Essays in agricultural economics

This thesis explores topics in Agricultural Economics and is composed of five papers. In the first paper (Chapter 2), a latent-class stochastic frontier model is used to estimate efficiency scores of farmers in Ethiopia. Compared to conventional models, which assume a unique frontier, much lower inefficiencies are found, suggesting that part of the inefficiencies uncovered in the literature could be an artefact of the methods used. The second paper (Chapter 3) revisits the link between cereal diversity and productivity using a panel dataset in Ethiopia. The results suggest that the positive effect between cereal diversity and productivity becomes much smaller when households who produce teff (a low-productivity and high-value crop) are excluded from the sample, hinting at the possibility that results could be driven by yield differentials between cereals, rather than diversity. The third paper (Chapter 4) estimates the labour impacts of the adoption of Soil and Water Conservation technologies (SWC) in Ethiopia. The results suggest that adopting SWC technologies leads to an increase in adult and child labour. Understanding the labour impacts is important in itself, but it also raises concerns about using impact evaluation methods that require no change in inputs as an identifying assumption of impacts. Paper 4 (Chapter 5), assesses the pertinence of a drought index that has recently been proposed in the literature by Yu and Babcock (2010) and argues that it defines drought too narrowly. An extension to this index is proposed and we show, using a dataset of Indian districts, that the original index is likely to underestimate the impacts of drought. In Paper 5 (Chapter 6), we identify data-driven ranges of rainfall for which the marginal effects of a rainfall-temperature index (RTI) are different and then we discuss how the impacts of drought have changed over the 1966-2009 period in India. Finally, Chapter 7 concludes.