Projected Benefits of CGIAR Research CGIAR is a global research partnership for a food-secure future dedicated to transforming food, land, and water systems in a climate crisis. www.cgiar.org The Alliance of Bioversity International and the International Center for Tropical Agriculture (CIAT) is part of CGIAR. It delivers research-based solutions that address the global crises of malnutrition, climate change, biodiversity loss and environmental degradation. The Alliance focuses on the nexus of agriculture, nutrition and environment. We work with local, national and multinational partners across Africa, Asia, Latin America and the Caribbean, and with the public and private sectors and civil society. With novel partnerships, the Alliance generates evidence and mainstreams innovations to transform food systems and landscapes so that they sustain the planet, drive prosperity and nourish people in a climate crisis. https://alliancebioversityciat.org CIMMYT works in the developing world to improve people’s livelihoods and promote more productive and sustainable maize and wheat systems. Our portfolio is squarely focused on critical issues such as food insecurity and malnutrition, climate change and environmental degradation. Through collaborative research, partnerships, and training, the center helps create and strengthen a new generation of national agricultural research and extension services in maize- and wheat-producing countries. As a member of the CGIAR, made up of 15 agricultural research centers, CIMMYT coordinates the Maize and Wheat Research Programs, which bring together and add value to the efforts of more than 500 partners and collaborators. https://www.cimmyt.org Projected Benefits of CGIAR Research Steven D. Prager Gideon Kruseman Alliance of Bioversity International and the International Center for Tropical Agriculture (CIAT) Americas Hub Km 17 Recta Cali-Palmira. C.P. 763537 A.A. 6713 Cali, Colombia Telephone: (+57) 602 4450000 Website: https://alliancebioversityciat.org/ Citation Prager SD; Kruseman G. 2022. Projected Benefits of CGIAR Research. CIAT Publication No. 529. International Center for Tropical Agriculture (CIAT). Cali, Colombia. 60 p. Steven D. Prager, Principal Scientist, Climate-Resilient Food Systems, Alliance of Bioversity and the International Center for Tropical Agriculture, and CGIAR Foresight Team Co-Leader. s.prager@cgiar.org Gideon Kruseman, Ex-ante Impact Assessment and Foresight Research Leader, International Maize and Wheat Improvement Center (CIMMYT). g.kruseman@cgiar.org © CIAT 2022. Cover photos: (1) Bioversity International, (2) Stevie Mann/ILRI, (3) P. Lowe/CIMMYT Some Rights Reserved. This work is licensed under a Creative Commons Attribution 4.0 International License (CC-BY 4.0) https://creativecommons.org/licenses/by/4.0/ Copyright © CIAT 2022. Some rights reserved. June 2022 Contributor acknowledgement This report involved the direct and indirect contributions of many individuals. Key contributors in addition to the authors include Keith Wiebe (CGIAR Foresight team co-leader); Ricky Robertson, Will Martin, Karl Pauw, Tim Sulser, and Nicola Cenacchi of the International Food Policy Research Institute (IFPRI); the Alliance of Bioversity and CIAT Foresight team, including Camila Bonilla, Ben Schiek, and Carlos González; and the CIMMYT Foresight team, including Adair Zepeda Villarreal, Kai Sonder, Jorge Andrés Redondo Arévalo, Melissa Bonilla Barillas, Marcelo Godoy Barillas, José David González Merlin, and María Erla Barillas Santos. Special recognition Special thanks go to Tek Sepkota and Ivan Ortiz Monasterio of CIMMYT for commenting on the fertilizer analysis results. We thank Julian Ramirez-Villegas (Alliance of Bioveresity and CIAT), Todd Rosenstock with the Center for International Forestry Research (CIFOR) and World Agroforestry (ICRAF), Peter Steward (CIFOR/ICRAF), Philip Thornton (the International Livestock Research Institute (ILRI), Ana María Loboguerrero and Andy Jarvis (Alliance of Bioversity and CIAT) for their contributions from a recent analysis titled “Projected climate adaptation benefits of One CGIAR,” much of the substance of which is presented in the Adaptation section and the corresponding Appendix. In-depth analysis of the projected benefits of poverty reduction was conducted by Channing Arndt (IFPRI), and these data were triangulated with additional support and models provided by Keith Fuglie (United States Department of Agriculture). Finally, we gratefully acknowledge both the financial and intellectual support of the CGIAR System Management Office, the guidance and insights from Action Area Technical Team leaders, and the critical supporting contributions of the extended CGIAR Foresight team. The extended CGIAR Foresight team is sponsored by the CGIAR Research Program on Policies, Institutions, and Markets (PIM) Cluster of Activity on Strategic Foresight and under the direction of Steven Prager and Keith Wiebe. Funding acknowledgement This work is a summative product of contributions and financial support from a variety of sources. The work was principally financed by the CGIAR Research Program on Policies, Institutions and Markets (PIM) Cluster of Activity 1.1 on Strategic Foresight. Major contributions also came from the Cluster of Activity 1.1 of both the CGIAR Research Programs on Maize and Wheat, as well as from the CGIAR Research Program on Climate Change, Agriculture and Food Security (CCAFS). Additional funding was provided by the CGIAR System Management Office. Neil Palmer / CIAT Contents List of acronyms 8 Executive summary 9 1. Introduction 12 1.1. Return on investment in agricultural research for development 14 1.2. The agrifood system 14 2. Foresight approaches for understanding potential investments 17 2.1. What is foresight? 17 3. Socio-technical innovation bundles 19 4. Inputs, models, and tools 21 4.1. The International Model for Policy Analysis of Agricultural Commodities and Trade 23 4.2. The Global Innovation Fund approach 24 4.3. The Agrifood Data Analysis Modeling framework 25 4.4. Geographic information system analysis of climate hotspots 25 4.5. Rural Investment and Policy Analysis 25 5. Results 27 5.1. Nutrition, health, and food security 28 5.2. Poverty reduction, livelihoods, and jobs 31 5.3. Gender, youth, and social inclusion 36 5.4. Climate adaptation and greenhouse gas reduction 41 5.5. Environmental health and biodiversity 47 6. Conclusion 52 References 53 Appendices 56 Appendix 1. Fish and aquaculture systems 56 Appendix 2. Employment generation 58 Appendix 3. Additional information regarding greenhouse gas emissions and adaptation 59 List of acronyms ADAM Agrifood Data Analysis Modeling IFPRI International Food Policy Research framework Institute ASEAN Association of Southeast Asian Nations IMAGE Integrated Model to Assess the Global BAU Environment Business as usual CGE IMPACT International Model for Policy Analysis of Computable general equilibrium Agricultural Commodities and Trade CH4 Methane K2O Potassium oxide CIFOR Center for International Forestry Research LAC Latin America and the Caribbean CIMMYT International Maize and Wheat Improvement Center LandSHIFT Land Simulation to Harmonize and Integrate Freshwater Availability and the CIS Climate information services Terrestrial Environment CO2 Carbon dioxide MAGNET Modular Applied GeNeral Equilibrium Tool CO2eq Carbon dioxide equivalent MIRAGE Modelling International Relationships in COMP Comprehensive investment scenario Applied General Equilibrium CSA Climate-smart agriculture N Nitrogen CWANA Central and West Asia and North Africa N2O Nitrogen dioxide ERA Evidence for Resilient Agriculture NoCC No climate change ESA East and Southern Africa P2O5 Phosphorous pentoxide FAO Food and Agriculture Organization of the PPP Purchasing power parity United Nations R&D Research and development GDP Gross domestic product RAFS Resilient Agrifood Systems action area GHG Greenhouse gas RIAPA Rural Investment and Policy Analysis GI Genetic Innovation action area RMM Reduced Marketing Margins (e.g., improved GIF Global Innovation Fund infrastructure) GLOBIOM Global Biosphere Management Model SA South Asia HIGH+NARS Futures scenario characterized by yield SAM Social accounting matrix improvement and investment in NARS SEA Southeast Asia and Pacific capacity SSP Shared Socioeconomic Pathway HIGH+RE Futures scenario characterized by yield improvement and research efficiency ST Systems Transformation action area improvements WCA West and Central Africa ICRAF World Agroforestry IFAD International Fund for Agricultural Development 8 Executive summary Two generations ago, the challenge facing agriculture was clear: the world needed to efficiently increase staple food production to meet rapidly rising demand and stave off mass starvation. CGIAR helped meet this challenge by developing high-yielding varieties, which spurred the green revolution. Over time, new problems arose, and current challenges facing food, land, and water systems are more numerous, daunting, and urgent than ever. A pressing example is the need to stay within planetary boundaries while meeting society’s demands for affordable, acceptable, adequate, and nutritious diets. The agrifood system is complex and dynamic. It is currently impossible to understand the agrifood system and all its details and emergent properties. Yet, we can make inroads by using abstractions, simplifications, and ordering principles that are fit-for-purpose. Foresight approaches offer the capability to understand the potential impacts and benefits of investments in international agricultural research for development. This report focuses on identifying plausible futures and potential returns on investment in the form of projected benefits for the One CGIAR research portfolio. Literature reviews indicate that returns are high. One meta-analysis of 292 studies finds an average rate of return of 100% on agricultural research after excluding extremely high outliers (Alston et al. 2000). In another study of 84 counties, the authors estimate that a 10-percentage-point increase in adoption of high-yielding varieties increases gross domestic product (GDP) per capita by about 15%. While macro-level returns on investment in agricultural research are clear, evidence linking technology adoption to improved agricultural productivity at the farm level is more difficult to find (Loevinsohn et al. 2013); studies are underway to enable estimations of returns on investment for specific crops and the corresponding socio-technical bundles. “Socio-technical innovation bundles” (Barrett et al. 2020a) are composed of elements of different initiatives and contributions from various non-governmental organizations, private-sector companies, national governments, and line agencies (Chapter 3). For One CGIAR, the four pillars of socio-technical innovation bundles include innovations in the following areas: (1) in core CGIAR technologies such as genetic improvements and enhanced agronomic practices; (2) in technology and systems that support these core contributions and offer additional, standalone functionality, such as digital capabilities and financial instruments; (3) in institutions, which may benefit from capacity building and enhanced collaboration and foresight; and (4) in the enabling environment, which derives advantages from better policy and broader inclusion. These four pillars of innovation are then strategically integrated through sets of activities in the six CGIAR regions that aim to achieve beneficial outcomes and impacts. The five crucial One CGIAR impact areas are: (1) nutrition, health, and food security; (2) poverty reduction, livelihoods, and jobs; (3) gender equality, youth, and social inclusion; (4) climate adaptation and mitigation; and (5) environmental health and biodiversity (CGIAR 2021). The findings presented in this report touch on each of these five areas. There are several different models and tools that we use to better understand plausible futures and corresponding projected benefits. These integrated assessment models are complex and use large sets of input data, with key parameters and mega-trends driving model outcomes. Though a representation of an uncertain future, these models have found broad acceptance in the scientific and policy communities as tools to better inform decision-makers. While many models exist, the International Model for Policy Analysis of Agricultural Commodities and Trade (IMPACT) is well suited for serves as the basis for this study, which also discusses other modeling and analytical options, some of which helped frame the present analysis—for instance, in anticipation of later application of the Global Innovation Fund approach. Other models and information resources also contributed key results, including the Agrifood Data Analysis Modeling framework, the Spatial Production Allocation Model database on crop location in combination with the Mink global gridded crop modeling framework, and the Rural Investment and Policy Analysis modeling and data system. The first CGIAR impact area, improvements in the nutrition, health, and food security of the population, is paramount to overall human well-being and economic development. When it comes to food security, IMPACT results demonstrate that a comprehensive, integrated investment strategy has the potential to shrink the population at risk of hunger and reduce the number of malnourished children in all CGIAR regions. In addition, as regards nutrition and health, in all regions, micronutrient availability improves, with proportionally larger projected benefits in South Asia and Africa. Poverty reduction, the second major CGIAR impact area, is a multi-faceted concept, and disentangling its components is challenging. The differences between the baseline and 2030 simulation results indicate a significant rise in the per-hectare gross value of food crop production that, depending on context, is considered a proxy for poverty alleviation. A review Projected Benefits of CGIAR Research 9 of scientific literature also suggests that poverty decreases as per-capita GDP increases. To gain additional perspective on potential benefits of the One CGIAR portfolio with respect to poverty, we also employed two absolute poverty lines: firstly, the well-known US$1.90 per day line, and secondly, separate, regional poverty lines set at 40% of per-capita income in 2030 for each of the six regions in focus for One CGIAR. As expected, percentage point poverty reductions for the region- specific absolute lines are much higher because many more people live “near” but below these higher poverty lines. Analysis suggests that increments to income delivered by CGIAR move a much larger share of people across regionally adjusted poverty levels. Gender, youth, and inclusion—the third major impact area—are key themes in agricultural development that are both affected by and influence the outcomes of investments in research. A recent study shows that for 8 of 14 analyzed countries in Africa, investments in agricultural research increased female employment more than male employment (Frija et al. 2020). The One CGIAR impact indicators call for targeted improvement in women’s empowerment and inclusion in the agricultural sector, in the number of women and youths benefiting from relevant CGIAR innovations, and in the number of women helped to exit poverty. These data support the idea that agriculture can reduce gender disparities. This is a key area where ongoing refinement to foresight models will support a great deal of additional insight. In keeping with One CGIAR’s fourth key impact area, opportunities to increase climate adaptation and reduce emissions are associated with agriculture from farm to fork. There are a wide variety of approaches—from climate services to improved agronomy—that enhance farmer productivity and resilience while helping the food system function within environmental limits (Springmann et al. 2018). Projected One CGIAR benefits result from the generally positive consequences of greater productivity per unit of input, offsetting land conversion, and changing diets. Negative results tend to be associated with heightened methane emissions from rice production and increased use of nitrogen. These favorable and unfavorable impacts are heterogeneous across regions, underscoring that effective mitigation of agriculture-related emissions will require regionally tailored investment strategies. For the vast majority of the six CGIAR target regions where CGIAR-mandate crops are grown, climate change will negatively impact agriculture without adaptation (Figures 5.4.2–5.4.9). In the CGIAR research portfolio, the genetic innovation action area focuses on breeding for climate change adaptation to generate varieties that meet new abiotic stress levels while maintaining or improving yield levels. This suggests that, with One CGIAR investment, the “worst case scenarios” are unlikely to materialize in our target areas (Figures 5.4.2–5.4.9). In terms of the effects of climate adaptation, at 25% and 24%, respectively, India and China contribute circa 50% of the total number of beneficiaries (Ramirez et al. 2021). By 2030, a total of 234.1 million people and 59.1 million households could benefit from CGIAR’s climate adaptation work. Environmental health and biodiversity—which together comprise the fifth and final CGIAR impact area—span many dimensions of the One CGIAR agenda. From agro-biodiversity and biodiversity more generally to deforestation, ecosystem services, and water use, a healthy natural environment supports the long-term sustainability of the food system. Results suggest modest decreases in water for irrigation and complimentary augmentations of green water use from precipitation throughout the CGIAR regions. High-level investment scenarios also substantially relieved pressure on biodiversity and reduced species loss to below the levels expected with climate change alone, and additionally had favorable impacts on deforestation (Table 5.5.2) (Rosegrant et al. 2017). Fertilizer use is also associated with environmental pressures. Across the six CGIAR target regions, the former Soviet Union, and other high-income countries, fertilizer consumption rates are veering upwards. While appropriate fertilizer use is vital for improving productivity, over and improper uses are common. These can have negative impacts on ecosystem health if mismanaged, however, due to run-off into freshwater systems, which can also cause algae blooms and fish kills when rivers drain into coastal areas. Further research is underway about the potential trade-offs for environmental health and biodiversity entailed by CGIAR investments. 10 Alfonso Cortés / CIMMYT Projected Benefits of CGIAR Research 11 1. Introduction Key messages A major challenge facing agriculture today is to stay within planetary boundaries while meeting society’s demands for affordable, acceptable, adequate, and nutritious diets. To this end, the One CGIAR research portfolio has five main impact areas: (1) nutrition, health, and food security; (2) poverty reduction, livelihoods, and jobs; (3) gender equality, youth, and social inclusion; (4) climate adaptation and mitigation; and (5) environmental health and biodiversity. Investment in agricultural research gives high rates of return and strong growth in per-capita gross domestic product (GDP) based on a literature review. The agrifood system, however, is complex, dynamic, and impossible to understand fully. Still, ex-ante assessment and foresight approaches enable us to project the benefits of investments in agricultural research, and conversely the potential negative consequences of not making these investments. A. Camacho / Bioversity International Two generations ago, the challenge facing agriculture Box 1.1. CGIAR was clear: the world needed to efficiently increase staple food production to meet rapidly rising demand and stave For fifty years, CGIAR has been combatting hunger through research off mass starvation. CGIAR helped meet this challenge by developing high-yielding varieties, which spurred the for development. Distinguished by world-class research facilities, green revolution. This innovation led to the CGIAR’s rise extensive partnerships, global reach, decades-long experience, and as a major player in international agricultural research for vast gene banks, CGIAR is now broadening its scope to offer innovative development (Box 1.1). Yet new problems have emerged, solutions that enhance global food systems while addressing the perils and current challenges facing food, land, and water systems of climate change. To achieve this goal, CGIAR targets five primary impact are more numerous, daunting, urgent, and complex than areas: (1) nutrition, health, and food security; (2) poverty reduction, ever before. They are also interrelated and span multiple livelihoods, and jobs; (3) gender equality, youth, and social inclusion; dimensions, including poor diets and health, poverty, (4) climate adaptation and mitigation; and (5) environmental health inequality and inequity, climate change and environmental and biodiversity. By striving to have positive impacts in these five areas, degradation, and pandemic diseases. CGIAR seeks to furnish outstanding science to improve human well-being worldwide. 