Social Impact Bonds

DC Public Safety Radio

See the main site at

See the radio show at

Leonard Sipes: From the Nation’s capitol this is DC Public Safety, I’m your host Leonard Sipes. Ladies and Gentlemen, Social Impact Bonds pay for success or making government more evidence based, more creative and more receptive to new ideas. Back at our microphones, John Roman, Senior Fellow Justice Policy Setter, Urban Institute,, John welcome back to DC Public Safety.

John Roman: Thank you very much.

Leonard Sipes: All right you’re my favorite guest because all I have to do is wind you up and let you go, so that’s what we’re going to do today. Social Impact Bonds, what the heck are Social Impact Bonds?

John Roman: This is a new idea and it’s probably the biggest idea that’s out there today in terms of reforming criminal and juvenile justice systems. Along with public health, workforce development, economic redevelopment, pretty much anything you can think of. What we’re trying to do here is we’re trying to get private companies to invest in traditionally public sector activities. Which will allow governments to do all kinds of things that they’ve wanted to do but never had the resources to pursue.

The basic idea is Goldman-Saks in the deal that we’re going to talk about in a minute. Invested in a program to help prisoners at Rikers Island, which is the New York City jail. Which is a very distressed little corner of the world.

Leonard Sipes: It’s crazy.[crosstalk 00:01:29]

John Roman: A place that needs new resources to invest in a new program that hadn’t been implemented there before with the idea that the government of New York city would only pay back Goldman-Saks if the investment met performance targets that all of the parties agreed to, before the transaction was implemented.

Leonard Sipes: What was the program? What were they trying to do?

John Roman: What they were to do was basically a program called, “Moral Reconnation Therapy” through a program called, “Able”. The idea was to work with 16 to 18 year olds. New York is 1 of only 2 states left in the country that 16 and 17 year olds automatically enter the adult criminal justice system and so you have certain responsibilities to treat these young people in different ways then you would adults. This program was an attempt to try and deliver more services to them. To the tune of over 9 million dollars worth of new services to hundreds of not thousands of young people in Rikers Island in ways that they otherwise wouldn’t have an opportunity to receive.

Leonard Sipes: Okay for the non-criminal justice people out there it’s a jail, so it’s not a prison so they are either awaiting trial or they’re serving shot sentences which category do these people fall in?

John Roman: The goal was, and this was really why this was very complicated, so the goal here was to try and serve people who were there on a sentence or were serving a long period of pre-trial detention. That just means they’re somebody who had done something that meant that the court wasn’t going to let them out on the street until their case was adjudicated in the courtroom or people who got a short sentence, less then one year 3, 6, 9, 12 months. Who were there for long enough that they had … There was the potential that you could actually deliver some services with some real dosage.

Leonard Sipes: Okay, that’s the point they were there long enough to deliver the services and were the services delivered as designed?

John Roman: Yeah, they really weren’t.

Leonard Sipes: Yes or no?

John Roman: No, they were not. The idea here is to say … We should take one gigantic step back if you like and say this is an idea that has been implemented around the world. There are Social Impact Bonds in the United Kingdom, there are 20 of them. There’s half a dozen in Australia. There are Social Impact Bonds, I was reading a piece today about Social Impact Bonds being developed in Brazil and Mexico. I know Israel is looking at them, Netherlands … They’re all around the world.

Leonard Sipes: Okay, so this is much bigger than any of us realize?

John Roman: That’s exactly right. This is the new idea, and the new idea is that we have all of these programs that have some evidence base right? People don’t really understand that we know so much more about how to help really disadvantaged populations then we did 20 years ago. I started at Urban in 1197 and what we know about what’s effective in serving people is completely exponentially grown since I started. What hasn’t grown is the resources to fund these programs to test whether they work and then implement them at scale.

Leonard Sipes: Why? When I talked to people about this program, getting ready do do the program they were saying, “Leonard, really intriguing idea, but if the ideas were so good why isn’t government funding?” Why isn’t government funding these programs if they are so evidence based, if they’re so impactful?

John Roman: I have two responses to that and one is really simple. They don’t. We’ve talked on this program before and you’ve probably talked a lot about drug courts.

Leonard Sipes: Yes.

