The elasticity of substitution between clean and dirty energy lies at the center of leading economic analyses of climate policy. Despite the importance, empirical assessments of this key parameter remain scarce. This paper fills the gap by estimating the elasticity of substitution between clean and dirty energy using data on French manufacturing firms. Using the elasticity estimates, I then measure the bias in technical change at the firm level and find that technical change was largely biased toward dirty energy, despite a recent shift toward clean technologies. The findings are rationalized by the historical movement of energy prices in France and provide a micro-empirical foundation for the theory of directed technical change. Finally, I recover the aggregate elasticity of substitution that combines substitution within firms and reallocation across firms from the micro elasticity estimates.
