TY - THES T1 - The role of social protection and agriculture for improved nutrition in Ethiopia A1 - Bahru,Bezawit Adugna Y1 - 2022/06/22 N2 - Nearly all nations embarked on a commitment to free the world from hunger and poverty by 2030 as part of the Sustainable Development Goals (SDGs). Considerable progress has been made in reducing poverty and hunger. Yet, a significant proportion of people in the global south, especially those living in rural areas, still live in poverty, go hungry, and suffer from malnutrition. Social protection and nutritionsensitive agriculture are among the strategies implemented by several countries to alleviate poverty, food insecurity, and malnutrition. While consensus exists on the impact of safety nets on poverty, evidence on their impact on agriculture, food security, and nutrition is not definitive. Moreover, the impact of nutrition-sensitive agricultural interventions, particularly the role of production diversification, in improving smallholders’ diet is not thoroughly examined. Ethiopia presents an interesting case study to investigate the role of social protection and nutrition-sensitive agriculture on household welfare. Ethiopia has one the largest social protection program in Africa called the Productive Safety Net Program (PSNP). More recently, Ethiopia has also made several efforts to make agriculture and social protection interventions nutrition-sensitive. Moreover, the country is not only one of the fastest-growing economies in the world, but has high rates of poverty, food insecurity, and malnutrition. Thus, using Ethiopia as a case study, this dissertation provides a rigorous assessment of the impact of PSNP on agricultural, food security and nutrition outcomes, and the impact of agricultural diversification on household and child nutrition. To this end, it uses two longitudinal data sets from Ethiopia—the Living Standard Measurement Study - Integrated Surveys on Agriculture and the Young Lives cohort study. This dissertation is organized into six chapters. Chapter 1 presents the conceptual framework, highlighting linkages between social protection, agriculture, and nutrition. Chapter 2 provides information about the data sources and the identification strategies used in the following three chapters. Chapter 3 evaluates the impact of a social protection intervention on a range of agricultural outcomes. Chapter 4 analyzes the role of agricultural diversification on household and child nutrition. Chapter 5 estimates the impact of social protection on household food security and child nutrition. Chapter 6 concludes the dissertation by provides policy and methodological implications, and outlines recommendations for future research. Since 2005, Ethiopia has shifted its social protection strategy from relief types of interventions of mainly an ad hoc distribution of food/cash following droughts into a development-oriented program called the Productive Safety Net Program (PSNP). The PSNP provides cash and in-kind transfers for labor contributions in public work projects such as roads, irrigation, schools, hospitals, and training centers for farmers. The program is complemented by asset building programs that transfer assets with the aim to improve livelihoods. Hence, household participation in the PSNP could improve agricultural outcomes by alleviating liquidity and credit constraints, improving certainty, acting as insurance against risks, improving access to inputs and agricultural markets, and building community agricultural assets. Chapter 3 evaluates the impact of PSNP on productive asset ownership, improved seed and fertilizer adoption, crop and livestock diversity, advisory services, and women’s control over resources by applying targeted maximum likelihood estimation. Doing so it contributes to the literature by considering a range of outcomes along the causal pathway and employing a novel method that uses machine learning algorithms and hence provides estimates that are less prone to model misspecification and outliers. Results show that PSNP participation increased ownership of agricultural tools, the value of livestock sales, the share of household income from non-farm sources, time spent on agricultural work, and access to credit at the household level. Moreover, PSNP participation improves community access to irrigation water and advisory services on natural resource management, as well as household credit access, crop production, and livestock production. However, PSNP participation has no impact on technology adoption, women’s control over income, crop and livestock counts, and access to extension services at the household level. Results also show that PSNP participants have a lower endowment of durable assets, human capital, and land, which might hinder improved community access to inputs and advisory services to improve agricultural outcomes. These results underscore the need to integrate household-level interventions that could lift household endowments to create an asset threshold that would allow the productive use of created community assets. One avenue to achieve this could be complementing cash/in-kind transfers with a well-designed productive asset transfer. Other studies have shown delays and underpayment of entitled public work transfers. Hence improving the timeliness and size of cash/in-kind payments are also critical to achieve impact on agriculture. Doing so may elevate the impact of the PSNP beyond improving community access to inputs and may promote agricultural development, ultimately improving rural livelihoods. Agriculture as a source of food and livelihood for the hungry and malnourished has an important role in alleviating malnutrition and its associated welfare consequences. Especially