Studying the latent demand of female labour

In the study, we focus on whether allowing women to work from home can induce them to participate in the labor force and increase their labor supply. The experiment involves offering work opportunities to women to produce crochet products. We implement a cross randomisation experimental design of 2 different interventions: 1) work from home vs work from workshop, 2) revelation treatment and control. Randomisation is conducted at the household level whereby participants from the same household will be alloted to the same arm.

2024-03-07

2025-06-30

The primary outcomes of interest are a) the absolute number of final items produced by each worker in a week; b) the number of hours worked by each worker in a day (self-reported);
Hypothesis, H1: Being assigned to the work from the home treatment arm will increase the number of items produced and the number of hours worked by each worker compared to being assigned to the work from the workshop arm.

The second type of outcomes of interest that we will explore is productivity of each worker defined as the number of products produced by each worker per unit of time.
Hypothesis, H2: Being assigned to the work from the home treatment arm will increase worker’s productivity compared to being assigned to the work from the workshop arm.
Hypothesis, H3: Assignment to the revelation treatment group lowers women’s labor supply and productivity compared to assignment to the control (no revelation) group.
Hypothesis, H4: The work from home treatment will have a larger positive impact on labor supply and productivity for women whose partners are more opposed to women’s work than for women whose partners are more receptive to women working

The third type of outcomes of interest we will examine will derive from the time use patterns of the respondent and aims to capture the complementary nature of certain in- home activities with home production. We will include an indicator variable equal to 1 if a respondent performs home activities simultaneously with economic production, and 0 otherwise; and the total duration that a respondent multi-tasks between home activity and economic production.

Hypothesis (H5): Assignment to work from the home treatment arm increases the likelihood of multitasking between economic production and home activities compared to assignment to work in the workshop arm.

In the study, we focus on whether allowing women to work from home can induce them to participate in the labor force and increase their labor supply. The experiment involves offering work opportunities to women to produce crochet products. We implement a cross randomisation experimental design of 2 different interventions: 1) work from home vs work from workshop, 2) revelation treatment and control. Randomisation is conducted at the household level whereby participants from the same household will be alloted to the same arm.

Each woman is randomly assigned to the following sub-groups as displayed in figure 1 for three weeks.
1. Home vs Workshop
The first randomisation involves work from home vs work from workshop. In the work from the workshop arm, women will be offered work opportunities in a workshop- center in the village to create crochet products. On the other hand, in the work from home arm, women will be offered the opportunity to work at home to produce the same crochet products.
2. Home vs Workshop
Participants who are randomly assigned to the ‘work from home’ and ‘work from workshop’ arm will be further cross-randomised into an identity-revelation sub-treatment. Half of the women in both the arms will be randomly assigned to the “revelation” (or treatment) arm. Women assigned to this arm will have their names displayed on posters around the village to encourage other women to join the program. On the other hand, in the “no revelation” (or control) arm, women’s names will not be put up on posters. The women whose information will be revealed will have prior knowledge of the revelation.

The primary purpose of this experiment is to understand whether a more flexible opportunity, which relaxes workers’ mobility constraints, affects female labor force participation rate and labor supply.

Not available

Randomization will be done through STATA, a statistical software.

Randomization will be conducted at the household level i.e. all women belonging to the same household will be assigned to the same arm

No

500 women

500 women

250 in work from home arm, 250 in work from workshop arm and 250 assigned to the poster arm (half from each work from home and work from workshop arm)