When delaying action becomes attractive: dynamic costs and coordination under collective risk

How to cite

Liang, Sijia; Weber, Till; Walker, Matt; Sharma, Smriti. 2026. "When delaying action becomes attractive: dynamic costs and coordination under collective risk." AEA RCT Registry. AEARCTR-0019099. https://doi.org/10.1257/rct.19099-1.0

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

Sijia Liang* , Till Weber , Matt Walker , Smriti Sharma

* Primary investigator

Registry ID
AEARCTR-0019099
DOI
https://doi.org/10.1257/rct.19099-1.0
First registered
Last updated
Status
In development
Trial dates
2026-07-04 – 2027-06-20
Countries
Private
Registry record
http://www.socialscienceregistry.org/trials/19099

Abstract

Using a live online experiment on Prolific Academic, this study investigates whether declining contribution costs affect the group success in a collective-risk social dilemma (CRSD; Milinski et al., 2008). Groups must make contributions toward a group account to reach a shared target to avoid a collective loss. We compare a constant-cost condition, in which contribution costs stay the same across rounds, with a declining-cost condition, in which contribution costs decrease across rounds, while holding the minimum total cost of reaching the target constant. We examine whether declining costs reduce the group's likelihood of reaching the target, and whether they delay contributions to later rounds.

Intervention

Participants will be randomly assigned to one of two experimental conditions: (i) a constant-cost control condition: each token costs £0.50 in every round; (ii) or a declining-cost treatment condition: the cost of each token declines by £0.05 per round, from £0.85 in round 1 to £0.40 in round 10.

Experimental design

Participants play an incentivised CRSD game in groups of four over ten rounds. Each participant begins with a private endowment of £10. In each round, participants decide how many tokens (0, 1, or 2) to buy and add to a group account, with the cost of each token deducted from their private account. If the group accumulates at least 40 tokens by the end of round 10, participants keep their remaining private balance. If not, each faces a 70% chance of losing it.

The study uses a between-subjects design. After providing informed consent, participants complete a background questionnaire and are randomly assigned to one of two contribution-cost conditions. They read the instructions for their condition and complete a comprehension check, then are grouped into fours with others in the same condition based on arrival time.

Design details

Randomization method
Randomization will be implemented in oTree.
Randomization unit
Individual level.
Clusters
N/A
Observations
480 participants
Arms
60 groups, 240 participants in the control condition; 60 groups, 240 participants in the declining-cost treatment condition.
Primary outcomes
Our primary outcome is the success rate, defined as the share of groups that reach the target by the end of the game, compared between conditions.
Secondary outcomes
Our secondary outcome is the back-loading share, measured at the group level as the total number of tokens contributed in rounds 6-10 divided by the total number of tokens contributed across all ten rounds. Any group with zero total contributions across all ten rounds will be excluded from the back-loading share analysis.

Cited by

No citation data available for this registration yet.

Keywords