nep-dcm New Economics Papers
on Discrete Choice Models
Issue of 2007‒01‒28
four papers chosen by
Philip Yu
Hong Kong University

  1. Sincere Voting with Cardinal Preferences: Approval Voting By Miguel �gel Ballester; Pedro Rey-Biel
  2. Dynamic Choice, Independence, and Emotions By Hopfensitz, Astrid; van Winden, Frans A.A.M.
  3. Understanding individuals’ decisions about vaccination: a comparison between Expected Utility and Regret Theory models By M Zia Sadique; John Edmunds; Nancy Devlin; David Parkin
  4. Observational Learning in the Motion Picture Market By Santugini, Marc

  1. By: Miguel �gel Ballester; Pedro Rey-Biel
    Abstract: We discuss sincere voting when voters have cardinal preferences over alter- natives. We interpret sincerity as opposed to strategic voting, and thus define sincerity as the optimal behaviour when conditions to vote strategically vanish. When voting mechanisms allow for only one message type we show that this op- timal behaviour coincides with an intuitive and common definition of sincerity. For voting mechanisms allowing for multiple message types, such as approval vot- ing (AV), there exists no conclusive definition of sincerity in the literature. We show that for AV, voters' optimal strategy tends to one of the existent definitions of sincerity, consisting in voting for those alternatives that yield more than the average of cardinal utilities.
    Keywords: sincere and strategic voting, approval voting
    JEL: D63 D71 D72 D80
    Date: 2007–01–15
  2. By: Hopfensitz, Astrid; van Winden, Frans A.A.M.
    Abstract: From the viewpoint of the independence axiom of expected utility theory, an interesting empirical dynamic choice problem involves the presence of a 'global risk', that is, a chance of losing everything whichever safe or risky option is chosen. In this experimental study, participants have to allocate real money between a safe and a risky project. Treatment variable is the particular decision stage at which a global risk is resolved: (i) before the investment decision; (ii) after the investment decision but before the resolution of the investment risk; (iii) after the resolution of the investment risk. The baseline treatment is without global risk. Our goal is to investigate the isolation effect and the principle of timing independence under the different timing options of the global risk. In addition, we examine the role played by anticipated and experienced emotions in the choice problem. Main findings are a violation of the isolation effect, and support for the principle of timing independence. Although behaviour across the different global risk cases is approximately the same, we observe clear differences in people's affective responses. This may be responsible for the conflicting results observed in earlier experiments. Dependent on the timing of the global risk different combinations of anticipated and experienced emotions influence decision making.
    Keywords: anxiety; background risk; emotions; global risk; investment; laboratory experiment; regret
    JEL: A12 C91 D81
    Date: 2007–01
  3. By: M Zia Sadique (Department of Economics, City University, London); John Edmunds (Health Protection Agency, Centre for Infections, London); Nancy Devlin (Department of Economics, City University, London); David Parkin (Department of Economics, City University, London)
    Abstract: This paper proposes two new theoretical models for examining individual decision-making regarding vaccination. In each case, individuals’ decisions are modelled as a binary choice (i.e. to accept or to reject an invitation to receive vaccination) which are a product both of the perceived risk of the preventable disease in question and of the perceived risk of adverse side effects of the vaccine itself. Individuals decisions are modelled in two ways – first, as expected utility maximising and second, as regret minimising – and the results compared. In both cases, the decision to vaccinate is explained by a threshold condition with respect to the risk of remaining exposed to the disease by rejecting vaccination, and the risk of experiencing adverse events from vaccination itself. Regret-averse individuals have a higher threshold – suggesting a lower propensity to vaccinate than that suggested by the expected utility models. Although the results are intuitively plausible, they rest on assumptions about the perceived severity of side effects as opposed to preventable disease. We conclude by identifying a number of theoretical issues that remain to be explored, and outlining the empirical research required.
    Date: 2005–07
  4. By: Santugini, Marc
    Abstract: I focus on the market for movies released in the theater to measure the extent to which consumers learn about the quality of a movie from observing its market share in the release week. I derive the demand for movies using a dynamic discrete choice model in which consumers are endowed with private information about a movie and engage in as well as anticipate learning. I also assume that consumers watch a movie at most once to account for demand saturation. I depart from previous applied work on estimating the demand for movies by incorporating forward-looking behavior and observational learning. I also propose a new approach to account for demand saturation. The approach allows the decay rate for a market share to depend on consumers' past decisions, past movie competition, as well as past observational learning. The decay rate also depends on consumers' anticipation of observational learning and future schedule of movies in the theater. Given the distributional assumptions, the corresponding market shares for movies are mixing multinomial logit probabilities, taking into account consumers' forward-looking behavior and heterogeneity due, in part, to their private information. Using US market-level data, I estimate the structural parameters of demand via the maximum simulated likelihood procedure. I recover reasonable estimates for the covariates such as MPAA ratings and studio indicators. I also find evidence of observational learning. I plan to measure the effect of observational learning on demand saturation as well as the effect of anticipation of observational learning on delaying consumers from watching movies in their release weeks.
    Keywords: Motion pictures; Information; Learning; Market saturation
    JEL: L15 L82 D83 D12 C13
    Date: 2006–11–08

This nep-dcm issue is ©2007 by Philip Yu. It is provided as is without any express or implied warranty. It may be freely redistributed in whole or in part for any purpose. If distributed in part, please include this notice.
General information on the NEP project can be found at For comments please write to the director of NEP, Marco Novarese at <>. Put “NEP” in the subject, otherwise your mail may be rejected.
NEP’s infrastructure is sponsored by the School of Economics and Finance of Massey University in New Zealand.