Competitive experimentation with private information: The survivor's curse

https://doi.org/10.1016/j.jet.2009.12.001Get rights and content

Abstract

We study a winner-take-all R&D race between two firms that are privately informed about the arrival rate of an invention. Over time, each firm only observes whether the opponent left the race or not. The equilibrium displays a strong herding effect, that we call a ‘survivor's curse.’ Unlike in the case of symmetric information, the two firms may quit the race (nearly) simultaneously even when their costs and benefits for research differ significantly.

References (23)

  • F. Gul et al.

    Endogenous timing and the clustering of agents' decisions

    J. Polit. Economy

    (1995)
  • Cited by (36)

    • Hiding and herding in market entry

      2022, Journal of Economic Theory
      Citation Excerpt :

      We use the continuous-time modeling approach developed in that literature; see, for example, Décamps and Mariotti (2004), Wagner (2018) and Wagner and Klein (2022). Our paper is further related to the literature on learning in R&D races, where firms gradually learn the rate of success of innovations and decide whether to drop out of the patent race; see Choi (1991), Malueg and Tsutsui (1997), Moscarini and Squintani (2010), Akcigit and Liu (2016) and Halac et al. (2017). Our model differs from the existing literature on social learning in common values environments and R&D contest in two dimensions.

    • Strategic investment and learning with private information

      2022, Journal of Economic Theory
      Citation Excerpt :

      A number of papers consider learning from others' actions and outcomes in models with additional application-specific payoff externalities. Moscarini and Squintani (2010) considers a model of a winner-takes-all R&D competition in which firms observe an initial private signal and decide when to exit irreversibly. They show that the aggregate duration of experimentation is longer under private information.

    • Inefficiency of sponsored research

      2021, Journal of Mathematical Economics
      Citation Excerpt :

      Interestingly, in the interval between the cutoff barrier for a winner-takes-it-all case and the one for the case when the laggard is rewarded, my model predicts a herding, or a “survivor’s curse” effect similar to the one in Moscarini and Squintani (2010). The driving force behind the “survivor’s curse” in Moscarini and Squintani (2010) is private information about the arrival rates of breakthrough. In my model, the “survivor’s curse” is due to a potential positive reward to the laggard.

    • Dynamic R&D competition under uncertainty and strategic disclosure

      2021, Journal of Economic Behavior and Organization
      Citation Excerpt :

      In comparison to these works, disclosure in our model bears both direct effects (the immediate reward and technological spillover) as well as indirect informational effects (informing the opponent about difficulty of early stage), which not only reflect a natural trade-off in the choice of when to disclose or conceal a success, but also lead to interesting results on further topics such as welfare and equilibrium behavior under firm heterogeneity. Other approaches for modeling payoff-related uncertainty in R&D competitions can be found in, for instance, Malueg and Tsutsui (1996), Chatterjee and Evans (2004) and Moscarini and Squintani (2010). A recent development in this field is Bobtcheff et al. (2017).

    • Undiscounted bandit games

      2020, Games and Economic Behavior
    View all citing articles on Scopus

    We thank Michele Boldrin, Chris Harris, Heidrun Hoppe, Paul Klemperer, David Levine, Tracy Lewis, Glenn Mac Donald, Meg Meyer, Mike Peters, Jennifer Reinganum, Jeroen Swinkels, Juuso Valimaki, and seminar audiences at Boston University, UC San Diego, Duke Fuqua Business School, Nuffield College University of Oxford, University of Bonn, University of Munich, Econometric Society Summer Meetings (2003), Society for Economic Dynamics (2003), ESSET at Gerzensee (2004), for valuable comments. The usual disclaimer applies.

    1

    Fax: +1 (203) 436 2626.

    View full text