For the poorest of the poor, putting money into savings is no small challenge. Strategies like direct deposit, automatic 401k deposits and employer matching, and default contributions are not typically available to the very poor, who are often self-employed and don’t have formal, regular paychecks to redirect default savings from. Even small levels of savings have been shown to net large benefits to savers, and policymakers recognize that increasing the savings rate is an important means to improving quality of life.
But how can the poor be encouraged to put just a little bit more of their income into savings?Professor Stephan Meier often investigates the intersections between psychology and economics, and one line of his research examines the effects of peers on behavior. Meier notes that while the poor may not have access to retirement accounts and default depositing, they do have access to peers. “Weight Watchers and Alcoholics Anonymous use peers as commitment devices, so we tried applying that idea to savings for the poor,” Meier says.
Peer commitment devices are thought to work in several possible ways. Reneging on a promise or goal declared to a group of peers can be met with direct punishment, such as a financial penalty, or with indirect consequences, such as a tarnished reputation. Gathering information from peers about the best way to succeed at losing weight or overcoming a challenge may help people meet those goals. And seeing peers succeed may lead people to believe they are able to achieve a goal they previously thought unfeasible.
Meier, working with former Chilean planning minister Felipe Kast and Dina Pomeranz of Harvard Business School, found research partners in Banco Credichile, Microdatos, and Chilean microcredit organization Fondo Esperanza.
Microcredit borrowers offered an ideal group of subjects: they typically have a low income and already meet regularly in groups designed to encourage loan repayment. In the researchers’ first experiment, borrowers were split into three groups and asked a series of questions about their individual savings goals and offered the chance to open a savings account through Banco Credichile (which made all deposit transaction data from these savers available to the researchers). The control group continued to meet as usual, reporting back on loan repayment. The second group likewise continued to meet and report as usual, but individuals also reported on progress toward their stated savings goals. The third group continued meeting and reporting back on their loan repayments, but rather than reporting back on savings goals, the group members were offered higher-interest savings accounts 5 percent, compared with 0.3 percent for the savings accounts offered to the other groups.
At the one-year mark, subjects in the group that reported their savings progress back to their peers had a savings rate 100 percent higher than those in the control group. For most participants in the high-interest rate group, the savings rate was not higher than in the control group (although a small number put away large amounts of money, so that on average, the savings rate in the high-interest rate group was about 50 percent higher than in the control group).
Why did the self-help peer-saving group have a greater effect on increasing savings rates than a remarkably high interest rate did? “If reputation preservation is an important aspect of peer effects,” Meier says, “then someone needs to know whether or not I save. Or it could be that going to the meeting is literally just a reminder oh yeah, I wanted to save and has little to do with peers.”
The researchers tested these possible explanations by taking the meeting out of the meetings. They asked participants from the first experiment to set a new savings goal. As before, a control group continued to meet as usual, without discussing their progress toward those savings goals, over a three-month period. A second group was offered automated cell phone text message reports, linked to their savings accounts, that congratulated the recipient on saving that period or noted if they had made no deposits. A third group was offered the automated reports combined with a savings buddy a friend or relative chosen to also receive the text messages about the saver’s progress and support the savers toward their goal.
The savings buddy triggered an impressive increase in savings they deposited more money more often into savings than the control group but so did the group that received only the automated text messages from the bank. Text messaging has distinct advantages over self-help peer groups and face-to-face settings, Meier notes. “First, it’s scalable, and doesn’t require attendance at meetings. Second, treatment is the default users must unsubscribe to stop text reminders from coming.”
The conclusion: peer self-help groups may work not so much because of peer pressure, but because they serve as a simple, effective reminder to take an action. “You can increase savings or deal with other self-control problems in ways that are extremely scalable, even for the poorest of the poor,” Meier says. “There are limits. Getting 50 other text reminders about other issues probably wouldn’t work. But prioritizing a few of the most important can.”
Stephan Meier is the Regina Pitaro Associate Professor of Business in the Management Division and a senior scholar at the Jerome A. Chazen Institute of International Business at Columbia Business School.
Read the Research
"Under-Savers Anonymous: Evidence on Self-Help Groups and Peer Pressure as a Savings Commitment Device"