Testing the Effects of Similarity on Risky Choice: Implications for Violations of Expected Utility

Our aim in this paper was to establish an empirical evaluation for similarity effects modeled by Rubinstein; Azipurua et al.; Leland; and Sileo. These tests are conducted through a sensitivity analysis of two well-known examples of expected utility (EU) independence violations. We found that subjective similarity reported by respondents was explained very well by objective measures suggested in the similarity literature. The empirical results of this analysis also show that: (1) the likelihood of selection for the riskier choice increases as the pair becomes more similar, (2) these choice patterns are consistent with well-known independence violations of expected utility, and (3) a significant proportion of individuals exhibit intransitive choice patterns predicted under similarity effects, but not allowed under generalized expected utility models for risky choice.     

Generalized Expected Utility, Heteroscedastic Error, and Path Dependence in Risky Choice

We evaluate the fit of several generalized expected utility models under homoscedasticity and three different heteroscedastic error structures for the data set first reported in Hey and Orme (1994). Standard chi-squared tests are used for nested tests, and both the Akaike (1973) information criterion and its consistent version (Hurvich and Tsai, 1989) are used for non-nested ranking of these models. A testing framework is developed that explicitly accounts for the path-dependent nature of the model selection problem. Not only does the selection of preference models depend on the error structure assumed, but the reverse is also true: the selection of the error structure depends on the preference structure assumed. An erratum to this article is available at .     

Individuals' Estimates of the Risks of Death: Part II—New Evidence

Using new survey data, we test the hypothesis that individuals' perceptions of health and safety risks are unbiased. While we find that respondents' estimates of those risks are sensitive to the information they are given to anchor their responses, we find no evidence to support the widely held view that people tend to underestimate the frequency of relatively common risks. The slight tendency for respondents to overestimate the frequency of extremely unlikely events can plausibly be interpreted as truncation bias. Overall, the accuracy of our subjects' estimates varies in a manner that is fully consistent with simple conjectures about the health and safety information that is of greatest relevance to them in the conduct of their lives. The marked difference between the results of our survey and those of previous studies appears to be attributable to the practical implications of costly information and the common failure of investigators to account for the fact that rational individuals will choose to acquire less than full information about many uncertain events.     

Performance of the similarity hypothesis relative to existing models of risky choice

Experimental studies have discovered behavior that is inconsistent with the expected utility model (EU) of risky choice (von Neumann and Morgenstern, 1953). The two approaches to address these paradoxes are tested: generalized expected utility models (GEU) and models incorporating decision-making limits or costs through question similarity. Tests are carried out over risky pairs related to well-known examples from Kahneman and Tversky's (1979) influential work. Statistical analysis reveals that GEU models of choice are significantly violated for choice patterns consistent with the similarity hypothesis. Additional tests point to shortcomings in the similarity approach that are consistent with fanning out behavior.