The Environmental Protection Agency has scrapped proposed regulations that would have restricted PFAS-contaminated sewage sludge — the byproduct of municipal wastewater treatment known as biosolids — on farmland used for food production, according to agency records and legal groups. In a federal register notice issued July 6, 2026, the agency discarded its own Biden-era draft health risk assessment that found PFAS levels above 1 part per billion increased cancer and other health risks for farmers, replacing it with a determination that sludge poses no widespread threat to the public. In the same notice, the EPA advised Americans to “research” sludge companies for home garden fertilizer products.

The 1% figure that cannot be verified

The EPA’s new assessment relies on a claim that sludge is spread on only 1% of U.S. farmland — a figure that purports to justify the conclusion that national regulation is unnecessary. But the agency’s own statements undermine this claim: according to the EPA, no government agency at any level has tracked where sludge has been spread, meaning the 1% figure cannot be verified or falsified. The article does not reconcile this contradiction, and no source in the public record explains how the agency arrived at the estimate from a system that tracks nothing.

The problem is compounded by the EPA’s own draft guidance document, published June 29, 2026 — the same month as the federal register notice — which states that “nearly 60% of sewage sludge is land applied as biosolids.” This figure refers to a different denominator (percentage of sludge generated, not percentage of cropland), but the agency has not reconciled the two numbers. A reader encountering both figures in the same month would find the 1% claim difficult to square with the 60% figure, and the agency has offered no explanation for the discrepancy.

This is not a neutral data gap. The 1% figure is the mechanism by which the deregulatory decision is operationalized: an unchallenged, unsupported baseline forecloses contaminant-specific rulemaking based on land-application footprint. No party can confirm or falsify the claim because no tracking infrastructure exists — and the absence of tracking is itself a policy choice, not a natural limitation.

Who benefits, who bears the cost

A frame audit shows that the current EPA’s account — that sludge is applied to only 1% of farmland and therefore does not warrant national regulation — advances the structural interests of the waste management industry, municipal wastewater treatment plants, and PFAS manufacturers (inferred from structural position and from their legal exposure under CERCLA; the article does not name them directly), each of whom benefits from the status quo. The waste management industry, which profits from sludge disposal, lobbied the EPA against regulations, according to the Guardian’s reporting on agency records — a circular sourcing chain in which the Guardian cites its own prior reporting, with no independent corroboration located in pre-flight verification. The decision’s structure — no PFAS testing requirement, no application tracking, an individual-responsibility framing — aligns with the industry’s interest in minimizing disposal costs regardless of whether lobbying directly caused the outcome.

Those who benefit from resolution — farmers on contaminated land, food consumers, and public health advocates at the Public Employees for Environmental Responsibility (PEER) and the Southern Environmental Law Center (SELC) — bear the health costs and have low institutional leverage. The asymmetry creates a self-reinforcing cycle: no tracking system exists, so contamination is invisible; because contamination is invisible, the EPA can claim risk is low; because risk is claimed low, no tracking system is built; and PFAS accumulate in agricultural soil year after year.

The Southern Environmental Law Center’s calculation — that more than 36 out of 1,000 children drinking milk from a farm where sludge containing 9.4 ppb of PFOA was spread could develop cancer later in life from that exposure alone, and that their risk of non-cancer health problems was more than 319 times higher — is an advocacy-group risk estimate, not a peer-reviewed finding. The article attributes the figure to SELC without supplying methodology, dose-response model, or statistical confidence interval. The figure should be read as an advocacy calculation whose underlying exposure scenario is not hypothetical: Maine has documented sludge exceeding 9.4 ppb PFOA on multiple farms. The broader exposure population is unmeasured because no tracking exists, leaving the true scope of harm unknown.

How the reversal was possible

A root-cause analysis traces the decision to three structural vulnerabilities in the regulatory framework.

First, a policy gap: the 1990s biosolids framework — 40 CFR Part 503 — permits sludge as fertilizer with monitoring required for only nine heavy metals, though biosolids can contain any of 90,000 human-made chemicals. No amendment cycle requires periodic reassessment as chemical knowledge evolves. When PFAS emerged as a recognized contaminant class, no statutory mechanism existed to incorporate them into the monitoring regime. The Clean Water Act has never been amended to mandate PFAS testing, automatic threshold triggers, or a requirement to act on completed risk assessments. PEER’s pending lawsuit, filed under the Clean Water Act, alleges precisely this statutory gap — that the agency is obligated to regulate PFAS in sludge but has failed to do so.

Second, a procedural vulnerability: the Biden-era draft health risk assessment was not a final agency action under the Administrative Procedure Act. Draft assessments carry no binding force and can be superseded by a successor administration without formal evidentiary review. The Administrative Procedure Act distinguishes between preliminary drafts and final agency actions, and no statute has overridden that default for biosolids risk assessments. The EPA’s current assessment was published via a federal register notice declaring sludge posed no widespread threat — but under current law, a future administration could reverse the reversal with the same procedural ease, making the policy vulnerable to whichever party holds the White House.

Third, an absence of regulatory trigger: there is no requirement that a completed EPA risk assessment for a biosolids contaminant automatically triggers a mandatory regulatory response. The 1990s framework was designed without a site-level application data requirement, which means the 1% claim is structurally unfalsifiable — the agency cannot know whether its estimate is accurate because it built no system to generate contrary evidence. This is causation by design, not co-occurrence.

