Personalizing or Reminding? How to better incentivize savings amongst low-income individuals
We use a randomized control trial to evaluate why text message (SMS) reminders may have worked in previous studies to incentivize savings. Using a sample of poor to middle-income Paraguayans, we measure whether SMSs impact savings because they act as reminders or because they are personalized to the target specified by participants. We find that salience and personalization on their own help little (and sometimes even reduce) savings behavior and only their combination significantly increases savings target and amounts, although the effects are relatively short-lived and through informal savings. Salience appears to be particularly detrimental for more sophisticated agents. Personalization reduces savings for those who are shown to have not been able to save previously. Combining both appears key, particularly for agents that are more financially experienced, suggesting that simple SMSs are unlikely to be successful on their own in this population.