Abstract

The USDA international baseline projections are crucial for US agricultural policy. Using a dynamic time warping algorithm, we examine whether corn, soybean, and wheat baselines exhibit a lack of significant distance indicating similarity in projections across countries. We then examine whether projections for countries with more similarity with the United States projections exhibit higher projection errors. We discover evidence of projections being similar to the United States, which is associated with lower errors for corn yield, harvested area, and exports and wheat import projections, but not for other variables. These findings offer valuable insights for USDA and stakeholders.