Strategic ignorance and discrimination

How to cite

Thunstrom, Linda. 2020. "Strategic ignorance and discrimination." AEA RCT Registry. AEARCTR-0005307. https://doi.org/10.1257/rct.5307-1.0

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

Linda Thunstrom*

* Primary investigator

Registry ID
AEARCTR-0005307
DOI
https://doi.org/10.1257/rct.5307-1.0
First registered
Last updated
Status
In development
Trial dates
2020-01-25 – 2020-01-31
Countries
Private
Registry record
http://www.socialscienceregistry.org/trials/5307

Abstract

This trial serves to complement results from previously collected data (not pre-registered). We examine the following: Do people have a taste for discrimination? We answer this question in a dictator game, allowing for Dictators to be strategically ignorant of the Recipients' payoffs (see Dana et al., 2007). Recipients are named either James or Mohammed -- names particularly associated with "Caucasian Christian" and "Muslim." We explore if people use the moral wiggle room created by strategic ignorance to discriminate against Mohammed.

Intervention

Participants will participate in a dictator game, of the kind developed in Dana et al. (2007), although with smaller amounts. Dictators will be recruited on Amazon Mechanical Turk. They will be randomized into one of two treatments: "treatment Mohammed" and "treatment James." If in treatment "Mohammed," they will learn that the Recipient in the game is called Mohammed, and if in treatment "James," they learn the Recipient is called James. They are thereafter offered the opportunity to learn or remain ignorant of the version of the game they are playing together with the Recipient -- they can focus only on their own payoff from the game (by choosing to remain ignorant of the Recipient payoff), or they can focus on both their own and the Recipient's payoff (by choosing to learn the version of the game they are playing).

Experimental design

Participants will participate in a dictator game, of the kind developed in Dana et al. (2007), although with smaller amounts. Dictators will be recruited on Amazon Mechanical Turk. They will be randomized into one of two treatments: "treatment Mohammed" and "treatment James." If in treatment "Mohammed," they will learn that the Recipient in the game is called Mohammed, and if in treatment "James," they learn the Recipient is called James. They are thereafter offered the opportunity to learn or remain ignorant of the version of the game they are playing together with the Recipient -- they can focus only on their own payoff from the game (by choosing to remain ignorant of the Recipient payoff), or they can focus on both their own and the Recipient's payoff (by choosing to learn the version of the game they are playing).

Design details

Randomization method
Randomization by computer software.
Randomization unit
Individual.
Clusters
0
Observations
800
Arms
400
Primary outcomes
We are interested in (i) the extent to which Dictators chose to ignore Recipient payoff across the two treatments, and (ii) overall distributions from Dictators to "Mohammed" versus "James."

Cited by

No citation data available for this registration yet.

Keywords