TRANSFORMATION OF FOOD SYSTEMS: HOW CAN IT BE FINANCED?

Highlights

● The analysis of financing for the transformation of food systems requires first a consideration of costs.

● Costs are then compared to six main financial flows, two internal and four external to food systems.

● There are enough potential financial resources available to fund the transformation of food systems.

● To mobilize the existing and potential funds adequate macroeconomic and overall incentive frameworks are needed, plus a variety of specific interventions discussed in the paper.

● Developing countries need to establish national structures to design, finance and coordinate comprehensive national programs for their food systems.

Abstract

The paper provides a review of the work on financing the transformation of food systems done by the author and colleagues at IFPRI. The analysis discusses the objectives of that transformation (related to the Sustainable Development Goals and the Paris Agreement) and shows estimates of the costs involved. Then it presents an evaluation of the existing financial flows and their potential scaling up, using a broader view of six main flows involved: two that are internal to food systems (consumer expenditures on food and related items, which are the main source of revenue (as sales) for the variety of actors on the supply side of food systems); and four that are external (international development funds, public budgets, banking systems, and capital markets). The paper notes that although current funding does not reach the scale needed to finance the desired transformation of food systems, there are sufficient potential financial resources available in the aggregate to achieve such transformation. However, to mobilize the existing potential funds the paper discussed several things that need to be done, starting with an adequate macroeconomic and overall incentive framework to guide both the internal flows related to consumption and production decisions, as well as the four external ones. The paper also suggested other possible interventions to mobilize, reorient, and increase the financial flows to the desired objectives. It further noted that the costs and financing, which in the paper were discussed at the global level, must be estimated at the country level, as part of the design and implementation of adequate national plans for equitable, healthy and sustainable food systems.

Graphical abstract

Keywords

financial markets / food systems / development financing / consumer expenditures / public budgets / Sustainable Development Goals

Only when all contribute their firewood can they build up a big fire (众人拾柴火焰高).

Horticultural crops are a major source of high value nutritious food, and new improved cultivars developed through breeding are required for sustainable production in the face of abiotic and biotic stresses, and to deliver novel, premium products to consumers. However, grower confidence in the performance of new germplasm, particularly across environmental variability, is important for commercial adoption and germplasm-environment matching to optimize production.

Knowledge on the stability of germplasm across environments for cultivar selection and breeding in horticultural crops is limited. Normally, evaluation of replicated (usually clonally) germplasm over several locations is done to generate information on germplasm-environment matching, however, these experiments are expensive in horticultural tree crops due to the large size and longevity of the experimental unit, the need for replication within and across sites, and high costs of phenotyping. In the absence of information for genotype-environment matching, breeding tends to focus on developing domestic cultivars for local environments using small experiments with limited or no overlap of germplasm between programs. In addition, confidence in the performance of imported germplasm in exotic regions is generally developed by assuming germplasm will be stable across environments, or from ad hoc knowledge accumulated by local risk-taking early adopters.

Genomic prediction might provide an efficient approach for combining historical data across breeding programs and selection trials to increase confidence in the performance of domestic germplasm in local environments, and imported germplasm in exotic environments. Genomic prediction exploits linkage between markers and QTLs in a training population (genotyped using genome-wide marker and phenotyped for traits of interest) to develop a prediction model that can then be applied to genotypic data for a selection population to predict genetic performance in the absence of direct phenotypic data^[1]. Usually, with the number of markers being greater than the number of germplasm entries (i.e., number of markers>>number of observations), a distribution of marker effects is assumed to reduce the number of parameters requiring estimation. A common distributional assumption is Gaussian (i.e., all loci have small effect). This model is equivalent to the standard definition of a quantitative trait, and the genomic realized relationship matrix (GRRM) estimated from the marker similarities is equivalent to direct modeling of individual marker effects on phenotype using this distribution^[2]. One advantage of the GRRM is that it can be directly incorporated into established linear mixed model approaches that are more flexible than the Bayesian approaches required to model other marker effect distributions. Given that the GRRM models track replication of individual alleles across germplasm, rather than replication of the whole genomes which occurs when individuals are vegetative propagated, the models can be used to connect unreplicated germplasm within and across locations to improve prediction accuracy in local and exotic environments. Therefore, here we define the term global genomic prediction to describe the hypothesis that “historical phenotypic data from multiple breeding programs are a sample of the experienced environment and their data sets can be connected through genome-wide multivariate prediction models”^[3].

