Panel Data Parametric Frontier Technique for Measuring Total-factor Energy Efficiency: Application to Japanese Regions

Using the stochastic frontier analysis (SFA) model, we estimate total-factor energy efficiency (TFEE) scores for 47 regions across Japan during 1996–2008. We extend the cross-sectional SFA model proposed by Zhou et al. (Applied Energy, 2012) to panel data models and add environmental variables. The results provide not only the TFEE scores, in which statistical noise is taken into account, but also the determinants of inefficiency. The three SFA TFEEs are compared with a TFEE derived from data envelopment analysis (DEA). The four TFEEs are highly correlated with one another. For the inefficiency estimates, the higher the manufacturing industry share and wholesale and retail trade share, the lower the TFEE score.