Hold the Phone? Social and Economic Implications of Improving Women's Agency and Control over Smartphones

1. Mobile phone training: Every woman who attended a camp was offered a basic digital literacy training. The training covered dialing a number, receiving an incoming call, taking photos, and using Google’s voice search feature. Women were trained in groups of 8-10.

2. Mor Awaaz: This is a phone service for women designed by the research team. It offers a use case that is in line with the prevailing norm that women should care for the household. Mor Awaaz sends women weekly push calls with pre-recorded information on good health practices, government health services, and updates on India’s workfare program. The service also includes monthly “pull calls”, where trained enumerators call women to conduct a short survey to measure women's phone usage and their awareness of, and access to, public health and nutrition-linked services.

3. Incentives: Incentives are linked to pull calls. Women earn INR 100 credit towards in-kind incentives for each pull call they complete. This credit can be redeemed for common consumables such as tea, lentils, and cooking oil.

4. High Saturation: Villages were randomly assigned to receive either a "high" or "low" saturation version of Mor Awaaz. In low saturation villages only 36 women per villages were eligible to enroll in the service. In high saturation villages the vast majority of women in the village were invited to enroll.

2019-07-30

2020-11-30

Our analysis will focus on the following family of primary outcomes.

1. Women's Phone Ownership and Use: Here we will rely on both self-reported measures (ownership, frequency and types of use) as well as objective measures (direct pick up on supplementary calls, participation in supplementary calls). We will also use administrative data from Mor Awaaz to study impacts of saturation and incentives among Mor Awaaz enrollees.

When we collect multiple outcomes per family we will aggregate them into a standardized index to address concerns regarding multiple hypothesis testing.

Our analysis will focus on the following families of secondary outcomes.

1. Gender Norms and Attitudes: Here we will consider impacts on several sub-families of gender outcomes including actual and perceived norms governing women's phone use and broader norms about women's role in society. We will analyze impacts on norms and beliefs held by women and their husbands separately.

2. Mental Health and Well Being: We will study impacts on measures of anxiety and depression (K6 index) and a woman's sense of self worth.

3. Empowerment: The measure of empowerment most directly related to our intervention is a woman's ability to use a phone independently, without family interference. We will also study impacts on downstream measures of women's empowerment, including mobility, economic activity, and self-reported decision making power.

4. Social Connections and Networks: We will measure the number of social contacts a woman has, how often she interacts with them, and whether these interactions are by phone or in person. We will also measure participation in women's self help groups.

5. Economic Activity: We will measure women's labor force participation and engagement in income generating activities.

When we collect multiple outcomes per family we will aggregate them into a standardized index to address concerns regarding multiple hypothesis testing.

First we randomized 212 villages into three groups: control (32 villages); low saturation (90 villages); high saturation (90 villages).

Then at the individual level within high and low saturation villages we randomly assign 12 "core sample" women to receive an offer of phone training only; 18 to receive an offer of training and Mor Awaaz; and 18 to receive an offer to training, Mor Awaaz, and incentives. In high saturation villages all "non-core" women eligible for a government phone distribution program were also invited to receive training and Mor Awaaz.

Randomization was conducted in office using Stata.

The control/high saturation/low saturation randomization is clustered at the village level.

Randomization to training only, training + Mor Awaaz, training + Mor Awaaz + incentives is conducted at the individual level.

Yes

212 clusters

1,696 core sample women participated in the baseline survey (8 women in 212 villages). We also interviewed their husband. At endline, we will enroll another 1,696 women and their husbands.

First we randomized 212 villages into three groups: control (32 villages); low saturation (90 villages); high saturation (90 villages).

Then at the individual level within high and low saturation villages we randomly assign 12 "core sample" women to receive an offer of phone training only; 18 to receive an offer of training and Mor Awaaz; and 18 to receive an offer to training, Mor Awaaz, and incentives. In high saturation villages all "non-core" women eligible for the government phone distribution were also invited to receive training and Mor Awaaz.

Pre-Analysis Plan: Hold the Phone? Social and Economic Implications of Improving Women's Agency and Control over Smartphones