Gender and Asymmetric Bargaining

Last registered on March 28, 2017

View Trial History

Pre-Trial

Trial Information

General Information

Title

Gender and Asymmetric Bargaining

RCT ID

AEARCTR-0002029

Initial registration date

February 22, 2017

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

February 22, 2017, 10:28 AM EST

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

Last updated

March 28, 2017, 8:28 AM EDT

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

Locations

Country

Spain

Region

The Basque Country

Primary Investigator

Name

Nagore Iriberri

Affiliation

University of the Basque Country

Contact Primary Investigator

Other Primary Investigator(s)

PI Name

Iñigo Hernandez-Arenaz

PI Affiliation

University of the Basque Country

Contact Investigator

Additional Trial Information

Status

In development

Start date

2017-02-27

End date

2017-09-30

Keywords

Labor

Additional Keywords

Bargaining, Gender, Asymmetry

JEL code(s)

Secondary IDs

Abstract

The gender wage gap is still a real and not fully understood phenomenon. Recently, there has been a shift from more classical explanations to more behavioral explanations.

Following this line, it has been argued that gender differences in the propensity to negotiate and in the actual behavior when negotiating can explain a sizable part of the observed gender wage gap. For example, Card et al. (2016) found that 8% of the existing gender wage gap in Portugal could be explained by men and women behaving differently on these aspects.

One underexplored issue is that of gender interaction effects under asymmetric bargaining situations with one bargaining party being in a stronger position than the other. For example, Hernandez-Arenaz and Iriberri (2017) used data from a TV show with large stakes to study gender differences and gender interaction effects in an alternating-offer bargaining settings between two asymmetric parties. They found that the only matching that shows to be different is that between male proposers (in a strong position) and female responders (in a weak position): in this matching responders demand less yielding bargaining outcomes that are more favourable for proposers. Similar results have been found by Ditrich et al. (2014) and Andersen et al. (2014).

We propose a laboratory experiment to explore gender differences and gender interaction effects in asymmetric bargaining situations. The main design consists of a lab experiment with 4 different treatments which vary the degree and source of the bargaining asymmetry in order to identify gender differences and gender interaction effects when bargaining in a clean randomized lab experiment.

External Link(s)

Registration Citation

Citation

Hernandez-Arenaz, Iñigo and Nagore Iriberri. 2017. "Gender and Asymmetric Bargaining." AEA RCT Registry. March 28. https://doi.org/10.1257/rct.2029-2.0

Former Citation

Hernandez-Arenaz, Iñigo and Nagore Iriberri. 2017. "Gender and Asymmetric Bargaining." AEA RCT Registry. March 28. https://www.socialscienceregistry.org/trials/2029/history/15499

Sponsors & Partners

Experimental Details

Interventions

Intervention(s)

We propose a laboratory experiment to explore gender differences and gender interaction effects in asymmetric bargaining situations. The main design consists of a lab experiment with 4 different treatments which vary the degree and source of the bargaining asymmetry in order to identify gender differences and gender interaction effects when bargaining in a clean randomized lab experiment.

Intervention Start Date

2017-02-27

Intervention End Date

2017-09-30

Primary Outcomes

Primary Outcomes (end points)

Our main variables of interest are of two different kinds. With regard to aggregate bargaining outcome variables we study three variables: whether negotiations end up in a deal or not, how long it takes to reach a deal and the share of the pie that is finally captured by Participant A. These variables allow us to test if there is any evidence for gender differences or gender interaction effects. In terms of bargaining behavior the outcome variables we study are four of them: offers by Participant A, demands from participant B and the probabilities of accepting them. These variables allow us to understand whether the differences (if any) come from the offer side or the demand side.

Primary Outcomes (explanation)

Secondary Outcomes

Secondary Outcomes (end points)

Secondary Outcomes (explanation)

Experimental Design

The experiment consists of the following treatments:

Symmetric bargaining (Benchmark):
1. Subjects perform a real effort task.
2. Subjects choose the gender avatar that corresponds to their gender.
3. Based on their performance in the real effort task subjects are ranked. The top third is endowed with a productivity of 15 euro, the ones in the intermediate third with a productivity of 10 euro, and the third lowest with a productivity of 5 euro.
4. Each subject is randomly matched with another subject. One is assigned the role of participant A (proposer) and the other the role of Participant B (responder). Participant A is the one with the highest productivity for facilitating comparison across treatments (see below) although this assignment protocol is not communicated to the subjects.
5. For each matching, the pie to bargain over is with probability .5 the productivity of participant A and with .5 the productivity of Participant B.
6. They have 3 minutes to reach a deal on how to split the resulting pie through an alternating-offer bargaining. During the bargaining process Participant A decides over offers to Participant B while Participant B decides over demands from Participant A. During the bargaining, all subjects can see their own avatar and the one of the opponent (their gender and the gender of their matched partner), the size of the pie to be shared and the bargaining history of offers and demands, but they do not observe their individual productivity or their opponent’s. If they reach a deal within the 3 minutes limit, the agreed split is implemented. Otherwise they get 0.
7. Repeat stages 4-6, 9 more times resulting in 10 bargaining periods in total.

Entitlement: Everything is as in the Symmetric treatment except for the fact that they can now observe individual productivities as well as their bargaining partners’productivities. This is public knowledge.

Empowerment: Everything is as in the Symmetric treatment except for the fact that if they do not reach a deal within the 3 minute limit Participant A has an outside option, some random quantity between 0.5 and 0.85 of the total pie while Participant B gets 0. This is public knowledge and the exact amount that Participant A gets as outside option is not disclosed until the end of the 3 minutes.

Information: Everything is as in the Symmetric treatment except for the fact that only Participant A can observe the actual amount to be shared while Participant B only knows that this amount can be 5, 10 or 15 euro. This is public knowledge.

Experimental Design Details

Randomization Method

Computer randomization across treatments and to match subjects in the bargaining stage.

Randomization Unit

For the randomization across treatments (symmetric, and three asymmetric) the unit is a bargaining match between two subjects (between bargaining match).
For the random matching across subjects within a treatment the unit is the subject: the same subject will be matched with different bargaining partners (within subject). Within a treatment there will be N/2*10 different negotiations, where N is the number of subjects within a treatment.

Was the treatment clustered?

Yes

Experiment Characteristics

Sample size: planned number of clusters

Minimum: 40 subjects per treatment (4 treatments): 160 subjects: 200 bargaining matchings per treatment: 20 independent observations per treatment.
Maximum: 80 subjects per treatment (4 treatments): 320 subjects: 400 bargaining matchings per treatment: 40 independent data points per treatment.
For the aggregate outcome variables the most conservative number of clusters will be doing a two-way clustering (subject level clustering) and for the bargaining outcome variables (offers, demands, probabilities of acceptance) the most conservative number of clusters will be at the subject level (i.e. proposer level for offers and responder level for the demands).

Sample size: planned number of observations

Minimum: 40 subjects per treatment (4 treatments): 160 subjects: 200 bargaining matchings per treatment: 20 independent observations per treatment. Maximum: 80 subjects per treatment (4 treatments): 320 subjects: 400 bargaining matchings per treatment: 40 independent data points per treatment. For the aggregate outcome variables the most conservative number of clusters will be doing a two-way clustering (subject level clustering) and for the bargaining outcome variables (offers, demands, probabilities of acceptance) the most conservative number of clusters will be at the subject level (i.e. proposer level for offers and responder level for the demands).

Sample size (or number of clusters) by treatment arms

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

Supporting Documents and Materials

IRB