by Dean Karlan
One reason why SMarT is such a powerful tool is that it allows people to bind themselves (not too tightly—one can always opt out) to doing the right thing. Economically, participants are changing the relative prices of good and bad behavior. Without SMarT, good behavior (like increasing retirement savings) is costlier than bad: It requires visiting the HR office and filling out a form. Once enrolled, the tables are turned. Good behavior is free, and bad behavior (like freezing or lowering retirement savings) is now the time-consuming option.
What if we had a way to change relative prices of other things in our lives?
With this question in mind, I created stickK.com, a Web site where anyone can make commitment contracts to achieve goals of his or her choosing. StickK.com lets users define what they want to accomplish, what’s at stake, and who will certify their success (or failure). This way people can change the relative prices of good and bad behavior directly, using dollars and cents. (You can also put your reputation up as the stakes by committing to automatically notify your friends and family of your success or failure.) Suppose, for instance, you wanted to work out at the gym once a week. With a stickK.com contract, you could fine yourself a hundred dollars for every week you failed to show up.
The plot thickens when you consider the recipient of the money you put on the line. When you make a stickK.com contract, you choose whether your hundred dollars will be transferred to a specific person, a charity, or an “anticharity”—that is, a charity you’d rather not support. Would you try harder to make it to the gym if failing meant sending a check to, say, the Bill Clinton or George W. Bush Presidential Library, as opposed to UNICEF? I suspect so! Other pairs of opposing “anticharities” include the National Rifle Association and the Educational Fund to Stop Gun Violence; Nature Conservancy and the National Center for Public Policy Research; Americans United for Life and NARAL Pro-Choice America. For those in England, stickK.com offers charities on both sides of the great, eternally contentious soccer rivalries: the fan clubs of Arsenal and Chelsea, Liverpool and Manchester United.
It has been a fun process to watch. We have seen plenty of contracts for weight loss, exercise, and smoking cessation. But the Internet has a way of bringing out people’s creative sides. With the power of incentives at their fingertips, people have used stickK.com to succeed in some unexpected ways. A few “custom” goals are worth noting:• I will not call him for two weeks
• No more dating losers
• No more porn
• No more afternoon five-dollar lattes
• No cutting my hair
• Spit my gum in a trash can and not out the window
And a few longer ones, complete with explanations:• I will wake up AND GET OUT OF BED each workday no later than six-thirty A.M. so that I will be at work no later than eight (goal being seven-thirty). I will e-mail my referee each workday morning so she can see that I am in by eight! I commit to taking a shower no longer than five minutes.
• Buy a car to replace the old POS I have. It will be manual transmission and it WILL be by the end of this year.
• Refrain from swearing even when Philly sports teams mess up if Katie is present. . . .
• Porn is destructive, unhealthy, and perverts my view of sex and relationships. It also can become obsessive-compulsive and an unhealthy release of stress. Watching porn also conditions you to the wrong types of women and the wrong kind of so-called physical “intimacy” instead of valuing their authentic feminine nature. So I commit to not watching porn more than once per week. [Note: Based on the vehement opening, I really thought this contract was heading for cold turkey on the porn!]
Gentler Nudges
Like SEED, stickK.com has some serious bite to it. There can be real money on the line, and money is a powerful tool for shaping our choices in the face of temptation. It makes sense: When you’re on the fence about going to the gym, the mere thought of losing a hundred dollars (or having to notify your friends that you skipped) is likely to push you in the right direction. Here’s a question, though: How far can we go with less bite? What if the problem is less our weakness than our forgetfulness—and what if the solution lies, accordingly, in the thought of losing a hundred dollars rather than the actual loss? Then you might change your behavior just by managing your attention. And maybe you could keep that hundred dollars after all.
That’s an idea we can test directly in the developing world. Maybe the poor don’t need the bite of restrictive commitment accounts (like SEED) in order to behave better. If they could just think of saving at the right times, maybe they would do more of it. Maybe a lot more.
Maggie McConnell (a post doc at Harvard and a former research assistant for IPA in Peru), Sendhil Mullainathan, Jonathan Zinman, and I set up RCTs to test this proposition in three countries: Bolivia, Peru, and the Philippines. Replicating experiments in differing contexts helps us address the ever-present question of “external validity”—whether results from one country or project site can be applied to another. If you really want to know, test the idea in a few different environments! Learn when it works, when it does not, and what the driving factors are for success. This principle was one of the central motivations for creating IPA, and we are testing a number of interventions in multiple locations for exactly this reason.
In all three places, we worked with clients who had recently opened “goal” savings accounts and had made plans to save every month for the year. We randomly chose some of them to get a little nudge toward their goals–the simplest nudge we could think of: We reminded them, once a month, to save. In the Philippines and Bolivia, we sent text messages. In Peru, where cell phone use was less widespread, we sent the messages through the physical mail.