12 We are challenged to stay within planetary boundaries (Steffen et al. 2015) while meeting society’s demands for affordable, acceptable, adequate, and nutritious diets. Staying within planetary boundaries means wisely using land and water resources, preserving biodiversity, and reducing our emissions of pollutants and greenhouse gases (GHGs) to acceptable levels. These goals must be realized in a global context of over-consumption and waste by developed countries, continuing global economic development, worldwide population growth, and accelerating climate change and their related asymmetrical impacts. These major drivers of change, often referred to as mega-trends, in turn contribute to secondary drivers of change that are of great consequence to agrifood systems. These secondary drivers include migration patterns, urbanization, and rural transformations such as increasing technology use and heightened prevalence of women in agriculture. Within this dynamic, the One CGIAR research portfolio is contributing to the sustainable development goals through five identified impact areas, which CGIAR defines as: (1) nutrition, health, and food security; (2) poverty reduction, livelihoods, and jobs; (3) gender equality, youth, and social inclusion; (4) climate adaptation and mitigation; and (5) environmental health and biodiversity (CGIAR 2021). CGIAR has developed targets for these impact areas (Figure 1.1). Figure 1.1. 2030 Global Collective Targets for Five Impact Areas. • End hunger for all and enable affordable healthy diets for the 3 billion people who do not currently have Nutrition, health access to safe and nutritious food. and food security • Reduce cases of foodborne illness (600 million annually) and zoonotic disease (1 billion annually) by one third. • Lift at least 500 million people living in rural areas above the extreme poverty line of US$1.90 per day Poverty reduction, (2011 PPP). livelihoods and jobs • Reduce by at least half the proportion of men, women, and children of all ages living in poverty in all its dimensions according to national definitions. • Close the gender gap in rights to economic resources, access to ownership, and control over land and Gender equality, natural resources for over 500 million women who work in food, land, and water systems. youth and inclusion • Offer rewarding opportunities to 267 million young people who are not in employment, education, or training. • Implement all National Adaptation Plans and Nationally Determined Contributions to the Paris Climate adaptation Agreement. and mitigation • Equip 500 million small-scale producers to be more resilient to climate shocks, with climate adaptation solutions available through national innovation systems. • Turn agriculture and forest systems into a net sink for carbon by 2050, with emissions from agriculture decreasing by 1 Gt per year by 2030 and reaching a floor of 5 Gt per year by 2050. • Stay within planetary and regional environmental boundaries: consumptive water use in food production Environmental of less than 2500 km3 per year (with a focus on the most stressed basins), zero net deforestation, health and nitrogen application of 90 Tg per year (with redistribution towards low-input farming systems) and biodiversity increased use efficiency, and phosphorus application of 10 Tg per year. • Maintain the genetic diversity of seeds, cultivated plants, and farmed and domesticated animals and their related wild species, including through soundly managed genebanks at the national, regional, and international levels. Projected Benefits of CGIAR Research 13 1.1. Return on investment and widely distributed facilities for germplasm management in agricultural research and the general importance of gene banks as a foundational element of these systems. However, the impact of research for development into natural resource and production system management has been site-specific. Its spread has been restricted The projected benefits analysis in this report focuses on because of policy and institutional constraints on the identifying plausible returns on investments in the form of transfer of technology (Pal 2011). projected benefits relative to the five CGIAR impact areas for the One CGIAR research portfolio. Before we go into While macro-level returns on investment in agricultural those details it is important to stress that investments research are clear, evidence linking technology adoption in international agricultural research have proven very to improved agricultural productivity at the farm level effective over the years, with a conservative estimate of is difficult to find (Loevinsohn et al. 2013). For individual return-on-investment ratios of approximately 10:1. In this commodities, return on investment has been calculated for section, we briefly review studies that have delved into wheat (Lantican et al. 2016) and maize (Krishna et al. 2021), the issue of return on investment in agricultural research. showing massive returns on investment in CGIAR research, The studies are varied and focused on a wide range of especially related to advancements in genetics and socio- topics. Investments in agriculture may result in some technical innovation bundles. As these initiatives progress, negative externalities associated with their outcomes, returns on investment in specific crops and corresponding which are treated both here and in the projected benefits socio-technical bundles can be estimated. study that follows. Most studies of return on investment in overall agricultural research date back several years, 1.2. The agrifood system and we did not identify any recent comprehensive studies. One meta-analysis of 292 studies finds an average rate of The context in which CGIAR operates – the agrifood system return of 100% on agricultural research after excluding – is dynamic and complex. This reality has been graphically extremely high outliers (Alston et al. 2000). Another represented by the ShiftN consultancy (Figure 1.2). Within study, using a modified internal rate of return, arrives the intricate agrifood system, there are specific subsystems at a more modest range of 10% to 40% (Hurley et al. and components, and within these, additional subsystems 2014). A more recent study points out that innovation has and components. These can be analyzed at different transformed agriculture and in doing so has contributed to scales and levels of aggregation. As a complex system, transforming whole economies (Alston and Pardey 2016). the agrifood system has both internal logic with causal The consequences have been profoundly important for relationships at the sub-component level and emergent lives and livelihoods, generally favorable but almost always properties because of the interactions involved. with undesirable effects for at least some people. The issue One of the most confounding challenges is that for human of insufficient funding for agricultural research despite beings, food and nourishment are basic needs (Maslow its huge benefits is attributed to an apparent disconnect 1943). Yet the agrifood system is part of and bounded by between evidence about the social payoffs of agricultural Earth’s ecological and atmospheric context. Humanity has innovation and the research that enables it, on one hand, been stretching this context, Earth’s planetary boundaries, and producer, public, and political perceptions on the other. and there is a clear need to shift the dynamic towards Another recent study examines the economic impact of agricultural systems that provide sufficient, affordable, high-yielding crop varieties in developing countries from nutritious, and acceptable food in a sustainable way, 1960 to 2000 (Gollin et al. 2016). It uses time variation in the while remaining within these boundaries (Raworth 2017). development and diffusion of high-yielding varieties of 10 Staying within planetary boundaries requires making major crops, spatial variation in agro-climatic suitability for projections based on major drivers of change such as growing them, and a differences-in-differences strategy to demographic growth, economic development, geopolitical identify the causal effects of adoption. In a sample of 84 considerations, and climate change, as well as projections counties, the authors estimate that a 10-percentage-point of changes in subsystem components. The complexity increase in the adoption of high-yielding varieties increases involved in developing useful projections is compounded GDP per capita by about 15%. by inherent stochastic processes and volatility that can lead to events with a profound influence on outcomes. It The success of CGIAR in varietal development programs is currently impossible to understand the agrifood system mainly stems from the free exchange of plant genetic and all its details and emergent properties, yet we can resources and from partnerships with national agricultural make inroads by using abstractions, simplifications, and research systems, underscoring the need for systematic ordering principles that are fit-for-purpose. 14 Figure 1.2. Global food system map. The Global Food System Source: ShiftN 2009 (used with permission). Despite this complexity, we seek to evaluate the potential future impact of different research investments in various dimensions of the agrifood system (either focused or broad and either individual or multiple investments occurring simultaneously). Both ex-ante assessment and foresight approaches offer the capability for analysis leading to improved understanding of the potential impacts and benefits of investments in international agricultural research for development. Foresight and ex ante analyses can help enumerate tradeoffs and interactions to inform prioritization and allocation of limited resources towards certain investments and not others. In addition, ex-post analysis of a past investment can support exploration of differential impacts on experimental and control populations who were and were not influenced by the investment, respectively. In this assessment of projected benefits, we compare a comprehensive (COMP) scenario that projects a full range of CGIAR research investments (Section 4.1), and a reference scenario that assumes no investments. We explore the following two components: • Additional benefits resulting from CGIAR research in various development pathways. • Negative consequences of not investing in international agricultural research for development. For many agricultural technologies, it takes anywhere from ten to twenty-five years or more from when research begins until it comes to fruition in farmers’ fields, and there are many intervening human, economic, policy, and environmental factors that will change during this period. For this reason, precise estimates of system-level changes are simply not possible. Therefore, this exercise of calculating the projected benefits of the CGIAR research portfolio is designed to roughly estimate, at least within an order of magnitude, the plausible futures and corresponding advantages associated with systematically targeting investments in agricultural development. With additional time, data, and clear programming objectives, these estimates can be further refined. Projected Benefits of CGIAR Research 15 The present report is structured as follows. In the next two chapters we delve deeper into how foresight and ex-ante impact assessment can be used to evaluate the potential of socio-technical innovation bundles (Barrett et al. 2020a). After this, Chapter 4 addresses foresight and ex-ante impact assessment methodologies, tools, and data sources. In Chapter 5, we then describe the findings, how these analyses can be interpreted, and the projected benefits within each of the five impact areas of the CGIAR research portfolio. Our results focus throughout on the six CGIAR regions: Central and West Asia and North Africa (CWANA), East and Southern Africa (ESA), Latin America and the Caribbean (LAC), South Asia (SA), Southeast Asia and Pacific (SEA), and West and Central Africa (WCA) (Figure 1.3). Figure 1.3. CGIAR’s six regional program areas. South East Asia and the Pacific (SEA) Central and West Asia and North Africa (CWANA) South Asia West and (SA) Central Africa (WCA) Latin America and the Caribbean (LAC) East and Southern Africa (ESA) Source: https://www.cgiar.org/how-we-work/strategy/ 16 2. Foresight approaches for understanding potential investments factors. These external factors include both foreseeable Key messages mega-trends and unexpected or unpredictable events. Development pathways of interest proceed from research Strategic foresight involves a collection of outputs through outcomes to the desired impacts for which approaches to interpret and better understand the we are calculating projected benefits (Figure 2.1.1). Spheres future implications of strategic decision making of control, influence, and interest can also be systematically based on a set of structured and relevant inputs. evaluated (Figure 2.1.1). Spheres of control are areas and activities over which one exerts control. Spheres of Foresight supports understanding of plausible influence encompass matters that can be influenced but future states and, depending on the scenarios, over which there is no direct control. Sphere of interest models, and inputs used, can provide insight into comprise impact areas of concern over which there is no policy outcomes, environmental externalities, and direct influence, although the activities in an investment interactions between different drivers of change. portfolio aim to generate changes in those areas. One CGIAR portfolio outcomes should play a crucial role in behavioral change and in the transition from the current to the desired future state, especially for lower income 2.1. What is foresight? countries. Meeting sustainable development goals along the five CGIAR impact areas will require ongoing evaluation and reevaluation of what has worked and, likewise, what The first question we answer is, “What is foresight?” could work to induce desired change. Projected benefits Likewise, we also delineate what foresight is not. are not targets, but ballpark figures indicating the direction Foresight—often characterized as strategic foresight—is of both desired benefits and foreseeable unintended an encompassing, systematic approach that uses a variety consequences. of methods to explore future possibilities associated with a specific strategic question; foresight thus facilitates Some of the terminology used here derives from impact thinking about the future to inform decisions today. assessment and evaluation literature. The core of Strategic foresight approaches are broadly defined and evaluation methodology is to test or check key assumptions may include combinations of horizon scanning, ex-ante that provide the “logic” and expectations of each different impact assessment, forecasts and modeling, scenario stakeholder. This “logic” is known by interchangeable analysis, and other quantitative and qualitative approaches. concepts as “logic models” (Mayne 2001), “results chains” Many foresight approaches, for example ex-ante impact (DCED 2010), “program theories” (Chen 1994; Rogers 2009), assessment, stand on their own for specific purposes, for or “theories of change” (Weiss 1997). These approaches instance, as an approach to start with a present condition direct attention to the impact of an investment and – and estimate likely future effects of a policy or other specific critically – explore the causal chain that is expected to intervention. A characteristic feature of foresight is that that result in that impact; that is, actions that make a difference it helps us to look ahead through a specific lens and serves in the ultimate outcomes. The combination of ex-ante to help paint the possible, plausible, and probable futures, (before/baseline), ex-durante (during/monitoring), and potentially including both desirable and undesirable states ex-post (after/results) evaluations provides quantitative of nature and society. Foresight thus helps to construct and qualitative mechanisms to better measure development pathways through necessary investments, performance or impact. policy adjustments, and interventions to reach a desired Ex-post impact assessment tends to be more specific and future. Not only does foresight help analyze and describe aligned with the data needs of stakeholders involved in potential futures, it can also help capture broader scale an intervention whereas impact evaluations tend toward system functions that consider a wide variety of drivers and net-impact calculations (Khandker et al. 2009). Impact interactions. assessment is a broad concept that encompasses both Plausible futures can be shaped by investments and ex-post and ex-ante analysis, looking at both contribution interventions under our control as well as by external and attribution, whereas impact evaluations are strictly Projected Benefits of CGIAR Research 17 Figure 2.1.1. Ex-ante and foresight approaches help inform how building blocks of research can work together for an integrated theory of change. Immediate Outcomes: Intermediate Outcomes: focus on expected / desired focus on the expected / behavioral change desired change in agrifood system components Outputs Impact: focus on the five impact GEOGRAPHY - areas of One CGIAR AGRI-FOOD SYSTEMS COMBINATION Sphere of Sphere of Sphere ofCONTROL INFLUENCE INTEREST Other Other influences Other influences, influences and outside factors outside factors and drivers of change ex-post. In general, impact assessments consider the that are likely to affect the process. Pertinent assumptions effectiveness and significance of a change, whereas and plausible and probable impact pathways capture both evaluations in the strict sense of the word examine positive the intended outcomes and key positive and negative and negative changes attributable to an activity. Both may externalities. They can therefore be used for rigorous inform foresight analyses, with impact assessment offering quantitative modeling to determine how technology insight regarding either initial conditions based on packages fit into farming systems, livelihood strategies, and ex-post insight or potential future impact via ex-ante more generally into dynamic, complex agrifood systems. impact assessment. Deriving plausible futures entails: a clear understanding of Conducting an ex-ante impact assessment and evaluation the current situation of the systems of interest, including of of how well promising technology packages might serve trends and mega-trends, and the development of scenarios future beneficiaries depends on several concepts. Firstly, it that, in general terms, consider the role of these trends requires developing a set of plausible and probable impact as drivers. Again, both impact assessment and impact pathways in which the technology proposition plays a role evaluation may be useful to understanding the systems and all underlying assumptions are made explicit. The in question. In turn, scenarios and driver information are impact pathways that are thus constructed, ideally using translated into models, and the relationship