Private investment in public service is a largely untapped profit channel that is just starting to be explored. Many attempts, often called “pay-for success” or “social impact bonds,” are designed to attract private dollars to meet public ends.
The concept is fascinating in that it could potentially solve a lot of social problems and provide return on investment to willing funders. However, many of the deals currently lack real quantitative outcomes—something that could stunt this industry’s growth for years to come.
A brief case study
The United Way of Salt Lake and its partners including StriveTogether are doing good work funding early childhood education through a pay-for-success program. They have reasonable numbers to suggest significant cost savings to the State of Utah while demonstrating a strong business case to scale the program further.
Of the 110 students attending this pre-school program who were “likely to need special education later on,” only one ended up requiring it in kindergarten. This translated into a cost savings of $2,607 per child—the marginal cost of providing special education in Utah.
These results suggest immediate outcomes of $281,550 in total savings that very next year. Because of the way the deal was structured, investors were to receive 95% of these savings after the one year, amounting to $267,473, leaving the state with a presumable profit of $14,077. Not bad for an immediate return.
However, after taking a quick look into Utah state budget documents, the actual results are quite different:
- Increased total expenditure from $22,432,300 in 2013 to $24,376,400 in 2015. This equals a compounded annual growth rate of 4.2%, which is higher than the total student enrollment growth rate of 2.4%.
- Increased per-unit expenditure from $2,607 per student in 2013 to $2,837 (compounded annual growth rate of 4.3%).
There are possibly very valid reasons for these increases. First, the need for special education could be growing generally and this trend could have at least been slowed by this program, thus making it a success in real terms. Also, the cost of providing this service could be going up generally, another trend that might have been slowed by the program – again, making it a success.
By these accounts, this program really is doing great. And that’s the point. The United Way has put together a legitimate program to help improve outcomes for early-age students. Almost every single program student in danger of needing special education ended up not needing it by kindergarten. That is a success.
Note: For simplicity in this analysis, we are avoiding things like selection bias and other aspects that might skew a program toward success to the detriment of potentially more needy program recipients.
However, because of this success, the State of Utah paid $267 thousand to investors after one year, despite seeing zero actual nominal savings on special education over that same period.
How does that inspire future governments to make these types of deals? How is this anything other than a public service, funded like everything else—with budgeted tax dollars?
The program works, but the deal doesn’t
A good program is one that generates positive outcomes—like the Utah program. A good deal, however, is one that generates positive financial rewards as well—something that is lacking so far.
The major problem with the way the Utah deal works is that it assumes the social outcomes are tied to lower spending on special education. The theory makes some sense, but practically speaking, this “outcome” could be achieved by slashing special education budgets, reducing amount of enrollees, or any number of additional manipulative tactics. And it is easy to see that these are not the same as actually improving the lives of those needing special education.
So how could we improve on a deal like this?
1. Identify the real outcome
Reduction in special education spending is not a social outcome, but reduction in special education need is. But how do we quantify this?
For one, it is probably safe to say that the more advanced our children are, the more ready for the world they become and the more successful they become as participants in the economy (getting good jobs, paying taxes, buying things, etc.).
If this is truly the sought-after outcome—creating productive citizens—then we have a much longer time horizon to wait than one year (more like 15-30 years). Instead of chasing incremental drops in special education spending, the state should focus on creating the best possible version of each student beyond just that first year after pre-school.
Who’s to say that the same reasons causing kids to “likely need” special education in the first place don’t resurface once they leave this rehabilitating pre-school program?
2. Determine the value
A Utah student in 2013 needing special education had a marginal additional cost of about $2,600, according to the Utah State Board of Education. However, these cost reductions were not achieved. There are probably legitimate reasons for this, including a proportion of fixed costs that we cannot immediately slash after one year (e.g., reducing the amount of students by 10% does not allow us to lay off 10% of a teacher).
If we take a step back and look at the real long-term outcome of such programs, we can identify educational attainment and career earnings as more valuable metrics. It is fair to say that educational attainment level contributes heavily to an individual’s earning power.
In 2014 for example, those with a high school diploma earned a median weekly income of $668 and unemployment rate of 6.0%. Those with a bachelor’s degree earned $1,101 with an unemployment rate of 3.5% (according to the U.S. Bureau of Labor Statistics).
Assuming all residents stay within the same community, the cost of unemployment insurance can be calculated (identifying cost savings) and the incremental gains in total tax dollars can be calculated (identifying increased revenues).
These are two major benefits and exclude any other secondary benefits such as increased buying power in the local economy, likelihood to raise children of higher caliber, and improved property values due to more desirable school districts (all of which can contribute greatly to social impact).
3. Match payments to benefits
Utah paid investors 95% of the “cost savings” after one year. These cost savings were not actual dollars, but underwritten savings at the time of the deal being signed.
Instead of doing this, Utah should have tracked special education need to things like graduation rates and educational attainment. From there, baseline costs of special-needs children into adulthood can be approximated.
If these children were subsequently tracked all the way through school and beyond, real costs and benefits can be tracked on a unit level and the benefits can be repaid based on when they are achieved.
For example, a child we will call “Timmy” has been identified as likely to need special education sometime during his school years so he participates in the pre-school program. The program works to significantly reduce the likelihood of him needing special education, so he progresses through school at a faster pace than previously expected.
He graduates high school and attends some college but leaves to work a full-time job earning $741 a week. This is compared to a special-needs baseline of someone not obtaining a high-school diploma and earning $488 a week. Assume both incomes fall in a state income tax bracket of 5%. This generates an additional $13 in tax income per week, or $658 per year. Plus, Timmy is 33% less likely to become unemployed and drain tax resources from that program.
Multiply the $658 over 10 years of working and the state earns an additional $6,578 in revenue on top of any cost savings from reduced need for special education. But because residents’ salaries are not budgeted the same way program funds are appropriated, it is easier to track expense progression and transfer some of these earnings back to the investors as they come in.
Here is a table outlining the net cash flows for the program investment ($ in thousands):
A couple disclaimers on these charts:
First – Let’s assume the program runs for the first five years, with equal $1m installments each year. Let’s also assume the $72k tax income increase starts when a class of students turns 18. So for a 4-year old in a pre-k program, this implies by year 14, he will be 18 and starting to earn more money than he would expect without the additional education.
Second – Inflation of 2% is assumed for the cost of program. The $1m invested right away would now cost $1.08m by the fifth year. Wage growth of 4.5% is assumed for earnings and tax increase, implying a jump from $72k to $133k by year 14 (when the first class turns 18).
As you can see, the major benefits of this program are not likely to be achieved until many years after the program is initiated, making immediate return to investors a very difficult challenge to achieve.
In fact, using these assumptions and only considering increased tax revenue, the program doesn’t break even until the 26th year. When factoring in ONLY the impact of increase tax receipts from higher earnings, the five-year program earns 2.6% annual ROI over a 50-year period.
To get a full picture of the program’s social impact, other outcomes should be quantified in this manner and payouts tied to specific years they are expected to take place. Without doing so, the program loses its pay-for-success nature.
Conclusion
To successfully construct a pay-for-success program, multiple projects should be pooled. Their results, on a combined basis, have the power to offer payback to investors at various times during the contract.
As illustrated above, a program like pre-school obviously wouldn’t warrant payment until closer to the end of such a contract. However, a program like prisoner reintegration might be more able to pay back investors on the front end of the deal.
The key is to diversify the types of programs and outcomes being achieved to stagger payback in a more realistic and reasonable manner to investors. Because after all, it’s their appetite for risk and reward that has the potential to drive capital toward the growing sector.