Avoiding Lemons in Search of Peaches: Designing Information Provision

The increasing amount of data available to consumers has most likely aided in decision-making. However, it has also created an opportunity for sellers to design the information landscape that consumers navigate. This paper develops a novel search model for alternatives with multiple characteristics, and reports estimation results for an online used car seller. The model allows search over alternatives with multiple characteristics with arbitrary marginal distributions and correlation structures. For example, more expensive vehicles may feature fewer past owners, and vehicles with higher mileage may reveal more issues in their inspection reports. The model also allows for a rich set of consumer search behaviors, including (but not limited to) sequential search within vehicles and characteristic-by-characteristic search across. The estimated fundamentals are then used to consider different information design policies. We find that the choice of the characteristics to be made available to consumers upfront has significant economic implications. For example, featuring variance-reducing information upfront (in our application, vehicle histories) instead of other characteristics translates into an approximate conversion rate increase of 20%, in relative terms. In light of our results, we provide intuition on how different information design policies affect consumer and seller welfares. Additional counterfactual analyses confirm our intuition. Finally, we show that a simplified approach based on traditional choice models would produce low quality recommendations about information design policies.