By Jesse Hajer

On Wednesday October 18th, the Province of Manitoba announced that it had hired the MaRS Centre for Impact Investing to design a “made-in-Manitoba” Social Impact Bond strategy. Social Impact Bonds (SIBs) are a relatively new type of contract between governments and the private sector to enable the delivery of social services in areas such as health, education, job training, early childhood development, child welfare, housing and crime prevention. SIBs focus on preventing undesirable social outcomes from occurring in the first place and can in theory end up saving the government money, although this has been downplayed more recently by proponents.
Evidence supports the idea that many social services interventions can ‘pay for themselves’. The Washington State Institute for Public Policy, for example, as of May 2017, had analyzed 343 social policy intervention models that were subject to rigorous experimental study, and found that 57 percent saved government more then they cost to deliver, and 76 percent produced a quantifiable benefit to citizens greater than their cost.
Under an SIB government contracts and pays based on outcomes, instead of for the delivery of a specific service. Under an SIB if the private partner fails to deliver on outcomes, its payment will be reduced, potentially to zero. SIBs are an incremental form of privatization, where financing and much of the design, management and decision-making power would otherwise have been under public sector control. With an SIB it is the private partner who manages the program, decides what interventions take place and who delivers services. The private partner also needs to finance the social program until the government is satisfied targeted results have been delivered. The main source of financing under an SIB come from private investors, including banks, investment funds, and charitable foundations.

Investments are paid back by government with a return if the project outcomes are achieved.
If this sounds complicated, it’s because it is, and this has been a concern with the SIB model. SIBs, if they are to have a shot at being cost effective for government, require extensive negotiations and contracts involving lawyers, experts and advisors on both sides. SIBs also often involve the hiring of an ‘intermediary’, as is the case in the Manitoba model, to coordinate between government, and the investors and service providers, and needs to be paid. These extra costs must be offset by improvements in service delivery if SIBs are to provide value for money for government.
Proponents of SIBs claim private sector participation will generate innovation and more efficient social services. By transferring risk and responsibility to the private sector, SIBs are said to incentivise better outcomes. Research on private management and paying for outcomes raises some doubts. Privatization of social service delivery has not been shown to generate significant efficiencies on average, and often reduces quality and involves more precarious work with lower wages and fewer benefits. Paying for outcomes has also a mixed record, and has created ‘gaming’ behaviours, ranging from pure fabrication of data to shifting of effort to meet short term targets, often at the detriment of service quality, requiring costly monitoring activity. SIBs have generated little innovation when it comes to new interventions, and actual risk transfer (based on failed projects where the government does not pay) has been minimal to date.
Manitoba can take steps to reduce the risk that provincial SIBs end up having detrimental impacts. Relying on experienced charitable funders with a solid track record in social service delivery and value-added evaluation, as opposed to private investors with no experience in social service delivery, is likely to generate program improvements and reinvestment when profits are paid out. Choosing indicators and a payment structure jointly with community agencies and providing sufficient funding for fair compensation that does not take advantage of or displace the public-spirited motivations of many NPOs could also help.
If Manitoba is truly interested in value for money and determining the cost effectiveness of SIBs, it could set up a counterfactual conventional model. SIBs are clearly generating more resources for management and evaluation, but these costs are being paid out in the end by governments. Allowing the public and non-profit sectors to compete directly with equal upfront resources could provide some useful insights. Transparent accounting of the funding and savings generated within government would also be beneficial.
SIBs have many desirable attributes, including their emphasis on prevention, longer-term funding for non-profit agencies, and the goals to scale-up and develop more effective social service delivery. SIB proponents have been judging SIBs a success when they meet their targets, and most SIBs appear to be on track to do so. But this should not come as much as a surprise given the added supports built into the model and the replication of proven interventions. The interesting question is not ‘are SIBs meeting their targets’, but are SIBs the best way to improve social outcomes. SIBs in the big picture may be delivering less-for-more rather than more-for-less. It would be nice to find out.

Jesse Hajer is a PhD candidate in Economics at the New School for Social Research. His research is focussed on SIBs.