This study proposes the use of partial least squares to determine the key parameters of the perpetual inventory method model of capital stock as a new approach to calculate research and development (R&D) knowledge stocks and R&D elasticities. This approach avoids most of the major problems encountered in the literature that lead to obtaining very high and implausible rates of return to agricultural R&D...Using this approach, we obtain an average R&D elasticity for low- and middle-income (LM) countries of 0.23 and an average rate of return to R&D investment of 6.0 percent, bigger than the average discount rate of 4.2 percent for these countries. Results show that 60 percent of LM countries in our sample are underinvesting in agricultural R&D, as they can get higher returns by investing in this activity than in activities that return the social discount rate.

Analyze alternative national and international strategies and policies for meeting foof needs of the developing world on a sustainable basis, with particular emphasis on low-income countries and on the poorer groups in those countries.

This analysis explores the relationship between agricultural R&D investments and rural poverty reduction, and the prevalence of undernourishment in sub-Saharan Africa (SSA). It uses a panel data set of internationally comparable poverty dis-aggregated by urban and rural areas, country level undernourishment, and ASTI data on R&D investments and derived indicators. The study uses agricultural R&D knowledge stocks (KS) to account for the lagged effects of research through depreciation and gestation period of investments, and applies causal mediation analysis to assess the impact of KS on poverty and hunger and measure the relative contribution of KS-induced agricultural productivity growth on those outcomes. Evidence suggests that, while SSA growth in KS has been relatively slow, it helped reduce rural poverty and undernourishment – the percentage point reduction in rural extreme and moderate poverty of a 1% annual increase in KS is 0.218 and 0.146 percentage points per year, respectively. Mediation analysis indicates that a fifth of the KS effect on extreme rural poverty, and a quarter of the KS effect on moderate rural poverty, can be attributed to KS driven gains in agricultural labor productivity. Likewise, KS growth reduces undernourishment – a 1% annual increase in KS leads to a drop of 0.132 percentage points per year in the prevalence of undernourishment, with about 40% of that effect mediated through gains in agricultural land productivity. These results indicate that KS supports poverty and hunger reduction through benefits on-farm and beyond it. They also suggest that there is room for strengthening the role of R&D KS productivity enhancing innovations. Given the current low levels of investments in R&D and resulting KS, increasing its levels will be critical, but that alone is not sufficient. Policy makers will have to rethink the way the innovations from R&D get scaled up and pay attention to the necessary complementary policies and investments that enable a sustainable pathway leading to greater productivity growth and development impacts.

This paper looks at the prioritization of agricultural value chains (VCs) for the allocation of R&D resources that maximize development outcomes (poverty, growth, jobs, and diets). Considering that growth in VCs affects those various outcomes differently, as expansion pathways result in the diverse use of production factors and inputs, trade-offs from linkages across sectors, and changes throughout the agri-food system, this analysis uses (i) the RIAPA dynamic computable general equilibrium model to identify which agricultural VCs, when expanded through TFP growth, provide the strongest effects on the development outcomes of interest; (ii) the perpetual inventory model (PIM) to represent the lagged effect of research through knowledge stocks of agricultural R&D investments; and (iii) information on the elasticities of VC agricultural activity TFP with respect to agricultural R&D knowledge stocks, to discuss the VC priority allocations of R&D resources in Senegal. Results indicate that no one VC (crop- or livestock-related) is the most effective at improving all development outcomes. When accounting for policy preferences that attribute relative priority weight to development objectives, results (based on a ranking scale) indicate that R&D investments for maximizing development objectives can be most effective in Senegal’s VCs for traditional export crops (growth, diets, jobs, and to some extent poverty), groundnuts (poverty, diets, and jobs), rice (poverty and jobs), poultry/eggs (diets and jobs), sorghum/millet (poverty and growth), and cattle (diets and growth). Other promising VCs with potential effects at scale if strategically targeted include vegetables (poverty, diets, and jobs), oilseeds (poverty and growth), and fruits (diets and jobs). While these results can inform strategies aimed at improving multiple development outcomes, future modeling needs to focus on deepening the standardization and integration of R&D investments costs into the framework, disentangle the relevance of different types of R&D investments sources, and bring together other factors and complementary agrifood system investment dimensions relevant to sustainable and inclusive agricultural VC growth.

The 2019 report of the Global Commission on Adaptation for accelerated action to adapt to climate change included a call for increased allocation of resources to international agricultural research. The production and adaptation challenges faced by agriculture will be most acutely felt in Africa and South Asia, focus regions of the CGIAR, the world’s largest public food systems research network. These challenges come at a time when the CGIAR is undergoing a transformation of its partnerships, knowledge, assets and global presence, emerging as One CGIAR, aimed at sharpening its mission and impact focus to 2030 and beyond, in line with the Sustainable Development Goals. Evidence on the impacts of CGIAR research since the 1980s have consistently found high rates of return to investment. How could this evidence on the performance of the CGIAR and its partnership with NARS in developing regions be used to inform investment priority setting and to achieve the One CGIAR goals in the coming years? We used detailed R&D investment data from the CGIAR, NARS (ASTI) and evidence from the literature on returns to CGIAR investment by crop and region to develop and calibrate a model of R&D investment that allows us to conduct priority-setting analysis of alternative CGIAR investment across research activities and regions. The model developed can be linked to global partial equilibrium and economy-wide forward-looking models to analyze the effect of different CGIAR investment options under alternative future scenarios. We checked the plausibility of the results obtained by the model calculating the Benefit-Cost ratio of historical CGIAR investments and found that each dollar invested by the CGIAR between 1971 and 2018 returned almost 10 dollars in output as the result of increased productivity, which is within the range of returns found by most recent meta-analyses impact of CGIAR investment. An application of the model to SSA shows that the best results for the CGIAR are obtained from investments in cassava and potato in Southern Africa; yams, sorghum, cassava and groundnuts in West Africa; cassava in East Africa and groundnuts and shoats in the Sahel.