Train the Trainer: Promoting Savings by Training Banking Business Correspondent Agents in Andhra Pradesh

Investigadores/as:
Laura Litvine
Santadarshan Sadhu
Trabajo de campo:
Ubicación:
Andhra Pradesh, India
Muestra:
1,250 business correspondent agents in three districts for agent-level intervention, 1,000 customers for customer-level intervention
Línea de tiempo:
2013 - 2014
Grupo objetivo:
  • Adults
Resultado de interés:
  • Earnings and income
Tipo de intervención:
  • Digital and mobile
  • Financial literacy
  • Savings
Número de registro del AEA RCT Registry:
AEARCTR-0000969
Artículos de investigación:
Socios Implementadores:

Hundreds of millions of previously unbanked individuals in the developing world now have access to formal financial services, but low levels of financial literacy and limited access to qualified advice make it difficult for people to make the most of these financial services. In India, many people access banks through business correspondent agents, who are locally-based third party intermediaries who have direct and constant contact with end users, and therefore may be able to increase financial capability more than traditional financial literacy training programs. In this ongoing evaluation, researchers are testing the relative effect on knowledge and savings behaviors of training customers through these agents and giving customers financial information directly, using for this purpose an innovative mobile-based information platform.

Problema de política pública

Formal banking has expanded dramatically in developing countries in the past few decades. Five hundred to eight hundred million previously unbanked individuals now have access to financial services thanks to the government-supported expansion of traditional banks, or innovations such as mobile banking or business correspondents. However, evidence shows that levels of financial literacy are low in both developing and developed countries, making it difficult for many customers to understand and use the range of financial products and services available to them. Many organizations have offered financial education classes in an attempt to address this need, but researchers have found that these courses are often expensive, poorly attended, and unreliable in terms of quality. Attendees in turn gain a limited amount of financial knowledge and the classes have little long run impact on financial behavior.

While improving customers’ financial capabilities may help them save more and make the most of the financial services available to them, innovations are needed to overcome the high cost and low long-term impact of traditional training interventions. These interventions could leverage technology and characteristics unique to local contexts to reach the most underserved and marginalized segments of society. In this project, researchers look at two such innovations: training banking agents to provide financial information to customers and using mobile phones as a medium for communicating financial information.

Contexto de la evaluación

Andhra Pradesh, a state in southeastern India, has a highly developed network of business correspondent agents (BCAs). BCAs, who are all women in the context of this evaluation, are not employed by banks but act as intermediaries connecting low-income, last-mile customers with banking services. Customers can open accounts, receive government electronic benefits transfers, make deposits, withdraw money, and in some cases purchase other financial products such as insurance policies or pensions. Customers generally live in rural or semi-urban areas far from formal bank branches, and work in agriculture, construction, or fishing. As many of these individuals access banking services solely through BCAs, the agents are often their primary source of information about financial products and practices.

Detalles de la intervención

This intervention is conducted in partnership with FINO Fintech Foundation, one of the largest providers of business correspondent services in India, and involves different ways to encourage customers to save actively.

The evaluation has two main arms. First, researchers are studying a BCA-level intervention and testing the effect of training and incentivizing BCAs on the savings and financial capability of the customers they serve. They are comparing the effect of disseminating information through BCAs to simply incentivizing BCAs to encourage their customers to save. The 1,250 BCAs from three districts in the state were randomly assigned to five different treatment groups, which will allow the researchers to disentangle the effects of three program components:

  • Training session, no mobile platform, no incentive program (250 BCAs)

  • Training session, mobile platform, no incentive program (250 BCAs)

  • Training session, no mobile platform, incentive program (250 BCAs)

  • Training session, mobile platform, incentive program (250 BCAs)

  • Comparison group with no training (250 BCAs)

All BCAs who were not in the comparison group attended a general training session, where BCAs learned about a new type of savings account made available to their customers. In addition, some BCAs learned how to use a new mobile platform, which allows them to browse information about financial products and ask questions in a forum. Finally, some were enrolled in an incentive program, which entered top-performing BCAs in terms of savings achievements in a lottery for prizes.

In the second arm of this evaluation, researchers conducted a customer-level intervention with 1,000 individuals. Half of these customers were invited to use a financial helpline similar to the mobile platform used by the BCAs. This arm of the intervention will allow the researchers to compare the effect of providing information through a trained BCA with allowing customers to directly access information about financial services.

Resultados y lecciones de la política pública

The division of Andhra Pradesh into two states in 2014 created unanticipated delays and logistical challenges for both local governments and the operating partners. This led researchers to end the study early, well before outcome data had been collected.