between inclusive stakeholder approaches, provide insights into trends, drivers, and variables of interest are systematically key processes, relevant states of nature, and behavioral quantified. It is beyond the scope of this document to changes that influence the adoption of new technologies. deeply explore all aspects of foresight models; for the Secondly, the pathways themselves help pinpoint critical purposes of this report, a core set of modeling systems nodes that indicate where, when, and how ex-durante are addressed, and the mega-trends of interest include evaluation, also known as monitoring, should take place. climate change, demographic growth, and overall economic Nodes are considered critical if relevant underlying development. assumptions are contentious or if there are outside factors 18 3. Socio-technical Box 3.1. Socio-technical Innovation Bundles innovation bundles for Agrifood Systems Transformation Socio-technical Innovation Bundles for Agrifood Systems Transformation are context-dependent, customized combinations of mutually Key messages reinforcing innovations (Barrett et al. 2020a, 2020b, 2021). These bundles consist of traditional agricultural innovations related to Socio-technical innovation bundles are genetic improvement and enhanced management practices, especially composed of elements of different initiatives and focused on creating production systems that stay within planetary contributions from various actors. boundaries. They also include innovations in alternative, land-saving nutrient production systems. Innovations is sustainable animal and CGIAR’s four pillars of socio-technical innovation plant production systems benefit from being combined with a variety bundles include innovations in core CGIAR of other advancements in digital solutions, financing, social protection technologies, in supporting technology and instruments, civic engagement and policy, supply chains, systems, in institutions, and in the enabling and health and nutrition. environment. Socio-technical innovation bundles are essential for balanced uptake. These four pillars of innovation are strategically Linking socio-technical innovation bundles with clear pathways to integrated through activities that aim to achieve impact and theories of change can help minimize perverse incentives beneficial outcomes and impacts across the six and negative externalities. Socio-technical innovation bundles have four CGIAR regions. pillars that encompass the variety of components necessary to ensure that innovations can have desired impacts, and for CGIAR, these pillars include innovations in the following areas: 1. Core CGIAR technologies such as genetic improvements and enhanced agronomic practices. Fitting the outputs of CGIAR research portfolios into 2. Technology and systems that support these core contributions development pathways hinges on the notion of Socio- and offer additional, standalone functionality, such as digital technical Innovation Bundles for Agrifood Systems capabilities and financial instruments. Transformation (Box 3.1; Table 3.1) (Barrett et al. 2020a, 2020b, 2021). Socio-technical innovation bundles are composed 3. Institutional settings, which may benefit from capacity-building of elements of different initiatives and contributions and enhanced collaboration and foresight. from various actors, such as the Food and Agriculture 4. The enabling environment, which derives advantages from Organization of the United Nations (FAO), the International better policy and broader inclusion. Fund for Agricultural Development, the World Bank, non- governmental organizations, private-sector companies, national governments, and line agencies. They are loosely analogous to a multitude of colors going into an inverse prism that produces a single bundle of white light, which is a regions. The overall theory of change of One CGIAR spans vector of intended impacts and also the basis of both positive this entire system (Figure 3.1). Each component will make a and negative externalities. For CGIAR, several innovation unique contribution to the overall set of projected benefits pillars support socio-technological bundling for integrative associated with CGIAR, including productivity enhancement, innovation (Box 3.1; Table 3.1). Ultimately, projected benefits improved water management, better infrastructure, derive from standalone and combined, past and present nutrition- and inclusion-focused policy, and improved innovations along these four pillars, accomplished within the technology, among other approaches. The overall potential CGIAR research portfolio and by the academic community to realize these advantages arises from the complimentary and implementation and scaling partners. nature of the combined activities spanning different action These four pillars of innovation are then strategically areas: Systems Transformation (ST), Resilient Agrifood integrated through sets of activities that aim to achieve Systems (RAFS), and Genetic Innovation (GI). beneficial outcomes and impacts across the six CGIAR Projected Benefits of CGIAR Research 19 Table 3.1. Innovation pillars required to support socio-technological bundling for integrative innovation. The “+” assessment indicates an approximate weight associated with each pillar in each action area based on an evaluation of inputs from action area tech teams. Pillar description ST RAFS GI Innovationtype Pillar 1 – Core technologies from increased research and development, such as genetic improvements, enhanced agronomic practices, bundling multiple complementary technologies, or other tailored or focused + + +++ technologies such as irrigation. “Hard” innovation Pillar 2 – Technology and systems innovations that both leverage core technologies and serve in a standalone manner, such as digital advisories, blockchain-enabled value chains, insurance and credit, improved water ++ +++ ++ use efficiency, and financial instruments more broadly. Pillar 3 – Institutional innovation, including fundamental shifts in institutional coordination, implementation approaches, and foresight, accompanied by heightened ambition and capacity to achieve the same or +++ ++ ++ improved market efficiencies with less waste. “Soft” innovation Pillar 4 – An enabling environment, entailing research to support investments and practices related to institutions, policies, gender, and +++ + + inclusion. For example, the GI program may involve innovation in activities to accelerate the development of new technologies such as drought- or heat-tolerant crops, whereas the RAFS area supports targeted dissemination of the same to those most affected by related climate hazards, and ST enables appropriate policy infrastructure to ensure gender-inclusive approaches and better risk management by farmers (see Figure 3.1). Figure 3.1. Socio-technical innovation bundles as packaged components of a development pathway. P1–P4 are the four pillars of innovation that are bundled to varying degrees to achieve the desired outcomes. Resilient Agrifood Systems PROJECTED BENEFITS CWANA Productivity Enhancement Genetic Innovation WCA Improved Water Management ESA Systems Transformation Improved Infrastructure SA Regional Integration SEA Focus Policy, Technology Digital Information and Approaches LAC P1 P4 P2 P3 The value proposition of the system derives from this complementarity and, as such, we assess projected benefits at the system level. The GI+RAFS+ST example highlights how a coordinated systems-level strategy can support greater impact, whereby the whole is greater than the sum of the parts. There is a clear relationship between system-level projected benefits, the socio- technical bundles required to achieve those benefits, and the contribution of each of the action areas. Socio-technical innovation bundles span genetic innovations, management practices aimed at creating resilient agrifood systems, and policy, institutional, and market innovations. Projected benefits at the systems level across the CGIAR research portfolio capture the breadth and width of these innovations. 20 Strategic Integration Specific Activities Holistic Activities 4. Inputs, models, and tools There are several different models and tools that may Key messages be used to better understand plausible futures and corresponding projected benefits. They all contain four A variety of modeling approaches can be used layers. The first layer is the evidence base of empirical to project the possible benefits of the OneCGIAR data related to the five impact areas, which obviously investment portfolio. does not include any projections into the future. Because this evidence base is constructed using available data These approaches entail varying levels of sources, it is crucial to be clear about the sources of the complexity and uncertainty. data and any relevant caveats. The second layer builds on this evidence base to provide a structural description The comprehensive investment scenario (COMP) of patterns of change in key indicators; this description involves heightened investment across the CGIAR includes trend analysis, econometric models using historic portfolio and targets better research efficiency, data, and big data analytics. irrigation expansion, water use efficiency, enhance The third layer contains projections into the future related soil water holding capacity, and infrastructure to vital aspects of the five impact areas. Examples include improvements. but are not limited to climate change projections based on well-established climate models and projections of The International Model for Policy Analysis of population growth and urbanization. These projections are Agricultural Commodities and Trade (IMPACT) widely accepted in the scientific community and have well simultaneously considers trends in crop described models, bandwidths, and caveats. The fourth production, climate, demand, and other factors. layer builds on the third and contains projections that may be more controversial due to their broad assumptions, high IMPACT undergirds many of the results presented levels of dependence on other models, and often longer in Chapter 5. time horizons. Examples include the outcomes of global economic models such as the International Model for Policy The Global Innovation Fund (GIF) approach can Analysis of Agricultural Commodities and Trade (IMPACT), help frame foresight results ans captures the the Modular Applied GeNeral Equilibrium Tool (MAGNET), social impact of innovative investment in terms of the Global Biosphere Management Model (GLOBIOM), and breadth, depth, and probability of success. Modelling International Relationships in Applied General Equilibrium (MIRAGE). These models combine information The Agrifood Data Analysis Modeling framework from a variety of different subsystems into an integrated (ADAM) is a flexible, extensible resource for assessment. Due to the non-linearity of interactions and foresight analysis related to agrifood systems. the difficulties in parametrization, the outcomes of these models provide a scenario for the future with multiple The Spatial Production Allocation Model database uncertainties. These four layers are quite general and can in combination with the Mink gridded crop be used for a wide variety of purposes, either individually modeling framework enables us to understand the or in combination with one another. effects of non-adaptation to climate change. These different layers furnish information characterized by varying levels of uncertainty and of acceptance in a The Rural Investment and Policy Analysis (RIAPA) broad scientific or policy community. Some outcomes modeling and data system is an excellent tool for are well received, and some are more contested. Yet forward-looking, economy-wide, country-level these layers provide a framework for discussions on analysis. prioritizing research, delineating key gaps in knowledge and offering insight into the extent that these gaps can be closed. The abovementioned approaches also operate at different temporal and spatial scales. Awareness of scale is important to choose the right approach for the appropriate level of aggregation and time horizon. Projected Benefits of CGIAR Research 21 Several approaches and tools can be used to inform and support priority setting (Figure 4.1). At the global scale, examples of integrated assessment models include IMPACT from the International Food Policy Research Institute (IFPRI), GLOBIOM from the International Institute for Applied Systems Analysis, MAGNET from Wageningen University and Research, or the Netherlands Environmental Assessment Agency’s Integrated Model to Assess the Global Environment (IMAGE). These models provide insights about the future and often use standard, intentionally designed, and internationally accepted scenarios regarding issues that are difficult to model but easy to understand by decision-makers. Examples of such scenarios are the Shared Socioeconomic Pathways (SSPs) developed by the Intergovernmental Panel on Climate Change (Riahi et al. 2017). As integrated models, they utilize outputs from other models, such global circulation models, that simulate future climate under climate change drivers. Integrated assessment models are quite complex and use large sets of input data, with key parameters driving model outcomes. Such models are constructed based on a wide variety of approaches, including econometric models, trend analysis, and expert guesstimates. Models, however, do not make decisions; decisions balance many factors and are the domain and responsibility of decision-makers. Involving people necessarily introduces diverse perspectives, and different people will interpret model results variously. To optimize their choices, decision-makers can endeavor to understand the basics of modeling and develop a shared understanding of how model results can inform decision processes. When model results are made public, transparently articulating assumptions and caveats associated therewith can support optimal use and interpretation of findings. Nevertheless, it is difficult to foresee all possible consequences. Figure 4.1. Temporal and spatial scales and associated approaches. Trend analysis Globalindicators IMPACT, Globium, MAGNET, etc. Scenario development History Global circulation models (GCMs) National statistics CGE models TIME TOA-MD, landscape Value chain simulation models analysis Econometric farm Farm household models simulation models SPACE For these reasons, other evidence-based approaches that are more transparent can also be used, such as comparison of key indicators for issues of interest and trend analysis of the most important indicators. There are also models operating at lower levels of aggregation (Kruseman et al. 2020), including farm-level models for ex-ante impact assessment of policies and technologies, landscape-level models, and trade-off models (Antle 2011; Capalbo et al. 2016). Though informed by these lower-level aggregation models, the current analysis does not utilize them because it focuses on the high-level projected benefits of the whole One CGIAR research portfolio. 22 Trend analysis has always been an integral part of foresight and ex-ante impact assessment, coupled with understanding of the current situation that serves as a benchmark. Ideally, projected benefits calculations would be done on a continuous basis, using a variety of modeling tools, at different scales, and covering various potential scenarios. While many models exist, the IMPACT model is one of the best-suited approaches for examining the CGIAR portfolio as a whole. However, there are certain aspects of the CGIAR impact areas, such as social inclusion, which are not adequately captured by IMPACT, and other approaches are also used. In the following sections, we describe how different modeling and analytical approaches help interpret and contextualize macro trends, and then we use this interpretation to assess projected benefits at the system level. 4.1. The International Model for Policy Analysis of Agricultural Commodities and Trade Georgina Smith / ILRI The IMPACT model projects benefits by drawing on a series of scenarios that are designated to describe plausible previously identified innovation pillars and comprising futures and are based on an earlier developed analysis in multiple socio-technical bundles. The COMP scenario support of the CGIAR portfolio design process (Rosegrant et assumes the following: al. 2017; Wiebe et al. 2020). IMPACT is a partial equilibrium model that was extended with a global computable • Considerable increases in investment and heightened general equilibrium model (GLOBE) and linked to climate, research efficiency across the CGIAR portfolio. crop, water, and economic models to explore worldwide • Irrigation expansion plus water use efficiency alternative futures for food and agriculture (Robinson investments. et al. 2015). IMPACT modeling involves generating two different types of insights or scenarios: firstly, baseline • Investments targeted to enhance soil water holding developments in global agricultural systems to 2050; and capacity. secondly, alternative investment options, including various • Infrastructure improvements to boost market scenarios that reflect investments in agricultural research, efficiency by reducing transportation costs and water resource management, irrigation, soil management, marketing margins. and infrastructure. The generated scenarios also include These scenarios capture specific components of the assumptions about growth in GDP and changes in One CGIAR research portfolio across action areas. Direct population, expressed through the SSPs (Riahi et al. 2017). projections of the impact of interventions closely related to The alternative investment options are then analyzed to the four innovation pillars and the One CGIAR impact area see how they could help achieve the goals of reducing indicators listed below are limited (Table 4.1.1). Nevertheless, poverty and meeting future food demands sustainably these model results give a general indication of the direction and equitably. For nutrition availability projections, we use and magnitude of the expected impacts associated with the results from Rosegrant et al. (2017) with the analytic investments in CGIAR research and, ultimately, are useful approach presented in Beach et al. (2019). for quantifying projected benefits of an enhanced One No new model runs with IMPACT were generated for this CGIAR portfolio. Given the different intellectual bases of projected benefits exercise. Instead, we use the COMP the IMPACT model and the One CGIAR impact indicators, scenario developed for the above-mentioned study. there are few direct connections, but the IMPACT results This COMP scenario is a portfolio of four investment are related to the One CGIAR impact areas, indicators, and components notionally spanning each of the four corresponding projected benefits (Table 4.1.1). Projected Benefits of CGIAR Research 23 Table 4.1.1. Impact areas, indicators, and projected benefits. Note that impact indicators highlighted with bold text are fundamental to the calculated projected benefits. Impact indicators in italic text are supported, albeit less directly, through the available results. Impact area Impact indicators Projected benefits approach Nutrition, health, and food security • #people benefiting from relevant CGIAR innovations • Population at risk of hunger • #people meeting minimum dietary energy • Undernourished children requirements • Nutrient availability of protein, vitamin A, • #people meeting minimum micronutrient requirements magnesium, zinc, and iron • #cases of communicable and non-communicable diseases Poverty reduction, livelihoods, and jobs • #people benefiting from relevant CGIAR innovations • Poverty-weighted value share by crop • #people helped to exit poverty Gender equality, youth, and social inclusion • #women’s empowerment and inclusion in the agricultural • Projected sex-disaggregated increases in sector employment due to agricultural investment in • #women benefiting from relevant CGIAR innovations 14 African countries • #youth benefiting from relevant CGIAR innovations • #women helped to exit poverty Climate adaptation and GHG reduction • #tonnes of CO2 equivalent emissions • CH4 rice emissions • #plans with evidence of implementation • CH4 enteric emissions • #$ climate adaptation investments • CO2 emissions from crops and animals • #people benefiting from climate-adapted innovations • CO2 land use change emissions • All agricultural emissions Environmental health and biodiversity • #ha under improved management • N2O fertilizer emissions • #km3 consumptive water use in food production • Water use • #ha deforestation • Land conversion • #Tg nitrogen application • Approximate species loss • #plant genetic accessions available and safely duplicated 4.2. The Global Innovation Fund approach The Global Innovation Fund (GIF) has developed an approach to capture the social impact of its investments in innovations based on the following components (Global Innovation Fund 2019): • Forecast the impact of prospective investments and use this information to guide investment decisions. • Track project performance and impact during implementation, using real-time information to adapt and adjust as necessary. • Evaluate investments after their completion to better understand how they fared and why they performed that way; then, use this evidence to guide future GIF decisions and inform decisions made by other development partners. The approach is also relevant for projected benefits because it focuses on three aspects of impact: breadth, depth, and probability of success. 