A Special Bond: How to Generate Social and Financial Returns

Over 40 percent of released prisoners in the U.S. return to prison within three years. In New York City, almost 50 percent of adolescents released return within a year, and that high recidivism rate presented a societal problem that New York City wanted to solve. The question was how.

Financial investment to solve problems facing society is nothing new. For centuries, governments, nonprofits, private philanthropies and businesses have worked to address social needs.

But given today’s environment, with its tight government and nonprofit budgets and limited capacity for philanthropies, Darden Professor Mary Margaret Frank became interested in examining innovative options that could spur private capital to take a more direct role in improving society.

Frank’s exploration led her to social impact bonds — a new and unique public-private partnership that raises private capital to invest in programs designed to improve social issues. Frank led a financial innovations roundtable at the 2016 Concordia Summit on 20 September, at which she framed how governments, nonprofits and private investors can work together to improve society through this new investment partnership.

“Social impact bonds are exciting because they are helping bring together sectors of the economy in a way that forces them to understand one another’s roles so that risk can be evaluated effectively and price can be assessed,” Frank said. “This collaboration brings a greater understanding of what works to measurably improve lives.”

The Old Model

The traditional model for addressing social issues involves three steps:

1. Identify the social issue: defined by governments
2. Operationalize an intervention: executed by the governments
3. Fund the actions: paid for by governments through tax revenue

However, Frank said the old model, while effective in certain situations, has limitations.

“Across the U.S., we spend trillions of dollars every year toward improving lives, but unfortunately know very little about whether those dollars are leading to meaningful outcomes.”

While the reasons for failure are complex, they often boil down to a lack of innovative solutions in Step 2 because governments’ incentives encourage risk-aversion or don’t address outcomes or Step 3, in which governments are limited by budget.

Overcoming the Obstacles

Social impact bonds are designed to overcome the limitations that have emerged with Steps 2 and 3 by creating a public-private partnership to come up with a plan of action and engaging the private sector as an investor to fund the plan. The updated model’s three steps remain, with different players involved through a partnership with the government, nonprofits and private investors:

1. Identify social issue: defined by governments
2. Operationalize an intervention: executed by a third party, typically a nonprofit
3. Fund the actions: paid for by private capital

Funding With a Focus

The premise is simple: Offer the private sector an opportunity to invest — and earn a return — and the market for capital to fix pressing societal issues will grow. But where does the money for the return on the investment come from?

With social impact bonds, private investors front the funding to put in place a plan of actions to achieve a positive social outcome that will save the government money. If the plan of actions results in an agreed upon outcome, which has been determined will provide the government cost savings, the government uses the cost savings to pay back that investment with a return.

“Through the framework, private investors and philanthropies pay for the program costs upfront,” Frank said. “The government only pays for achieved outcomes on the back end after an evaluator determines whether the program met its target outcomes.”

The Challenge of Assessing Risk — Rikers Island Case in Point

While the premise is simple, social impact bonds are more like private equity investments than actual bonds. The practice of defining risk — and, thereby, appropriate returns — among the parties in the partnership is extremely difficult. In each step, risk factors can include:

• Step 1: The parties must be able to estimate how much improving a social issue will achieve in cost savings for any government before it agrees to potentially pay a return to an investor. Based on imperfect historical data, the quantification of projected cost savings are a risk.
• Step 2: The parties must be able to develop and execute a plan that will achieve agreed-upon outcomes within three to five years, because most investors are not willing to wait 10 years or more for a return. These plans need to have evidence of success to reduce risks, but success in one setting does not necessarily translate to another.
• Step 3: Improved outcomes are not enough. They must be improved relative to a control group. The choice of the control group imposes a risk to the investors because it determines success and thus whether the investors are paid. Negotiation over what outcomes are considered success is important.

Take the case of reducing adolescent recidivism at the Rikers Island prison. The social challenge brought together many parties, including New York City Corrections, nonprofits including MDRC and the Vera Institute, the Bloomberg Philanthropies and Goldman Sachs to tackle the issue, which would mean better outcomes for former prisoners and savings on correction costs for the city.

Goldman Sachs agreed to invest $9.6 million over six years in exchange for receiving a maximum of $11.7 million — a return of $2.1 million — by 2017, if recidivism at Rikers were reduced by at least 20 percent. Goldman would break even if recidivism dropped by 10 percent. To limit the bank’s financial risk given the limited return, Bloomberg Philanthropies pledged to cover up to $7.2 million of Goldman’s losses if the project failed, as well as paid the intermediary to reduce the transactions costs.

Ultimately, however, the partnership failed to achieve the agreed upon minimum reduction, as recidivism at Rikers fell only 8.3 percent in the first three years, leading Goldman to close out its investment early.

Frank said this case highlights the difficulty of properly establishing risk and return in a social impact bond. She said, essentially, the government won by paying nothing while achieving a still-solid drop in recidivism. However, Frank argues that no savvy investor like Goldman would be willing to sign up for such an investment again after it did not earn anything, despite a positive social outcome that fell short of the break-even outcome.

Shaping Social Impact Bonds for Success

“Partnerships need to be very careful about how they structure returns — financial returns and the social returns,” Frank said. Initial ideas to improve the nascent asset class include:

• Choose social issues for which there is strong evidence in a variety of settings that an intervention will result in a positive social outcome.
• Chose social issues for which improvement brings large, clear, measurable value to governments.
• Given limited returns, create new ways, such as guarantees by willing donors, to limit downside risk to private investors.
• Find creative ways to reduce high transaction costs without increasing risks.

“Structuring the appropriate return for the risk taken is still in its infancy,” Frank said. “The more social impact bonds that are started, the more we learn about how to structure them appropriately, the more comfortable people will become with this type of financing — thus reducing the uncertainty to investors and lowering the return they require.”