Evan Magnusson
Job Market Candidate

Stanford University
Department of Economics
579 Jane Stanford Way
Stanford, CA 94305


Applied Microeconomics,
Industrial Organization,
Economics of Digitization

Expected Graduation Date:
March, 2020

Thesis Committee:
Frank Wolak (Primary):

Brad Larsen:

Peter Reiss:


Unboxing the Causal Effect of Ratings on Product Demand: Evidence from Wayfair.com (Job Market Paper)
I estimate the causal effect of increasing displayed product ratings by a half star on Wayfair.com to be a five percent increase in demand for products. This increase is driven by an increase in sales, not a change in price. Products with more ratings or that are sold by lesser known brands see a large effect of star ratings on orders, while products with fewer ratings or that are sold by well known brands see no effect of ratings on orders. Furthermore, I find the causal effect of ratings to be heterogeneous across products with different ratings–with the highest rated products seeing the largest benefit to having a marginally higher rating–but not across customers with varying degrees of experience purchasing products from the site. This heterogeneity in demand effects shows that the benefits of higher ratings are not equivalent across old and new products and sellers, indicating potential policy responses by e-commerce sites.

Seller Responses to Changes in Product Ratings
How do sellers respond to exogenous increases in demand for their products? I find that when the value of product star ratings increase by a half star, which on average increases product orders by 4.1% (see Magnusson (2019)), sellers on Wayfair.com do not adjust their wholesale prices. This is explained by competition in both the downstream and upstream markets in which sellers market their goods. Many sellers offer their products on other sites, and thus altering their price in response to demand increases would likely result in customers buying their product elsewhere. Even when this is not the case, there is sufficient competition between sellers of similar products on Wayfair.com to make holding wholesale prices constant the optimal strategy.

Competition, Reputation Inflation, and the Price Gap Between High and Low Rated Sellers
This paper seeks to quantify the difference in prices between products sold by high and low quality sellers, and how that difference is affected by the number of top-rated sellers in the market. On Amazon.com, 5 star sellers set prices around eight percent higher 4.5 star sellers, though this premium varies widely by the condition and category of the product the seller is listing. In markets with few sellers, five star sellers charge up to 12.5% more than 4.5 star sellers. However, as the number of five star sellers increases (holding constant the total number of sellers in the market), the price difference between their listings and those of 4.5 star sellers disappears.