Empirical Properties of Diversion Ratios
NBER Working Paper No. 24816
The diversion ratio for products j and k can be interpreted as the result of an experiment: increase the price of product j, measure the number of consumers who leave product j, and then measure the fraction of leavers who switch to a substitute product k. In theory, this is expressed as the ratio of demand derivatives that enter the multi-product Bertrand-Nash first-order condition for a firm. In practice, diversion ratios are also measured from second-choice data or customer switching surveys. We establish a LATE interpretation of diversion ratios, and show how diversion ratios can be obtained from different interventions (price changes, quality changes, assortment changes) and how those different measures relate to one another and to the underlying properties of demand.
Document Object Identifier (DOI): 10.3386/w24816