This is the 26th in this year’s series of posts by PhD students on the job market.
In low- and middle-income settings, many available jobs are undesirable. Wages are low and workers face a host of challenges, including long commutes, injury risk, and wage theft. Workplaces can be particularly tough for women, who face higher risk of harassment. When reliable information about workplace conditions is scarce, the possibility of landing in an unpleasant workplace may not only deter women from working but can also discourage working women from pursuing new opportunities, limiting their outside options and wage growth.
My job market paper studies a strategy women can use to manage these risks: working in the same establishment as their husbands. Using new survey data from India’s garment sector, I first document that “spousal co-working” is widespread: 46% of married women in the garment sector work in the same factory as their husbands, despite living near hundreds of potential employers.
Why do couples co-work and what are the labor market implications? Co-working may help couples navigate uncertainty about working conditions. Husbands can screen workplaces and refer their wives if conditions are good. In fact, about 60% of couples working together say the husband joined the factory first. If couples switch together, the husband’s presence can deter harassment, help women negotiate with supervisors for better treatment, and make commutes safer or more pleasant. Of course, couples may co-work for other reasons, such as spending time together. Co-working may actually be detrimental for the wives if some husbands do not permit their wives to work independently or use co-working as a monitoring tool. Do employers also recognize that couples value co-working and how do they respond?
To answer these questions, I draw on rich survey data and three field experiments with workers and firms in Tirupur, Tamil Nadu, which employs 31% of India’s garment workforce. On the labor supply side, I estimate how much workers value switching to a new job with a spouse rather than alone and probe the mechanisms driving this phenomenon. On the labor demand side, I test whether firms adjust hiring and wages to capture rents when couples value co-working.
Setting and co-working prevalence
India’s textile and garment industry, the country’s second-largest employer after agriculture, often entails harsh working conditions. I recruited garment workers through household surveys near three factory clusters in Tirupur. In a survey of over 500 workers, 20–35 percent reported various challenges: unsafe conditions, delayed salaries, wage theft, and long hours. Moreover, raises and promotions are infrequent, so job switching is the primary pathway for wage growth. A second survey covering more than 4,000 workers shows that 46% percent of married women in the garment sector co-work with their husbands.
Both women and men value co-working
To estimate the value of co-working, I ran an incentivized Job Choice Experiment with nearly 800 married garment workers, including couples working together and separately. Participants chose between several pairs of hypothetical jobs with randomized pay and attributes. In each pair, one job guaranteed co-working (the spouse could be hired immediately at the same factory); the other had a single vacancy. To introduce real stakes, we recommended candidates profiles to HR managers at real factories based on some of their survey responses.
Figure 1: Willingness to pay for co-working, by gender and co-working type
Women value co-working as much as a 36% increase in monthly wages. Men value it too, but less (24%). When men were asked to choose between hypothetical jobs for their wives, their valuation increases to 38%, closing the gender gap.
Co-working mitigates risk in job transitions
I consider three explanations for co-working preferences: co-working may protect wives when moving to a new, unfamiliar workplace (risk mitigation); couples may value working with someone they know (familiarity); and some husbands may only allow their wives to work if they can monitor workplace interactions or control their wages (spousal control).
A model of workplace choice yields testable predictions. If risk mitigation dominates, women should be less willing than men to switch alone; couples may move together or let husbands screen first; and husbands should be more willing to search alone once the wife’s workplace is known to be safe. If familiarity or control dominate, joint moves should be preferred to either spouse switching alone.
I experimentally test these predictions. 338 couples in the garment sector, both co-working and not, were offered referrals for new jobs at nearby factories. I randomized whether both spouses were recommended to the same factories (“Joint” arm with 171 households), or only one spouse was recommended (“Husband Only” and “Wife Only” arms with 80 and 87 households respectively). I then measured whether participants called a hotline to apply for the referral.
Results suggest risk mitigation is the primary motive for co-working. Only 12% of households in the Wife Only arm applied, compared with 32% in the Husband Only arm and 26% in the Joint arm. Among co-working couples, husbands were significantly more interested in being recommended alone when couples reported higher satisfaction with their current workplace than when they reported low satisfaction. Thus, once the couple learns that the workplace is suitable, husbands are comfortable searching for better opportunities.
Figure 2: Applications for job referrals for new garment jobs, by treatment arm
Turning to other mechanisms, I find that spousal co-working offers benefits beyond working with someone familiar, but control may be a relevant driver. In another job choice experiment, I find that working with a spouse is valued twice as much as working with a same-sex friend or relative. Willingness to pay for co-working is 36%-50% higher among women with below-median mobility (measured by solo trips outside the home) and among men whose wives describe them as more jealous.
Even so, co-working husbands are willing to switch alone during the experiment and 60% of co-working couples report that the husband joined the factory first, suggesting that the control and familiarity mechanisms are less important.
Implications for job mobility and wages
Women are more willing to switch jobs when they can move with a spouse than when they have to move alone. Co-working therefore increases women’s job mobility, which matters because job-to-job moves are the main engine of wage growth in this setting. Relatedly, in a survey experiment with women who are not working I find that fear of unsuitable workplaces may limit employment for women, and co-working opportunities may enable them to work.
Note this does not only affect women: I find that men working with their wives in lower-quality factories were reluctant to switch jobs alone, limiting their wage growth as well. Observationally, men typically switch first and refer their wives; if men can’t search alone, couples must wait till they can move together or remain in place, potentially making them less mobile than other workers.
Do employers recognize that co-working exists? Do they take advantage and underpay co-working couples? I run an employer-side hiring experiment, asking 65 garment sector employers to choose between pairs of hypothetical job applicants, and incentivize meaningful responses by sending real job seeker profiles matching respondents’ reported preferences. Employers strongly prefer hiring co-working couples over other applicants, believing they are less likely to move. Perhaps unsurprisingly, they then offer co-working women 6% lower wages than female applicants seeking jobs alone for skilled machine operator positions. There is, however, no wage markdown for husbands by coworking status.
Policy implications
Co-working can ease barriers to women’s work. However, it remains a second-best solution. It raises household income risk, especially in an industry vulnerable to export demand shocks. By tying spouses to the same sector, co-working may divert each from their comparative advantage, contributing to misallocation.
Three lessons follow. First, in the short run, creating jobs in industries where couples can work together may boost female labor force participation. Second, since spousal co-working allows women to tolerate harsh conditions, it may weaken factories’ incentives to improve. Robust government policy is needed to ensure safe, fair workplaces. Finally, policies to improve working conditions and better enforcement of labor standards may reduce the risk of switching jobs and couple’s dependence on coworking, promoting access to more job opportunities and career growth for both women and men.
Shreya Tandon is a PhD student at Harvard University.
Join the Conversation