Win Some, Lose Some: The Effect of Chronic Losses on Decision Making Under Risk

Abstract

Losses, including those that are chronic in nature, are a fact of life. The research reported here was designed to examine, using a controlled experiment, the effect of chronic losses in a given contextual domain on subsequent decisions with uncertain outcomes that take place in the same and in unrelated domains. Randomly selected adult subjects who took part in the experiment were randomly assigned to one of three treatment groups: One group, chronic losers, was exposed to chronic financial losses as part of a controlled, multi‐round gambling simulation. Groups two and three were exposed to chronic wins and random outcomes, respectively, as part of the same gambling simulation. Results from the experiment revealed that chronic losses, in contrast to random outcomes and chronic wins, had clear effects on decision making in the domain where the initial losses were incurred. Subjects who were exposed to the chronic loss induction demonstrated a significantly higher level of risk aversion when compared with subjects who were exposed to either random outcomes or chronic wins. Subjects exposed to chronic losses also displayed a depressed affective state and a tendency to accept less as an outcome of future decisions, and still consider it to be a satisfactory result, when compared to subjects in the two control conditions. There appears to be no spillover, however, of a similar degree of risk aversion when considering similar kinds of decisions in unrelated contextual domains. These results seem consistent with prospect theory and the theory of learned helplessness, and have implications for risk communication and management in a variety of contexts.

Keywords:

• Chronic loss
• learned helplessness
• representativeness
• affect
• decision making

Acknowledgments

This research was supported by the National Science Foundation under grant numbers SES‐0519551 and SES‐0350777. Any opinions, findings, and conclusions or recommendations expressed in this publication are those of the authors and do not necessarily reflect the views of the sponsors. The research reported here benefited from the comments and insights of Melissa Finucane, Robyn Wilson, Hal Arkes, and Paul Slovic.

Notes

1. This situation led to the filing—by a group of African American farmers—of a class action lawsuit citing discrimination against the USDA; the case was settled in 1999 for $3 billion. Since 1999, however, 90% of those awaiting compensation from the USDA were denied payment resulting in the filing of a second lawsuit—also citing discrimination—in which the plaintiffs are seeking $20.5 billion.

2. Previous research (e.g., Bohm et al., 1997) has shown that the opportunity to practice a BDM‐type questions increases the validity of the final result.