24 • Breadth is the number of low-income people projected 4.4. Geographic information to benefit 10 years after the initial investment. This ten-year horizon is an index of progress towards the system analysis of climate long-run maximum scale of the innovation and can hotspots be applied to other indicators such as the area under sustainable management or GHG reduction factors To understand the benefits of the investments in climate linked to improved practices. change adaptation in the CGIAR research portfolio, we also identify the effects of non-adaptation. Using the • Depth of impact is a universal measurement of IFPRI Spatial Production Allocation Model database on economic and noneconomic benefits. It captures relative crop location in combination with the Mink gridded changes in consumption or standards of living, or in the crop modeling framework developed at IFPRI (Müller nd case of planetary boundaries, relevant indicators for Robertson 2014; Robertson 2017), we can project changes those variables. in productivity related to climate change, keeping other • Probability of success is assessed by considering factors constant. the risks along the path to scale. This aspect can be a The outputs of the Genetic Innovation Action Area and dichotomic choice between success or failure or can of the Resilient Agrifood Systems Action Area are aimed span a probability distribution. at counteracting the negative effects of climate change Though the GIF approach is not used directly in combination through improved varieties that are tolerant of abiotic with the IMPACT model, IMPACT results serve as a suitable stress and through climate-smart agricultural practices. backdrop to inform estimations of GIF parameters. The Offsetting yield decreases marks the bounds of the GIF approach is especially useful when there is at least projected benefits of climate change adaptation. some clarity about the different aspects of impact. In the Complementary approaches are used to estimate the current high-level projected benefits analysis, this step is projected benefits associated with climate adaptation. For not yet relevant. The GIF approach is mostly used to be example, geographic information system analysis can be able to prioritize between different investment options. applied to understand a population’s exposure to climate Here we look at the portfolio as a whole without a priori hazards and later projected potential benefits by using prioritization. One of the primary objectives of this report prior estimates of research impact for different types of is to support later use of the GIF method in the internal climate exposures and extrapolating that impact with an assessment of the different proposed initiatives that will be adoption rate (Ramirez et al. 2021). the basis of the new One CGIAR strategic approach. 4.3. The Agrifood Data Analysis 4.5. Rural Investment Modeling framework and Policy Analysis The Rural Investment and Policy Analysis (RIAPA) modeling Understanding complex, dynamic agrifood systems and data system is IFPRI’s primary tool for forward- requires nuanced analysis of information from many looking, economy-wide, country-level analysis. It serves as sources that are not necessarily compatible. Open data a simulation laboratory for experimenting with different sources are expanding and provide important building interventions and shocks and assessing their implications blocks. Open and transparent systems can generate for economic development and inclusive agricultural information that can be used for decision making. To do transformation. At its core is a dynamic computable general this, the foresight team at the International Maize and equilibrium (CGE) model that captures the interactions of all Wheat Improvement Center (CIMMYT) have developed producers and consumers in an economy, both within and a flexible, extensible resource for generating actionable outside the agrifood system. RIAPA also includes various information related to agrifood systems, the Agrifood Data add-on modules that inform simulation design or model Analysis Modeling framework (ADAM). The focus of ADAM inputs, including, for example, how public agricultural is to provide parsimonious fit-for-purpose approaches investments affect crop yields; how livestock policies shape to quickly furnish metrics for foresight purposes. Where herd dynamics and animal productivity; or how newly IMPACT did not generate results, ADAM was used as established value chains, once brought to scale, impact a complement related to the impact areas of poverty the broader economy. In addition to standard economic reduction, livelihoods, and jobs; environmental health and variables such as sectoral production, household incomes, biodiversity; and to some extent gender equality, youth, and market prices, various add-on modules produce and social inclusion. Projected Benefits of CGIAR Research 25 Secondly, RIAPA helps policymakers better understand the trade-offs associated with policy and investment choices. The reality is that not all policies, investments, or innovations are equally effective in achieving development outcomes. A given policy choice inevitably creates winners and losers; for example, research that contributes to the improved functioning and profitability of one value chain may harm others. Trade-offs can arise from competition over scarce resources and from structural differences between economic sectors, for instance in terms of technologies, labor skill requirements, or downstream linkages to processing and services. Differences in the way households interact with economic sectors as workers or consumers can also require trade-offs. Ultimately, in a world of scarcity, policy choices must reflect society’s priorities, and RIAPA facilitates that prioritization process. Third, by capturing macro-micro linkages, RIAPA helps analysts measure the impacts of interventions for individual households. RIAPA’s social inclusion outcome indicators on poverty, diet quality, and women’s inclusion are computed using survey-based “microsimulation” models linked to the core CGE model. Thus, although the CGE model itself uses aggregated data, including broad sectors and representative household groups, it also links economic outcomes such as relative prices, consumption choices, income, and employment changes to real households in the survey. These features allow for a more Peter Lowe/CIMMYT nuanced assessment of distributional and welfare effects. In total, RIAPA’s outcome indicators are closely aligned to CGIAR’s impact areas, including (i) nutrition and food security; (ii) poverty reduction, livelihoods, and jobs; and specialized outcome indicators. For example, agrifood (iii) gender equality, youth, and social inclusion. RIAPA’s system GDP and employment are indicators well suited to set of outcomes is currently being expanded to include capture agricultural transformations, and indicators such agrifood system resilience, including variability in food as poverty, diet quality, and the engagement of women or consumption due to climate and world price risks; agrifood youths in structural transformations provide information system water consumption; and GHG emissions. This about social inclusion. expansion will allow RIAPA to address key aspects related RIAPA includes useful features that make it ideal for tracking to the other CGIAR impact areas, including (iv) climate the economy-wide benefits of interventions such as One adaptation and mitigation and (v) environmental health CGIAR’s research portfolio. Firstly, RIAPA’s structure allows and biodiversity. it to measure impacts throughout the agrifood system and Fourth, RIAPA’s country coverage has expanded rapidly in involving the broader national economy and global markets. recent years to over 30 countries, due to a new streamlined RIAPA can therefore highlight how the development of a process for constructing social accounting matrices (SAMs). new agricultural technology, for example, not only affects The Nexus SAM Building Toolkit adopts common data farmers but also has spillover implications for input suppliers, standards and classification systems, ensuring greater traders, agro-processors, and food service providers such transparency and consistency across countries. Nexus SAMs as restaurants. Likewise, off-farm interventions or policy include at least 86 economic sectors, including 35 in primary changes can affect farmers indirectly. RIAPA allows analysts agriculture and 18 in agro-processing, permitting more to anticipate these consequences by modeling the complex nuanced analysis of impacts across the agrifood system. linkages that exist between economic actors. 26 5. Results Key messages CGIAR investments are projected to result in benefits across the five impact areas. For the nutrition, health, and food security impact area, a comprehensive, integrated investment strategy can shrink the population at risk of hunger, reduce the number of malnourished children, and augment protein and micronutrient availability. Investments in CGIAR research can therefore support the following One CGIAR impact indicators: #people meeting minimum dietary energy requirements and #people meeting minimum micronutrient requirements. Poverty reduction, livelihoods, and jobs together comprise the second impact area; poverty is a multifaceted phenomenon requiring multiple modeling approaches to capture its various contributing causes and aspects. Productivity and GDP increases associated with CGIAR investments support decreases in poverty. Regional poverty lines set at 40% of per-capita income in 2030 yield heightened numbers of beneficiaries moved out of poverty through increments to income delivered by CGIAR. Gender, youth, and social inclusion together make up the third impact area; further research and foresight analysis that explicitly tackles these considerations are urgently needed. Agriculture is a promising mechanism to boost female employment and reduce gender disparities in economic development. Projections on female and youth employment in agriculture are difficult due to gaps in existing data. Climate adaptation and greenhouse gas reduction – the fourth impact area – showed varying results across CGIAR regions, underscoring the necessity of tailored investment strategies. Generally positive consequences follow upon greater productivity per unit of input, offset land conversion, and changing diets; negative climate adaptation results tend to be associated with heightened methane emissions from rice production and increased use of nitrogen. CGIAR genetic research into varieties adapted to climate change can help avert the worst possible climate change- related futures in the regions under study. For climate adaptation work, at 25% and 24%, respectively, India and China contribute circa 50% of the total number of beneficiaries of CGIAR investments. Environmental health and biodiversity is the fifth impact area and there are potential trade-offs; high-investment scenarios had generally positive impacts on biodiversity, species loss, and deforestation. Pressure on biodiversity was substantially relieved, and species loss was reduced to below the levels expected with climate change alone for high-investment scenarios. Throughout all the One CGIAR regions, the use of blue water, indicative of water consumption, is also somewhat reduced. Fertilizer use entails many trade-offs and much regional variation; in general, rates are increasing worldwide. Projected Benefits of CGIAR Research 27 CGIAR investments are projected to result in benefits across children (Figure 5.1.1b), and improve protein intake (Figure the five impact areas: nutrition, health, and food security; 5.1.1c). In all cases, the results are favorable in all CGIAR poverty reduction, livelihoods, and jobs; gender, youth, and regions. Fish has potential to increase available dietary social inclusion; climate adaptation and GHG reduction; protein (Appendix 1). and environmental health and biodiversity. This chapter presents these projected benefits; each section breaks down one of CGIAR’s five impact areas. For each, we provide context to link the IMPACT and related results to the impact Figure 5.1.1. CGIAR Investment Impact for #people meeting minimum dietary energy requirements (COMP shows with area itself and to the corresponding impact indicators. We investment, REF assumes climate change). then characterize the projected benefits and support the interpretation thereof relative to a potential CGIAR research portfolio. In general, high-investment scenario is projected a. Population at risk of hunger to have positive impacts across each of these five areas relative to a future characterized by climate change alone, 150 as detailed in the results described below. 100 5.1. Nutrition, health, and food security 50 The nutrition, health, and food security of the population 0 CWANA ESA LAC SA SEA WCA are paramount to overall human well-being, economic CGIAR Regions development, and national security. Given the integrated COMP REF and heterogeneous nature of many of the socio-technical bundles required to respond to the challenges being addressed by One CGIAR, the precision of projections at the system level is limited. The following findings draw b. Undernourished children from core IMPACT outputs and demonstrate the potential for an integrated investment scenario spanning the four 60 innovation pillars to support improved outcomes related to nutrition, health, and food security. IMPACT results show 40 that under the COMP scenario with CGIAR investment, as opposed to the reference (REF) scenario that assumes 20 only climate change, fewer people are at risk of hunger, fewer children are malnourished, and availability of protein 0 CWANA ESA LAC SA SEA WCA and micronutrients is better across CGIAR regions. These CGIAR Regions results demonstrate how investments that comprise the COMP REF COMP scenario support the following One CGIAR impact indicators: #people meeting minimum dietary energy requirements and #people meeting minimum micronutrient requirements (Figures 5.1.1a, b, and c and 5.1.2a, b, c, and c. Mean daily protein availability d). For additional clarity about potential benefit targets, 200 throughout this chapter, the relevant information is also presented in tabular form in order to indicate the absolute 150 and relative differences between the COMP scenario reflective of systematic One CGIAR investment, and the 100 reference scenario which presupposes only climate change 50 (Tables 5.1.1 a, b, c and 5.1.2 a, b, c, d). When it comes to food security, the IMPACT results 0 CWANA ESA LAC SA SEA WCA demonstrate the potential for a comprehensive, integrated CGIAR Regions investment strategy to shrink the population at risk of COMP REF hunger (Figure 5.1.1a), reduce the number of malnourished Source: IMPACT model results based on Rosegrant et al. 2017. 28 Grams / capita Million Million Table 5.1.1. Absolute and numeric targets for potential benefits related Where nutrition and health are concerned, the identified to dietary energy requirements and CGIAR Investment Impact. impact indicators call for specific numbers of people meeting minimum micronutrient requirements. As a. Population at risk of hunger in millions: micronutrient uptake and use depend on various factors Differences between the COMP and reference scenarios including food access, nutrient bioavailability, hygiene, and adequate food diversity, the pathways to achieve these CGIAR Investment Impact quotas may span everything from biofortification to social Difference % protection and safety nets; they can integrate multiple Region Reference COMP in millions Difference pillars. IMPACT results overlaid with nutritional information and multipliers show improvements in the availability of CWANA 64.015 49.576 -14.439 -22.6 micronutrients such as vitamin A, iron, magnesium, and ESA 134.7 100.475 -34.225 -25.4 zinc (Figures 5.1.2 a, b, c, d). LAC 35.881 29.156 -6.725 -18.7 In all regions, micronutrient availability improves, with proportionally larger projected benefits in SA and all SA 154.875 80.458 -74.417 -48.0 three African regions (Tables 5.1.2 a, b, c, d). These figures SEA 100.642 84.244 -16.398 -16.3 illustrate potential improvements in nutrient availability. This information should be taken in context, recognizing WCA 76.148 44.902 -31.246 -41.0 that additional innovation is required to translate heightened nutrient availability to growth in the number of people meeting minimum requirements. b. Undernourished children in millions: Differences between the COMP and reference scenarios Difference % Figure 5.1.2. Support for #people meeting minimum micronutrient Region Reference COMP in millions Difference requirements. CWANA 9.99 9.098 -0.892 -8.9 ESA 17.853 16.656 -1.197 -6.7 a. Mean daily vitamin A availability LAC 2.984 2.592 -0.392 -13.1 1200 SA 61.271 58.402 -2.869 -4.7 800 SEA 10.732 10.021 -0.711 -6.6 WCA 23.315 21.255 -2.06 -8.8 400 0 CWANA ESA LAC SA SEA WCA CGIAR Regions c. Mean daily protein availability in grams per capita: Differences between the COMP and reference scenarios COMP REF Region Reference COMP Difference % in millions Difference b. Mean daily iron availability CWANA 119.15 126.083 6.933 5.8 50 ESA 69.29 75.231 5.941 8.6 40 LAC 90.104 94.445 4.341 4.8 30 SA 81.001 87.44 6.439 7.9 20 SEA 97.79 102.941 5.151 5.3 10 WCA 69.699 76.64 6.941 10 0 CWANA ESA LAC SA SEA WCA CGIAR Regions Source: IMPACT model results based on Rosegrant et al. 2017. COMP REF Projected Benefits of CGIAR Research 29 Milligrams / capita Micrograms / capita c. Mean daily magnesium availability b. Mean daily iron availability in milligrams per capita: Differences between the COMP and reference scenarios 1500 Region Reference COMP Difference % in millions Difference 1000 CWANA 27.598 29.287 1.689 6.1 500 ESA 24.584 26.627 2.043 8.3 0 LAC 19.75 20.657 0.907 4.6 CWANA ESA LAC SA SEA WCA CGIAR Regions SA 22.425 24.381 1.956 8.7 COMP REF SEA 23.299 24.606 1.307 5.6 WCA 25.535 28.056 2.521 9.9 d. Mean daily zinc availability 30 c. Mean daily magnesium availability in milligrams per capita: Differences between the COMP and reference scenarios 20 Region Reference COMP Difference % 10 in millions Difference CWANA 918.953 975.218 56.265 6.1 0 ESA 655.498 712.606 57.108 8.7 CWANA ESA LAC SA SEA WCA CGIAR Regions LAC 587.217 613.373 26.156 4.5 COMP REF SA 613.009 663.94 50.931 8.3 Source: IMPACT model results based on Rosegrant et al. 2017. SEA 558.191 591.504 33.313 6.0 WCA 686.997 756.097 69.1 10.1 d. Mean daily zinc availability in milligrams per capita Region Reference COMP Difference % in millions Difference Table 5.1.2. Absolute and relative potential benefit targets related to micronutrient availability. CWANA 20.399 21.678 1.279 6.3 ESA 11.309 12.309 1.0 8.8 