A Field Experiment on Group Competition and Charitable Giving

How to cite

Oslislo, Christoph; Schwerter, Frederik. 2021. "A Field Experiment on Group Competition and Charitable Giving." AEA RCT Registry. AEARCTR-0007724. https://doi.org/10.1257/rct.7724-1.2

Download: BibTeX · RIS — for reference managers without a browser translator. Zotero and most others pick up this page's citation metadata directly.

Christoph Oslislo* , Frederik Schwerter

* Primary investigator

Registry ID
AEARCTR-0007724
DOI
https://doi.org/10.1257/rct.7724-1.2
First registered
Last updated
Status
Completed
Trial dates
2021-06-14 – 2022-07-31
Countries
Germany ()
Registry record
http://www.socialscienceregistry.org/trials/7724

Abstract

In collaboration with a large charitable organization, we conduct a large-scale field experiment designed to investigate the role of group identity for donation decisions in a natural setting. In particular, we test the hypothesis that perceived group membership has the potential to increase charitable giving. The interventions takes place in June and July 2021. In addition to the main research question, the project aims at shedding light on (I) whether it needs explicit incentives for salient group membership to have a positive effect on individual donation decisions, (II) where to locate the appeal to be most effective, and (III) whether effects of group identity on charitable giving can be further exploited by a supplementary peer-to-peer fundraising method. After approximately twelve months, we measure long-term effects of our treatment manipulation.

Intervention

In a regular donation request, we vary whether there is salient group membership and whether there is competition on the group level. The requests will be sent out via postal mail and supplemented with an e-mail (if applicable).

Experimental design

In the first experiment, the intervention takes place in June 2021. Regular donors receive a donation request via postal mailing. All mailings are identical except from the precise wording of the appeal. In this experiment, we ask whether a salient group membership alone affects donations. In order to investigate this question, we prime a natural group identity of subjects in the treatment group without explicitly mentioning competition. Subjects in the control condition receive no group membership prime and no information on any competition. In addition, we test where the appeal needs to be located to be most effective. Using a 2x2 factorial design, we vary the appeal (Group vs. Individual) along one factor. Along the other factor, we vary where subjects are confronted with the appeal for the first time (Envelope vs. NoEnvelope). In the second experiment, we implement a third treatment variation in addition to the "Group" and the "Individual" treatment of the first experiment. In addition to receiving the group identity prime, participants in this third treatment group are informed that we will announce the "most cooperative" groups as part of the charity's next regular newsletter and online.

Design details

Randomization method
Randomization done in office by a computer
Randomization unit
Randomization on individual-level
Clusters
1) approximately 42.000 donors 2) approximately 71.000 donors
Observations
1) approximately 42.000 donors 2) approximately 71.000 donors
Arms
1) 10.500 donors in each of the four treatment arms 2) 23.700 donors in each of the four treatment arms
Primary outcomes
Key variables of interest will be donation decisions at the extensive margin as well as donation sizes and total donations raised (also considering cancellation rates) under different treatment conditions and at different points in time.
Secondary outcomes
To shed light on the behavioral mechanisms, we will additionally measure heterogenous treatment effects with respect to city characteristics (e.g. sizes), gender, and individual donation history, e-mail open rates and clickthrough rates on info webpages, donation decisions resulting from a P2P fundraising strategy, as well as long-term effects of the intervention (measurement of primary outcomes one year later - before and after having published tournament results).

Cited by

No citation data available for this registration yet.

Keywords