Assessing power stations performance using a DEA-bootstrap approach

Purpose - The purpose of this paper is to assess the performance of Greek fossil fuel-fired power stations employing a data envelopment analysis (DEA) model combined with bootstrapping. Design/methodology/approach - DEA is used to derive aggregate performance indicators using data on inputs and desirable and undesirable outputs for a sample of fossil fuel-fired power stations. The statistical significance of the derived aggregate performance indicators is assessed via the bootstrapping approach. Findings - The results suggest that the power stations in the sample are considerably more inefficient than revealed by the initial point estimates of inefficiency. Moreover, the non-lignite-fired stations of the sample are on an average more efficient than the lignite-fired stations. Research limitations/implications - DEA represents a useful framework for exploring the current state to derive aggregate performance indicators of power stations, and moreover, the statistical properties of these metrics can be assessed via the bootstrapping approach. Practical implications - The bootstrapping approach in DEA shows its superiority over DEA models that do not address the uncertainty surrounding point estimates. The DEA bootstrapping model used in this study to model environmental performance in the power station electricity production setting provides bias correction and confidence intervals for the point estimates and it is therefore more preferable. Originality/value - The derivation of aggregate performance indicators of Greek fossil fuel-fired power stations is an important addition to the existing literature on energy economics. The paper is also innovated in providing the statistical properties of the derived performance metrics.