Abstract
This paper examines how productivity changes contribute to short-run costvariations over time and across railroads. Using an unbalanced panel of US Class I railroad data for the period 1996-2003, decomposition models were used to attribute intertemporal and multilateral cost variations to their causal factors. We find that (i) railroads were able to reduce their physical capitals in recent years; (ii) railroads experienced technical progress over time, but they were technically and allocatively inefficient in some time periods; and (iii) the benchmarking railroads could learn from the low-cost and cost-efficient benchmark when it comes to cost savings. (Author/publisher).