The Rich Man's Bet

In the movie Trading Places, two rich old guys place a bet on a social experiment-- switching poor Eddie Murphy and rich Dan Ackroyd. It was a striking premise for the film precisely because it seems callous to place monetary bets on peoples' lives. Who would do such a thing?

Well, let's talk about social impact bonds, aka pay for success.

We've talked about this before, because SIBs have been waiting in the wings for their big chance on the education stage for a while now. The premise seems benign enough, but SIBs must inevitably lead to some highly undesirable consequences.

The Basic Idea

Think of SIBs as a rich man's bet. Let's say that I'm a Goldman Sachs type financial institution (Goldman Sachs is a big fan of SIBs). I look at your local pre-school program on which you are spending $2 million, and I say, "I'll bet you that I can run that program for $1.6 million." So you turn the program over to me to operate. We agree on a metric that we will use to decide whether I won the bet or not; if I win, some portion of the $400K that was saved goes to me.

Yes, I'm simplifying the whole business, especially all the various financial tools used to play with the money involved. But you get the idea.

So what's the problem? Well, there are several Bad Things that are likely to occur as a result of our little bet.

Oversimplification of Goals

If we're going to settle a bet, we will need a simple and clear metric for determining success. After all, a goal like "pre-school graduates will be healthy and happy and excited about continuing their education." So we could pick something like "few pre-schoolers will require special services in grades 1-6." Or we could set a cut score for success on a Big Standardized Test or a success level score for a battery of smaller, daily, personalized [sic] learning style mini-tests. We could use a system of microcredentials and set success as students achieving a certain number of edu-badges.

Because we need something clear and specific, we are bound to create two problems. One is that clear and specific goals flatten education; the higher order, more important goals in education are neither simple nor easily measured. So we end up with "gets at least five out of seven multiple choice questions right on quiz about the excerpt" rather than "able to apply critical thinking to deep discussion of themes used throughout the novel."

And because we focus on flat, narrow data goals that are really proxies for our real goals, we open up the door to Campbell's Law and start working to game the numbers rather than educate the children.

Vast Hunger for Data

To settle the bet, we need all the data. In fact, the more data we collect, the more things we can find to make bets about. The personalized [sic] learning model doesn't have to be part of the SIB picture, just as blockchain-based permanent records need to be- but both are very conveniently positioned to push SIBs, and help create a multitude of revenue streams. It's like a mining company that says, "We'd like to dig up this whole area to collect pyrite" and then takes everything, including the gold and silver that are right next to the pyrite.

But for my immediate purposes, having the maximum amount of data lets me better push the students across the finish line that I've set for them. I'll just go ahead and monitor the students and watch the steady flow of data. It's like watching and making bets on the flow of data generated by the stock market, only this data is generated by tiny humans. SIBs, armed with vast data gathering capabilities, turn students into a human commodities market.

If it doesn't bother you to consider children as a commodity, then consider this-- the regular market has had a destructive effect on US business precisely because too many companies are now run poorly by CEOs whose philosophy is "We're not here to make widgets; we're here to provide stockholders with a good return on investment. Our job is all about getting those stock numbers to look good." The market too easily loses sight of what the actual point is supposed to be; a school where that happens, where the students are there not to be educated, but to generate profitable numbers for Goldman Sachs-- that's a school that has lost its way.

Privatization

SIBs are not just a bet; they are also a way for Goldman Sachs et al to buy the profitable part of a public entity, while letting the public still carry the risks and liabilities. It is a way to buy up public goods without actually owning it outright.

Investors still get all the important parts. They get to dictate how the school will be run, how the programs will be operated, how the students will be treated. They get to skim whatever profits they are able to squeeze out. But if the school, say, burns down-- well, that's not the investor's problem. Because while I'm characterizing this as a bet, it's different in one important feature. When I bet you that I can run your school more cheaply than you can, here's the deal-- if I win the bet, you give up a bunch of tax dollars, and if I lose the bet, I give up... well, I give up nothing. I give up the chance to collect some of the money. (And really, if I lose this bet, it's because I didn't do my homework and I somehow set terms that didn't favor my success,)

In the meantime, I get to treat the school like my own possession. And the taxpayers have given up control of the school that they still technically own.

Your Homework

If you'd like to read about social impact bonds until your eyeballs bleed, I recommend this list and this collection. The whole business can seem painfully technical at times, but there are people in the world who are very very very excited about seeing this happen, and we should be paying attention.