Below is the uncorrected machine-read text of this chapter, intended to provide our own search engines and external engines with highly rich, chapter-representative searchable text of each book. Because it is UNCORRECTED material, please consider the following text as a useful but insufficient proxy for the authoritative book pages.
Appendix B Environmental Protection Agencyâs Regulatory Impact Analysis for the Final Ozone National Ambient Air Quality Standard Table B-1 (EPA 2008b, Table 7-14) shows a variety of assumptions about the association between ozone exposure and mortality. The table is presented here to illustrate that EPA had included the assumption of no causal association between estimated reductions in the incidence of premature mortality and reduc- tions in ozone exposure. The committee recommends that future regulatory im- pact analyses (RIAs) give little or no weight to that assumption unless new in- formation that refutes the interpretation of this association as causal emerges (see Chapter 6). Presentations like that included in Table 7-14 should be revised in light of this recommendation. 211
TABLE B-1 (TABLE 7-14) Illustrative Strategy to Attain 0.075 ppm: Estimated Annual Reductions in the Incidence of 212 Premature Mortality Associated with Ozone Exposure in 2020 (Incremental to Current Ozone Standard, Arithmetic Mean, 95% Confidence Intervals in Parentheses)b, c, d, e Model or Assumptiona Reference National Full Attainment NMAPS Bell et al. 2004 71 (27-110) Meta-Analysis Bell et al. 2005 230 (120-340) Ito et al. 2005 310 (200-430) Levy et al. 2005 320 (230-420) Assumption that association is not causal 0 a Does not represent equal weighting among models or between assumption of causality vs no causality (see text in section 6.3.2.1 [of EPA 2008b]). b With the exception of the assumption of no causal relationship, the arithmetic mean and 95% credible interval around the mean estimates of the annual number of lives saved are based on an assumption of a normal distribution. c A credible interval is a posterior probability interval used in Bayesian statistics, which is similar to a confidence interval used in frequentist statistics. d All estimates rounded to two significant figures. As such, confidence intervals may not be symmetrical. e This table reflects full attainment in all locations of the U.S. except two areas of California. These two areas, which have high levels of ozone, are not planning to meet the current standard until after 2020. The estimates in the table do not reflect benefits for the San Joaquin and South Coast Air Basins. Source: EPA 2008b.