摘要：In this paper, we contribute to the existing literature of global yield gap analysis by using geo-spatial dataset on crop yields, agro-climatic factors and selected socio-economic variables. In line with the literature, we link yield gaps to technical efficiencies. By treating each grid-cell as a farm unit, we employ data envelopment analysis to calculate relative farm efficiencies with respect to a technically efficient global production frontier. We then apply spatial econometric techniques to relate the calculated efficiency scores to population, irrigation, fertilizer use, market access, institutional strength and market influence. We find that the effects of the socio-economic variables on the efficiency scores are not consistent and will typically vary by geographic region, crop type and on the scope of analysis (all areas, irrigated areas only, rainfed areas only). However, we can see some general trends. Most of the total impacts of socio-economic variables on efficiency scores are positive. For example, across all crops, regions and all areas, estimates of the total impact of irrigation, fertilizer use, institutional strength and market influence are generally positive. In regions wherein efficiency scores are low, the key variables which positively affect these scores are market influence and fertilizer use. By changing the coverage from all areas to either rainfed or irrigated areas, we generally see changes in the statistical significance of some variables.
