![Sample our Global Development journals, sign in here to start your access, latest two full volumes FREE to you for 14 days]({"type":"image" , "src" : "/sda/5929/TOC-Subject-Sample-2021-Global-Development.jpg"})
Abstract
We extend the individual dynamic model of lifetime resource allocation to assess the monetary value given to the increase in survival probabilities for every member of a household induced by improved air quality. We interpret this monetary value as VPF (value of a prevented fatality), which can also be expressed as a flow of discounted VOLY (value of life years) lost, and account for potential altruism towards other household members. We use a French air pollution contingent valuation survey that provides a description of the life-length reduction implied by a change in air pollution exposure. By privatising the public commodity air pollution, we succeed in ruling out any form of altruism (towards others living today and towards future generations) except altruism towards one's family. We estimate a mean VOLY of €2001140,000, a 30% premium for VOLY in perfect health w.r.t. average expected health status, and a mean VPF of €20011.45 million for the respondent, all context-specific. In addition, we find an inverted U-shaped relationship between his/her age and VOLY/VPF, and significant benevolence only towards children under 18.
JEL codes:
Acknowledgements
Financial support from the French Environment Ministry (Primequal grant number 36/98), the French programme ANR-08-RISKNAT (Riskemotion grant number 007-01) and technical support from the Provence-Alpes-Côte-d’Azur Regional Council are gratefully acknowledged. We thank Sylvia Medina, Ari Rabl, Marjorie Sweetko, Miriam Teschl and two anonymous reviewers for helpful suggestions. We are also very grateful to Kip Viscusi who gave us excellent advice leading to major improvements to the paper.
Notes
1. This terminology is the most widely used, along with the ‘value of a statistical life’ (VSL), despite criticisms. Cameron Citation(2010) proposes for instance ‘willingness to swap (WTS) alternative goods and services for a microrisk reduction in the chance of sudden death’, while US-EPA Citation(2010) recommend the term ‘value of mortality risk’ (VMR) as a more accurate description of the typically small changes in individual-level risks.
2. We did a structured literature search with keywords in Scopus, and a systematic review of abstracts (or full-text) of every article cited in one of the five recent reviews on the value of life (224 studies in ASCC Citation2008; 75 in Lindhjem et al. Citation2011; 30 in Chestnut and De Civita Citation2009; 45 in Kochi, Hubbell, and Kramer Citation2006; and 27 in Dekker et al. Citation2011).
3. Note that Soguel and van Griethuysen Citation(2000) used a sample of 199 Swiss respondents to estimate an implicit VOLY based on a scenario that isolates the WTP for a one hour of life per year gain. Although their VOLY estimate is of reasonable magnitude (€39,000), we choose not to consider this study in the review because assuming that a VOLY equals 24×365 value of a life hour is disputable.
4. Economists have long known that people discount future outcomes, and ‘future years should be no exception’ (Viscusi, Hakes, and Carlin Citation1997).
5. See Chanel et al. Citation(2004) and Aprahamian, Chanel, and Luchini Citation(2007) for an overview of the other aspects of air pollution.
6. Air quality in Marseilles, the largest city of the district, was used as a reference point for all respondents.
7. Moreover, the scenario implies a good familiarity with the hypothetical market, which resembles those that respondents are used to dealing with in ‘real’ life, and better reflects the market situation. Even though other criteria are relevant in real decisions to move, the scenario makes apparent the trade-off between two criteria only (air quality and cost of living) by constraining the choice set to two cities similar in their other characteristics. This allows for a better understanding of the exact boundaries of the environmental change, and may reduce embedding effects. However, respondents might consider the less polluted city as an ideal city with no negative amenities but numerous positive amenities, which is reinforced by the fact that this city is presented as already existing.
8. Detailed descriptions of the underlying rationale as well as the exact computations from epidemiological and air pollution exposure data are given in supplementary online material 2. As noted by a referee, some respondents may have mistakenly interpreted this wording in terms of a 10-year LE gain, which would correspond to an expected gain of 1.2 month per individual for a 50% reduction.
9. The survey was designed to explore additional topics not considered here (whether respondents were influenced by public opinion, the effect of information procurement, anchoring effects, the respective weights of each type of effect of air pollution).
10. Note that as dμ is lower than 10−3, choosing the duration D = 1 year makes the approximation error lower than 10−6.
11. The trade-off was between using more tractable survival functions (see for instance Boucekkine, de la Croix, and Licandro Citation2002), which relatively poorly fit observed death rates, or keeping the Gompertz function, which fits them well but requires the proposed approximation.
12. Protest responses are respondents who express nil WTP and give a reason in open comments that can be described as protests (for instance, ‘I do not agree with the principle of paying’, ‘I would not pay since I will only move to live in the country’ or ‘I do not want to pay because the factories are the major polluters’). See Ami et al. Citation(2011) for a comprehensive discussion of protest responses in different hypothetical scenarios intended to value air pollution long-term health effects.
13. Income per consumption unit is computed based on the OECD equivalence scale, as in Leung and Guria Citation(2006) for instance. This consists in dividing household income by the square root of household size (see Atkinson, Rainwater, and Smeeding Citation1995).
14. As a sensitivity analysis, a specification based on Household income divided by household size (instead of the square of household size) was also estimated. Results however indicate that this specification performs poorly when compared to non-parametric VPF estimates (detailed results upon request).
15. Current health and expected health at age 75 were elicited using a visual analogue scale from 0 to 10. This allows us to make relative comparisons on subjective health status as well as computations of VPF based on different subjective health status, perfect health (i.e. 10) in particular, as we will see later on.
16. This is in line with Bateman et al. Citation(2008) who suggested that framing is related to lack of familiarity with the elicitation mechanism rather than preference uncertainty. Using repeated DBDC questions, they find evidence of starting-point bias for the first DBDC question but no evidence of starting-point bias in responses to subsequent WTP questions.
17. Note that for single respondents, the monetary counterpart of the expected remaining present value utility and VPF coincide by definition.