John Roman: Drug courts are an idea where you take people who are drug involved and it’s their drug usage that’s causing them to offend and if you address the underlying substance abuse issue you could get them to stop offending and lead more productive lives and that would save tax payers money.

Leonard Sipes: Drug courts are generally seen as effective.

John Roman: Drug courts are seen as effective, there’s an enormous amount of research around drug courts that they have an effect.

Leonard Sipes: Right.

John Roman: We’ve been doing these things for 25 years.

Leonard Sipes: Right.

John Roman: They’re in every major county, there are 3,000 counties in America, they’re are probably in 2,000.

Leonard Sipes: Right, and that’s just the point. I mean they’ve shown to be effective and government has implemented them in hundreds of jurisdictions throughout the United States and Canada and throughout the world.

John Roman: That’s right where you hit the problem. The problem is we did a study about 5 years ago where we went and looked at of the 1.5 million people who enter the criminal justice system every year, which is an astonishingly large number.

Leonard Sipes: Yes.

John Roman: We estimated that of the 1.5 million offenders who enter the American criminal justice system, who are at risk of substance abuse disorders, of that 1.5 million, maybe 3% got a drug court. Then some really smart researchers at the University of Maryland came on behind us, redid our studies and different data and they estimated that we were wrong that in fact it’s less than 1%. Here we have an idea that is universally accepted, this is a way to stop future offending, this is a way to help disadvantaged people. This is a way to make America a better place and we don’t do it in any large numbers.

Leonard Sipes: I do want to get on the Social Impact Bonds because it’s intriguing and you’re saying it’s a national effort or an international effort. We should really be paying attention to it. Again the people I’ve been talking to are saying, “Leonard if it’s so dag-gone effective, why aren’t we getting the funding for it? Why do we have to go to Goldman-Saks, why do we have to go with hat in hand?” I mean, why don’t we just start kick-starter programs for drug treatment and for mental health? Why doesn’t the government, if government is saying they want fewer people in the criminal justice system, if government is saying we cannot stand the strain of the correctional budget as it currently is. We want to have fewer people recidivate, fewer people enter the criminal justice system, why the heck doesn’t government pay for it?

John Roman: Well government doesn’t pay for it.

Leonard Sipes: Why?

John Roman: There’s a bunch of reasons for it. One of the reasons for it is the results of these interventions occur way down the road. They’re not today, they’re distant. The administration that funds it today won’t be around when those benefits accrue so they’d rather fund things where the benefits will occur while they’re in office. That makes total sense. The populations we’re talking about are very difficult to serve. A drug involved population is … Drug courts are a great example, drug abuse treatment is a great example.

It’s not effective for everybody every time right? Relapse is part of recovery and so you have to know that you’re going to make the overall average population better but you’re not going to help every person. What that means is you’re going to be putting people out on the street instead of in prison, which is what you normally do with these folks and some of them are going to commit new offensives and that’s politically untenable for some populations. Then at the other end of the spectrum a lot of the people that we’re talking about helping with these kinds of programs just don’t have a political constituency they aren’t very sympathetic.

Leonard Sipes: All right, I don’t want to take away from the program on Social Impact Bonds but I promised others when I was doing the research for the program that I would ask that question. Social Impact Bonds very big, they are happening throughout the world. Why Social Impact Bonds? You may have just answered the question. Government unto itself really does not want to take these programs on.

John Roman: There are 3 things going on here that are really important. If you think about the research, I’m a researcher so I come at it from an evidence base and I come at it from the perspective of trying to get government to invest more in outcomes rather then good intentions. What I’d like to see for social service providers to get good evidence around their good works instead of just good intentions. Which is where we are today. There’s a lot of stuff that goes on out there that sounds good on paper that does not have any research support around it. Part of what these bonds do is because you have to go to an investor who’s going to write a check and in some cases these are really big checks right?

The Department of Labor, The US Department of Labor has funded a couple of these transactions to the tune of 15 million dollars so we’re talking about real large investments. If you’re going to get somebody to write a 15 million dollar check for you, you’d better have some sound evidence to demonstrate to them what their expectations should be for how effective this thing is and whether you’re going to be able to achieve the performance goals that you’ve outlined in the agreement.

