The effect of perceived prosocial impact on the relationship between monetary incentives and prosocial work

Last registered on May 30, 2025

View Trial History

Pre-Trial

Trial Information

General Information

Title

The effect of perceived prosocial impact on the relationship between monetary incentives and prosocial work

RCT ID

AEARCTR-0015825

Initial registration date

May 26, 2025

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 30, 2025, 9:23 AM EDT

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

Locations

Country

United Kingdom of Great Britain and Northern Ireland

Region

Primary Investigator

Name

Hadar Gafni

Affiliation

King's College London

Contact Primary Investigator

Other Primary Investigator(s)

PI Name

Lars Bo Jeppesen

PI Affiliation

Copenhagen Business School

Contact Investigator

Additional Trial Information

Status

In development

Start date

2025-05-26

End date

2025-06-09

Keywords

Behavior

Additional Keywords

prosocial motivations, incentives, behavioral economics

JEL code(s)

Secondary IDs

Prior work

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

Abstract

This study examines how monetary incentives influence people’s willingness to help others, and how that effect changes depending on who is being helped. Organizations often want to encourage prosocial behavior—actions people take to help others—but relying only on people’s goodwill is often not enough. One common approach is to offer money as a reward. However, previous research has shown that in some cases, offering money can actually reduce people’s motivation to help, rather than increase it. This study explores when money helps and when it backfires.
We use an online survey-experiment with a sample of individuals living in the United Kingdom, aged 18-70. In the first step, participants read a short scenario where someone is asked to help load a sofa into a van. Across different versions of the scenario, we change two things: whether the helper is offered no money, a small amount ($0.50), or a larger amount ($5), and who the beneficiary is—ranging from a wealthy person to someone who is elderly and in a wheelchair. Participants are then asked how likely the “average person” would be to help in that situation.
We also include additional questions to better understand why people think others would help or not, based on existing psychological theories of motivation. This will help us learn more about what types of motivation matter most—and when money helps or hurts. In a second step, participants are shown a new scenario where the incentive is changed, allowing us to examine how people believe motivation shifts when financial rewards are introduced or removed. We anticipate writing at least one paper based on this study. If the second step yields informative and novel results, a second paper may be developed. All outcomes and analyses are included here regardless of final publication format.

External Link(s)

Registration Citation

Citation

Gafni, Hadar and Lars Bo Jeppesen. 2025. "The effect of perceived prosocial impact on the relationship between monetary incentives and prosocial work." AEA RCT Registry. May 30. https://doi.org/10.1257/rct.15825-1.0

Sponsors & Partners

Experimental Details

Interventions

Intervention(s)

Intervention (Hidden)

Intervention Start Date

2025-05-26

Intervention End Date

2025-06-09

Primary Outcomes

Primary Outcomes (end points)

Participants’ ratings (on a scale from 1 = Extremely unlikely to 11 = Extremely likely) of how likely the average person would be to help in the described scenario – willingness to help.

Primary Outcomes (explanation)

Secondary Outcomes

Secondary Outcomes (end points)

Secondary Outcomes (explanation)

Experimental Design

This is a survey-experiment examining how monetary incentives and perceived prosocial context jointly shape judgments about others’ willingness to help. Participants are randomly assigned to one of 12 conditions in a 4×3 between-subject design. The first dimension manipulates monetary incentives for helping (no incentive, $0.50, or $5). The second dimension manipulates the perceived prosocial need of the person receiving help: (1) a neutral, unspecified beneficiary (replicating Heyman & Ariely, 2004), (2) a wealthy individual, (3) an elderly person from a poor neighborhood, or (4) an elderly person in a wheelchair.
Participants read a brief description of a scenario in which someone is asked to help load a sofa into a van. Based on the version they receive, they are asked how likely they think the average person would be to help, rated on a scale from 1 to 11 (11 being extremely likely to help). Asking about of the likelihood of others to help rather than the likelihood of the participants themselves is essential to potential reduce social desirability biases (Epley & Dunning, 2000; Fisher, 1993; Fisher & Katz, 2000).
In a second step, participants are presented with a revised version of the scenario in which the monetary incentive changes (to one of the two conditions they did not initially receive), allowing us to examine how people believe willingness to help changes when financial rewards are added or removed. This step makes the experiment a within-subject experiment.
Between the two steps, all the participants answer a questionnaire about the motivations of the “average person” to help (or not) by adapting questions from the Multidimensional Work Motivation Scale (MWMS), Ryan and Connell (1989), Millette & Gagné (2008), Gagné (2015), Fenigstein, Scheier, and Buss (1975), Grant (2008), Duffy & Kornienko (2010), Hartmann et al (2017), Leuker (2021), Bandura (2006), Cuddy, Fiske, Glick (2007), Erlandsson et al (2015), Jie (2020), and others. The answers will be used to understand the mechanisms that explain the answers in the first stage, and are not part of the experiment itself.

Experimental Design Details

Randomization Method

Random allocation by Qualtrics

Randomization Unit

Individuals.

Was the treatment clustered?

Experiment Characteristics

Sample size: planned number of clusters

720 individuals

Sample size: planned number of observations

720 individuals living in the United Kingdom, aged 18-70.

Sample size (or number of clusters) by treatment arms

60 individuals for each of our 12 cells, overall 720 individuals. In the second step, each cell is divided in two, resulting 24 cells, 30 in each one.

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

The primary outcome is measured on a Likert-type scale from 1 to 11. Based on pilot data and prior replications, we estimate the standard deviation of this outcome to be approximately 3.0. The study includes 720 participants, randomly assigned to 12 experimental conditions (60 participants per cell) in a 4×3 between-subject design. Assuming α = 0.05 (two-tailed) and power = 0.80, the minimum detectable raw difference between two groups is approximately 1.55 points on the 1–11 scale. Considering the results of our pilot, this difference should be large enough. This corresponds to a Cohen’s d of approximately 0.52. In the second step, we have 30 participants per cell, which increases the minimum detectable raw difference between two groups to approximately 2.21 points There is no clustering in this study design; randomization occurs at the individual level.

Supporting Documents and Materials

IRB