Talent as luck? An experimental study on inequality acceptance

How to cite

Skarpeid, Ingvild Lindgren. 2017. "Talent as luck? An experimental study on inequality acceptance." AEA RCT Registry. AEARCTR-0002597. https://doi.org/10.1257/rct.2597-2.0

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

Ingvild Lindgren Skarpeid*

* Primary investigator

Registry ID
AEARCTR-0002597
DOI
https://doi.org/10.1257/rct.2597-2.0
First registered
Last updated
Status
On going
Trial dates
2017-11-20 – 2017-11-30
Countries
Private
JEL codes
C9, D01, D9
Secondary IDs
Luck, talent, inequality, fairness
Registry record
http://www.socialscienceregistry.org/trials/2597

Abstract

People are highly skeptical towards inequalities that arise due to pure luck, such as the toss of a coin. In contrast, they largely accept inequalities stemming from talent differences. However, talent may be fundamentally understood as coming about from a lucky draw in the ``genetic lottery." The question arises why inequalities from talent are accepted even when an individual has done nothing to develop her talent. We investigate two potentially morally relevant distinctions between a ``talent" lottery and a pure luck lottery. The first is whether the indivisibility of talent-luck from its owner implies that it is a personal trait with a legitimate claim on the outcome. The second is whether it is the act of exerting effort on a lucky draw that makes talent a more legitimate source of inequalities.

Experimental design

We run a real effort dictator game with a spectator design. The experiment has two separate parts, where the first is a real-effort task run in an online labour market by workers. We then recruit a spectator who makes a dictator decision (implemented with ten percent probability, as the ratio of workers to spectators is one to ten) and thus may have real monetary consequences for two of the workers who have completed the real effort assignment.

The workers in Part 1 of the experiment will be recruited from the international online market place Amazon Mechanical Turk (mTurk). We will recruit 400 workers to obtain 200 unique pairs of workers. In Part 2, the recruitment of spectators will be done by Research Now, using Qualtrics as survey provider. We will recruit 2000 US participants who are nationally representative (above 18 years old) on observable characteristics (gender, age and geography).

Design details

Randomization method
Survey randomization into treatment, at the individual level.
Randomization unit
Individual level
Clusters
2000 unique spectators from a representative sample of US citizens
Observations
2000 unique spectators from a representative sample of US citizens
Arms
We have four treatment arms, where we vary both the personal/impersonal nature of the lottery as well as the timing of the lottery. The sample sizes for each of the treatment arms are the following : 500 spectators personal ex ante, 500 spectators impersonal ex ante, 500 spectators personal ex post, 500 spectators impersonal ex post.
Primary outcomes
The main outcome variable of interest is the experimentally implemented inequality (worker-pair gini) by the spectator.

Cited by

No citation data available for this registration yet.

Keywords