Research

Downloads: paper (updated 2024/12),  additional appendices,  AEA trial registry

Abstract: Do higher aspirations for the future motivate people living in poverty to make long-term investments? Do their aspirations increase when economic conditions improve? To answer these questions, we run a 415-village field experiment with 8,300 women living in poverty in rural Kenya. We design an 80-minute workshop to help people set higher aspirations and plan to achieve them. We cross-randomise this with large unconditional cash transfers. The workshop substantially raises aspirations, labour supply, investment, revenue, and living standards 17 months later, relative to a placebo workshop. Increases in aspirations are the most likely mechanism to explain the economic effects. Cash transfers also raise aspirations, which might help to explain why transfers increase labour supply and investment. We conclude that aspirations respond to both economic and psychological interventions, contribute to investment decisions and living standards, and are important considerations for development policy.

Downloads: paper (updated 2024/11),  AEA trial registry

Coverage: IGC, VoxDev

Abstract: Understanding specific barriers to job search and returns to relaxing these barriers is important for economists and policymakers. An experiment that changes the default process for initiating job applications increases applications by 600% on a search platform in Pakistan. Perhaps surprisingly, the marginal treatment-induced applications have approximately constant rather than decreasing returns. These results are consistent with a directed search model in which some jobseekers miss some high-return vacancies due to psychological costs of initiating applications. These findings show that small reductions in search costs can substantially improve search outcomes in environments with some relatively inactive jobseekers.

Downloads: published paper,  final preprint version

Downloads:  published paper (open access)

Abstract: This VoxDevLit summarises evidence documenting the labour market barriers faced by urban jobseekers and firms in low and lower-middle income countries, highlighting the role of information frictions on both sides of the market: jobseekers’ monetary and psychological search costs; jobseekers’ limited networks; and recruitment costs for firms. We also evaluate the impacts of a wide set of interventions designed to address these barriers. We find a fairly strong positive evidence base for interventions that provide information about worker skills, both to firms and to workers. There is also evidence that interventions relieving search costs can boost employment for populations that are liquidity-constrained and with realistic job targets, but effects may not persist. There is mixed evidence on interventions altering jobseekers’ beliefs, encouraging jobseekers’ use of job platforms and altering firms’ recruitment strategies. There is limited evidence on interventions targeting other constraints, such as psychological costs of search or limited networks, and generally less evidence on frictions facing firms than those facing workers. Existing evaluations are generally not informative on (i) displacement effects (i.e. negative employment effects for untreated workers), and (ii) general equilibrium effects that would only materialise when interventions are offered at scale. Evidence suggests search barriers may differ by gender but we have very little evidence showing how job-search policies could be adapted to address this issue.

Downloads: published paper (open access)

Abstract: This paper summarises evidence from a rapid review of international findings on the effects of cash transfers. The learnings were used to inform the design of urgent response social protection during the COVID-19 pandemic. The summary demonstrates that in response to widespread disruption, cash transfers have broad benefits for children, adults, and the wider economy. First, cash grants for child support have been shown to reduce hunger, increase dietary diversity, and reduce secondary school dropout. Transfers may also reduce child malnutrition. Second, there is some evidence that cash grants could encourage job search and economic activity. By contrast, there is little evidence that grants discourage adults from working, or increase spending on temptation goods (alcohol, tobacco). Third, for the wider economy, there is little evidence that grants will increase inflation, while some studies find that transfers create a fiscal multiplier and stimulate the local economy. Finally, we review evidence on design considerations and find that unconditional cash transfers (UCTs) are particularly well suited to rapid response when compared to conditional cash transfers (CCTs). Outside crisis settings, there is some evidence that getting recipients to enrol children in school or attend health check-ups improves these outcomes more than unconditional grants. However, the differences are small and the additional costs of implementing conditions outweigh these benefits in circumstances where response is required urgently. Comparing evidence on cash grants and food vouchers suggests both achieve similar improvements in nutrition; however, cash transfers are likely to be more cost-effective for governments, especially where a system to distribute grants is already set up.

Downloads: published paper,  final preprint version,  replication code, data & questionnaires,  AEA trial registry,  guide for implementing skill certification interventions

Coverage: GlobalDev, J-PAL, VoxDev

Abstract: We assess South African workseekers' skills and disseminate the assessment results to explore how limited information affects firm and workseeker behavior. Giving workseekers assessment results that they can credibly share with firms increases workseekers' employment and earnings and better aligns their skills, beliefs and search strategies. Giving workseekers assessment results that they cannot easily share with firms has similar effects on beliefs and search, but smaller effects on employment and earnings. Giving assessment results only to firms shifts interview decisions. These findings show that getting credible skill information to the right agents can improve outcomes in the labor market. 

Downloads: published paper,  final preprint version,  replication code, data & questionnaires,  AEA trial registry,  training curriculum

Abstract: Online professional networking platforms are widely used and may help workers to search for and obtain jobs. We run the first randomized evaluation of training work seekers to join and use one of the largest platforms, LinkedIn. Training increases the end-of-program employment rate by 10 percent (7 percentage points), and this effect persists for at least 12 months. The available employment, platform use, and job search data suggest that employment effects are explained by work seekers using the platform to acquire information about prospective employers and perhaps by work seekers accessing referrals and conveying information to prospective employers on the platform. 

Downloads: published paper,  final preprint version,  replication code

Abstract: Researchers routinely use datasets where outcomes of interest are unobserved for some cases, potentially creating a sample selection problem. Statisticians and econometricians have proposed many selection correction methods to address this challenge. We use a natural experiment to evaluate different sample selection correction methods’ performance. From 2007, the state of Michigan required that all students take a college entrance exam, increasing the exam-taking rate from 64% to 99% and largely eliminating selection into exam-taking. We apply different selection correction methods, using different sets of covariates, to the selected exam score data from before 2007. We compare the estimated coefficients from the selection-corrected models to those from OLS regressions using the complete exam score data from after 2007 as a benchmark. We find that less restrictive semiparametric correction methods typically perform better than parametric correction methods but not better than simple OLS regressions that do not correct for selection. Performance is generally worse for models that use only a few discrete covariates than for models that use more covariates with less coarse distributions.

