x

Please fill out this short user survey of only 3 questions in order to help us improve the site. We appreciate your feedback!
Evaluating the Impact of Commitment Savings Accounts Linked to Mobile Money
Last registered on February 12, 2019

Pre-Trial

Trial Information
General Information
Title
Evaluating the Impact of Commitment Savings Accounts Linked to Mobile Money
RCT ID
AEARCTR-0001904
Initial registration date
March 15, 2017
Last updated
February 12, 2019 2:15 PM EST
Location(s)

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

Request Information
Primary Investigator
Affiliation
World Bank
Other Primary Investigator(s)
PI Affiliation
World Bank
PI Affiliation
Dartmouth College
PI Affiliation
World Bank
PI Affiliation
World Bank
Additional Trial Information
Status
On going
Start date
2018-10-12
End date
2020-12-31
Secondary IDs
Abstract
This study is designed to increase understanding of how intra-household dynamics and gender impact savings. As part of this study, we offer mobile-phone based bank savings accounts that are linked to mobile money to married men and women in Ghana’s Volta region.

In order to answer our research questions, we plan to randomly vary three elements: (i) whether a liquidity restriction on the account is binding; (ii) spousal information regarding which liquidity restriction is in place; and (iii) the value of incentive payments for choosing a binding versus non-binding liquidity restrictions.

The first set of research questions will look at the demand curve for hard vs soft commitment; the correlation between baseline characteristics and demand for hard vs soft commitment; the extent to which the spousal information treatment impacts demand for hard vs soft commitment; and how all of these vary by gender. The second set of research questions will look at impacts on savings, including: how the spousal information treatment impacts savings; how the hard commitment treatment impacts savings; the interaction between the hard commitment treatment and the spousal information treatment; and the impact of perceived (but not actual) binding commitment from the perspective of the spouse.
External Link(s)
Registration Citation
Citation
Goldstein, Markus et al. 2019. "Evaluating the Impact of Commitment Savings Accounts Linked to Mobile Money." AEA RCT Registry. February 12. https://doi.org/10.1257/rct.1904-4.0.
Former Citation
Goldstein, Markus et al. 2019. "Evaluating the Impact of Commitment Savings Accounts Linked to Mobile Money." AEA RCT Registry. February 12. http://www.socialscienceregistry.org/trials/1904/history/41407.
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)
Study participants will be told about two forms of accounts: accounts with a hard, fixed, and mandatory withdrawal restriction feature (henceforth referred to as “hard commitment”), and accounts with a soft, flexible, and optional withdrawal restriction feature (henceforth referred to as “soft commitment”). We will randomize which account they are offered, and we will cross-randomize incentive payments and a spousal information treatment. After account activation we may also randomly switch some of the people in hard commitment accounts to soft commitment accounts.

Account offer and take-up: After study participants declare their preferred account (hard or soft commitment), we will randomize which account they are offered, with the randomization weighted by their declared preference. Then they will be offered the account to which they have been randomized and asked whether they wish to take-up this account. For half of the sample, this will be done in the presence of their spouse, and for the other half this will be done privately, such that their spouse cannot hear what account is offered nor whether they chose to take up the account. For those who wish to take-up the account, a team will be in place to immediately assist them with completing all steps associated with account opening and set-up.

Target-setting: The hard commitment group will be required to put in place a target during account opening. The target can either be an amount, or a date. When a target is in place, an account holder is unable to withdraw funds from the m-savings account until the target is achieved. The soft commitment group will be offered the opportunity to put a target in place during account opening but will not be forced to set a target. And the soft commitment group, in contrast to the hard commitment group, will have the capability to remove a target after it has been set but before it has been achieved, and can thus cancel the liquidity restriction at any time.
Intervention Start Date
2018-10-12
Intervention End Date
2020-12-31
Primary Outcomes
Primary Outcomes (end points)
Account choice; account take-up; savings behavior; goal attainment.
Primary Outcomes (explanation)
Account choice: Prior to being offered an account, which account type would the study participant prefer to be offered? A hard commitment or a soft commitment account? Varied financial incentives for selecting one vs the other will allow us to map a demand curve.

Account take-up: After being offered an account, does the study participant proceed with opening that account, and complete account opening?

Savings behavior: deposits and withdrawals into mobile money and mobile-money based savings accounts, other cash savings, and non-cash savings, and how savings allocation across different places shifts in response to treatment.

