Restrictive social norms (especially norms that impose female seclusion) and safety concerns (like the threat of harassment in public spaces) limit women's ability to move freely in many developing countries. It can also have significant economic consequences for women’s lives. One important yet unexplored consequence of these norms and concerns is that they compel women to travel only with companions. For many, traveling with family members or female friends is a way to navigate seclusion norms and feel safer. In India, where this study is based, this practice is widespread: 44% of women report that they are not allowed to travel outside their neighborhoods alone (DHS 2019-21).
Traveling with companions is not an issue if a woman can always find someone to accompany her. But when she cannot travel alone and a companion is not available, her ability to participate in economic activities—particularly in job search—can be constrained. Without someone to accompany her, she may be unable to visit potential employers to inquire about jobs or attend interviews. A natural solution could be to coordinate travel with other job-seeking women, who, due to their shared purpose, are more likely to be reliable travel companions. But this, too, can be challenging: in India, women's social networks are often limited to family members and rarely include job-seeking women.
In my job market paper, co-authored with Smit Gade, we experimentally evaluate whether connecting job-seeking women to coordinate their travel improves their job-seeking efforts. We partnered with five garment factories in North India, located in the suburbs of Delhi (Faridabad and Noida) to conduct a field experiment. Our primary goal is to test whether matching job-seeking women and facilitating them to coordinate their travel improves their attendance in interviews at the partner factories. Our secondary goal is to understand what may be driving these effects. Expanding women's networks to include other job-seeking women can also improve women's attendance in interviews by, for example, shifting beliefs about women's work or by increasing the amenity value of factory jobs. We focus on separating the effects of women coordinating their travel from the possible impacts of matching job-seeking women.
Study setting
The study is based in India, where women’s labor force participation rate is 33% compared to 77% for men (The World Bank, 2023). Despite barriers like household and childcare responsibilities, 28% of women in urban India report a willingness to work but do not search for jobs. Even when they do, they report looking for jobs less intensively than men and are less likely to find their preferred jobs. As per our survey data, approximately two-thirds of women reported traveling with companions in the past week to mitigate safety concerns compared to 18% of men.
Experimental design
Our field experiment was conducted in 106 lower-income neighborhoods near the partner factories. The neighborhoods were randomized into two treatment groups and a control group, stratified within the city and by distance to the nearest factory. We conducted the baseline survey starting March 2024, and follow-up surveys six weeks later, starting June 2024. The sample consists of 693 women (on average, 7 women per neighborhood) who, at baseline, were not employed outside their homes, were interested in working at partner factories, and were skilled in operating sewing machines. Women, on average, were 28 years old, married, and lived in nuclear households at baseline.
Following the factories’ usual hiring policy, women from the study sample had to show up at the factory gate to participate in the job interviews. Our empirical analysis relies on comparing the interview attendance rates of women across the following three groups:
Matching & Coordinated Travel (N = 241 women) - Within a treated neighborhood, participants were matched together through group meetings and were invited to attend interviews at the nearest partner factory. Their interviews were scheduled on the same dates and women were informed that this scheduling would enable them to travel together to interviews.
Only Matching (N = 224 women) – In this treatment, participants within a neighborhood were matched through group meetings (identical to Matching & Coordinated Travel), but their interviews were deliberately scheduled on different dates to minimize travel coordination.
Control (N = 228 women) – Participants from the same neighborhoods were scheduled for interviews on the same dates but they were invited to interviews through individual meetings rather than group meetings.
We first compare the interview attendance rates of women in the Matching & Coordinated Travel treatment with those in the control group to estimate the effect of facilitating travel coordination to factories. Second, we compare the attendance rates between the Matching & Coordinated Travel treatment and the Only Matching treatment to isolate the effect of travel coordination from the matching component (i.e., group meetings).
Impacts on interview attendance and role of coordinating travel
Figure 1: Effect on interview attendance
Among women from the Control group, only 15.4% attended interviews. In contrast, women from the Matching & Coordinated Travel treatment attended interviews at much larger rate — their interview attendance increases by 13.1 percentage points or by 85% relative to the control group (p-value = 0.034). The treatment effects are particularly pronounced for the subgroups of women who, at baseline, reported knowing fewer women living nearby and feeling unsafe while traveling: the treatment increases their interview attendance by 17 percentage points (155%) and by 34 percentage points (310%), respectively.
We present two pieces of evidence demonstrating that the effects of Matching & Coordinated Travel treatment on interview attendance stem from women being able to travel together to interviews. First direct evidence comes from the comparison of interview attendance rates across the two treatments groups. Compared to the Only Matching treatment, Matching & Coordinated Travel increases the interview attendance rate by 13.7 percentage points (p-value <0.01). Second, women in the Matching & Coordinated Travel treatment are 9.8 percentage points (or 100%) more likely to travel with study women from their neighborhoods to the interviews compared to the control women.
Notably, most women in the control group (83%) traveled to interviews with adult companions, and 63% of these women traveled with other control group participants from their neighborhoods. This suggests that the control group women actively sought out each other as travel companions to the interviews. However, their lower interview attendance rate relative to the Matching & Coordinated travel treatment indicates that many women were unable to make such connections on their own. The Matching & Coordinated Travel solves this coordination problem by connecting women who want to attend interviews but don't know other job seekers in their neighborhood.
We also find that the women in Matching & Coordinated Travel are 2.4 percentage points more likely than women in the control group (mean of 1%) to accept jobs at the partner factories, and those in the Only Matching are 1.2 percentage points more likely to do so.
Impacts on job search efforts beyond the interview experiment
Figure 2: Effect on job search beyond the interview experiment
Six weeks after the interviews, we find that the effects of the Matching & Coordinated Travel treatment extend beyond the interview experiment: women in this group are 12.6 percentage points (or 78%) more likely to make a job search trip outside their homes compared to the 16% of women from the control group. It also more than doubled the average number of job search trips made by women. The Only Matching treatment also had a positive though smaller and statistically insignificant effect. We provide direct evidence that these effects are driven by women continuing to coordinate their travel: women in the Matching & Coordinated Travel treatment are 8.9 to 11.1 percentage points (111% to 150%) more likely than those in the Only Matching treatment or control group to make job search trips with another woman from their neighborhoods.
Policy implications
Our findings suggest a key policy implication: firms requiring in-person interviews could expand their pool of female applicants by inviting or mobilizing women in groups for interviews on the same days. This is particularly relevant for garment factories, where hiring for production roles primarily takes place through in-person visits, with many factories even requiring workers to visit job sites to inquire about openings. Our study shows that matching job-seeking women into groups and enabling coordinated travel can help them make more frequent visits to factories and employers, thereby improving their employment.
We also compare the effects of this study with a pilot intervention that fully covered round-trip costs to factories for interviews and find that its impact on interview attendance was only half as large as that of the Matching & Coordinated Travel treatment. This suggests that combining coordinated travel initiatives with public policies like providing free public transport for women, as implemented in Delhi, could be effective in improving women's access to labor markets.
Rolly Kapoor is a PhD student at University of California Santa Cruz.
Join the Conversation