April 2020

I have a new paper with Steve Bond, Arshia Hashemi and Piotr Zoch called Some Unpleasant Markup Arithmetic: Production Function Elasticities and their Estimation from Production Data. We highlight some commonly overlooked issues about the identification and estimation of markups using the ratio estimator - the ratio of the output elasticity of a variable input to that input’s cost share in revenue. This is the estimator that has underpinned a recent literature that seeks to estimate heterogeneity and trends in markups. Most importantly, we show that it is essential that one uses an output elasticity, as opposed to a revenue elasticity, in the numerator of the ratio estimator. When the revenue elasticity is used, then the ratio estimator contains no information about markups at all. We then show that without data on output quantities (as opposed to sales or revenue) it is not possible to obtain consistent estimates of the output elasticity, and hence markups, using any current methods, when firms have market power and there is heterogeneity in markups