Leonard Sipes: I mean people watching the Shark Tank program, I would imagine you have that sort of atmosphere. They’re saying, “Hey, prove to me that there is a research base behind this program. We’re not going to write you a check for anything until you come along and provide us with enough evidence that leads us to believe that we’re going to get a return on our investment.” There has to be an evidence base or is the private sector better at understanding evidence based procedures the the government?

John Roman: There’s two questions in there, one is am I willing to put my scholarship aside and admit that I watch Shark Tank? I do.

Leonard Sipes: It’s my wife’s favorite program.

John Roman: It’s a wonderful show and it really is exactly the sort of thing we’re talking about here, except that the scale is way smaller then what we’re talking about here. There they’re talking about 6 figures, hundreds of thousands, here we’re talking tens of millions.

Leonard Sipes: Right.

John Roman: Maybe more, down the road. You have to be able to go to the same kind of investors, high-worth investors and ultra high-worth investors, right? People with 500 million dollars that they want to invest and you have to convince them that you’re going to reinvent Cleveland. You are going to do urban redevelopment in Akron. You are going to solve asthma in Fresno. You’re going to cure blight in Baltimore and if you can make a case that is evidence based to these folks they are going to invest in a way where they don’t actually get a market rate of return.

Leonard Sipes: Because if the program doesn’t succeed and are endless issues in terms of whether or not a program succeeds, implementation is one of the hardest things on the face of the Earth. If it doesn’t succeed it may not have anything to do with the evidence it may just be how it’s implemented that may be how faithful were to the original program design. They don’t get anything at all, they only get the return on their dollars. The program is paid for by government if it works. In this case the question becomes recidivism, in this case it didn’t reduce recidivism there by Goldman walks away from the table with nothing in it’s pocket.

John Roman: I want to come back to the Rikers Island deal, because it’s really important to the development of this concept to understand what happened in New York City.

Leonard Sipes: Okay.

John Roman: What I want to say about your point of implementation fidelity. Fidelity to best practice is really critical and that’s actually part of what’s going on here and it’s part of what’s so exciting about this whole prospect. Department of Labor funded the state of Massachusetts to do a program for young adult offenders in Massachusetts. Called “Roka” what they’re are trying to do there is to understand if they reduce the number of jail bed days that you would have expected from new offending compared to people who don’t get this program.

Leonard Sipes: Right.

John Roman: The other thing that they’re doing that’s really important is … The Urban Institute is doing that validation. Another group Apt Associates is doing an implementation, the question they’re asking is, “Have you implemented this program as close to best practice as possible?”

Leonard Sipes: A process evaluation.

John Roman: A process evaluation yes, but more than that an implementation science is what we call it, right? What we’re learning is, we’re learning whether government cam come together and get the data together to do this big system reform. We’re learning whether this kind of program helps young people involved in the criminal system, and we’re learning how to put these things in place in a way that will inform the next time we try and replicate this in another place in another time.

Leonard Sipes: Any answers to any of those questions?

John Roman: We don’t have them yet, but what we think is really important in all of this process is the process itself is a reform. What we want to do is we want to go to a juvenile and adult criminal justice systems and we want to say to them, “Who are the drivers of your cost and populations? Who are the people you serve over and over again?”

Leonard Sipes: Right.

John Roman: “Why have you failed with this population?”

Leonard Sipes: Right.

John Roman: Getting people to admit that they’ve failed is really, really hard. Why is it that the same people keep coming in over and over, because that is evidence of failure. “Is there evidence out there about a program that you could be implementing that would help these people be more successful with their lives and save your tax payers resources and is a Social Impact Bond the way to finance that?”

Leonard Sipes: Now is a Social Impact Bond part of that Massachusetts initiative, I heard Department of Labor, who else?

John Roman: It is funded, it’s third sector is the intermediary there is private financing that’s associated with it. All of the dollars that are paying for this up front come from somebody’s pocket other than the Massachusetts state government. The federal government is paying for some of the implementation pieces and the evaluation pieces and the data integration and [crosstalk 00:14:52]

Leonard Sipes: So the private sector is putting up money?

John Roman: The private sector is putting up another 15 million dollars so it’s about a 30 million dollar transaction.

Leonard Sipes: That’s a huge investment.

John Roman: It’s a huge investment and so what we’re doing here is we’re thinking about investing in chronically disadvantaged populations on a completely different scale then we’ve thought about it at all. If you think about the Second Chance Act that was a hundred million dollars across all 50 states. That’s 2 million dollars a state. Here we’re talking about 30 million for one state.

