Financial results at Goldman Sachs are going to look a little bit better this quarter because of the educational success of 100 or so kindergarten pupils in Utah.
The students were part of a relatively new financial experiment in which Goldman put up money to pay preschool costs for students who had been expected to need special education services.
When the students were tested this year — after a year in preschool — and found not to need extra help, the State of Utah paid Goldman most of the money it would have spent on special education for the children.
The payment represented the first time a so-called social impact bond paid off for investors in the United States.
The idea of social impact bonds is still very new. The first one was started in England in 2010; Goldman started the first in the United States in 2012.
The bonds are already being talked about as one of the most promising ideas to come out of finance recently — providing a new way to fund social programs in an era of government budget cuts.
But the effectiveness of the bonds has been unclear. Only one of the British efforts has returned results so far. (It was a success.) Goldman’s first social impact bond to publish outcomes — a program that attempted to reduce recidivism among inmates on Rikers Island in New York — failed this summer.
For people studying social impact investing, the results in Utah are exciting — even more so given the children’s success. Among the 110 students who had been expected to need special education had they not attended preschool, only one actually required it this year.
“We’re in the beta testing stage of this whole endeavor, and now we have one example where the concept worked as intended in terms of delivering the outcomes and executing the contract,” said John Roman, a senior fellow at the Urban Institute who is researching social impact bonds.
Gov. Gary Herbert of Utah and Goldman’s chief executive, Lloyd C. Blankfein, lauded the outcome on Wednesday as a victory for public-private partnerships. It is also a public relations victory for Goldman, which has been trying to reform its reputation as an institution focused solely on the bottom line.
But these are still early days for social impact bonds. The experiment in Utah has only operated for one year and involved 110 schoolchildren. It and other social investments Goldman has made could end up more like the discontinued Rikers program. Even if that happened, the sums involved — in the millions of dollars — are tiny for the megabank.
But Goldman has structured its more recent social impact bonds more like the one in Utah, which expanded on existing programs that had a successful track record. The Rikers program, by contrast, paid for a service that had never been tried before in the prison.
“The tool of ‘pay for success’ is much better suited to expanding an existing program,” Andrea Phillips, vice president of Goldman’s urban investment group, said in an interview on Wednesday. “That is something we’ve already learned through this.”
The $4.6 million put up by Goldman — and the $2.4 million invested by the Pritzker Family Foundation — went toward expanding an existing preschool program for poor children in Salt Lake County.
The program had already been shown to decrease the need for special education, but it had not been able to expand to meet all of the demand.
The Goldman money — which is more of a loan than a bond — allowed another 595 3- and 4-year-olds to attend preschool last year, in addition to the 2,400 or so other children who were already enrolled. Of those new students, tests indicated that without preschool, 110 were likely to need special education.
This year, teachers determined that 109 of those pupils did not need special services. For each one of them, Goldman and its partner got about $2,500, and will receive that each year, through the six grade, that the students avoid special education, with the amount decreasing in the years after that.
The State of Utah, and some local private charities, are still paying 95 percent of what it would have cost for the special education. But Mr. Roman, at the Urban Institute, said the state would reap significant other savings if the students continued to avoid special education, which is generally associated with higher levels of truancy, juvenile crime and other problems.
The state and county were also pushed to quantify the results of their services, something governments are frequently able to avoid.
Mr. Roman said that when he started looking at social impact bonds, he assumed they would need to get most of their money from philanthropists, given the risk-return profile of the investment. While social impact bonds pay out a relatively healthy interest rate when they succeed — between 5 and 7 percent a year in the Utah program — the investors lose all the money when they fail.
The investments, though, have drawn a remarkable amount of interest from institutions and wealthy individuals eager to put their money into something they can feel good about. After the Utah program began, Goldman raised $150 million from clients for a social impact fund, and a number of other social impact bonds are moving ahead with backing from other financial institutions.
There is still a lively debate about whether these sorts of investments will come with unintended consequences, perhaps tilting funding toward programs that can provide easily quantifiable results and making it easier for governments to cut budgets.
But for now, there is a bottleneck in expanding this sort of investing. It is not with investors; instead, it is the time-consuming difficulty of structuring deals like the one in Utah.
Mr. Roman is looking at ways to make the process easier — and he imagines it will happen as the idea catches on, especially if there are more results like those in Utah.
“The goal of this whole thing was to keep kids out of special ed, and on that metric it seems to have done so spectacularly,” he said.