Nix the U-MACT

The U.S. Senate will vote Wednesday on legislation (S.J.Res.37) to overturn the EPA’s Utility MACT Rule, a regulation establishing first-ever maximum achievable control technology (MACT) standards for hazardous air pollutant (HAP) emissions from power plants.

Like EPA’s New Source Performance Standards (NSPS) for greenhouse gases, the U-MACT would effectively ban the construction of new coal-fired power plants. That is a policy Congress has not approved. Indeed, Congress declined to pass a milder variant of the same anti-coal agenda when Senate leaders pulled the plug on cap-and-trade legislation in 2010.

In September 2011, the House of Representatives passed the Transparency in Regulatory Impacts on the Nation (TRAIN) Act (H.R. 2401) by 249-169. The TRAIN Act would nullify the U-MACT Rule and direct EPA to adopt less burdensome regulations to control power plant HAP emissions. Clearly, had the U-MACT been introduced as legislation, it would have been dead on arrival. Yet supporters of S.J.Res.37 are struggling to get the 50 votes needed for passage. Why is that?

Partisanship is an obvious factor. Some Democrats will vote against S.J.Res.37 simply because the U-MACT is the Obama administration’s handiwork. Another factor is the elitist notion that the ‘best and brightest’ are entitled to enact their policy preferences via backdoor regulation when they can’t persuade Congress to pass ‘progressive’ legislation.

Supporters of S.J.Res.37 must also overcome years of hype about power plant emissions of mercury, the principal HAP targeted by the U-MACT Rule. A typical example is a recent Sierra Club billboard that shows a pregnant woman cupping her belly in her hands. An arrow pointing to her womb contains the following text: “Just where does all that mercury from our nation’s coal-fired power plants end up?”

The ad implies that all, most, or at least much of the mercury emitted by power plants ends up poisoning unborn children. Fact check! The “majority” of U.S. industrial mercury emissions are deposited outside the U.S. (EPA, Report on the Environment, p. 2-46). A fraction of the emissions deposited domestically ends up in lakes and streams. A “small fraction” of the aqueous mercury is converted to the organic form, methylmercury, which can accumulate in aquatic food chains (EPA, Regulatory Impact Analysis, p. 4-16). A fraction of the methylmercury ends up in fish tissues. And a small fraction of fish in U.S. lakes and rivers is actually eaten by pregnant women.

EPA’s messaging is less fanciful but still alarmist. EPA contends that pregnant women in subsistence fishing households consume enough mercury in self-caught fish to impair their children’s cognitive and neurological development. Although that is theoretically possible, in the 22 years since Congress tasked the EPA to study the health risks of mercury, the agency has not identified a single child whose learning or other disabilities can be traced to prenatal mercury exposure.

The EPA’s December 2000 “appropriate and necessary” determination, the trigger for the Utility MACT Rule, depicted power plant mercury emissions as a significant and growing public health threat. That was sheer exaggeration.

The EPA projected that power plant mercury emissions would increase from 46 tons per year (TPY) in 1990 to 60 TPY in 2010. In fact, emissions declined to 29 TPY in 2011 – 50% below EPA’s projection.

Citing a 1999-2000 Centers for Disease Control and Prevention (CDC) survey, EPA’s “appropriate and necessary” determination stated that 7% of childbearing age women in the U.S. (one in every 14) had blood mercury concentrations exceeding the agency’s “reference dose” (the ‘safe’ exposure level). But the relevant subpopulation is pregnant women, not women of childbearing age (defined as 15 to 44 years). Childbearing age women are older, on average, than the average pregnant woman, and blood mercury levels increase with age. In the CDC’s 2001-2004 surveys, only 0.4% of pregnant women (one in every 250) had blood mercury levels exceeding the reference dose (Schwartz and Hayward, Air Quality in America, p. 169). More critically, the EPA’s reference dose is not a valid measure of actual health risk. The reference dose is 1/15th the lowest exposure level – a value known as the “benchmark dose” – associated with mild, subclinical effects in any epidemiological study. The highest exposure measured in any pregnant woman by the CDC during 1999-2004 was 3.7 times the reference dose – about 1/4th the benchmark dose. Serious harm such as neurological disorders requires exposures much higher than the benchmark dose.

In the U-MACT Rule, the EPA assumes that any increase in prenatal mercury exposure above the reference dose produces a corresponding decrease in the child’s IQ. Here the EPA relies on a single study funded by the agency and led by an EPA scientist. The study purports to be an “integrative assessment” of epidemiological studies conducted in the Faroe Islands, the Seychelles, and New Zealand. However, the Seychelles study – the most reliable of the three – found no association between prenatal mercury exposure and IQ even though exposures were as high as 22 times the reference dose.