Several studies have evaluated global genomic prediction in horticultural tree crops. The first study was undertaken using data on crispness of apples evaluated across 662 (unreplicated) entries across three sites in the USA (Wenatchee, WA; St Paul, MN; and Geneva, NY) that had been genotyped using an 8K SNP array^[4]. A GRRM was estimated from the SNP data, with each site-by-trait-by-age considered as a unique attribute (sensu Falconer^[5]), and a factor analytic model was used to estimate the entry-by-attribute covariance matrix. This multivariate approach allowed the combining of data collected by different methods. The analysis suggested genomic effects for texture were highly correlated across the Geneva and Wenatchee sites, but were less well correlated with effects at St Paul. A subsequent study in sweet cherry ^[6] with 597 entries phenotyped for timing of fruit maturity across four sites (Prosser, WA, USA; Balandran and Bourran, France; and Forli, Italy) (and two years at each site) using a similar statistical model estimated high additive genomic correlation (>0.9) among environments and high prediction accuracy (0.9 to 1.0). In recent unpublished research on peach with 506 entries phenotyped for sweetness across ten environments (four sites and multiple years within sites), we observed an increase of about 20% in prediction accuracy by combining data into a single analysis compared to only using data from a single location to train the prediction model.

Several challenges to implementing global genomic prediction have been encountered. Considerable effort is required to collate, standardize, and curate datasets contributed from different sources. Names of genetically identical entries need to be standardized, particularly if language differences exist. The identity of SNP loci and alleles across multiple datasets needs to be well defined, particularly if SNP genotypes are to be imputed across different genotyping platforms to increase marker density. Correct description of complex sampling designs in different environments is required (e.g., Hardner et al.^[7]) to reduce non-genetic phenotypic noise. Often collinearity is observed in the GRRM, and therefore bending (i.e., replacing negative Eigen values with slightly positive values) is required to obtain model solutions. Confounding of population structure with testing location might also be an issue for some datasets, with further research is required to develop models that account for heterogeneity in allele frequencies across subpopulations.

Several challenges also exist for translation of global genomic prediction to genetic improvement in practice. While genome-assisted parental selection (GAPS) using additive effects is a credible alternative to phenotypic selections^[8], performance of GAPS across different sites/environments remains to be investigated. For prediction of clonal values (i.e., predicted phenotype based on total genetic variation captured by vegetative propagation of superior individuals – which is the important for identifying elite cultivars), relationship matrices of non-additive genetic effects are generally more sparse. Thus, the genetic architecture of non-additive effects is likely to be estimated with less precision and might be confounded with additive genetic effects, meaning the accuracy of predicted clonal values is expected to be lower than for breeding values. While accuracy of breeding values was not greatly improved by inclusion of non-additive effects, bias was reduced^[9]. Global genomic prediction supports prediction of performance of entries into environments in which they have not been tested through the use of correlated genetic effects, however, interpretation of these predictions needs qualification for traits that require expression of other traits (e.g., expression of fruit texture requires trees to flower and set fruit). Lastly, the intellectual property of contributors of data and germplasm to combined datasets needs to be respected and protected.

Global genomic prediction fundamentally relies on collaboration to leverage, and add to, the latent value of existing data of individual breeding and selection programs. A preliminary online portal has been developed with the Genome Database for Rosaceae to predict peach sweetness at four sites in the USA (Fresno, CA; College Station, TX; Clarksville, AR; and Seneca, SC) using genome-wide genotypic data uploaded by any user into the unpublished model used for the analysis of peach described above. Although crops from the family Rosaceae have been the focus of initial development, this approach can be readily applied to other crops. The approach is being extended into other traits and larger datasets, improving understanding of the drivers of genotype-by-environment interaction and enabling practical tools in genetic improvement for responding to climate change.