a. Mean daily vitamin-A availability in micrograms per capita: LAC 14.78 15.534 0.754 5.1 Differences between the COMP and reference scenarios SA 13.366 14.451 1.085 8.1 Region Reference COMP Difference % SEA 14.17 15.005 0.835 5.9in millions Difference WCA 11.985 13.225 1.24 10.3 CWANA 593.657 607.154 13.497 2.3 ESA 595.903 627.152 31.249 5.2 Source: IMPACT model results based on Rosegrant et al. 2017 and the methods LAC 972.567 992.919 20.352 2.1 described in Wiebe et al. 2020. SA 349.226 370.976 21.75 6.2 SEA 707.718 733.306 25.588 3.6 WCA 1000.567 1038.865 38.298 3.8 30 Milligrams / capita Milligrams / capita 5.2. Poverty reduction, livelihoods, and jobs To understand the poverty alleviation potential of investments in agricultural research, we define poverty as a multi- faceted concept linked to individual households and the livelihood strategies that they employ. General conclusions can be drawn from past research, and specific considerations correspond to the different CGIAR target regions. Projections about poverty reduction at scale tend to focus on the macroeconomic perspective and overall total factor productivity. Long-term perspectives on poverty reduction effects associated with agriculture are less prevalent, and we bring several approaches here to offer a sense of what is realistically possible as a function of One CGIAR investments. The overall conceptual framework that indicates how major changes shape decision-making for CGIAR’s target beneficiaries hinges on livelihood strategies; hence, it is intimately linked to poverty and poverty alleviation (Figure 5.2.1). Livelihood strategies include but are not limited to agricultural activities and off-farm employment. This conceptual framework allows us to build a narrative that addresses rural transformation, investment strategies, and migration issues, amongst other factors. Generally, we use a systems approach focusing on drivers, pressures, states, impacts, and responses to structure the narrative, placing it in a foresight context. On the one hand, this approach allows us to highlight plausible and probable alternative development pathways available to resource-poor people in low- and middle-income countries. On the other hand, it also enables us to explicitly consider social inclusion and inequality issues. Pressures in the form of policies affecting agriculture, aquaculture and fisheries, food system value chains, and resource use shape the responses of individual farmers, livestock keepers, and fisherfolk as regards the livelihood strategies they pursue. Figure 5.2.1. Drivers and factors affecting agricultural income and employment. Macroeconomic and other major drivers / factors Access to input capitals (land, human, finance, resources) 1 Capital endowment affectingincome and employment Constraints, opportunities Investment / Production • Climate change Production system • Scale economy • Desertification resilience and Household systems • Technologies • Loss of soil fertitily stability • Risk management / financial services• Education / extension services 2 Technology, financial services, Livelihood strategies affecting and extension services affecting 3 income and employment income and employment y inab ilit Sust a Markets systems and value Food Production Markets systems / 4 chain integration affecting Integration, value chains income and employment Projected Benefits of CGIAR Research 31 Population growth leads to relative scarcity of land Figure 5.2.2. People lifted out of poverty by CGIAR research, 1971-2015, resources in the absence of movement out of agriculture. in millions. In high-income countries, reduction in the number of farms and hence an increase in average farm size have 35 been ongoing since the mid-20th century (Kislev and 30 Peterson 1982; MacDonald et al. 2013; Hermans et al. 25 2017). Economic development is partly associated with 20 structural change, occurring when members of the 15 10 workforce, including youths and women, leave agriculture 5 for more productive jobs in other sectors (Kilby and 0 Johnston 1975; Chenery, et al. 1975; Mueller and Thurlow China Former Latin Africa Central South Middle South Soviet America South of America East Asia East & Asia 2019). There is a correlation between GDP and average Union Sahara North farm size (Adamopoulos and Restuccia 2014). The key Asia question is what this means for farm size development in Source: Laborde et al. 2020. low-income countries in Sub-Saharan Africa and Asia under simultaneous population and GDP growth. • While urbanization is expected to influence food Figure 5.2.3. Global and regional poverty impacts of MIRAGRODEP- consumption patterns, farm size is likely to affect food COVID-19 scenario by region. supply (Fan et al. 2013). 160 • Land fragmentation and small farm size are common in 140 23% many developing countries. The average size of farms in 20% Bangladesh, for example, is about 0.68 ha. In Egypt, 50% 120 of farmers are cultivating less than one ha of land, while 100 15% 15% 15% 14% the average farm size is about 0.4 ha in China, 0.8 ha in 80 Indonesia, and 0.7 ha in Vietnam (Larson et al. 2016). 60 • Generally, farmers with less than one ha of landholdings 40 struggle to fulfil their subsistence requirements through 20 agriculture (Niroula and Thapa 2005). 0 Total Rural Total Rural Total Rural Such factors can be addressed through a combination of population population population population population population resources: for example, the global CGE model known as World Africa South of Sahara South Asia MIRAGRODEP and its linked tools; the 300,000 households Increase in number of poor people Relative increase in the number (US$1.90 poverty line) of poor people represented in the POVANA model; and various physical datasets such as the FAOSTAT emissions database. This set Source: MIRAGRODEP and POVANA simulations. of tools has been used in several recent studies to assess impacts of major shocks, events, and trends, such as the effects of agricultural productivity growth arising from There are no clear-cut methods with proven rigor to project CGIAR investments in research and development (R&D) poverty reduction as result of targeted investments in to address global poverty. A recent study estimated that agricultural R&D (Section 1.1. Poverty is a multi-faceted CGIAR research had reduced the number of people living concept, and disentangling its components is challenging. in poverty by over 70 million, primarily in SA and in Africa In this section, therefore, different approaches to projecting south of the Sahara (Figure 5.2.2) (Laborde et al. 2020). the effects of investments in agricultural R&D on poverty This global framework can also be used to assess the reduction are highlighted. impacts of shocks such as COVID-19 on global poverty Using the parsimonious ADAM modeling framework in (Figure 5.2.3). Recent work concluded that the pandemic conjunction with the IMPACT model, we can identify a proxy caused an increase in global poverty of around 150 value for the poverty implications of food crop production. million people, mostly in SA and Sub-Saharan Africa This proxy focuses on the poverty of smallholder farmers (Laborde et al. 2021). This exacerbation resulted from a who depend primarily on food production for home range of shocks, including morbidity and mortality, social consumption and the marketing of surplus. They comprise distancing, closure of some business activities, and the one of the traditional target groups of CGIAR. The current associated loss of income. situation based on the ADAM dataset encompasses the gross 32 production value of key food crops per hectare for all countries in the various CGIAR regions, split out into high- and low-income countries using the World Bank income group classification (Figure 5.2.4). Using the ADAM modeling framework in combination with key IMPACT outcomes, we can identify differences in production value and, by extension, the potential effect on poverty when comparing the reference scenario and the COMP scenario that serves as an analog for the One CGIAR research portfolio. Figure 5.2.4. Current situation for gross production value of basic food crops. Basic Food Crops: Gross Production Value by Region 20000 10000 0 All countries CWANA ESA LAC SA SEA WCA Regions Income group: High Low Source: Own calculations with ADAM modeling framework. A different analysis focuses on food crop production as a proxy for the production strategy of poor farmers and does not consider cash crops or animal production systems (Figure 5.2.5). The indicator used here is the gross value of production of food crops per hectare harvested. The differences between the baseline and the 2030 simulation results indicate a significant rise in the per-hectare gross value of food crop production, which is a proxy for poverty alleviation. The IMPACT model results show that the scenario that best represents the integrated One CGIAR research portfolio will not necessarily directly reduce poverty in gross production terms relative to the reference scenario, however, because heightened crop productivity may not lead to lower prices of food crops. It is also possible to differentiate between on the one hand, high-income and upper middle-income countries, and on the other hand, low-income and lower middle-income countries (Figure 5.2.6). This distinction is made for all countries in the world as well as for the six CGIAR target regions. The comparison of the reference IMPACT model scenario with the comprehensive IMPACT model scenario indicates that major productivity increases are likely to accrue benefits for consumers rather than producers. The trends in Asia and Latin America indicate that this situation in turn can lead to consolidation of agricultural assets and increases in the scale of production (Jayne et al. 2014; Jayne et al. 2016; Jayne et al. 2018). Projected Benefits of CGIAR Research 33 GPV (2015-2017 USD / ha) Figure 5.2.5. Comparison of the distribution of gross production value per hectare of food crops by region and simulation scenario for the IMPACT base year 2005, the 2030 results for the reference scenario, and the simulation scenario that best reflects the One CGIAR research portfolio, using the ADAM framework. Basic Food Crops: Gross Production Value by Region and Simulation Scenario 100000 75000 Income group: 50000 Base 2005 - High income Base 2005 - Low income One CGIAR investment 2030 - High income One CGIAR investment 2030 - Low income Reference 2030 - High income Reference 2030 - Low income 25000 0 All countries CWANA ESA LAC SA SEA WCA Regions There is evidence in the scientific literature of possible associations between total factor productivity growth, GDP per-capita growth, and poverty reduction. We can draw from this literature to implicitly link IMPACT and related scenario projections about productivity and GDP growth—namely the SSPs—to potential reductions in poverty. Poverty decreased as per-capita GDP increased (Table 5.2.1). This is demonstrated by changes in semi-elasticities, with one study finding that the semi- elasticity varies from -0.5% to -0.1% as we move from very poor to less poor countries (Ivanic and Martin 2018) (Figure 5.2.6); increases in agriculture offer the greatest relative poverty reduction effect. Figure 5.2.6. The relationship between per-capita gross domestic product and poverty change given a 1% productivity change, generated through single-country simulations using validated computable general equilibrium models. 0.4% 0.3% 0.2% 0.1% 0.0% GPD / person -0.1% -0.2% -0.3% -0.4% -0.5% 190 760 3,040 Agriculture Log. (Agriculture) Industry Log. (Industry) Services Log. (Services) Source: Graphic from Ivanic and Martin 2018. 34 GPV (Constant 2017 USD / ha) The complex linkages between agricultural growth and of CGIAR research into poverty reduction. Overall, these are poverty reduction require precise treatment of what is meant high returns in terms of poverty reduction relative to the by poverty and, likewise, of the mechanisms whereby sectoral investment level. dynamics can contribute to poverty reduction. Poverty can To address these complexities, we employ two absolute be defined broadly or narrowly. When considered broadly, poverty lines. First, we deploy the well-known US$1.90 per poverty is a multidimensional phenomenon that accounts day line. We also utilize separate, regional poverty lines set at for multiple deprivations including, for example, a lack of 40% of per-capita income in 2030 for each of the six regions adequate access to services such as education and healthcare in focus for One CGIAR.1 The IMPACT model suite considers that are often publicly provided. One CGIAR both focuses the implications of increased investment in CGIAR R&D as attention on elements of well-being and uses a dashboard of well as the heightened efficiency of One CGIAR investments indicators related to those elements. Hence, for the purposes for per-capita GDP by region in 2030 (Rosegrant et al. 2017). considered here, poverty is narrowly defined as consumption Poverty-growth semi-elasticities are then used to estimate in line with the cost-of-basic-needs approach that underlies poverty impacts (Klasen and Misselhorn 2008) (Table 5.2.1).2 the $1.90 per day poverty line used by the World Bank. In this approach, a given level of welfare as it relates to the Prospects for projecting and tracking the implications of the One consumption of goods and services is deemed minimally CGIAR program for more rigorous assessment of key indicators, acceptable. This level is converted to purchasing power for including but not limited to poverty, are excellent. By the end a basket of goods and services, such as US$1.95 per day. of 2021, the complete initial research program will be fully Households with sufficient purchasing power are deemed developed, allowing for comprehensive assessment of targeted non-poor, while those with insufficient purchasing power are outputs and outcomes. The impacts for growth, jobs, poverty, categorized as poor. diets, inclusion, GHG emissions, and other key metrics can then be rigorously evaluated using the full array of approaches Within the context of One CGIAR, poverty is distinct from available within One CGIAR. Over time, this trajectory can nutrition in at least two ways. First, under the cost-of-basic- be compared to realized impacts while accounting for other needs approach, the welfare levels deemed minimally important factors, such as global pandemics. acceptable are typically rooted in calories. However, it is now clear that nutritious diets cost more than diets delivering adequate calories. Using the US$1.90 poverty line, fewer than Table 5.2.1. Implications of CGIAR research for poverty by region in 2030 one billion people were considered poor worldwide prior to the using two absolute poverty lines. pandemic; however, recent estimates indicate that up to three Changes in poverty given Changes in poverty billion people are unable to afford an adequately nutritious diet region-specific absolute given an absolute (Herforth et al. 2020). Second, poverty is defined in relation to poverty lines set at 40% poverty line at US$1.90 purchasing power or capability. When determining whether a of the baseline per- in 2011 PPP per daycapita income in 2030 household is categorized as poor or non-poor, no reference is made to the household’s actual consumption patterns. Only Region % point Million % point Million observed purchasing power is considered. Clearly, for nutrition, reduction of people reduction of people actual consumption patterns are crucial. LAC -0.36 -2.5 -0.02 -0.2 One CGIAR operates across all major regions of the developing WCA -1.58 -9.8 -0.34 -2.1 world, and measured poverty rates vary dramatically across ESA -0.66 -3.3 -0.12 -0.6 regions. For example, the most recent poverty headcount ratio measures using US$1.90 per day in 2011 purchasing power CWANA -0.63 -4.5 -0.01 -0.1 parity (PPP) from the World Development Indicators are 27.3, SA -1.50 -27.9 -0.23 -4.3 36.8, and 53.5 in Ethiopia, Kenya, and Nigeria, respectively. In contrast, the corresponding numbers in Brazil, Egypt, India, SEA -0.96 -20.2 -0.01 -0.2 and Indonesia are 4.8, 1.3, 21.2, and 5.7, respectively. Generally, using the regional definitions deployed by CGIAR, the US$1.90 1 For reference, 40% of per capita income was almost exactly US$1.90 per day in 2011 PPP in poverty line is relevant in WCA, ESA, and SA. However, because Ethiopia in 2017. consumption levels are higher in other regions—such as LAC, 2 Klasen and Misselhorn (2008) estimate a matrix of semi-elasticities which vary according to the location of the poverty line relative to mean per-capita income and the Gini coefficient. A convenience CWANA, and SEA—the relatively small share of the population of the 40% threshold is that estimated poverty growth semi-elasticities are essentially insensitive to the living below the US$1.90 poverty line potentially disguises value of the Gini coefficient within relevant ranges. For this poverty line, we apply a growth-poverty the implications of One CGIAR activities for the broader semi-elasticity of 0.37 for all regions. For the US$1.90 poverty line, we assume a Gini coefficient of 0.5 for LAC and SA and of 0.4 for all other regions. We calculate the US$1.90 per day value, about population. Differences in results across regions principally $694 per year, relative to projected per-capita GDP in 2030 under the baseline, and we then draw arise from CGIAR’s regional focus and the efficient translation poverty-growth semi-elasticities from the matrix, interpolating linearly between nearest values. Projected Benefits of CGIAR Research 35 As expected, percentage point poverty reductions for the agriculture increase overall employment, it also reduces region-specific absolute lines are much higher because the gender disparity (Frija et al. 2020). many more people live “near” but below these higher Note that the approach in the Frija et al. study (2020) departs poverty lines. The increments to income delivered by slightly from the from the core results used throughout CGIAR move a much larger share of people across these most of this report. Rather than the highly coordinated higher lines. By the same token, poverty reduction is and optimistic COMP scenario, Frija et al. evaluated a much lower when the US$1.90 line is employed because larger set of the many building block scenarios that make relatively few people live “near” but below this lower line. up much of the COMP scenario. Also, while the majority of A related model that links agricultural R&D investments the results presented in the other sections of this analysis to total factor productivity growth similarly captures the use a climate change scenario as the “business-as-usual” potential poverty-reducing effects of agricultural growth (BAU) counterfactual, another strategy is to use a “no as result of One CGIAR investments (Table 5.2.2) (Fuglie climate change” (NoCC) scenario. The BAU used by Frija et 2018a, 2018b; Fuglie et al. 2019). The results of the two al. follows the latter approach and assumes current levels approaches are reasonably consistent and offer a range of productivity without climate-related yield shocks (Frija of potential poverty-reducing effects associated with One et al. 2020). For the purposes of understanding projected CGIAR investments in high-yielding varieties and improved benefits, of the scenarios modeled by Frija et al. (2020), the research efficiency under conditions of climate change. one that most closely maps to the COMP scenarios used elsewhere in this report is the “HIGH+RE” scenario. This Table 5.2.2. Simulated total factor productivity growth of the One CGIAR scenario is characterized as involving a “[h]igh increase in portfolio relative to business as usual. R&D investment across the CGIAR portfolio plus increased research efficiency,” and thus accounts for much of the Total factor Change in poverty investment impact seen in COMP (Rosegrant et al. 2017).Region productivity growth (%) levels (%) Considering trends in female employment relative to total Sub-Saharan 0.83 -0.25 employment and comparing them to changes in rural Africa development together indicate the gendered