Experimental Design
We conducted a two-month-long baseline survey with female students in 130 Industrial Training Institutes (ITIs) located in 29 districts of the states of Delhi, Haryana, and Uttar Pradesh, India, between June and August 2022. We selected female students who, at the time of the baseline survey: i) were enrolled in the ITI, ii) were in the final year of their studies (i.e., whose year of graduation was 2022-2023), and iii) had access to a personal or a shared smartphone. At the end of baseline data collection, our sample included 4,411 women who met our selection criteria. We collected data on women’s demographic characteristics, socioeconomic status, education, labor market outcomes, job search behavior, career aspirations, preferences, personality traits, empowerment status, gender attitudes, and perceptions of social norms. Roughly 22% of the surveys were conducted over the phone, and the rest of the respondents were interviewed in person.
Using baseline survey data, we formed 1,838 clusters of female vocational trainees based on their potential social connections in the sample (i.e., a woman’s friends and friends of friends). Next, we randomly divided these clusters into a “control” group and a “treatment” group. We randomized at the cluster level to account for plausible intervention spillovers between sample women who are socially connected to each other. Women assigned to the treatment group will be contacted via phone to a) inform them about the job portal and b) facilitate and encourage them to download, install, and use the app on their phones. We stratified randomization based on the following baseline cluster-level characteristics: group size of the cluster, the district where the ITI is located, women’s beliefs about the shortage of jobs, and their expected salary.
We expect to run the intervention for six months. During this period, our partner organization will monitor the intervention by recording information related to women’s registration to the job portal, the creation of job profiles, and the number of job applications, among other app-usage indicators. We will implement a follow-up survey at the end of the intervention period.