This paper develops a dynamic micro-economic land use model that maximizes social welfare and internalizes externality from greenhouse gas emissions to obtain the optimal land use allocation for traditional row crops and bioenergy crops (corn stover, miscanthus and switchgrass), the mix of cellulosic feedstocks and fuel and food prices. We use this carbon tax policy as a benchmark to compare the implications of existing biofuel policies on land use, social welfare and the environment for the 2007-2022 period. The model is operationalized using yields of perennial grasses obtained from a biophysical model, county level data on yields of traditional row crops and production costs for row crops and bioenergy crops in Illinois. We show that a carbon tax policy that is directly related to carbon intensity of fuels can generate the highest social welfare among alternative policy scenarios. The existing ethanol tax credits result in substantial deadweight losses and higher GHG emissions as compared to the baseline. Ethanol blending mandates with subsidies lead to further welfare losses and higher GHG emissions. To meet advanced biofuel blending mandates, corn stover and miscanthus are used but the mix of viable cellulosic feedstocks varies spatially and temporally. Corn stover is viable mainly in central and northern Illinois while miscanthus acres are primarily concentrated on southern Illinois. The blending mandates lead to a significant shift in acreage from soybeans and pasture to corn and a change in crop rotation and tillage practices.