Reference points in sequential bargaining

How to cite

Helland, Leif; Ciccone, Alice; Brekke, Kjell Arne; Heggedal, Tom-Reiel. 2020. "Reference points in sequential bargaining." AEA RCT Registry. AEARCTR-0006446. https://doi.org/10.1257/rct.6446-1.3

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

Leif Helland* , Alice Ciccone , Kjell Arne Brekke , Tom-Reiel Heggedal

* Primary investigator

Registry ID
AEARCTR-0006446
DOI
https://doi.org/10.1257/rct.6446-1.3
First registered
Last updated
Status
In development
Trial dates
2020-10-05 – 2021-08-15
Countries
Denmark (Copenhagen), Norway (Oslo)
JEL codes
C72, C91, D03
Registry record
http://www.socialscienceregistry.org/trials/6446

Abstract

We introduce loss aversion in the infinite horizon, alternating offers model. When outside options serve as reference points, the equilibrium of our model follows the standard Rubinstein subgame perfect equilibrium. However, when reference points are given by the
resources players contribute to the pie, the bargaining outcome changes such that a player's share increases in her contribution.
The model is to be tested in two laboratory experiments. In experiment 1 random and earned outside options are compared. In experiment 2 random and earned contributions are compared.

Intervention

Experiment 1: In alternating offers bargaining over a pie of 70 experimental currency units (ecu's) second movers are allocated an outside option of 20 or 60 ecu's respectively. Value of the pie is reduced by 10 percent each round of bargaining. Treatments: outside options are randomly allocated vs. allocated based on performance in a costly effort task.

Experiment 2: In alternating offers bargaining the pie size is the sum of contributions brought to the table by the bargaining parties. Bargainers are allocated a contribution of 20 or 60 ecu's, respectively. Contributions have zero value outside of the bargaining. Value of the pie is reduced by 10 percent each round of bargaining. Treatments: contributions are randomly allocated vs. allocated based on performance in a costly effort task.

Experimental design

For treatments, see "Intervention". Data consists of 8 independent matching blocks in each treatment. Each matching block consists of 8 subjects playing 6 bargaining games. In each matching block 4 bargaining pairs are formed randomly prior to each new bargaining game. Subjects are unique in each matching block. Roles as first and second mover are fixed in games 1-3, and reversed in games 4-6. Within each experiment treatment effects are based on averages taken over all six games, using matching blocks as units of observation. Based on a pilot the design provides a power of >90 percent for our main predictions, using a (non-parametric) Wilcoxon rank sum test and a significance level of 5 percent.

Design details

Randomization method
Randomization done by computer (using the software zTree).
Randomization unit
Matching blocks (see "Experimental design").
Clusters
32 matching blocks (16 in Experiment 1 and 16 in Experiment 2)
Observations
256 subjects
Arms
Experiment 1: 8 matching blocks in treatment random allocation, 8 matching blocks in treatment earned allocation. Experiment 2: 8 matching blocks in treatment random allocation, 8 matching blocks in treatment earned allocation.
Primary outcomes
Average share of the pie obtained by the responder in the final agreement, conditioned on treatment and value of outside option/relative contribution to the pie.
Secondary outcomes
Delay in reaching agreement

Cited by

No citation data available for this registration yet.

Keywords