Jakša Cvitanić, Dražen Prelec, Blake Riley and Benjamin Tereick. 2019. "Honesty Via Choice-Matching." American Economic Review: Insights, 1 (2): 179-92.
Abstract: We introduce choice-matching, a class of mechanisms for eliciting honest responses to a multiple choice question (MCQ), as might appear in a market research study, opinion poll, or economics experiment. Under choice-matching, respondents are compensated through an auxiliary task, e.g., a personal consumption choice or a forecast. Their compensation depends both on their performance on the auxiliary task, and on the performance of those respondents who matched their response to the MCQ. We give conditions for such mechanisms to be strictly truth-inducing, focusing on a special case in which the auxiliary task is to predict the answers of other respondents.
Improving Information Aggregation Through Meta-Cognitive Judgments
Abstract: How best to form a single judgment out of the many judgements existing in a group of individuals is an age-old problem in decision theory. In this paper, I evaluate methods which make use of an understudied resource for approaching this problem: the meta-beliefs (beliefs about beliefs) of the members themselves. In particular, I propose “self-aggregation” (SELF), a new aggregation scheme, and compare its performance with simple and confidence-weighted majority voting and with the Surprisingly Popular Algorithm (SPA) recently proposed by Prelec et al. (2017). SELF asks individuals to vote for an option and to simultaneously provide a threshold of the number of people that would convince them of the opposite. SELF then picks an option if more people vote for it than the average threshold provided in the group.
In a model in which individuals update their beliefs in a Bayesian fashion, I show that SELF is predicted to outperform alternatives. The model further gives an additional theoretical explanation for the previously found empirical success of the SPA. In an experimental test of the model, respondents solve a binary decision problem in a stylized urn experiment in which responses and aggregation results can be directly compared to the Bayesian prescription.
In the experiment, SELF compares favorably to (simple and confidence-weighted) majority voting, but does not realize its theoretical advantage over the SPA. The results show that while the meta-cognitive abilities of individuals are challenged by complex methods such as SELF and the SPA, responses contain sufficient information to outperform methods based on less challenging questions.
Follow the Money (under review, Joint with Aurélien Baillon and Tong V. Wang)
Abstract: For some questions, such as what the best policy to address a problem is, it is uncertain if the answer will ever be known. Asking experts yields two practical problems: how can their truth-telling be incentivized if the correct answer is unknowable? And if experts disagree, who should be trusted? This paper solves both problems simultaneously. Experts decide whether to endorse a statement and trade an asset whose value depends on the endorsement rate. The respective payoffs of buyers and sellers indicate whom to trust. We demonstrate theoretically and illustrate empirically that ``following the money" outperforms selecting the majority opinion.
Market Scoring Rules in a Bayesian Framework
Abstract: Hanson (2003, 2007) proposed the use of the logarithmic market scoring rule (LMSR) for eliciting private information about a future, verifiable, event. A market maker sets a baseline probabilistic forecast of the event and subsequent market participants report their own forecasts. Each participant is paid the logarithmic score of each of their forecasts, and pays the logarithmic score of the previous forecast. I show that the LMSR admits a Perfect Bayesian Nash equilibrium in truthful strategies in a setting in which agents receive information dynamically over several periods. This generalizes previous results in which information arrival is static, strengthening the status of the LMSR as an attractive payment scheme for prediction markets and forecasting tournaments.
Human Crowds versus Language-based Prediction in Lie Detection (under revision) (Joint with Aurélien Baillon, Sophie van der Zee and Tong V. Wang)
Intuitive forecasting in the short and long run: an experimental study (inactive) (Joint with Maxime Cugnon de Sévricourt)