Monetary Valuation of Air Pollution Mortality: Current Practice, Research Needs and Lessons from a Contingent Valuation

The current situation of the monetary valuation of air pollution mortality is reviewed, with particular attention to the valuation of a gain in life expectancy (LE) achievable by a reduction of ambient pollution. This paper presents exploratory research centered around the application in France of a contingent valuation (CV) questionnaire developed by Krupnick et al [2002] which has been applied in the USA, Canada, Japan, and more recently in France, Italy and the UK. The original version was administered to 300 individuals, but by contrast to the application elsewhere, in France an open question was added after each set of bids and at the end of the questionnaire the WTPs (willingness-to-pay) were recalled to give the respondents the opportunity to correct their values. In addition several variants were tested on samples of about 50 each, including variants phrased in terms of LE gain. All the interviews (self-administered with a computer) were followed by written debriefing, and for the LE gain variants by face-to-face debriefing and discussions in groups of three or four, in order to better understand the perception of the questionnaire and the reasons for the responses. The results are used to provide estimates for the value of statistical life (VSL) and for the value of a life year (VOLY): they range from 0.4 to 4.1 M€ for VSL and from 0.021 to 0.206 M€ for VOLY. The WTP for a risk reduction between the ages of 70 and 80 yields a time preference rate of about 2% for mortality. However, the most important results are not the numbers but the lessons learned by debriefing and by the variants of the questionnaire. The wide scatter of the results is a reflection of the enormous difficulties that the respondents have in understanding risk reductions and replying to the WTP (willingness-to-pay) question. Thanks to the open question it was possible to measure the bias due to the starting bid: it is very large, on the order of 50% for the bids that were used. Thus the recommendation of the NOAA Panel on contingent valuation, that only the closed question should be used, is not appropriate for small risk reductions.