The consequences of frequently used price expectation models are analyzed by comparing the responsiveness of U.S. corn and soybean acreages to six alternative formulations. The trade-off between bias and variance associated with these forecasts is investigated. The results of this study have important implications for future research on supply analysis.
