A reply to ‘Multiyear versus single-year drought: a comment on Peck and Adams’

Abstract

We appreciate Tihomir Ancev’s interest in our research on the economic consequences of multiyear droughts (Peck and Adams 2010). Ancev raises two concerns about our paper in their comment. First, he is concerned that our main conclusion (i.e. drought in one year can influence the effects of drought in subsequent years) may be an artefact of our representation of crop interdependencies as ‘agronomic rules.’ He asks, specifically, whether our conclusion would change if we allowed the model to make trade-offs between drought management and pest and disease management. Ancev’s second concern is that our conclusion does not hold for all drought scenarios we analysed. Because our conclusion does not hold universally (across all states of nature in the analysis), Ancev questions its validity. He closes by recommending an alternative approach that would compare the effects of single versus multiyear drought events. Ancev is correct that agricultural producers facing drought are not rigidly bound by agronomic rules, but instead face trade-offs between crops that help manage scarce water resources and those that help manage pests and diseases. Producers may violate agronomic rules if they are more concerned about mitigating the financial impact of drought in the current year than the financial impact of increased pests and diseases in future years. Continuous production functions would capture the yield effects of alternative crop sequences more precisely and allow marginal trade-offs between drought and pest/disease management. Unfortunately, we found no such production functions in the literature for our study area’s major crops (wheat, corn, sugar beets and onions). El-Nazer and McCarl (1986) estimated production functions for wheat and corn in the Pacific Northwest, but their study did not include beets or onions, so their production functions do not capture the yield effects of sugar beets or onions on wheat and corn, or vice versa. Nor were adequate production data available from the study area to estimate functions that capture crop interdependencies.