Understanding Mechanisms of Persistence in Prosocial Behavior

How to cite

Haenni, Simon; Jiang, Lingqing; Bruhin, Adrian; Goette, Lorenz. 2021. "Understanding Mechanisms of Persistence in Prosocial Behavior." AEA RCT Registry. AEARCTR-0007366. https://doi.org/10.1257/rct.7366-1.0

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

Simon Haenni* , Lingqing Jiang , Adrian Bruhin , Lorenz Goette

* Primary investigator

Registry ID
AEARCTR-0007366
DOI
https://doi.org/10.1257/rct.7366-1.0
First registered
Last updated
Status
Completed
Trial dates
2015-03-01 – 2016-12-31
Countries
Switzerland ()
JEL codes
C93, D04, D91, C36
Registry record
http://www.socialscienceregistry.org/trials/7366

Abstract

We test whether asking individuals to donate blood leads to a persistent change in behavior, and examine the underlying mechanism. In a field experiment, we randomize a phone call, asking blood donors to turn out, and follow them over up to 18 months. We observe significant behavioral persistence for at least one year. We use naturally occurring rainfall as a second instrument for donor turnout to test whether behavioral persistence is due to habit formation (Stigler and Becker, 1977) or a persistent increase in motivation independent of past donation. Our results strongly favor habit formation as the underlying mechanism.

Intervention

Phone call asking pre-registered donors two days before the upcoming blood drive to make a donation.

Experimental design

4 experimental conditions: In the first condition, C1, donors receive a phone call only in period 1. In the second condition, C2, donors receive a phone call only in period 2, i.e., six months after the donors in condition C1. In the third condition, C12, donors receive a phone call in both periods 1 and 2. In the control condition, donors do not receive any phone calls. There are no further phone calls in periods 3 and 4, i.e., 12 and 18 months later. However, we keep observing donation outcomes in periods 3 and 4.

Design details

Randomization method
Randomization done by a computer
Randomization unit
Individual
Clusters
1400
Observations
5600
Arms
350 individuals in each treatment condition
Primary outcomes
Blood donation turnout (binary)

Cited by

No citation data available for this registration yet.

Keywords