Willingness to pay estimates for wildfire smoke health impacts in the US using the life satisfaction approach

ABSTRACT

Wildfire smoke is harmful to human health, imposing economic costs on society. Understanding wildfire smoke health costs is increasingly important in evaluating the true burden of wildfire on society, including the difficult to measure individual-level costs of smoke exposure. However, extant research on the individual-level health damages of smoke exposure are based on geographically limited case studies and make strong assumptions on agent rationality and the functioning of markets. Given that US wildfire policy is often set at the national-level, more representative estimates based on weaker and more realistic assumptions are needed. In this paper, the life satisfaction approach is applied for the first time to estimate willingness to pay (WTP) to avoid wildfire smoke induced health effects. Using nationally-representative data from the US Behavioral Risk Factor Surveillance System over 2006–2010, we estimate that US adults have a WTP of $129 [95% CI: $23, $235] to avoid one day of wildfire smoke induced health effects. WTP is largest for young adults (18–24 years) and among urban residents. The baseline WTP result is larger than the only other comparable extant estimate, suggestive that health damages may vary both geographically and according to the set of assumptions made on exposed individuals’ behavior.

KEYWORDS:

• Wildfire smoke
• health effects
• life satisfaction
• willingness to pay
• general health status

Disclosure statement

No potential conflict of interest was reported by the author.

Notes

1. For reviews of the economics literature on life satisfaction see Clark et al. (2008) and Kahneman and Krueger (2006).

2. Technically, the marginal rate of substitution estimates from this approach do not represent any specific individual's WTP, but represents an estimate of the trade-offs between income and environmental quality that will leave people, on average, equally happy (Levinson, 2012). The term WTP has been traditionally applied for simplicity, which we will follow in this study.

3. Extending Jones (2017) to smoke-induced health impacts is not as simple or straightforward as it might otherwise seem. This is because we must first isolate the relationship between wildfire smoke and health status and then separately the relationship between health status and life satisfaction. Finally, we must relate together the two separate associations in order to provide a meaningful result. To do so requires more sophisticated empirical methods than those used in Jones (2017).

4. As a further check on respondent heterogeneity by US region, we include region fixed effects in our baseline empirical specifications. This will capture any unobservable time invariant regional heterogeneity in health and life satisfaction outcomes that might otherwise bias results.

5. This censoring results in a loss of 848 counties out of a total of 3108 counties in the contiguous US identified over 2006-2010.

6. Missing, refused, and not sure responses were dropped from the analysis.

7. Models were also explored that contained macroeconomic covariates for unemployment, poverty, median income levels, and race/ethnicity. However, these variables were insignificant predictors of health status and life satisfaction after accounting for individual-level determinants and fixed effects, and thus were not included in the final models.

8. We also explored models that accounted for tobacco smoking behavior, but the main results were negligibly changed as a result.

9. Results of the Hausman specification test for each model are: (i) health status model: , and; (ii) life satisfaction model: . A significant p-value provides evidence to reject the null hypothesis that the assumptions underlying the random effects estimators are met.

10. This is because is no longer explicit in Equation (2)(2) due to the substitution. If we think of implicitly as a function of smoke event frequency, , then we are technically solving for where the last equality holds after cancelling like terms. This is the MRS between income and smoke-induced changes to general health.

11. The only other wildfire smoke WTP estimate in the US, Jones et al. (2016), is not directly comparable to ours because it is the willingness to pay to avoid any smoke-induced health effect and not a one-day effect, as estimated here.

12. Though other interpretations are also possible. Individuals living in rural areas may experience a greater frequency of wildfire smoke and at a greater intensity than urban residents owing to their closer proximity to the flame zone. In such a case, may be the more likely outcome. The extant smoke health cost literature is at present too nascent to conclude in favor of one direction over another. This is yet another reason to compare urban and rural populations in our data.

13. Urban residents are defined as those reporting to live in a city center or within a metropolitan statistical area (MSA) as defined by the US Census Bureau. Rural residents are those living outside of MSAs.

14. As one reviewer pointed out to us, while it is not possible to acclimate to smoke, it is possible that rural residents see wildfire smoke as a part life, and thusly, may have psychologically adapted. Hence, smoke exposure in rural areas may have less of an impact on self-reported health status due to such an adaptation.