AI Summary of Scholarly Research

This page presents an AI-generated summary of a published research paper. The original authors did not write or review this article. [See full disclosure ↓]

Choice models can be estimated despite consumer search

Research area:economics-policyvaluation-environment

What the study found

The study finds conditions under which choice data can identify preferences even when consumers may not know all product attributes. It also reports that the method can be used to test for full information, forecast responses to new information, and conduct welfare analysis when consumers are imperfectly informed.

Why the authors say this matters

The authors say the study suggests a way to test whether consumers have full information, forecast how they will respond to information, and conduct welfare analysis when they are imperfectly informed. They also conclude that the method can help identify which attribute was not immediately visible in search results and compute the value of additional information.

What the researchers tested

The researchers developed a method for estimating discrete choice models, which are statistical models of choices among alternatives, under consumer search. They applied it in a lab experiment and in data from Expedia.

What worked and what didn't

In the lab experiment, the method successfully forecast the average response to new information when consumers engaged in costly search. In Expedia data, it identified which attribute was not immediately visible in search results and allowed computation of the value of additional information.

What to keep in mind

The abstract does not describe specific limitations or failures of the method. The summary also does not provide details on the exact conditions required for identification beyond stating that such conditions exist.

Key points

  • The study says choice data can identify preferences even when consumers are not fully informed about goods' attributes.
  • The method can be used to test for full information and to forecast responses to new information.
  • In a lab experiment, the method successfully forecast average responses when search was costly.
  • In Expedia data, the method identified an attribute that was not immediately visible in search results.
  • The method also allowed the authors to compute the value of additional information.

Disclosure

Research title:
Choice models can be estimated despite consumer search
Authors:
Jason Abaluck, Giovanni Compiani, Fan Zhang
Institutions:
Universidade Nova de Lisboa, University of Chicago, Yale University
Publication date:
2026-01-07
OpenAlex record:
View
AI provenance: This post was generated by gpt-5.4-mini (OpenAI). The original authors did not write or review this post.