by Dean Karlan
Each camp claims prominent economists as adherents: Jeffrey Sachs of Columbia University, an adviser to the United Nations, and Bill Easterly of New York University, a former senior official at the World Bank. Sachs and his supporters regale us with picture-perfect transformational stories. Easterly and the other side counter with an equally steady supply of ghastly the-world-iscorrupt-and-everything-fails anecdotes. The result? Disagreement and uncertainty, which leads to stagnation and inertia—in short, a train wreck. And no way forward.
Jake and I propose that there actually is a way forward. My hunch is that, at the end of the day, even Sachs and Easterly could agree on the following: Sometimes aid works, and sometimes it does not. That can’t be all that controversial a stand!
The critical question, then, is which aid works. The debate has been in the sky, but the answers are on the ground. Instead of getting hung up on the extremes, let’s zero in on the details. Let’s look at a specific challenge or problem that poor people face, try to understand what they’re up against, propose a potential solution, and then test to find out whether it works. If that solution works—and if we can demonstrate that it works consistently—then let’s scale it up so it can work for more people. If it doesn’t work, let’s make changes or try something new. We won’t eradicate poverty in one fell swoop with this approach (of course, no approach yet has managed to do that), but we can make—and are making—real, measurable, and meaningful progress toward eradicating it. That’s the way forward.
To get there, we need a two-pronged attack.
The first prong is to understand the problems in the first place. Some problems are systemic, in the way entire populations interact and exchange information, and in the way they buy, sell, and trade. Increasingly we are recognizing that the problems are also with us as individuals, with the way we make decisions. Here we turn to behavioral economics for insight.
In the past, economists would have thought about the monks in a pretty wooden, mechanical way. They would have talked about the cost of the fish, the value the monks imputed to their survival, the opportunity cost of the fishermen’s time, and the social impact of running the boat on diesel fuel. They would have put you to sleep. More important, at the end of the discussion the monks probably would still be dumping tubs of fish into the Marina del Rey causeway.
This is a narrow view of what makes us tick. Traditional economics gives us economic humans, the archetypes for rational decision-making. Borrowing a term from Richard Thaler and Cass Sunstein (from their book Nudge), I call these folks Econs. When they need to choose between two alternatives, Econs weigh all the potential costs and benefits, compute an “expected value” for each, and choose the one whose expected value is higher. In addition to keeping a cool head, they are very methodical and reliable calculators. Given accurate information about their options, they always choose the alternative most likely to give them the greatest overall satisfaction.
Behavioral economics expands on narrow definitions of traditional economics in two important ways. The first is simple: Not everything that matters is dollars and cents. In a sense, this is nothing new. For instance, Gary Becker—by many accounts a “traditional” economist—has been using economic analysis to think about marriage, crime, and fertility for years. The second expansion is a bit more radical. Behavioral economics recognizes that, unlike Econs, we do not always arrive at decisions by calculating a cost-benefit analysis (or even act as if we had done so). Sometimes we have different priorities. Other times we are distracted or impulsive. We sometimes slip up on the math. And, more often than we’d like to admit, we are shockingly inconsistent. To mark all of the ways we are different from Econs, Thaler and Sunstein use the powerfully simple term Humans. I will do the same.
Behavioral economics incorporates more nuanced behavior, and sometimes inconsistent behavior—like when we continue to sneak an occasional candy bar when we say we want to lose weight, or when we still eat dinners out while we try to pay down our credit card debt. It might suggest that the monks don’t care what traditional economics has to say. Maybe they throw the fish back because paying for not-fishing wouldn’t serve their purpose. Maybe it’s important to them to hear that silvery splash, or to see the minnows dart away like a bursting firecracker. Maybe there is something psychological about the salience of seeing, with one’s own eyes, fish jump free. And maybe the monks simply are willing to accept a less efficient solution in exchange for that moment of spiritual connection.
The breakthrough of behavioral economics has been to claim that if we want to understand the monks, then we must know how and why they make the decisions they do. Instead of deducing a way to think from a core set of principles, behavioral economics builds up a model of decision-making from observations of people’s actions in the real world. As we will see throughout this book, this way of thinking can help us design better programs to attack poverty.
This does not imply that we should throw out the old models. Behavioral economics is a powerful tool, but the proverb still applies: Just because you have a hammer, doesn’t mean everything is a nail. The inspiration for some of the antipoverty programs we’ll see comes straight from nuts-and-bolts economics. Combining the old and new approaches gives us the best chance to understand exactly what problems we’re up against, and to design and implement the best solutions.
This first prong of the attack—understanding the problems we face—is a start, but it’s not enough. Imagine you are stranded on a desert island with a rusted-out rowboat. Understanding the problem, even deeply, is like understanding why boats full of holes don’t float. That alone will not get you home. You need to find a way to build a better boat.
Hence, the second prong of the attack: rigorous evaluation. Evaluation lets us compare competing solutions—like different boat designs or plugs for the holes—and see which one is most effective. Creative and well-designed evaluations can go even further, and help us understand why one works better than another.
