Productivity and efficiency in urban railways: Parametric and non-parametric estimates

This paper formulates and compares parametric productivity estimates and non-parametric efficiency scores for urban rail firms. It outlines a decomposition of total factor productivity (TFP) change and proposed some hypotheses about how this relates to an analysis of efficiency using data envelopment analysis (DEA). TFP is empirically evaluated using estimates from a translog production function. The estimation also tests some hypotheses about the elements that comprise firm specific 'technology'. The results show that while estimates of returns to scale differ using the TFP and DEA methods, the ranking of urban rail efficiency is broadly similar.