Lab Experiments in Developing Country Contexts

Date:  Spring 2025
WP#: 636
Authors: Sara Lowes and Nathan Nunn

Abstract: Lab-in-the-field experiments, in which lab experiments are conducted in more naturalistic settings, are increasingly being implemented in developing country contexts. In this chapter, we outline the conceptual and logistical challenges typically associated with lab-in-the-field experiments in non-Western settings. We describe the importance of worldviews and how researcher preconceptions may inadvertently shape the types of research questions that are asked. We emphasize the importance of increasing diversity of subject pools and researchers. We also suggest a set of best practices when implementing lab-in-the-field experiments in developing countries.

The Social Consequences of Traditional Religion in Contemporary Africa

Date:  Spring 2025
WP#: 635
Authors: Etienne Le Rossignol, Nathan Nunn, and Sara Lowes

Abstract: In sub-Saharan Africa, despite the adoption of Christianity, traditional religious beliefs remain widely held. We examine the social consequences of holding traditional religious beliefs among urban and rural populations in central Africa. Using a variety of lab-in-the-field experiments that randomize partner characteristics, we test whether individuals who believe in traditional religion are treated or viewed differently by others. We find that participants act less prosocially towards partners known to hold traditional religious beliefs. We find that this behavior is supported by norms and by negative perceptions and stereotypes of traditional believers. The effects are economically important, ubiquitous, and are amplified by economic development. Individual-level data from across the African continent reveal patterns consistent with our experimental findings. Individuals who believe in witchcraft have lower incomes, and the effect is stronger in countries that are more developed. Our final analysis speaks to the origins of these effects. Within our experimental sample, the negative effects are stronger in rural villages with more historical missionary activity, and across the continent, the negative relationship between belief in witchcraft and income is stronger in regions with more colonial missionary presence. Both findings are consistent with descriptive accounts of Christianity leading to the stigmatization of African traditional religion.

The Complexity of Multidimensional Learning in Agriculture

Date:  Spring 2025
WP#: 634
Authors: Rachid Lajaaj and Karen Macours

Abstract: The adoption of new technologies often requires farmers to re-optimize production decisions over various inputs. Yet few studies examine how exposure to information signals on different inputs translate into dynamic learning and adoption decisions. We randomly invited farmers in Kenya to participate in agronomic research trials, giving them an opportunity to conduct side-byside comparisons of different combinations of inputs during three consecutive seasons. Drawing from detailed data collection during six seasons, we study farmers’ learning process. Farmers react to the exogenous signal by increasing experimentation and their farming know-how increases rapidly. High skill farmers experiment the most as a response to the treatment and learn faster, but also make new mistakes and endure a reduction in short term profit. We present a theoretical model with a multidimensionality of input and practice decisions where complementarities and substituabilities make searching for a better equilibrium costly, and then test the model’s implications.

The Colonial Legacy in India: How Persistent are the Effects of Historical Institutions?

Date:  Spring 2025
WP#: 633
Authors:  Lakshmi Iyer and Coleson Weir

Abstract: Using updated data, we analyze the long-run effects of two British colonial institutions established in India. Iyer (2010) showed that areas under direct colonial rule had fewer schools, health centers, and roads than areas under indirect colonial rule. Two decades later, we find that these differences have been eliminated, and that the gaps in poverty, health and education attainment are also smaller. Banerjee and Iyer (2005) found lower agricultural investments and productivity in areas with landlord-based colonial land tenure systems. Our updated data finds that only some of these differences have been eliminated, while others have remained constant and even widened. Consistent with this lack of convergence, we find that non-landlord areas continue to have higher education attainment and lower poverty rates six decades after the end of colonial rule. We conclude that the impact of colonial institutions can eventually fade away under the influence of targeted policies.

Testing Fractional Doses of COVID-19 Vaccines

Date:  Spring 2025
WP#: 632
Authors: Witold Więcek, Amrita Ahuja, Esha Chaudhuri, Michael Kremer, Alexandre Simoes Gomes, Christopher M. Snyder, Alex Tabarrok, and Brandon Joel Tan

Abstract: Due to the enormous economic, health, and social costs of the COVID-19 pandemic, there are high expected social returns to investing in parallel in multiple approaches to accelerating vaccination. We argue there are high expected social returns to investigating the scope for lowering the dosage of some COVID-19 vaccines. While existing evidence is not dispositive, available clinical data on the immunogenicity of lower doses combined with evidence of a high correlation between neutralizing antibody response and vaccine efficacy suggests that half- or even quarter-doses of some vaccines could generate high levels of protection, particularly against severe disease and death, while potentially expanding supply by 450 million to 1.55 billion doses per month, based on supply projections for 2021. An epidemiological model suggests that even if fractional doses are less effective than standard doses, vaccinating more people faster could substantially reduce total infections and deaths. The costs of further testing alternative doses are much lower than the expected public health and economic benefits. However, commercial incentives to generate evidence on fractional dosing are weak, suggesting that testing may not occur without public investment. Governments could support either experimental or observational evaluations of fractional dosing, for either primary or booster shots. Discussions with researchers and government officials in multiple countries where vaccines are scarce suggest strong interest in these approaches.

