Measuring Railway Performance With Adjustment of Environmental Effects, Data Noise and Slacks

Conventional data envelopment analysis (DEA) approaches (e.g., CCR model, 1978; BCC model, 1984) do not adjust the environmental effects, data noise and slacks while comparing the relative efficiency of decision-making units (DMUs). Consequently, the comparison can be seriously biased because the heterogeneous DMUs are not adjusted to a common platform of operating environment and a common state of nature. Although Fried et al. (2002, Journal of Productivity Analysis, 17, 157–174) attempted to overcome this problem by proposing a three-stage DEA approach, they did not account for the slack effects and thus also led to biased comparison. In measuring the productivity growth, Färe et al. (1994, American Economic Review, 84, 66–83) proposed a method to calculate the input or output distance functions. Similarly, they did not take environmental effects, statistical noise and slacks into account and thus also resulted in biased results. To correct these shortcomings, this paper proposes a four-stage DEA approach to measure the railway transport technical efficiency and service effectiveness, and a four-stage method to measure the productivity and sales capability growths, both incorporated with environmental effects, data noise and slacks adjustment. In the empirical study, a total of 308 data points, composed of 44 worldwide railways over seven years (1995-2001), are used as the tested DMUs. The empirical results have shown strong evidence that efficiency and effectiveness scores are overestimated, and productivity and sales capability growths are also overstated, provided that the environmental effects, data noise and slacks are not adjusted. Based on our empirical findings, important policy implications are addressed and amelioration strategies for operating railways are proposed.