An updated assessment of technical efficiency and returns to scale for US electric power plants

This paper utilizes cutting-edge panel stochastic frontier electricity production models to measure the impact of state and federal regulations on United States (U.S.) natural gas fired power plants from 1994 to 2016. Using an expansive dataset, I simultaneously account for plant specific heterogeneity as well as time varying and persistent inefficiency. Previous studies of electricity generation in the U.S. have ignored at least one of these components. The ability to control for unobserved heterogeneity is important in assessing long-term managerial effectiveness. Deploying a translog functional form, I extract firm specific information on returns to scale (RTS). For the preferred model, results indicate that mean persistent technical efficiency (TE) for natural gas fired power plants is 71 percent, transient TE is 72 percent, and RTS are 0.94. I find an inverse relationship between RTS and plant output level, but a positive relationship between TE and output level. For a given plant, the regulatory variables under consideration tend to increase plant efficiency, along the lines of the Porter hypothesis. Furthermore, plants with higher transient TE have significantly lower costs per kWh.