Here’s how it might work with the monks. I could propose setting up a new market, a market for hiring fishermen to not-fish, which would enable the monks to save fish more efficiently. It might sound good in theory, but then we’d go to the field and test.
Sometimes things that sound good fail. Suppose the monks actually don’t care about seeing the splash of the minnows and would be happy to pay fishermen to not-fish; maybe they are simply up against a problem that makes it unfeasible. It could be a trust problem, where the monks fear the fishermen would accept payment for not-fishing and go out fishing anyway. Or maybe it is a monitoring problem, where there just aren’t enough monks to tail around all the fishermen on not-fishing days to ensure they keep their word. A rigorous evaluation could point us to the specific problem that keeps the not-fishing market from saving more fish.
In the context of development, rigorous evaluation can help resolve the debate about how best to attack global poverty, by going to the field and finding out whether specific projects work. (It turns out that some projects work better—sometimes much, much better—than others.) You might think this goes without saying. You might assume that aid organizations have always routinely conducted careful and rigorous evaluations to see if they’re doing the best they can. If so, you would be surprised.
Until recently, we knew astonishingly little about what works and what doesn’t in the fight against poverty. We are now beginning to get the hard evidence we’ve lacked for so long, by measuring the effectiveness of specific development programs, many of which you’ll read about in these pages. The next chapter will go into a bit more detail about how we do this.
Microcredit, the provision of small loans to the poor, is a perfect example of an idea that generated tremendous enthusiasm and support long before there was evidence on its impact. The excitement is largely understandable, for the very design of microcredit is appealing. It strikes a lot of chords: Microcredit often targets women, and many believe that the economic empowerment of women redo
unds to the benefit of the entire family; microcredit often focuses on entrepreneurs, and many believe that such individuals, given access to a modicum of working capital, are capable of dramatically improving their lives through their ingenuity and enterprising spirit; microcredit often involves communities, and many believe that by involving the community rather than just individuals, we are more likely to succeed.
But in some sense the enthusiasm is surprising: It seems to be predicated on a double standard about the useful role of high-interest debt. At the same time that we see millions of dollars pouring into microcredit programs to lend to poor microentrepreneurs at rates ranging from 10 to 120 percent APR (all in the hope of alleviating poverty), we also see millions of people outraged at payday-loans outfits at home, which lend at similar rates to the poor in America.
Without some basic facts about whether these loans actually make people better off, I would not know which side to believe, much less how to reconcile the two positions. But rigorous evaluation can—and does—help. Many were surprised by a study in South Africa, which we will see in chapter 4, that found that access to consumer credit, even at 200 percent APR, made people much better off on average. This does not imply that all credit is good for all people, but it should make us look critically at our strong opinions about what works and what doesn’t, about what’s good and what’s bad. Do we have concrete facts to back them up?
The two-pronged attack we’ll see throughout this book is a powerful economic tool. I use it (albeit in a slightly different form) whenever I teach development economics, both to undergraduates and doctoral students. Three questions organize our discussions. First: What is the root cause of the problem? Using both behavioral and traditional economics to answer this question is exactly the first prong of our attack in this book. Then two more questions: Does the “idea” at hand, whether a government policy, NGO intervention, or business, actually solve the problem? And how much better off is the world because of it? Using rigorous evaluations to answer these two questions together is the second prong of our attack.
Jump in Singer’s Lake
Even in the absence of hard evidence about specific programs, people find compelling reasons to engage in and support the fight against poverty. One such reason comes from ethics, plain and simple: Suppose you are walking down a street by a lake on your way to a meeting, and if you miss the meeting you will lose two hundred dollars. You see a child drowning in the lake. Do you have an ethical obligation to stop and jump in and save the child, even though it will cost you two hundred dollars?
Most people say yes.
Don’t you then also have an ethical obligation to send two hundred dollars right now to one of many organizations delivering aid to the poor, where it can save a child’s life? Most people say no—or, at least, they don’t cut that check.
The example comes from Peter Singer, a utilitarian philosopher at Princeton University and a hero of mine. I tend to think of it at some very specific moments, like when I am in a store and tempted to buy something that I don’t really need. Couldn’t that money go toward something better?
Singer’s basic idea resonates, at least with me, but the logical conclusion of his argument is hard to swallow. The implication of his strict utilitarian reasoning is that we should all give away our money until we are so hard up that we honestly couldn’t spare two hundred dollars to save a drowning child. Maybe an Econ would feel compelled by the cold force of logic to do so (assuming, of course, he’d had the heart to save the drowning child in the first place). But no Human I know of—not even Singer himself, a tireless advocate for doing more—goes that far.
Because the conclusion to the lake analogy makes us uncomfortable, we grope for holes in the logic. We raise objections. Often people’s first response is to point out that when you dive in and pull the child to safety, there is no question that you’ve made a difference in the world. You can see with your own eyes that you’ve saved a life. But when you cut a check to an aid organization, the link is much less clear. How can you know your two hundred dollars is really doing good?
