Has the Averch-Johnson Effect Been Empirically Verified?

Abstract

Recently three studies have been published which claim to confirm the existence of the Averch-Johnson effect in the electric power industry. Each of these papers uses a distinctly different methodology. This paper examines the general problem of what the nature of the A-J effect might be and what sort of data would be required in order to confirm its presence. The other studies are then critically examined on the basis of this discussion. A modification of the method used in one previous study is then used to test the A-J hypothesis, and no evidence of capital bias is found. The principal conclusion of this study is that if the A-J effect is significant in distorting input choices in the electric utility industry, very different sorts of data than those that have been used thus far are going to be required in order to verify its presence. Mechanical usage of gross input and output numbers, without understanding of the technological processes involved, leads only to erroneous conclusions.