Research

Access to smartphones and mobile internet is increasingly necessary to participate in the modern economy. Yet women significantly lag men in digital access, especially in lower-income settings with gender gaps that span other dimensions - and where digital gaps threaten to deepen existing analog inequities. We study the short- and long-term effects of a large-scale state-sponsored program in India that aimed to close digital gender gaps by transferring free smartphones to women while constructing 4G towers to bring rural areas online. The program was well implemented, reversing gender gaps in smartphone ownership in the short run. However, many women lost ownership and gender gaps in use quickly worsened as men made use of the new phones. Nearly 5 years after the program began, we find limited evidence of persistent effects across a range of outcomes, including phone ownership and use, gender norms, access to information, and local economic activity, although we do find some evidence of sectoral reallocation in the labor market. Despite widespread increase in smartphone adoption across households, digital gender gaps persist and were not affected by the program. Our findings suggest that in gender-unequal, resource-constrained settings, addressing affordability alone may not close digital gender gaps.

Modern video and social media platforms host an unprecedented collection of easily accessible digital learning materials, yet this content remains out of reach of many skill seekers in low- and middle-income countries. This project asks whether mobile internet can support on-the-job skill acquisition in small informal firms when key barriers to digital learning are relaxed.

Focusing on tailoring firms around Kampala, Uganda, the project will pilot interventions that address three constraints to learning from online video platforms: affordability, access to relevant and understandable content, and the ability to identify and use useful content in a noisy digital environment. Candidate treatment arms provide (i) free mobile data plus an informational nudge, (ii) free data plus curated, language-accessible instructional videos, or (iii) free data plus training on how to search for, assess, and apply online content. The pilot will proceed in two stages: a logistical pilot to refine intervention delivery and measurement, followed by a mini-RCT with a comparison group and the most feasible treatment arm(s). Primary outcomes are profits and applied skill acquisition, the latter measured through an incentivized demand-drop exercise in which firms complete a novel production task that is independently scored.

Implemented in partnership with Dyadic Research Impacts, a local survey research firm, the project tests a scalable, low-cost approach to skill development that leverages existing global instructional content, while identifying frictions that prevent mobile internet access from translating into on-the-job human capital formation.

We evaluate the economic and social impacts of cash transfers and complementary entrepreneurship training for refugee and host communities in urban Nairobi, Kenya. In a randomized, controlled trial that varies transfer saturation levels among hosts and refugees in Nairobi neighborhoods, we examine impacts on refugee livelihoods, on social cohesion, and on consumer, producer, and input prices, among other outcomes. Outcomes are measured in quarterly follow-up rounds. This design sheds light on the relative importance of liquidity and human capital constraints for refugees, and on the relative importance of these constraints among comparably situated host-community members. Exploiting randomized saturation, we test whether local demand constrains refugees' entrepreneurial growth, and whether refugees respond differently to demand from their own community relative to hosts residing in their neighborhoods. Results will have implications for the design of refugee livelihood programs, and will shed light on the economic incentives shaping ethnic enclaves among refugee communities.

In many low-income settings, wage employment at small firms does not fully insure workers against income risk: workers are often paid piece rates or through other output-contingent arrangements, leaving them exposed to firms' demand risk. This may make wage employment less attractive, increase turnover, and reduce firms' incentives to train and retain workers. This study tests these hypotheses through a randomized experiment with 900 tailoring, welding, and carpentry firms in Kampala, Uganda. Using a mystery-shopper approach, the research team will place standardized product orders with firms, randomly varying whether induced demand is stable or variable over time and, among firms facing variable demand, whether the order schedule is known in advance. By holding total induced demand constant across the treatment arms, this design isolates the roles of volatility and predictability on a range of firm- and worker-level outcomes, which are measured through high-frequency follow-up surveys during and after the intervention. The study will estimate impacts on firm performance, worker earnings and earnings stability, risk-sharing, worker retention and upgrading, and firms' organization of production. In doing so, it sheds light on how demand volatility shapes worker turnover and how within-firm labor dynamics may constrain small-firm growth.

Urban labor markets in developing countries are marked by instability, weak social protection, and widespread informality, making risk-sharing a central concern for workers and policymakers. This paper examines the nature of insurance and informal networks in urban settings using longitudinal data from Ghana and Uganda, supplemented by original survey evidence from manufacturing workers in Kampala, Uganda. Tests for full insurance reveal that while urban households are not fully insured, their consumption is only weakly sensitive to idiosyncratic income shocks, with elasticities between 0.01 and 0.05, indicating effective partial insurance. Assessing the relevance of risk-pooling networks, I compare geographically proximate networks to those consisting of communities in places of origin. Evidence suggests that urban co-location provides stronger pooling than rural-origin networks, though data granularity limits definitive conclusions. To supplement these findings, I conduct a survey of workers capturing detailed information on shocks, insurance mechanisms, and coping strategies. Results highlight the predominance of self-insurance via precautionary savings, with over 60% of workers reporting a shock in the past month and 90% relying on savings to cope. This strategy, however, delays transitions into self-employment, potentially contributing to aggregate misallocation. Finally, I document polarized demand for unemployment insurance, shaped heavily by workers’ beliefs about job search prospects. The findings underscore the importance of understanding informal arrangements in shaping worker welfare, labor market trajectories, and the design of effective safety nets in rapidly urbanizing economies.

Government across developing countries have addressed the jobs problem by investing in skilling and vocational training. Prior research has identified multiple demand-related channels to explain tepid returns to such investments. However, little is known about the supply of vocational training, especially the kind of trades offered and the quality of training provided. By constructing a novel dataset on the universe of vocational training institutes (VTIs) in India, I am able to establish five crucial stylized facts. First, I find that government training institutes perform 0.6 SD better than private counterparts based on a holistic quality assessment score. Second, better quality institutes tend be older, located in urban settings, run by the government, and offer more trade courses. Third, I find no preliminary support for competition driving up the quality of training provided. Fourth, in post economic liberalization India, the supply of private vocational institutes is responsive to literacy and population growth—in particular of Scheduled Castes. Finally, the types and array of trade courses offered remain unresponsive to evolving local economic and demographic conditions. Taken together, these results point to lack of dynamism in trades offered and private provision of VTIs as significant barriers to the supply of quality vocational training in India.