in a context where subsistence agriculture constitutes a large share of agricultural production and market participation is hindered by market failure, own production plays a crucial role in household food security and nutritional outcomes. Evidence as to whether production diversity improves nutritional outcomes is not definitive. It largely draws on studies that did not address the endogenous relationship between production and consumption decisions. To fill this research gap, Chapter 4 analyzes the impact of production diversity on household diet and child undernutrition. It uses instrumental variable approach to account for endogeniety in production and consumption decision among smallholders. Results show that production diversity is associated with improvements in household diet, but not child chronic undernutrition. Positive effects could come through three possible pathways: consumption, improved agroecology due to production diversification, and improved risk-bearing capacity. We find higher consumption of more nutritious foods (egg, fish, fruits, legumes/nuts/seeds, meat, roots, tubers, milk, milk products, and vegetables) among households with higher levels of production diversity. Nevertheless, production diversification above 7.5 and 7 food groups, for household with access to community market and those with no access respectively, is negatively associated with household nutrition. This could be due to foregone income benefits from specialization. As far as the role of markets is concerned, we found that market access improves household dietary diversity even when production diversity is high. Our analysis of the role of market participation shows a positive impact of market participation on household diets, but finds no impact on child nutrition. Given that an average farmer in our sample produces about six groups of foods, promoting more production diversity is likely to result in negative effect on household diets. Hence, policies that aim to improve smallholders’ nutrition should focus on improving conditions for market participation over increasing the diversity of production. While there is ample evidence on the impact of safety nets on poverty, their role in addressing the root causes and manifestations of poverty, such as child malnutrition, is not well documented. Moreover, the available evidence on the impact of safety nets on household food security and child nutrition emanates from programs in Latin American countries that were implemented under different institutional capacity and implementation modalities. Moreover, evidence as to whether safety nets impact child nutrition is not conclusive. In addition, in Ethiopia, the very few studies examining the impact of safety nets on child nutrition and household food security provided mixed evidence. These studies used methods that are prone to bias due to time-variant confounders that are pertinent to the program design and implementation. In Chapter 5, we address this gap by evaluating the impact of the PSNP on household food security and child nutrition outcomes. We did this by applying marginal structural models that adjust not only for timein-variant, but also time-variant confounders. We find that although PSNP participation increases child meal frequency, it has no impact on household food security and child dietary diversity, height-for-age z-scores, body-mass-index, the likelihood of stunting, and the likelihood of being underweight. We also find important predictors of child nutrition, such as maternal education, child dietary diversity, household food security, durable assets, expenditures, and nutritional status during the 1,000 days window to be lower among children in PSNP participant households. Moreover, while experts note that integrating safety nets with other sectoral programs is critical to result in impacts on nutrition, such integrations are lacking in the PSNP’s implementation. Therefore, we recommend integrating PSNP with other sectoral programs that are nutrition-specific and nutrition-sensitive. Some proven interventions include, but are not limited to, the promotion of access to clean water and sanitation, access to health services, women’s empowerment, nutrition education, and agricultural technology adoption. To summarize, using Ethiopia as a case study, this dissertation contributes to the growing body of literature on impact evaluation of programs aimed at achieving the SDG targets. It does so by applying novel techniques in impact evaluation using observational studies including those that integrate learning algorithms. It shows that a higher production diversity negatively affects household nutrition while market access and participation improve household diets. Hence, improving conditions for market participation and increased participation in markets might be one of the pathways through which agriculture improves nutrition, especially in contexts where production is already diversified enough. It also shows that despite the potential of safety nets to address the root cause of poverty by improving household livelihoods and breaking the inter-generational cycle of poverty, these impacts are mostly nonexistent in one of Africa’s largest social protection program. To alleviate these constraints and elevate the contribution of PSNP beyond the limited impact on nutrition and agriculture, we recommend integrating nutrition-sensitive interventions and well-designed asset transfer programs. KW - Landwirtschaft KW - Ernährung KW - Ernährung KW - Zusammenhänge Landwirtschaft-Ernährung CY - Hohenheim PB - Kommunikations-, Informations- und Medienzentrum der Universität Hohenheim AD - Garbenstr. 15, 70593 Stuttgart UR - http://opus.uni-hohenheim.de/volltexte/2022/2028 ER -