Social reference dependence

Last registered on June 24, 2024

View Trial History

Pre-Trial

Trial Information

General Information

Title

Social reference dependence

RCT ID

AEARCTR-0013697

Initial registration date

June 12, 2024

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

June 24, 2024, 1:41 PM EDT

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

Locations

Country

United States of America

Region

Primary Investigator

Name

Erik Ø. Sørensen

Affiliation

Norwegian School of Economics

Contact Primary Investigator

Other Primary Investigator(s)

PI Name

Alexander W. Cappelen

PI Affiliation

Norwegian School of Economics

Contact Investigator

PI Name

Samuel Hirshman

PI Affiliation

Norwegian School of Economics

Contact Investigator

PI Name

Matthew Rabin

PI Affiliation

Harvard University

Contact Investigator

PI Name

Bertil Tungodden

PI Affiliation

Norwegian School of Economics

Contact Investigator

Additional Trial Information

Status

In development

Start date

2024-05-29

End date

2024-06-26

Keywords

Behavior

Additional Keywords

fairness, reference points

JEL code(s)

D63

Secondary IDs

Prior work

This trial does not extend or rely on any prior RCTs.

Abstract

This study aims to test an extension of the reference dependence theory of Köszegi and Rabin (AER, 2007) to a social domain. In an online experiment, we create a situation where two people have worked under provided expectations of a bonus payment: With 45% probability, bonuses will be (a,b), with 45% probability bonuses will be (c,d), and with 10% probability something else will happen, and a decision will be made by either a one of the two worker participants (a "stakeholder") or a third party (a "spectator"). A treatment difference in this decision will be informative on the existence of reference dependence.

External Link(s)

Registration Citation

Citation

Cappelen, Alexander W. et al. 2024. "Social reference dependence." AEA RCT Registry. June 24. https://doi.org/10.1257/rct.13697-1.0

Sponsors & Partners

Experimental Details

Interventions

Intervention(s)

The interventions differ in terms of the what the payments to participants are and also by whether the decision maker is a stakeholder or a spectator.

Intervention Start Date

2024-05-29

Intervention End Date

2024-06-26

Primary Outcomes

Primary Outcomes (end points)

Proportion choosing equality over self interest (in stakeholder treatment). Proportion choosing more equal over larger total amount (in spectator treatment).

Primary Outcomes (explanation)

Decision makers make a single binary choice between allocations that will be implemented with 10% probability (when the intervention expectations won't be implemented). In the Stakeholder treatment, the binary choice is between (4,4) and (7,1), (6,2), or (5,3). Each stakeholder is faced with only one set of alternative (with 1/3 probability of them facing each). The allocation amounts are in US dollars.

In the Spectator treatments, the alternatives are (3,2) and either (8,1), (7,1) or (6,1). Each spectator is faced with only one set of alternative (with 1/3 probability of them facing each).

For the main analysis we plan to pool the share choosing the more equal outcome across the different alternatives.

Secondary Outcomes

Secondary Outcomes (end points)

Secondary Outcomes (explanation)

Experimental Design

In an online experiment, we create a situation where two people have worked under provided expectations of a bonus payment: With 45% probability, bonuses will be (a,b), with 45% probability bonuses will be (c,d), and with 10% probability something else will happen, and a decision will be made by either a one of the two worker participants (a "stakeholder") or a third party (a "spectator").

Experimental Design Details

In the Stakeholder treatment, decisionmakers are matched 1:1 to the passive worker in the pair (who makes no decisions). In the Spectator treatment, 1 out of 10 decisionmakers are randomly matched to a pair of workers and will potentially have their decision implemented.

The decision makers and the workers are recruited on different platforms. The decision makers are recruited via the survey provider NORSTAT, while the workers are recruited via Amazon mTurk.

The main analysis is to test the hypothesis (derived from theory) that without social reference dependence, the proportion choosing the more equal alternative should be the same in the Risky Equality and the Risky Inequality arms. Under the alternative hypothesis (loss aversion in social utility), the proportion choosing the more equal alternative should be larger in the Risky Equality arm. The main outcome will not be compared across the Stakeholder and Spectator arms.

There are attention checks participants must satisfy to take part in the experiment, and comprehension checks before the main decision to be made.

Randomization Method

Randomization is implemented by the survey software.

Randomization Unit

Each individual is independently assigned to treatment.

Was the treatment clustered?

Experiment Characteristics

Sample size: planned number of clusters

Not applicable.

Sample size: planned number of observations

4000 decision makers matched to 2400 workers (because of the 1:10 matching in the Spectator treatment).

Sample size (or number of clusters) by treatment arms

In each of the four cells of {Stakeholder, Spectator} X {Risky Equality, Risky Inequality} we aim to recruit 1000 decision makers.

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

We don't plan to compare outcomes across the Stakeholder and Spectator arms. Within each of the Stakeholder and Spectator arms, the main outcome (a proportion), under an assumption of the baseline proportion being 0.5, the minimum detectable effect size (power=0.8, alfa=0.05) is 0.055.

Supporting Documents and Materials

IRB