Randomized evaluations of the agronomic productivity gains from new crops or agricultural techniques have been common in the agricultural field for many years. More recent is an approach to agriculture that aims to conduct ‘effectiveness’ trials, incorporating real-world issues of access and adoption among smallholder farmers, rather than the idealized ‘efficacy’ trials produced using experimental test plots. Tackling the impacts of agricultural interventions outside of the test plot introduces issues at the heart of economics, such as transaction costs, social interactions, marketing, finance, and contracting as we think carefully about the decision to adopt. Thinking of the smallholder farm as a small business, this decision should be driven by profitability. The core contribution of RCTs is their ability to clearly trace causality between the constraints to agricultural technology adoption, adoption itself and final outcomes . Randomized experimental evaluations allow researchers to isolate the causal impact of a program from other confounding factors—such as price, weather, or access to credit—which are simultaneously changing over time and across regions 2 . Carefully designed experiments allow us to identify whether specific constraints to adoption are binding, and measure the impacts of a technology when adopted in farmers’ actual fields. These evaluations speak to the effectiveness of specific approaches to achieving agricultural technology adoption for improved smallholder productivity and welfare.The Agricultural Technology Adoption Initiative was founded in 2009 to increase the quantity and quality of experimental evidence in developing-country agriculture. ATAI aims to serve as a mechanism to generate, aggregate,livestock fodder system and summarize research for policy outreach on the adoption of agricultural innovations by smallholders in Sub-Saharan Africa and South Asia.
ATAI exclusively funds randomized controlled trials, and pilot work that lays the groundwork for future RCTs, and was organized intellectually around understanding how a set of specific constraints held back technology adoption. Because of this methodological focus, the resulting evidence is primarily on interventions targeted at the individual or household level, although we also report on studies in areas such as input and output markets that attempt to drive outcomes at more aggregated levels. Even within this domain, we have a distribution of studies that is purposive, driven by the questions asked by our affiliated investigators, and by the technical feasibility of running randomized trials. We use the structure of the ATAI constraints to adoption to help summarize the experimental evidence, aggregating individual, internally valid studies around these common themes. This produces an evidence base that is far from comprehensive in terms of the important issues in agricultural development, but is broader than would have been produced by a more tightly structured replication-focused research initiative and does provide a relatively clear guide to what makes specific interventions attractive in terms of evidence-based funding. Throughout the world, 63% of those living under $1.25 per day are working in agriculture .Ligon and Sadoulet show the importance of economic growth in the agriculture sector for the livelihoods of the poorest households: a one percent growth in GDP that originates from agriculture correlates with a 5.6 percentage point increase in expenditures among the poorest decile of the population, a 4.45 percentage point increase for the bottom 30%, while “growth from non-agriculture sectors does not appear to have a significant effect on expenditure growth for the poorest 50%.” The Green Revolution of the 1960s saw the spread of agricultural technologies to less industrialized nations, and large agricultural productivity gains particularly in East Asia.
Yet technological innovations have not similarly spread to transform agricultural productivity in Sub-Saharan Africa and parts of South Asia as evident in the lagging adoption of modern varieties and a persistent yield gap between regions. Many African countries have rising private sectors developing agricultural technologies, and research and implementation groups including the CGIAR centers and AGRA continue to develop improved inputs and interventions designed to improve the resilience, profits, and nutrition of African smallholders in particular. Yet these innovations do not appear to have translated into meaningful improvements in yields at the macro-level. FAOSTAT data shows a large gap between low per hectare cereal yields in Africa and South Asia which are on average roughly one third of the per hectare yields in East Asia and OECD countries. Sub-Saharan Africa is particularly lagging behind. In South Asia, land use for cereal production has increased 20% while yields have tripled. In Sub-Saharan Africa, land use for cereal production has more than doubled, while yields have increased by just 80% . The macro picture of fertilizer use over time similarly looks unchanged, with low and stagnant use of fertilizers in mainly rainfed areas like SubSaharan Africa. Fertilizer consumption remains extremely low in SubSaharan Africa compared to other regions. Roughly 16 kilos of fertilizer are used per hectare in SubSaharan Africa, and among all developing countries the average is 26.75 kg/hectare. This figure is much higher in other regions: 344 kg/hectare in East Asia/Pacific, and 159 kg/hectare in South Asia.This clearly demonstrates that the status quo of agricultural production, particularly in Sub-Saharan Africa, remains far below the technological frontier, suggesting missed potential in terms of yields, income, and welfare improvements to food security and nutrition. The specific reasons behind lagging adoption of productivity enhancing technological innovations and persistent yield gaps in rainfed Sub-Saharan Africa and South Asia relative to the rest of the world have been a puzzle in need of policy solutions. Field experiments help us move beyond test plots to explain the continuing puzzle of low technology adoption by smallholder farmers in rainfed areas where agriculture is performing well below the technological frontier. Focusing at the micro-economic level of this challenge, we focus on technology adoption as an outcome that inherently requires smallholder farmers to change their practices.