Reassessing India’s Poverty Decline over the Missing Decade: 2011-12 to 2022-23

Date: Spring 2025
WP#: 631
Authors: Maitreesh Ghatak and Rishabh Kumar

Abstract: This paper addresses the lack of official data on Indian poverty between 2011-12 and 2022-23, a period known as the "missing decade". It critiques two major attempts to estimate poverty during this time: one using national account growth rates and the other using a private survey. These methods are criticised for their assumptions and data limitations. The paper proposes a new method, imputing consumption from official labour force surveys using a wage-based model. The authors find that poverty reduction was not as dramatic as previously suggested, with about 20% of Indians living in poverty on the eve of the pandemic. The paper's analysis aligns with the fact that India's structural economic transformation was limited, with agriculture output stagnant and regional convergence lacking. The paper concludes that while poverty may have gone down, but its rate of decline has been slow in the decade after 2011-12. 

Meta-Analysis and Public Policy: Reconciling the Evidence on Deworming

Date: Spring 2025
WP#: 630
Authors: Kevin Croke, Joan Hamory Hicks, Eric Hsu, Michael Kremer, Ricardo Maertens, Edward Miguel, and Witold Więcek

Abstract: The WHO recommends mass drug administration (MDA) for intestinal worm infections in areas with >20% infection prevalence. Recent Cochrane meta-analyses endorse treatment of infected individuals but recommend against MDA. We conducted a theory-agnostic random-effects meta-analysis of the effect of multiple-dose MDA and a cost-effectiveness analysis. We estimate significant effects of MDA on child weight (0.15 kg, 95% CI: 0.07, 0.24; p <0.001), mid-upper arm circumference (0.20 cm, 95% CI: 0.03, 0.37; p=0.02), and height (0.09 cm, 95% CI: 0.01, 0.16; p=0.02) when prevalence is over 20%, but not on Hb (0.06 g/dl, 95% CI: -0.01, 0.14; p=0.1). These results suggest that MDA is a cost-effective intervention, particularly in the settings where it is recommended by the WHO. 

Can Digital Aid Deliver During Humanitarian Crises?

Date: Winter 2024
WP#: 629
Authors: Michael Callen, Miguel Fajardo-Steinhäuser, Michael G. Findley, and Tarek Ghani

Abstract: Can digital payments help reduce extreme hunger? Humanitarian needs are at their highest since 1945, aid budgets are falling behind, and hunger is concentrating in fragile states where repression and aid diversion present major obstacles. In such contexts, partnering directly with governments is often neither feasible nor desirable, making private digital payment platforms a potentially useful means of delivering assistance. We experimentally evaluated digital payments to extremely poor, female-headed households in Afghanistan, as part of a partnership between community, nonprofit, and private organizations. The payments led to substantial improvements in food security and mental well-being. Despite beneficiaries’ limited tech literacy, 99.75% used the payments, and stringent checks revealed no evidence of diversion. Before seeing our results, policymakers and experts are uncertain and skeptical about digital aid, consistent with the lack of prior evidence on digital payments for humanitarian response. Delivery costs are under 7 cents per dollar, which is 10 cents per dollar less than the World Food Programme’s global figure for cash-based transfers. These savings can help reduce hunger without additional resources, demonstrating how hybrid partnerships utilizing digital payment platforms can help address grand challenges in difficult contexts.

Outsourcing Policy and Worker Outcomes: Causal Evidence from a Mexican Ban

Date: Winter 2024
WP#: 628
Authors: Alejandro Estefan, Roberto Gerhard, Joseph P. Kaboski, Illenin O. Kondo, and Wei Qian

Abstract: A weakening of labor protection policies is often invoked as one cause of observed monopsony power and the decline in labor's share of income, but little evidence exists on the causal impact of labor policies on wage markdowns. Using confidential Mexican economic census data from 1994 to 2019, we document a rising trend over this period in on-site outsourcing. Then, leveraging data from a manufacturing panel survey from 2013 to 2023 and a natural experiment featuring a ban on domestic outsourcing in 2021, we show that the ban drastically reduced outsourcing, increased wages, and reduced measured markdowns without lowering output or employment. Consistent with the presence of monopsony power, we observe large markdowns for the largest firms, with the decline in markdowns in response to the ban concentrated among high-markdown firms. However, we also find that the reform reduced capital investment and increased the probability of market exit.

