Current research

Here are some of the projects that I am currently working on, ranging from projects that are still in the field to those for which data analysis and writing are still underway, all the way to papers that are in revise-and-resubmit at various journals.

  1. “Transformative change through Payments for Ecosystem Services (PES): A conceptual framework, and application to conservation agriculture in Malawi” (with Andrew R. Bell, Tim Benton, Klaus Droppelmann, Lawrence Mapemba, and Oliver Pierson).

Payments for ecosystem services (PES) schemes provide a mechanism to connect beneficiaries of ecosystem services with those whose actions could provide them. Recent work on PES has demonstrated a priority on matching costs and benefits at the margin, where, on the whole, payments are also low. We draw on dynamic systems theory to demonstrate that reinforcing feedbacks that benefit ES producers may warrant much higher initial investments in PES programs, and provide evidence of behavioral drivers from a PES trial in Malawi to support these claims. Specifically, in our study, peer effects and improvement to soil structure are processes that can encourage adoption of sustainable land management practices, alongside or in the absence of other incentives. Under this framing, PES programs can be vehicles to shift systems between basins of attraction over a time-limited period, rather than programmes necessary for long term maintenance of services. (revise and resubmit)

  1. “Drought-tolerant rice, weather index insurance, and comprehensive risk management for smallholders: Evidence from a multi-year field experiment in India” (with Simrin Makhija and David J. Spielman)

In rainfed production systems throughout India, agricultural activities are intimately tied to the summer monsoon, and any aberration in monsoon rainfall patterns—late onset or early cessation of rainfall, prolonged dry spells, or flooding—can have severe consequences for rice production. There is considerable public policy interest in designing programs to lower small-scale farmers’ exposure to these types of risk given the regularity with which adverse monsoon events occur. This paper introduces a field experiment conducted with two risk management options in the state of Odisha: a drought-tolerant (DT) rice cultivar and a weather index insurance (WII) product designed to complement the performance profile of the DT cultivar. Uptake rates for the DT cultivar itself and for the joint DT-WII product are compared across two years alongside an analysis of factors that predict uptake. Results indicate high levels of demand for both the DT cultivar and DT-WII product, albeit with a significant degree of price sensitivity. Sustained demand between years one and two is primarily explained where individuals received a payout in year one and had a large number of peers within the village also purchasing the product. Results also indicate that the withdrawal of discounts introduced in year one did not affect demand significantly in year two, providing some of the first evidence of moderate take-up of WII at rates at or above the actuarially fair cost. (manuscript being prepared for submission)

  1. “Ex ante and ex post effects of hybrid index insurance Bangladesh” (with Ruth Vargas Hill, Neha Kumar, Simrin Makhija, Nicholas Magnan, Francesca de Nicola, and David J. Spielman)

This study assesses the demand for and effectiveness of an index insurance product designed to help smallholder farmers in Bangladesh manage crop production risk during the monsoon season. Villages were randomized into either an insurance treatment or a comparison group, and discounts and rebates were randomly allocated across treatment villages to encourage insurance take-up and permit the estimation of the price-elasticity of demand. Among those offered insurance, demand was fairly price elastic, with discounts significantly more successful in stimulating demand than rebates. Purchasing insurance generates both ex ante risk management effects and ex post income effects on agricultural production. Risk management effects lead to an expansion of cultivated area and increases in input expenditures during the monsoon season. Income effects lead to more intensive rice production during the subsequent dry season, with more intensive use of both irrigation and fertilizers, resulting in higher yields and higher total rice production. (revise and resubmit)

  1. “Farmer preferences for Cotton Leaf Curl Virus-resistant cultivars: Empirical evidence from Pakistan” (with Shahzad Kouser and David J. Spielman)

Cotton leaf curl virus (CLCuV) is one of the major biotic constraints to cotton production in Pakistan. Many cotton producers consider CLCuV a significant threat to their crop, rivaling only bollworm infestation, and various research stations are struggling to develop resistant cultivars. This study aims to help crop developers by investigating farmers’ preferences for different attributes of CLCuV-resistant cultivars. We conduct a discrete choice experiment and a household survey of 551 cotton farmers from all cotton-producing zones of Punjab and Sindh Provinces in eastern Pakistan. A mixed logit model is used to incorporate preference heterogeneity in the stated choice analysis. Empirical findings suggest that farmers are willing to pay for cultivars that protect cotton yields against increasing disease stresses. We suggest that considerable scope exists for the rapid development and widespread adoption of such cultivars, even if marketed at prices significantly higher than current cottonseed prices. (under review)

  1. “Can differences in individual learning explain patterns of technology adoption? Evidence on heterogeneous learning patterns and hybrid rice adoption in Bihar, India” (with Jared Gars)

Much empirical research that has shown that an individual’s decision to adopt a new technology is the result of learning — both in personal experimentation as well as observing the experimentation of others. Yet even casual observation would suggest significant heterogeneity learning processes, manifesting itself in widely varying patterns of adoption over space and time. In this paper we explore this heterogeneity in the context of early adoption of hybrid rice in rural India. Using specially-designed experiments conducted as part of a primary survey in the field, we are able to identify which of four broad learning heuristics most accurately reflects individuals’ information processing strategies. Linking these learning heuristics with observed use of rice hybrids, we demonstrate that pure Bayesian learning is well suited for the tinkering and marginal adjustments that would be required to learn about a technology like hybrid rice, but is also more cognitively taxing, requiring a longer memory and more complex updating processes. Consequently, only about 25 percent of the farmers in our sample can be characterized as pure Bayesian learners. Present-biased learning and relying on first impressions will likely hinder adoption of a technology like hybrid rice, even after controlling for access to credit and a rudimentary proxy for intelligence. (under review)