Behavior changes can include, for example, the adoption of resilient and high-yielding crop varieties or a shift to high-value crops, the purchase and application of complementary inputs such as fertilizers, and the adjustment of farm labor allocated toward specific agronomic practices. Many smallholder farmers face barriers to adopting effective agricultural technologies. These constraints to adoption may be driven by standard economic factors , or may be behavioral . Standard economic explanations consider smallholder farmers as economic agents, building from the conception that “in a well-functioning economy where markets perfectly capture all costs and benefits, and individuals are fully informed and unconstrained, farmers will adopt a technology if they make a profit from adopting it” . This is an important distinction from a world where farmers focus their efforts to maximize their productivity, for example, their crop yields, given increased yields do not necessarily lead to improved welfare. Profitability can be limited by input costs, credit constraints, and market access. Information and labor constraints are also relevant — how well do farmers understand the properties of new technologies, in the absence of opportunities to experiment? What are the additional labor requirements for the use of these new technologies, and how do farmers value their time in input decisions? Jack reviews in detail other dimensions that mediate whether certain technologies “meet the expected profitability condition” for specific farmers. This varies temporally and spatially . This also varies between and within households,hydroponic nft gully particularly when complementary asset or capital investments are needed, or new technologies challenge individual tastes and preferences. Even where markets are functioning well, accessible and profitable technologies may not be adopted for behavioral reasons, such as risk or uncertainty aversion or procrastination, which challenge decision-making even in the best of circumstances. Smallholders’ decision-making is highly complex and conducted in risky and low resource environments. Farmers make interconnected choices over long time frames that are characterized by risks and uncertainty. One of many choices is among a range of potential inputs to production , in contexts with highly variable land, wide ranging and seasonal climatic variation that is growing increasingly extreme given climate change, and unpredictable shocks to their livelihood. New technologies may change the risk or payoff profiles of farming in ways that require us to incorporate other social science insights, for example expected utility theory and behavioral economics, in order to understand perceived benefits at the farmer level. Motivated by addressing the constraints hindering the adoption of new agricultural technologies, ATAI has worked to fund and structure the experimental evidence base across seven primary market inefficiencies that constrain adoption. These are credit5 , risk, information, input and output markets, labor and land market inefficiencies, as well as externalities . These may operate through supply or demand channels, for example by limiting the availability of technologies, information, or financing, and/or dampening demand by lowering expected profits. Lessons from psychology and behavioral economics are considered where they are particularly relevant. Jack motivates the focus on constraints to adoption, rather than specific technologies, as a framework that helps identify effective strategies to address common inefficiencies and constraints in order to encourage the adoption and use of more than one technology. ATAI uses this conceptual framework of seven constraints to drive its research competitions.
Randomized evaluations are selected for ATAI funding based not only on methodological rigor, logistical viability, and innovation, but also on their potential for both a significant contribution to public knowledge, and practical influence and scalability in related contexts. Field experiments require, by their very nature, durable partnerships with real-world implementation groups that are working directly with smallholder farmers in order to randomize interventions and deliver credible results. Partner organizations may work as agro-dealers, contract farming groups, extension agents, financial service providers, technology developers, or otherwise. ATAI views more favorably studies that evaluate questions of key importance to large-scale program and policy partners, particularly those that are difficult to address without causal evidence, and those that have received less research attention to date. To meet these criteria, technologies under investigation are those where there is credible field data signaling that adoption would prove neither distasteful nor ineffective in target farmers’ contexts, and that the take-up and use of a technology is likely to prove utility-enhancing, profitable, and welfare-increasing for smallholder farmers and others along agricultural value chains. For such promising under-adopted technologies, ATAI funds social science field experiments to provide evidence on the strategies that work in helping farmers adopt, and ultimately benefit from, these technologies. In the sections that follow, we summarize particular components of the evidence base given the accumulation of ATAI-generated experimental evidence in four areas: credit and savings, risk, information, and input and output market inefficiencies. This does not imply that the latter three constraints to adoption, i.e. externalities and land and labor markets, are excluded from this chapter because they do not bind or do not deserve further investigation. These topics are not covered here simply because there is less rigorous micro-evidence given the difficulty of examining them through the lens of RCTs6 . This is not intended to be an exhaustive review. ATAI-funded studies are often presented in greater detail given our familiarity with their contributions. Each section begins by motivating the specific constraint to Agricultural income streams are characterized by large cash inflows once or twice a year that do not align well with specific times when farmers need access to capital to either make agricultural investments or, for example, pay school fees. If there is limited access to credit in an area, farmers may not have cash on hand to make agricultural productivity investments unless they are able to save, or can afford the potentially high interest rates of informal lending. However, saving can be difficult for farmers given their limited resources, a variety of demands on their money, and the seasonal cycle of production and prices of their agricultural production. Credit and saving products could help farmers make investments in inputs and other technologies by making cash available when needed. Yet many developing countries, and particularly rural areas, have limited access to formal financial services that could provide this liquidity. Credit constraints have been reflected in farmers self-reports , and are associated with less use of productive inputs like high-yielding varieties . On the supply side, formal financial service providers are often unwilling or unable to serve smallholders.