Can Science Make People Save Money?
Friday afternoon is one of the busiest times at RiteCheck, a 24-hour check casher in New York’s South Bronx. At the register, Dayana Machin, 29, jokes with customers as she runs their checks through a computer system—ribbing a regular for having a rival check casher’s loyalty card on his keychain, accepting compliments on the golden lion hanging around her neck. Every time she hands a customer a wad of cash she says, “Anything else, my love?”
And sometimes, with the right customer, she asks something else: “Would you like to open a savings account?”
Most of the customers at RiteCheck have never had a savings account, so Machin takes a few minutes to explains how it works: You have to come to a branch to deposit or withdraw money, but there are no overdraft fees and it takes only $5 to open an account.
“Gratis?” one man asks at least five times.
“You have to always remind them it’s totally free,” Machin says, after assuring the man that it’s “totalmente gratis.” He opens up an account and puts $300 in it. “I can tell who the person willing to open an account is,” she says. “He’s nice. You tell him something, he will trust you.”
The savings accounts at RiteCheck, called Cash and Stash, are an experiment—both for the check cashing business and for a group of social scientists with the research and policy nonprofit Innovation for Poverty Action (IPA). Originally, RiteCheck wanted to offer savings accounts as a loyalty program: If people have to come in to make withdrawals, they won’t go to a competitor. But check cashers can’t hold money for customers. So RiteCheck collaborated with IPA, along with a local credit union, to run a pilot program. In exchange for helping RiteCheck, IPA is studying how customers behave. The goal: to see if they can get people to save money by using behavioral psychology.
“You accept this fundamental principle that people are flawed,” says Jeff Mosenkis, a senior policy communications associate at IPA. “How can you get flawed people to do what they want to do: save instead of spend? You build financial products that take advantage of these flaws.”
Americans are bad at saving. In an annual survey by the Fed, almost half said they couldn’t come up with $400 in an emergency. The savings rate of the bottom 90 percent of American households hovers just above 1 percent.
There are many theories for why Americans don’t save, from poverty to debt to conspicuous consumption. But the most enticing comes from behavioral economics: It’s easier not to. Inertia is strong, and putting money away requires overcoming what economists call present bias. “The idea comes back to this human flaw, that we’re wired to live in the moment,” says Mosenkis. “Life used to be very short; your task was making it to the next day, not making it to 30 years from now.”
The good news, according to behavioral economists, is that we can just as easily be tricked into overcoming that psychology with “nudges” that reframe incentives. Just post calorie counts next to unhealthy food, and people won’t order cheeseburgers. Or, make 401(k) plans opt-out, and more people will save money for retirement. Suddenly, with one oh-so-simple tweak, making bad decisions becomes the harder option.
That’s the theory behind the Cash and Stash study: A small push in the right direction can get people saving.
If only it were that easy.
It’s hard to get anybody to do anything,” says Dean Karlan, 46. We’re sitting in a dark room that’s all mahogany and library lamps at the Yale Club, about 12 subway stops from RiteCheck.
Along with Dartmouth economics professor Jonathan Zinman, Karlan, an economist at Yale, is the architect of the RiteCheck study. At any given time, Karlan, a founder of IPA, is working on several different projects across several different continents. “Nobody has ever used the word focus and me in the same sentence,” he says. After seeding a research empire in developing countries abroad, he started working on pilots and research projects in the U.S. in 2002, he estimates.
Karlan is a fidgety person, and he can’t sit still in his burgundy leather chair as he explains the RiteCheck study. It’s an elegant experiment: After 1,000 people signed up for savings accounts, IPA split them into four groups. The first got a verbal reminder, or “nudge,” every time they came into a branch to do a transaction. (A pop-up on the teller’s computer screen would prompt her to remind the customer to put some money away.)
The second group got a different incentive: what’s called a prize-linked savings account. For every dollar they saved, they got a lottery ticket for a $50 pot doled out each month. Those people also got a verbal nudge, every time they came into a RiteCheck, about their savings account and the lottery.
The third group got RiteCheck reward points for every dollar saved, as well as a nudge from the teller. Points can be redeemed for appliances, electronics, and toys. Among the most popular items: a toaster (185,000 points), a digital photo frame (560,000 points), and a portable DVD player (700,000 points).
And finally, the last group—the control—got no nudge and no incentive at all.
In his previous research, Karlan has proven that these kind of reminders and incentives work. Studies in the Philippines, Peru, and Bolivia found that text message reminders increased the likelihood that a participant would reach the savings goal by 3 percentage points, from 55 percent to 58 percent. Prize-linked savings, for example, turn the lottery on its head: Instead of draining people’s spare cash—Americans spent $70 billion on lottery tickets in 2014—it helps them save it.
“The idea is tying the lottery instinct to deposits,” explains Mosenkis, sitting across from Karlan. “You harness a harmful instinct and figure out how to make it for good.” It has proven to work abroad. Over 20 countries have offered prize-linked savings accounts, according to a paper from the National Bureau of Economic Research. The accounts have gained traction in the U.S. only slowly, because of strict lottery and gaming regulations.