effects of CWANA rural transformation. We present this analysis of selected 0.66 -0.20 countries for which information is available through SA 2.40 -0.72 International Labour Organization statistics. An often- SEA 0.69 -0.21 stated hypothesis holds that there is a tendency towards increased prevalence of women in agriculture, especially LAC 2.49 -0.75 in low-income countries with emerging out-migration Industrialized 1.5 -0.45 patterns (Slevchevska et al. 2019; Ravula 2019). However, in countries macro-level ILO data, we observe many different patterns. The emergent patterns are not directly linked to GDP per capita or to region. Among the countries under analysis, 5.3. Gender, youth, and social clear feminization of agriculture combined with increased inclusion urbanization only occurs in Bangladesh, Colombia, Costa Rica, and Afghanistan, and to a lesser extent in Ecuador. Gender, youth, and inclusion are key themes in agricultural In contrast, feminization of agriculture is declining in development that are both affected by and influence the many other countries, irrespective of whether the rural outcomes of investments in agricultural research. Our population is changing. Yet there are differences in whether aim in this analysis was to identify implications of CGIAR the percent of women employed in agriculture is higher or investments on gender-related issues. These themes are lower than the percent of the rural population. Preliminary often omitted from studies that do not otherwise focus findings from the countries analyzed suggest that percent specifically on them. Though the IMPACT model does of women working in agriculture are higher than the rural not explicitly consider sex-disaggregated data, a recent population in Armenia, Georgia, and South America, but study building on the Rosegrant et al. (2017) IMPACT lower in Central America, the Middle East, and Asia. analysis shows that for 8 of 14 analyzed countries in Africa, Returning to Frija et al. (2020), the relative potential for investments in agricultural research served to increase agriculture investment to improved outcomes in a sex- female employment more than male employment (Frija et disaggregated manner is demonstrated via three countries, al. 2020). It finds that not only does targeted investment in spanning WCA, ESA, and CWANA (Figures 5.3.1 a, b, c). Again, 36 these illustrate a plausible future associated with substantial increases in investment across the CGIAR portfolio and, likewise, better research efficiency compared to the NoCC scenario. The One CGIAR impact indicators call for targeted improvement in women’s empowerment and inclusion in the agricultural sector, in the number of women and youths benefiting from relevant CGIAR innovations, and in the number of women helped to exit poverty. A comparison of the COMP and REF scenarios based on data from the IMPACT assessment supports the idea that agriculture can serve as a mechanism to reduce gender disparities in agricultural development though additional research is merited. Figure 5.3.1. Support for the role of agriculture in addressing gender-related impact indicators. a. NIGERIA 2.50 2.00 1.50 1.00 0.50 0.00 2025 2035 2045 2025 2035 2045 Female Male High + RE REF - NoCC b. KENYA 2.50 2.00 1.50 1.00 0.50 0.00 2025 2035 2045 2025 2035 2045 Female Male High + RE REF - NoCC c. TUNISIA 0.80 0.70 0.60 0.50 0.40 0.30 0.20 0.10 0.00 2025 2035 2045 2025 2035 2045 Female Male High + RE REF - NoCC Neil Palmer / CIAT Projected Benefits of CGIAR Research 37 Annual increase of employment (in %) Annual increase of employment (in %) Annual increase of employment (in %) As shown, in key instances, the annual increase in of employment in agriculture can be substantially higher for women. These results must be contextualized in each specific instance, however, in that greater agricultural employment can be either a positive or negative outcome demanding on context, the specific value chains, and a variety of other intervening factors. The full set of results is shared in Table 5.3.1, with greater relative change for women indicated in the last column. Table 5.3.1. Relative changes in employment by gender. Female (% change) Male (% change) Country Scenario Relative % change > for women 2025 2035 2045 2025 2035 2045 HIGH + RE 0.99 1.54 2.05 0.91 1.40 1.87 * Nigeria REF-NoCC 0.56 0.74 1.31 0.51 0.68 1.20 * HIGH + RE 0.60 0.91 1.15 0.63 0.95 1.21 Ghana REF-NoCC 0.34 0.44 0.74 0.35 0.46 0.77 HIGH + RE 0.42 0.70 1.05 0.43 0.70 1.06 Niger REF-NoCC 0.24 0.34 0.67 0.24 0.34 0.68 HIGH + RE 0.31 0.50 0.71 0.31 0.50 0.71 Mali REF-NoCC 0.17 0.24 0.45 0.17 0.24 0.45 HIGH + RE 0.37 0.55 0.71 0.35 0.51 0.66 * Senegal REF-NoCC 0.21 0.27 0.45 0.19 0.24 0.42 * HIGH + RE 0.57 0.90 1.30 0.29 0.46 0.74 * Ethiopia REF-NoCC 0.35 0.47 0.89 0.18 0.24 0.50 * HIGH + RE 1.03 1.56 2.09 0.97 1.47 1.96 * Kenya REF-NoCC 0.63 0.81 1.42 0.60 0.77 1.34 * HIGH + RE 0.17 0.25 0.34 0.11 0.16 0.21 * Sudan REF-NoCC 0.10 0.13 0.23 0.06 0.08 0.14 * HIGH + RE 0.10 0.17 0.26 0.10 0.17 0.26 Uganda REF-NoCC 0.06 0.09 0.18 0.06 0.09 0.18 HIGH + RE 0.40 0.67 0.97 0.43 0.72 1.03 Tanzania REF-NoCC 0.25 0.35 0.66 0.26 0.37 0.70 HIGH + RE 0.50 0.65 0.69 0.23 0.30 0.32 * Tunisia REF-NoCC 0.21 0.24 0.35 0.10 0.11 0.16 * HIGH + RE 0.08 0.10 0.12 0.27 0.37 0.42 Morocco REF-NoCC 0.03 0.04 0.06 0.12 0.14 0.21 HIGH + RE 0.11 0.13 0.12 0.04 0.05 0.04 * Algeria REF-NoCC 0.04 0.04 0.05 0.02 0.02 0.02 * HIGH + RE 0.23 0.32 0.37 0.18 0.25 0.29 * Egypt REF-NoCC 0.10 0.12 0.19 0.08 0.09 0.15 * Source: IMPACT model results based on Rosegrant et al. 2017. 38 Figure 5.3.2. Additional employment under the investment scenario. Additional Accumulated Employment: HIGH + RE scenario implementation in selected countries 10000 9000 8000 7000 6000 5000 4000 3000 2000 1000 0 SSA-Nigeria SSA-Ghana SSA-Niger SSA-Mali SSA-Senegal SSA-Ethiopia SSA-Kenya SSA-Sudan SSA-Uganda SSA-Tanzania MEN-Tunisia MEN-Morocco MEN-Algeria MEN-Egypt Additional women employed (in 1000s) by 2030 Additional men employed (in 1000s) by 2030 Source: Results presented from Frija et al. 2020. Youth unemployment is a major issue globally (Figures 5.3.3–5.3.5). Rural youth unemployment data, both in total and divided between young men and women, clearly indicate the hotspots where agricultural research for development should have a strong youth focus. The information is not widely available, and therefore we were unable to generate meaningful trends. Figure 5.3.3. Rural youth unemployment in general (latest figures where available). Youth unemployment rate 2019 % No data 0.1 - 5.0 5.1 - 10.0 10.1 - 25.0 Source: International Labour Organization 2020. 25.1 - 50.0 Data source: ILO, FAO GAUL 50.1 - 61.9 Projected Benefits of CGIAR Research 39 Figure 5.3.4. Rural male youth unemployment (latest figures where available). Male youth unemployment rate 2019 % 0.0 0.1 - 5.0 5.1 - 10.0 10.1 - 25.0 Source: International Labour Organization 2020. 25.1 - 50.0 Data source: ILO, FAO GAUL 50.1 - 57.0 Figure 5.3.5. Rural female youth unemployment (latest figures where available). Female youth unemployment rate 2019 % No data 0.1 - 5.0 5.1 - 10.0 10.1 - 25.0 Source: International Labour Organization 2020. 25.1 - 50.0 Data source: ILO, FAO GAUL 50.1 - 73.1 40 Young people are increasingly linked to targeted agriculture projected benefits associated with the COMP scenario and food security interventions; however, the evidence of underscores that effective mitigation of agriculture-related the effect of agricultural R&D on the youth is missing for emissions will require regionally tailored investment all intents and purposes (Ripoll et al. 2017). Gender and strategies. Various factors affect total emissions across youth were both recently highlighted as the biggest gaps regions (Appendix 3). Net emissions for agriculture are in foresight during a commissioned Independent Science generally expected to decrease, although less in SA overall for Development Council review on the topic. These gaps and in SEA as regards crop and animal emissions (Tables require a focused research agenda (Ripoll et al. 2017). 5.4.1 a, b, c). The increases in SA and SEA are largely attributable to rice production and to considerable jumps 5.4. Climate adaptation and in the expected use of and emissions from nitrogen greenhouse gas reduction fertilizers (Appendix 3). These results suggest that, with some exception, COMP scenario investments contribute to Climate adaption and GHG emissions reduction are a the relevant One CGIAR impact indicator by reducing CO2eq core priority in agricultural development. The agricultural emissions (Figures 5.4.1 a, b, c). sector has profound impacts on the environment by nature of the activity itself. Biodiversity loss relates to agricultural practices that place exceptional pressure on the environment above and beyond the inevitable Figure 5.4.1. Support for #tonnes CO2 equivalent emissions. pressures associated with agriculture itself. Opportunities to increase climate adaptation and reduce emissions are linked to agriculture from farm to fork, and there is a wide a. Crops & Animals Emissions 2030 variety of approaches—from climate services to improved agronomy—that enhance farmer productivity and 1.0 resilience while helping the food system function within environmental limits (Springmann et al. 2018). As with 0.8 many of the other potential impacts throughout this report, 0.6 computation of projected benefits for specific approaches to adaptation and mitigation is context-specific. 0.4 The results presented here summarize anticipated agriculture-related emissions under the conditions 0.2 associated with the COMP scenario within the IMPACT model. Generally positive consequences follow upon 0.0 greater productivity per unit of input, offset land CWANA ESA LAC SA SEA WCA CGIAR Regions conversion, and changing diets. Negative results tend to be associated with heightened methane emissions from rice COMP REF production and increased use of nitrogen. In the IMPACT model, these figures were calculated using loosely coupled GLOBE and Land Simulation to b. Land Use Change CO2 Emissions 2030 Harmonize and Integrate Freshwater Availability and the 1.0 Terrestrial Environment (LandSHIFT) models to understand environmental and mitigation aspects. Unlike many other 0.8 projected benefits, the favorable and unfavorable impacts associated with the COMP scenario are heterogeneous 0.6 across regions. There exists regional differentiation in the favorable and unfavorable impacts associated with the 0.4 COMP scenario in tonnes of carbon dioxide equivalent (CO2eq) by region for crops and animals, land use change, 0.2 and all agricultural emissions, respectively (Figures 5.4.1 a, b, c). There is also differentiation in methane emissions 0.0 CWANA ESA LAC SA SEA WCA from ruminants and rice as well as in nitrous oxide CGIAR Regions emissions (Appendix 3). The regional heterogeneity in the COMP REF Projected Benefits of CGIAR Research 41 Gt CO2eq Gt CO2eq c. All Agriculture CO2 Emissions 2030 c. All agricultural emissions: Differences between the COMP and reference scenarios 2.0 Region Reference COMP Difference % 1.5 (Gt CO eq) difference2 CWANA 0.4274 0.4238 -0.0036 -0.8 1.0 ESA 0.649 0.6254 -0.0235 -3.6 LAC 1.1933 0.8291 -0.3642 -30.5 0.5 SA 0.6604 0.6837 0.0233 3.5 SEA 1.4736 1.2867 -0.1869 -12.7 0.0 CWANA ESA LAC SA SEA WCA WCA 0.6232 0.5981 -0.0252 -4 CGIAR Regions COMP REF Source: IMPACT model results based on Rosegrant et al. 2017. Table 5.4.1. Absolute and relative difference in GHG emissions. For the vast majority of the six CGIAR target regions where CGIAR-mandate crops are grown, the effects of climate a. Crop and animal emissions: Differences between the COMP and reference scenarios change will be negative without adaptation according to spatial modeling (Figures 5.4.2–5.4.9). Areas where we Region Reference COMP Difference % observe positive effects from climate change tend to be (Gt CO eq) difference2 high-income. In the CGIAR research portfolio, the genetic CWANA 0.3967 0.3962 -6e-04 -0.1 innovation action area focuses on breeding for climate change adaptation to generate varieties that meet new ESA 0.1607 0.1597 -0.0010 -0.6 abiotic stress levels while maintaining yield levels. This LAC 0.4919 0.4723 -0.0197 -4.0 effort means that the direst possibilities (Figures 5.4.2– 5.4.9) will not materialize in our target areas. Other results SA 0.6251 0.6606 0.0355 5.7 agree with previous analysis of the impacts of climate SEA 0.8626 0.8831 0.0205 2.4 change on maize in Africa (Figure 5.4.2) (Tesfaye et al. WCA 2015). The potential benefits of climate change adaptation 0.156 0.1524 -0.0035 -2.2 have also been documented (Tesfaye et al. 2017; 2018). Our findings for wheat are also in line with other studies, and the adaptation potential is clearly documented as well (Figure 5.4.3) (Chenu et al. 2017; Hernandez-Ochoa et al. b. Land use change emissions: Differences between the COMP and reference scenarios 2018; Pequeno et al. 2021). The data presented below offer preliminary indications of the importance of maintaining Region Reference COMP Difference % and further developing the gene bank capacity of One (Gt CO eq) difference2 CGIAR (Figures 5.4.2–5.4.9). These findings underscore CWANA 0.0307 0.0277 -0.0030 -9.8 the multiplier effects of adequate germplasm acquisitions (Castañeda-Álvarez et al. 2016; Galluzi et al. 2016). ESA 0.4883 0.4657 -0.0226 -4.6 LAC 0.7013 0.3568 -0.3445 -49.1 SA 0.0353 0.0231 -0.0122 -34.6 SEA 0.6111 0.4036 -0.2074 -33.9 WCA 0.4672 0.4456 -0.0216 -4.6 42 Gt CO2eq Figure 5.4.2. Irrigated and rain-fed maize yield changes without improved, climate-resilient varieties. Figure 5.4.3. Irrigated and rain-fed wheat yield changes without improved, climate-resilient varieties. Figure 5.4.4. Irrigated and rain-fed sorghum yield changes without improved, climate-resilient varieties. Projected Benefits of CGIAR Research 43 Figure 5.4.5. Irrigated and rain-fed rice yield changes without improved, climate-resilient varieties. Figure 5.4.6. Irrigated and rain-fed bean yield changes without improved, climate-resilient varieties. Figure 5.4.7. Irrigated and rain-fed chickpea yield changes without improved, climate-resilient varieties. 44 Figure 5.4.8. Irrigated and rain-fed groundnut yield changes without improved, climate-resilient varieties. Figure 5.4.9. Irrigated and rain-fed potato yield changes without improved, climate-resilient varieties. To explore the effects of climate adaptation, we rely on results from a working report by Ramirez et al. (see Special Recognition in the Acknowledgements). Ramirez et al. first calculate the number of potential climate adaptation beneficiaries among rural people and households. They base their calculations on geospatial datasets about climate hazards and rural population as well as on a conservative estimate of 2% annual adoption of climate-smart agriculture (CSA) practices and climate information services (CIS) within these hazard areas. Second, they compute the average mean yield and yield stability benefits using Evidence for Resilient Agriculture (ERA; http://era.ccafs.cgiar.org) (see Box 5.4.1). Projected Benefits of CGIAR Research 45 The projected beneficiaries of climate adaptation work, including people and households, were also aggregated by region in a sister report. The largest numbers of beneficiaries are in SA and SEA. At 25% and 24%, respectively, India and China contribute circa 50% of the total number of beneficiaries. Individuals and households projected to benefit from CSA and CIS work were aggregated by region (Figures 5.4.10 a and b). We also present the projected beneficiaries by year in non- cumulative values at the regional and global scales in terms of the total population, individuals, and households (Tables 5.4.2 a and b). Ramirez et al. also provide a detailed breakdown of the yield benefits associated with adaptation measures. By 2030, a total of 234.1 million people and 59.1 million households could benefit from CGIAR’s climate adaptation work. Figure 5.4.10. Cumulative projected beneficiaries of CGIAR investments in climate adaptation from 2022–2030 in terms of rural individuals (a) and households (b). Jeffery M Walcott / IWMI 200 Box 5.4.1. 150 The Evidence for Resilient Agriculture (ERA) harmonizes and synthesizes 40 years of research on crop, livestock, and tree management in Africa. 100 ERA details the impact of 300-plus combinations of management practices on 75 indicators of productivity, resilience, and greenhouse 50 gas emissions. The dataset includes about 115,000 observations drawn from more than 2,000 studies. Meta-analysis of ERA offers essential 0 information about potential benefits when changing farming practices, 2022 2023 2024 2025 2026 2027 2028 2029 2030 showing both the most likely outcomes and the less likely, but possible, extremes. ERA data can be used to evaluate the potential of multiple 60 options against each other, examine synergies when using practices in combination, assess tradeoffs between social and environmental outcomes, and more. The data are georeferenced. Therefore, it is possible to combine them with publicly available climate and soils data 40 and predict suitability of technologies in locations other than where they were studied and repurpose historical agricultural data to answer today’s pressing questions such as quantifying the resilience benefits 20 when using improved farming technologies, filling critical knowledge gaps. ERA is also a data source, providing accessible, transparent, and coherent input data for economic and financial analysis and risk assessment. ERA has informed national investment plans, project 0 proposals, adaptation policies, research agendas, and financial service 2022 2023 2024 2025 2026 2027 2028 2029 2030 products. ERA currently only includes data collected in Africa but future extension in geographic and topical scope are forthcoming. SEA SA CWANA LAC ESA WCA 46 Cumulative beneficiaries (million households) Cumulative beneficiaries (million people) Table 5.4.2. Annual numbers in millions of people (a) and of households (b) benefiting from CGIAR’s climate adaptation work from 2022–2030, aggregated by region. a. Annual numbers, millions of people, benefiting from CGIAR’s climate adaptation work from 2022–2030 Region 2022 2023 2024 2025 2026 2027 2028 2029 2030 Total adopters Total population CWANA 3.09 3.02 2.96 2.90 2.85 2.79 2.73 2.68 2.63 25.7 154.3 ESA 2.18 2.14 2.09 2.05 2.01 1.97 1.93 1.89 1.85 18.1 108.9 LAC 2.01 1.97 1.93 1.89 1.85 1.82 1.78 1.75 1.71 16.7 100.5 SA 9.60 9.41 9.22 9.03 8.85 8.68 8.50 8.33 8.17 79.8 479.9 SEA 9.11 8.93 8.75 8.58 8.41 8.24 8.07 7.91 7.75 75.8 455.7 WCA 2.17 2.13 2.09 2.04 2.00 1.96 1.92 1.89 1.85 18.1 108.6 Total 28.16 27.60 27.05 26.50 25.97 25.46 24.95 24.45 23.96 234.1 1,408.0 b. Annual numbers, millions of households, benefiting from CGIAR’s climate adaptation work, 2022–2030 Region 2022 2023 2024 2025 2026 2027 2028 2029 2030 Total adopters Total population CWANA 0.667 0.653 0.640 0.628 0.615 0.603 0.591 0.579 0.567 5.5 33.3 ESA 0.520 0.509 0.499 0.489 0.479 0.470 0.460 0.451 0.442 4.3 26.0 LAC 0.577 0.566 0.554 0.543 0.532 0.522 0.511 0.501 0.491 4.8 28.9 SA 2.005 1.964 1.925 1.887 1.849 1.812 1.776 1.740 1.705 16.7 100.2 SEA 2.906 2.848 2.791 2.735 2.680 2.627 2.574 2.523 2.472 24.2 145.3 WCA 0.439 0.430 0.422 0.413 0.405 0.397 0.389 0.381 0.373 3.6 21.9 Total 7.113 6.971 6.831 6.695 6.561 6.430 6.301 6.175 6.052 59.1 355.7 5.5. Environmental health and biodiversity Environmental health and biodiversity span many dimensions of the One CGIAR agenda. From agro-biodiversity and biodiversity more generally to deforestation, ecosystem services, and water use, a healthy natural environment is critical for the long-term sustainability of the food system. CGIAR has identified impact indicators specifically addressing improved land management, water use, deforestation, application of nitrogen, and agrobiodiversity. Though many studies offer indicators of projected climate change impacts on environmental systems, there is a substantial opportunity to improve estimates to understand the potential direct and indirect benefits of the One CGIAR portfolio on these systems. As with many other indicators, the