In Peru, we also tried what we thought was a neat idea: to give people pieces of a jigsaw puzzle each time they made a deposit, so that after twelve deposits they would complete a picture of their savings goal, such as a vehicle, a home, or a student at graduation. We suspected that the simple act of depositing money was too abstract, and that a jigsaw puzzle piece would make their goal more salient—and thus provide an extra incentive to make the deposits. Every time we told people about this idea, we were enthused by positive feedback. Some early results even made us hope they would work.
But the jigsaw puzzles did not work. They didn’t cause savings to increase or make savers any more likely to achieve their goals. Where did they fail? Perhaps they got the timing wrong: The salience of receiving a piece occurred after the deposit was made. So if the problem was one of attention—that individuals weren’t thinking of their future goals—then the salience-enhancer needed to be done before they were to make their deposit, not as a reward for doing so.
Fortunately, the tiny little reminders really did work. Total savings increased by 6 percent, and people were also 6 percent more likely to reach their savings goals. The effect was even bigger for savers randomly assigned to receive mention of their specific goal (versus a vague reference to savings) in their reminder.
These messages cost the banks pennies apiece. It was practically a free lunch.
A practically free lunch is great, but how about an honest-to-God, zero-cost free lunch? Let’s look at one more nudge—one that’s even simpler (and cheaper) than sending text messages and still makes a big difference in people’s behavior. This one is right in our backyard, at H & R Block tax preparation offices in St. Louis.
In 2005, economists Esther Duflo, William Gale, Jeffrey Liebman, Peter Orszag, and Emmanuel Saez partnered with H & R Block to see how offering a one-time matching IRA contribution would affect retirement savings. They found, as expected, that larger matching contributions (say, 50 percent versus 20 percent) generate larger contributions from savers. But they also noticed something strange.
The tax code already contained a program called the federal saver’s credit, which was economically equivalent to the matching IRA contributions they were studying. But the federal saver’s credit was framed as a rebate instead
of a match. Since the two were ultimately the same in dollar terms, they were surprised to find that people reacted to them very differently. In particular, an increase in the matching level generated far more savings than an equivalent increase in the rebate.
Emmanuel Saez returned to St. Louis in 2006 to see what was happening. Again, he partnered with H & R Block and designed an RCT on retirement savings, but this one was different. While the first study was mainly concerned with responses to a range of economic incentives (i.e., the different matching levels), this one focused on a single one. The only thing that varied was the way that incentive was presented.
Some tax filers would be offered a 50 percent match on IRA contributions, and some others an equivalent rebate. In the 2005 study, the federal saver’s credit rebate was buried deep in the dense, forbidding tax code—so we might excuse filers if they were unaware of it. But this time he laid it out simply. All the offers were direct and transparent. Still, the results were similar: 10 percent of those offered the match made contributions, compared with 6 percent of those offered the rebate, and 3 percent of those receiving no special offers.
The findings of Saez’s 2006 study fit into a growing body of evidence that matching is more effective than equivalent credits or rebates at increasing contributions—to charities, to give one example. Knowing that gives us an easy opportunity to shape better products and policies. Just think: It costs nothing to describe a rebate as a matching contribution, but doing so could increase participation by two-thirds!
That’s one reason why strong results about the little things are so exciting—because they point to ways in which we can do better without reinventing the wheel. We are harnessing our irrationality to improve lives.
8
TO FARM
Something from Nothing
Farming is all about growth, about making something out of nothing (or very little). But the sad economic reality is that many farmers make nothing out of something. There is an old joke: A farmer wins the Kansas state lottery and is visited by a reporter from his local newspaper. The reporter wants to know what he plans to do with the money. He asks, “Will you buy a fancy car? Build a bigger house? Quit your job and move to Miami?” The farmer thinks for a moment and replies, “No, I think I’ll just keep farming until it’s all gone.”
Like most good jokes, this one contains a kernel of truth. Lots of farmers lose money. In America, where they often have the benefit of hybrid seeds, fertilizer, piped irrigation systems, perfect roads, mechanized equipment, futures contracts, and easy access to export markets, many struggle mightily to stay afloat. So how about their counterparts in the developing world? Armed with the creaky tools of an earlier age, they work the same earth their grandparents worked—often in the same way their grandparents worked it. They trudge out to the fields at sunup just like American farmers, but they push wheelbarrows of cow dung instead of fertilizer spreaders. They carry wooden hoes instead of tractor keys. Is it any wonder that prosperity eludes so many?
In the landscape of poverty, agriculture is far too big a region to ignore. Well over a billion of the world’s poor are farmers. If it is tempting to focus on other issues, perhaps the reason is that the challenges surrounding agriculture in the developing world are so numerous, so varied, and so tightly intertwined that the whole thing appears as an intractable knot.
Let’s start to unravel it. First, there are the environmental hazards that all farmers, rich and poor alike, face: droughts, floods, bugs, blights, and the like. Second is the technology gap, in both equipment and cultivation practices. When farmers work without drought- and disease-resistant seeds and rich fertilizers, without testing the soil chemistry to choose which kinds of crops to grow, and without sophisticated irrigation and drainage systems, they are more likely to lose harvests to bad weather and pests. Finally, there are structural challenges. Limited information about, and access to, profitable markets, swings in commodity prices, and the high costs of transporting and storing produce hamstring even those farmers who manage to grow a bountiful crop.
