The economic value of seasonal forecasts stochastic economywide analysis for East Africa

Rodrigues, Joao · Thurlow, James · Landman, Willem · Ringler, Claudia · Robertson, Richard D. · Zhu, Tingju

8/2016 · IFPRI Discussion Paper 1. kniha · Intl Food Policy Res Inst

Táto e‑kniha

There is growing interest within the climate change and development community in using seasonal forecast information to reduce the losses to agriculture resulting from climate variability, especially within food-insecure countries. However, forecast systems are expensive to establish and maintain, and therefore gauging the potential economic return to investments in forecast systems is crucial. Most studies that evaluate seasonal forecasts focus on developed countries and/or overlook agriculture’s economywide linkages. Yet forecasts may be more valuable in developing regions such as East Africa, where climate is variable and agriculture has macroeconomic importance. We use computable general equilibrium and process-based crop models to estimate the potential economywide value of national seasonal forecast systems in Kenya, Malawi, Mozambique, Tanzania, and Zambia. Stochastic seasonal simulations produce value distributions for forecasts of varying accuracy and varying levels of farm coverage. A timely and accurate forecast adopted by all farmers generates average regional income gains of US$113 million per year. Gains are much higher during extreme climate events and are generally pro-poor. The forecast value falls when forecast skill and farm coverage decline. National economic and trading structures, including the importance of agricultural exports, are found to be major determinants of forecast value. Economywide approaches are therefore needed to complement farm-level analysis when evaluating forecast systems in low-income agrarian economies.