Abstract
One of the most important effects that railways have on the environment is noise pollution, notably in Europe. The purpose of this study is to evaluate the environmental efficiency of railways in 22 European countries, considering two factors; a country's response in retrofitting their wagon fleet with more silent braking technology and the number of people affected by railway noise. The railway transport process efficiency is decomposed into assets and service efficiency. The additive decomposition network Data Envelopment Analysis (NDEA) approach is customised to account for intermediate and undesirable outputs. Results suggest that Estonia, Germany and Poland are overall environmentally efficient and that except for Finland, asset efficient countries are also service efficient; the inverse does not hold. Sensitivity analysis revealed that efficiency rankings are robust to alterations in the decomposition weight restrictions. This is the first study that uses DEA to incorporate the noise-pollution problem in railway efficiency measurement.