Modeling and Forecasting Agricultural Commodity Support in the Developing Countries

We use econometric models to study the links between the evolution of agricultural support for six agricultural commodities and economic development as measured by real income per capita. Each commodity has a panel dataset with around 30-50 countries including developed, developing and less developed countries over 1961-2011. We investigate more complicated nonlinear relationships between income and support as measured by Nominal Rates of Assistance (NRAs) than previously examined and employ fixed effects to capture heterogeneity across countries. We find that a significant relationship exists between income measures and measures of border protection, but that the link between income and domestic support is generally weaker. Using these estimates and projections of macroeconomic variables, projections of future agricultural commodity support are generated for Brazil, Russia, India, and China. The projections of economic measures of support are then be compared to the commitments made by these countries to the World Trade Organization (WTO). There is a clear distinction between actual policies in place, aggregate estimates of them (such as NRAs), and WTO notifications. We do not forecast WTO notification data but compare in a general way long-run trends in development and associated support to these countries’ multilateral commitments. We use the projections derived from the empirical models to discuss the drivers of agricultural support, how these might have implications for WTO notifications, and how these effects relate to WTO commitments.