Essays on reference-dependence in natural resources and agriculture

Abstract

This dissertation studies reference dependent producers in the primary industry. I investigate empirical evidence of how goal-oriented motivations shape producer responses to realized shocks and provide theoretical insights into how adaptability influences decisions prior to the realization of productivity shock. Chapter 2 presents an analysis of expected and realized fishing behaviors among Alaskan halibut skippers using unique survey data. The surveys are matched with logbook records during a trip, enabling me to compare pre-trip fishing plans and realized behaviors from 444 trips. The findings suggest the influence of reference dependence preferences on fishing strategies, such as extended exploration at sea when faced with unexpectedly low catches. Additionally, this chapter examines factors that influence the formulation of reference points and emphasizes the importance of context-specific data in constructing proxy reference points. Chapter 3 further confirms the presence of reference dependence among skippers using the same dataset. By exploiting the reported planned harvests as subjective reference points, the study assesses the effect of reference dependence on the skipper's working duration. A hazard model analysis demonstrates the skipper is significantly less likely to continue working once the harvest goal of a fishing trip is achieved. Furthermore, a mixed logit model provides additional evidence of the prevalence of reference dependence among skippers. Chapter 4 develops a theoretical model that incorporates a multi-stage production framework with reference dependence utility. This model explores sequential input choices before and after uncertainty resolution and uncovers two key findings. Firstly, reference dependence leads to divergent responses to price or productivity changes locally before these changes assure to surpass the reference outcome, as trade-offs between profits and meeting the reference occur to achieve the reference. Secondly, the analysis suggests that loss aversion may contribute to the low adoption of high-yielding inputs due to a higher probability of falling short of the reference outputs.