Set Identification of Price Elasticities of Demand - A Narrative Sign Restriction Approach

Providing insight into both consumer behaviour and market dynamics, information about the price elasticity of demand enables companies to optimise pricing strategies and improve their overall competitiveness. As prices and quantities are determined jointly, the empirical evaluation of price elasticities of demand requires a sound understanding of structural (i.e., truly causal) relations. External information required for the econometric assessment of structural relations is often scant, weak or subject to debate. As an alternative, we employ structural vector autoregressive models that are endowed with consensual sign restrictions to identify price elasticities for a total of 31 categories of consumer packaged goods and about 3,000 stores in the US. Rather intuitive narrative restrictions imposed on the functioning of sales promotions turn out to effectively sharpen the identified sets, and result in elasticity estimates that are sufficiently precise to improve the design of retail pricing strategies, and to offer interesting views on their (i) spatial profile across US states, (ii) dependence on demographic factors, and (iii) viable patterns of positive correlations (i.e., spillovers) across related product categories.