activities that influence water use will span much of the CGIAR portfolio and the innovation pillars required to maximize potential benefits. This complex set of interactions results in a limited range of possibilities for the projected benefits calculated. In this section, we focus on water use results derived from the IMPACT model, draw on the LandSHIFT analysis of nitrogen-related emissions, and discuss other findings related to the LandSHIFT model. As mentioned, LandSHIFT is an external model that was loosely coupled with the IMPACT model runs associated with the Rosegrant et al. 2017 study that serves as the basis of the present analysis. Water use sustainability issues occur throughout the food system. The COMP scenario from the IMPACT model used here includes consideration of irrigation expansion, water use efficiency investments, and investments targeted to increased soil water holding capacity (see also Section 4.1). The projected benefits therefore relate to “irrigation water use” and “precipitation on crops”. These indicators approximate overall efficiency in agricultural water use. Irrigation water use, also referred to as “blue water”, indicates the amount of water used in irrigation. Precipitation on crops, also referred to Projected Benefits of CGIAR Research 47 as “green water”, indicates the amount of precipitation Table 5.5.1. Absolute and relative changes in water use. that falls on and can be utilized by crops; this indicator may account for crop locations optimized for climate or a. Irrigated water use in cubic kilometers: improved soil water holding capacity via R&D mechanisms Differences between the COMP and reference scenarios (Figures 5.5.1 a and b; Tables 5.5.1 a and b). Region Reference COMP Difference % Throughout each of the One CGIAR regions, the use of blue difference water, which we believe is at least moderately consistent CWANA 328.995 298.6417 -30.3533 -9.2 with consumptive use, is somewhat reduced under the COMP scenario reflective of One CGIAR investments. ESA 53.1536 49.1057 -4.0479 -7.6 Less blue water consumption is partially due to potential LAC 164.8272 147.7557 -17.0715 -10.4 increases in green water use resulting from investments SA 695.2581 651.7459 -43.5122 -6.3 in heightened water holding capability. Though somewhat modest in percentage terms, decreases in water for SEA 395.7726 343.2039 -52.5687 -13.3 irrigation and complimentary augmentations of green WCA 18.8328 17.5525 -1.2803 -6.8 water use occur throughout the CGIAR regions. b. Rainfall on crops in cubic kilometers: Differences between the COMP and reference scenarios Figure 5.5.1. Support for #km3 consumptive water use in food production. Region Reference COMP Difference % difference CWANA 37.3239 40.2726 2.9487 7.9 a. Irrigated water use ESA 27.0192 36.1733 9.1541 33.9 LAC 119.4276 146.1003 26.6727 22.3 600 SA 492.406 518.3546 25.9486 5.3 400 SEA 589.0776 622.2695 33.1919 5.6 200 WCA 11.4455 15.0988 3.6533 31.9 0 Source: IMPACT model results based on Rosegrant et al. 2017. CWANA ESA LAC SA SEA WCA CGIAR Regions COMP REF The LandSHIFT model can offer additional insights into b. Precipitation on crop area the potential benefits of the COMP scenario (Rosegrant et al. 2017). The model showed that low levels of investment resulted in large amounts of agricultural extensification, 600 likely leading to increased deforestation and a variety of other negative feedbacks including higher water 400 consumption and GHG emissions (Rosegrant et al. 2017). Additionally, again via the LandSHIFT model, high-level 200 investment scenarios substantially relieved pressure on biodiversity and even reduced species loss to below the 0 CWANA ESA LAC SA SEA WCA levels expected with climate change alone (Rosegrant et al. CGIAR Regions 2017). In general terms, according to the LandSHIFT model, COMP REF the COMP scenario has favorable impacts on deforestation (Table 5.5.2) (Rosegrant et al. 2017). Across the board, the COMP scenario results in less forest coverage lost than the reference scenario by 2030. As with any macro-level investments, however, there are trade-offs, and the road expansion associated with the infrastructure improvement component of the COMP scenario somewhat diminishes the 48 Cubic km Cubic km positive outcome associated with easing the pressure on As the above section illustrates, fertilizer is an area where forestland conversion. The LandSHIFT model results also several tradeoffs can occur in relation to the overall showed that due to heightened expansion of agricultural agricultural development agenda. Insufficient fertilizer lands, investments in irrigation and other infrastructure results in poor agricultural productivity, while excessive may lead to increased losses of bird species (Rosegrant et or mistimed fertilizer can result in both environmental al. 2017). contamination and lower profitability. In the approximately fifteen-year period between 1999 and 2014, fertilization rates varied by crop and region for each of three macro- nutrients: nitrogen (N), phosphorous pentoxide (P2O5), and Table 5.5.2. Changes in forest cover by FAO regions: potassium oxide (K2O) (Figures 5.5.2-5.5.4). Each composite Differences between the COMP and reference scenarios. graph provides extant fertilization data for a major crop or crop group. Each crop graph, meanwhile, contains boxplots Forest loss of fertilizer rate distributions per region. The six CGIAR Region Data Reference COMP avoided (km2) target regions, the former Soviet Union, and other high- income countries are included for comparison. Cereals East Asia and 2020 2990666 2990666 account for the highest fertilizer consumption globally, as the Pacific, only including 2030 2977254 2982346 5092 they also occupy large production areas. The data are not low- and middle- complete or comprehensive but do indicate key trends, income countries Change (km2) -13412 -8320 specifically that consumption rates are veering upwards. 2020 9715197 9717064 The optimal range of fertilization for different plant LAC 2030 9703347 9715099 9885 macro-nutrients can in general be defined as follows, Change (km2) -11850 -1965 obviously depending upon soil types, prior fertilization rates, and hence preexisting soil fertility, as well as crop 2020 5871615 5871784 type. Legumes and pulses generally require less nitrogen Sub-Saharan Africa 2030 5865455 5866880 1256 fertilization than other crops. Change (km2) -6160 -4904 N: 50-150 kg/ha P2O : 20-80 kg/ha K O: 20-60 kg/ha2020 332094 332169 5 2 SA 2030 331317 331959 567 Change (km2) -777 -210 2020 136914 136914 0 Middle East and North Africa 2030 136847 136847 0 Change (km2) -67 -67 0 Fertilizer use is another factor in environmental health that is associated with environmental pressures in two ways. Firstly, overuse of fertilizers leads to emissions of plant nutrients into the environment. Secondly, nitrate leaching is well documented and can decimate life in estuaries, coastal areas, seas, and oceans. Especially in higher-income countries, where the cost of fertilizer is low relative to other production costs, overuse of fertilizers is rampant. This is also the case in countries where fertilizers are heavily subsidized. On the other hand, in many poorer countries, structurally low fertilization rates in agriculture contribute to soil mining with heightened erosion risks in consequence; this situation ultimately results in Chris de Bode / CGIAR desertification. Projected Benefits of CGIAR Research 49 Figure 5.5.2. Fertilizer rates for key crops and crop groups from 1999-2004. CWA=Central and West Asia; DW-Hi=high income countries; and FSU=Former Soviet Union. Rate in surface by crop type and region, 1999-2004 Maize Other cereals Pulses SEA LAC SA LAC LAC ESA FSU ESA ESA DW-HI DW-HI CWANA DW-HI CWA CWA 0 50 100 150 200 0 50 100 150 0 50 100 150 Rice-paddy Roots and tubers Wheat SEA SA SEA LAC LAC FSU FSU SA ESA ESA DW-HI DW-HI LAC CWA CWA 0 50 100 150 0 50 100 150 0 50 100 150 Nutrition rate (Kg/ha) Nutrient: K2O N P2O5 Figure 5.5.3. Fertilizer rates for key crops and crop groups from 2005-2009. Rate in surface by crop type and region, 2005-2009 Maize Rice-paddy Wheat SEA SEA SEA SA SA SA LAC LAC LAC FSU FSU FSU ESA ESA DW-HI DW-HI DW-HI CWANA CWANA CWANA CWA CWA CWA 0 50 100 150 0 50 100 150 200 0 50 100 150 Nutrition rate (Kg/ha) Nutrient: K2O N P2O5 50 CGIAR region CGIAR region Figure 5.5.4. Fertilizer rates for key crops and crop groups from 2010-2014. Rate in surface by crop type and region, 2010-2014 Maize Rice-paddy Wheat SEA SEA SEA SA SA SA LAC LAC LAC FSU FSU FSU ESA ESA DW-HI DW-HI DW-HI CWANA CWANA CWANA CWA CWA CWA 0 50 100 150 200 0 50 100 150 0 50 100 150 200 Nutrition rate (Kg/ha) Nutrient: K2O N P2O5 Projected Benefits of CGIAR Research 51 Crop type 6. Conclusion This report represents the first installment of the report utilized two different poverty lines – the familiar assessment of projected benefits in support of the One US$1.90 in 2011 PPP benchmark and regional poverty lines CGIAR design and planning process. Several opportunities set at 40% of per-capita income in 2030. This latter method of remain to improve this report: providing additional approach yielded larger numbers of beneficiaries moved out information on livestock’s contribution to both positive and of poverty through increments to income delivered by CGIAR. negative externalities, deeper analysis of the biodiversity Gender, youth, and social inclusion together make up the impacts associated with different investment scenarios, third impact area. This report highlights the urgent need and enhanced evaluation of the water impacts associated for further research and foresight analysis that explicitly with the same. The authors will consider feedback from the tackles these considerations. Agriculture is a promising review process and further develop these features, along mechanism to boost female employment and reduce with the further in-depth examination of key variables as gender disparities in economic development. Yet due to required. gaps in existing data, it is difficult to make projections Given that evaluation of One CGIAR investments is an about trends in female and youth employment in ongoing process, this analysis has also taken steps to agriculture. support further studies in a similar vein. IMPACT supported In terms of climate adaptation and greenhouse gas by ADAM, GLOBE, and LandSHIFT, enabled us to explore reduction – the fourth impact area – results varied across a COMP scenario that projected the potential effects of CGIAR regions, underscoring the necessity of tailored CGIAR investments. The IFPRI Spatial Production Allocation investment strategies. Generally positive consequences Model database3 on crop location in combination with follow upon greater productivity per unit of input, offset the Mink gridded crop modeling framework, meanwhile, land conversion, and changing diets; negative climate supported evaluation of the scenario characterized only adaptation results tend to be associated with heightened by climate change, keeping other factors constant. This methane emissions from rice production and increased analysis is also designed to dovetail with the GIF approach use of nitrogen. Across the regions under analysis, CGIAR and RIAPA, which furnish opportunities to undertake genetic research can help avert the worst possible climate deeper analysis into the potential benefits and trade-offs change-related futures by furnishing varieties that are associated with One CGIAR investments. adapted to climate hazards. When it comes to climate These investments are projected to result in benefits adaptation work, at 25% and 24%, respectively, India across the five CGIAR impact areas as compared with a and China contribute circa 50% of the total number of future scenario characterized only by climate change. The beneficiaries of CGIAR investments. first impact area is nutrition, health, and food security. A The fifth impact area is environmental health and comprehensive, integrated CGIAR investment strategy can biodiversity. With some potential trade-offs, under high- shrink the population at risk of hunger, reduce the number investment scenarios, pressure on biodiversity was of malnourished children, and augment protein and substantially relieved, and species loss was reduced to micronutrient availability. Investments in CGIAR research below the levels expected with climate change alone. can therefore help grow the number of people meeting Deforestation was also reduced, which in turn could reduce minimum dietary energy requirements and minimum pressure on biodiversity even further. Throughout all the micronutrient requirements worldwide. One CGIAR regions, the use of blue water, indicative of Poverty reduction, livelihoods, and jobs together comprise water consumption, was also somewhat reduced. Fertilizer the second CGIAR impact area. Poverty is a multifaceted use entails many trade-offs and much regional variation; phenomenon requiring multiple approaches to capture in general, rates are increasing worldwide. Proper fertilizer its various contributing causes and aspects. In addition use can positively contribute to intensification, as long as to IMPACT and ADAM, this report also draws on literature awareness of potentially adverse impacts is maintained in review, plus MIRAGRODEP and its linked tools, the POVANA order to reduce negative impacts on the environment. model, and datasets including FAOSTAT. Results suggest Overall, the integrative strategy of the One CGIAR has the that productivity and GDP increases associated with CGIAR potential to result in substantial benefit to both people and investments may be linked with decreases in poverty. To planet. This report on the projected benefits of the CGIAR further capture the complexities of defining poverty, this research portfolio will hopefully serve to inform the future of 3 https://www.mapspam.info/ the CGIAR and agrifood system development more broadly. 52 References Adamopoulos T; Restuccia D. 2014. The Size Distribution of Farms Chenery HB; Syrquin M; Elkington H. 1975. Patterns of Development and International Productivity Differences. American Economic 1950-1970. Oxford University Press London. 252 p. Review 104(6):1667–1697. https://doi.org/10.1257/aer.104.6.1667 Chenu K; Porter JR; Martre P; Basso B; Chapman SC; Ewert F; Bindi Alston JM; Marra MC; Pardey PG; Wyatt TJ. 2000. Research Returns M; Asseng S. 2017. Contribution of Crop Models to Adaptation Redux: A Meta-Analysis of the Returns to Agricultural R&D. in Wheat. Trends in Plant Science 22(6):472–490 The Australian Journal of Agricultural and Resource Economics https://doi.org/10.1016/j.tplants.2017.02.003 44(2):185–215. https://doi.org/10.1111/1467-8489.00107 DCED (The Donor Committee for Enterprise Development). 2010. Alston JM; Pardey PG. 2016. Antipodean Agricultural and Resource The DCED Standard for Measuring Achievements in Private Economics at 60: Agricultural Innovation. The Australian Journal Sector Development: Control Points and Compliance Criteria. of Agricultural and Resource Economics 60(4):554–68. Donor Committee for Enterprise Developement. https://doi.org/10.1111/1467-8489.12162 https://bit.ly/3ng02Vu Antle JM. 2011. Parsimonious Multi-Dimensional Impact Delgado CL. 2003. Fish to 2020: Supply and demand in changing Assessment. American Journal of Agricultural Economics global markets (Vol. 62). WorldFish. 93(5):1292–1311. https://doi.org/10.1093/ajae/aar052 Dorward A. 2009. Integrating Contested Aspirations, Processes and Barrett CB; Benton TM; Cooper KA; Fanzo J, Gandhi R; Herrero M, Policy: Development as Hanging in, Stepping up and Stepping James S; Kahn M; Mason-D’Croz D; Mathys A; Nelson RJ; Shen J; Out. Development Policy Review 27(2):131–146. Thornton P; Bageant E; Fan S; Mude AG; Sibanda LM; Wood S. https://doi.org/10.1111/j.1467-7679.2009.00439.x 2020a. Bundling Innovations to Transform Agri-Food Systems. Fan S; Brzeska J; Keyzer M; Halsema A. 2013. From Subsistence to Nature Sustainability 3:974–76. https://go.nature.com/3ubL32X Profit: Transforming Smallholder Farms. Food Policy Report. Barrett CB; Benton T; Fanzo J; Herrero M; Nelson RJ; Bageant E; International Food Policy Research Institute. 30 p. Buckler E; Cooper k; Culotta I; Fan S, Gandhi R; James S; Kahn M; https://bit.ly/3HRZP4l Lawson-Lartego L; Liu J; Marshall Q; Mason-D’Croz D; Mathys Frija A; Chebil A; Mottaleb KA; Mason-D’Croz D; Dhehibi B. 2020. A; Mathys C; Mazariegos-Anastassiou V; Miller A; Misra K; Mude Agricultural Growth and Sex-Disaggregated Employment AG; Shen J; Sibanda LM, Song C; Steiner R; Thornton P; Wood in Africa: Future Perspectives under Different Investment S. 2020b. Socio-Technical Innovation Bundles for Agri-Food Scenarios. Global Food Security 24. Systems Transformation. Report of the International Expert https://doi.org/10.1016/j.gfs.2020.100353 Panel on Innovations to Build Sustainable, Equitable, Inclusive Food Value Chains. Ithica, NY and London: Cornell Atkinson Fuglie K. 2018a. R&D Capital, R&D Spillovers, and Productivity Center for Sustainability and Springer Nature. Growth in World Agriculture. Appl. Econ. Perspect. Policy 40, https://go.nature.com/3OpsRun 421–444. https://doi.org/10.1093/aepp/ppx045 Barrett CB; Beaudreault AR; Meinke H; Ash A; Ghezae N; Kadiyala Fuglie KO. 2018b. Is agricultural productivity slowing? Glob. Food S; Nigussie M; Grove Smith A; Torrance L. 2021. Foresight and Sec. 17, 73–83. http://doi.org/10.1016/J.GFS.2018.05.001 Trade-off Analyses: Tools for Science Strategy Development Fuglie K; Gautam M; Goyal A; Maloney WF. 2019. Harvesting in Agriculture and Food Systems Research. Q Open 1(1):1–7. Prosperity: Technology and Productivity Growth in Agriculture. https://doi.org/10.1093/qopen/qoaa002 License: C. Washington, DC: The World Bank. Capalbo SM; Antle JM; Seavert C. 2017. Next Generation Data http://doi.org/10.1596/978-1-4648-1393-1 Systems and Knowledge Products to Support Agricultural Galluzzi G; Halewood M; Noriega IL; Vernooy R. Twenty-five Producers and Science-Based Policy Decision Making. years of international exchanges of plant genetic resources Agricultural Systems 155:191-199. https://bit.ly/3bDE8cm facilitated by the CGIAR genebanks: a case study on global Castañeda-Álvarez, NP; Khoury CK; Achicanoy HA; Bernau V; interdependence. Biodiversity and Conservation 25:1421–1446. Dempewolf H; Eastwood RJ, et al. 2016. Global conservation https://doi.org/10.1007/s10531-016-1109-7 priorities for crop wild relatives. Nat. Plants 2, 16022. Global Innovation Fund. 2019. Practical Impact: GIF’s Approach to http://doi.org/10.1038/nplants.2016.22 Impact Measurement. London. https://bit.ly/3xUtHZx CGIAR System Organization. 2021. CGIAR 2030 Research and Gollin D; Hansen C; Wingender A. 2016. Two Blades of Grass: The Innovation Strategy: Transforming food, land, and water systems Impact of the Green Revolution. CEPR Discussion Papers. in a climate crisis. Montpellier, France: CGIAR System Organization. https://hdl.handle.net/10568/110918 Herforth A; Bai Y; Venkat A; Mahrt K; Ebel A; Masters WA. 2020. Cost and affordability of healthy diets across and within Chan CY; Tran N; Dao DC; Sulser TB; Philips MJ; Batka M; Wiebe KD; countries: Background paper for The State of Food Security Preston N. 2017. Fish to 2050 in the ASEAN Region. WorldFish and Nutrition in the World 2020. FAO Agricultural Development Center and Intl Food Policy Res Inst. Economics Technical Study No. 9 (Vol. 9). Food & Agriculture Chan CY; Tran N; Pethiyagoda S; Crissman CC; Sulser TB; Phillips MJ. Organization. 2019. Prospects and challenges of fish for food security in Africa. Hermans F; Chaddad F; Gagalyuk T; Senesi S; Balmann A. 2017. Glob. Food Sec. 20, 17–25. The Emergence and Proliferation of Agroholdings and Mega https://doi.org/10.1016/j.gfs.2018.12.002 Farms in a Global Context. International Food and Agribusiness Chen HT. 1994. Theory-Driven Evaluations. Thousand Oaks, CA: SAGE. Management Review 20(2):175–86. Projected Benefits of CGIAR Research 53 Hernandez-Ochoa IM; Asseng S; Kassie BT; Xiong W; Robertson Larson DF; Muraoka R; Otsuka K. 2016. On the Central Role of Small R, Pequeno DNL; Sonder K; Reynolds M; Babare MA; Molero Farms in African Rural Development Strategies. Policy Research A; Hoogenboom G. 2018. Climate Change Impact on Mexico Working Paper No. 7710. World Bank, Washington, DC. Wheat Production. Agricultural and Forest Meteorology 263(32): https://bit.ly/3A3p2H7 373–87. https://doi.org/10.1016/j.agrformet.2018.09.008 Loevinsohn M; Sumberg J; Diagne A; Whitfield S. 2013. Under What Hurley TM; Rao X; Pardey PG. 2014. Re-Examining the Reported Circumstances and Conditions Does Adoption of Technology Rates of Return to Food and Agricultural Research and Result in Increased Agricultural Productivity? A Systematic Development. American Journal of Agricultural Economics Review. Institute of Development Studies. https://bit.ly/3O67LS2 96(5):1492–1504. https://doi.org/10.1093/ajae/aau047 MacDonald JM; Korb P; Hoppe RA. 2013. Farm Size and the International Labour Organization. 2020. ILOSTAT database Organization of US Crop Farming. USDA. https://bit.ly/3QOqpQ7 [database]. Available from https://ilostat.ilo.org/data/ Maslow AH. 1943. A Theory of Human Motivation. Psychological Ivanic M; Martin W. 2018. Sectoral productivity growth and poverty Review 50(4):370–396. https://doi.org/10.1037/h0054346 reduction: National and global impacts. World Development, Mayne J. 2001. Addressing Attribution through Contribution 109, 429–439. Analysis: Using Performance Measures Sensibly. Canadian Jayne TS; Chamberlin J; Headey DD. 2014. Land Pressures, the Journal of Program Evaluation 16(1):1–24. Evolution of Farming Systems, and Development Strategies in Mueller V; Thurlow J. 2019. Youth and Jobs in Rural Africa: Beyond Africa: A Synthesis. Food Policy 48:1–17. Stylized Facts. IFPRI, NY, USA. 323 p. https://bit.ly/3A1Tv8x Jayne TS; Chamberlin J; Traub L; Sitko N; Muyanga M; Yeboah FK; Müller C; Robertson RD. 2014. Projecting Future Crop Productivity Anseeuw W; Chapoto A; Wineman A; Nkonde C; Kachule R. for Global Economic Modeling. Agricultural Economics 45(1): 2016. Africa’s Changing Farm Size Distribution Patterns: The 37–50. https://doi.org/10.1111/agec.12088 Rise of Medium-Scale Farms. Agricultural Economics 47(S1): 197–214. https://doi.org/10.1111/agec.12308 Niroula GS; Thapa GB. 2005. Impacts and Causes of Land Fragmentation, and Lessons Learned from Land Consolidation Jayne TS; Chamberlin J; Benfica R. 2018. Africa’s Unfolding in South Asia. Land Use Policy 22(4):358–72. Economic Transformation. The Journal of Development Studies https://doi.org/10.1016/j.landusepol.2004.10.001 54 (5):777–87. https://doi.org/10.1080/00220388.2018.1430774 Pal S. 2011. Impacts of CGIAR Crop Improvement and Natural Khandker S; Koolwal GB, Hussain S. 2009. Handbook on Impact Resource Management Research: A Review of Evidence. Evaluation: Quantitative Methods and Practices. Learning. Agricultural Economics Research Review 24:185–200. World Bank. https://doi.org/10.1596/978-0-8213-8028-4 Pequeno DNL, Hernandez-Ochoa IX; Reynolds M; Sonder K; Kilby Peter; Johnston BF. 1975. Agriculture and Structural Molero-Milan A; Robertson R; Lopes MS; Xiong W; Kropff M; Transformation: Economic Strategies in Late-Developing Asseng S. 2021. Climate Impact and Adaptation to Heat and Countries. Oxford University Press London. 474 p. Drought Stress of Regional and Global Wheat Production. Kislev Y; Peterson W. 1982. Prices, Technology, and Farm Size. Environmental Research Letters 16(5). https://bit.ly/3HSTZj3 Journal of Political Economy 90(3):578–95. Ramirez-Villegas J; Rosenstock T; Steward P; Thornton P; Klasen S; Misselhorn M. 2008 : Determinants of the growth semi- Loboguerrero A; Jarvis A. 2021. Projected climate adaptation elasticity of poverty reduction, IAI Discussion Papers, No. 176, benefits of One CGIAR. CGIAR Research Program on Climate Georg-August-Universität Göttingen, Ibero-America Institute Change, Agriculture and Food Security (CCAFS). for Economic Research (IAI), Göttingen. https://hdl.handle.net/10568/115780 Krishna VV; Lantican M; Prasanna BM; Pixley KV; Abdoulaye T; Ravula P. 2019. Feminization of agriculture. ICRISAT. Menkir A; Banziger M; Erenstein O. 2021. Impacts of CGIAR https://hdl.handle.net/20.500.11766/10859 Maize Improvement in Sub-Saharan Africa, 1995-2015. CIMMYT. Raworth K. 2017. A Doughnut for the Anthropocene: Humanity’s 44 p. https://bit.ly/3I5Cyw9 Compass in the 21st Century. The Lancet Planetary Health Kruseman G; Bairagi S; Komarek AM; Molero Milan A; Nedumaran 1(2):E48–49. https://doi.org/10.1016/s2542-5196(17)30028-1 S; Petsakos A; Prager S; Yigezu YA. 2020. CGIAR Modeling Riahi K; van Vuuren DP; Kriegler E; Edmonds J; O’Neill B; Fujimori S; Approaches for Resource-Constrained Scenarios: II. Models Bauer N; Calvin K; Dellink R; Fricko O; Lutz W; Popp A; Crespo for Analyzing Socioeconomic Factors to Improve Policy J; KC S; Leimbach M; Jiang L; Kram T; Rao S; Emmerling J; Ebi Recommendations. Crop Science 60(2):568–81. K; Hasegawa T; Havlik P; Humpenöder F; Da Silva L; Smith S; https://doi.org/10.1002/csc2.20114 Stehfest E; Bosetti; Eom J; Gernaat D; Masui T; Rogelj J; Strefler Laborde DW; Tokgoz MS; Vos R. 2020. Contributions of CGIAR R&D J; Drouet L; Krey V; Luderer G; Harmsen M; Takahashi K; to Global Poverty Reduction. Mimeo. Baumstark L; Doelman J; Kainuma K; Klimont Z; Marangoni G; Laborde D; Herforth A; Headey D; de Pee S. 2021. COVID-19 Lotze-Campen H; Obersteiner M; Tabeau A; Tovoni M. 2017. pandemic leads to greater depth of unaffordability of healthy The Shared Socioeconomic Pathways and Their Energy, Land and nutrient-adequate diets in low- and middle-income Use, and Greenhouse Gas Emissions Implications: An Overview. countries. Nat. Food 2, 473–475. Global Environmental Change 42:153–168. http://doi.org/10.1038/s43016-021-00323-8 https://doi.org/10.1016/j.gloenvcha.2016.05.009 Lantican M; Braun H; Payne TS; Singh RP; Sonder K; Baum M; Ripoll S; Andersson J; Badstue L; Büttner M; Chamberlin J; Erenstein Ginkel M; Erenstein O. 2016. Impacts of International Wheat O; Sumberg J. 2017. Rural Transformation, Cereals and Youth Improvement Research, 1994-2014. CIMMYT, Mexico D.F. in Africa: What Role for International Agricultural Research? https://bit.ly/3OGpxLu Outlook on Agriculture 46(3):168–177. https://doi.org/10.1177/0030727017724669 54 Robertson RD. 2017. Mink: Details of a Global Gridded Crop Wiebe K; Sulser TB; Dunston S; Rosegrant MW; Fuglie K; Modeling System. IFPRI, Washington D.C. 261 p. Willenbockel D; Nelson GC. 2020. “Modeling Impacts of Faster https://bit.ly/3NjTEHs Productivity Growth to Inform the CGIAR Initiative on Crops to Robinson S; Mason-D’croz D; Islam S; Sulser TB; Robertson R; Zhu End Hunger.” Submitted to PlosOne. T; Gueneau A; Pitois G; Rosegrant M. 2015. The International World Bank. 2013. “Fish to 2030: Prospects for fisheries and Model for Policy Analysis of Agricultural Commodities and aquaculture.” Washington, D.C.: World Bank. Trade (IMPACT) Model Description for Version 3. IFPRI, Washington DC. 128 p. https://bit.ly/3A1Y1nv Rogers PJ. 2009. Matching Impact Evaluation Design to the Nature of the Intervention and the Purpose of the Evaluation. Journal of Development Effectiveness 1(3):217–226. https://doi.org/10.1080/19439340903114636 Rosegrant MW; Sulser TB; Mason-D’Croz D; Cenacchi N; Nin-Pratt A; Dunston S; Zhu T, Ringler C; Wiebe K; Robinson S; Willenbockel D; Xie H; Kwon H-Y; Johnson T; Thomas TS; Wimmer F; Schaldach R; Nelson GC; Willasrts B. 2017. Quantitative Foresight Modeling to Inform the CGIAR Research Portfolio. IFPRI, Washington, D.C. 225 p. https://bit.ly/3b5s7vW ShiftN. 2009. Global Food System Map. https://bit.ly/3HxWFCK Slavchevska V; Kaaria S; Taivalmaa S. 2019. The Feminization of Agriculture: evidence and implications for food and water security. In: Allan T; Bromwich B; Keulertz M; Colman A (eds.). The Oxford Handbook of Food, Water and Society. Oxford University, Ney York, USA. pp. 268–284. Springmann M; Clark M; Mason-D’Croz D, Wiebe K; Bodirsky BL; Lassaletta L; De Vries W; Vermeulen SJ; Herrero M; Carlson KM; Jonell M; Troell M; DeClerck F; Gordon LJ; Zurayk R; Scarborough P; Rayner M; Loken B; Fanzo JH; Godfray CJ; Tilman D; Rockström J; Willett W. 2018. Options for Keeping the Food System within Environmental Limits. Nature 562:519–525. Steffen W; Richardson K; Rockström J; Cornell SE; Fetzer I; Bennett EM, et al. 2015. Planetary boundaries: Guiding human development on a changing planet. Science. Vol. 347, Issue 6223. http://doi.org/10.1126/science.1259855 Tesfaye K; Gbegbelegbe S; Cairns JE; Shiferaw B; Prasanna BM; Sonder K; Boote K; Makumbi D; Robertson R. 2015. Maize Systems under Climate Change in Sub-Saharan Africa. International Journal of Climate Change Strategies and Management 7(3):247–271. https://doi.org/10.1108/IJCCSM-01-2014-0005 Tesfaye K; Kruseman G; Cairns JE; Zaman-Allah M; Wegary D; Zaidi PHH; Boote KJ; Rahut D; Erenstein O. 2018. Potential Benefits of Drought and Heat Tolerance for Adapting Maize to Climate Change in Tropical Environments. Climate Risk Management 19:106–119. https://doi.org/10.1016/j.crm.2017.10.001 Tesfaye K; Zaidi PH; Gbegbelegbe S; Boeber C; Rahut DB; Getaneh F; Seetharam K; Erenstein O; Stirling C. 2017. Climate Change Impacts and Potential Benefits of Heat-Tolerant Maize in South Asia. Theoretical and Applied Climatology 130(3–4):959–970. https://doi.org/10.1007/s00704-016-1931-6 United Nations, Department of Economic and Social Affairs, Population Division. 2019. World Population Prospects 2019: Highlights. ST/ESA/SER.A/423. Weiss CH. 1997. “How Can Theory-Based Evaluation Make Greater Headway?” Evaluation Review 21 (4): 501–24. https://doi.org/10.1177/0193841X9702100405 Projected Benefits of CGIAR Research 55 Appendices Appendix 1. Fish and aquaculture systems IMPACT fish model In 2003, the IMPACT fish model was first used to produce projections of global food fish production, consumption, and trade from 1997 to 2020 (Delgado et al. 2003). This study, entitled “Fish to 2020: Supply and Demand in Changing Global Markets”, was also the first to include fish in a major global agricultural production and trade model. Approximately 10 years later, the World Bank commissioned a follow-up study in collaboration with the IMPACT modeling team at IFPRI, the Fisheries and Aquaculture Department of the FAO, and the University of Arkansas at Pine Bluff. Incorporating the lessons learned in “Fish to 2020”, a new fish module in the IMPACT model was developed and used to create the “Fish to 2030” report covering 2000–2030 (World Bank 2013). “Fish to 2030” also incorporated new developments in global seafood markets and the aquaculture sector. The model differentiates capture and aquaculture production of 16 fish commodities in 115 regions. Under the auspices of the Global Futures and Strategic Foresight project, IFPRI shared the IMPACT fish model with WorldFish in 2015. WorldFish made further significant updates to the model to reflect the latest FAO historical trends and to calibrate future trends to 2050, considering specific biophysical and socioeconomic factors for fisheries and aquaculture and fish management and production targets defined by national governments. Using the updated model, in 2017 WorldFish and IFPRI released “Fish to 2050” as an Association of Southeast Asian Nations (ASEAN) regional report (Chan et al. 2017), and two years later, also published a foresight special issue paper on fish in Africa in 2019 (Chan et al. 2019). Currently, the IMPACT model for fish is a stand-alone module in the process of integration into the latest version of the crop and livestock IMPACT model. Fish projections for major CGIAR regions Using the IMPACT model, we conducted a future projection to 2030 using the Shared Socioeconomic Pathway (SSP) 2 scenario, a middle-of-the-road scenario that follows historical trends. Under SSP 2, economic development persists but is not uniform, and environmental degradation continues but at a slowing pace. Generally, there is improvement in terms of the global environment and economy, but it is much slower than under SSP 1. Climate change presents moderate challenges to both adaptation and mitigation (World Bank). Table A1.1. Regional per-capita fish consumption in 2017 and 2030. Per-capita fish consumption (kg/person/year) Africa SSA WCA ESA CWANA Asia ASEAN SA 2017 9.9 8.8 11.7 5.7 9.8 24.1 39.5 8.3 2030 10.9 8.6 11.8 5.2 12.6 27.3 49.4 10.8 Under this scenario, global fish production will increase 24% by 42 million tonnes from 2017 to 2030. Compared to 2017, the per-capita fish consumption in Africa’s sub-regions, namely SSA, WCA, and ESA, are almost stagnant in 2030. Other regions in Asia such as ASEAN, SA, and CWANA show an increasing trend (Table A1.1). Overall, the total fish production and consumption will increase by 2030 compared to 2017 for all regions. GDP growth, meanwhile, ranges between 3.5% to 5.4% in the listed regions (Table A1.2). A negative supply-demand gap reflects a fish supply deficit, whereas positive values represent a fish supply surplus. The population in sub-Sahara Africa will rise from 1.1 billion in 2020 to 1.4 billion people in 2030 (United Nations 2020). By 2030, this region will face a significant shortfall in its domestic fish supply totalling 2.9 million tonnes, with almost stagnant per capita fish consumption relative to 2017 at 8.8 kg, less than half of the global average. Sub-Saharan Africa will require an additional 17.4 million tonnes of fish supply to reach the current global fish per-capita consumption of 21 kg and meet domestic demand by 2030. Major investments in fisheries and aquaculture could provide sufficient aquatic foods for future populations. 56 Georgina Smith / CIAT The ASEAN region exhibits a high annual per-capita fish consumption of 49 kg, more than double the global average. Overall, the ASEAN has become a net exporter and will remain so in the coming decades. Regional fish and fishery products will continue to be broadly traded, fuelled by higher fish consumption and globalization of the food system. There are, however, several principal constraints to the sustainable supply of fish: limited access to quality freshwater, competition in land use, the challenges of regional and national governance and regulation, environmental degradation, and climate change. By 2030, SA will experience a surplus of fish production due to lower fish consumption per capita. The number of vegetarians in India is estimated at around 30%, and animal source food consumption is growing, including the consumption of fish. In SA, per-capita fish consumption is as low as in Africa, about half the global average, but is expected to grow. The population in SA will surge from 1.9 billion in 2020 to 2.1 billion people in 2030. To attain the average global fish consumption of 21 kg and meet its demand by 2030, SA will require 21.8 million tonnes of fish. Table A1.2. Regional supply and demand gap by 2030 under the SSP 2 scenario. Supply and demand in 2030 (thousand tonnes) Africa SSA WCA ESA CWANA Asia ASEAN SA Total 2018 fish production 12,268 8,629 5,262 3,367 5,685 122,404 31,578 19,379 Total 2030 fish production 14,511 9,714 6,221 3,493 6,083 156,614 43,427 29,104 Total 2030 fish consumption 18,592 12,584 8,904 3,680 8,820 132,904 36,055 23,164 Supply-demand gap in 2030 -4,081 -2,870 -2,683 -187 -2,737 23,709 7,373 5,941 Projected Benefits of CGIAR Research 57 Appendix 2. Employment generation Different investment scenarios result in different levels of employment generation in 2025, 2035, and 2045 for a selected set of countries, presented in 1000s employed, derived from Frija et al. 2020. Country Year HIGH + NARS HIGH + RE HIGH MED RMM REF-NoCC 2025 712.33 2162.55 555.78 277.77 754.66 1213.20 Nigeria 2035 7767.99 8594.98 6132.33 3220.22 1404.63 4154.95 2045 18016.78 17799.58 14321.14 7606.15 1806.52 11389.92 2025 138.03 419.35 107.51 53.83 146.30 235.19 Ghana 2035 1472.82 1629.75 1163.04 610.37 266.25 787.56 2045 3264.93 3225.66 2595.58 1377.97 327.30 2063.57 2025 81.56 249.63 62.46 31.83 86.82 139.57 Niger 2035 956.79 1059.58 757.74 395.27 172.47 510.18 2045 2515.31 2485.77 2002.44 1058.64 251.59 1586.24 2025 43.30 132.51 33.16 16.89 46.09 74.09 Mali 2035 498.70 552.27 394.95 206.02 89.90 265.92 2045 1232.19 1217.72 980.95 518.60 123.25 777.06 2025 46.45 143.86 34.57 18.14 49.78 80.04 Senegal 2035 493.66 547.38 392.77 202.91 88.58 262.03 2045 1110.10 1097.58 885.75 465.13 110.64 697.55 2025 397.29 1146.03 338.96 154.82 637.06 700.82 Ethiopia 2035 4208.10 4649.32 3303.60 1805.01 1369.87 2431.89 2045 10527.84 10428.48 8408.78 4633.62 2139.95 7103.54 2025 283.47 815.94 242.87 110.44 453.93 499.37 Kenya 2035 2858.37 3157.35 2242.57 1227.12 931.19 1653.11 2045 6644.34 6581.14 5305.72 2926.57 1351.35 4485.79 2025 28.56 86.24 22.13 11.18 47.15 51.87 Sudan 2035 300.40 333.47 238.92 126.54 96.26 170.90 2045 702.80 697.27 564.07 304.48 141.13 468.48 2025 46.52 137.11 38.00 18.17 75.62 83.19 Uganda 2035 550.81 609.94 435.10 234.26 177.98 315.97 2045 1467.36 1454.60 1174.71 641.05 296.56 984.44 2025 161.04 467.13 135.89 62.79 259.14 285.07 Tanzania 2035 1824.05 2016.49 1434.32 780.66 592.64 1052.09 2045 4548.45 4506.41 3635.12 1998.04 923.16 3064.44 2025 21.73 63.39 17.68 8.61 34.61 26.90 Tunisia 2035 189.84 211.89 147.27 82.31 68.55 79.20 2045 351.63 351.14 278.26 158.39 89.24 178.01 2025 76.43 218.80 63.66 30.38 121.11 94.14 Morocco 2035 675.60 751.96 516.77 296.47 246.94 285.31 2045 1336.04 1332.19 1046.61 608.31 342.57 683.31 2025 35.15 113.60 24.54 13.66 57.58 44.76 Algeria 2035 286.25 324.23 238.37 116.29 96.77 111.81 2045 457.82 460.79 381.65 194.43 109.84 219.10 2025 172.66 500.17 141.67 68.49 274.44 213.34 Egypt 2035 1580.14 1761.90 1219.61 688.14 573.12 662.17 2045 3198.29 3192.12 2521.69 1446.30 814.73 1625.11 58 Appendix 3. Additional information regarding greenhouse gas emissions and adaptation The results presented below span emissions from animals, crops, and land use change (Figures A3.1 a, b, c, and Tables A3.1 a, b, c). For the CO2eq emissions in Section 6.4, we used CH4 rice, CH4 enteric, and N2O fertilizer emissions to compute crop and animal GHG emissions in Gt CO2eq for each CGIAR region as follows: (0.416 * N2O fertilizer emissions) + (0.028 * CH4 rice emissions) + (0.028 * CH4 enteric emissions). The results from Rosegrant et al. (2017) were disaggregated from the FAO regions then re-aggregated using the One CGIAR regions and supporting data from IFPRI and the FAO. Additional details are available on request. Figure A3.1. CH4 ruminant emissions, CH4 rice emissions, and N2O fertilizer emissions under the COMP and reference scenarios. a. CH4 Ruminant Emissions 2030 b. CH4 Rice Emissions 2030 15 14 12 10 10 8 6 5 4 2 0 0 CWANA ESA LAC SA SEA WCA CWANA ESA LAC SA SEA WCA CGIAR Regions CGIAR Regions COMP REF COMP REF c. N2O Fertilizer Emissions 2030 0.5 0.4 0.3 0.2 0.1 0.0 CWANA ESA LAC SA SEA WCA Projected Benefits of CGIAR Research 59 Tg CH4 Tg N2O-N Tg CH4 Tables A3.1. Enteric emissions , rice emissions, and N2O emissions. a. Enteric emissions: Differences between the COMP and reference scenarios Region Reference COMP Difference % (Tg CH4) difference CWANA 11.9914 11.5716 -0.4198 -3.5 ESA 5.0973 4.8155 -0.2819 -5.5 LAC 15.3511 14.3698 -0.9812 -6.4 SA 9.8524 9.6207 -0.2317 -2.4 SEA 13.1378 12.5816 -0.5563 -4.2 WCA 5.1676 4.8871 -0.2805 -5.4 b. Rice emissions: Differences between the COMP and reference scenarios Region Reference COMP Difference % (Tg CH4) difference CWANA 0.4741 0.4612 -0.0129 -2.7 ESA 0.4501 0.6303 0.1802 40.0 LAC 0.9589 1.0176 0.0587 6.1 SA 9.0489 9.5002 0.4513 5.0 SEA 12.4151 12.9381 0.5230 4.2 WCA 0.3013 0.4191 0.1178 39.1 c. N2O emissions: Differences between the COMP and reference scenarios Region Reference COMP Difference % (Tg N2O ) difference CWANA 0.1146 0.1424 0.0278 24.3 ESA 0.0128 0.0174 0.0045 35 LAC 0.0848 0.0995 0.0148 17.5 SA 0.2306 0.301 0.0705 30.6 SEA 0.3536 0.4051 0.0516 14.6 WCA 0.0069 0.0093 0.0024 35 60 Bioversity International and the International Center for Americas Hub https://alliancebioversityciat.org Tropical Agriculture (CIAT) is part of CGIAR, a global research www.cgiar.org partnership for a food-secure future. Km 17 Recta Cali-Palmira. C.P. 763537A.A. 6713 Bioversity International is the operating name of the Cali, Colombia International Plant Genetic Resources Institute (IPGRI). Telephone: (+57) 602 4450000