Intra-household Communication and Female Labor Force Participation

The treatment is a training in effective communication, designed by an NGO we are partnering with.

2022-10-08

2022-11-07

(1) Women's employment: (i) any work for income, (ii) sign-up for training and employment program offered by partner firm, (iii) attendance in vocational training, and (iv) attendance in employment post-training.

(2) Index of outcomes husbands and wives disagree about, including women's employment, purchase of a saree, women visiting the market, and women visiting their natal villages.

We will test for heterogenous effects in 1. by baseline disagreement between preferences of husbands and wives regarding the wife’s employment
We will test for heterogenous effects in 2. by baseline disagreement index between preferences of husbands and wives regarding each particular behavior (split at the median)
We will also test for heterogenous effects in 1. and 2. by age of the woman, hypothesizing that younger women may benefit more from the communication training.

1. Survey and vignette questions on behaviors undertaken when preferences diverge, to distinguish effective communication behaviors from other types of communication as well as from direct bargaining.
2. Wages in the last 90 days
3. Reported preferences regarding female labor force employment and vocational training (measured as interest in implementation partner's vocational training, employment as well as general female labor force employment). We will make an index of these outcomes. This may not change if the program does not change their preferences but is a potential mechanism if it does.
4. Who has the final say in decisions such as food, clothing, large purchases, spending husband's earnings, whether the woman works, and spending her earnings. We will create an index with these variables.
5. Women's interest in own employment, purchase of a saree, visiting the market, and visiting their natal villages. We will create an index with these variables. This may not change if the program does not change their preferences but is a potential mechanism if it does.
6. Whether the woman gives input into decisions such as food, clothing, large purchases, spending husband's earnings, whether the woman works, and spending her earnings. We will create an index with these variables.
7. Gender attitudes regarding female labor force employment
8. Self-reported happiness
9. Lab in the field games, including:
i) Variants of the dictator game:
a) One version of the game will only have the husband decide to share an endowment, with a subset of the couples allowed to communicate first.
b) A second version of the same will include a component where surplus shared by the husband doubles and the wife decides the final split. A subset of the couples allowed to communicate first.
10. Self-efficacy
11. Knowledge of a key component of communication training (see-feel-want concept).

We will conduct a randomized controlled trial with one treatment arm and one control arm. The treatment is a training in effective communication, designed by an NGO we are partnering with. Women will be put into groups with a facilitator who will train them in effective communication techniques. In the control group, women will also be put into groups with a facilitator that will meet on the same schedule as the treatment groups, but with no focused discussion on communication. At the end of the intervention period, women will be invited to participate in a weaving training and employment program set up by our partner firm, a large carpet manufacturer in Uttar Pradesh.

Randomization done in office by a computer

Group

Yes

240 meeting groups

1540 women

120 groups (with 779 women) treatment, 120 groups (with 761 women) control