The Use of Self-Help Groups as a Savings Commitment Device in Chile

Stephan Meier
Felipe Kast
2,687 micro-entrepreneurs
2007 - 2009
Target group:
  • Entrepreneurs
Outcome of interest:
  • Savings/deposits
Intervention type:
  • Nudges and reminders
  • Savings
  • Social networks

Self-help peer groups are thought to be an effective way to encourage savings, but it is possible that the same dynamics could be simulated by more convenient programs, such as text-message feedback. Researchers in Chile compared the effects of self-help peer groups and text-message feedback on entrepreneurs’ ability to make regular deposits into a savings account. Receiving feedback by text message increased savings by almost as much as being a member of a self-help group, suggesting that the physical aspect of self-help groups may not be as important as previously thought.

Policy issue

The ability to save is important to allow individuals to maintain their consumption levels when faced with fluctuations in their income or expenditure needs, such as business downturns or health emergencies, or to help them save towards a large investment or expenditure, such as a new tool or school supplies. However, many people may find it difficult to save regularly due to self-control problems and lack of access to formal financial services. Self-help peer groups are often used to hold people accountable for reaching certain goals such as quitting smoking or losing weight. This study investigates the effect of self-help peer groups and of feedback by text message on the ability to make regular deposits into a savings account. This type of deposit commitment device can be particularly important for independent entrepreneurs and those who work in the informal sector and therefore often do not have access to mechanisms that can help them save, such as direct deposits of salaries into a savings account.

If self-help peer groups are an effective way to encourage savings, it is possible that certain dynamics of self-help peer groups can be simulated through other means, such as text messages, for example. The use of cell phones and text messaging has increased dramatically in recent years throughout the developing world, and mobile phones have become a popular delivery mechanism for many services to the poor. Since text messages do not require as much coordination and time investment as group meetings, they may be able to reach a larger population in a more convenient way. Knowing which of the many elements of self-help peer group meetings are required for their effectiveness can provide insight for designing alternative policies to scale up the benefits of self-help peer groups for savings.

Context of the evaluation

Fondo Esperanza (FE) is a microfinance institution that provides micro-credit and support to entrepreneurs in Chile. Members of FE are mostly urban micro-entrepreneurs, who receive small loans to invest in their businesses and meet regularly in groups of 10-20 to repay their loans. Individuals who participated in this study had on average ten years of education, and around two-thirds of them did not have a savings account. In focus groups conducted before the intervention, FE members expressed interest in increasing their level of precautionary savings for emergencies, and 68 percent said they frequently regretted not having saved more.

Details of the intervention

Researchers conducted two related experiments to assess the impact of self-help peer groups and feedback text messages as a commitment device to help people save.

Peer Group Experiment:

In the first experiment, 196 FE groups, comprising 2,687 members, were randomly assigned to receive one of three types of savings accounts.

  1. Basic Savings Account: Depositors received a real interest rate of 0.3 percent, comparable to the highest available alternative in the Chilean market. The interest rate was guaranteed for two years, and savings could be withdrawn at any point.

  2. Self-Help Peer Group Account: In addition to the basic savings account, depositors had the option of announcing a weekly savings goal for the next three months to the group, and the peers monitored in the meetings who was meeting their goal.

  3. High Interest Rate Account: Depositors received a real interest rate of 5 percent, which was significantly higher than other options available in the market. After the account was introduced, an FE leader illustrated the potential returns and compounded interest rates during a one-hour workshop.

Feedback Message Experiment:

One year after savings accounts were opened, researchers introduced a second experiment to shed light on the mechanisms through which self-help groups affected savings, and to test whether follow-up and feedback through text messages could achieve similar impacts without requiring in-person meetings. 873 participants who had opened an account during the first experiment and owned a cell-phone were randomly assigned to one of three groups:

  1. Peer Pressure Treatment: Participants set a weekly savings goal for themselves and chose a "savings buddy" to monitor their savings. Both the participant and their buddy received weekly text messages informing them whether the participant had made their weekly deposit.

  2. Peer Information Treatment: Participants set a weekly savings goal for themselves and received weekly text messages informing them whether they fulfilled their weekly commitment, and what proportion of other participants had made their deposit.

  3. Comparison Group: Participants were asked to set a weekly savings goal for themselves, but did not receive any text message feedback.

Results and policy lessons

Impact of Self-Help Peer Groups: Those assigned to the self-help peer groups deposited money into the savings accounts 3.5 times as often and saved almost twice as much as those in the comparison group. The increased number of deposits was not offset by a corresponding increase in withdrawals, even though participants were allowed to withdraw money at any point during the program. Raising interest rates, in contrast, did not increase the number of deposits or the amount deposited for the vast majority of participants, suggesting that in certain contexts, social nudges can be more effective than financial rewards.

Impact of Text Message Feedback: By simulating some of the functions of self-help peer groups, the text message feedback program was able to separate the effect of physical meetings from the effect of regular feedback on savings behavior. Those who received feedback text messages made almost three times as many deposits and saved about 7,400 pesos (approximately eight percent of monthly per capita income) more on average, during the three months following the introduction of the feedback messages. Holding people accountable through weekly feedback text messages increased savings almost as much as self-help peer groups, suggesting that physical meetings may not be as central to the effect of self-help peer groups as previously thought. While both types of feedback text messages increased savings, researchers did not find significant differences between the two text message treatments, suggesting that text messages are effective even without a savings buddy and that peer pressure is not required.

The overall results indicate that regular feedback and follow-up could be more important to the success of self-help groups than the peers themselves. In contexts where cellphones are readily available, text message feedback programs are also likely to be more easily scaled, since they require minimal coordination and do not rely on physical proximity.

Kast, Felipe, Stephan Meier, Dina Pomeranz. 2018. "Savings More in Groups: Field Experimental Evidence from Chile." Journal of Development Economics, 133: 275-294.