Recent evidence suggests formal savings accounts can lead to increased savings for the poor, but uptake of bank accounts has been low. There is also evidence suggesting that savings for specific goals can potentially be increased by enabling people to commit money towards goal-specific “mental accounts”. In Malawi, Innovations for Poverty Action is working with researchers to investigate alternative ways to encourage saving by offering business owners various savings products, including single or multiple lockboxes or mobile money accounts. They are measuring the impacts of the products on entrepreneurs’ ability to reach savings goals, as well as on overall savings, business decisions and labor supply. 

Policy Issue 

Most people in developing countries are self-employed and must save for competing demands simultaneously, including household expenses, emergencies, business investment, and inventory restocking. Keeping track of saving and spending for these various “accounts” can be challenging. While offering people formal savings accounts seems like a promising solution, usage of these accounts is limited by high fees, limited operating hours, and limited branch locations. Instead, many entrepreneurs keep cash at home. However, money stored at home may be insecure and overly accessible, making it vulnerable to theft, overspending, or use by others in the household.

Mobile money makes banking more accessible at a substantially lower cost than conventional banking. Since the density of mobile money agents in this part of the world is significantly greater than that of banks, mobile money not only decreases the travel time and cost required to make a deposit or a withdrawal, but also the safety risk inherent in traveling with cash. This research investigates the impact of mobile money and lockboxes, and how providing people with multiple accounts for different savings goals may affect goal-specific and overall savings, business decisions, and labor supply. 

Context of the Evaluation 

This study takes place in Blantyre (the largest city in Malawi) among self-employed business owners including retailers, petty traders, and artisans.  Among this population, 42 percent have bank accounts. However, 70 percent of those with accounts withdraw money only once a month or less, suggesting that they do not use these accounts for regular expenses such as inventory restocking.

Researchers partnered with Airtel, one of the two largest mobile money providers in Malawi, to deliver this intervention. For research purposes, IPA has worked with Airtel to reimburse mobile money withdrawal fees (usually about 5 percent of the withdrawal value) on a weekly basis among those in the study sample.

Details of the Intervention 

Innovations for Poverty Action is working with researchers to evaluate  the impacts of various savings products (single or multiple mobile money accounts or lockboxes) on savings behavior, consumption, business decisions, and labor supply, among other outcomes. Specifically, they will compare the impacts of offering access to multiple lockboxes and mobile money accounts compared to one account as well as no account. Researchers will randomly assign 800 business owners to receive one of the following:

  1. Offer of one mobile money account
  2. Offer of up to two mobile money accounts
  3. Offer of one lockbox
  4. Offer of up to three lockboxes
  5. No savings device offer (comparison group)

The mobile money accounts are “as-is”, with the sole added feature of waived withdrawal fees.  Participants will also be trained on how to use the accounts.

Across each of the five groups, half will also be randomly assigned to receive logbooks to record daily income, expenditures, employment, investment, economic shocks (such as emergency medical care), usage of savings sources, and several other key outcomes. Logbooks will assist researchers in understanding how various savings mechanisms may affect business owners’ abilities to cope with uneven income, spending needs and employment. 

Results and Policy Lessons 

Project on-going; results forthcoming.