ple dimensions of interest, and improving how uncertainties are characterized and communicated. The issues of multidimensional decision-making and addressing uncertainty in complex systems are discussed below. EPA’s economists are cognizant of the controversies and challenges in conducting benefit-cost studies and of the frontiers of economic research in environmental benefit—cost analysis.
Even if benefit—cost analysis were implemented based on the recommendations from Harrington et al. (2009), there are important gaps in the scope of available work on the valuation of benefits, and the literature is becoming dated. For example, a value-of-a-statistical life (VSL) approach is used to assign monetary values to reductions in mortality risk. EPA typically bases its VSL values on a 1992 synthesis of 26 published studies (Viscusi 1992). Although EPA does provide more recent references to frame the discussion, including studies of how VSL may vary as a function of life expectancy or health status, the core quantitative value remains based on old studies that are not necessarily relevant to the people most vulnerable to air-pollution health effects. Inasmuch as analyses have consistently shown that uncertainty in VSL dominates the overall uncertainty in benefit—cost analyses and given that policy choices may hinge on this value, it seems incumbent on EPA to invest in intramural and extramural research specifically on it. Similarly, with respect to morbidity outcomes, the most recent willingness-to-pay study that was incorporated into the analysis of the Clean Air Act Amendments (EPA 2011f) was conducted in 1994. In that benefit-cost analysis, multiple key health outcomes were valued by using only cost-of-illness information.
Valuation of the ecologic and welfare benefits of air-pollution reductions is similarly lacking; the only dimensions monetized are the effects of reductions in agricultural and forest productivity on the price of related goods, the willingness to pay for visibility improvements (based on studies conducted 20–30 years ago), damage to building materials, and effects on recreational fishing and timber in the Adirondacks. A recent workshop on the use of ecologic nonmarket valuation in EPA benefit—cost analysis work concluded that “perhaps the most surprising outcome was the realization of how few nonmarket ecological valuation studies are used by the EPA” (Weber 2010).
Funding for valuation research has been reduced, and disciplinary interest in valuation research, once a major topic in environmental-economics journals, has diminished. Assessing and addressing gaps in the environmental-benefits estimates should have high priority and can be tackled through research designs that produce statistically representative samples for EPA regulatory impact assessments (for the importance of standardization and sampling strategies for water see, for example, Bruins and Heberling 2004; Van Houtven et al. 2007; Weber 2010). The challenges in addressing these gaps are not trivial given budget constraints and logistics barriers to collecting public data.
Two recent EPA documents discuss ecologic-valuation challenges and strategies for the agency (EPA 2006b; EPA SAB 2009). The stated goal of EPA’s Ecological Benefits Assessment Strategic Plan is to “help improve