Leaving it to Chance: The Effect of a Coin Flip on Happiness and Guilt in Other-Regarding Decisions

How to cite

Goff, Sandra. 2018. "Leaving it to Chance: The Effect of a Coin Flip on Happiness and Guilt in Other-Regarding Decisions." AEA RCT Registry. AEARCTR-0002971. https://doi.org/10.1257/rct.2971-1.0

Download: BibTeX · RIS — for reference managers without a browser translator. Zotero and most others pick up this page's citation metadata directly.

Sandra Goff*

* Primary investigator

Registry ID
AEARCTR-0002971
DOI
https://doi.org/10.1257/rct.2971-1.0
First registered
Last updated
Status
In development
Trial dates
2018-05-09 – 2019-05-04
Countries
Private
JEL codes
C70, C91, D91
Registry record
http://www.socialscienceregistry.org/trials/2971

Abstract

Research on repeated choosing and choice complexity has demonstrated that decision-making by individuals uses up scarce psychological resources and can affect both decision outcome and an individual’s satisfaction with their choice. Given this, people will sometimes avoid deciding, opting instead to use an external device such as a coin flip (Dwenger 2014). Levitt (2016) argues that when people are facing big life decisions and find themselves at an impasse, letting a coin decide can increase happiness. We assert that responsibility aversion plays a role in this outcome and design an experiment to assess this effect of the use of external decision devices. Participants engage in a dictator game with a “take” framing with two potential choices for the dictator: (1) take $7 from the other player’s endowment, leaving the partner with only $3, or (2) take $3, leaving the other player with $7. We hypothesize that those who are directed by a coin flip to take $7 will score lower on a measure of negative affect and higher on a measure of positive affect than those who made the same decision of their own accord. We also assert that positive self-assessment will be greater in the coin flip condition and that this will free these participants from the need to engage in subsequent moral cleansing. Finally, we test to determine whether people are willing to pay a non-zero amount to transfer the decision-making power to the coin flip in anticipation of the avoidance of guilt feelings.

Experimental design

The study is designed as an online between-subjects experiment. Participants are randomly assigned to one of three treatments: (1) free choice (no coin flip), (2) free coin flip (virtual coin is flipped on their behalf, but they are not restricted by the result of the flip), and (3) pay-to-flip (subject can choose to flip a virtual coin for a $1 fee, but they are not restricted by the result of the flip).
Each participant performs three tasks, a Dictator Game task, an Ultimatum Game task, and a Survey task. These three tasks are followed by a set of demographic questions. Participants are informed that they will be paid for only one of the tasks, that the task for which they will be paid will be selected at random, and that the average task payoff is $5. Participants are also paid a $2 completion fee. The treatment to which a participant is randomly assigned affects only the Dictator Game task as described below in the Task Descriptions section.
The Dictator Game task is the target task. The Ultimatum Game task is used as a comparison and to assess participant’s attitudes towards norms of fairness. The survey task is unrelated to the current study. Rather, this task is included because the Dictator Game and Ultimatum Game tasks are short and the compensation participants receive is substantial given the time needed to complete the work. This provides a great opportunity to collect preliminary data on a subject of interest to the PI, political consumerism. The order in which participants receive the Dictator Game and Ultimatum Game are randomized. The survey is always received as the third task to ensure that the questions do not interfere with the target task.

Design details

Randomization method
Randomization is performed automatically by the survey software.
Randomization unit
Individual
Clusters
NA
Observations
452 participants
Arms
176 free-flip, 176 no flip, 100 pay-to-flip
Primary outcomes
PANAS (Both positive and negative affect), Moral identity (Aquino & Reed, 2002), pro-social behaviors (donation, volunteer, consumption)

Cited by

No citation data available for this registration yet.

Keywords