Ethanol and sugarcane expansion in the Brazilian Cerrado: farm, industry, and market analyses

Abstract

Brazil is one of the leading producers of ethanol, sugar, and sugarcane. Increasing demand for biofuels aligned with public policies prompted the expansion of sugarcane into the Brazilian Cerrado, particularly, into the states of Goiás and Mato Grosso do Sul. The overall purpose of this dissertation, comprised of three essays, is to understand the impacts from the sugarcane expansion on farmers, processors, and the market. At the market level, the first essay, estimates the impacts of public policies and market factors on ethanol and sugar, supply and demand, in Goiás and Mato Grosso do Sul, using three-stage least squares. Results show that ethanol supply is sensitive to public policies whereas the sugar supply is sensitive to market prices. Sugar and ethanol were found to be complementary outputs. For ethanol expansion to be sustainable the ethanol market must be developed to the extent that it relies on market factors and is no longer dependent on public policies. At the farmer level, the second essay, examines farmers' willingness to sign a sugarcane contract with a mill in the Brazilian Cerrado. A hypothetical stated choice experiment was conducted with farmers in Goiás and Mato Grosso do Sul. Respondents choose between three contracts (land rental, agricultural partnership and supply) and two optout options ("keep current contract" or "not grow sugarcane"). A single and a two opt-out random parameters models were estimated. The two opt-out model allowed for a better interpretation of the status quo. Willingness to pay, direct and cross-elasticity measures for contract attributes were calculated. Results showed that farmers prefer contracts with higher returns, shorter duration and a lower probability of late payments. Farmers seemed to prefer to renting out their land to the mill than to produce sugarcane themselves, which could lead to consequences for rural development and the sustainability of sugarcane expansion. At the processor level, the third essay investigates the impact of vertical coordination on input-oriented technical efficiency using data envelopment analysis (first stage) and a Tobit censored model (second stage). 204 Brazilian mills were considered. The second stage controlled for vertical integration as well as other characteristics of the mill. Vertical integration was measured as the percentage of total sugarcane used, supplied by mills. A negative, though minimal, relationship between vertical integration and technical efficiency was found. Hence, technical efficiency is not the major driver of vertical integration. Other vertical coordination strategies may bring more benefits in terms of technical efficiency (e.g. contracts). Drivers of vertical integration seem to vary according to the characteristics of the location of the mill.