Abstract

Bundled loyalty discounts are considered exclusionary whenever an equally efficient competitor in the competitive (“tied”) market cannot earn a profit while compensating the buyer for forgone discounts on the monopolized (“tying”) product. This “discount attribution” standard has been applied in cases involving share-based loyalty rebates for a single product; in those cases, the “contestable” portion of demand plays the role of the tied product, and the non-contestable portion constitutes the tying product. Economists have developed an alternative test that predicts the impact on consumer welfare by comparing the standalone price of the tying product to the profit-maximizing price that would prevail the absence of the bundled (or share-based) rebate. Economists have proposed to implement this test by comparing the price of the tying good before and after the implementation of the loyalty rebate. We propose a framework that builds on these concepts, expanding the analysis to model fluctuations in market conditions. In a review of three recent cases in the pharmaceutical industry, we find that comparisons based on a simple before-after approach are often clouded by confounding factors. By explicitly modeling shifts in demand and cost structures, our proposed framework provides a more robust mechanism for applying these consumer-welfare-based tests.

You do not currently have access to this article.