Goal attainment: The baseline survey will include a goal setting exercise. Then, at the time of the account set-up, the goal expressed during baseline will be used to prime the study participant to set up a liquidity restriction. This outcome will capture whether the study participant achieved their goal, irrespective of target set within the m-savings product.
Secondary Outcomes
Secondary Outcomes (end points)
Secondary Outcomes (explanation)
Experimental Design
Experimental Design
As part of this study, we offer mobile-phone based bank savings accounts that are linked to mobile money to married men and women in Ghana’s Volta region.

In order to answer our research questions, we plan to randomly vary three elements: (i) the liquidity restriction; (ii) spousal information regarding which liquidity restriction is in place; and (iii) value of incentive payments for choosing different levels of liquidity restrictions.

The first set of research questions will look at the demand curve for hard vs soft commitment; the correlation between baseline characteristics and demand for hard vs soft commitment; the extent to which the spousal information treatment impacts demand for hard vs soft commitment; and how all of these vary by gender. One of the baseline characteristics that we will use for this heterogeneity analysis is spousal alignment on spending and savings priorities (well versus poorly aligned). The second set of research questions will look at impacts on savings, including: how the spousal information treatment impacts savings; how the hard commitment treatment impacts savings; the interaction between the hard commitment treatment and the spousal information treatment; and the impact of plausible (but not necessarily actual) commitment.

The study sample will be comprised of 1,000 married couples (for a total of 1,000 men and 1,000 women) – or more, if budget constraints allow for it. All 2,000 individuals must be age 18 or older; must be in a monogamous marriage or union and live together with their spouse and partner; and must have a valid ID. Sample creation will begin with a listing of people interested in the product. The spouse who initiates the couple’s entry into the sample by participating in the listing (we will refer to this person as the “primary spouse”) must also must also already have their own MTN sim card that is registered in their own name, and that is registered for MTN mobile money – as both of these are pre-requisites for activating the m-savings product. The study sample will be recruited from the communities near partner bank branches. The branches are located in Nkwanta, Krachi, Kadjebi, and Jassikan Districts within Ghana’s Volta Region.

We will begin by conducting public outreach sessions in communities surrounding our partner bank’s branches. The outreach sessions will address in general terms the features of the mobile-savings accounts. During these, as well as other mobilization efforts, interested individuals will be screened for eligibility, and their contact information will be recorded. Eligible, interested individuals will then be scheduled for an appointment during which they and their spouse will participate in an information session about the products, will both complete baseline surveys, and will both be offered m-savings accounts.

All randomizations will be stratified by gender, community, and whether or not the primary spouse has a pre-existing account with a formal bank.
Experimental Design Details
Not available
Randomization Method
Randomization done in advance by computer in office for spousal information treatment, and opening bonus incentive levels.

Randomization done on the spot by tablet for commitment type (hard or soft)
Randomization Unit
Individual person level - commitment type (hard or soft) and opening bonus incentive levels

Couple level - spousal information treatment
Was the treatment clustered?
No
Experiment Characteristics
Sample size: planned number of clusters
N/A
Sample size: planned number of observations
The study sample is anticipated to include approximately 1,000 couples, for a total of approximately 2,000 individuals (1,000 women, and 1,000 men).
Sample size (or number of clusters) by treatment arms
We will have 3 main experimental cells: (i) Hard commitment; (ii) Soft commitment; (iii) Hard -> Soft commitment. Within each of these cells we will have approximately 666 individuals.

And within each treatment cell, we will also randomly vary whether a study participants spouse receives an information treatment (with half receiving, and half not). This will thus result in a total of 6 treatment cells, each with approximately 333 individuals.
Minimum detectable effect size for main outcomes (accounting for sample design and clustering)
Our power calculations target 80 percent power using a two-sided test of size 0.05. For the information treatment we assume the within-couple correlation in outcomes is 0.60, and that attrition from baseline to endline is 0.10 (this is only relevant for outcomes that use survey data). We will have power to detect a 0.16-0.18 standardized effect size of the information treatment on demand for commitment/savings outcomes, depending on whether genders are pooled or analyzed separately. This translates into a 7-8 percentage point effect on demand for the hard commitment account, assuming 30 percent of individuals prefer hard commitment in private. For the account type treatments we will be able to detect a 0.22 standardized effect size on savings outcomes when comparing one arm to another: e.g., hard versus soft commitment (this assumes we pool information arms, limit attention to one gender, and that one third of the sample prefers soft to hard commitment). This amounts to a 10 percentage point increase in monthly active use of the account, relative to a 30 percent mean in the soft commitment group that we observed in our pilot.
IRB
INSTITUTIONAL REVIEW BOARDS (IRBs)
IRB Name
University of Ghana: Office of research, innovation and development: ethics committee for humanities
IRB Approval Date
2015-05-12
IRB Approval Number
ECH 085/14-15