This dissertation focuses on mortgage defaults in Southern California during the housing bubble of the 2000s. The rapid decline in the housing market that precipitated the current recession has been accompanied by an unprecedented number of loan defaults and foreclosures. Recent studies have identified two major theories of default--the "double trigger" hypothesis, where negative equity and an income shock are necessary conditions for default--and "strategic default" where negative equity is a sufficient condition for default. This paper adds to the default literature by adding short sale as another possible outcome of mortgage default. The primary goal is to analyze the determinants of mortgage default to assist in understanding the conditions under which strategic behavior of home sales is most likely to occur. Data from Los Angeles County was analyzed from 2007 to 2010 for every closed sale, then coded into three possible sales outcomes: 1) Organic 2) Short Sale 3) Real Estate Owned (REO). A multinomial probit model was used to model homeowner decision-making based on the sale outcome. The model rejected the "double trigger" hypothesis, as it was found that income shocks do no have a significant effect on impacting the predicted probability for distressed sales. Education levels, the sales price of homes, credit card debt, and market price reductions were found to be significant variables in determining distressed sales outcomes, thereby confirming the strategic default hypothesis. The next section studied spatial association of short sales and REO to see if any contagion effects were present. It was found that both short sales and REO form into clusters of hot and cold spots. Social stigma is believed to impact consumer behavior, the theory was confirmed through the findings of contagion and spatial lag. The final section constructed a hedonic price model to capture the price effects that distressed sales have on neighborhood pricing. Foreclosures were found to have three times the negative impact on neighborhood pricing compared to short sales.

Rainfed agriculture is inherently risky, with climate change expected to intensify its variability. In the West African Sahel, where agriculture is crucial not only for subsistence but for national and household incomes through cotton production, the need to safeguard farmers' livelihoods against risk is essential. Formal crop insurance providers in such contexts cannot easily rely on traditional models, where indemnifications are based on realized losses, and have instead proposed a stream of index-based insurance products which indemnify clients based on a predefined, and yet objective parameter (the index). One promising product for Burkinabe cotton farmers is, the Double-Trigger Index-Based Insurance (2TIC), whose two-tier triggering mechanism has the potential of reducing moral hazard and minimizing basis risk. This dissertation uses three essays to consider a farmer-centric approach to assessing the implications of this double trigger mechanism for index-based insurance. The first essay explores cotton farmers' judgments of fairness vis-à-vis the 2TIC indemnification system by using Principal Component Analysis (PCA) and Logistic Regression Analyses, and examines if and how these judgments affect decisions to subscribe. The second essay assesses the impact of 2TIC on farmers' cotton-derived net income by employing Coarsened Exact Matching (CEM). The third essay compares the actuarially fair premium of the 2TIC with the commercial premium paid by cotton farmers, by using statistical approaches. The study provides important evidence-based insights into how 2TIC can be improved and promoted by incorporating farmers' needs and perspectives.