Sources of Discrimination - The Effect of Information

How to cite

Eyting, Markus. 2021. "Sources of Discrimination - The Effect of Information." AEA RCT Registry. AEARCTR-0008609. https://doi.org/10.1257/rct.8609-4.0

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

Markus Eyting*

* Primary investigator

Registry ID
AEARCTR-0008609
DOI
https://doi.org/10.1257/rct.8609-4.0
First registered
Last updated
Status
In development
Trial dates
2021-12-05 – 2021-12-31
Countries
Private
Registry record
http://www.socialscienceregistry.org/trials/8609

Abstract

While the existence of discrimination has long been established, little is known about the actual sources of discrimination. Researchers typically categorize discrimination as either taste-based or belief-based discrimination based on either accurate or inaccurate beliefs. While any kind of discrimination can have fatal consequences for the discriminated, a precise identification of the source of discrimination has important implications for effective policy interventions. I suggest a careful analysis of the information acquisition and processing stage that precedes potentially discriminatory decisions. In particular, imitating a hiring decision in an online experiment, I analyze the effect of biased information acquisition and processing on discriminatory behavior and investigate whether people selectively attend and adhere to different kinds of information.

Experimental design

I set up a hiring situation in which 'employers' are repeatedly asked to select one of two potential 'workers'.

In a 5X2 design, employers are randomly allocated into one of four information groups and within each information group into either the treatment groups "neutral" and "real". In group 'neutral' any information about whether or not the workers belong to a minority group is hidden, while in the 'real' group, the ethnicity of each worker is displayed.

The information groups:
"Oneplusinfo": In each hiring decision, employers first receive a randomly drawn piece of information about the two workers. They may then choose to gather additional information signals before hiring one of the two workers.
"Oneinfoonly": In each hiring decision, employers first receive a randomly drawn piece of information about the two workers. They do not have the opportunity to gather additional information signals before hiring one of the two workers.
"Allinfo": In each hiring decision, employers receive several pieces of information about the two workers. They do not have the opportunity to gather additional information signals before hiring one of the two workers.
"Noinfo": In each hiring decision, employers do not receive any pieces of information about the two workers.
"Signals": In each hiring decision, employers receive the name of the better of the two workers. However, this information comes from either a True News or a Fake News source. The True News source always tells the truth, whereas the Fake News source never tells the truth. For each decisions, employers can request up to 9 additional signals, each from either the True News or the Fake News source. Employers do not know from which source a message is delivered.

Analyses are mainly within information groups, and between treatment groups.

Design details

Randomization method
Randomization into one of the 10 (5 X 2) groups is done by a computer
Randomization unit
individual
Clusters
~2500 individuals
Observations
~ 50,000 hiring decisions
Arms
~ 400 individuals in "Allinfo" and "Oneinfo" groups, ~ 600 individuals in "Oneplusinfo" and "Noinfo" groups, ~ 500 in "Signals" group
Primary outcomes
"Discrimination", "information acquisition" and "information processing"
Secondary outcomes
Decision times Beliefs

Cited by

No citation data available for this registration yet.

Keywords