This website uses cookies.

By clicking the "Accept" button or continuing to browse our site, you agree to first-party and session-only cookies being stored on your device to enhance site navigation and analyze site performance and traffic. For more information on our use of cookies, please see our Privacy Policy.

  1. Home
  2. ASSA Annual Meeting
  3. 2024
  4. Program
  5. Inferring Preferences from Decision Time
« Back to Results

Inferring Preferences from Decision Time

Paper Session

Sunday, Jan. 7, 2024 8:00 AM - 10:00 AM (CST)

Grand Hyatt, Travis B
Hosted By: Econometric Society
  • Chair: Stefano DellaVigna, University of California-Berkeley

Fast or Slow? Uncovering Expert Views from Time to Decision

David Card
,
University of California-Berkeley
Stefano DellaVigna
,
University of California-Berkeley
Dmitry Taubinsky
,
University of California-Berkeley

Abstract

Cognitive scientists and psychologists stress the informativeness of decision time, e.g., in drift-diffusion models; yet, economists have made limited use of it outside of laboratory experiments. We show that decision time provides valuable information about the preferences and decision-making of experts, focusing on the decisions of referees and journal editors to give a revise and resubmit. A simple two-period model outlines two predictions. First, decision time should be inverse U-shaped in the perceived relative value of the two options. Second, the accuracy of the decision depends predictably on the decision time as a function of two forces—selection based on signal and learning over time. We bring these predictions to the editorial decision setting, taking advantage of the fact that we observe a proxy for quality of decisions, citations accumulated years later, as well as the decision inputs and the number of days taken at each step of decision. We document that the decision time is indeed inverse U-shaped in the signals received by the editor and referees. Second, we show that selection based on signal dominates the effect of additional learning. Consequently, rejected papers on which editors and referees take longer accumulate more citations ex post, all else constant. Instead, papers that received revisions after a longer delay receive fewer citations than revise-and-resubmits decided sooner. We provide estimates of a simple structural model that can be adapted to other settings.

Response Times in the Wild: eBay Sellers Take Hours Longer to Reject High Offers and Accept Low Offers

Miruna Cotet
,
Ohio State University
Ian Krajbich
,
Ohio State University

Abstract

Hesitation in the marketplace has the potential to betray private information. Recent results from lab experiments have confirmed that subjects’ response times reveal their strength-of-preference or belief, even in strategic settings. What remains unclear is whether these results extend beyond the lab to markets with experienced agents. Here we address this question using a dataset consisting of millions of bargaining exchanges from eBay. We find that the time it takes sellers to accept or reject offers is strongly related to the size of the offer. Sellers are quick to accept good offers and to reject bad offers, and slow to accept bad offers and to reject good offers. These response-time differences are on the order of hours. These findings apply to a majority of the exchanges on eBay, and are present with more experienced sellers. Overall, these results indicate that there is information in response-time data from non-lab markets, and that a majority of agents do not have prepared strategies but instead evaluate offers on the spot in a way that reveals their values for the goods.

Happy Times: Measuring Happiness Using Response Times

Nick Netzer
,
University of Zurich

Abstract

Surveys measuring happiness or preferences generate discrete ordinal data. Ordered response models, which are used to analyze such data, suffer from an identification problem. Their conclusions depend on distributional assumptions about a latent variable. We propose using response times to solve that problem. Response times contain information about the distribution of the latent variable through a chronometric effect. Using an online survey experiment, we verify the chronometric effect. We then provide theoretical conditions for testing conventional distributional assumptions. These assumptions are rejected in some cases, but overall our evidence is consistent with the qualitative validity of the conventional models.
JEL Classifications
  • D9 - Micro-Based Behavioral Economics
  • D8 - Information, Knowledge, and Uncertainty