Abstract: We study how peers affect in-group bias. Exploiting several umpiring reforms in international cricket matches —where two umpires make independent decisions in each other’s presence—we show that home-team umpires are less biased when working with a neutral colleague, i.e. one who is neither a national of the home nor the foreign team. This temporary debiasing is driven by the social pressure umpires feel to be impartial in the presence of neutral peers.Performance evaluation by visually non-salient monitors does not reduce bias, suggesting that physical presence is an important component of debiasing and peer influence.

Abstract: Firms accustomed to hiring through networks may face sparse applicant pools, limiting the quantity or quality of hires. Online job portals expand their pool of potential employees, but such firms may be reluctant to trust workers outside their networks. We sample over 1,500 firms posting vacancies on an online job portal in India and randomly: (i) increase the size of their applicant pool, doubling the number of applicants; (ii) provide background verification services for their applicants; or (iii) both. Firms assigned to the combined treatment increase hiring from the portal relative to control and the individual treatments, which contributes to greater hiring overall. The hires from the portal are good matches as measured by employee retention after six months. We show how combining the treatments can generate a complementarity when firms value both ability and trustworthiness, but place greater relative weight on the latter. Our results suggest the effectiveness of online job portals can be enhanced through the provision of screening tools that induce greater employer trust in job seekers.

Abstract: The inheritance of wealth promotes occupational choice, but may restrict it where its use is constrained by limited markets and cultural norms. This paper investigates the effects of inheriting agricultural land in rural India and finds that while larger inheritances, on average, increase future household consumption, first-born sons do not experience these gains. For first-borns, inheriting land reduces urban migration and entry into non-agricultural work. In contrast, inheriting land does not influence occupational choice or migration for latter-born sons. I attribute these differences, in part, to a cultural norm of parental support incumbent on first-borns and its interaction with inherited land. Patrilineal Hindu inheritance customs, whereby sons inherit an equal amount of their parents’ land, motivate sibling sex composition as an instrument for land. In support of instrument exogeneity, I find no reduced form effects for individuals with landless parents.

Abstract: For many in South Asia, international migration to the Gulf countries provides access to lucrative employment opportunities without domestic comparison. Yet higher wages in Gulf countries are often coupled with poor working conditions, employer malpractice, and abuse. We utilize a unique administrative dataset on Sri Lankan migrant workers and complaints made by them to Sri Lankan Consulates to analyze the prevalence of workplace abuse and employer malpractice along this complex supply chain. Our analysis reveals that Sri Lankan migrants are systematically exposed to fraud and abuse that can, in part, be attributed to factors specific to the local recruitment agency who placed them. Understanding why recruitment agencies systematically differ in the quality of their placements may inform regulatory policy aimed at reducing the risks faced by migrants in destination countries.

Abstract: Do Information and Communications Technologies (ICTs) ‘disrupt’ in-person peer interactions that generate and spread local knowledge? To investigate, I randomize access to a mobile phone-based agricultural extension service and find that while it reduces reliance on peer agricultural advice, it does not crowd-out peer interactions. Instead, treated farmers are more likely to recommend inputs to their peers, who, in turn, prioritize interacting with them. Consequently, exposure to the treatment, directly or via peers, increases willingness-to-pay for the service. Overall, evidence on complementarities between treated respondents suggest ICT-based services may encourage peer interactions and technology diffusion at scale.

Abstract: Mobile phones promise to bring the ICT revolution to previously unconnected populations. A two-year study evaluates an innovative voice-based ICT advisory service for smallholder cotton farmers in India, demonstrating significant demand for, and trust in, new information. Farmers substantially alter their sources of information and consistently adopt inputs for cotton farming recommended by the service. Willingness to pay is, on average, less than the per-farmer cost of operating the service for our study, but likely exceeds the cost at scale. We do not find systematic evidence of gains in yields or profitability, suggesting the need for further research.