“The process here in the U.S. is different than overseas,” Karlan says. “It’s hard to get people’s attention. People worry about privacy issues, there’s a lack of trust in institutions.” In Uganda, he says he gets 90 percent completion rates with surveys. IPA could barely sign 1,000 people up for RiteCheck savings accounts.
And that was just the beginning of IPA’s problems. At every step of the way, the study ran into a web of competing incentives and pesky human flaws that hurt its goal of getting poor people to save money. “The first six months were basically a bust,” Karlan adds.
If you don’t look too closely, RiteCheck looks like a bank. Its storefront has the cleanest sign on the block. Inside, flat screen TVs flash the news. One customer called the branch “beautiful.”
Look closer and you’ll see that week’s $111,000 Mega Millions state lottery potential prize winnings flashing across a ticker. Below, there’s what looks like a fast-food menu of financial services. Most people come in to cash paychecks, but customers can also pay ConEd utility bills or purchase money orders, lottery tickets, and subway cards. RiteCheck collects fees: $30 for a $1,500 check, 89 cents for a money order, $1.75 for paying a bill.
Check cashers profit from and, at times, take advantage of poor people, who don’t have many options when it comes to banking. In spring 2016, the Consumer Financial Protection Bureau (CFPB) filed a lawsuit against All American Check Cashing for misleading customers. The suit alleges that All American instructed employees to hide its fees and make it difficult for customers to cancel transactions after learning about charges. CFPB estimated that All American makes $1 million each year in fees.
“It’s a misunderstood business,” protests RiteCheck President Joe Coleman, who is usually one of the only white people at RiteCheck’s headquarters inside the 138th Street branch in the Bronx. “Most middle-class people never have to use us. They imagine it must be something untoward, because we are profit-making businesses in low-income neighborhoods.” (Coleman, who lives in Westchester, N.Y., says he used check cashing himself in his youth. “I was a starving actor for awhile.”)
Coleman argues that for low-income people, his model is cheaper than going to a bank that charges overdraft fees and has a minimum requirement for opening an account. RiteCheck also tells customers exactly what everything costs up front. That may be true, but RiteCheck sure sells a lot of lottery tickets.
All this makes RiteCheck and IPA curious partners: RiteCheck makes money off poor people; IPA wants to help them spend less and save more. Check cashers cluster in poor neighborhoods with few banks. The South Bronx has only one bank per 20,000 residents; as of 2008, more than half the residents in Bronx Community Board 1, which includes RiteCheck’s Mott Haven neighborhood, had no bank account. There are two more RiteCheck branches no more than 5 blocks from the 138th Street location.
But, evil or not, check cashers are exactly where IPA wants to be. “The idea is deliberately targeting people who need it most, where they do their financial transactions” says Mosenkis.
It turns out that the people who use RiteCheck often have very good reasons for not having a bank account. Research assistants spent months trying and failing to sign people up for Cash and Stash. “People were pretty unimpressed overall,” says Kim Smith, a researcher at IPA who worked on the project. “They wouldn’t believe me. What’s the catch?” Others thought it was too good to be true. That’s because in some ways, it was: If customers have outstanding tax liens, the government can garnish those funds. Other people lack proper identification or a Social Security number.
It’s the tellers—and the relationships they forge with their customers—that keep many people coming back to RiteCheck. Lisa Servon, a professor of urban policy at the New School, spent four months working as a RiteCheck teller in 2013. “I interviewed 50 RiteCheck customers after my stint as a teller,” she wrote for a piece in the Atlantic’s City Lab. “When I asked them why they brought their business to RiteCheck instead of the major well-known bank three blocks away, they often told me stories about the things the RiteCheck tellers did for them,” Servon wrote. One customer, Jason Rivera, told me he came to RiteCheck because that’s where he went as a child. He had his son with him.
But it was also the tellers who often stymied IPA’s research. Once they got enough people signed up to have a reasonable sample size for the study, they found that the nudges worked—but only when tellers followed the prompts, which they often didn’t.
During the few days I spent at RiteCheck, I witnessed multiple tellers ignore the prompts. One afternoon, for example, a customer with prize-linked savings came up to the window to cash his paycheck. The teller reminded him that he had a Cash and Stash account, but failed to mention the lottery. The customer didn’t put any money in the account until a second RiteCheck employee chimed in to mention the possibility of winning $50 at the end of the month. The customer changed his mind and put $30 in the account.
It was incredible to see the power of nudges in action, but I’d also witnessed another psychological phenomenon: friction. “If you want to discourage a behavior, you increase frictions,” Mosenkis explains to me.
The savings accounts aimed to reduced frictions for customers by offering easy access and an easy sign-up process. But IPA didn’t account for other frictions: “They [tellers] just want to be quick and not read the message,” Machin tells me. “If it’s busy, there’s people, like, ‘It’s busy, I don’t have time to be talking.’”
