by Dean Karlan
But a less obvious dynamic also stood out. The strongest predictor of account usage was participation in a homegrown savings scheme, like the Rotating Savings and Credit Associations mentioned earlier. That is, people who were already saving (albeit not in a formal bank) at the time of the offer were significantly more likely to become active users of the new accounts. That’s sort of a strange finding. Participants in homegrown saving schemes had already found a way to save; why did so many of them rush to sign up for the new accounts? One explanation is that the new accounts were actually better than the existing alternatives. The way to find out was to see how the village cooperative accounts had changed people’s daily lives.
And so Dupas and Robinson came to the detailed daily logs, simple foolscap notebooks filled out in ballpoint pen and dull pencil. Despite their unassuming appearance, these were vast treasure troves of information, containing records of everything from purchases of business inventory to payment of hospital bills. Taken together, the logbooks told a coherent story: Saving in the village cooperative accounts, women invested more in their businesses and increased their spending on food and other goods.
There was also evidence that suggested the accounts improved women’s abilities to cope with illness, both to themselves and to other household members. Women without accounts in Dupas and Robinson’s study reacted to serious illness by working less, drawing down their working capital, and giving away goods on credit (which was likely a tactic to prevent spoilage). This compounds the effect of sickness by interrupting the business—and the income it generates—precisely when it is needed most.
In contrast, women who had been offered accounts did better. They could draw on savings to pay for treatment right away, so they tended to keep their working hours up even during weeks with sickness. As a result, they were not forced to dip into their working capital or give goods away on credit. Those are steps in the right direction. But why did this happen only for the women in Dupas and Robinson’s study? That is an open question worthy of further research. Their finding could be a fluke, or it could be a deep and important truth about the differences between women and men—a truth that has implications about how we should design future programs. Replicating their experiment is the best way to find out. To this end, Dupas and Robinson are currently working on follow-up studies in Kenya, and I am also working with them to replicate and scale their work in Uganda, Malawi, Chile, and the Philippines, through the generosity of the Bill & Melinda Gates Foundation.
Pascaline Dupas and Jonathan Robinson’s study in Kenya showed us two things. First, aspiring savers were in a bad spot to begin with. That so many signed up for an account with a negative effective interest rate attests to the amount of obstacles they faced. Second, the fact that village cooperative accounts really did improve people’s lives suggests that available alternatives were weak.
There is a much more positive way to look at these findings. Through their use of the village cooperative accounts, the people of western Kenya showed they had both the will and desire to save. If such a middling savings solution can do so much good, then how much more could we do with a better product? The impacts Dupas and Robinson found are encouraging. Now we have to look for ways to get more people on board.
Sunny Saves
Even without a drunkard husband like Vijaya’s to tear through our pocket money, many of us find ways to avoid saving. I recently performed a substantial feat of not-saving when I took a friend out for a fancy dinner to celebrate nothing in particular. That expense is easily dismissed as frivolous, but often our reasons for spending are, well, reasonable—repairs or improvements to a house, back-to-school shopping, sending money home to support extended family. These are not splurges.
Unfortunately, the view from inside the piggy bank is stark and simple: Either money gets deposited into the slot on top, or it doesn’t. The account balance is no respecter of reasons or justifications. Just shake the thing and see how much is jingling inside. Clink, clink. The coins are few; their sound is sad and small. Buyer’s remorse sets in. We wonder which recent purchases were unnecessary, which could have waited.
So it was with Sunny.
I met Sunny in the Philippines, in Butuan, a city in the north of Mindanao. Sunny had a few goals for her home. She wanted new plaster on her walls. She wanted to build a nice patio. She wanted to fix her bathroom. Each improvement would cost about two hundred dollars.
Sunny wasn’t the poorest person on her block, nor was she the richest. She certainly did not have two hundred dollars. What Sunny did have was a savings account at the Green Bank of Caraga. So she set out depositing money into it, with an eye on her goal. She put in five dollars at a time, and slowly but surely her account balance rose. But every time she reached fifty dollars, something seemed to come up. Nothing earth-shattering: One time her kids needed clothes. Another time her husband was agitating for a fancy new TV. What happened? There was a withdrawal, and Sunny went back to zero.
One day something changed.
The Green Bank came along and offered Sunny a new product called SEED (short for “Save Earn Enjoy Deposit”). SEED was identical to her existing savings account, with one exception: It had a commitment feature, which locked away her money until she reached a goal amount of her choosing.
Sunny signed up and set her goal at two hundred dollars. She saved two hundred dollars, withdrew it, and promptly signed up again. When I met her during a qualitative follow-up survey, she was on round three. She had never managed to save so much before in her life.
SEED was new to the Green Bank and to Sunny, but commitment savings is an old trick. In the United States, Christmas Club accounts have long helped people save little by little toward a big goal. What defines a commitment savings scheme is that deposits are locked up, either until a given date or until a specific account balance is reached.
