Answers are grounded in this paper and cite their sources.
Results indicate that, when comparing the unconditional derived-demand elasticities to the unconditional consumer demand elasticities, significant differences emerge due to the differences in the first-stage estimation procedure between the differential production approach and the Rotterdam model. In comparing the consumer demand price/cross-price elasticities to the derived-demand price/cross-price elasticities, it is clear that use of the Rotterdam model when a production approach should be used can lead to overestimation, underestimation, and incorrect signs in deriving unconditional price effects.
No takes yet. Share an insight, caveat, or question.
Washington et al. (2002) studied this question.
Synapse has enriched 10 related papers. Consider them for comparative context: