Believing in forecasts, uncertainty, and rational expectations

Efe A. Ok, Andrei Savochkin

    Research output: Contribution to journalArticlepeer-review


    We model situations of choice under uncertainty where one is exogenously given information about the unknown states as a “suggested prior” (as in weather forecasts, betting odds provided by bookmakers, success likelihoods provided by medical doctors, estimates given by financial analysts, etc.). We wish to understand when a decision maker would adopt the suggested prior as her own subjective beliefs, yielding fully to the power of suggestion. We find that this happens under surprisingly weak conditions: If a preference relation, may it be complete or incomplete, (1) uses the information it is given consistently (in the sense of being state-neutral) and (2) believes that events that are suggested to occur with zero probability will indeed not occur, then it is not only probabilistically sophisticated, but also holds the suggested beliefs as actual beliefs. If the agent is a (subjective) expected utility maximizer, this happens even in the absence of condition (2).

    Original languageEnglish (US)
    Pages (from-to)947-971
    Number of pages25
    JournalEconomic Theory
    Issue number3
    StatePublished - Oct 2022


    • Distortion of probabilities
    • Power of suggestion
    • Probabilistic sophistication
    • Rational expectations

    ASJC Scopus subject areas

    • Economics and Econometrics


    Dive into the research topics of 'Believing in forecasts, uncertainty, and rational expectations'. Together they form a unique fingerprint.

    Cite this