x

We are happy to announce that all trial registrations will now be issued DOIs (digital object identifiers). For more information, see here.
“Train the trainer”: Promoting savings by training Banking Correspondent Agents in Andhra Pradesh
Last registered on December 16, 2015

Pre-Trial

Trial Information
General Information
Title
“Train the trainer”: Promoting savings by training Banking Correspondent Agents in Andhra Pradesh
RCT ID
AEARCTR-0000969
Initial registration date
December 16, 2015
Last updated
December 16, 2015 11:51 AM EST
Location(s)
Primary Investigator
Affiliation
HBS
Other Primary Investigator(s)
PI Affiliation
Harvard Business School
PI Affiliation
University College London
Additional Trial Information
Status
Withdrawn
Start date
2013-10-21
End date
2014-10-21
Secondary IDs
Abstract
Studies show that 500 to 800 million of the world’s poor have gained access to finance in the last few decades (Deb and Kubzansky, 2012). However, evidence indicates that many of these individuals lack the skills to interact with the increasingly complex financial products and services. In this study, we plan to conduct an intervention that includes (1) training and incentivizing Banking Correspondent Agents, and (2) providing financial information via mobile phone to customers in Andhra Pradesh, India. In so doing, we aim to test whether we see changes in financial knowledge, habits, and outcomes both among the BCAs and among the customers. This study was abandoned in 2014 due the unanticipated separation of the state of Andhra Pradesh into two states. Due to this seperation, the integrity of our randomization could not remain intact, nor were our partner organizations willing to proceed.
External Link(s)
Registration Citation
Citation
Cole, Shawn, Shawn Cole and Laura Litvine. 2015. "“Train the trainer”: Promoting savings by training Banking Correspondent Agents in Andhra Pradesh." AEA RCT Registry. December 16. https://doi.org/10.1257/rct.969-1.0.
Former Citation
Cole, Shawn, Shawn Cole and Laura Litvine. 2015. "“Train the trainer”: Promoting savings by training Banking Correspondent Agents in Andhra Pradesh." AEA RCT Registry. December 16. https://www.socialscienceregistry.org/trials/969/history/6329.
Sponsors & Partners

There are documents in this trial unavailable to the public. Use the button below to request access to this information.

Request Information
Experimental Details
Interventions
Intervention(s)
FINO, a business and technology platform catering to “last mile” customers through Business Correspondent Agents (BCAs), has a bundled product linking a No-Frills Savings Account to a beneficiary’s EBT account. This product could indeed offer a powerful tool in expanding effective access to finance, as it combines an easily accessible savings vehicle with frequent interactions with the formal banking system. However, this product alone does not address the issue of low average financial capability of low-income populations, and it is probable that one would see the history of dormancy of NFAs repeat itself with no intervention being targeted at encouraging usage of financial products.
Motivated by other successful examples we will therefore encourage customers holding an EBT-NFA account to divert some benefits payments to build savings and become more “financially active”. While this will still be done through financial literacy campaigns and incentives schemes, we intend to contribute to the financial capability space by addressing some of the major constraints of typical interventions, namely their high cost and low long-run effects, by studying the efficiency of two innovations: (1) training and incentivizing BCAs, and (2) providing financial information via mobile phone.
Intervention Start Date
2013-12-30
Intervention End Date
2014-01-31
Primary Outcomes
Primary Outcomes (end points)
Financial knowledge, habits, outcomes
(in particular, levels of saving)
Primary Outcomes (explanation)
Secondary Outcomes
Secondary Outcomes (end points)
Secondary Outcomes (explanation)
Experimental Design
Experimental Design
We assess the impact of the intervention described above by conducting a clustered randomized controlled trial on a sample drawn from FINO’s pilot client base in Andhra Pradesh, India.
Our study has two main arms: On one hand we provide training, incentives and access to a mobile platform we have developed to selected BCAs during group training sessions conducted at the block level; and on the other hand, we invite a subsample of end-customers to use a similar financial information mobile platform directly. We describe each of these arms successively.
1: BCA-level intervention
FINO’s pilot happened in 117 blocks of 4 districts of Andhra Pradesh, and includes 2600+ BCAs. In the first arm of this study, we randomly select and assign, at the block level, 1000 BCAs to one of five treatment statuses:
1. 16% of blocks are assigned to receiving no information
2. Sampled BCAs in 21% of blocks are assigned to receiving information about the new EBT-NFA product only
3. Sampled BCAs in 21% of blocks are assigned to receiving information about the new EBT-NFA product, plus are incentivized based on the savings achievements of their customers
4. Sampled BCAs in 21% of blocks are assigned to receiving information about the new EBT-NFA product, plus are given access to the financial information mobile platform and in-person training on how to encourage one’s customers to save
5. Sampled BCAs in 21% of blocks are assigned to receiving information about the new EBT-NFA product, given access to the mobile platform and training on how to encourage savings, and incentivized based on the savings achievements of their customers.

