Adaptation to Weather Shock through Harvesting Time Adjustments: Evidence from Field-level Data in the U.S

This paper examines how weather affects corn harvest timing and harvest duration using a field-level panel dataset of approximately 1.25 million corn field-year observations from 2007 to 2020. We focus on two margins of the harvest operation: growing-season weather determines when crops are ready to harvest, while weather during the harvest window itself determines how efficiently machinery can operate. We estimate how growing-season heat and precipitation affect farm harvest start dates, and how harvest-period precipitation and temperature affect harvest working hours per acre. Using within-farm variation across years, we find that each additional GDD in the range 8–29◦C during the growing season (April–August) advances farm harvest start by 0.053 days; moving across the interquartile range of within-farm heat accumulation is associated with roughly a 6-day earlier start. Harvest-period precipitation extends harvest duration with diminishing returns; at the sample mean of 7.5 mm, the marginal effect is 0.096 working hours per acre per millimeter. Both minimum and maximum temperatures on the cold end are consistent with a grain drydown mechanism. Minimum temperatures above 15◦C are associated with 1.2 fewer working hours per acre than the near-freezing reference bin (0–5◦C). Maximum temperatures below 5◦C require 0.38 more hours per acre than moderate maximum temperatures (15–20◦C).