Source: InsideEPA.com, 1/26/11

EPA's environmental economics advisers are resisting EPA's proposed new risk-based measure for assessing the benefits of its regulations, urging the agency to adopt a more-unique benefits' estimate for each rule but struggling with the practicality of how to do that.

The advisers, members of the Science Advisory Board's augmented Environmental Economics Advisory Committee (EEAC), are reviewing a white paper by EPA environmental economists, who proposed an approach for updating the existing "value of a statistical live" (VSL) metric by which the agency measures regulatory health benefits with a new "Value of Mortality Risk" (VMR) measure for calculating the relative health benefits of agency rules.

The white paper also suggests weighting those rules that protect against cancer effects -- such as air pollution rules -- with a 50 percent multiplier.

"Our current policy focuses on accidental or general risk," NCEE's Associate Director Nathalie Simon told EEAC. "Many EPA policies [address] cancer risk, so there's a mismatch. Maybe people value cancer risks differently from other types of risk."

But during their Jan. 20-21 meeting in Washington, DC, EEAC members signaled that they will recommend much broader changes to how the agency estimates the health benefits of its rules. Several panelists said they did not like VMR as a measure, suggesting that it was too complicated for non-specialists to understand.

Several members also argued that there are many diseases besides cancer that should also be weighted because they also lead to lingering, painful illness before death, such as emphysema or Alzheimers. And several of the panelists recognized that it may be difficult for the agency to adopt their preferred approach of using two different kinds of economic benefits' studies -- willingness to pay (WTP) and stated preference studies -- when calculating benefits.

Some even suggested that the agency adopt a different risk measure -- Value of Risk Reduction (VRR) -- that varies across population groups and risk characteristics to better calculate the value of risks reduced. "[VMR] sounds like you like risk. It seems backwards," said Nicholas Flores, an EEAC member from the University of Colorado at Boulder.

By the end of the meeting, however, members were stuck upon technical difficulties of the proposal, and it is unclear to what extent they will recommend its adoption or how quickly they recommend EPA attempt to implement it. The group is already scheduled to hold two more conference calls to iron out recommendations they will submit to EPA Administrator Lisa Jackson.

The advisers' difficulty in reaching consensus could delay its inclusion in the agency's new Guidelines for Preparing Economic Analyses, a document that had been delayed since 2008. Al McGartland, director of the agency's National Center for Environmental Economics (NCEE) told the panel that EPA economists intend to replace the VSL in the Guidelines with whatever changes are proposed, McGartland said.

In response to a question from the audience, McGartland added that he intends to replace the VSL soon, depending on how quickly EEAC is able to provide its recommendations, perhaps within a year, he added. "Ultimately this will be dropped into our guidelines, and that will be a very high priority for us," McGartland said.

VSL Controversy
EPA's economic policy -- a practice for more than a decade -- has been to base its benefits calculations on the VSL metric -- a calculation of the value of a human life, which is then used to calculate the cost of environmental exposures and a rule's benefits. EPA set the value at $6.3 million in 2000 and adjusted periodically for inflation with EPA's new Guidelines, published late last month, providing an estimate of $7.9 million.

But the VSL has long been controversial because of the idea that it places a price on human life, while specialists view the estimate as a means of calculating reduction in risk to individuals in a population. EPA sources explain that the agency "withheld" certain key updates from the latest versions of the Guidelines because they have yet to be settled, including the consideration of how to update the VSL. The controversial VSL issue was also withheld from the Guidelines when they were released in draft form at the end of the Bush administration in 2008.

But NCEE's Simon told the panel that since its adoption in 1990, the VSL "has caused a fair amount of confusion and angst." She said that "economists recognize [the VSL is] an aggregation of small risks across many individuals," rather than literally placing a dollar value on a person's life.

As a result of the controversy and confusion, EPA is proposing to change its benefits metric from VSL to VMR. The white paper proposes more closely linking health benefit assessments with risk assessment, by creating a new unit for VSL, in terms of micro-risk reduction, equivalent to the 1 in 1 million or 1 in 10,000 risk levels that EPA uses in its cancer risk assessments and seeks to prevent related regulations.
McGartland told EEAC that the agency economists are "trying to merge with our colleagues in risk assessment . . . because they view risk in terms of [1 in 1 million]."

At the EEAB meeting, panelist agreed that EPA should change the metric from VSL, but suggested that the agency use VRR instead, since the benefit of value is reducing risk, rather than risk itself. "Yes, you definitely should" change the measure, said Jinhua Zhao, a professor at Michigan State University.

 "Mortality risk is what we are trying to measure. VMR does that much, much better than VSL. I strongly believe this will ease misunderstandings, but this is not enough."

Advice Unclear
But beyond the metric change, EEAC's recommendations remain unclear. They disagreed, for example, with EPA's 50 percent differential for estimating the benefits of rules that protect against cancer -- in part due to the ongoing research of one EEAC member, Trudy Ann Cameron, a professor at the University of Oregon.

Cameron indicated that her research indicates that people also value higher risk reduction for other diseases besides cancer, and that for some cancers, such as skin cancer, people are less concerned that for other types. "We've got 14 different endpoints" in our study. "There's a lot of overlap. People care about breast cancer, prostate cancer, heart disease and heart attacks," Cameron said. "Dead is dead but the way you get there matters. The name doesn't matter," but the duration of the disease and its symptoms matter.

Based on Cameron's comments, several members suggested that EPA replace the notion of one single, universal VSL with the idea of creating a new VRR for each rule or benefits estimate, tailored to the type of risk and specific population addressed in the rule. An EPA staff member, however, questions how agency economists will be able to use the suggestion.

"It's getting away from the notion there is a VSL and we just have to look for it," Cameron said. "The number should be [based on] the type of risk and the [population]. The central estimate may be right but more likely it is not, and its use can be dreadful."

"Are we constrained by the agency . . . to use the same number for transparency?" questioned Laura Taylor, an EEAC member and professor at North Carolina State University.

"This is what burned EPA before. But we need to stop sticking our heads in the sand," Cameron said, adding that it is time to acknowledge that people face different risks and have different values.

"We're already doing this with all these other things," agreed Catherine Kling, an Iowa State University professor and chair of the EEAC. "It doesn't benefit anybody even though it feels great. We've got to stick with what's theoretically and economically right."

George Parsons, an EEAC member and professor at the University of Delaware, praised the new methodology proposed by Taylor. "Take three or four studies and create a VSL specific to the project," he said. "We'll generate some variability this way. I like the idea of more judgment."

But McGartland cautioned the group that it may be difficult to incorporate the two types of studies that are used to calculate the VSL -- stated preference studies and WTP, or revealed preference studies. Stated preference studies as a group produce significantly lower numbers than willingness to pay studies, McGartland reminded the group.

"If you march down that road, there is a problem with internal consistency," McGartland said, noting that a stated preference study of a "really painful death" would have a much lower value than a revealed preference study with a "quick death." He continued, "It's inescapable if you look at the data -- stated preference [studies] are lower. If you're going to force us into a situation where we sometimes rely on stated preference studies and other times rely on revealed preference" there is no explanation for that difference.

EEAC member Maureen Cropper, a professor at the University of Maryland, suggested a solution. "We do want to move in the direction of having a VRR that varies with population and risk characteristics . . .The baseline VRR for immediate accidental death of adults, you want that number. It's useful for comparison with the current number and is relevant to other agencies. You really want to do both."

After the meeting, Kling told Inside EPA, "We really would like to push them towards doing" the specific VRRs "in the shorter term. It becomes a question of do they really have enough data."