References

[1] Díaz-Bonilla E. Financing SGD2 and ending hunger. Center for Development Research (ZEF) in cooperation with the Scientific Group for the UN Food System Summit 2021, 2021

[2] Díaz-Bonilla E, Swinnen J, Vos R. Financing the transformation to healthy, sustainable, and equitable food systems. In: Global Food Policy Report 2021: Transforming Food Systems after COVID-19. Washington DC: International Food Policy Research Institute (IFPRI), 2021, 21–23

[3] Díaz-Bonilla E, McNamara B, Njuki J, Swinnen J, Vos R. The UN food systems summit 2021: lessons of the gender and finance levers. In: Agricultual & Applied Economics Association (AAEA) Annual Meeting. Anaheim: AAEA, 2022

[4] Diaz-Bonilla E, Echeverria R G. Climate finance: funding sustainable food systems transformation. In: 2022 Global Food Policy Report: Climate Change and Food Systems. Washington DC: International Food Policy Research Institute (IFPRI), 2022, 48–57

[5] Díaz-Bonilla E, Echeverria R, Vos R. Global action for climate finance and investments for agrifood system transformation. In: Food Security and Sustainable Agriculture Task Force Brief. Washington DC: International Food Policy Research Institute, 2022

[6] Food and Agriculture Organization of the United Nations (FAO), International Fund for Agricultural Development (IFAD), UNICEF, World Food Programme (WFP), World Health Organization (WHO). The state of food security and nutrition in the world 2022. Repurposing food and agricultural policies to make healthy diets more affordable. Rome: FAO, 2022

[7] Crippa M, Solazzo E, Guizzardi D, Monforti-Ferrario F, Tubiello F N, Leip A . Food systems are responsible for a third of global anthropogenic GHG emissions. Nature Food, 2021, 2(3): 198–209

[8] Díaz-Bonilla E, Callaway V. The G20, global trade and investment regimes, and a sustainable food future. In: Bacchus J, Callaway V, Ciuriak D, Díaz-Bonilla E, Hepburn J, van Rensburg S J J, Campos S L, Nakatomi M, Narlikar A, Biasutti C N, Panzini F S, Parry A, Torres H R, Viviers W, Wang H, Wang W, eds. How the G20 Can Help Sustainably Reshape the Global Trade System: A Compilation of Analysis. Geneva, Switzerland: International Centre for Trade and Sustainable Development (ICTSD), 2018, 83–92

[9] Fan S, Díaz-Bonilla E, Cho E E, Rue C. SDG 2.1 and SDG 2.2: why open, transparent, and equitable trade is essential to ending hunger and malnutrition sustainably? In: Hepburn J, Bellmann C, eds. Achieving Sustainable Development Goal 2: Which Policies for Trade and Markets? Geneva, Switzerland: International Centre for Trade and Sustainable Development (ICTSD), 2018, 17–28

[10] Valin H, Thomas H, Bodirsky B L, Tomoko H, Elke S. Achieving zero hunger by 2030: a review of quantitative assessments of synergies and trade-offs among the UN Sustainable Development Goals. In: United Nations Food Systems Submit 2021. Bonn: Center for Development Research (ZEF) in Cooperation with the Scientific Group for the UN Food System Summit 2021, 2021

[11] von Braun J, Chichaibelu B B, Torero M, Laborde Debucquet D, Smaller C. Ending hunger by 2030—Policy actions and costs. Bonn: Center for Development Research (ZEF), 2020

[12] Center for Development Research of the University of Bonn (ZEF) and Food and Agriculture Organization of the United Nations (FAO). Investment Costs and Policy Action Opportunities for Reaching a World without Hunger (SDG 2). Bonn and Rome, 2020

[13] Laborde D, Parent M, Smaller C. Ending hunger, increasing incomes, and protecting the climate: what would it cost donors? In: Ceres2030: Sustainable Solutions to End Hunger. Washington DC: International Food Policy Research Institute; Winnipeg, Canada: International Institute for Sustainable Development; Ithaca, New York: Cornell University, 2020

[14] Rosegrant M W, Sulser T B, Cenacchi N, Wiebe K, Willenbocke D. Estimating the global investment gap in research and innovation for sustainable agriculture intensification in the Global South. Colombo, Sri Lanka: International Food Policy Research Institute (IFPRI) and Commission on Sustainable Agriculture Intensification (CoSAI), 2021

[15] Sulser T, Wiebe K D, Dunston S, Cenacchi N, Nin-Pratt A, Mason-D’Croz D, Robertson R D, Willenbockel D, Rosegrant M W. Climate change and hunger: estimating costs of adaptation in the agrifood system. Washington DC: International Food Policy Research Institute (IFPRI), 2021

[16] The Food and Land Use Coalition (FOLU). Growing better: ten critical transitions to transform food and land use. FOLU, 2019

