Doctoral Dissertation Research in Economics: Regret in Games: When It Is Not (Only) Your Fault

New York University, New York NY

Investigators

Abstract

This project lies at the interface of economics and psychology. It will examine how feelings of regret and blame shape human behavior in situations where people interact strategically so that each person’s welfare also depends on the actions of others. It poses that people regret their actions when they feel responsible for a bad outcome. On the other hand, they blame other people when they believe them to be responsible for the bad outcome. This project will examine how this tension between regret and blame affects behavior and economic outcomes. The project will (i) emphasize the importance of accounting for emotions in economic models, (ii) enhance our understanding of environments where reduced cooperation and inefficient outcomes may arise due to division of responsibility, and (iii) produce a previously nonexistent type of data that other researchers in economics and psychology can use. This proposed work will examine regret in games. It will propose a theory where strategically-interacting agents make choices taking into account both material payoffs and the regret that their choices may generate. This theory accounts for how the magnitude of a player’s (anticipated) regret (for her action in a game and her resulting payoff) is affected by the division of responsibility between the players for the outcome of the game. This is the first work to consider a strategic form of regret, since earlier works have analyzed regret as if in a single-agent context with the opponents’ actions treated as the state of the world. The intuition is as follows. In a decision-theoretic context an outcome is exclusively a result of the decision-maker’s (DM) choice and ”luck” (the at first unknown state of the world). Feelings of regret can naturally arise given the power the DM has over the outcome. On the other hand, the outcome of a game is the result of the strategic interaction of multiple agents. Thus, it is postulated that a player may not (expect to) experience feelings of regret to the same degree, as she may feel less responsible for the combined result of all the players’ actions. Specifically, it is posed that a player’s regret tends to arise when her chosen action is proven sub-optimal after the other players’ actions are revealed. However, her regret is mitigated if it is also revealed that the other players could have chosen differently (making everyone better off), as part of the blame is attributed to them. The project will examine both theoretically and experimentally how such regret and blame considerations affect economic outcomes and whether accounting for these consideration leads to theoretical predictions closer to experimental findings. The project will have three main benefits. First, it will contribute to our understanding of the circumstances under which attribution of responsibility and blame can affect outcomes of strategic interactions. It will also help us understand whether this effect will be welfare enhancing or damaging, and gauge the magnitude of the effect. Second, it will inform theoretical work by emphasizing the gains from modifications to models of choice when those are applied in games. Last, it will produce experimental data on regret and blame in games that other researchers can use. This award reflects NSF's statutory mission and has been deemed worthy of support through evaluation using the Foundation's intellectual merit and broader impacts review criteria.