Skin in the Game: The Effect of Recipient Contribution on Charitable Giving

Last registered on December 20, 2016

View Trial History

Pre-Trial

Trial Information

General Information

Title

Skin in the Game: The Effect of Recipient Contribution on Charitable Giving

RCT ID

AEARCTR-0001850

Initial registration date

December 20, 2016

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

December 20, 2016, 6:39 PM EST

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

Anya Samek

Affiliation

University of Southern California

Contact Primary Investigator

Other Primary Investigator(s)

PI Name

Sally Sadoff

PI Affiliation

University of California-San Diego

Contact Investigator

Additional Trial Information

Status

On going

Start date

2016-12-16

End date

2017-01-31

Keywords

Welfare

Additional Keywords

charitable giving, field experiment

JEL code(s)

H41, C93

Secondary IDs

Abstract

A growing literature explores the effect of donor provided matching funds on charitable giving. However, no previous study has examined the effect of matching funds provided by the recipient of the donation. Such “donee matching” can potentially increase giving because donors may believe that if recipients have “skin in the game,” their charitable contribution is being distributed to those who value it most. We will run a randomized field experiment by inviting approximately 5,000 members of an online representative panel of Americans and asking them whether they would like to donate to a program that provides food for low income households. Within the experiment we will compare the impact on of varying the solicitation message to include either a low recipient monetary contribution (10% of costs) or a high monetary contribution (50% of costs) compared to no contribution. We also test the effect of low and high recipient time contributions. To uncover mechanisms, we will run additional treatments that include contributions from an anonymous donor and using any food (rather than restricted to healthy). Finally, we will study heterogeneous treatment effects by correlating our results to a rich set of data already available on the panel, with particular focus on political ideology. Together, our results will allow us to examine whether recipient contributions encourage giving and how donors respond to the level of the donee match.

External Link(s)

Registration Citation

Citation

Sadoff, Sally and Anya Samek. 2016. "Skin in the Game: The Effect of Recipient Contribution on Charitable Giving." AEA RCT Registry. December 20. https://doi.org/10.1257/rct.1850-1.0

Former Citation

Sadoff, Sally and Anya Samek. 2016. "Skin in the Game: The Effect of Recipient Contribution on Charitable Giving." AEA RCT Registry. December 20. https://www.socialscienceregistry.org/trials/1850/history/12719

Sponsors & Partners

Experimental Details

Interventions

Intervention(s)

Respondents to the survey will be invited to donate up to $8 (in any amount) of their $8 earnings from the survey to University of Southern California's food basket program.

Intervention Start Date

2016-12-16

Intervention End Date

2017-01-31

Primary Outcomes

Primary Outcomes (end points)

whether a donation is made, amount of the donation, amount of donation conditional on donating, beliefs (how much of the food is eaten, income of recipients, race of recipients, deservingness of recipients), interaction of political leanings with donation amount

Primary Outcomes (explanation)

Data from the experiment will tell us whether a donation is made and the amount. The political leanings data will be taken from the USC Dornsife/LA Times "Daybreak Poll" which included weekly questions on voting intentions (Trump versus Clinton) during the election.

Secondary Outcomes

Secondary Outcomes (end points)

Secondary Outcomes (explanation)

Experimental Design

There are 11 treatments, varying how much/who pays for the food basket and the varying the contents of the basket.

Experimental Design Details

There are 11 treatments. We will randomly assign approximately 5,000 survey participants to treatment in the following numbers: about 600 participants to each of treatments 1-3, about 500 to each of treatments 4-6, and about 300 to each of 7-11.

Main treatments (Healthy food baskets: fruit and vegetables only)

1. No money
2. Low money
3. High money
4. No money+no time
5. No money+low time
6. No money+high time

Testing mechanisms (foods of their choice)

7. No money
8. Low money
9. High money

Testing mechanisms (healthy food, anonymous donor)

10. Low money
11. High money

Treatment messages below:

Treatment 1 & 7
Families pay nothing for the basket. Donations provide the full $10 cost.

Treatment 2 & 8
Families contribute $1 for the basket. Donations provide the other $9.

Treatments 3 & 9
Families contribute $5 for the basket. Donations provide the other $5.

Treatment 4
Families receive the basket with no additional registration process. Families pay nothing for the basket. Donations provide the full $10 cost.

Treatment 5
Families receive the basket following a 5 minute registration process. Families pay nothing for the basket. Donations provide the full $10 cost.

Treatment 6
Families receive the basket following a 25 minute registration process. Families pay nothing for the basket. Donations provide the full $10 cost.

Treatment 10
Funding from a private donor provides $1 for the basket. Donations provide the other $9.

Treatment 11
Funding from a private donor provides $5 for the basket. Donations provide the other $5.

Randomization Method

Randomization done by NubiS software on the computer.

Randomization Unit

Individual

Was the treatment clustered?

Experiment Characteristics

Sample size: planned number of clusters

5,000 respondents

Sample size: planned number of observations

5,000 respondents

Sample size (or number of clusters) by treatment arms

Treatments 1-3: About 600 respondents per treatment
Treatments 4-6: About 500 respondents per treatment
Treatments 7-11: About 300 respondents per treatment

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

The minimum detectable effects size (MDES) of these tests is 0.17 standard deviation (SD) effect size for continuous outcomes at conventional levels of 80% power and 5% significance. For binomial outcomes (i.e., whether donated), the effect size ranges from 2.5-8.4 percentage points depending on baseline donation rates (assuming a minimum rate of 1%).

Supporting Documents and Materials

IRB