Leonard Sipes: I want to talk about how expensive these programs are and I’m not quite sure the average person realizes when in terms of what we do at the court services and the defendant supervision agency for high risk people is to put them in a 28 day residential program then put them into a 90 day residential treatment program them put them in an after care program. All that carries an enormous expense, but ladies and gentlemen, back at our microphones John Roman, Senior Fellow Justice Policy Setter, Urban Institute,, Talking about Social Impact Bonds.

John, this is exciting, so what you’re saying is we can sell hard-nosed, demanding individuals who only invest their money if they see the potential for return on their dollars and there’s enough evidence out there or they’re willing to do social change or intervention programs on a very large scale.

John Roman: Right, so let me give you one example and I can give you others if you want to hear them. There is a new drug therapy that actually cures Hepatitis C. That’s important for our discussion here because something like two thirds of people with Hepatitis C in America contact the criminal justice system at some point right? You get Hepatitis C because of prostitution or intravenous drug use or something like that.

Leonard Sipes: Right.

John Roman: We actually have this medical therapy that is a 3 month course of pills basically, that you take that at the end of the 3 months you no longer have Hepatitis C. The only other cure for Hepatitis C is a liver transplant which is like a million bucks.

Leonard Sipes: Right.

John Roman: The course of treatment is about $85,000 over 3 months right? We’re talking something like-

Leonard Sipes: For 1 person?

John Roman: For 1 person.

Leonard Sipes: Oh my God.

John Roman: We’re talking about 7% of the criminal justice population has Hepatitis C.

Leonard Sipes: Right.

John Roman: We’re talking hundreds of thousands of people.

Leonard Sipes: Right.

John Roman: Almost a hundred thousand dollars per person.

Leonard Sipes: Got it.

John Roman: When you talk about needing resources on that scale there is no solution, right?

Leonard Sipes: Right.

John Roman: This is the solution, you could come along and you could go to Goldman, or JP Morgan-Chase, or Bank of America, or Deutsche Bank, or whoever, these folks have all expressed interest in this concept and say to them, “Look, we will as the government have this enormous financial benefit for having treated these people for this condition, because we won’t have to treat them. We won’t have prisoners that need liver transplants that we have to pay for so we’ll bay you back for a profit.”

Leonard Sipes: Is this actually happening?

John Roman: This one is on the books but we’re talking about these things around, another big one is asthma. Very similar idea.

Leonard Sipes: All right the medical part of it I understand, the criminal justice system often times to me strikes me as being a hard sell, which is one of the reasons why the data that I’ve seen in the past is that fewer than 10% of all people in prison have access to substance abuse treatment.

John Roman: Right.

Leonard Sipes: Even the substance abuse treatment that’s there is not very good and the numbers are tiny, so if 80% have histories of substance abuse, 10% are getting it. You say to yourself, “Is it that government doesn’t care?” What you’re talking about is revolutionary. What you’re talking about is not Social Impact Bonds, what you’re talking about is a process, a sea change, a different way of conducting criminal justice, a different way of conduction public health, it’s not government that’s driving the boat anymore it’s industry. It’s investments, it’s people who are very demanding, very specific, very evidence based. There are the people that possibly could bring upon significant change within public health and the criminal justice system crawls to the board. What you’re talking about is monumental.

John Roman: Yeah, I mean I think we’re talking-

Leonard Sipes: Am I exaggerating or not?

John Roman: NO I think it’s a totally different way of thinking about the world and we’ve talked about it in Ottawa, to the Canadian government, in Israel, in London, all over the world and people see the opportunity here. The idea here is really simple, we want to turn the problem upside down. The problem has been, how do we get resources to help people with a particular problem. What we want to say instead is, “Hold on, we have evidence on how to solve some particular problems, how so we get resources to those things instead of the things that we are funding as business as usual now?” In some respects what we want to say is really simple. Some things that we do don’t really work, right? I mean we’ve talked about, you and I have talked about-

Leonard Sipes: What don’t work very well.

John Roman: What don’t work like DARE, right, every schools got a Dare, a police officer comes in.

Leonard Sipes: People love it.