Long-run Impacts of Forced Labor Migration on Fertility Behaviors: Evidence from Colonial West Africa

Date: Winter 2024
WP#: 627
Authors: Pascaline Dupas, Camille Falezan, Marie Christelle Mabeu, and Pauline Rossi

Abstract: Is the persistently high fertility in West Africa today rooted in the decades of forced labor migration under colonial rule? We study the case of Burkina Faso, considered the largest labor reservoir in West Africa by the French colonial authorities. Hundreds of thousands of young men were forcibly recruited and sent to work in neighboring colonies for multiple years. The practice started in the late 1910s and lasted until the late 1940s, when forced labor was replaced with voluntary wage employment. We digitize historical maps, combine data from multiple surveys, and exploit the historical, temporary partition of colonial Burkina Faso (and, more specifically, the historical land of the Mossi ethnic group) into three zones with different needs for labor to implement a spatial regression discontinuity design analysis. We find that, on the side where Mossi villages were more exposed to forced labor historically, there is more temporary male migration to Cote d’Ivoire up to today, and lower realized and desired fertility today. We show evidence suggesting that the inherited pattern of low-skill circular migration for adult men reduced the reliance on subsistence farming and the accompanying need for child labor. We can rule out women’s empowerment or improvements in human and physical capital as pathways for the fertility decline. These findings contribute to the debate on the origins of family institutions and preferences, often mentioned to explain West Africa’s exceptional fertility trends, showing that fertility choices respond to changes in modes of production.

Digital Information Provision and Behavior Change- Lessons from Six Experiments in East Africa

Date: Winter 2024
WP#: 626
Authors: Raissa Fabregas, Michael Kremer, Matthew Lowes, Robert On, and Giulia Zane

Abstract: Mobile phone-based informational programs are popular worldwide, though there is little consensus on how effective they are at changing behavior. We present causal evidence on the effects of six mobile-based agricultural information programs implemented in Kenya and Rwanda. The programs shared similar objectives but were implemented by three different organizations and varied in content, design, and target population. With administrative outcome data for over 156,000 people across all experiments, we are sufficiently powered to detect small effects in real input purchase choices. Combining the results of all experiments through a meta-analysis, we find that the odds ratio for following the text-based recommendations is 1.20 (95% CI: 1.14, 1.26). We cannot reject similar effects across experiments and for different agricultural technologies. We do not find evidence of message fatigue or crowd-out of other inputs. The effects, however, seem to diminish over time. Providing more granular information, supplementing the texts with in-person calls, or varying the messages’ framing did not significantly increase impacts, but message repetition had a modest positive effect. While the overall effect sizes are small, the low cost of text messages can make these programs cost-effective.

Revisiting the Eswaran-Kotwal Model of Tenancy

Date: Winter 2024
WP#: 625
Authors: Dilip Mookherjee, Maitreesh Ghatak

Abstract: Persistence of sharecropping tenancy and increases in farm productivity following regulations protecting tenant rights have been observed in many developing countries. This paper examines if these can be explained by alternative models of sharecropping with two sided efforts/investments, namely, complete contract models either without wealth constraints (Eswaran-Kotwal (1985)), or with a wealth constrained tenant (Mookherjee (1997), Banerjee-Gertler-Ghatak (2002)); and incomplete contract holdup models without wealth constraints (Grossman-Hart (1986)). In the absence of wealth constraints the complete contract model always results in (incentive constrained) surplus-maximizing productivity, so there can be no scope for tenancy regulations to raise productivity. In the incomplete contract model, tenancy regulations would raise productivity only if the tenant’s investments are more important than the landlord’s investment. But in that case sharecropping tenancy would not persist in the absence of wealth constraints as the tenant would have purchased the land right ex ante from the landlord. The model with wealth constraints helps explain both the persistence of tenancy and productivity/surplus enhancing effects of tenancy regulations.

Public Service Delivery, Exclusion and Externalities: Theory and Experimental Evidence from India

Date: Fall 2023
WP#: 624
Authors: Alex Armand, Britta Augsburg, Antonella Bancalari, and Maitreesh Ghatak

Abstract: This study explores the interaction between the quality of public services, the implementation of user fees, and the resulting potential for exclusion, that can lead to negative externalities. Our theoretical framework takes account of the possible externalities that result from excluded users accessing alternative options in the context of sanitation, i.e., open defecation, and challenges the conventional wisdom that higher quality unequivocally leads to increased use. Instead, it highlights the ambiguity that results from a simultaneous increase in usage due to improved services (quality effect) and a decrease caused by the fees (price-elasticity effect). We then provide empirical evidence from a randomized controlled trial, where we incentivized the quality of water and sanitation services in the two largest cities of Uttar Pradesh, India. We show that higher service quality increases fee compliance but excludes some users, leading to unintended negative health externalities. Our detailed data provides evidence that results are driven by changes in caretaker behaviour. This finding highlights the need to be cautious regarding user fees, especially for public services involving significant externalities, and in settings where the users are very poor.