What’s surprising, from the perspective of standard economic theory, is that people actually sign up for these things. The reasoning behind the standard view is that having more options—like the freedom to withdraw money whenever you want—is better. After all, you’re never forced to withdraw from a regular savings account. You are free to save toward a goal, or not. You can do what suits you. On the other hand, a saver in a commitment scheme has given up the withdrawal option. What could be good about that?
For one, it silences the voices of temptation. It snatches at the windpipes of all the things calling out to us and to our wallets. We cannot make impulse purchases when our money is locked away in a vault. That fact alone was enough to solve Sunny’s problem. Is the same true for others? Can SEED, or other commitment savings schemes, help improve the lives of the poor?
Getting the Poor to Save
The fact is that most poor people do not have savings accounts, commitment-based or otherwise, and conventional wisdom has a simple explanation for this. They’re not saving because they spend every last cent on the bare essentials. (They’re poor, remember?)
I’ll admit, there is something attractive about that reasoning. It’s clean and it makes a kind of intuitive sense. But it leads to a dead end. If not-saving is simply an inescapable fact of poverty, then even if we find ways to help poor people control their impulses and resist temptation spending, we won’t fix the root problem. They still won’t have anything left after taking care of their immediate needs. So we should not waste time and money finding better ways for them to save.
Fortunately, we need not accept that line of reasoning on faith; we can test it in the real world.
While I was studying at MIT, I met Mary Kay Gugerty, then a Ph.D. student at the Harvard Kennedy School of Government. Mary Kay was writing her dissertation on informal savings clubs that Kenyan women used to overcome self-control problems. I had also been thinking a lot about temptation—not just with savings, but in other areas of life as well. We went to talk with David Laibson, a behavioral economist at Harvard who had done some pioneering research on self-control.
&nbs
p; We went into the meeting hoping to find out how we could test his theories, but we walked out with a much more practical question: How can we use these theories to improve people’s lives? Specifically, can we design a product that will help solve the temptation problem?
A week later, Mary Kay and I joined up with Nava Ashraf, another graduate student, and started looking for a product to test and a place to test it. We wrote up a brief proposal and blasted it out to our e-mail list. It turned out a lot of people were interested in temptation and self-control. We got dozens of responses.
Nava and I went to the Philippines in August 2002 to follow up on some of them. John Owens, manager of a USAID initiative to assist rural banks, and microfinance guru, set up a meeting with a dozen potential partners. I’ll always remember that trip for one simple reason: Nava and I both had temptation on the brain. I thought about it every morning, when I had to fight to pass up a breakfast of French toast and banana bread while Nava effortlessly opted for a healthy bowl of fruit and cup of yogurt. She thought about it when we talked about clothes. It came out that I hadn’t spent more than ten dollars on a shirt as long as I could remember (I usually wear tie-dyes and batiks from West Africa or kurtas from India), while Nava had to hide her credit cards from herself to keep from splurging.
So we had a lot to say when we met with the banks we might work with. The very next day we had a strong partner in the Green Bank of Caraga, the same bank I later worked with to study group- and individual-liability loans. They were eager to find and test new ways to mobilize savings.
Nava, Wesley Yin (a graduate student at Princeton), and I worked with the Green Bank to create SEED, the savings account that would later help Sunny make her home improvements. And in the process we designed an RCT to find out whether a self-control product could really work. We wanted to know what kinds of people would sign up, and how SEED would affect overall savings.
We started by surveying about eighteen hundred current and former clients of the bank. Once the survey was complete, we randomly divided the respondents into three groups. Those in the first group were visited by a Green Bank employee, who spoke to them about the importance of saving and invited them to open a SEED account. Those in the second group were also visited by an employee and given an identical marketing spiel about saving, but were not offered SEED accounts. Finally, those in the third group received no visit and no offer. They were monitored for control.
Our initial survey captured more than the usual demographic and household information. It also included a battery of questions that measured “time-preferences”: a person’s willingness to pass up gains today for larger gains in the future. These are questions like “Would you prefer to get five dollars today, or six dollars a month from today? How about five dollars in a month versus six dollars in two months?” By probing across a variety of time frames and dollar amounts, we can sketch a good picture of a person’s preferences for immediate versus delayed gratification. Or, in plain English, we can find out about patience. Impatient people, for instance, choose the five dollars now over the six dollars in a month.
Things get interesting when preferences change over different time frames. Consider people who are impatient now (i.e., prefer five dollars today over six dollars in a month), but claim they will be patient later (i.e., prefer six dollars in two months over five dollars in one month). You know people like this. They are the ones who never have time to begin their new exercise regimen this week, but are sure they will next week. Or who are always going to start diverting a bigger part of their paycheck to retirement savings—next month. They have identified goals and ways to achieve them, but always find an excuse when the time comes to start. If you knew yourself to be such a person, you might jump at the chance to accomplish a goal by locking yourself into a commitment scheme. That is exactly what happened with SEED.
