Skip directly to content

Minimize RSR Award Detail

Research Spending & Results

Award Detail

Doing Business As Name:Cornell University
  • Michael W Macy
  • (607) 255-4555
Award Date:03/13/2013
Estimated Total Award Amount: $ 115,412
Funds Obligated to Date: $ 115,412
  • FY 2013=$115,412
Start Date:03/15/2013
End Date:08/31/2016
Transaction Type:Grant
Awarding Agency Code:4900
Funding Agency Code:4900
CFDA Number:47.075
Primary Program Source:040100 NSF RESEARCH & RELATED ACTIVIT
Award Title or Description:Generalized Reciprocity: Can Generosity Become Contagious?
Federal Award ID Number:1260348
DUNS ID:872612445
Parent DUNS ID:002254837

Awardee Location

Street:373 Pine Tree Road
Awardee Cong. District:23

Primary Place of Performance

Organization Name:Cornell University
Street:373 Pine Tree Road
Cong. District:23

Abstract at Time of Award

Every day, millions of people donate blood, stop to help a stranded motorist, participate in unpaid surveys, review restaurants, leave online product ratings, answer questions on, contribute code to open-source software, and edit Wikipedia articles, often anonymously, without social recognition or opportunity to benefit from direct reciprocation. This widespread willingness to help a stranger poses a fascinating theoretical puzzle with enormous practical implications. An important key to the puzzle may be the possibility that helping behavior can be contagious--people respond to being helped by "paying it forward" to a third person. Known as "generalized reciprocity," this behavior has the potential to trigger a chain of helping that reaches far beyond the original act. Generalized reciprocity bridges two largely disconnected existing literatures on social contagion and altruistic behavior. Integrating these two research traditions, this project identifies and tests four specific conditions expected to alter the probability that a recipient of help will pay it forward: group size, observation of generalized reciprocity by others, the number of others from whom help has been received, and the number of others who will benefit. Data will be collected from a large field experiment in an online labor market and validated with a small-group experiment run in a traditional laboratory setting. The results will advance theoretical knowledge about helping behavior and its diffusion, as well as new methodologies for online research, both within sociology and across multiple fields, including social psychology, behavioral economics, moral philosophy, computer science, communication, and information science. Broader Impact The potential ripple effect of generalized reciprocity means that when we help others, we may also be indirectly helping many -- both online and offline -- to mobilize voluntary contributions by more effectively targeting those who have benefited in the past. In addition, the project will have practical benefits for the scientific community by developing and demonstrating an innovative method for large online field experiments that can be adapted for a wide range of other studies of decision-making and diffusion.

Publications Produced as a Result of this Research

Note: When clicking on a Digital Object Identifier (DOI) number, you will be taken to an external site maintained by the publisher. Some full text articles may not yet be available without a charge during the embargo (administrative interval).

Some links on this page may take you to non-federal websites. Their policies may differ from this site.

Milena Tsvetkova Michael W. Macy "The Social Contagion of Generosity" PLOS ONE, v.10.1371, 2014, p.0087275. doi:PLOS ONE 

Project Outcomes Report


This Project Outcomes Report for the General Public is displayed verbatim as submitted by the Principal Investigator (PI) for this award. Any opinions, findings, and conclusions or recommendations expressed in this Report are those of the PI and do not necessarily reflect the views of the National Science Foundation; NSF has not approved or endorsed its content.

The health regime we follow, the music we listen to, the new technologies we adopt, the news stories we read, and even the likelihood that we vote in an election are all to a large degree influenced by our friends and peers. Many human behaviors spread through social contact, including social behavior. Previous research has suggested that generosity and prosocial behavior, as well as “broken windows” and other minor antisocial infractions may spread between individuals. But how does such contagion occur? In this project, we distinguished between two mechanisms for the contagion of social behavior among strangers: generalized reciprocity (a recipient of social behavior is more likely to pay it forward) and third-party influence (an observer of social behavior is more likely to emulate it). We conducted two online experiments and a computational study to investigate these mechanisms and their effects.

First, we investigated the contagion of prosocial behavior. We conducted an online experiment with 574 participants recruited worldwide using Amazon Mechanical Turk. Participants were given the option to donate part of their earnings to a future participant. Most participants could receive a donation, but “seeds” could not. Participants who received a donation from a previous participant were more likely to donate to others, but observing donation reduced the willingness to help among “seeds” but not among those who had received a donation.

Second, we focused on the contagion of antisocial behavior. In a second online experiment, we recruited 750 participants from Amazon Mechanical Turk and gave them the option to steal part of the earnings from a future participant. Most participants could be victimized, but “seeds” could not. Those who were victimized by a previous participant were more likely to victimize others.  Observing honesty made seeds more honest, but observing theft had no effect, and observation had no effect among victims.

If receiving and observing generosity increase generosity, under what conditions can the spread of prosocial behavior become self-sustaining? We used an agent-based model to investigate how successful voluntary-contribution communities emerge among anonymous individuals. We found that contributions which benefit many individuals (“nonrival” contributions) are much more likely to spread than those which benefit only a single individual (“rival”). This finding offers explanation for the fact that cultures of kindness are rare for anonymous face-to-face interactions but common online, for example in the form of user-generated-content communities.

The project furthered current understanding of prosocial behavior, antisocial behavior and social contagion. It also contributed an innovative experimental design and a new online platform for studying the diffusion of behavior in large social groups under controlled conditions. Our findings have implications for strategies for encouraging group-oriented contributions and controlling aggressive and self-serving actions and are applicable to large-scale online collaborative projects and content communities.


Last Modified: 12/09/2016
Modified by: Michael W Macy

For specific questions or comments about this information including the NSF Project Outcomes Report, contact us.