This dissertation applies experimental methods to study how information, incentives, and institutional context shape behaviors across health decision-making, pro-social behavior, and cooperation in social dilemmas.

In Chapter I, I test whether monetary incentives improve the accuracy of elicited beliefs about personal health risks and whether confidence moderates the link between beliefs and preventive behavior. I construct a 2×2 experimental treatment design where subjects provide either point estimates or complete belief distributions, which I compare to highly-personalized statistical benchmarks. Monetary incentives reduce mean squared error by 32% for belief distributions and have no effect for point estimates. Beliefs with higher confidence are more strongly associated with preventive care consumption, suggesting that eliciting both risk perceptions and confidence produces valuable additional information for researchers.

In Chapter II, by leveraging more than 30,000 coffee shop transactions, I compare information nudges with a choice-architecture change to tipping defaults. Messages communicating local descriptive norms do not reliably change the frequency or size of tips. By contrast, Square’s “Smart Tipping” feature, which sets dollar rather than percent defaults for bills under $10, consistently raises tip percentages without reducing the likelihood of tipping. The results indicate that simple default design outperforms information provision in this setting and provide practical guidance for firms and workers seeking low-cost, scalable ways to increase tip revenues.

Finally, in Chapter III, I examine how the origin of shared resources affects behavior in social dilemmas by distinguishing endogenous resources created through collective action from exogenous resources that exist independently of participants’ choices. A novel two-stage design links a public-goods provision stage to a subsequent mixed game over the resulting group account, and a between-subjects triadic structure separates trust, reciprocity, and unconditional other-regarding preferences. Subjects are less cooperative when group balances are endogenously low, and 50% display conditional other-regarding preferences when balances are exogenous, showing that resource origins shape cooperation.