Reference Dependence ad Games

Abstract

Pareto efficiency is a central notion in welfare analysis. In games, players have to coordinate their strategies towards Pareto efficient outcomes. However, coordination failure leads to Pareto inefficiency. The loss of welfare motivates policy interventions to facilitate coordination. In the thesis, I extend behavioural biases known in the individual choices to games to explore the potential use of the biases as a policy intervention and their relevance. In chapters 4 and 5, I extend the asymmetric dominance (AD) effect to games; and experimentally examine the relevance of the bias. This effect predicts that the addition of a dominated strategy increases the comparative attractiveness of the corresponding dominating strategy. I focus on the Battle of Sexes game which embeds the concept of coordination and an extension of the game to include a dominated strategy (AD-strategy) for one player (a row player). Further, I investigate if the AD-effect in risky choices predicts the effect in games at an individual level. To this aim, I complement the game experiments with a standard individual choice under risk. The finding of Chapter 4 indicates that AD-strategies affects the row subjects’ through the column subjects’ payoff in AD-strategies. Chapter 5, incorporates this finding into the experimental design to isolate the AD-effect to row players payoffs. Results show that 19% of the row subjects alter their strategy choices as predicted by the AD-effect, while 16% change their choice in the opposite direction. The column subjects appear to anticipate the row subjects’ choices and coordinate following the row subjects’ choices. This enhances both players welfare. Moreover, the results show that the bias is present in risky choices. Nonetheless, within-subjects, there is no correlation between the effect between games and risky choices, confirming a context effect. The next experiment (Chapter 6) provides evidence that biases attributed to the perception of probabilities affect cooperation levels in repeated games. Subjects completed a prisoner’s dilemma game that continues with a fixed probability. Under the standard assumption of (constant) discounted expected utility, such a probability can be interpreted as time discounting. The presence of probability biases leads to deviations from constant discounting, as shown in Halevy (2008, AER). The hypothesis is that this affects the cooperate-defect decision of subjects. Using an incentive-compatible mechanism based on scoring rules, I quantify the subjects’ probability biases. I find that 53% of subjects are expected utility (EU) subjects; 21% of subjects reveal biases that accord with prospect theory: small probabilities are overweighed and medium to large ones are underweighted (inverse-S); 26% of subjects underestimate small probabilities and overestimate medium to large ones (S-shape). I find that the cooperation level among EU subjects is approximately 50% for all continuation probabilities. Inverse-S subjects cooperate 19:6 percentage points more than EU subjects, and S-shape subjects cooperate 27:6 percentage points less than EU subjects, for all continuation probabilities. I explain this behaviour by adopting Halevy’s impatience index. Keywords (Chapters 4 and 5): Asymmetric dominance effect, coordination, preference reversal; (Chapter 6) Cooperation, impatience, repeated games, time preferences.

Date of Award	31 Dec 2020
Original language	English
Awarding Institution	The University of Manchester
Supervisor	Horst Zank (Supervisor)