Risk Perception and the Gender Perspective. Hans Wolfgang Brachinger, University of Fribourg (CH); and Renate Schubert, Swiss Federal Institute of Technology (ETH), Zürich
Risk perception has been investigated under many different aspects. However, until now, gender aspects have been rarely integrated. This paper shows in which way such integration can be undertaken. Furthermore, first results concerning women’s risk perception as compared to men’s are presented.
Insights into the risk perception of women and men are important for an explanation of observable gender differences in risk behavior. In many contexts, women seem to choose alternatives which are characterized by a lower risk than those chosen by men. As a rule, alternatives with lower risk show lower average results. This means that gender differences in the riskiness of choices would give rise to differences in economic results. Therefore, in order to influence economic results, one should take into consideration men’s and women’s behavior with respect to risky choices. Being able to influence this behavior requires knowledge on the determinants of risk attitudes as well as of risk perceptions. In this paper, the main focus is on the perception aspect.
The key issue of an analysis of gender differences in risk perception consists in the construction and estimation of risk functions. In such risk functions, risk values which individuals attach to different alternatives are explained by a well defined vector of variables. Differences in risk perception could, then, be represented by differences in the variables values. Differences in risk attitudes would correspond to differences in the variables coefficients. Yet, there are mainly two approaches how to define the variables and how to construct risk functions.
According to the first approach, the so-called prescriptive or experimental approach, essentially two explaining variables are considered: one loss variable and one probability-of-loss variable. Based on lottery experiments which include these two variables, theoretical risk functions are constructed. In a next step, their relevance has to be evaluated by means of empirical fitting tests.
In the second approach, the so-called descriptive or psychometric approach, a multitude of variables which possibly determine the risk perception are transformed by means of a factor analysis into uncorrelated factors. Using a simple regression analysis model, risk functions are estimated from risk-value vectors on the one side and from vectors of the factors just mentioned on the other side.
In this paper, groups of female and male students as well as female and male managers are compared. For these groups, relevant risk dimensions are identified by means of each of the aforementioned approaches. Furthermore, the explicative power of both approaches will be analyzed. Finally, the relevance of a hybrid model built up from the two approaches will be tested. The resulting insights are important from a methodological point of view as well as from a practical one.
It turns out that dread and knowledge play an important role for risk perception. In order to study gender differences in depth, the dread-knowledge trade-off is explored in a series of experiments on how to mitigate risks.
References:
Brachinger, H.W., Weber, M. (1996): Risk as a Primitive. OR-Spektrum, to appear.
Holtgrave, D.R., Weber, E.U. (1993): Dimensions of Risk Perception for Financial and Health Risks. Risk Analysis 13, 553 - 558.
Slovic, P. (1987): Perception of Risk. Science 236, 280-285.