Essays on agricultural technology, resource allocation and the value of information

• Ano de defesa: 2020
• Autor(a) principal: Queiroz, Pedro Wesley Vertino de
• Tipo de documento: Tese
• Tipo de acesso: Acesso aberto
• Idioma: eng
• Instituição de defesa: Universidade Federal de Viçosa
• Palavras-chave em Português: Inovações agrícolas; Implementos agrícolas; Agricultura de precisão; Sistema de informação geográfica; Sistema de recuperação da informação - Agricultura; Economias Agraria e Recursos Naturais
• Link de acesso: https://locus.ufv.br//handle/123456789/28159

Understanding agricultural technology adoption and resource allocation is crucial to achieve agricultural development and to adapt to changes in current institutional, environmental, and climatic conditions. In addition, decision-making can be improved because of new information and timely data available from precision agriculture (PA) and satellite imagery technologies. The objective of this dissertation was to, first, study agricultural technology adoption and resource allocation due to changes in the relative supply of land (land-and-labor ratio) in South American agriculture. Second, the payoff of a precision agriculture (PA) technology was assessed for U.S. farms. Third, I used remote sensing information to obtain a better measurement of the effect of droughts in Brazilian municipalities. Chapter 1 studied traditional (e.g., land and labor) and commercial (e.g., machinery and fertilizers) inputs in South American agriculture. Acemoglu’s directed technical change framework was used to estimate the process of induced innovation in agriculture using deforestation patterns as source of exogenous variation for the agricultural land supply. The results indicated that deforestation was important to provide agricultural land in South America and because of a larger availability of land in intensive deforestation countries, more land-complementary inputs (machinery) were used relative to labor-complementary inputs (fertilizers). The induced innovation in South America was mainly driven by a larger “market size effect” indicating that technical change was biased towards land. Chapter 2 studied nitrogen fertilizer application in U.S. agriculture. Soil information (signal) obtained from the PA technology allowed the adoption of variable rate (VR) applications of nitrogen specific to the different plots (cells) of the field, with potential to increase the farmers’ profitability and decrease the environmental damages due to excessive use. I provided a Bayesian structural model, based on the Expected Value of Sample Information (EVSI) approach, with a direct application using data from the Data-Intensive Farm Management (DIFM) project from University of Illinois to evaluate the expected returns of the VR technology. The results from the studied U.S. farms showed that the information from soil electroconductivity (EC) provided low expected returns. The insights from the model revealed that the low returns can be explained by EC being “poorly” correlated with the true soilconditions and/or the quality of the soil may be uniform across the fields, hence, not supporting the VR technology adoption. Chapter 3 used satellite remotely sensed information to estimate the effects of droughts on agriculture for Brazilian municipalities. First, the effect of droughts for all the corn- and soybeans-producing Brazilian municipalities was estimated, then a model adding remote sensing data was estimated for the municipalities from a soybeans-producing region of Southern Brazil, both for the 2002-2016 period. The results implied that the lack of biophysical variables in the model, reflecting the interaction among the soil, the plant, and the atmosphere, would bias the drought effects. This is an important result because economic decisions are made based on the effects of climate conditions in agriculture and remote sensing information can provide more reliable estimates of the true climatic effects. Keywords: Agricultural technology. Innovation. Agricultural input use. Precision agriculture. Remote sensing information. Value of information.