How Goldman Sachs Made Money Investing in Preschool in Utah (National Journal)

By Emily DeRuy 

The res­ults are in on a Gold­man Sachs-backed early-child­hood-edu­ca­tion in­vest­ment, and the firm is set to make money.

Gold­man is the first ma­jor Wall Street in­vest­ment bank to em­ploy a so­cial-im­pact bond in the early learn­ing realm—and the first to earn money at it. Es­sen­tially, Gold­man (in part­ner­ship with ven­ture cap­it­al­ist J.B. Pritzker’s found­a­tion) agreed to loan sev­er­al mil­lion dol­lars to the United Way of Salt Lake, the pro­ject in­ter­me­di­ary, to send hun­dreds of low-in­come chil­dren in Utah to preschool through a pay-for-suc­cess agree­ment in 2013.

At the out­set, every­one in­volved agreed on a meas­ure of suc­cess—in this case, spe­cial-edu­ca­tion status. If the preschool pro­gram worked, and few­er chil­dren than ex­pec­ted re­quired spe­cial edu­ca­tion in ele­ment­ary school, Utah would pay Gold­man and the oth­er in­vestors back, plus in­terest. If it failed, Utah was off the hook, and Gold­man would be out a few mil­lion dol­lars, a drop in the pro­ver­bi­al buck­et for the bank.

The strategy is re­l­at­ively new (the first so­cial-im­pact bond was launched in Eng­land in 2010), but in­creas­ingly ap­peal­ing both to in­vestors and to cash-strapped cit­ies and states.

In the Utah pro­gram’s first co­hort, 595 3- and 4-year-olds at­ten­ded preschool. Ini­tial test­ing sug­ges­ted that 110 of them would likely need spe­cial edu­ca­tion in grade school. Of those stu­dents, just one needed spe­cial-edu­ca­tion ser­vices. In oth­er words, the find­ingssug­gest a 99-per­cent suc­cess rate, well bey­ond the 80-per­cent bench­mark. The 110 stu­dents will be mon­itored through sixth grade, and Utah will con­tin­ue to make pay­ments based on how many of those chil­dren avoid spe­cial edu­ca­tion each year. The pay­ment marks the first time that in­vestors have profited from a so­cial-im­pact bond in the United States. The part­ner­ship with Utah is ex­pec­ted to ul­ti­mately serve more than 3,000 chil­dren in five co­horts.

“This pi­on­eer­ing ef­fort is a per­fect ex­ample of data-driv­en col­lect­ive im­pact with the lar­ger goal of as­sur­ing every child is ready for school,” Bill Crim, pres­id­ent and CEO of United Way of Salt Lake, said in a state­ment. “When busi­ness, gov­ern­ment and the so­cial sec­tor work to­geth­er in this way, large- scale so­cial change is pos­sible.”

The the­ory is that every­body wins be­cause, even though they’re pay­ing in­terest, Utah will save money be­cause it won’t have to of­fer ex­pens­ive spe­cial edu­ca­tion courses to as many chil­dren. The ini­tial sav­ings for the Utah bond come in around $281,550 (based on a $2,607-per-child spe­cial-edu­ca­tion add-on cost). In­vestors re­ceived 95 per­cent of that fig­ure, around $267,000.

Yes, Utah still pays most of what it would have spent, but since re­search sug­gests that stu­dents who re­ceive spe­cial edu­ca­tion are more likely to end up in costly ju­ven­ile de­ten­tion pro­grams, the long-term sav­ings add up.

Fol­low­ing the ini­tial res­ults, Pritzker and the Bridgespan Group pub­lished a “fun­ders guide” that out­lines early-child­hood-in­vest­ment op­por­tun­it­ies, in­clud­ing so­cial-im­pact bonds. As the re­port lays out, the Rock­e­feller Found­a­tion and oth­ers have furthered the so­cial im­pact bond arena by back­ing, in Rock­e­feller’s case, Har­vard’s So­cial Im­pact Bond Tech­nic­al As­sist­ance Lab, which of­fers free as­sist­ance to states and cit­ies pur­su­ing the bonds.

Still, the total so­cial-im­pact bond mar­ket is small, around $80 mil­lion. Gold­man’s share of that is around $30 mil­lion. Bank of Amer­ica Mer­rill Lynch entered in­to a so­cial-im­pact part­ner­ship in 2013 with New Yorkto fund a five-and-a-half-year work­force reentry pro­gram for formerly in­car­cer­ated people. Gold­man de­clined to provide de­tails on in­vestors, but both found­a­tions and in­di­vidu­als have ex­pressed in­terest in these types of in­vest­ments, and in­terest ap­pears to be grow­ing, and is not lim­ited to early child­hood edu­ca­tion. The concept is be­ing con­sidered as a way to ad­dress en­vir­on­ment­al is­sues. The bank said it has fielded in­quir­ies from gov­ern­ments and private found­a­tions in Asia.

While ini­tial res­ults in Utah ap­pear to be pos­it­ive, so­cial-im­pact bonds don’t al­ways yield a profit for in­vestors. Gold­man’s first for­ay, to re­duce re­cidiv­ism at Rikers Is­land pris­on in New York, was scrapped after poor ini­tial res­ults. That ex­per­i­ence, An­drea Phil­lips, a vice pres­id­ent in Gold­man Sachs’ Urb­an In­vest­ment Group, told Next Amer­ica, in­dic­ated that the mod­el is “bet­ter suited to the ex­pan­sion of ex­ist­ing pro­grams…rather than tak­ing a pro­gram that has been used else­where and bring­ing it to a new op­er­at­ing en­vir­on­ment.”

Crit­ics of the idea ques­tion the in­volve­ment of big banks in early-child­hood edu­ca­tion and say gov­ern­ments might feel like they can cut budgets for badly needed pro­grams and rely on the bonds in­stead. The fed­er­ally fun­ded Head Start pro­gram was slashed as Gold­man ramped up preschool fund­ing, PBS noted. But Gold­man doesn’t dic­tate cur­riculum, and pro­ponents say the bonds could lend real weight to ef­forts to ex­pand early-child­hood edu­ca­tion and con­ser­va­tion ef­forts, and force every­one in­volved to fo­cus on meas­ur­able mark­ers, which are too of­ten ig­nored. A bi­par­tis­an group of ad­voc­ates, in­clud­ing Save the Chil­dren Ac­tion Net­work and Re­pub­lic­an Sen. Or­rin Hatch, who in­tro­duced an amend­ment in the No Child Left Be­hind reau­thor­iz­a­tion to ex­pand pay-for-suc­cess pro­grams, have praised the idea.

For early-child­hood-edu­ca­tion ad­voc­ates and back­ers of oth­er causes that so­cial-im­pact bonds might serve, the concept of­fers a path to­ward pro­gress that is of­ten slow or nonex­ist­ent through tra­di­tion­al gov­ern­ment chan­nels. For in­vestors, the bonds are a chance to make money and feel good about the change they are im­pact­ing at the same time.