Downloads: published paper,  final preprint version,  replication code, data & questionnaires,  AEA trial registry

Abstract: This study analyzes the effects of differences in survey frequency and medium on microenterprise survey data. A sample of enterprises were randomly assigned to monthly in-person, weekly in-person, or weekly phone surveys for a 12-week panel. The results show few differences across the groups in measured means, distributions, and deviations of measured data from an objective data-quality standard provided by Benford’s Law. However, phone interviews generated higher within-enterprise variation through time in several variables and may be more sensitive to social desirability bias. Higher-frequency interviews did not lead to persistent changes in reporting or increase permanent attrition from the panel but did increase the share of missed interviews. These findings show that collecting high-frequency survey data by phone does not substantially affect data quality. However, researchers who are particularly interested in within-enterprise dynamics should exercise caution when choosing survey medium.

Downloads: published paper,  final preprint version,  replication code

Abstract: I study the relative academic performance of students tracked or randomly assigned to South African university dormitories. Tracking reduces low-scoring students' GPAs and has little effect on high-scoring students, leading to lower and more dispersed GPAs. I also directly estimate peer effects using random variation in peer groups across dormitories. Living with higher-scoring peers raises students' GPAs, particularly for low-scoring students, and peer effects are stronger between socially proximate students. This shows that much of the treatment effect of tracking is attributable to peer effects. These results present a cautionary note about sorting students into academically homogeneous classrooms or neighborhoods. 

Downloads: paper (updated 2025/04)

Abstract: We study whether commuting barriers constrain women’s labor supply in urban Pakistan. We randomize offers of gender-segregated or mixed-gender commuting services at varying prices. Women-only transport more than doubles job application rates, while mixed-gender transport has minimal effects on men’s and women’s application rates. Women value the women-only service more than large price discounts for the mixed-gender service. Results are similar for baseline labor force participants and non-participants, suggesting there are many “latent jobseekers” close to the margin of participation. These findings highlight the importance of safety and propriety concerns in women’s labor decisions.

Downloads: paper (updated 2025/03)

Abstract: Employment in developing countries is often short and disrupted, generating costly search and limiting the potential for workers to accumulate firm-specific human capital. We study the incentives guiding firms’ use of short-term relative to long-term contracts in Nairobi, Kenya, using novel survey data on firms’ hiring and contracting practices, and hypothetical vignettes measuring their beliefs and preferences. Our key finding is that the use of short-term labor is governed by a trade-off between managing demand variation versus minimizing adjustment costs and incentivizing worker performance. We first document that firms face considerable variation in demand for goods and services across time, much of which they pass on to workers through short-term contracts: higher demand variation is associated with a greater use of short-term labor. Second, we show that bringing on short-term workers involves adjustment costs: it takes time searching for, hiring, and on-boarding workers, potentially offsetting the gains from flexibility. We show both that median adjustment costs are low, making short-term contracts feasible for many hires, but that hires with greater adjustment costs are more likely to be on long-term contracts. Finally, we show that firms believe contract type incentivizes worker performance: the same worker is expected to perform better when hired on a long-term basis. We incorporate these features -- variation in demand, on-boarding costs, and incentives -- into a model of firm hiring, through which to interpret contract choice and turnover in low-income countries.

Downloads: paper (updated 2024/12),  AEA trial registry

Abstract: Worker sorting into tasks and occupations has long been recognized as an important feature of labor markets. But this sorting may be inefficient if jobseekers have imperfect information about their skills and therefore apply to jobs that poorly match their skills. To test this idea, we study two field experiments that give young South African jobseekers information on their results from standardized assessments of job-relevant skills. Information redirects jobseekers’ search toward jobs that value skills where they score relatively highly, without raising their search effort. Information also substantially raises earnings, consistent with inefficient sorting due to imperfect information.

Downloads: paper (updated 2019/08)

Abstract: Most governments provide some public education and must choose how to finance it from fiscal transfers and user fees. I study the effects of a shift from mixed funding to pure fiscal transfers in South African secondary schools. Fee elimination increased enrollment in early grades but decreased enrollment in later grades and both pass counts and pass rates on graduation exams. The latter patterns can be explained by crowding that generated reductions in proxies for instructional quality and hence learning outcomes. This demonstrates that price ceilings without supply-side investments may have limited or even negative effects on education outcomes.

Downloads: paper (updated 2016/12)

Abstract: We present some of the first evidence of within-sample variation in the efficiency of intrahousehold resource allocation. In a sample of rural, low-income Mexican households, observed consumption patterns are Pareto efficient for households with relatively old heads, but not households with relative young heads. This variation in efficiency has important welfare implications: younger households invest less in children’s education and their education investments are less sensitive to the receipt of cash transfers. We believe this is the first empirical link between inefficient household resource allocation and lower investment in children’s education, though this link follows naturally from standard models of investment in household public goods. Heterogeneity in efficiency by household head age may be due to cohort and/or lifecycle effects. The specific patterns of heterogeneity we find are more consistent with cohort than lifecycle effects, which means that average efficiency may decline over time in this population. From a policy perspective, these results highlight a limitation of cash transfers relative to other forms of poverty alleviation policy in settings where at least some households engage in inefficient resource allocation. Many papers have already established that the distributional effects of development policy depend on whether households behave as unitary actors; our results emphasize that it is also important to assess whether (some) households behave as efficient collective units.