A panel cointegration analysis and dynamic causality test of agricultural supply response to prices: The case of rice in West Africa

Abstract

The paper provides a regional estimate of rice yield and acreage supply response to prices in West Africa. Using balanced panel data of seven countries covering 1991–2006, the study employs panel cointegration estimators, namely FMOLS and DOLS to estimate the long-run rice yield supply function and using one-step Blundell-Bond system GMM to estimate the short-run acreage supply function in the region. While the empirical findings from both estimators provide similar results, the estimates show that price of maize is a key determinant of rice yield and acreage and there is rice-maize substitutability in both the long-run and short-run in the study. The short-run Granger causality runs only from maize price to rice acreage, while long run causality of joint effect of the prices on the acreage supply is also evident in the study. Our results also show that the rice yield and acreage supply functions are unresponsive to own price in the short -run and long -run, respectively.