Amazon's Effect on Prices: The Case of Mexico

Amazon’s efficient logistics chain lowers costs, which may result in lower prices for consumers on and off its website. In this paper, we analyze whether Amazon’s presence in Mexico has had a pro-competitive effect in lowering brick-and-mortar prices. We find that the entry of Amazon could have had caused an important procompetitive effect by reducing brick-and-mortar retail prices; and this was found on the entry of Amazon platform but also to the product level: each time a product started selling through the Amazon platform, brick-and-mortar prices of that product decreased. The cities with the highest e-commerce penetration show the strongest decreasing price effects. We first study the Consumer Price Indexes (CPIs) for groups of products and cities. We find that cities with more e-commerce consumers tend to have lower CPIs in furniture and clothing, which are groups of products in which e-commerce has a greater presence. This difference is found to be important and statistically significant. We also analyze a novel database we created by merging average brick-and-mortar prices for a selected set of products (using public information from the National Institute of Statistics and Geography, INEGI) with information on products sold on Amazon (obtained through a third-party app, Keepa). We find that in general, when a product started selling on Amazon, brick-and-mortar prices decreased (though to different degrees, depending on the type of product). This pro-competitive decreasing price effect is observed with the introduction of products either sold and delivered by Amazon or sold and delivered by third parties on Amazon’s website. Using a difference-in-differences empirical strategy, we find that this effect is more important in cities with larger numbers of e-commerce consumers. Finally, we conservatively estimate the welfare gains of this price decrease (ignoring other important welfare gains like increased product variety and increased ease of purchase) and find statistically significant welfare gains in states with higher absolute and proportional numbers of ecommerce consumers.