who purchase a State of Minnesota Appropriations Bond. Unlike the SIB model, these investors aren’t investing in social impact by taking a risk on whether or not the outcomes are achieved. The funds received from the sale of the bonds are to be held in a fund at the State until the contract terms have been met and payment is made. The only risk for these investors is whether or not future legislatures appropriate funds to repay the bond. Since the State of Minnesota is highly likely to repay any bonds, the risk is pretty low. There were efforts made to reduce the risk to the service providers by creating an intermediary to take some of the risk, or to create a working capital loan fund, but each of these ran aground for the same reason. Someone has to bear the risk of loss. Acting as if the risk isn’t there doesn’t make it go away.
I’m not ready to conclude that there is no place for pay for success financing in Minnesota, but the 2011 pilot design is not the model I’d recommend. We will be following up on this report with additional opportunities to discuss the pilot, the lessons learned, and the broader topic of pay for success financing. Please stay tuned, and let me know what you think.