The Not So Big Squeeze: Modeling Future Land Use Patterns of Miscanthus for Bioenergy Using Fine Scale Data

We predict land coverage in the Midwest for a new bioenergy crop, Miscanthus x Giganteus (miscanthus), using fine scale panel data and an econometric model. To explain farmers ’ current crop choices we use a local, limited dependent variable regression based on soil and weather characteristics. To this model we add miscanthus as a new crop, based on its place dependent BioCro model-predicted yield. We then predict in the Midwest where miscanthus will be grown assuming that the total production of miscanthus there will be 100 million ton (Mt), proportional to the scale of production mandated by the federal Renewable Fuel Standard. We find that the vast majority of all land used to grow miscanthus will come from land now used for non-major crops, pasture, woodland, etc. We also find that, given the total production of miscanthus, when the average cost of growing miscanthus decreases faster with yield, higher yield land will be selected; miscanthus hectares will shift to the south; less corn and soy land and more cotton land will be diverted