Essays on Human Behavior
Loading...
Date
2024-12-18
Authors
Holloway, Marcus
Journal Title
Journal ISSN
Volume Title
Publisher
University of Oregon
Abstract
This dissertation explores human behaviors that may confound current economic measurements in the environmental and behavioral contexts.
Chapter 2 examines disparate wildfire-smoke avoidance responses across demographic and socioeconomic groups in the western United States. Cellphone location data is combined with census block group level data to uncover heterogeneity in communities' responses to smoke. This heterogeneity is found to be correlated with historical margins of disadvantage. These communities have in many cases been documented to face higher levels of exposure to other pollutants and are more negatively impacted by these exposures. This suggests that those that have a higher benefit of avoiding wildfire-smoke also face higher costs of avoidance, leading to substantial equity concerns in terms of climate adaptation.
Chapter 3 focuses on the presentation and development of theory on time-varying time preferences, results from a novel experiment designed to measure these preferences, and structural estimation and evaluation of models allowing for these preferences. While standard economic models acknowledge the agents' ability to exhibit preference reversals, these reversals are largely attributed to a present- or future-bias in agents' underlying preferences. However, instead of myopic reversals, agents may instead be making decisions according to time preferences that drift across measurement dates. To test for this, an experiment that records subjects' decisions at seven points in time across twelve weeks is conducted. It is found that time-varying preferences are present in the sample studied and that models allowing for these preferences outperform traditional and behavioral discounting models in both aggregate fit and number of subjects best described.
Chapter 4 builds on the analysis conducted in Chapter 3 to determine what margins drive the time-varying behavior observed. Characteristics of subjects such as their demographics, socioeconomic status, risk preferences, cognition, and other behavioral factors, along with dynamic data on subjects' financial and psychological well-being were collected as part of the experiment conducted in Chapter 3. While little evidence of correlation between demographics or preferences and discounting is found, subjects do accurately perceive their time-variance. This suggests time-variance may itself be a preference that can be modeled, or that subjects can learn about their discounting process over time. In an attempt to capture this, the theory developed in Chapter 3 is used to develop a discount function that asymptotically approaches time-invariance. This model does not outperform the non-asymptotic, time-invariant, behavioral discounting model it approaches in the limit for the aggregate data. However, it does better explain a large portion of our subjects' behavior.
This dissertation includes unpublished co-authored material.
Description
Keywords
Behavioral, Discounting, Environmental, Experimental