Chasing Diamonds and Crowns: Consumer Limited Attention and Seller Response

Abstract

Online platforms often assign sellers summary symbols based on whether their ratings pass certain thresholds. Consumers may focus on the symbols and be limited attentive to the ratings. This bias leads to discontinuously increased demand at the thresholds. I use a theoretical model to illustrate that sellers will lower the prices before their ratings reach the thresholds and increase their prices afterward due to the positive demand shock. I collect data from Taobao to test the theoretical predictions. Using regression discontinuity, I find that on the demand-side, the hourly sales increase significantly when a seller passes the threshold, even conditional on the same item. On the supply side, the prices indeed exhibit a V-shaped pattern with respect to the ratings. Furthermore, sellers preemptively increase prices shortly before reaching thresholds, supporting the theoretical predictions.

Keywords: Limited Attention, Behavioral IO, Online Reputation, Regression Discontinuity

JEL Classification: D22, L81, M31