Kyoungwon Seo


Working papers
  • Ambiguity and Second-Order Belief, Revised November 2007, Job Market Paper, conditionally accepted in Econometrica

    • The Ellsberg Paradox has raised questions about the subjective expected utility model. In one version of the paradox, there are two urns, each containing 100 balls that are either red or black. Subjects know that urn A contains 50 red (r) balls and 50 black (b) balls, but nothing is known about urn B. A ball is randomly drawn from each urn. The bet Ar gives $100 if the ball drawn from urn A is red and nothing otherwise. The bets Ab, Br and Bb are defined similarly. Most subjects report the following preference ranking : Ar=Ab>Br=Bb. Subjective Expected Utility (SEU) cannot accommodate this behavior.

    One way to rationalize the above rankings is to assume that the agent has in mind a second-order belief, or a probability measure on probability measures (a belief over "the number of red balls" in the above example). Existing models of preference involving a second-order belief presume that the analyst can observe more than just the ranking of bets on the color of the drawn ball - the ranking of bets on the true probability (bets on the composition of urn B in the above example) must also be observable.

    The importance of the domain assumption can be illustrated in the context of an asset market. Consider a simple model where the asset price may go up (H) or go down (L). In this setting, a bet on H corresponds to buying the asset and a bet on L to selling the asset - decisions that are observed in many data sets. On the other hand, we do not observe bets on the true probability that the price goes up - for example, the payoffs of real-world securities depend on realizations of prices, and not separately on the mechanism that generates these realizations.

    This paper characterizes preferences having a second-order belief, adopting a domain consisting of lotteries over acts defined over a basic state space. This domain is close to the set of choices involved in the Ellsberg Paradox and thus provides foundations that are testable in principle. Besides, the domain in this paper is the same as that in Anscombe and Aumann (1963), one of the classic papers on SEU.

  • Subjective States: a More Robust Model, with Larry Epstein, Revised September 2007

    • Following Kreps (1979), Nehring (1996, 1999) and Dekel, Lipman and Rustichini (2001, DLR), we study the demand for flexibility and what it reveals about subjective uncertainty. These papers model choice under uncertainty without positing a Savage-style primitive state space. We interpret them as addressing the question: when can we view an agent choosing an action under uncertainty as though she foresees a set of possible contingencies or a subjective state space?

    DLR obtain a unique subjective state space by assuming that the agent satisfies alternative independence-style axioms. We argue that these axioms preclude ambiguity and thus capture too much - they not only justify thinking of the agent as using a set of subjective states, but they also compel us to view her as being completely confident about their likelihoods.

    Our principal goal is to identify those properties of (suitably defined) preference that justify the noted "as if" view of the agent, without otherwise restricting preference unduly, and while also permitting identification of a unique subjective state space.


    Publication
  • Coarse contingencies and ambiguity, Theoretical Economics 2007, with Larry Epstein and Massimo Marinacci

    • Consider an agent who must choose an action today under uncertainty about the consequence of any chosen action but without having in mind a complete list of all the contingencies that could influence outcomes. She conceives of some relevant contingencies or states of the world but she is aware that these contingencies are coarse - they leave out some details that may affect outcomes. Though she may not be able to describe these finer details, she is aware that they exist and this may affect her behavior. How does one model such an agent?

    In this paper, we describe axioms for preference that capture coarse perceptions and that characterize functional forms for utility representing preference. We propose, and our model formalizes, that the impossibility of fully describing all relevant contingencies is one reason, an important one in our view, why decision makers may not be able to quantify uncertainty about future payoffs with a single probability measure.


    CV

    Department of Economics
    University of Rochester
    email : kseo at troi.cc.rochester.edu