The EPA estimates that implementing the U-MACT Rule will cost $9.6 billion in 2016, and that the required mercury reductions will provide $0.5 to $6 million in health benefits in the same year. The agency does not even try to estimate the benefits of the Rule’s MACT standards for other HAPs such as chromium, nickel, and acid gases. For the HAP reductions that are the U-MACT’s statutory purpose, estimated costs exceed estimated benefits by 1,600 to one or even 19,200 to one.

Even those numbers give the U-MACT too much credit. The EPA’s benefit estimate assumes that mercury emission reductions achieved in 2015 yield proportional reductions in fish tissue concentrations in 2016. Yet the agency admits that fish tissue concentrations may not decrease for “years to decades” (EPA, RIA, p. 4-18).

The benefit estimate further assumes that the hoped-for reduction in fish mercury concentrations will avert the loss of 0.00209 IQ points per child born in 2016 in a guesstimated population of 240,000 subsistence fishing households (EPA, RIA, p. 4-56), and that those ‘saved’ IQ points will boost aggregate lifetime income by $0.5 million to $6 million. This is not verifiable even in principle. IQ cannot accurately be measured out to five decimal places. Consequently, it is also impossible to determine whether any relationship exists between income and IQ for increments as tiny as 0.00209 IQ points. In short, the EPA’s benefit estimate for the U-MACT’s mercury reductions is a statistical figment.

The EPA nonetheless assures us the Rule will pay for itself many times over. This supposedly is due to the “co-benefits” of coincidental reductions in non-HAP emissions, particularly sulfur dioxide, which is a precursor of fine particulate matter (PM2.5). The EPA estimates that in 2016, the Rule’s coincidental PM2.5 reductions will avert 4,200 to 11,000 premature deaths, generating co-benefits of $33 billion to $89 billion, or $3 to $9 in health benefits for every dollar of cost. None of this is credible. As Anne Smith of NERA Economic Consulting points out, almost all of the projected 11,000 premature deaths averted are in areas already in attainment with the EPA’s National Ambient Air Quality Standard (NAAQS) for PM2.5. By law, NAAQS are set at a level “requisite to protect public health” with an “adequate margin of safety.” Last week, the EPA proposed to revise the NAAQS for PM2.5, to make it more stringent. Still, more than 94% of the U-MACT’s estimated co-benefits are attributable to PM2.5 reductions below the revised NAAQS. Anne Smith notes that the EPA attributes up to 89% of the Rule’s co-benefits to PM2.5 reductions below the lowest exposure associated with mortality risk in any epidemiological study.

In short, the U-MACT Rule’s advertised health benefits exist only in EPA’s modeling, not in the real world.

In stark contrast, the Rule’s costs are real and substantial. U-MACT will raise electricity and natural gas prices, imposing a regressive de facto energy tax on low-income households. The utility sector will become more dependent on natural gas, a fuel with a history of price volatility and a future clouded by the environmental movement’s hostility to hydraulic fracturing. Premature retirements of up to 50,000 megawatts and more combined with hundreds of thousands of megawatts that must be taken offline to install new pollution control equipment create significant reliability challenges for grid operators, increasing the risk of power failures, rolling blackouts, and brownouts.

Most importantly, U-MACT will do more harm than good to public health. NERA Economic Consulting estimates that U-MACT and three other EPA regulations could reduce annual average disposable income by $34 billion from 2012 to 2020. That is money households won’t be able to spend on health care, nutritious food, and other priorities. NERA also estimates that the four EPA regulations could reduce net employment by an average of 183,000 jobs per year. Many people who lose their jobs also lose their health insurance. Numerous studies demonstrate that unemployment increases the risk of sickness and death.

To sum up, the U-MACT Rule would have zero chance of passing were its mandates introduced as legislation and put to a vote. The Rule effectively bans new coal generation – again, a policy Congress would reject if proposed in legislation. Along with other EPA regulations, U-MACT will make energy more costly, make electricity prices more volatile, jeopardize electric supply reliability, and reduce GDP growth and job creation. The alarms over mercury and PM2.5 that EPA and others use to promote the Rule are unscientific hype. Because households generally use income to enhance their well-being, regulations with large net costs are likely to have a negative impact on public health. For all these reasons, Congress should nix the U-MACT.