Essays in Behavioral and Experimental Economics
William Hickman
Advisor: Johanna B Mollerstrom, PhD, Department of Economics
Committee Members: Daniel Houser, Kevin McCabe, Joy Buchanan
Vernon Smith Hall (formerly Metropolitan Building), #5075
March 28, 2025, 09:00 AM to 11:00 AM
Abstract:
This dissertation is composed of three experimental studies that investigate human behavior in diverse economic settings, offering insights into labor supply, trust in information sources, and household decision-making.
Chapter 1 examines how redistributive systems affect individual effort provision. Through a series of controlled online experiments involving 1,600 participants, the study tests whether withholding a portion of labor earnings leads to reduced work effort, while holding net wages constant. The findings reveal a robust aversion to “working for others” when autonomy is limited. However, when individuals are provided with the option to opt out of having a portion of their earnings go to others, the negative impact on effort is mitigated. These results suggest that increasing perceived autonomy can enhance the efficiency of redistributive policies and improve labor outcomes.
Chapter 2 investigates trust in human and AI-produced text. Using an incentivized survey experiment, the study compares the credibility attributed to statements produced by human authors and those generated by large language models, such as ChatGPT. Results indicate that when participants are uninformed about the source, they tend to have higher trust in content they believe is human-written. In contrast, explicit disclosure of authorship leads to comparable levels of skepticism toward both human and AI-produced statements. This work highlights the context-dependent nature of trust in AI and provides insights into the role of AI systems in sharing information.
Chapter 3 explores household decision-making in rural Rwanda by assessing how spouses value control over income and time use. Through a lab-in-the-field experiment with agricultural cooperative members, the analysis reveals that while women value control over household income more than men, this valuation is relatively modest for both women and men. At the same time, both women and men place a high value on their time. The findings suggest that policy interventions focused on time-saving practices may offer more substantial benefits than those solely aimed at increasing income control.