Identifying the elasticity of substitution with biased technical change

Despite being critical parameters in many economic fields, the received wisdom, in theoretical and empirical literatures, states that joint identification of the elasticity of capital-labor substitution and technical bias is infeasible. This paper challenges that pessimistic interpretation. Putting the new approach of "normalized" production functions at the heart of a Monte Carlo analysis we identify the conditions under which identification is feasible and robust. The key result is that the jointly modeling the production function and first-order conditions is superior to single-equation approaches in terms of robustly capturing production and technical parameters, especially when merged with "normalization". Our results will have fundamental implications for production-function estimation under non-neutral technical change, for understanding the empirical relevance of normalization and the variability underlying past empirical studies.