  1. “Understanding information processing heuristics to target development interventions: The case of drought risk management interventions in Bangladesh” (with David L. Ortega and Vincenzina Caputo)

We use stated preference data to study farmer demand for two alternate strategies to manage risks associated with drought in Bangladesh. A choice experiment regarding drought tolerant rice varieties and weather index insurance was administered to 2,306 farmers and the data were analyzed using an equality constrained latent class model specification which captures both preference heterogeneity and information processing behavior. Two overarching preference classes or meso-groups were identified, each containing various patterns of attribute non-attendance. We utilize group membership probabilities and covariates to identify and describe farmers in these two groups as well as derive program design and policy implications. (revise and resubmit)

  1. “Valuation and aspirations for drip irrigation in Punjab, Pakistan” (with Andrew R. Bell, Muhammad Ashfaq, and Stephen Davies)

Drip irrigation is a high-efficiency technology that improves water-use efficiency while simultaneously transforming areas that are not otherwise irrigable in practice (too distant or too high to be reached by surface waters). While drip irrigation is expanding rapidly in India, adoption remains low in neighboring Pakistan. We employed a discrete choice experiment framed around the hypothetical subsidized purchase of a drip irrigation system in four districts within Punjab province in eastern Pakistan. Our results suggest a clear increase in valuation of drip systems in the first several years following adoption, suggesting that farmers may initially be unaware of the opportunities for the use of drip on their farms or the benefits that may accrue. We observed that aspirations for cropping systems under drip were better predictors of the valuation of drip systems than were current cropping patterns, implying that a different agricultural landscape might reasonably emerge under more widespread adoption of drip. Aspirations differed across the different agro-ecological zones and water regimes captured by our study. Aspirations to substitute high-value fruits and vegetables for low-value wheat crops were associated with higher valuation of the subsidy level, while aspirations to expand wheat were associated with higher valuation of the area covered by the drip subsidy; together, these imply a degree of control over the extent of wheat production in the landscape via careful design of the drip subsidy program. While the penetration of drip irrigation is not sufficient yet to draw inferences from a representative sample, these results suggest a number of ways through which drip irrigation may transform Pakistan’s agricultural landscape.  (under review)

  1. “Can information help reduce imbalanced application of fertilizers in India? Experimental evidence from Bihar” (with Jared Gars, Ram Fishman, Avinash Kishore, and Yoav Rothler)

The imbalanced application of chemical fertilizers in India is widely blamed for low yields, poor soil health, pollution of water resources, and large public expenditures on subsidies, amounting to about 1 percent of India’s gross domestic product. To address the issue, the government of India is investing in a large-scale, expensive program of individualized soil testing and customized fertilizer recommendations, with the hope that scientific information will lead farmers to optimize the fertilizer mix. We conducted a randomized controlled trial in the Indian state of Bihar in what we believe to be the first evaluation of the effectiveness of the program as currently implemented. We found no evidence of any impact of customized fertilizer recommendations on fertilizer use. The lack of impact can be attributed to several factors, including a lack of understanding, lack of confidence in the information’s reliability, or other factors such as fertilizer costs that inhibit farmers from optimizing fertilizer application ratios even if the information shifts their underlying preferences. We provide evidence that suggests lack of confidence is the main factor inhibiting farmers’ response. (manuscript being prepared for submission)

9. “Intrahousehold valuation, preference heterogeneity, and demand of an agricultural technology in Bihar, India” (with Kajal Gulati, Travis Lybbert, and David J. Spielman)

Both women and men within an agricultural household potentially have a role in the adoption of labor-saving agricultural technologies, particularly when the technology differentially their on-farm labor. One such technology is mechanical rice transplanting (MRT), a new agricultural technology that significantly reduces labor demand during rice transplanting — an agricultural activity primarily reserved for women in many parts of India. In this paper we examine some of the intrahousehold decisionmaking dynamics that shape the household’s decision of whether or not to adopt this technology. First, we elicit individual willingness-to-pay for MRT from male and female decisionmakers from the same households. Second, we use a village-level experimental auction to measure the household’s willingness-to-pay for the technology. Women value MRT more than men and this demand heterogeneity is not driven by differences in their individual characteristics. Women also tend to select plots that save their unpaid labor more as compared to plots chosen by men. Despite women valuing MRT more than men, they do not significantly influence the household’s technology adoption decision. The intrahousehold differences in valuation disappear when women engage in hired wage work, suggesting that women value MRT in order to potentially reallocate farm labor to other unpaid family work. These results have implications for rural labor market welfare because agricultural and labor productivity gains due to MRT adoption may push women into more traditional gendered labor divisions. (manuscript being prepared for submission)

10. “Behavioral constraints to the adoption of nutrition-sensitive food production systems” (with Muzna Alvi, Simrin Makhija, and David J. Spielman) (research ongoing)

11. “Leveraging network externalities to promote soil conservation and ecosystem services: A randomized evaluation of the impacts of agglomeration payments on the diffusion of conservation agriculture in the Shire River Basin in southern Malawi” (with Andrew R. Bell, Lawrence Mapemba, Klaus Droppelmann, and Tim Benton) (research ongoing)

12. “Bundling technology with insurance: Evidence from a randomized evaluation in Odisha, India” (with David J. Spielman and Simrin Makhija) (research ongoing)

13. “Innovative financial products to relax quantity and risk rationing in access to rural credit in Kenya” (with Liangzhi You, Yanyan Liu, and Apurba Shee) (research ongoing)