A bio-economic assessment of the spatial dynamics of U.S. corn production and yields

University of Minnesota Ph.D. dissertation. February 2012. Major: Applied Economics. Advisors: Philip G. Pardey, Terrance Hurley. 1 computer file (PDF); viii, 172 pages, appendix p. 172-153.This dissertation reports on an investigation into the effects of location on corn production and productivity. The landscape of crop production is dynamic--where crops are produced changes dramatically over time. The answers to important questions about the potential impacts of global climate change and whether agriculture will be able to meet the world's increasing need for food are affected by the moving footprint of production. However, most studies of agricultural productivity and the effects of global warming do not consider that agriculture moves, and that the concomitant changes in natural services have important effects. A full set of county-level census data on corn production and area in the United States have been digitized and assembled for the first time, and new methods have been applied to account for changing geopolitical boundaries. Concepts adapted from economic index number theory are used to show that some 15 to 20 percent of the change in U.S. corn output over the past 130 years has come about due to shifts in where corn is produced. A newly developed, long-run, corn-specific weather dataset is used with the county data to show that, because of changes in the location of production, U.S. corn is now grown in cooler climates than it was a century ago, possibly offsetting some of the potential impacts of climate change. Finally, methods from ecological modeling, spatial econometrics, and crop modeling are combined to create a corn yield model that is then used to develop a location- and season-specific crop suitability indicator that takes into account the intra-seasonal dynamics of weather and the complex relationships between weather and yields. It will be shown that the suitability metric developed in this study gives results that are both consistent and more interpretable than more traditional approaches