Investigating the drivers of railway performance: Evidence from selected Asian countries

Although efficiency measurement has become relevant for logistics infrastructure planning, research on railway efficiency still remain scarce and rather focused on discussing rankings to the detriment of possible improvement paths. In fact, while the use of multi-activity models is increasing in railway efficiency research, previous studies fail short to assess their drivers at each operational stage. Here, we develop a novel super-efficiency Multi-activity Network DEA (MNDEA) model - based on directional distance functions (DDFs) and capable of handling undesirable outputs - to assess how different contextual variables impact railway efficiency levels in Asia. Two case studies are provided: one focused on six different countries, taken in aggregate (Japan, Thailand, Vietnam, Malaysia, Myanmar, and Indonesia). The other, on major state-owned Chinese railways. Differently from previous research, Generalized Additive Models for Location, Scale and Shape (GAMLSS) are used for the first time to regress super-efficiency scores on the contextual variable set. Findings reveal that the Asian railways are strongly marked by heterogeneity, the Chinese railways need to improve passenger-operation efficiency, while the other countries need to increase the cargo-operation efficiency. It also sheds lights on the design of policies for efficiency improvement in several different areas for the Asian railway system.