Round-Number Effects in Bargaining: Bias vs. Focal Point

How to cite

Lauf, Alexander; Schlereth, Lars. 2020. "Round-Number Effects in Bargaining: Bias vs. Focal Point." AEA RCT Registry. AEARCTR-0006823. https://doi.org/10.1257/rct.6823-1.1

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

Alexander Lauf* , Lars Schlereth

* Primary investigator

Registry ID
AEARCTR-0006823
DOI
https://doi.org/10.1257/rct.6823-1.1
First registered
Last updated
Status
Completed
Trial dates
2020-07-22 – 2021-01-15
Countries
United States of America ()
Registry record
http://www.socialscienceregistry.org/trials/6823

Abstract

It is well-established that round numbers play an important role in bargaining situations. In a first step, using data from Backus, Blake and Tadelis (2019) we show that a large fraction of successful negotiations end with a round price, and that round final prices are associated with quicker agreement. These effects might be driven by a simple preference for round numbers or by round numbers serving as focal points. In a second step, we design a theory-guided experiment to disentangle these explanations. To this end, we plan to run two treatments on Amazon MTurk with roughly 1000 participants. In treatment One, a single decision-maker might be affected by round-number bias. Treatment Two extends the setting in a one-change-at-a-time fashion to two-player bargaining where additionally round numbers as focal points might facilitate coordination. Comparing these two treatments allows to differentiate between the two explanations.

Experimental design

We deploy a multi-period bargaining game with a shrinking-cake design. There are two treatment groups. Each participant makes a total of 10 decisions to accept or reject a random offer.

Design details

Randomization method
The recruited participants are randomly assigned a treatment by their time of arrival. We randomly draw offer sequences for groups of 4 participants. One group consists of two participants from each treatment.
Randomization unit
Each participant will randomly be assigned to one of the 2 treatments. A group of 4 randomly partnered participants consisting of two participants from each treatment see the same offer sequence.
Clusters
250 (groups with same sequence) and 500 (per treatment)
Observations
We plan to recruit 1000 participants.
Arms
1000 participants in 250 groups of 4.
Primary outcomes
We measure the number of accepted offers per treatments and by the offer type.

Cited by

No citation data available for this registration yet.

Keywords