In this paper we compare the preferences of the general public with the preferences of public administrators working in the area of safety. We are interested in three different aspects of risk reductions: (i) large versus small accidents, (ii) actual versus subjective risks, and (iii) trade-off between avoiding fatalities and injuries for different age groups. We use stated preference surveys where respondents assume the role of a policy maker. In particular, respondents were asked to choose between different public infrastructure projects that resulted in different outcomes. When responding they were asked to take the role of a public policy-maker.



For the general public, we use survey responses from two mail questionnaires sent out between May and June 2007 to a random sample of 1400 and 2600 Swedish citizens. For the administrators, we used survey responses from an internet survey sent out in September 2008 to a random sample of 330 administrators. The administrators were working in the field of fire and rescue services both in the national and local level.



Large and small accidents

The question concerned the choice between avoiding one large accident with many deaths, or many smaller accidents with fewer dead people per accident. Both projects would in total save an equal number of lives.



A large fraction of the citizens are indifferent to both projects, but the most common response is that many small accidents should be avoided instead of one large accident. Among the administrators, there is almost an equal split between preferring to avoid many small and one large accident, and fewer think that the two projects are equally good. Thus, administrators are more likely to choose the project that will avoid one large accident, and they are less likely to say that the two projects are equally good.



Actual versus perceived risks

Respondents were asked to choose between projects with different effects on the actual and perceived risks. In one case, peoples perception of the risk is correct. In the other case people overestimate the risk. A majority of both citizens and administrators chose the alternative where the actual and subjective risk decreases in equal ratio, but a higher percentage of administrators (over 30 %) opt for the alternative where the decrease in subjective risk is higher.



Saving different groups

We find that saving the life of 1.43 10-year olds is equivalent to saving one 40-year-old from dying in accidents. Likewise, saving the life of one 70-year-old is equivalent to saving 3.31 10-year olds from dying. The social marginal substitution rate between saving a life and avoiding a serious injury is between 3.2 and 3.8 for the different age groups, thus one saved life is equivalent to avoiding around 3.5 seriously injured, which is significantly lower than the officially used value of 6 by the Swedish Road Administration. We find only a few differences between citizens and administrators preferences with respect to different age groups.

We design a donations vs. own money choice experiment and compare the results from three different treatments. In two of the treatments the pay-offs are hypothetical. In the first of these, a short cheap talk script was used and subjects were required to state their own preferences in this scenario. In the second treatment, subjects were asked to state how they believed the average student would respond to the choices. In the third treatment the pay-offs were real, allowing us to use the results to compare the validity of the two hypothetical treatments. Our hypothesis is that when subjects are asked to state how they believe an average person would respond, they will use their own preferences in their responses without using the survey situation for self-enhancement. However, we find a strong hypothetical bias in both hypothetical treatments where the marginal willingness to pay for donations is higher when subjects state their own preferences but lower when subjects state what they believe are other peoples preferences. Our explanation is that subjects use the survey situation to bolster their self-image. We also find that it is mainly women who are prone to hypothetical bias in this study

One of the more difficult ethical questions from a public decision-making perspective is whether the estimation of benefits from risk reducing projects should be influenced by factors such as age groups and risk domains. For example, should a project that saves the lives of elderly people be assigned a more different benefit value in cost-benefit analyses than one that saves the same number of children's lives? This paper examines the preferences of the general public in Sweden on these issues. We design a choice experiment in which subjects are required to make six pair-wise choices where the characteristics of each choice are accident type (fire and traffic), number of fatalities and serious injuries avoided, and age of those saved (515-, 3545- and 6575-year-olds). We find that avoiding the fatality of one 515-year-old is equivalent to avoiding 1.4 fatalities of 3545-year-olds. Likewise, avoiding the fatality of one 515-year-old is equivalent to avoiding 3.3 fatalities of 6575-year-olds. We find no significant differences between the causes of accident. One avoided fatality is found to be equivalent to around 3.5 avoided severe injuries, which is lower than the official value of 6 used by the Swedish Road Administration

The purpose of this study is to compare the value of statistical life (VSL) estimates for traffic, drowning, and fire accidents. Using a choice experiment in a mail survey of 5,000 Swedish respondents we estimated the willingness to pay for risk reductions in the three accidents. In the experiment respondents were asked a series of questions, whether they would choose risk reducing investments where type of accident, cost of the investment, the risk reduction acquired, and the baseline risk varied between questions. The VSLs for fire and drowning accidents were found to be about 1/3 lower than that for traffic accidents. Although respondents worry more about traffic accidents, this alone cannot explain the difference in VSL estimates. The difference between fire and drowning accidents was not found to be statistically significant

One of the more difficult ethical questions from a public decision making perspective is whether the estimation of benefits from risk reducing projects should be influenced by factors such as age groups and risk domains. For example, should a project that saves the lives of elderly people be assigned a different benefit value in cost-benefit analyses than one that saves the same number of childrens lives? This paper examines the preferences of the general public in Sweden on these issues. We design a choice experiment in which subjects are required to make six pair-wise choices where the characteristics of each choice are accident type (fire and traffic), number of fatalities and serious injuries avoided, and age of those saved (5-15-, 35-45- and 65-75-year-olds). We find that avoiding the fatality of one 5-15-year-old is equivalent to avoiding 1.4 fatalities of 35-45-year-olds. Likewise, avoiding the fatality of one 5-15-year-old is equivalent to avoiding 3.3 fatalities of 65-75-year-olds. We find no significant differences between the causes of accident. One avoided fatality is found to be equivalent to around 3.5 avoided severe injuries, which is lower than the official value of 6 used by the Swedish Road Administration

We design a donations vs. own money choice experiment comparing three different

treatments. In two of the treatments the pay-offs are hypothetical. In the first of these, a short

cheap talk script was used, and subjects were required to state their own preferences in this

scenario. In the second, subjects were asked to state how they believed an average student

would respond to the choices. In the third treatment the pay-offs were real, allowing us to use

the results to compare the validity of the two hypothetical treatments. We find a strong

hypothetical bias in both hypothetical treatments where the marginal willingness to pay for

donations are higher when subjects state their own preferences but lower when subjects state

what they believe are other students preferences. The explanation is probably a self-image

effect in both cases. We find that it is mainly women who are prone to hypothetical bias in

this study