Spatial dependence in production frontier models

In recent years, spatial dependence in production frontier models has attracted considerable research attention. This is explained, among other things, by the observation that in many economic sectors, decisions on taxes, expenditure, research and development, etc., are likely to be interconnected across production units. Consequently, the assumption that (benchmarked) production units operate in isolation from their peers may no longer be valid. In this article, we review and summarize this literature. First, we identify a set of analyses in the production frontier field that take into account spatial dependence based on a systematic search in the Web of Science and Scopus databases from 1975 to 2022, and provide global insights into this subfield. We find that this literature is relatively young: we trace the first contribution back to 2004 and, until 2011, the number of analyses per year remained relatively low (one per year at most). Second, we analyze identified articles in depth, classify them, and briefly outline avenues for future research. We hope that this will enable researchers to navigate this subfield and develop it further. We observe that the majority of these articles use parametric methods (e.g., stochastic frontier analysis) over nonparametric ones (e.g., data envelopment analysis). We also note that while in the nonparametric framework, the objective of the vast majority of contributions is to distinguish the intrinsic performance of productive units by the contribution of their territories, contributions in the parametric framework are more diverse.(This abstract was borrowed from another version of this item.)