Most of this book is an attempt to respond to that objection. I hope that seeing some successes (and failures) up close convinces you that we can know we’re doing good—if we commit to rigorously testing aid programs and supporting the ones that are proven to work.
The second objection people raise to Singer’s lake analogy is about the “identifiable victim”—a vague sense that there’s something morally significant about seeing the child flailing around in the lake, whereas we can’t see the child our two-hundred-dollar check would be saving in, say, Madagascar. Logically, this objection is easy to refute. If someone runs to your house to tell you there is a boy drowning in the lake, you still have to go save him even though you haven’t seen him with your own eyes. Wearing a blindfold does not solve ethical conundrums, and we cannot confine our responsibilities to a specific geographic area simply by narrowing our field of vision. A child is a child, wherever he is in the world, even if we cannot see him.
The trouble is that, while this refutation might be logically valid, it isn’t viscerally compelling. We cannot simply reason our way into having a feeling of compassion, of responsibility for others. We need to be moved to act.
Behavioral Solutions Right Under Our Noses
Aid organizations, which depend for funding on our feelings of compassion, know from experience that appealing only to people’s ethical obligations doesn’t pay the bills. That’s why tactics like the identifiable victim are longstanding staples of fund-raising. Think of Save the Children, which promises a photograph and a handwritten letter from your sponsored child in exchange for thirty dollars a month. Rather than approaching donors with facts, figures, and tables—which is what might sway an Econ—aid organizations take full advantage of the fact that we’re Humans. They capitalize on our emotions.
This is exactly behavioral economics applied to the marketing of charities. Once you get inside the minds of those who give, you can come up with clever strategies to raise more money.
One such fund-raising strategy takes the financial sting out of giving by tacking donations onto other purchases. Recently I was in the checkout line at Whole Foods Market when the cashier asked me if I wanted to donate a dollar to the Whole Planet Foundation. She pointed to a small flyer on the counter. If I wanted to donate, she could scan a bar code on the flyer and add a dollar to my bill.
With a hundred dollars of groceries already rung up, an extra dollar is a tiny hit to take—so tiny you’d hardly notice it. And you get a lot of bang for that buck. Suddenly, you feel good walking out of Whole Foods Market with your bags of groceries. You’ve done something positive. It’s not hard to see why the Whole Planet Foundation has been awash in donations.
Another behavioral approach to fund-raising involves separating the good parts of contributing (i.e., the satisfaction of doing a good deed) from the bad (i.e., the pain of parting with your money). Giving becomes much easier if you can enjoy the satisfaction up front, unencumbered by that irksome feeling that your wallet is thinner, and pay later.
That’s exactly what happened in the phenomenally successful “Text to Haiti” campaign in January 2010. In the weeks following the devastating earthquake, people rose up in unprecedented numbers and acted to help those in need. Small individual contributions—the vast majority of them ten dollars or less—piled up at an unbelievable pace. Text donations from the first three days alone totaled more than ten million dollars.
Giving by text message takes a few seconds and is utterly gratifying. You type in “HAITI,” press Send, and get an instant response thanking you for your generosity. You hardly have time to think of the phone bill coming at the end of the month. When it does arrive, your ten dollars is easier to part with because it’s tacked onto the cost of phone service—a cost you’re already prepared to bear.
Unless, that is, you are Cara. The following was pulled from a real Facebook page:
Ca
ra’s profile said: “I’ve texted Haiti to 90999 over 200 times . . . over $2000 dollars [sic] donated to Haiti relief efforts. Join me!”
COMMENTS
Noah: Your parents might not like your cell phone bill this month
Cara: It’s not my money! Hah
Cara: Wait a second . . . this doesn’t get added to the phone bill does it? I thought it was just a free thing . . .
Aaron: Cara shooot. No every text is $10!!!
Cara: Oh wow, are u sure? This could be very bad for me.
Aaron: Yeah I saw it on the football game they bill it to your cell phone bill.
Chloe: Yeah. Every text is 10 bucks. It said so when the Health and Human Services lady came on and told people about it on the Colbert Report. Uhh, Ask for people to help you pay your phone bill?
Cara: Thanks for letting me know! Haha Haiti must love me!
Kyle: A 2000 dollar phone bill? this is sitting in its own special zone of hilarious.
Aaron: Well . . . you may be screwed but, in this case there’s a big upside at least
Cara: Just counted my texts . . . grand total is 188 texts. $1,880 phone bill . . . this is not hilarious Kyle!!!!!!
Never mind, Cara—there are worse ways to make a mistake with $1,880. And this really doesn’t happen very often—in the vast majority of cases, people know exactly what they are giving when they give it.
But behavioral marketing approaches can make it so donors may not always know exactly what, or whom, they are giving to, and this is more disquieting.
As an example let’s look at Kiva.org, a tremendously popular Web site that raises money for microlenders around the world. Ask a user of the site how it works and this is what you’re likely to hear: You log on and read through the stories of people who need loans. When you find one you like, you can fund her loan by clicking and sending money through Kiva. When the client repays her loan, you get your money back.
