Summary: New research offers a computational explanation for a familiar idea: teenagers can seem self-centered. A study using a repeated Prisoner’s Dilemma game shows that adolescents are just as capable as adults of recognizing cooperative behavior, yet they are much less likely to reciprocate that cooperation. The findings suggest the intrinsic “reward” for returning kindness is weaker in adolescent brains than in young adults.
Researchers tested how willingness to cooperate develops from adolescence into adulthood. The study indicates that teens prioritize immediate personal gain in social dilemmas, even when cooperating would produce mutually better outcomes with a willing partner.
Key Findings
- Reciprocity gap: Adolescents and adults were equally accurate at judging how cooperative a partner was. The difference lies in behavior—teens see cooperation but are less motivated to mirror it.
- Selective response: Both groups reduced cooperation after a partner defected. But unlike adults, adolescents did not increase cooperation when their partner remained consistently cooperative.
- Computational results: Mathematical models revealed that the reward value assigned to reciprocating cooperative acts is substantially lower in 14–17-year-olds than in 18–30-year-olds.
- Self-interest emphasis: Adolescents favored immediate personal benefits over the longer-term advantages of mutual cooperation, even when the partner’s behavior made cooperative gains predictable.
Source: eLife
Overview of the study and its implications
Published in its final Version of Record after an earlier reviewed preprint, this study reinforces earlier reports that adolescents cooperate less than adults and explains why that difference may emerge. Instead of failing to recognize others’ kindness, adolescents appear to place lower intrinsic value on returning it.
“Adolescence is a time when young people learn to navigate friendships and group cooperation,” says Xiaoyan Wu, co-first author and postdoctoral researcher. “Our aim was to move beyond the simple observation that teens cooperate less and to understand the decision processes behind that pattern.”
Wu and colleagues recruited 261 participants: 127 adolescents (ages 14–17) and 134 adults (ages 18–30). Participants played a repeated Prisoner’s Dilemma game in which each round gave the choice to cooperate for shared benefit or to defect in order to gain more immediately at the partner’s expense. The researchers designed the partners’ behavior with pre-determined cooperation patterns to test how participants adapted across rounds.
Behavioral results showed adolescents cooperated less overall, but the gap was particularly clear after partners had demonstrated cooperative behavior in prior rounds. Crucially, ratings and self-report measures indicated that adolescents were just as good as adults at estimating how cooperative their partner was—so perception was not the limiting factor.
To unpack the decision mechanisms, the team built computational models that combined expectations about a partner’s future cooperation with a variable representing the intrinsic reward of reciprocity. Modeling revealed a clear developmental difference: adults raised their internal reward for reciprocating when partners stayed cooperative, whereas adolescents did not show this increase. Both groups, however, reduced cooperation in response to consistent defection, indicating shared sensitivity to being exploited.
“These results point to development in the valuation of reciprocity rather than a failure to predict others’ actions,” says co-first author Hongyu Fu. “Adolescents can learn about others’ tendencies, yet they often prioritize immediate self-interest over the social payoff of returning kindness.”
The authors acknowledge limitations. The study used artificial opponents with fixed cooperation patterns, so behavior might differ in more naturalistic, face-to-face interactions. Participants came from Beijing and nearby areas, which reduced regional variation but did not eliminate socioeconomic and social experience differences that could influence cooperation development.
Despite these caveats, the work offers a meaningful step toward understanding how cooperative motivation matures. If adolescents assign less value to reciprocating kindness, interventions could focus on emphasizing the long-term benefits of fairness and teaching why returning cooperation strengthens social relationships.
Key Questions Answered
A: No. The study shows adolescents accurately perceive cooperative behavior. The gap appears in decision-making: they recognize cooperation but often prefer the immediate personal gain of not reciprocating.
A: Not necessarily. The finding is better explained by developmental changes in how the adolescent brain values long-term social rewards versus short-term gains. Adolescents react to uncooperative partners similarly to adults, but proactive reciprocation develops more slowly.
A: The authors suggest interventions could help adolescents better interpret others’ intentions and strengthen the value placed on fairness and reciprocity, emphasizing why returning kindness benefits them over time.
Editorial Notes
- This article was edited by a Neuroscience News editor.
- Journal paper reviewed in full.
- Additional context added by editorial staff.
About this research
Author: Emily Packer
Source: eLife
Contact: Emily Packer – eLife
Image: Image credited to Neuroscience News
Original Research: Open access. “The self-interest of adolescents overrules cooperation in social dilemmas” by Xiaoyan Wu, Hongyu Fu, Gökhan Aydogan, Chunliang Feng, Shaozheng Qin, Yi Zeng, and Chao Liu. DOI: 10.7554/eLife.106840.4
Abstract
The self-interest of adolescents overrules cooperation in social dilemmas
Cooperation is a cornerstone of social success. Prior work consistently shows adolescents are less cooperative than adults, a pattern often attributed to underdeveloped mentalizing and social expectations. This study explores the internal computations behind that reduced cooperation.
Using a repeated Prisoner’s Dilemma game, researchers compared adolescents and adults. Adolescents cooperated less overall, particularly after their partner had cooperated. Computational modeling indicated adults increased the intrinsic reward for reciprocation when partners remained cooperative—a dynamic absent in adolescents. Self-report and modeling also showed adolescents formed similar expectations about partners’ behavior as adults.
The results suggest that lower adolescent cooperation stems from a reduced intrinsic reward for reciprocity and a stronger focus on immediate self-interest, rather than an inability to predict others’ cooperative choices. These insights clarify how cooperative motivation develops from adolescence into adulthood and point to directions for interventions that promote fairness and reciprocal behavior in youth.