A Spatial Stochastic Frontier Model with Omitted Variables: Electricity Distribution in Norway
将空间计量与随机前沿分析结合，利用相邻企业的地理信息作为代理变量，控制未观测成本驱动因素，以挪威电力配送企业2004-2011年数据验证该方法可弥补天气和地理条件数据的缺失，适用于公用事业效率分析。
An important methodological issue in efficiency analysis for incentive regulation of utilities is how to account for the effect of unobserved cost drivers such as environmental factors. We combine a spatial econometric approach with stochastic frontier analysis to control for unobserved environmental conditions when measuring efficiency of electricity distribution utilities. Our empirical strategy relies on the geographic location of firms as a source of information that has previously not been explored in the literature. The underlying idea is to utilise data from neighbouring firms that can be spatially correlated as proxies for unobserved cost drivers. We illustrate this approach using a dataset of Norwegian distribution utilities for the 2004-2011 period. We show that the lack of information on weather and geographic conditions can be compensated with data from surrounding firms. The methodology can be used in efficiency analysis and regulation of other utilities sectors where unobservable cost drivers are important, e.g. gas, water, agriculture, fishing.