Cooperation and Contagion in Web-Based, Networked Public Goods Experiments (1008.1276v2)

Abstract: A longstanding idea in the literature on human cooperation is that cooperation should be reinforced when conditional cooperators are more likely to interact. In the context of social networks, this idea implies that cooperation should fare better in highly clustered networks such as cliques than in networks with low clustering such as random networks. To test this hypothesis, we conducted a series of web-based experiments, in which 24 individuals played a local public goods game arranged on one of five network topologies that varied between disconnected cliques and a random regular graph. In contrast with previous theoretical work, we found that network topology had no significant effect on average contributions. This result implies either that individuals are not conditional cooperators, or else that cooperation does not benefit from positive reinforcement between connected neighbors. We then tested both of these possibilities in two subsequent series of experiments in which artificial seed players were introduced, making either full or zero contributions. First, we found that although players did generally behave like conditional cooperators, they were as likely to decrease their contributions in response to low contributing neighbors as they were to increase their contributions in response to high contributing neighbors. Second, we found that positive effects of cooperation were contagious only to direct neighbors in the network. In total we report on 113 human subjects experiments, highlighting the speed, flexibility, and cost-effectiveness of web-based experiments over those conducted in physical labs.

• The paper finds that network topology does not significantly affect average contributions in public goods games, contrary to theoretical predictions.
• It demonstrates that conditional cooperation is symmetrical, with responses to low contributions equaling those to high contributions among participants.
• The study shows that cooperative behavior fails to spread beyond direct neighbors, prompting a re-evaluation of contagion assumptions in networked experiments.

Cooperation and Contagion in Networked Public Goods Experiments

The paper by Suri and Watts closely investigates the role of network topology in public goods games through a series of web-based experiments conducted on Amazon Mechanical Turk (AMT). This paper uniquely contrasts established theoretical expectations about cooperation in social networks with empirical evidence gathered by deploying controlled experiments with a diverse participant pool.

Experimental Setup

The authors explore how variations in network topology—ranging from highly clustered cliques to random regular graphs—affect cooperation among participants engaged in public goods games. Participants were grouped into networks of 24 individuals under five different topologies: disconnected cliques, paired cliques, cycles, small-world networks, and random regular graphs. Each participant played a local public goods game, modulating their contributions across several rounds.

Key Findings

Contrary to expectations informed by existing literature, the experiments demonstrated no significant effect of network topology on average contributions. This contradicts the hypothesis that highly clustered networks foster greater cooperation through mutual reinforcement among neighboring conditional cooperators. Two potential explanations were tested:

1. Conditional Cooperation: Experiments involving artificial "seed" players making consistent full or zero contributions revealed that players do behave as conditional cooperators. However, responses to low-contributing neighbors were as strong as responses to high-contributing ones, negating the effect of positive reinforcement.
2. Contagion: Another hypothesis—that cooperation or defection could spread contagiously beyond immediate neighbors—was debunked as seeds influenced only direct neighbors. Adjusting the positioning of cooperative seeds did not amplify contributions among second-degree connections, indicating limited propagation through the network.

Implications

The findings challenge prevalent theories suggesting network structure significantly influences cooperation, emphasizing the dual nature of conditional cooperation which can reinforce either cooperative or defecting behavior based on initial conditions. The lack of contagion signifies the need to re-evaluate assumptions about behavior propagation in structured networks.

Future Directions

The paper opens avenues for further investigation into the theoretical conditions under which contagion might manifest in networks, suggesting an exploration of different network settings or introducing dynamic elements such as network rewiring or evolving strategies. Additionally, understanding how targeted interventions, like strategic placement of unconditional cooperators, can optimize network-wide outcomes without modifying fundamental game incentives remains a pertinent field of inquiry.

Conclusion

By leveraging the scale and flexibility of web-based platforms like AMT, the research provides robust evidence that questions the assumed relevance of network topology on cooperative behavior in game theory contexts. These insights are critical not only for theoretical advancements but also for crafting effective strategies in promoting cooperation in real-world networked environments.