Kuethe, Todd H.

By: Kuethe, Todd H. ; Bora, Siddhartha ; Katchova, Ani
The USDA Economic Research Service's (ERS) farm income forecasts play an important role in decision making and planning across the agricultural sector, yet recent studies suggest that their initial farm income forecasts are biased. This study examines the degree to which the initial forecast of net cash income and its components can be improved using information from USDA 10-year agricultural baseline (AB) projections. We apply several forecast evaluation tools to a unique set of ERS forecasts, AB projections, and official estimates from 1997 through 2019. Our forecast-encompassing tests show that the AB projection provides important information for predicting livestock receipts, direct government payments, farm-related income, and cash expenses. Our findings are potentially useful for both ERS forecasters and a variety of farm income forecast users.
By: Kuethe, Todd H.; Hubbs, Todd; Sanders, Dwight R.
The USDA produces four forecasts of net farm income for each year; these forecasts are closely monitored by decision makers across the agricultural sector. However, little is known about the performance of these forecasts. Traditional forecast evaluation tests suggest that between 1975 and 2016, the long-horizon forecasts systematically under-predicted realized values. In addition, the shorter-horizon forecast revisions overreact to new information. The findings suggest that forecast users should adjust their expectations and that the USDA may want to consider other forecast approaches to supplement current procedures.