[17] Laborde D, Mamun A, Martin W, Piñeiro V, Vos R. Modeling the impacts of agricultural support policies on emissions from agriculture. IFPRI Discussion Paper, No. 1954. Washington DC: International Food Policy Research Institute (IFPRI), 2020

[18] United Nations Environment Programme (UNEP). Adaptation gap report 2021: the gathering storm—Adapting to climate change in a post-pandemic world. UNEP, 2021

[19] International Energy Agency (IEA). Net zero by 2050. A Roadmap for the Global Energy Sector. IEA, 2021

[20] United Nations Framwork Convention on Climate Change (UNFCCC). UNFCCC standing committee on finance. Fourth (2020) biennial assessment and overview of climate finance flows. UNFCCC, 2021

[21] Macquarie R, Naran B, Rosane P, Solomon M, Wetherbee C. Updated view of the global landscape of climate finance 2019. London: Climate Policy Initiative (CPI), 2020

[22] Watson C, Schalatek L. The global climate finance architecture. Climate Finance Fundamentals. No 2. Washington DC: Climate Funds Update, Heinrich Böll Stiftung; London, UK: Overseas Development Institute, 2021

[23] Díaz-Bonilla E. Macroeconomics, agriculture, and food security: a guide to policy analysis in developing countries. Washington DC: International Food Policy Research Institute (IFPRI), 2015

[24] Parry I W H, Black S, Vernon N. Still not getting energy prices right: a global and country update of fossil fuel subsidies. International Monetary Fund (IMF), 2021

[25] SIFMA. Capital markets fact book, 2022. SIFMA, 2022

[26] Hand D, Dithrich H, Sunderji S, Nova N. 2020 Annual impact investor survey. Global Impact Investing Network (GIIN), 2020

[27] Díaz-Bonilla E. Clarifying some of the confusion about climate finance in COP26. Washington DC: International Food Policy Research Institute (IFPRI), 2021

[28] Díaz-Bonilla E. Using the new IMF special drawing rights for larger purposes: guaranteeing ‘pandemic recovery bonds’. Washington DC: International Food Policy Research Institute (IFPRI), 2021

[29] von Braun J, Díaz-Bonilla E. Letter: perpetual bonds can help states fight hunger. Financial Times (FT), 2021

[30] Fan S. Public expenditures, growth, and poverty in developing countries: lessons from developing countries. Washington DC: International Food Policy Research Institute (IFPRI), 2008

[31] Asia D. Funding agricultural innovation for the global south: does it promote sustainable agricultural intensification? Colombo, Sri Lanka: Commission on Sustainable Agriculture Intensification, 2021

[32] von Braun J, Afsana K, Fresco L O, Hassan M . Food systems: seven priorities to end hunger and protect the planet. Nature, 2021, 597(7874): 28–30

[33] Andrews Colin, de Montesquiou A, Sánchez I A, Dutta P V, Paul B V, Samaranayake S, Heisey J, Clay T, Chaudhary S. The state of economic inclusion report 2021: the potential to scale. Washington DC: World Bank, 2021

[34] Díaz-Bonilla E, Fernández-Arias E, Piñeiro V, Prato B, Arias J. Forum for analysis of financing and financial inclusion for agricultural, rural and food systems development in Latin America and the Caribbean: summary document. LAC Working Paper 5. Washington DC: International Food Policy Research Institute (IFPRI), 2019 (in Spanish)

[35] Díaz-Bonilla E, Loboguerrero A M, Verchot L, Viglizzo E, Mirzabaev A. Financing a sustainable food future. T20 Argentina: Food Security and Sustainable Development Task Force Brief. Washington DC: International Food Policy Research Institute (IFPRI), 2018

[36] Apampa A, Clubb C, Cosgrove B E, Gambarelli G, Loth H, Newman R, Rodriguez Osuna V, Oudelaar J, Tasse A. Scaling up critical finance for sustainable food systems through blended finance. Discussion Paper. CGIAR Research Program on Climate Change, Agriculture and Food Security (CCAFS), 2021

Compliance with ethics guidelines

Eugenio Díaz-Bonilla declares that he has no conflicts of interest or financial conflicts to disclose. This article does not contain any studies with human or animal subjects performed by the author.

RIGHTS & PERMISSIONS

The Author(s) 2023. Published by Higher Education Press. This is an open access article under the CC BY license (http://creativecommons.org/licenses/by/4.0)