Division of the Humanities and Social Sciences California Institute of Technology Pasadena, California 91125 Detecting Failures of Backward Induction: Monitoring Information Search in Sequential Bargaining
نویسندگان
چکیده
We ran three-round sequential bargaining experiments in which the perfect equilibrium o er was $1.25 and an equal split was $2.50. Subjects o ered $2.11 to other subjects, $1.84 to \robot" players (who are known to play subgame perfectly), and $1.22 to robots after instruction in backward induction. Measures of information search showed that subjects did not look at the amounts being divided in di erent rounds in the correct order, and for the length of time, necessary for backward induction, unless they were specifically instructed. The results suggest that most of the departure from perfect equilibrium is due to limited computation and some is due to fairness. The nancial support of NSF 88-09299 and NSF 90-23531 to the rst two authors is gratefully acknowledged. We also thank Rachel Croson, Brian Becker, Barry Blecherman, Gerard Cachon, David Goldstein, Teck-Hua Ho, Keith Weigelt, Ben Wilner, and many colleagues at Penn for their work and ideas. We also have received helpful comments from several referees and seminar participants at many universities including Harvard, Cornell, New York University, Penn State, Rochester, Toronto, Minnesota, MIT, and the International Conference on Game Theory (Florence) and BoWo IV (Bonn). E-mail: [email protected].
منابع مشابه
Dynamic unstructured bargaining with private information and deadlines: theory and experiment
We study dynamic unstructured bargaining with deadlines and one-sided private information, via theory and experiment. We predict the incidence of bargaining failures (“strikes”) and payoffs in each state by combining mechanism design and focal point approaches. Strikes are common in states with lower surpluses (“pies”) and strike incidence is decreasing in the pie size. Subjects reach equal spl...
متن کاملDivision of the Humanities and Social Sciences California Institute of Technology Pasadena, California 91125 Sequential Entry in Many-to-one Matching Markets
We study sequential bargaining in many-to-one matching markets. We show that there is an advantage to entering late in the market, and that the last agent to enter the market will receive his or her best partner in a stable matching, extending the results of Blum and Rothblum (2002) and Cechlárová (2002) for the marriage model. We also discuss the relation between sequential bargaining and a po...
متن کاملNber Working Paper Series Experimenting with Measurement Error: Techniques with Applications to the Caltech Cohort Study
Measurement error is ubiquitous in experimental work. It leads to imperfect statistical controls, attenuated estimated effects of elicited behaviors, and biased correlations between characteristics. We develop simple statistical techniques for dealing with experimental measurement error. These techniques are applied to data from the Caltech Cohort Study, which conducts repeated incentivized sur...
متن کاملDivision of the Humanities and Social Sciences California Institute of Technology Pasadena, California 91125 on Incentives and Updating in Agent Based Models
متن کامل
Division of the Humanities and Social Sciences California Institute of Technology Pasadena, California 91125 Finite Perfect Information Extensive Games with Generic Payoffs
In nite perfect information extensive (FPIE) games, backward induction (BI) gives rise to all pure-strategy subgame perfect Nash equilibria, and iterative elimination of weakly dominated strategies (IEWDS) may give di erent outcomes for di erent orders of elimination. Duggan recently posed several conjectures in an e ort to better understand the relationship between BI and IEWDS in FPIE games. ...
متن کامل