The effect of cooperative group composition on monetary contributions for public goods. A lab experiment from Ecuador.

How to cite

Mangraviti, Nicola. 2026. "The effect of cooperative group composition on monetary contributions for public goods. A lab experiment from Ecuador.." AEA RCT Registry. AEARCTR-0019295. https://doi.org/10.1257/rct.19295-1.1

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

Nicola Mangraviti*

* Primary investigator

Registry ID
AEARCTR-0019295
DOI
https://doi.org/10.1257/rct.19295-1.1
First registered
Last updated
Status
In development
Trial dates
2026-09-03 – 2026-09-21
Countries
Private
Registry record
http://www.socialscienceregistry.org/trials/19295

Abstract

This research project aims at investigating the effect of unconditional cooperation on monetary contributions to public good. According to existing evidence on cooperative types, especially based on influential works such as Fischbacher et al, 2001; Gachter, 2006 and Fischbacher and Gachter, 2006, individuals are mainly classified into two categories that is conditional cooperators and free riders. Conditional cooperators are individuals who contribute based on the observation of the others’ contributory levels or based on the belief that the others will contribute. Free riders are by definition individuals who never contribute and therefore reap the benefits of the others’ commitment and efforts. These represent the two most common cooperative types. Less observed cooperative types are the triangular cooperator, who are individuals who contribute only until certain levels and then gradually or suddenly withdraw from cooperation, and the unconditional cooperator (Thoni and Volk, 2018). Unconditional cooperators are individuals who provide constant positive levels of contributions in public goods, regardless of the level provided by other peers. These types have been observed and detected also in other studies (Hermann and Thoni, 2009; Thoni and Volk, 2018).
The composition of social groups deeply affects cooperation levels. Based on existing literature on cooperation in public good games, within groups cooperation is expected to decline over time not just because of the presence of pure free riders but also because of conditional cooperators. In fact, according to Fischbacher and Gachter, conditional cooperators are imperfect in the sense that they do not tend to reciprocate in a proportional way. If they observe a certain level of monetary commitment from other peers, they give a little less. This eventually leads to general free riding for what regards public good outcomes. This study draws from existing evidence on cooperative types and focus on a rare type of cooperative agent, that is the unconditional cooperator, and its effect on groups’ trend in cooperation levels. We argue that unconditional cooperators may serve as social anchors for other peers, especially conditional cooperators, and their presence in groups may alter the tendency of cooperation to decline over time. To the best of our knowledge, no previous study has focused on the specific effect of positive unconditional giving on groups’ cooperative trajectories

Experimental design

The laboratory experiment consists of a two-stage design. In the first stage the participants are asked to fill in the Contribution Table from Fischbacher et al, 2001 where they are asked to indicate for each average contribution level of other group members how much they want to contribute to the public good. In this way we elicit individual preferences for cooperation. This subsequently allows us to analyze results from the second stage by cooperative type.
In the second stage participants are randomly assigned to control and treatment sessions. In each session, participants are randomly matched in groups of 4 people, have the same endowment of z tokens and play a repeated public good game with a certain marginal per capita return MPCR (the return for 1 token contributed by a participant) for n rounds. Groups compositions are constant over rounds as well as the parameters z, n and MPCR (which need to be discussed of course). At the end of each round all members of a group are told the other members’ contributions.
Treatment sessions differ from control sessions by the way groups are composed. While participants in the control sessions are assigned to groups randomly, participants in the treatment sessions are matched according to the cooperative preferences elicited in the first stage. Specifically, participants classified as conditional cooperators, unconditional cooperators, free riders, or other behavioral types are allocated to groups with predetermined compositions. Depending on the treatment, groups may consist of members with similar cooperative preferences (e.g., all conditional cooperators, all free riders, or all unconditional cooperators) or deliberately heterogeneous compositions combining different behavioral types. This design allows us to examine how the composition of cooperative preferences within a group influences contribution dynamics, the evolution of cooperation over repeated interactions, and overall public good provision relative to randomly formed groups.

Design details

Randomization method
Randomization by computer.
Randomization unit
University students.
Clusters
40 groups of students.
Observations
160 students divided in 40 groups, playing for 20 rounds, for a total of 800 obs.
Arms
20 control and 20 treatment (group composition).
Primary outcomes
Monetary contributions in the public good game.

Cited by

No citation data available for this registration yet.

Keywords