BCAs located in blocks assigned to group 1, or non-sampled BCAs of blocks assigned to groups 2 to 5 will not be contacted, and no identifiable information will be obtained from them. We will only use administrative data provided by FINO including unidentifiable transaction data from their customers, and exploit average variation in savings achievements between groups.
The 1000 sampled BCAs assigned to groups 2-5 will first be invited by phone to participate in a training session of the same format as the training sessions usually held by FINO: They will be called up by our team and informed that they have been randomly selected to take part in a pilot conducted jointly by FINO and CMF.
BCAs who agree to participate will then be informed of the time and place of the session to be held; all sessions will be held within the block the BCA lives in, conducted jointly by CMF staff and FINO trainers, and transportation and food for the day will be paid for, even if participants leave before the end of the day. Sessions will include 6-12 participants depending on the block, and will take place on a single day, from 9.30am to 6pm at the most. Participants completing the 1h-long end-of-day survey will be compensated with INR 30 (~USD 0.50), which corresponds to approximately the income from 2 hours work of a BCA.
BCAs will then be informed of the features of the NFA-EBT bundled account, and taught how to practically provide these new services. BCAs assigned to groups 3 and 5 will be provided with explanations on a new incentive scheme applying to this pilot product and consisting of lotteries for gifts that they will be able to enter depending on the savings achievements of their customers (number of active accounts, average balance, etc.). The gifts will be adapted to the local population, and not exceed a value of INR 3000 (~USD 50).
In addition to the basic training on the product, groups 4 and 5 will be trained on the benefits of savings and how to encourage people to save, using for this purpose a module developed and piloted by CMF in the course of a previous study. They will also be provided with access to and training on how to use the mobile platform we have developed with our technology partner, Awaaz.De. This platform, a voice-based interactive on-demand product, includes: (1) a browsing system to access information about financial products and practices, (2) a Q&A forum where they are able to ask any question to financial experts, and (3) the possibility to comment on or grade the messages delivered. It is also enabled to push information, either through recorded voice messages or SMS. The messages to be sent are developed by our financial experts, and carry important general financial information, as well as answers to relevant questions posted on the Q&A forum. Messages will be pushed no more than once a week, and participants will be informed that they are free to opt out of push calls. Beyond the cost of owning a mobile phone, the only user-cost of the platform is airtime, which participants are informed will be covered by the CMF-FINO team for the length of the study.
Finally, the session participants from all groups will be asked to (1) complete a self-assessed financial knowledge test, which lasts approximately 30 minutes and includes basic mathematics and financial literacy questions, and (2) answer a survey conducted individually and in-person by trained CMF surveyors. The survey questionnaire lasts approximately 45 minutes, and includes questions about basic demographics, financial habits, experience as a BCA, post-training financial knowledge, mobile usage and sources of financial information.
BCAs from all 5 groups will then be observed for approximately 10 months. During this period, sampled BCAs will be contacted up to twice for follow-up phone surveys of maximum 20 minutes conducted by CMF surveyors. During these follow-up surveys they will be asked questions about their practice of the new EBT-NFA product, financial habits, sources of information and financial knowledge, as well as questions directly related to the platform (for those having access to it). Participants completing the full phone surveys will be compensated with INR 20 (USD 0.30) given in the form of immediate mobile top-ups.
2. Customer-level intervention
In the second arm of this project, we select 1000 customers of the sampled BCAs, of which 50% are directly invited to use a financial helpline similar to the one described above, and the remaining 50% serve as control and receive no additional information. In order to ensure comparability between these two groups, the 1000 customers are selected so that they all own a mobile phone.
During a face-to-face meeting with a CMF surveyor, we first explain to sampled customers that they have been randomly selected to participate in a study which is fully voluntary, and present all study features to them, including duration, compensation, and the fact that they might win free access to an innovative information platform for the length of the study. We emphasize that participation in this study implies accepting that researchers will access individual bank account data provided by FINO for the duration of the study, while of course making sure to separate all identifiable data from the rest of the data and securely storing the identified key code on IFMR’s cold-room computers and unidentified data on IFMR’s servers. All customers who agree to participate are then asked to answer a short in-person survey conducted by CMF surveyors, which includes baseline questions about socio-demographic characteristics, financial knowledge and habits, needs and sources of information and mobile usage. It should not last more than 30 minutes. Participants who complete the survey are compensated with INR 20 (USD 0.35) in immediate mobile top-ups.
Once the survey is completed participants are provided with a card to scratch off in order to reveal whether or not they have won free access to the new mobile platform for the duration of the study. Participants assigned to the treatment group (those who won) are then given a description and training on how to use the mobile platform, during which an emphasis will be put on the fact that access to the platform will be free for the duration of the study.
Participants are also informed during this initial visit that they will be contacted once more for a follow-up phone survey where similar questions will be asked, and identical data security precautions will be applied. Participants completing the follow-up phone survey will be compensated with INR 20 (USD 0.30) given in the form of mobile top-ups and sent immediately to their phone.
Experimental Design Details
Randomization Method
Randomization in office by computer.
Randomization Unit
Randomization at the Business Correspondent Agent Group level
BCA groups assigned to one of four treatment groups
Was the treatment clustered?
Yes
Experiment Characteristics
Sample size: planned number of clusters
N/A (trial abandoned)
Sample size: planned number of observations
500 treatment end customers and 500 control end customers
Sample size (or number of clusters) by treatment arms
N/A (trial abandoned)
Minimum detectable effect size for main outcomes (accounting for sample design and clustering)
N/A
IRB
INSTITUTIONAL REVIEW BOARDS (IRBs)
IRB Name
Harvard University- Committee On the Use of Human Subjects in Research
IRB Approval Date
2013-10-21
IRB Approval Number
IRB13-2146
Post-Trial
Post Trial Information
Study Withdrawal

This section is unavailable to the public. Use the button below to request access to this information.

Request Information
Intervention
Is the intervention completed?
No
Is data collection complete?
No
Data Publication
Data Publication
Is public data available?
No
Program Files
Program Files
Reports and Papers
Preliminary Reports
Relevant Papers