Cheating with Externalities and a Regular Audience

How to cite

Chen, Siyu; Dufwenberg, Martin. 2024. "Cheating with Externalities and a Regular Audience." AEA RCT Registry. AEARCTR-0013366. https://doi.org/10.1257/rct.13366-1.0

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

Siyu Chen* , Martin Dufwenberg

* Primary investigator

Registry ID
AEARCTR-0013366
DOI
https://doi.org/10.1257/rct.13366-1.0
First registered
Last updated
Status
In development
Trial dates
2024-04-16 – 2024-07-31
Countries
Private
Registry record
http://www.socialscienceregistry.org/trials/13366

Abstract

Reports, for example, tax returns, research findings, or peer evaluations, can in some instances be falsified. We develop an experiment to study how externalities affect an individual’s self-reporting behaviors. We modify the cheating game (Fischbacher & Föllmi-Heusi, 2013) by matching a reporter with a peer subject and manipulate the externality. Specifically, the reporter privately rolls a die and then issues a report on the outcome, which determines the earnings of both the reporter and the audience. Another design arm compares treatments with and without the experimenter observing the die-rolls. In the existing literature on the cheating game, the experimenter is typically assumed to serve as a relevant audience, which contrasts with the perceived role of the experimenter in most other laboratory experiments. We aim to explore whether the observability of die-rolls by the experimenter affects the decision-maker’s behavior when a regular audience is present.

Experimental design

The primary objective of this experiment is to investigate whether externalities influence individuals' self-reporting behaviors, with a secondary focus on assessing the impact of experimenter observability on these behaviors. We adapt the die-rolling game originally developed by Fischbacher & Föllmi-Heusi (2013) by incorporating a regular audience into the setup. We employ a 3 × 2 between-subjects design, varying the type of externality imposed on the audience by the reporter (negative, zero, or positive) and the visibility of the initial draw to the experimenter (non-observed vs. observed).

Design details

Randomization method
Approximately 216 undergraduate and graduate students from the University of Arizona will be randomly assigned to six treatments, irrespective of gender, ethnicity, or any demographic characteristics. Each treatment will include three sessions, with 12 participants per session. The randomization of participants will be handled by the online recruitment system at the Economic Science Laboratory (ESL) of the University of Arizona.
Randomization unit
Individual
Clusters
216 students
Observations
6 treatments*36 participants/treatment = 216 participants
Arms
36 participants in treatment Negative Externality -- Non Observed, 36 participants in treatment Zero Externality -- Non Observed, 36 participants in treatment Positive Externality -- Non Observed, 36 participants in treatment Negative Externality -- Observed, 36 participants in treatment Zero Externality -- Observed, 36 participants in treatment Positive Externality -- Observed
Primary outcomes
This paper aims to explore whether externalities and experimenter observability influence individuals' decisions regarding their reporting behavior.

Cited by

No citation data available for this registration yet.

Keywords