I've always been fascinated by the challenge of finding funding for front-end investment for programs that promise downstream, long-term savings. Early childhood education, geriatric fall prevention, prisoner recidivism, permanent supportive housing -- are examples of programs that appear to pay for themselves.

One set of challenges is programmatic: selection of an evidence-based intervention; identification of a sufficiently narrow group of high-risk individuals to avoid prohibitively high costs; execution of the intervention with sufficient fidelity to achieve expected outcomes; and rigorous evaluation to determine whether cost avoidance has been achieved.

The second set of challenges reflects the complexity of funding streams. Often, the agency that may benefit from front-end investments sees no incentive to redirect its own funds to another agency offering the front-end service. Rarely do leadership and incentives line up well enough for a public agency or agencies to provide risk capital on the confidence that savings or benefits will materialize.

In the last decade, a confluence of interests across several sectors has given rise to a different approach to capitalizing these up-front investments. Interchangeably called pay-for-performance bonds or social impact bonds, these lending vehicles are instruments through which high-net-worth individuals, foundations and financial institutions with community lending obligations invest in promising interventions while accepting higher risk and lower returns than they might from conventional investment vehicles. (An excellent toolkit for the practice is found at payforsuccess.org.) Public agencies pay back the loans only with proof of desired outcomes or cost avoidance as spelled out in the contract. Intermediary entities such as Social Finance US have formed to provide necessary connections and support in structuring these transactions.

The field is young and proliferating. A Stanford Innovation Review paper from 2014 traces an acceleration of social impact bonds globally but notes challenges resulting from the practice's immaturity. Youthful enthusiasm in the field may well outpace development of accepted standards of "exemplary" programs, public officials' understanding of the rewards and risks, and commonly accepted and pragmatic definitions of success that would trigger repayment of investments.

In Philadelphia, two pay-for-performance bonds are being explored for feasibility. The first is with the Philadelphia prison system to reduce the rate of recidivism of young adult males; the second, with the city's human-services department and the Philadelphia school district, aims to reduce placements outside of the city of for children in the child-welfare system.

The city chose these two areas for examination in part because of their high financial and human cost. They are also problems for which there are tested approaches for addressing effectively. The feasibility studies underway with an intermediary contractor involve deep dives into current programming, current costs and effective strategies that either have not been tried or need scaling up to have significant impact. The aim is to propose cost-effective interventions that are "bankable" to potential investors.

For recidivism, possible strategies include workforce and educational programs that are available pre-sentencing, in prison and post-release. To reduce out-of-town placements of troubled children and youth, the hope is that strong behavioral health supports in the community for both children and families can slow foster-care and institutional placements, which disrupt family ties as well as educational continuity.

Pay for performance and evidenced-based programming are the currency of social-service practice today. One might ask what this innovation really brings to the field. One advantage, of course, is "new money" from prospective investors who have traditionally shied away from direct support of governmental programs. The structuring of multi-year contracts, identification of high-quality providers and independent verification of results provide a level of assurance that proffered funding is not simply flowing ineffectually into general funds.

But more importantly, the process brings expertise to guide a rigorous analysis of evidenced-based interventions and their costs and benefits, an assessment of provider and agency readiness, and independent verification of outcomes. This is a level of focused discipline that is often missing in public agencies. If it succeeds, it could become a broader best practice which could ultimately provide more confidence in front-end investments that are hard to fund in times of tight budgets.