Do revisions improve agricultural baselines?

Abstract

Long-term agricultural baseline projections are revised annually to incorporate new information. We evaluate the effectiveness of the US Department of Agriculture and Food and Agricultural Policy Research Institute crops and farm income baseline revisions to inform policymakers. The revisions are effective in improving the baseline projections in comparison to the initial projections, but they do not outperform the naïve projections. Upward revisions are found to be more effective than downward revisions. We find evidence of predictability and therefore a violation of weak efficiency for the revisions, which results from information rigidity for the crop variables and from information rigidity and strategic smoothing for the farm income variables.