Goal Setting for Remittances

The goal setting intervention is divided into two parts. Part A is an additional module in the baseline survey (for both migrants and recipients) that facilitates individual goal setting, and Part B is an extension of this activity that takes place in a preferred group chat platform (either WhatsApp or a messaging application popular in Bangladesh called imo). Both sender and recipient will be invited to the group chat, which will encourage goal sharing and joint goal setting. This goal setting intervention is provided only to the recipient/migrant pairs that are assigned to the treatment group.

We will also send monthly follow-up messages to the goal setting chats over the course of the intervention period to remind participants of their goals and to provide encouragement to keep saving.

The goal setting intervention is divided into two parts. Part A is an additional module in the baseline survey that facilitates individual goal setting, and Part B is an extension of this activity that takes place in a preferred group chat platform (either WhatsApp or a messaging application popular in Bangladesh called imo) and encourages goal sharing. This goal setting intervention is provided only to the recipient/migrant pairs that are assigned to the treatment group.

Part A.

Directly following the migrant and recipient baseline survey, enumerators will conduct a 20-minute individual goal setting intervention.

Steps

1. Following a brief introductory script, enumerators introduce the importance of goal setting and openly communicating these goals with people you share finances with. They stress the following points.

- Goal setting refers to the development of an action plan designed to motivate and guide an individual—or groups of individuals—towards a goal.
- Research in psychology shows that goal setting can help people overcome mental barriers to making better financial decisions.
- Goal setting can help people avoid impulse consumption and build towards important investments.

Then, the enumerator asks if the respondent is comfortable sharing their goals with their counterpart through a group chat. If the respondent does not agree, the enumerator still proceeds with Part A, but does not create the group chat at the end. If the respondent agrees, the enumerator conducts Part A and creates the group chat immediately afterwards.

2. Enumerators begin the goal setting exercise. They ask respondents about their priorities for how the household should spend remittance income. Each is asked to rank given expenditure categories, such as education, medical, business, etc., in order of importance.

3. Once the respondent decides on the most important category, the enumerator asks them to think of a specific investment in this category that they would like to have four months in the future. If the respondent is having trouble thinking of a concrete investment, the enumerator provides examples of possible investment goals, e.g., paying for a child's private school tuition, purchasing fertilizer, or paying off a debt.

4. After a single investment is chosen, the enumerator asks the respondent to estimate the total cost X. Given an intervention period of four months, the enumerator says that if they expect to reach the chosen target on time, they would need to save $X per month. Since the enumerator will also have access to the household's estimated monthly income as reported in the baseline survey, they can use this information to ensure that the monthly savings target represents a reasonable goal for the household. Reasonable goals should not be so difficult that they lead to frustration, but ambitious enough that they produce meaningful investments. For this reason, enumerators will encourage savings targets between 10% and 30% of the household's total monthly income. If the chosen goal falls outside of this range, the enumerator will encourage participants to choose a slightly larger/smaller investment goal, though they are not required to do so if they feel that the chosen goal is appropriate. If the respondent agrees to revise their target amount, steps 2 and 3 are repeated.

5. Once both a long-term investment goal and a proximal (monthly) savings amount have been established, the enumerator summarizes what has been agreed upon so far (while also recording all this information for their own records).

6. This is the end of the goal setting activity. Respondents are encouraged to record their goal information for their own records, but they will also receive a text message that contains the same essential information. The enumerator thanks the respondent for their participation, asks if they have any additional questions, and explains that we will be following up within the next few days.

7. The group chat is created on a shared messaging app, and both recipient and migrant are invited to join the group. Below we describe how this group chat will be used to share goal information between migrants and recipients.

Part B.

A few days after the baseline survey and goal setting activity are conducted for the recipient/migrant pair, the enumerator will follow up via the previously created group chat. The purpose of the chat is to communicate each respondent's individual goal to their remittance counterpart, and to facilitate agreement on a shared goal. In the group chat, the enumerator shares both the migrant and recipient's savings goal (if consent to share was given), and asks the respondents to choose which goal they want to purse as their joint goal. They can also choose to pursue both goals simultaneously or create a new shared goal together. If they choose to create a new goal, the steps outlined above are repeated via text within the group chat. Whether the respondents choose a new or existing goal, the enumerator will then summarize the details of the finalized joint goal.

