Railroad productivity analysis: case of the American Class I railroads

Purpose – The purpose of this paper is to provide an economic assessment of the productivity growth and technical efficiency of US Class I railroads for the period of 2002‐2007. Design/methodology/approach – The US railroad industry has become increasingly concentrated with seven Class I railroads accounting for over 90 percent of the industry's revenue. Because the small sample size creates a dimensionality problem for data envelopment analysis (DEA) with contemporaneous frontiers, the authors use sequential DEA and calculate the Malmquist productivity indexes using sequential frontiers. Through a decomposition process, changes in productivity are attributed to technical efficiency change, technical change, and scale efficiency change. Findings – Burlington Northern Santa Fe (BNSF) led the industry in terms of productivity growth (4.6 percent) and consistently stayed on the production frontier in every period studied; both BNSF and Union Pacific (UP) are top innovators in the industry, but UP trailed BNSF in both productivity growth and technological innovations by wide margins; and Grand Trunk Corporation was very good at “catching up” or leading its peers in efficiency improvements. Research limitations/implications – Railroads have invested heavily in technology over the years to enhance efficiency and productivity. However, two recent economic studies find that railroad productivity has slowed in recent years. The authors' benchmarking analysis sheds light on how individual railroads performed relative to their peers, and what they could learn from industry best practice. Originality/value – The benchmarking study enables the authors to report each railroad's performance instead of reporting industry‐wide aggregate indexes or industry averages which tend to mask performance variations. The paper also examines the causal factors of recent productivity growth and provides useful information for the industry and its regulators.