Is Risk Preference Induction a Reliable Method of Controlling Risk Preferences?

Economic theory provides the theoretical support for many of the recent experiments studying budgeting and other management control techniques. An important assumption in many of these models specifies the risk preferences of principals and agents. In laboratory experiments, however, the measurement of risk preference has proved problematic. An alternative method, the Risk Preference Induction (RPI) technique [Berg et al., 1986], may provide a means to control risk preferences. The purpose of this study is to evaluate the RPI method, to suggest future research, and to raise issues that affect future uses of RPI in managerial accounting research. Our results indicate that RPI apparently worked better for risk-averse and risk-seeking attitudes than for risk-neutral attitudes, but the relationships among the lottery points and probabilities appear most critical to the induction's success — RPI appears to control all risk preferences except when choices are "close." Future research should subject RPI to more exploratory tests, some of which are outlined here.