For tellers, engaging with the savings accounts often invites unwanted friction of another kind into their workdays. On one Friday, a man comes into RiteCheck complaining that he is owed money. A RiteCheck teller patiently checks his account as he sighs and whines and yells at her. “We levied your account, which means money you owe,” she says, looking at his record on a computer. “They took $15 out of your account and sent it to the department of taxes. Then you withdrew $100. That’s the only withdrawal we’ve had. Right now you have zero balance.” She prints out his transactions for him and writes her number on a piece of paper.
He refuses to believe her, yells back that she owes him money, and walks out with the store’s black plastic garbage can as collateral. (RiteCheck has since upgraded to stainless steel cans.) “I love your garbage can. It’s beautiful in my room. It’s real nice,” he says when he come into RiteCheck a week later.
To counteract teller aversion to Cash and Stash, RiteCheck engaged in its own psychological warfare, dispatching “secret shoppers” who would randomly come into stores to test the employees. If workers followed protocol, they won some money. If they didn’t, they “got in trouble,” says Machin, bragging that she got a perfect score.
After a year and a half, it seems customers aren’t really using their savings accounts to save. RiteCheck reports that only 20 percent of 872 accounts were “active,” meaning they saw a deposit within a given month. And people who did use their accounts treated them more like checking accounts than savings accounts. A typical user would put in $50 one week, take out $40 the next, and then add $10 a third week.
“Sometimes you tell them, and they still don’t want to put money in the savings account. ‘Oh, next week, next week,’” Machin explains. It also doesn’t help that a cork board selling lottery tickets—a powerful nudge in the wrong direction—sits right behind check cashers as they ask customers if they want to save money.
A big reason RiteCheck customers don’t save, despite the nudges, is that they don’t have money. “I don’t think we affluent researchers fully appreciate the financial pressures on low-income families,” says George Loewenstein, a professor of economics at Carnegie Mellon University, where he studies behavioral economics. With the Bronx’s unemployment rate at 7.8 percent, around 30 percent of the borough’s 1.4 million residents live at or below the poverty line. “It simply may not be possible for them to save.”
The problem goes beyond a sheer lack of funds. The psychology of poverty is hard to overcome with a dainty nudge. “Every time you buy a cup of coffee or a sandwich, it’s just a lot of mental work,” says Eldar Shafir, a behavioral scientist at Princeton and the author of Scarcity: Why Having Too Little Means So Much. “We can assume we can live OK without the extra money, and they cannot.”
Nudges can’t overcome that psychology. At times, they can even make it worse. One of Loewenstein’s studies used a series of reminders to see if people would put money in a federally funded, matched-savings program for low-income families. The results, posted in February to the Social Science Research Network, found that “none of our four interventions had the desired effect of increasing savings.” Some, in fact, discouraged saving because they were overly complicated and added stress to the savings process.
Politicians and policy makers love nudges because of their simplicity. They’re easy, cheap, bipartisan-supported solutions to complicated problems. Last year, President Obama signed an executive order directing federal agencies to apply behavioral science to “better serve the American people.” Last month, the Social and Behavioral Science Team released its annual report, outlining the many ways the government is using nudges to improve policy—from automatically enrolling kids in free or reduced-price lunch programs to sending out e-mail reminders about income-driven repayment plans for student debt.
“Nudges can seem like quick fixes to problems that are much more substantive,” Loewenstein says. “And if nudges offer false hope, they give policymakers an excuse for not trying out more substantive solutions.” Both Loewenstein and Shafir now prefer heavier-handed, paternalistic approaches.
Karlan was hoping the study would show a clear method for turning check cashers into savers. “That’s something Consumer Financial Protection Bureau would be very interested in knowing,” he had said, his eyes lighting up. “This could affect regulation about how they go about regulating check cashers. And rather than treat the check cashers like evil entities, partner with them. They have people’s attention and control over the cash flows. They’ll make more money, you’ll get people into savings accounts.”
But even as he spoke, he knew that this was not how it would work out. The CFPB declined to comment, but Karlan hasn’t ended up with the clear-cut results he wanted to to show them.
Instead, the study’s preliminary results were muddy. They suggested that the nudge method did get some people to save more: Deposits increased when people got some kind of reminder. But they didn’t show whether one type of nudge worked better than any other (possibly because of teller error), and they provided no evidence that the savings accounts helped people build up money over time.
Nonetheless, Karlan says that he’s not disappointed at how the study went. “There’s skepticism but hope,” Karlan says. “We’re skeptical optimists.”
The study officially ended in July. While RiteCheck will keep the savings accounts, it will stop using the reminders for now. “No more prompts. It’s over,” says Coleman, the president. “We have a marketing calendar, where every month we push a particular program. This will be a good one to push, actually,” he says. (As it turned out, the study did show that the savings accounts increased customer loyalty—just as RiteCheck originally wanted.)
More exciting to Coleman is the new RiteCheck outlet opening up in the Throggs Neck Shopping Center, also in the Bronx. “A brand-new shopping center over by the Whitestone Bridge! It’s got a Target, Five Guys, a Sleepy’s Mattress. We’re hittin’ the big time,” he says, laughing.
Corrects the roles of architects of the study in the 13th and 14th paragraphs.