Date of Award
Doctor of Philosophy (PhD)
Paul J. Ferraro
Charles J. Courtemanche
Glenn W. Harrison
Susan K. Laury
Michael K. Price
This dissertation comprises three essays that empirically evaluate the effects of policy interventions that have been inspired by insights from behavioral economics. The essays apply experimental or quasi-experimental designs to environmental, health and labor issues, in field and laboratory settings. They analyze if these nudges are effective in market environments, and if they are likely to continue to have the impact as their popularity continues to grow. The first chapter analyzes policy banning pharmaceutical representatives from providing off-site meals to doctors. It examines the effect on the type of medications prescribed, pharmaceutical and overall medical spending, and self-reported health measures, using the Medical Expenditures Panel Survey, and individual level difference-in-differences regressions. The analysis focuses on medications which are still on-patent but are in the same class as a generic medication, and finds that the regulation significantly decreased the likelihood that an individual took one of these medications, implying savings. There is no adverse impact on rates of physical or mental health, or whether they report that health limits daily activities.
The second and third chapter use laboratory experiments to test how awareness impacts nudges. Awareness has been suggested as a way to make defaults more ethical, but could make these nudges less effective. Awareness of the impacts of a nudge could also occur as subjects gain experience, providing information simulates experience.
The second employs a loss framing nudge in a real effort task. I replicate the standard empirical finding that loss framing (penalties) increases performance (in comparison to bonuses), but show that this effect disappears when participants are aware of the nudge. Additionally, I find that the majority of subjects prefer gain framing. Awareness increases the preference for gain framing.
In the third a default nudge is employed in a carbon offsets purchase decision. All subject are given the option to buy an offset. One option is randomly selected. Subjects in the Aware condition then see information about the typical impacts of defaults. I find that the default was not effective at increasing purchases, and that information did not make it more likely that subjects change their selection.
Tracy, Josef, "Essays on Behavior, Experimental Design, Health and the Environment." Dissertation, Georgia State University, 2017.