Anders Munk-Nielsen, University of Copenhagen

"Markups on Dropdowns: Lexicographically Ordered Search and Drug Prices"

Abstract

When doctors in Denmark write prescriptions for generic drugs, they choose from a lexicographically (alphabetically) sorted dropdown list. This results in a competitive advantage to being ranked alphabetically first. We show that firm prices respond to alphabetical rank in a statistically significant way and that the relationship is robust to using vastly different sources of variation, relying on our unique data setting with many markets and many time periods. We then explore whether prices are increasing or decreasing in rank. Theoretical work has found opposite conclusions depending on whether there is heterogeneity in search costs or product match values. Our findings suggest that the answer depends on the level of competition: the price-rank gradient is negative in markets with few firms and gradually increases and eventually becomes positive in markets with many firms. Leveraging transaction-level data linked to consumer demographics on the full population, we show that this competition channel is far stronger than the previously emphasized channels. 

Keywords: Ordered search, pharmaceutical prices, market power, status quo bias, firm names.