John Roman: People love it, or scared strait, there’s absolutely no evidence that changes anybody’s behavior and in someways it might cause worse behavior. When the DARE guy cones in and shows your kid his suitcase full of drugs, some of these kids are going to be like, “Oh, so that’s what cocaine looks like. Cool, that’s not what I thought.” That’s not good for them.

Leonard Sipes: The weird thing about government is because we do all sorts of things that make people feel good but really don’t have an evidence base behind it.

John Roman: Right, what we want to do here is say “look, I could give you a laundry list of things that I have in evidence, that I believe have an evidence base that is overwhelming, that we should do at scale. That everybody that needs this help should get it right?

Leonard Sipes: Yeah, of course.

John Roman: We’re talking about in our world we’re talking about high-risk adolescents, we’re talking about family based therapies, we’re talking about multi-systemic therapy. We’re talking about programs with evidence bases that are so compelling that the most sophisticated researchers in the world, the most skeptical researchers in the world are saying, “$15 in benefit for every dollar in cost, only they’re really expensive how do we raise the money to get those benefits?” This is the answer.

Leonard Sipes: It’s not just the answer, it’s not just the matter of money, from dealing with business people in the past, and dealing with government throughout my entire career, the business people have an entirely different mind set. Very polite, very nice, but very demanding.

John Roman: Yes.

Leonard Sipes: Very precise, they are … especially I worked with marketers, in the past, and marketing in the private sector is a different world. It’s very precise it’s like driving a jet plane. It’ like driving a 747 it’s a very precise set of rules and understanding and operations. If you don’t follow these precise, rules, procedures, and understandings they ground you. It’s do not pass go, do not collect $200. Government is far mushier then the private sector so if you’re telling me that the private sector is convinced of evidence based procedures and they’re going to start coming in and they’re going to start funding this it’s a seed change in terms of what we’re talking about not just for the criminal justice system, but for the delivery of services to undeserved populations across the board.

John Roman: I think what’s really been interesting in this is that when we approach this, and we started thinking about this in 2009. The Rikers deal happened shortly after that, the first deal in the United Kingdom happened at about that time, and every year it grows more and more quickly. When we first thought about this we knew there were 3 players in this equation, there was the government, there was private philanthropy, people who give grants, give away their money, and commercial investment bankers. We assumed the government would love this idea because what it says to the government, “Here’s some money to do something you can’t do, and you only have to pay us back if it works.”

Leonard Sipes: The government loses control.

John Roman: You’re getting ahead of me but that’s where I’m headed. Philanthropy, we thought this would be very interesting to them because they, this allows them to leverage their giving and give the same dollar over and over again. We thought investment bankers would be the hardest sell. We thought these people would be very hard headed and they wouldn’t be interested in the rates of return that they were getting. The actual operational practical realities, the absolute reverse. The investment bankers are extremely interested in this idea, in fact there are hundreds of million of dollars if not a billion dollars or more on the sidelines looking for deals to invest in. They cannot find the deals.

The philanthropists have been a little scratching their heads at this whole concept and haven’t weighed it in as deeply as we had expected. The governments, probably for the reason that you just raised, which is when they get engaged in these deals they lose control. Have really had people come to them and say, “I will give you a hundred million dollars and I can solve this problem for your entire population in your state. Would you like me to do it? You only have to pay me back if it works.” and they said, “No thank you.”

Leonard Sipes: You’re not … I’m not the first person to suggest to you as a lifelong researcher that government doesn’t trust [inaudible 00:24:17] Government likes to do what government wants to do. We do that from a very common sense perspective, because we believe that the delivery of this particular service is going to work, We’re not interested in somebody coming along and measuring it and telling us whether it does or it does not. We believe that it works thereby, do not pass go, do not collect $200. We’re comfortable with that and that’s how we proceed. You’re talking about a different world, you’re talking about a different mindset, you’re talking about a “prove it to me and prove it to me consistently and we’re going to hold you to these markers. Have you met these markers, at the eighth point, at the quarter point, and if you didn’t why not?” It puts government under a spotlight and we’re not happy about that.

