Zhao, Shuoli2020-09-222020-09-222018-07https://hdl.handle.net/11299/216334University of Minnesota Ph.D. dissertation. July 2018. Major: Applied Economics. Advisors: Chengyan Yue, Terry Hurley. 1 computer file (PDF); viii, 108 pages.This dissertation aims to address a few recent theoretical and methodological developments to better understand individual decision-making. Consumer purchases of community supported agriculture are decision-making under risk. In this case, essay one incorporates prospect theory in behavioral economics to a discrete choice experimental design, and simultaneously estimate consumer preferences for product attributes and risk parameters including loss aversion, diminishing sensitivity, and probability weighting. Not only consumer purchasing decisions involve risk and uncertainty, producers’ production decision-making is also affected by individual risk preferences. Thus, in the second essay, we explore the preference differences between conventional crop producers and specialty crop producers using behavioral economics models, and the results shed lights on risk mitigation and policies related production contract design. Given the recent developments in estimation methods, the third essay assesses US households’ organic produce purchases using the method of machine learning. This study compares the predictive accuracy of organic budget share between econometric models and machine learning methods, which provides some initial insights into the effectiveness of using machine learning methods to estimate household demand.enCommunity Supported AgricultureMachine LearningProspect TheoryRisk PreferenceSpecialty CropThesis or Dissertation