An Empirical Model of Quantity Discounts with Large Choice Sets

Abstract We introduce a Generalized Nested Logit model of demand for bundles that can be estimated sequentially, eliminating the challenge of dimensionality due to large choice sets. We use it to investigate quantity discounts for carbonated soft drinks by simulating a counterfactual with linear pricing. The prices of small-volume UPCs (<2L) decrease by −29.2% while those of large-volume UPCs (≥2L) increase by +16.8%. Purchased quantities decrease by −19.5% and industry profit by −6.3%. Consumer surplus however reduces only moderately, suggesting that a ban on quantity discounts on sugary drinks may be a simple and effective policy to limit added sugar intake. Our calculations confirm that such a ban would be as effective as a sugar tax of 1.5/oz of added sugar and reduce added sugar intake by −17.4%. A nested logit model that mistakenly ignores joint purchases of UPCs within shopping trips would instead simulate a reduction in added sugar intake of only −3.9%.