The EPA’s decision activates a shifting-the-burden dynamic: biosolids application dampens the immediate symptom of waste-stream volume while generating a displaced symptom — agricultural PFAS contamination — that loops back through the food chain and water sources. The balancing loop of regulation that would have constrained this displacement was deactivated by the policy reversal. The information flow that would have powered that balancing loop — the risk assessment itself — was eliminated and replaced with an individual-vigilance substitute, converting a collective regulatory problem into a private burden.

What the decision does not resolve

Three structural features of the problem mean that no policy announcement can eliminate the underlying harm.

The first is chemical irreversibility. PFAS compounds do not break down naturally. Each year of continued biosolids application adds permanent contamination to agricultural land that persists regardless of future regulatory action. The Maine example — more than 100 farms affected, with chemicals spread decades ago still contaminating land, including organic farms where farmers unwittingly grew crops in chemical-laden soil — demonstrates that contamination outlasts any policy window.

The second is the destroyed evidentiary basis. No federal, state, or local agency tracks sludge application nationally. The historical contamination from the untracked decades cannot be retroactively mapped. Future soil-sampling may reconstruct some sites, but the full geographic scope of PFAS deposition during the years biosolids were permitted as fertilizer is partly irrecoverable. There is no future policy action that brings the system to a state of known safety, because the contamination history is partly irrecoverable.

The third is the privatized-risk shift. The EPA’s instruction for Americans to “research” sludge companies themselves converts a risk created by federal permitting policy into an individual vigilance burden. Farmers lack access to sludge testing data the agency does not collect, do not have the technical capacity to interpret PFAS concentrations in parts per billion, and have no enforcement mechanism if a company’s product tests high. Laura Dumais of PEER characterized this as “do your own research — we’re not going to regulate anything, so if you think it’s dangerous then figure it out for yourself.” The agency’s actual language from the federal register notice is not reproduced in full by the Guardian, creating a sourcing gap that a full verification would need to resolve.

The evidence that exists

PFAS compounds are linked to cancer, liver disease, kidney issues, high cholesterol, birth defects, and decreased immunity, according to published health research cited by EPA in prior assessments. The FDA recently rejected a petition to set PFAS limits in food. A separate analysis found approximately half of California’s tested waterways were contaminated with PFAS pesticides. These documented exposure pathways demonstrate that the chemical class is already in the environment at measurable levels — the question is not whether PFAS contamination exists, but whether regulatory response will follow the evidence or substitute political judgment for it.

The question that remains open — and the answer the evidence forces

The evidentiary asymmetry between the two framings of this policy decision is not a neutral analytical feature — it is the story. The precautionary framing can cite partial data: Maine contamination evidence, the Biden-era sludge testing, SELC’s risk calculation. The regulatory-efficiency framing requires data that does not yet exist: comprehensive national tracking, longitudinal health-outcome studies on farming populations, economic modeling of disposal alternatives. The evidentiary gap does not create genuine uncertainty about the policy’s merits — it creates the appearance of uncertainty that serves the power-holding framing. The regulatory-efficiency position cannot survive evidentiary scrutiny because it requires data that was never collected, which is itself a consequence of the policy framework it defends.

The EPA owes the public answers it has not provided: What methodology produced the 1% figure from a system that tracks nothing? Why does the agency’s own guidance document, issued the same month, state that nearly 60% of sewage sludge is land-applied as biosolids — a figure that contradicts the 1% claim? What does the full federal register notice say about citizen responsibility? What do health-outcome records show for farming populations on biosolids-applied land — the data that could make the epidemiological case that the 1% framing either underestimates or accurately describes the scope of harm?

The question that remains is not whether the EPA’s decision was science-based — it was not, by the agency’s own admission, since the science it discarded found a clear harm threshold. The question is whether the political position that substituted for that science can withstand the scrutiny the agency has refused to invite.

What is not known

The source article is the sole repository of key claims in this analysis. The Guardian (Tier 2, July 20, 2026) is the sole source for the 1% figure, the no-tracking statement, the industry lobbying characterization, the SELC calculation attribution, the EPA’s “research” advisory language, and the characterization of the new assessment as “thin and watered-down.” No direct EPA document, federal register text (beyond the existence of the July 6, 2026 notice), or independent scientific source is quoted in the available package. The 1%/60% land-application discrepancy remains unreconciled in any source consulted. The SELC calculation methodology — dose-response model, exposure assumptions, statistical confidence intervals — is internal to SELC and not published in a peer-reviewed venue. The EPA’s biennial review records — whether subsequent §405(d) reviews after PFAS awareness grew specifically identified PFAS as a candidate for regulation — are not in the available package. Internal EPA deliberation on the assessment reversal is not documented. PFAS manufacturers’ lobbying is inferred from structural position and CERCLA liability exposure; specific manufacturers are not named in the Guardian article. The full geographic scope of historical biosolids application during the years the practice was permitted without tracking is partly irrecoverable — soil sampling may reconstruct some sites, but the complete deposition record does not exist and cannot be reconstructed retrospectively.

Analytical techniques used in this piece

This analysis applies the methods below. Each links to a short, plain-English explainer you can read and reuse.

Red-Team Assessment
Models a capable adversary probing a plan for the seams they would exploit.
Root-Cause Analysis
Traces a symptom back along its causal chain to the conditions that actually generated it.
Wicked Problems
Treats a problem as wicked — no stopping rule, no clean test of success, every attempt consequential.