John Roman: What I think what’s really important what’s happened and we started out by talking about the Rikers deal and the Rikers deal ended up not going all the way to the end because it wasn’t meeting the performance measures that had been agreed upon at the outset by Goldman-Saks and Bloomberg Foundation who were the one’s who underwrote it. Who basically insured the deal. What Mayor Mike Bloomberg did and Deputy Mayor Linda Gibbs did in New York City was really brave and really critical. Because what they said was, “No, were going to take the biggest, I think it’s the biggest city government in the United States, and we’re going to get behind this concept and we’re going to test it. We know there’s a chance it’s going to work, but that will provide a precedent for other cities and other counties and other states to think about investing in this.” and that propelled this field forward.

Now you have Salt Lake County Mayor Ben McAdams is really pushing this and Corporation for National Community Service is 11 million dollars behind this and Department of Labor and Department of Justice. It’s probably a hundred million dollar industry and none of that would have happened without some government having the courage to step forward knowing it might fail, and that happened and that sparked everything that has gone through.

Leonard Sipes: We’re not talking about Social Impact Bonds, I said this before and I’m probably being repetitive to the point of fault, but we’re not talking about Social Impact Bonds, we’re not talking about delivery of services. We’re talking about fundamentally changing the way that government operates through a public/private partnership.

John Roman: I think the metaphor that I always tell and I think it works, is we’re sitting in the heart of Washington DC, there’s a big commercial corridor right around the corner, 7th and 8th street, there’s a big stadium there. There’s tons of bars and restaurants and at some point tonight I can just about guarantee you that a metropolitan police department officer is going to engage with somebody who is in crisis. They might be drunk, they might be having a mental health problem, they might have a substance abuse problem. They might just be unhappy with their station in life.

Leonard Sipes: Yes.

John Roman: That police officer is going to have to serve as a social worker to try and guide that person to engage in better behavior so they don’t end up in the DC jail that night.

Leonard Sipes: Sure, of course.

John Roman: They shouldn’t be in that position, we should have the community infrastructure to help serve that person. If it’s mental health care, if it’s drug abuse treatment, if it’s alcohol, whatever it is that that person needs, we don’t have the infrastructure to do it. What this does to create a way to fund that infrastructure in ways that just simply aren’t going to happen if we don’t go down this road. The idea isn’t to privatize policing, the idea is to let that officer go back to being a police officer.

Leonard Sipes: Right.

John Roman: Which is what they are trained to do and they know how to do and they’re good at. We’ve sort of forced all these people throughout our government to take on jobs that they aren’t trained for and we’ve over the last couple of years we’ve seen the results of that. A lot of it’s been not very pretty.

Leonard Sipes: Final minute of the program. What percentage who aren’t getting substance abuse treatment to day are going to get substance abuse treatment? What percentage of people who desperately need mental health treatment are going to be getting mental health treatment? What percentage can we say the people who are undeserved, who are not getting services now will get services say in 10 years through a private/public partnership.

John Roman: If you think about the family based therapies that I talked about, which are ways to treat high-risk young people in their homes with mom, with grandma, with their cousins, with their siblings, with their friends.

Leonard Sipes: Which is the most effective program I’ve seen out there.

John Roman: It’s the most effective program that’s out there and it probably serves less then 10% of all the people it could.

Leonard Sipes: Yes.

John Roman: We could get to 70% by using this a s a way to raise funding, to package together all of these things. The idea is called “Scale finance” Steve Goldberg, Caffeinated Capital, is pushing this idea. It’s a wonderful thing. We could get a long ways to getting to 100%.

Leonard Sipes: What are the odds of really doing that?

John Roman: I think we can do 1 state in the next 2 years.

Leonard Sipes: That’s amazing, that’s amazing. The most effective program that I’m aware of, that does provide fundamental change and has a long term impact on criminality and use of the criminal justice system. We could fund 75% in terms of one state.

John Roman: I think within the next few years, some state will sign on to this and we’ll be able to take this to every young person that needs it.

Leonard Sipes: We proposed to have a program today, a program on Social Impact Bonds. It’s turned out to be much more than that and I hope everybody stuck with us through the program to the final conclusion. John Roman, Senior Fellow Justice Policy Setter, Urban Institute,, He has 3 articles talking about the situation in Rikers Island and I’ll put all 3 in the show notes. Ladies and gentleman this is DC Public Safety, we appreciate your comments, we even appreciate your criticisms and we want everybody to have yourselves a very pleasant day.

%d bloggers like this: