Weather Extremes and the Costs of Cotton Acreage Abandonment

This study uses a double hurdle model to estimate the relative importance of weather extremes, irrigation, cotton variety choice, and changes in cotton prices on rates of county-level cotton acreage abandonment. The model’s first stage estimates the probability that a county will have any cotton acres abandoned. The second stage estimates severity of abandonment, given positive abandonment.  The protective effects of irrigation were quantified, while seasonal temperature and precipitation variables were significant predictors of abandonment. The probability that a county had some cotton acres abandoned increased with the number of cotton farms. The rate of abandonment for counties with abandoned acres, however, declined with the number of cotton farms. This result justifies use of the flexible double hurdle model over a tobit model of crop abandonment.