On Risk Preferences and Curvilinear Utility Curves

Do Risk and Time Experimental Choices Represent Individual Strategies for Coping with Poverty or Conformity to Social Norms?

“…The wealth effects debate came to the attention of many anthropologists in 2001 and 2002 in a series of papers and replies published in Current Anthropology by Kuznar (2001Kuznar ( , 2002b and by Henrich and McElreath (Henrich 2001;Henrich andMcElreath 2002a, 2002b). The discussion started with Kuznar's (2001) experimental study of risk preference among Aymara herders in highland Peru.…”

“…This led to a lengthy debate within experimental economics as to the existence and salience of "wealth effects" or "income effects" on preferences (Bosch-Domènech and Silvestre 1999;Henrich andMcElreath 2002a, 2002b;Kahneman and Tversky 1979;Kuznar 2001Kuznar , 2002aKuznar , 2002bKuznar , 2007Kuznar and Frederick 2003;Lybbert and Just 2007;Rabin 2000;Winterhalder, Lu, and Tucker 1999). "Preference" refers to an individual's subjective evaluation of some objectively measurable quantity.…”

“…Kuznar interpreted this inverse-U-shaped relationship between wealth and risk preference as evidence that people's total utility functions are sigmoidal, changing from decreasing to increasing marginal utility near the borders of social status categories so that one accepts more risk when a win advances one to a superior social status. Henrich (2001) criticized that, for a decision maker at a given wealth status, a sigmoidal marginal utility curve can be drawn to predict any risk preference, depending on the size of the stakes involved. Henrich and McElreath (2002a) then published the results of risk preference experiments among four populations: Sangu agropastoralists of Tanzania, Mapuche farmers of Chile, their Huinca neighbors, and University of California, Los Angeles (UCLA), undergraduates.…”

Field Experiments and Control

“…Even if those projects do not increase average welfare for an individual, since proximity to the poverty line must presumably increase risk aversion, 15 even modest reductions in variance will be of value. Early work by Binswanger [1980][1981] in India has not been followed up in any systematic way, although field experiments in developed 16 For example, see Kuznar [2001aKuznar [ ][2001b and Henrich [2001].…”