Cropland use, yields, and droughts: spatial data modeling for Burkina Faso and Niger
运用动态空间面板数据模型，分析布基纳法索和尼日尔主要粮食作物产量和价格的空间溢出效应及对降雨的响应，发现多数作物产量对价格反应弱，而干旱导致高粱产量显著下降。
For countries with recurrent droughts, the design of drought impact mitigation measures could benefit from analyses of determinants of yields and prices of local crops at regional and district level. This study applies dynamic spatial panel data regression models to yields and prices of four major food crops across regions of Burkina Faso and Niger, over sample periods between 1984 and 2006. Results lend support to mainly simultaneous spatial spillovers, particularly for millet and cowpea prices and sorghum yields in Niger, and maize yields in Burkina Faso. After accounting for these effects, most crop yields are found to be weakly price-responsive, as envisaged by a supply-side geographical diffusion hypothesis. Seasonal rainfall elasticity estimates suggest that dominant food crops have slight advantage margins in terms of relative resilience to rainfall shortages. However, this result is to be weighed against low millet yields in Niger, and marked drops in sorghum yields during officially declared droughts in Burkina Faso.