Building upon early theoretical work that established the underlying principles of decision making in settings involving risk (e.g. Arrow 1965; Pratt 1964; Von Neumann and Morgenstern 1944), a substantial body of experimental literature has developed investigating the theory of risk in laboratory and field settings in order to better understand agent behavior and market outcomes. These experimental explorations are motivated by the ubiquity of risk across the spectrum of decisions that agents make on a daily basis. In agricultural production, where farmers' yields and revenue are dependent upon numerous largely exogenous factors such as weather conditions and price fluctuations, risk and uncertainty is omnipresent in farming decisions. Ultimately, risk and uncertainty influence crop-selection and crop-rotation schemes (El-Nazer and McCarl 1986), technology adoption (Purvis et al. 1995), environmental degradation and policy compliance (Ozanne, Hogan, and Colman 2001), and crop insurance markets. Early experimental studies focusing on decision making under uncertainty revealed that the predominance of individuals are risk averse (Binswanger 1980). Recent experimental evidence has provided a deeper understanding of the individual-specific characteristics related to differing degrees of observed levels of risk aversion and the relationship with economic outcomes. Examples include an assessment of the link between individual-level risk preferences and socio-demographic characteristics (Harrison, Lau, and Rutström 2007), cognitive abilities (Dohmen et al. 2010) and personality attributes (Eckel and Grossman 2008). In this study we build upon this experimental evidence and explore the linkage between risk attitudes and subjective beliefs of an uncertain outcome occurring with a specific focus on agricultural losses due to weather events. Under subjective expected utility theory (Savage 1954), an agent's optimal decision in a risky setting is determined not only by their attitude towards risk, but also their subjective belief regarding the probability of an uncertain outcome occurring. This framework recognizes that in many risky settings individuals do not know the probability of uncertain events occurring, and thus make decisions based upon subjective beliefs which may not necessarily correspond with true probabilities. The underlying cognitive processes, heuristics, and individual-specific factors that shape how agents formulate their subjective beliefs, particularly in complex settings with limited information, is an open question (Gilboa, Postlewaite, and Schmeidler 2008). In particular, as we focus upon in this study, it is unclear if and how risk attitudes are related to the subjective probabilities agents perceive of uncertain outcomes occurring. That is, does an agent's degree of risk aversion influence their perception of the probability of uncertain events occurring? We assess whether there is a relationship between risk aversion and subjective probabilities by using an experimental approach with a random sample of relatively homogenous (in terms of agricultural operations) farmers facing weather risks. As detailed in the following section, lottery-choice tasks (Eckel and Grossman 2008) were used to elicit each farmer's risk attitudes, and a simple structured smoothing method (Norris and Kramer 1990) was used to elicit each farmer's subjective belief of the probability of crop losses due to weather events. Using the experimental outcomes and controlling for a number of factors that would be hypothesized to influence subjective probabilities (e.g. past experiences with crop losses, information, and communication with other farmers), regression analysis is used to assess the relationship between risk attitudes and subjective belief     «

Building upon early theoretical work that established the underlying principles of decision making in settings involving risk (e.g. Arrow 1965; Pratt 1964; Von Neumann and Morgenstern 1944), a substantial body of experimental literature has developed investigating the theory of risk in laboratory and field settings in order to better understand agent behavior and market outcomes. These experimental explorations are motivated by the ubiquity of risk across the spectrum of decisions that agents...     »