Fairness in Winner-Take-All Markets: A Representative Study

How to cite

Ekström, Mathias; Sørensen, Erik Ø.; Cappelen, Alexander W.; Bartling, Björn; Tungodden, Bertil. 2018. "Fairness in Winner-Take-All Markets: A Representative Study." AEA RCT Registry. AEARCTR-0003155. https://doi.org/10.1257/rct.3155-1.0

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

Mathias Ekström* , Erik Ø. Sørensen , Alexander W. Cappelen , Björn Bartling , Bertil Tungodden

* Primary investigator

Registry ID
AEARCTR-0003155
DOI
https://doi.org/10.1257/rct.3155-1.0
First registered
Last updated
Status
On going
Trial dates
2018-07-04 – 2018-12-31
Countries
Private
JEL codes
C91: D63
Registry record
http://www.socialscienceregistry.org/trials/3155

Abstract

This trial is a replication study, based on a representative US sample, of an earlier study published in a working paper that was based on a student sample (see SSRN-link). For details of the experimental design we refer the reader to the working paper, here we only report the main differences. The purpose of the replication study is to generate results based on a more policy relevant sample of subjects --- a nationally representative sample (in terms of gender, age and region). The results from the new data collection will be added to the paper and presented in the same set of figures and tables. We have no particular prior in terms of whether the treatment effects will be larger or smaller relative to the results from the student sample in the working paper, also because the student population and the representative population are recruited from different societies (Norway and the US). We therefore view the present study more broadly as providing evidence of the generality of the initial findings reported in the working paper. In addition we will use the larger representative data set to explore heterogenous treatment effects with respect to political orientation and views on income inequality.

Intervention

The treatments in this study are the same four treatments described in the working paper (WTA, WTA - No Choice, WTA - No Expectations, and Base).

Experimental design

The experimental design closely follows the design reported in the working paper. The three main differences are: i) the spectators in the new study are recruited via Research Now / SSI and should be representative of the US population in terms of gender, age and four broad regions of the country while the initial study recruited a student sample from Norway, ii) we aim to recruit 4,000 spectators in the new study, while we had 349 spectators in the initial study, and (iii) we aim to have a one-to-five matching of spectators in the new study (every fifth spectator decision will be randomly selected and implemented), while we had one-to-one matching in the initial study (every spectator decision was implemented).

Design details

Randomization method
Done via computer program
Randomization unit
Individuals
Clusters
0
Observations
4,000 individuals
Arms
1,000 individuals per treatment
Primary outcomes
Same two outcome as in the working paper: 1) Share given to the winner, 2) Dummy variable equal to one if all earnings given to winner.

Cited by

No citation data available for this registration yet.

Keywords