The EPA Stopped Pricing Lives Saved

A push alert from The New York Times circulated widely this year: the EPA "will stop considering lives saved when setting pollution limits and instead calculate only the cost to businesses." The agency's administrator called it fake news within hours. The rule behind the story is public, and it settles most of the argument.

What the rule actually says

The document is a final rule on pollution standards for new gas-fired turbines, published January 15, 2026 at 91 FR 1910.^[1] Its benefits section explains the change and then states it:

"Therefore, to rectify this error, the EPA is no longer monetizing benefits from PM2.5 and ozone but will continue to quantify the emissions until the Agency is confident enough in the modeling to properly monetize those impacts." 91 FR 1910, January 15, 2026

The stated reason precedes it: the agency's "analytical practices often provided the public with a false sense of precision and more confidence regarding the monetized impacts of fine particulate matter (PM2.5) and ozone than the underlying science could fully support."^[1]

The other half of the ledger was not touched. The rule's cost table reports net costs of negative $87.0 million to negative $33.8 million at a 3 percent discount rate, meaning the rule saves industry money, and describes those savings as the benefits of the action. Health effects appear under the heading of potential non-quantified impacts.^[1][2] What continues on the benefits side is a count of tons of pollutant, not of deaths avoided.

The corroboration question

Administrator Lee Zeldin's response was that the story was "another dishonest, fake news claim courtesy of the New York Times," and that "the EXACT OPPOSITE of this headline the actual truth."^[8][7] On a second post he called the report entirely untrue and the headline "Cute BS."^[6]

Three things bear on whether that is a rebuttal. First, the Associated Press reported the same finding on January 13 without relying on the Times, describing an agency that will stop calculating how much is saved in health care costs avoided and deaths prevented, and will focus the soot and ozone rules on industry cost.^[3][5] NPR reported it the same day from the rule and its economic analysis.^[4] Second, the rule text quoted above is the agency's own. Third, EPA's spokesperson did not deny the substance. The agency's line, given to multiple outlets, was that "Not monetizing DOES NOT equal not considering or not valuing the human health impact."^[3][6]

A Times spokesperson noted that the agency "did not deny" the reporting, which found that EPA "is no longer calculating the health benefits of reducing fine particulate matter and ozone pollution when writing clean-air regulations."^[6]

So the dispute is narrower than the word "opposite" implies. Nobody contests that the dollar figure was dropped. The disagreement is about whether an agency that no longer prices or counts deaths avoided can be said to be considering them. On that question the rule is the better witness than the press release, which is why Zeldin's claim that the reverse is true rates False while the alert itself rates Mostly True.

What "considering" means now

EPA's position has a defensible core, which is why it rates Mixed rather than false. The agency does still describe the health effects of the pollutants in its rules, it still quantifies emissions reductions in tons, and some hazardous air pollutant rules have long assessed health effects without attaching a dollar value.^[4]

Jeffrey Holmstead, who ran EPA's air office under George W. Bush, gave the strongest version of the agency's legal footing and its practical limit in the same breath. Courts require EPA to weigh costs and benefits, he told NPR, but "there is no legally enforceable requirement for them to do it in any particular way." Then the qualifier: "you can't do a sophisticated cost-benefit analysis without trying to monetize both the costs and the benefits," and this would be the first time in a long while that EPA has not tried for PM2.5 and ozone.^[4]

Richard Revesz of NYU named the mechanism that follows. "It looks good only because you ignore the main consequence of the rollback, which is the additional negative impact on public health," he said. "By just saying we are assuming no harm doesn't mean there is no harm."^[4] W. Kip Viscusi of Vanderbilt, who developed much of the underlying valuation literature, called a zero value unprecedented.^[5]

The uncertainty rationale, tested

EPA's justification is that the estimates carried more precision than the science supports, particularly the benefit-per-ton screening values, and it lists real sources of uncertainty: concentration-response functions, mortality valuation, discount rates, threshold and linearity assumptions.^[1]

