Vertical Integration and Consumer Choice: Evidence from a Field Experiment

Platforms, retailers, and other firms often offer their own products alongside products sold by competitors, but this form of vertical integration has become a target of regulation in digital markets. We study the effects of this practice through a field experiment that hides brands owned by Amazon (i.e., private labels) from shoppers on Amazon.com. We first consider the effects of this removal on three aspects of consumer behavior: substitution to other products, changes in search effort, and substitution to other retail websites. In the absence of Amazon brands, our results indicate that consumers substitute toward products that are similar along most observable dimensions. We find no evidence that treated consumers change their search effort, nor that they shift their shopping behavior to other retail websites. To evaluate a fourth mechanism—how the presence of Amazon brands affects equilibrium prices—we estimate a structural model of demand and simulate counterfactual prices when removing such products. Our estimates imply that, for the categories we study, removing Amazon brands would reduce consumer surplus by 3.8 percent in the short run, and roughly one quarter of the impact is due to equilibrium price increases by other products. The effects are heterogeneous, with consumer surplus reductions exceeding 10 percent in some categories, while other categories realize no change or even positive increases in consumer surplus when Amazon brands are removed.