Abstract: Knowledge of consumer demand is important for firms, policy-makers, and economists. One common tool for incentive-compatible demand elicitation, the Becker-DeGroot-Marschak (BDM) mechanism, has been widely used in laboratory settings but rarely evaluated for reliability at large scale in the field. In two field experiments (for a new agricultural information service and rainfall index insurance) we compare demand curves estimated from BDM implementations with demand curves estimated from choices at individually randomized fixed prices. In the test (for an agricultural information service) we obtain demand curves from BDM and fixed prices that align closely and are unable to reject equivalence of the demand curves. For the second test (of rainfall index insurance) the results are mixed, with the distributions lining up well at certain points of the demand curve and deviating at others. Overall, we find no evidence for systematic bias. Our evidence suggests that a ‘reframed’ version of BDM that uses discrete prices may minimize non-standard bidding behavior and is well suited to future experimental work in low-literacy environments.

Abstract: Migrant workers and prospective employers rely on intermediaries to facilitate complex, lucrative, and often risky labor market matches. However, if information frictions obscure their reputation, intermediaries may have little reason to invest in the quality of their placements. Using data on over 1.5 million Sri Lankan migrants to the Gulf region, we test whether an intermediary rating program designed to ease information frictions can improve placement quality. The regulator first announced the rating criteria and then publicly revealed a rating. Agencies eligible for the program—especially previously under-performing agencies—invest in the rating criteria and are less likely to exit the market. Migrants placed by these agencies receive higher salaries, are matched with less abusive employers, and receive better non-wage amenities even prior to the ratings being revealed. Once revealed, higher ratings increase foreign demand, but elicit limited migrant response. Collectively, our results suggest the program induced intermediaries to improve placement quality by screening employment opportunities abroad.

Abstract: International migration offers access to lucrative but, often risky, labor market opportunities. Concerns about migrant abuse have led some sending-country governments to impose sweeping restrictions on migration with potentially important welfare consequences to would be migrants. We study a policy implemented by the Sri Lankan government in 2013 that restricted women’s migration opportunities both because of their own age and the age of their children. The ban’s intent was to reduce the potentially harmful costs of migrant mothers who leave young children behind. We find that the policy drastically reduced female migration to the Gulf region, suggesting strong compliance with the ban. Second, we find that households subject to the ban experience a reduction in remittance income lowering household consumption. Mothers subject to this ban reduce their fertility— an effect driven by behavioral response to the ban as opposed to compositional changes resulting from returning mothers and those restricted by the ban. The reform does not improve child outcomes. Rather, children in households affected by the ban are less likely to report attending school and have fewer years of education.

Abstract: We assess the search and employment impacts of providing job seekers with information about local labor market conditions on a job portal in India. Specifically, for their desired city and occupation, we give truthful information about vacancies (Demand), searchers (Supply), or a combination of the two (Tightness). A control group receives only general job search advice. We find that treated respondents— who receive any local labor market information— are 10.2% more likely to be employed than control respondents six months later. Effects are largest for the group that receives Tightness information. The overall positive effect masks heterogeneity by baseline employment status and priors about job finding via the portal. Those employed at baseline are more likely to be employed six months later. Among the unemployed at baseline, there is a small positive, but not significant, effect on employment; however, those unemployed individuals with more realistic job finding beliefs are more likely to be employed. Altogether, our results shows how access to broader information about local labor markets can reduce frictional unemployment.

Abstract: Social learning is an important component of the process of technology diffusion. If actions are not perfectly observable, learning may instead require information sharing by word-of-mouth. This project presents results from cluster-randomized controlled trial of a mobile phone-based agricultural extension program in rural India. The intervention was implemented in a setting in which farmers were already participating in bi-weekly agricultural meetings in learning groups. The intervention was randomized at the village-level but with varying degrees of saturation at the learning group level. We find that the demand by farmers who were randomly invited to participate was high, and that their agricultural knowledge and practices changed one year later. However, untreated members of their agricultural group were no more likely to have higher knowledge scores nor change their agricultural practices and they were less likely to report that they relied on others for agricultural information. These results suggest that information spillovers may be small and, at least in some contexts, relying on farmers to discuss agricultural information with others might not be an effective strategy for technology diffusion.