Overall, the product was a hit. More than a quarter of those invited to enroll did open an account. If that doesn’t seem like a strong response, consider what the Green Bank was offering: a savings account identical to the normal one (which anyone was free to open), but without the option to withdraw. Seen in that light, the 28 percent take-up rate is impressive: It means 28 percent of people wanted to lock up their money where even they couldn’t reach it.
We also confirmed our suspicions about the kinds of people who were opting in. Women with impatient-now-patient-later preferences were about 50 percent more likely to sign up as women without. So, on the whole, SEED was reaching the right people.
We were excited to see people signing up, but the big question was about impact. Would SEED really help people save? In a word, yes. We found that simply offering a SEED account caused the typical client’s balance to increase by 47 percent over six months. That figure rose to 82 percent after twelve months. Keep in mind that this was the average change in savings balances for everyone offered the account—regardless of whether they accepted. In fact, the impact of SEED only on those who actually opened accounts was a remarkable 318 percent. That is, we found that the effect of offering a SEED account to a client who will open it is to increase her savings balance fourfold!
The findings from the SEED study are important for two reasons. First, they confirm that Sunny’s experience is no mere anomaly: SEED accounts helped many Green Bank clients increase their savings. Second, and more fundamentally, they undermine conventional wisdom by demonstrating that, if given the right tools, the poor can save more, even without a boost to their overall incomes. This is a powerful and encouraging result, for it suggests that people can do better with their existing resources.
Maybe all they need is a behavioral push—or, to borrow a term from Richard Thaler and Cass Sunstein’s eponymous book on behavioral solutions to everyday problems, a little nudge.
Insights from the Home Front
We have seen what’s happening in a south Indian market, in the Kenyan bush, and amid the vast, wet checkerboard of Filipino rice paddies. But what about Decatur, Illinois? What about Astoria Boulevard, Queens, New York? Practical applications of behavioral economics are still, relatively speaking, in their infancy even in the developed world—and it’s worth taking a look at the kind of nudges and commitment devices that are catching on over here, to inspire new ideas for what might be tried over there. What’s more, as we find that behavioral solutions are adaptable to both rich and poor settings, we strengthen the case that they actually respond to something fundamental and shared, something that transcends the poverty line.
Evidence from the home front supports this view. It turns out that, when it comes to savings, even financial types and (gasp!) economics professors aren’t immune to biases and shortcuts in their thinking. The behavioral economists Richard Thaler and Shlomo Benartzi noticed that most people—including their fellow academics—tended to approach retirement savings contributions as they would a certain countertop rotisserie oven: Set it and forget it. When first joining a firm, employees chose a contribution level and an investment plan and tended not to change it. Ever.
From the standpoint of classical economics, this is puzzling. One’s needs and resources vary enough over the course of a career that it’s highly unlikely any single plan would be best throughout. So these people, savvy as they were, were not making the best economic decisions. What’s more, they were failing by their own assessment. When surveyed, many said they were dissatisfied with the size and allocation of their monthly contributions. This came as a surprise; after all, they were the ones who originally set those parameters, and they were free to change them anytime.
What was to blame for these mistakes? Thaler and Benartzi could see procrastination and inertia, two behavioral barriers to saving we saw earlier in the chapter, at work. They devised a plan called Save More Tomorrow (SMarT) that turned these barriers on their heads.
Under SMarT, participants agree now to a series of future savings increases coincident with pay raises, so that saving steps up over time without ever decreasing take-home p
ay. Because savings will not increase until a pay raise, signing up for SMarT is painless now—that’s good news for procrastinators. The plan is voluntary, so participants can opt out of the scheduled increases at any time. Of course, doing so requires some initiative and positive action; thus, for participants, inertia helps, rather than hinders, saving.
Thaler and Benartzi suspected they were onto something, so they partnered with a firm that agreed to offer SMarT and track its employees’ saving progress. The implementation went as follows: First, all employees eligible for the firm’s retirement savings plan were offered a free meeting with a financial consultant. For those who accepted, the consultant computed a desired savings rate and recommended an appropriate on-the-spot savings increase to get there. Twenty-eight percent of those who met with the consultant took his recommendation; the rest were offered the SMarT plan. An impressive 78 percent signed up. After four pay raises, the picture was striking. Eighty percent of those who signed up were still enrolled in the SMarT program, and signers-up were saving 55 percent more than those who had accepted the consultant’s advice.
Since the first implementation, SMarT has caught on. Fidelity Investments and Vanguard, two of America’s largest retirement savings plan operators, now offer a version of the plan to their corporate clients. As a result, millions of employees have agreed to Save More Tomorrow.