Zero-Sum Thinking, the Evolution of Effort-Suppressing Beliefs and Economic Development

Date: Fall 2023
WP#: 623
Authors: Augustin Bergeron, Jean-Paul Carvalho, Jonathan Weigel, Joseph Henrich, and Nathan Nunn

Abstract: We study the evolution of belief systems that suppress productive effort. These include concerns about the envy of others, beliefs in the importance of luck for success, disdain for competitive effort, and traditional beliefs in witchcraft. We show that such demotivating beliefs can evolve when interactions are zero-sum in nature, i.e., gains for one individual tend to come at the expense of others. Within a population, our model predicts a divergence between material and subjective payoffs, with material welfare being hump-shaped and subjective well-being being decreasing in demotivating beliefs. Across societies, our model predicts a positive relationship between zero-sum thinking and demotivating beliefs and a negative relationship between zero-sum thinking (or demotivating beliefs) and both material welfare and subjective well-being. We test the model’s predictions using data from two samples in the Democratic Republic of Congo and from the World Values Survey. In the DRC, we find a positive relationship between zero-sum thinking and the presence of demotivating beliefs, such as concerns about envy and beliefs in witchcraft. Globally, zero-sum thinking is associated with skepticism about the importance of hard work for success, lower income, less educational attainment, less financial security, and lower life satisfaction. Comparing individuals in the same zero-sum environment, we observe the divergence between material outcomes and subjective well-being predicted by our model.

Globalization and Inequality in Latin America

Date: Summer 2023
WP#: 622
Authors: Brian K. Kovak and Rafael Dix-Carneiro

Abstract: We survey the recent literature studying the effects of globalization on inequality in Latin America. Our focus is on research emerging from the late 2000s onward, with an emphasis on empirical work considering new mechanisms, studying new dimensions of inequality, and developing new methodologies to capture the many facets of globalization’s relationship to inequality. After summarizing both design-based and quantitative work in this area, we propose directions for future work. Our overarching recommendation is that researchers develop unifying frameworks to help synthesize the results of individual studies that focus on distinct aspects of globalization’s relationship to inequality.

On the Importance of African Traditional Religion for Economic Behavior

Date: Summer 2023
WP#: 621
Authors: Aimable Amani Lameke, Lewis Dunia Butinda, Max Posch, Nathan Nunn, and Raul Sanchez de la Sierra

Abstract: Within the field of economics, despite being widespread, African traditional religions tend to be perceived as unimportant and ignored when studying economic decision-making. This study tests whether this presumption is correct. Using daily data on business decisions and performance of beer sellers in the eastern Democratic Republic of the Congo, we study the importance of traditional religious beliefs for economic behavior and outcomes. Beer sellers perceive the risk of theft in their shops to be higher than it actually is, causing them to hold lower inventories, more frequent stock-outs, and reduced profits. We facilitate randomly-timed access to commonly-used, but typically prohibitively expensive rituals, which reduce the perceived risk of theft. We find that the rituals partially correct the beliefs about the risk of theft for sellers who report believing in the ritual’s efficacy. These sellers purchase more inventory, experience fewer stock-outs, and have larger sales, revenues, and profits. To distinguish the belief in the efficacy of the ritual from other incidental effects of participation, we analyze these outcomes for sellers who do not believe in the ritual. For these individuals, we find none of the observed effects. The findings provide evidence of the importance of African traditional religions, demonstrating that they can influence behavior and outcomes that are important for economic development.

Chat Over Coffee? Diffusion of Agronomic Practices and Market Spillovers in Rwanda

Date: Spring 2023
WP#: 620
Authors: Esther Duflo, Daniel Keniston, Tavneet Suri, and Celine Zipfel

Abstract: Agricultural extension programs often train a few farmers and count on diffusion through social networks for the innovation to spread. However, if markets are imperfectly integrated, this may also inflict negative externalities. In a two-step experiment of an agronomy training program among Rwandan coffee farmers, we first randomize the concentration of trainees at the village level and then randomly select within each village. Knowledge increased, and yields were 6.7% higher for trained farmers. We find no evidence of social diffusion; instead, control households experienced negative spillovers in high treatment concentration areas, likely because of competition for a scarce input, fertilizer.