Is the EPA Lowballing the Mercury Risk?

The Environmental Protection Agency recently issued rules to reduce the amount of mercury released into the atmosphere from U.S. coal-burning power plants. Some of those emissions make their way into the fish we eat and have been linked to cognitive impairment in children. They may also cause heart attacks in adults.

The new rules require an emissions reduction from the current 48 tons a year down to 15 tons by 2018. But critics say the EPA didn’t go far enough. They say the agency’s cost-benefit analysis minimized the benefits of reductions and ignored a study by the Harvard Center for Risk Analysis that estimated the benefits would be much greater. Greater benefits could justify even further reductions.

More is at stake here than mercury emissions. Cost-benefit analysis is required for nearly all major government regulations. It is a valuable method that can help identify optimal choices. But although it can be used to inform, it can also mislead, as critics charge in this case. The credibility of the government’s use of this tool for all regulations may have been damaged.

I direct the Harvard risk-assessment center. Our analysis found that cutting power plants’ mercury emissions by the amount the EPA required could yield benefits -- including reduced cognitive impairment in infants and children and fewer heart attacks in adults -- of between $100 million and $5 billion per year, depending on what share of the population suffers what degree of health damage.

In contrast, the EPA estimated less than $4 million in benefits specifically related to mercury reductions (with an additional $40 million to $44 million because mercury-emissions controls would reduce other pollutants). The EPA estimated the costs of its new caps vastly exceeded the benefits, so it wouldn’t call for further emissions cuts. Yet our study suggests that larger or faster cuts might be justified.

Two factors explain much of the difference between the estimates. First, our study included the effects of emissions reductions on the mercury we get through eating saltwater fish such as tuna and cod, which is how most of us are exposed. The EPA included only the effects on freshwater fish such as catfish, citing legitimate scientific uncertainty about how (and how quickly) reductions in emissions translate into lower mercury levels in ocean fish.

Second, the higher estimates of benefits in our report included the effects of mercury on fatal heart attacks as well as on children’s cognitive development. The EPA included only the effects on children. The evidence that mercury raises the risk of heart attacks is not conclusive, but we think it is sufficiently persuasive that the effect should be counted. The EPA deemed the science too uncertain to quantify these benefits.

Rather than omit these factors, a better approach would be to estimate how much the benefits might change if these factors were real and to express the results in a range, which is what we’ve done. The White House recently called for such an approach in analyses. Mercury may indeed cause heart attacks. That should be included as a possible outcome. We do get most of our mercury from saltwater fish. That should not be left out just because science cannot determine exactly how changes in emissions translate into changes in the levels in fish.

Although the EPA can be criticized for failing to quantify some of the potentially important effects of reducing mercury, it should be applauded for the approach it recommended for cutting emissions: a cap-and-trade solution. That means that plants are allowed to buy and sell emissions credits, which achieves a larger reduction at the same cost as setting plant-by-plant standards. It does not matter much which power plants reduce their emissions, as long as total emissions go down.

Still, it can fairly be asked whether the new EPA emissions standard is the right one, because its analysis was not sufficiently inclusive. It is also fair to ask what such a limited approach does to public trust in the process of environmental rule making. For credibility and public confidence, the government practitioners of cost-benefit analysis should be honest about the imprecision of their estimates by evaluating the full range of plausible outcomes, and by acknowledging how much their results depend on including and excluding uncertain but potentially important factors.