Social Network Formation among Female Marriage Migrants

1. Inviter Treatment (n=600) – This group will view the Sharing List and can invite up to five women on the list to share a meal (invitation delivered by enumerators).
2. Vouchers + Inviter Treatment (n=600) – Recipients will receive 7 USD (8% median monthly income) vouchers, redeemable at several local vendors selling perishable produce, and the Inviter Treatment.
3. Control (n=400) – This group will not receive any intervention.

Within treatments 1 and 2, participants will view one of three types of Sharing Lists: within-SES (all Invitees are in the bottom 80% of an SES index), cross-SES (all Invitees are in the top 20% of an SES index), or a random selection of women (for each of these three groups, n=200 in treatment 1, and n=200 in treatment 2).

2024-07-14

2024-08-14

1. Psychological well-being index: UCLA-3 score, CESDR-10 score, "Is there someone you can speak to if you have a quarrel with your husband?"
2. Social capital index: of number of relative social network connections, number of non-relative social network connections, and average strength of these relationships (measured by the inclusion of self in others score, ability to trust, laugh with, and confide in network connection)
1. Savings and consumption index: maize stores, meals eaten yesterday, meals eaten day before yesterday, number of animal protein meals in past month

All indexes will be constructed by standardizing each variable in the index, taking the average of those variables, and then standardizing the index in terms of standard deviations from the control mean. I will also construct two other indexes for robustness: the Anderson Inverse Covariance weighted and first component of the variables using Principal Components Analysis

1. Female empowerment index: autonomy over decision-making about children's health, children's schooling, disciplining children, family planning, and household savings.
2. Labor market outcomes: methods of income-earning, income earned in past month
3. Risk-sharing outcomes: number of people lent to and borrowed from in past month, number of people feels comfortable to lend to or borrow from

All indexes will be constructed by standardizing each variable in the index, taking the average of those variables, and then standardizing the index in terms of standard deviations from the control mean. I will also construct two other indexes for robustness: the Anderson Inverse Covariance weighted and first component of the variables using Principal Components Analysis

My RCT will include two intervention arms that each attempt to solve barriers to inviting somebody to share a meal: a food price subsidy to solve budgetary concerns, and information about who else in the village would like to share meals to solve social concerns, such as fear that an invitation might be rejected. I will first recruit a sample of 3,600 women who moved to the village after the age of 14 but no more than 20 years ago (Baseline Sample). I will select a 1,600-person subset of the Baseline Sample for the intervention (Intervention Sample), selected if they are in the bottom 80% of an SES index (proxy for low-income) since budgets are more likely to constrain low-income women. I will offer 1,000 Baseline Sample participants who are not selected for the Intervention Sample the opportunity to be an Invitee, who add their name to a list of women who are interested in sharing meals with other women in the village more often (Sharing List). The rest of the study will be conducted with the Intervention Sample, who will be randomized into the Inviter, Inviter + Voucher, or Control treatment arms.

Data:
Baseline Survey: collected with all 2600 participants
Treatment Allocation Survey (one month after Baseline): collected with the Intervention Sample
Endline Survey (one month after treatment): collected with the Intervention Sample
Food Diaries: Intervention Sample participants will fill out a pen-and-paper food diary, where they can mark from a list of food options what they consume each day over the course of two weeks. They can also mark if each meal was shared on the same diary.

Not available

500 women in the bottom 80% of an SES index will be randomly assigned to the Invitee treatment; and 500 women in the top 20% of the SES index will be randomly assigned to the Invitee treatment
Women in the bottom 80% of an SES index will be randomized into Inviter, Inviter + Voucher, Invitee, or Control via randomization in office by a computer
The remaining women will also be in the Control group, but will not be interviewed in the Treatment Allocation survey, or one-month follow-up Endling.

Geographic clusters will be randomized to high- or low-treatment saturation. Geographic clusters are determined by k-means clustering
Treatment assignment will be randomized at the individual level

No

3600

3600

Invitees: 1000
Control: 400
Inviters + Vouchers -- Within-SES Invitees: 200
Inviters w/o Vouchers -- Within-SES Invitees: 200

Control group without follow-up: 1,000
Inviters + Vouchers -- Cross-SES Invitees: 200
Inviters w/o Vouchers -- Cross-SES Invitees: 200
Inviters + Vouchers -- Random Invitees: 200
Inviters w/o Vouchers -- Random Invitees: 200