Follow-up and Monitoring:

We will follow-up with each created group chat four times (once a month for the four-month intervention period). These follow-up messages will contain a reminder about the chosen goal and target amount, a note about how much the respondents should have already saved to be on track, and a message of encouragement (even if they have fallen behind their target). According to existing goal setting literature, the most effective kinds of goal feedback contain both encouraging messaging and informative progress updates. At the end of each follow-up message, the enumerator will also remind the participant that the goal target date is in approximately 4 months, but they will not give any reminders about the endline survey, as it is important to maintain the impression of independence between the survey team and the goal setting team. For control respondents, enumerators will also send monthly text messages, but these messages will simply contain generic text reminders about the expected date of the endline survey (but make no mention of goal setting, savings, or remittances).

2023-05-09

2024-03-02

Savings (from administrative data)

The primary outcome variable will be an objective measure of savings from BRAC Bank administrative data. The primary sample is the 1,017 matched pairs of remittance recipients and migrants for whom we have a verified BRAC bank account number. Our outcome is the maximum weekly balance over the entire 4-month intervention period between administration of the baseline and endline survey. We use the maximum rather than the final balance or average because the goal-setting intervention encourages treated individuals to save towards a spending target. Using average or final balance could penalize respondents that invest in their goal before the end of the intervention period, possibly leading to underestimation of treatment effects.

Savings (from survey data), Long-term Savings (from administrative data), Remittances

Subjective Savings Measure: Since the administrative data does not cover the entire sample, and only captures one type of savings instrument (BRAC Bank account holdings), we will also ask a comprehensive set of savings questions to both migrant and recipient during the endline survey. As a secondary savings outcome, we will use the self-reported value of combined cash savings and bank balances (a stock measure) at the time of the endline survey. Within each sender-recipient pair, we will take the sum of this reported value for the migrant and recipient. Using this outcome, we can estimate treatment effects for the full matched sample of 1,556 matched pairs.

Long(er)-term Savings: Though active implementation and monitoring of the goal setting intervention will end after 4 months, we will continue to have access to the administrative data beyond this point. As such, we can also estimate effects on savings over a longer period to see if individuals sustain their savings behavior even after the initial goal deadline has passed. To do this, we will use the maximum weekly balance from each recipient's bank account, constructed analogously to our primary outcome, but over a period of 6 months from baseline rather than 4. As with our primary outcome, this long-term effected can only be estimated for the 1,017 pairs with verified BRAC Bank accounts.

Remittances: In addition to testing whether joint goal setting increases the portion of income that respondents allocate to savings, we will also estimate its effect on the total amount of remittances sent by migrants. For this secondary research question, we take use as an outcome the total amount of remittances sent by the migrant to the recipient over the 4-month intervention period, as reported by the migrant at endline. We will use sender---rather than recipient---reports, as these have been shown to be more accurate.

Our intervention will be evaluated through a randomized controlled trial. In this study, there is just a single treatment group, into which half of the 1,017 sender-recipient pairs will be randomly assigned. After an implementation period of 4 months, average outcomes in the treated group will be compared to average outcomes in the control group to establish an estimate of the treatment effect. Outcome data will be drawn both from an endline survey conducted 4 months after enrollment, as well as administrative bank data provided directly by BRAC Bank.

Randomization will occur at the individual level during administration of the baseline survey. Each remittance recipient that meets basic eligibility criteria and consents to the study will be automatically assigned to either the treatment or control group by survey software, with an equal probability of being assigned to either group. Remittance senders are always assigned to the same experimental group as their recipient counterparts. Since we will not have a definite list of potential participants, randomization will be conducted in real time on a rolling basis. Therefore, no stratification criteria will be used in the randomization algorithm.

Individual

No

70 Agent Banking Locations (BRAC Bank branch with limited services)

1,017 remittance sender-recipient pairs

492 pairs in the treatment group and 525 in the control group