The measurement of people's values for public or non-marketed goods - or for the loss of these goods - is playing an increasing role in benefit-cost analysis for public policy decisions and regulations and in assessment of liability for environmental damage. A widely used measure of these values, judgment of willingness to pay (WTP) for public goods, seems to be susceptible to a number of biases and irrelevant influences, particularly when it is elicited by a survey instrument. An example is the embedding effect, in which the valuation of a good such as disaster preparedness depends on whether the entire program containing that good (environmental services) has been evaluated first. The research will conduct questionnaire studies designed to isolate crucial features of WTP with particular attention to the role of moral norms. Alternative elicitation methods, such as direct ratings of goods, will be explored, with a view to understanding how unbiased evaluation might be achieved. Other variables for investigation will include provision of context, type of choice rule, type of good (environmental or consumer), number of other people affected, fairness of the payment vehicle, and expectation of others' responses.