Communicating Probabilistic Forecasts to Decision Makers: A Case Study of Zimbabwe

Abstract

Seasonal climate forecasts offer the possibility of helping people to change their decisions in response to scientific information. With an improving ability to model and predict the El Niño / Southern Oscillation, climatologists are able to issues seasonal forecasts that in some places are quite reliable. One such place, is Zimbabwe, lying in the semi-arid tropics of southern Africa, and with an economy highly dependent on rain-fed agriculture. Starting in 1997, there have been efforts to apply seasonal forecasts to decision making in Zimbabwe. The success of these efforts has been mixed. This study examines these efforts, and attempts to explain why they may have been more or less successful. Drawing off literature in environmental assessment, risk communication, and behavioral economics, this study offers guidance for ways to improve the forecast applications process, particularly with respect to the communication of probabilistic information. Additionally, this study seeks to test whether the recommended course of action-a highly participatory assessment process examining uncertainties in great detail-could succeed, through the undertaking of a behavioral economic experiment in rural villages throughout the country. The experimental results suggest the approach could work.

Seasonal climate forecasts offer the possibility of helping people to change their decisions in response to scientific information. With an improving ability to model and predict the El Niño / Southern Oscillation, climatologists are able to issues seasonal forecasts that in some places are quite reliable. One such place, is Zimbabwe, lying in the semi-arid tropics of southern Africa, and with an economy highly dependent on rain-fed agriculture. Starting in 1997, there have been efforts to apply seasonal forecasts to decision making in Zimbabwe. The success of these efforts has been mixed. This study examines these efforts, and attempts to explain why they may have been more or less successful. Drawing off literature in environmental assessment, risk communication, and behavioral economics, this study offers guidance for ways to improve the forecast applications process, particularly with respect to the communication of probabilistic information. Additionally, this study seeks to test whether the recommended course of action-a highly participatory assessment process examining uncertainties in great detail-could succeed, through the undertaking of a behavioral economic experiment in rural villages throughout the country. The experimental results suggest the approach could work.