Fairness and Risk in Ultimatum Bargaining

Last registered on June 15, 2020

View Trial History

Pre-Trial

Trial Information

General Information

Title

Fairness and Risk in Ultimatum Bargaining

RCT ID

AEARCTR-0005846

Initial registration date

May 13, 2020

Initial registration date is when the trial was registered.

It corresponds to when the registration was submitted to the Registry to be reviewed for publication.

First published

May 13, 2020, 3:38 PM EDT

First published corresponds to when the trial was first made public on the Registry after being reviewed.

Last updated

June 15, 2020, 2:37 PM EDT

Last updated is the most recent time when changes to the trial's registration were published.

Locations

Country

United States of America

Region

Primary Investigator

Name

Kyle Hyndman

Affiliation

Naveen Jindal School of Management, The University of Texas at Dallas

Contact Primary Investigator

Other Primary Investigator(s)

PI Name

Matthew Walker

PI Affiliation

Durham University Business School

Contact Investigator

Additional Trial Information

Status

In development

Start date

2020-05-15

End date

2020-12-31

Keywords

Other

Additional Keywords

Risk, Bargaining, Fairness

JEL code(s)

D91

Secondary IDs

Abstract

Using online experiments on Amazon Mechanical Turk (AMT), we will investigate what constitutes a fair allocation of risk in ultimatum bargaining. There are two agents, a proposer and a responder. We are interested in how offer and acceptance decisions differ when the proposer offers a probabilistic allocation of an indivisible asset, versus the standard case in which the asset is divisible and the proposer proposes a deterministic surplus allocation. We are also interested in how the timing of the resolution of uncertainty affects behavior.

External Link(s)

Registration Citation

Citation

Hyndman, Kyle and Matthew Walker. 2020. "Fairness and Risk in Ultimatum Bargaining." AEA RCT Registry. June 15. https://doi.org/10.1257/rct.5846-1.1

Sponsors & Partners

Experimental Details

Interventions

Intervention(s)

The purpose of our intervention is to understand how risk and intentions affect the fairness perceptions of agents in simple ultimatum bargaining environments. To study this, we will compare behavior of both proposers and responders in standard ultimatum games and in risky ultimatum games, where in the latter, we will vary the timing of the resolution of uncertainty.

Intervention Start Date

2020-05-15

Intervention End Date

2020-12-31

Primary Outcomes

Primary Outcomes (end points)

Offers by proposers and accept/reject decisions by responders

Primary Outcomes (explanation)

Secondary Outcomes

Secondary Outcomes (end points)

Secondary Outcomes (explanation)

Experimental Design

Subjects will participate in two ultimatum games as either a proposer or a responder: a standard ultimatum game (Standard UG) and a "lottery" ultimatum game (Lottery UG) in which there is an indivisible prize and the object of negotiation is the probability of winning the prize. We will conduct treatments with direct response by responders (Risk-Direct) and using the strategy method (Risk-Strategy). In order to also study the role of intentions, we consider a treatment variation in which the responder makes his/her accept/reject decision in the Lottery UG after the resolution of uncertainty (Intent-Direct).

Experimental Design Details

On AMT, subjects complete two ultimatum game (UG) tasks:

Standard UG. Each subject is randomly and anonymously paired with another participant. One of the subjects in the pair is assigned to the role of a Proposer and the other to the role of a Responder. Each pair has $6 to split between the Proposer and the Responder. The Proposer must decide how much of the $6 to offer to his/her matched Responder. Proposals can be in increments of $0.1. The Responder will observe the Proposer's offer and either accept or reject this proposal. If the Responder accepts, then if the Proposer offers $X to the Responder, the Proposer will earn $(6 - X) and the Responder will earn $X. If the Responder rejects, then both the Proposer and Responder will earn $0.

Lottery UG. Each subject is randomly and anonymously paired with another participant. One of the subjects in the pair is assigned to the role of a Proposer and the other to the role of a Responder. Each pair has 100 lottery tickets, numbered from 1, 2, ..., 99, 100, to split between the Proposer and the Responder. The Proposer must decide how many tickets to offer to his/her matched Responder. If the Proposer offers x tickets to his/her matched Responder, then the Responder will have tickets 1, 2, ..., x, while the Proposer will have tickets x+1, x+2, ..., 100. The Responder will observe the Proposer's offer and either accept or reject the proposal. If the Responder accepts, then the experimental software will randomly draw (with equal chance) a number between 1 and 100. The person who has the number drawn by the computer will earn $6, while the other person will earn $0. If the Responder rejects, then both the Proposer and Responder will earn $0.

Subjects are paid the outcome of one of the two tasks as a bonus. The task used for payment is determined at random by the computer after completion of both tasks and is paid in addition to a participation fee of $0.50.

We conduct three different treatment variations:
1. Risk-Direct. Subjects complete the two tasks as described above.

2. Risk-Strategy. Subjects complete the two tasks using a strategy method elicitation for the Responder’s decision. We use a two-level method to refine the minimum acceptance threshold.

3. Intent-Direct. Subjects complete the two tasks as described above except that, in the Lottery UG, the outcome of the lottery is revealed to the Responder before the acceptance decision.

We choose not to employ the strategy method for treatment Intent-Direct because of concerns about inducing an experimenter demand effect.

Randomization Method

Randomization done in office by a computer.

Randomization Unit

Groups of participants (sessions) are randomized to experimental treatments. We control for the order in which the two tasks are presented by randomly assigning half of subjects within each treatment to complete the Standard UG first and the other half to complete the Lottery UG first. In the first task, roles are assigned randomly in the pair. To permit within-subjects comparisons, participants retain the same role for the second task and the new pairing.

Was the treatment clustered?

Experiment Characteristics

Sample size: planned number of clusters

Based on our current plan, we estimate conducting 6-7 sessions per treatment, for a total of 18 - 21 sessions.

Sample size: planned number of observations

50 complete pairs per treatment, with 3 treatments for a total of 300 subjects.

Sample size (or number of clusters) by treatment arms

50 in Risk-Direct
50 in Risk-Strategy
50 in Intent-Direct

Minimum detectable effect size for main outcomes (accounting for sample design and clustering)

This sample size gives us 75% power to detect an offer effect size of 3 percentage points and 99% power to detect an offer effect size of 5 percentage points, assuming an average offer of 40%, standard deviation of 16 percentage points and within-subjects correlation of 75%, based on prior experiments.

Supporting Documents and Materials

There is information in this trial unavailable to the public. Use the button below to request access.

Request Information

IRB