Taken together, these obstacles account for one fact we know for certain about farming in the developing world: It’s a tough row to hoe.
DrumNet and the Kitchen Sink Approach
The snarl of problems entangling poor farms hasn’t proved much more tractable to aid organizations or policymakers than to farmers themselves. But there have been some successes.
One idea is to provide an integrated suite of services that address many challenges at once. On the ground, this means doing more for farmers than just giving training sessions, or just offering loans for agricultural inputs. In some cases it means nudging them toward new cultivation techniques, or even entirely new crops.
DrumNet, a comprehensive agricultural development program in central Kenya, did just that. DrumNet was the brainchild of PRIDE AFRICA, a U.S.-based microfinance and agriculture nonprofit that serves some two hundred thousand clients across East Africa. With all its experience in the region, PRIDE AFRICA had developed the local expertise to impart some useful information to Kenyan farmers; surprisingly, the most valuable fact on offer was that Europeans had a taste for French beans and baby corn.
When DrumNet came on the scene in 2003, many Kenyan smallholder farmers were growing crops that they could use themselves or sell locally, like maize, potatoes, kale, and bananas. Some knew about the rich European export markets, but those who wanted to sell overseas faced significant obstacles.
First, information was a problem. The farmers, living in the rural villages of Gichugu Constituency, Kirinyaga District, at the foot of Mount Kenya, were out of the loop. They didn’t have access to current prices in the world markets for dozens of crop varieties. Second, trust was a problem. Relations with exporters—who were scarce in the first place—were fraught with mutual suspicion. Farmers feared that exporters would arbitrarily slap low grades on their produce or find other ways to shortchange them. Exporters, in turn, feared that farmers would refuse to sell at the agreed-upon price or would simply fail to produce an adequate crop. Third, credit constraints played a role. Selling for export in tightly regulated European markets meant paying for certification and grading in addition to the usual agricultural inputs. Without loans, these extra investments were out of reach for most farmers. The final hurdle to exporting was transportation—arranging and paying for trucks to move produce from farm to port.
PRIDE AFRICA saw in all these obstacles the intertwined strands of a single rope binding farmers’ hands, and they designed DrumNet to cut all those strands at once. The program would assist recipients with each step of the transition to exporting: training in farming practices and European agricultural standards, liaising with exporters, opening savings accounts, and providing in-kind loans of agricultural inputs. Although DrumNet was a nonprofit program, it was designed to be “sustainable”—to generate sufficient revenues to cover its costs.
Understanding the impacts of such a rich, multifaceted program was a tall (and exciting) order. With so many moving parts, it’s hard to see clearly what each one is doing. But sometimes it is good to start off with a kitchen-sink approach, see if the whole thing works, and then zoom in to figure out the value of each component. In April 2004, Nava Ashraf (from the SEED commitment savings project we saw in the last chapter), Xavier Giné (from the group-liability project in the Philippines, discussed in chapter 6), and I partnered with PRIDE AFRICA to do just that, and evaluate DrumNet with an RCT.
Once the program got under way, different kinds of shoots began to poke out of the furrowed fields of the Gichugu Constituency. French beans and baby corn crept in where maize and kale had dominated the season before. Farmers were responding with gusto to the opportunity presented by DrumNet’s suite of services. Those who had been invited to join DrumNet were almost 50 percent more likely to grow export crops than those assigned to the control group.
Interestingly, farmers who were already growing exportable produce at the outset tended to
stand pat, even if they enrolled in DrumNet. They didn’t dedicate any more of their arable land to export crops. Most of the increase in French bean and baby corn production came from switchers—farmers who had actively made the leap to exporting from subsistence farming or local cash crops as a result of the program. In accordance with the proverb, fortune favored the bold: By the end of the year, switchers’ household incomes had increased by almost a third compared to the control groups’.
Better Farming Through Chemistry
DrumNet was humming right along, getting money into Kenyan pockets and vegetables onto European plates, until it ran into the economic equivalent of an oncoming freight train. It was a messy and dispiriting crash, of which we will see a full account later. But before we look at that train wreck, let’s consider what the program got right: It hit on a viable (at least initially viable) solution to the specific problems Kenyan farmers faced. Its early success was not serendipity but the result of thoughtful planning and extensive local knowledge. PRIDE AFRICA knew, from its prior experience in the area, that famers were up against an information deficit, a tenuous relationship with exporters, and financial constraints. It had also got its hands dirty. It understood the soil of Gichugu Constituency and knew that French beans and baby corn could thrive there.
Had DrumNet been rolled out elsewhere in the country—in the low tropical plains of the coast, in the arid north, or even just on the far side of Mount Kenya—circumstances might have molded a radically different program. Even if those farmers were grappling with the same economic impediments, maybe the soil couldn’t have supported French beans and baby corn; maybe PRIDE AFRICA would have encouraged them to grow something else instead.