DECOMPOSING REGIONAL EFFICIENCY*
研究使用稳健离群值的数据包络分析结合地理加性回归，分解了439个德国区域在基础设施和人力资本使用上的效率，发现效率受空间和非空间结构性因素驱动，建议区域资助方案应区分这两种因素。
ABSTRACT Applying an outlier robust extension of the data envelopment analysis (DEA) followed by a geoadditive regression analysis, this study identifies and decomposes the efficiency of 439 German regions in using infrastructure and human capital. The findings show that the regions’ efficiency is driven by a spatial and a nonspatial, arguably structural factor. As a consequence, concrete regional funding schemes, shaped by best practice results, might not be appropriate for all regions. Instead, a more differentiated funding scheme that accounts for both spatial and structural factors seems more promising.