1. Ian Savage (1993). An empirical investigation into the effect of psychological perceptions on the willingness-to-pay to reduce risk. Journal of Risk and Uncertainty 6(1):75-90.
[Journal Website]  [Manuscript Version]

The research investigates the hypothesis that the willingness-to-pay to reduce risks for specific hazards is related to the psychometric perceptions of that hazard. The motivation for the research was that economist have empirically found a wide range of "values of life" in studies that analyze willingness to pay to reduce various different risks. It is possible that some of this variation may be explained by the observation by psychologists that risk perceptions vary greatly across hazards.

The data for this study was collected as part of a wide-ranging random-dial telephone survey of 1,000 individuals in metropolitan Chicago in 1991. The study investigated attitudes towards four hazards: aviation accidents, auto accidents, fires in the home, stomach cancer. Opinions were solicited on a 1-7 scale on the respondent's personal exposure, level of knowledge (referred to as the "unknown factor" by psychologists), and nervousness ("dread") towards the four hazards, and how they would divide up $100 between organizations working to reduce the risks of each of these hazards.

Mean values for the four hazards were:

An initial conclusion is that there are certain kinds of hazards which are in a class apart from all others. These risks engender considerable fear and a high willingness-to-pay to reduce them. Previous work by the current author have found that nuclear power is one such example. The current work has found that stomach cancer also fits into this category, especially for older people. This hazard shares with nuclear power the fact that people feel that the threat is unknown.

Amongst the other three hazards considered a more consistent pattern emerged. Regression analysis found that hazards which make people become nervous when thinking about them ("dread") are generally associated with a higher willingness-to-pay to reduce that hazard. However for hazards about which the risk is considered unknown engender a lower willingness-to-pay. This latter result need not be regarded as being counter-intuitive. When people think that a hazard is unpredictable in occurrence, and that scientists do not have a good understanding of the risks, then people would prefer to spend their income on research into other hazards where there is a greater chance that preventive measures may be discovered.

The practical implications of the findings are firstly that researchers should not be disappointed at the wide variation in the valuation of life found in empirical studies. Different hazards will produce different values of life. Secondly, practitioners should not strive for, and use, a common valuation of life when making policy decisions regarding priorities for spending to reduce hazards.

A related article in a trade journal applied the findings of this research to the issue of spending on improved safety on railroads versus highways.

Ian Savage (1993). The price of saving lives. Developing Railways 1993 (A Railway Gazette International Yearbook) 23-24.
[Manuscript Version]

2. Ian Savage (1993). Demographic influences on risk perceptions. Risk Analysis 13(4):413-420.
[Journal Website]  [Manuscript Version]

The survey instrument also collected demographic information on respondents, including data on age, years of schooling, income (in six bands), sex and race. In a separate paper regression analysis was used to investigate the effects of these demographic variables on perceptions of personal exposure, knowledge and nervousness about the four hazards. The principal findings were that

• Dread and perception of personal exposure falls with: being older, having a higher income, obtaining more schooling, being male, and being black.
• Blacks feel more informed about risks of fires and autos.
• Women are better informed about cancer.
• Schooling and income unrelated to being informed about risks