Corruption and Cooperation: an experimental study from a public good game

This thesis explores the intricate relationship between corruption, cooperation, and transparency through both a theoretical analysis and an experimental investigation based on a public good game. The first part provides a comprehensive overview of corruption, focusing on its definitions, causes, and consequences for economic development and social capital. Particular attention is devoted to the role of transparency and the rule of law in fostering cooperative norms and curbing opportunistic behavior. Methodologically, the research relies on game theory and employs a public good game under two experimental settings: FULL INFORMATION, where participants are granted complete transparency from the outset, and PENDOX, where citizens can either purchase information about the monitor’s behavior or opt out of the game altogether, receiving a fixed payoff. Conducted over twenty periods, with groups composed of three citizens and one monitor, the experiment was designed to capture dynamics of contributions to the public good, instances of bribery, and punishment decisions. The working hypotheses predicted that PENDOX would foster higher contributions and lower corruption levels, as the exit option and the choice to acquire information reduce the monitor’s arbitrary power and diminish incentives for bribery. The data analysis largely supports these expectations. In the FULL INFORMATION setting, although transparency initially promotes cooperation, contributions tend to decline over time while bribery attempts increase, reflecting the monitor’s stronger punitive leverage. Conversely, in PENDOX, the availability of an exit strategy and optional information purchase create a more balanced environment: citizens display higher levels of cooperation, resort less frequently to bribery, and are subject to milder punishment from the monitor. Results indicate higher average contributions, fewer accepted bribes, and a progressive decrease in the number of sanctions imposed. These findings suggest that transparency alone is insufficient to sustain cooperation unless it is complemented by mechanisms that enhance individual choice and accountability. The exit option, despite its costs, acts as a deterrent against abuse of power and strengthens trust in the institutional framework. The research therefore highlights the effectiveness of combining transparency with agency-enhancing mechanisms as a means to mitigate the corrosive effects of corruption on social cooperation. In conclusion, the thesis underscores the need to rethink anti-corruption strategies not only in terms of transparency but also in terms of empowering citizens to monitor, choose, and actively influence institutional behavior. Although limited by sample size and the controlled laboratory setting, the experiment provides valuable insights for the design of governance tools aimed at reinforcing social capital and promoting cooperative behavior in contexts where corruption threatens collective welfare.