Bayesian learning in social networks

Douglas Gale, Shachar Kariv

    Research output: Contribution to journalArticle

    Abstract

    We extend the standard model of social learning in two ways. First, we introduce a social network and assume that agents can only observe the actions of agents to whom they are connected by this network. Secondly, we allow agents to choose a different action at each date. If the network satisfies a connectedness assumption, the initial diversity resulting from diverse private information is eventually replaced by uniformity of actions, though not necessarily of beliefs, in finite time with probability one. We look at particular networks to illustrate the impact of network architecture on speed of convergence and the optimality of absorbing states. Convergence is remarkably rapid, so that asymptotic results are a good approximation even in the medium run.

    Original languageEnglish (US)
    Pages (from-to)329-346
    Number of pages18
    JournalGames and Economic Behavior
    Volume45
    Issue number2
    DOIs
    StatePublished - 2003

    Fingerprint

    Bayesian learning
    Social networks
    Private information
    Speed of convergence
    Connectedness
    Approximation
    Optimality
    Social learning
    Uniformity

    Keywords

    • Herd behavior
    • Informational cascades
    • Networks
    • Social learning

    ASJC Scopus subject areas

    • Finance
    • Economics and Econometrics

    Cite this

    Bayesian learning in social networks. / Gale, Douglas; Kariv, Shachar.

    In: Games and Economic Behavior, Vol. 45, No. 2, 2003, p. 329-346.

    Research output: Contribution to journalArticle

    Gale, Douglas ; Kariv, Shachar. / Bayesian learning in social networks. In: Games and Economic Behavior. 2003 ; Vol. 45, No. 2. pp. 329-346.
    @article{beb4ec7e49f94ed9b3019e6afe440556,
    title = "Bayesian learning in social networks",
    abstract = "We extend the standard model of social learning in two ways. First, we introduce a social network and assume that agents can only observe the actions of agents to whom they are connected by this network. Secondly, we allow agents to choose a different action at each date. If the network satisfies a connectedness assumption, the initial diversity resulting from diverse private information is eventually replaced by uniformity of actions, though not necessarily of beliefs, in finite time with probability one. We look at particular networks to illustrate the impact of network architecture on speed of convergence and the optimality of absorbing states. Convergence is remarkably rapid, so that asymptotic results are a good approximation even in the medium run.",
    keywords = "Herd behavior, Informational cascades, Networks, Social learning",
    author = "Douglas Gale and Shachar Kariv",
    year = "2003",
    doi = "10.1016/S0899-8256(03)00144-1",
    language = "English (US)",
    volume = "45",
    pages = "329--346",
    journal = "Games and Economic Behavior",
    issn = "0899-8256",
    publisher = "Academic Press Inc.",
    number = "2",

    }

    TY - JOUR

    T1 - Bayesian learning in social networks

    AU - Gale, Douglas

    AU - Kariv, Shachar

    PY - 2003

    Y1 - 2003

    N2 - We extend the standard model of social learning in two ways. First, we introduce a social network and assume that agents can only observe the actions of agents to whom they are connected by this network. Secondly, we allow agents to choose a different action at each date. If the network satisfies a connectedness assumption, the initial diversity resulting from diverse private information is eventually replaced by uniformity of actions, though not necessarily of beliefs, in finite time with probability one. We look at particular networks to illustrate the impact of network architecture on speed of convergence and the optimality of absorbing states. Convergence is remarkably rapid, so that asymptotic results are a good approximation even in the medium run.

    AB - We extend the standard model of social learning in two ways. First, we introduce a social network and assume that agents can only observe the actions of agents to whom they are connected by this network. Secondly, we allow agents to choose a different action at each date. If the network satisfies a connectedness assumption, the initial diversity resulting from diverse private information is eventually replaced by uniformity of actions, though not necessarily of beliefs, in finite time with probability one. We look at particular networks to illustrate the impact of network architecture on speed of convergence and the optimality of absorbing states. Convergence is remarkably rapid, so that asymptotic results are a good approximation even in the medium run.

    KW - Herd behavior

    KW - Informational cascades

    KW - Networks

    KW - Social learning

    UR - http://www.scopus.com/inward/record.url?scp=0242654875&partnerID=8YFLogxK

    UR - http://www.scopus.com/inward/citedby.url?scp=0242654875&partnerID=8YFLogxK

    U2 - 10.1016/S0899-8256(03)00144-1

    DO - 10.1016/S0899-8256(03)00144-1

    M3 - Article

    VL - 45

    SP - 329

    EP - 346

    JO - Games and Economic Behavior

    JF - Games and Economic Behavior

    SN - 0899-8256

    IS - 2

    ER -