An Improved Three-stage DEA as Business Intelligence for Benchmarking Luxury Hotels with Environmental Effect

Abstract

Investors need efficient method to effectively evaluate true performance in the big data era. An improved three-stage slack-based DEA method is proposed as business intelligence to overcome the independently identically distributed (i.i.d.) restriction on small sample size in the popular three-stage DEA proposed by Fried. Later, it is applied to find the relative efficiencies for international hotels. By treating environmental variable (near airport) as non-discretionary input, each hotel can consider non-discretionary input as exogenous variable to adjust input by filtering out environmental effects and statistical noise. The improved relative efficiencies can be resulted from referred hotels. The results show the environmental effect is significant but not as important as internal resources.