Bryan Hubbell and Alan Krupnick of Resources for the Future, both veterans of this modeling, answered it point by point. EPA has characterized those uncertainties for decades through peer review, confidence intervals and sensitivity analysis, and multiple national advisory bodies endorsed the approach. EPA's own economics guidelines call for "a central or best estimate in addition to a range or confidence interval," so dropping point estimates cuts against the agency's own methodology. The value-of-a-statistical-life approach has decades of peer-reviewed support and appears in OMB guidance across administrations.^[12]

A legal analysis in the Yale Journal on Regulation made the administrative-law version of the objection: a framework that catalogs costs exhaustively while declining to quantify benefits invites an arbitrary-and-capricious challenge, and the reversal of forty years of practice was justified in about four pages.^[16]

Thirty-two senators wrote to Zeldin on February 12, 2026 with the shortest form of the argument: "Even where health benefits are 'uncertain', what is certain is that they are not zero." The letter quoted his own confirmation testimony that EPA's mission is "to protect human health and the environment," and added that you cannot protect human health if you do not count it. Documents were requested by February 26.^[13][14] We found no public EPA response.

What the change does not do

One reading of the alert goes further than the facts. The change does not mean the national air quality standards for soot and smog will now be set on industry cost. Under Whitman v. American Trucking, those health-based standards cannot consider cost at all, and technology standards like the turbine rule weigh cost among competing control options rather than against lives.^[17] What changed is the Regulatory Impact Analysis that Executive Order 12866 requires alongside major rules, the document that tells Congress, the courts and the public what a rule is worth.^[17] That is a consequential document, and it is not the standard-setting formula. The stronger claim rates False.

Nor did EPA abolish cost-benefit analysis. It removed one side of it for two pollutants.

Why the number mattered

The omitted figure was not a rounding detail. EPA's own study of the Clean Air Act found the law's benefits exceeded its costs by more than 30 to one and avoided over 230,000 early deaths in 2020 alone.^[11] When the agency tightened the annual soot standard in 2024, it projected preventing up to 4,500 premature deaths and 290,000 lost workdays, yielding up to $46 billion in net health benefits in 2032, and stated that "for every $1 spent from this action, there could be as much as $77 in human health benefits in 2032."^[10] That claim was EPA's own and it is True as the agency stated it.

Under the January practice, an analysis of a rule repealing that standard would show the dollar of cost and not the $77 of benefit. Former EPA staff at the Environmental Protection Network ran the arithmetic the agency has stopped performing: across 12 of the 31 rollbacks announced in March 2025, the prior analyses had shown roughly $254 billion a year in public benefits against $39 billion in industry cost, with nearly 200,000 premature deaths projected through 2050 if they proceed.^[20] That is an advocacy group's calculation using EPA's earlier inputs, and it is the calculation EPA now declines to make.

Interim, or permanent?

The rule frames the change as temporary, lasting "until the Agency is confident enough in the modeling."^[1] Eight months later that claim cannot be verified in either direction, which is why it rates Unproven, and the trend runs against it.

EPA said it would seek peer review for any new method consistent with OMB guidance.^[17] None has been published. Meanwhile the Environmental Defense Fund counted five rules issued without monetized health benefits by May 2026: the turbine rule, the rescission of the greenhouse gas endangerment finding and vehicle standards, a proposed delay of vehicle criteria-pollutant standards, mercury and air toxics standards for coal plants, and hazardous air pollutant standards for marine tank vessels.^[15] Harvard's regulatory tracker records the practice as a standing change rather than a one-rule anomaly.^[18]

One thing did not go the agency's way. On June 26, 2026, the D.C. Circuit upheld the 2024 soot standard, rejecting challenges from industry, states and EPA itself.^[19] The standard Zeldin moved to reconsider in March 2025 remains in force, and the analytical method for weakening standards through individual rules now omits the number that historically made them hardest to weaken.

The verdicts

EPA's mission statement is unchanged and mentions human health first.^[9] The agency has not stopped saying that pollution harms people. It has stopped saying how much that harm is worth, on one side of a ledger whose other side stays fully priced.