Titolo della tesi: Essays on Agricultural Innovations
Adopting new technologies and agricultural practices can help smallholder farmers improving their livelihoods, increasing their income and other aspects of their well-being, such as nutrition. Despite this, adoption rates in developing countries are still low, especially in Africa. Factors affecting the adoption and the use of these practices are manifold and include elements such as access to information, risk aversion and sufficient liquidity. Among them, special attention in the recent debate has been paid to information and market access, which will be two core elements in this thesis. Indeed, I try to provide a comprehensive picture of technology adoption, analyzing its determinants and the specific consequences on household food security and nutrition for women. Innovation will be defined throughout the essays as something not necessarily new to the context under analysis but as an improved product (good or service) or process (OECD, 2005).
This work contributes to a deeper understanding of the debate on technology adoption in developing countries in several ways. Firstly, this thesis wants to contribute by pointing up benefits deriving from the use of technology using a gender perspective and evaluating the impact of agricultural practices on women's nutrition and household food security, improving evidence that is still scarce. Secondly, it contributes to the debate around the determinants of the adoption of agricultural practices with peculiar reference to access to information and to the market for small farmers. Another relevant contribution of this thesis is the use, along the three essays, of several approaches to analyze transmission channels through formal or informal mediation analysis. Mediation analysis is a statistical framework used to analyze the mechanism through which a variable of interest, the treatment, affects an outcome through one or more intermediate variables called mediators. This methodology is commonly used in sociology, psychology, and epidemiology, but despite the importance of knowing causal mechanisms of economic phenomena, few studies use this approach in economics. Indeed, many studies estimated the magnitude and the significance of an impact but cannot disentangle what are the causes of that impact, leaving the causal effect as what was called a “black box” in the mediation analysis literature. The analyses will be performed using both cross-sectional and panel data.This thesis is built by three independent, although related, empirical essays investigating both the determinants and the impact of agricultural technology and agricultural practices. 1 examines the causal link between agricultural technology use, food security and nutrition, using new panel data on smallholder oil palm growers in Ghana collected by the Agricultural Policy Research in Africa (APRA) consortium in 2017 and 2019. In this essay, I introduce a gender perspective, focusing on women and the possible mediating role of women's empowerment that could be an important mechanism in driving the results. Here, technology use will be defined firstly as the use of one practice among agrochemicals, irrigation, and intercropping, three relevant practices for oil palm cultivation in the country. Secondly, I narrowed the definition to the use of irrigation and agrochemicals only due to possible heterogeneity when including also intercropping. Using fixed effect models with interaction terms and a Heckman’s model, the core results are: i) for oil palm producers in south-western Ghana, the use of at least one agricultural practice among irrigation and agrochemicals is significantly linked with women’s dietary diversity; ii) women empowerment appears to be a positive factor for household food security, regardless of the technological use status and iii) women empowerment mediates the relationship between technology use and women’s dietary diversity.
Essay 2 shifts the focus to another country and a different type of technology. One of the main barriers to adoption is limited knowledge about the technology, a barrier that could be reduced by increasing and improving extension services. In this essay, I assess the role of extension services on the adoption of laser land leveling (LLL) among 604 households in the Indian state of Karnataka using cross-sectional data collected by the South Asia Regional (SAR) division of IFPRI. The empirical analysis includes propensity score matching and causal mediation analysis. The core results are: i) having visited at least once the extension center (or received a visit by its officials) increases the likelihood of using LLL; ii) after explaining the advantages of the technology and its cost, farmers develop a perception about the affordability of laser land leveling that mediates the treatment effects of the extension service on laser land leveling adoption.
Finally, essay 3 includes the role of market access in determining the use of innovation. This essay wants to shed light on the nexus between market access, the three main constraints to technology adoption detected in the literature (i.e., limited knowledge, farmers’ risk aversion and limited liquidity and access to credit), and the final adoption. Using the four waves of the LSMS-ISA for Nigeria, I firstly identify the local governmental areas (LGAs) that can be classified as hot spots and cold spots for the main crops grown in the country - cereals, cassava or tubers. To do so, I use the Getis and Ord statistic to detect how the geographical concentration of certain crops and agricultural commercialization are linked. Then,I employ two approaches for instrumental variables mediation analysis to account for non-random selection and possible simultaneity between market access and the use of agrochemicals. Results show a positive correlation between selling on the market and the outcome variable, and the causal effect is confirmed also by the instrumental variable mediation. The primary transmission channel identified seems to be the possibility to access to credit.