I appreciate the technology that lets me join you.
I am Elizabeth Lower-Basch. I'm the policy coordinator at CLASP, the Center for Law and Social Policy. We seek to improve the lives of low income people by analyzing and advocating for good practices and policies.
Earlier this year we released a report I wrote that looked at social impact bonds, or SIBs, from the perspective of assessing whether they're a good way to expand services for the disadvantaged populations that we care about. The paper looks at the stylized model of a social impact bond, reviews how the projects that are currently under way in the U.S. and in the United Kingdom compare to this model, and lays out the possible benefits and drawbacks of using SIBs. It is based on our review of the literature on SIBs, as well as on our knowledge and experience with performance measurement systems, performance-based contracting, and generally, strategies to link public policy and implementation with research evidence for programs serving low income and other disadvantaged populations.
Let me begin with a quick review of the main features of the stylized model to make sure that we're all on the same page. At the core of a SIB is a prevention or early intervention service. An intermediary organization, not part of the government, lines up private investors to provide upfront funding, which it uses to hire organizations to offer the preventive service.
At the same time, the government enters into a contract with that intermediary, in which it promises to pay a certain amount at the end of the performance period—typically five to seven years—if specified impacts have been achieved. This amount is enough to allow the investor to recoup its upfront investment, and also to make a profit, to compensate it for both the use of the funds, and for the risk involved in the project.
There is risk, because if the specified impacts are not met, the private investor is not paid and loses its investment. In some cases, the amount of the payment is also based on savings that the government expects to realize as a result of the project meeting those specified outcomes.
I'm going to focus on two of the claims that are made about SIBs. One is that they will increase the focus on outcomes of services, rather than inputs, and the other is that they will save governments money.
It's definitely true that SIBs increase the focus on outcomes of services. It pushes to the forefront the question of what are the outcomes we care about, and how much we are willing to pay to achieve them. Because there’s a contract involved that's paid for success, the government has to be clear about what success would look like.
SIBs also force people to think hard about the impacts of programs—the difference between what happens because of the program and what would have happened without them—not just outcomes. This can involve a random assignment evaluation, but doesn't have to. But it does require identification of a counterfactual, in some way, based on either a comparison group or underlying trends with controls. I think everyone agrees that it's a good thing to focus on the impacts of programs in these ways. It’s also important to be careful that what is incentivized is actually what we care about at the core, and not just what's easy to quantify and measure.
I'm not going to get into details today, but in the paper I spend a fair amount of time on the possible distortions that can be caused when we attach high stakes incentives to measures that are only part of what we actually care about. What do I mean by this? If you're looking at, say, a job training program for young adults who are at risk, you might see that it reduces incarceration rates as well as promoting employment. And you might consider entering into a social impact bond to support replicating this program, with the payments tied to incarceration rates. This could have unintended consequences. For example, the program might realize that very few of the women in the control group wind up in jail, so it might decide to only serve young men. Or it might decide that it doesn't really want to focus on job training at all, but only on mentoring or other services aimed at the incarceration, if that's what the payment is tied to. So you just want to think carefully about what your goals are and what would be acceptable.
The second claim that people make is that SIBs will save the government money. It's really important to distinguish here between two distinct claims. The first is that prevention focused services can save the government money, and the second is that SIBs themselves save money.
In the idealized version of the SIB, these go together. The services save the government so much money that government spending can decline even after repaying the investors. There may be some cases where this is possible, but SIBs have been proposed in a range of areas, and in only some of them is there likely to be this sort of savings.
First, it's important to recognize that in not all cases does investing in prevention save the government money. That said, criminal justice is probably one of the areas where the case is strongest that preventive services can save money in a short timeframe. Putting people in jail is very expensive, so there's a lot of potential for savings, even in the short term.
In many other areas, such as early childhood or job training, the preventive services are definitely beneficial to both the participants themselves and to society as a whole, but they may not directly lead to government savings, or they may lead to savings only over a very long time period, and it is not clear that investors are willing to wait that long to be repaid. This is why so many of the early SIBs are focused on justice populations, and particularly on ex-offenders and preventing the incarceration.
But it's also important to understand that for any possible services SIBs are going to be more expensive than funding those same services and activities through a more traditional financing mechanism just because the SIB itself has incremental costs. There are the returns that have to be paid to the investors, there are also the costs of the intermediary and the evaluation, and there's also some upfront costs in negotiating and figuring out what your baselines are and what the contract's going to look like.
I'm not going to get into all the details, but McKinsey & Company has done a pro forma analysis of the costs of a SIB, which I thought was really helpful.
I will note that SIBs can potentially save money if the programs fail to meet the outcomes, and therefore the government does not have to pay the intermediary. This is obviously not desirable. In a mature market, of course, profit motivated investors will demand the profit that accounts for this risk of failure, so it gets built into the overall cost across the portfolio.
There's also a risk that when a SIB is failing to meet its target, it would be rational for the intermediary to cut their losses and stop providing services. In some cases that may be okay, in other cases the government will need to step back in to fill that gap, which itself incurs costs.
It would seem you could save the most money by investing in more preventive services without the social impact bond. Even if you had to borrow the funds, the interest rates would be much lower than the carrying costs of the SIB, but the reality is that in many cases policy-makers, advocates, have not been able to make that successful case for expanding investments in prevention. That seems crazy. It can be very frustrating to those of us who have worked in these areas, so if SIBs are able to overcome this problem of underinvestment in prevention, it may well be worthwhile to pay the incremental cost.
I have a few recommendations.
The first is to be realistic about what can be accomplished by SIBs, or social impact bonds. Don't oversell them. Recognize it's a new approach, that it's still in the experimental stages. None of them in the world have reached the payment stage yet, so there's still a lot to learn. For that reason, I'd say they should not displace existing spending, and that you should recognize that they're not a panacea to cure the problem of underinvestment in social programs.
I'd also specifically urge you to start with the analysis of the desired outcomes, and how much you are willing to pay to achieve them, whether because of savings that will be achieved or because of the value to society. This framework is core to a successful SIB. You can't move forward without it. But it's also a really important conversation to have that adds value to your budgeting and to your decision-making process, whether or not you decide to move ahead with social impact bonds.
Thank you.