A Philanthropic Case for PFAS

I used Claude to help draft/edit this post; all arguments were reviewed and modified by me.

Last year I asked on this forum if anyone in EA was exploring PFAS. Unfortunately I wasn't able to identify anyone so I wrote this substack piece laying out the case. The piece is not just for the EA audience so I've highlighted more of the EA pieces below. But please check out the whole piece if you're interested in learning more and would love to hear more insights on this subject. 

PFAS (per- and polyfluoroalkyl substances) are a class of more than 12,000 synthetic chemicals used in firefighting foams, food packaging, textiles, nonstick cookware, and a wide range of industrial applications. They are persistent in the environment and in the human body, they bioaccumulate, and have been detected in over 95% of Americans tested in every NHANES cycle since 1999.

Scale. An analysis covering only four PFAS across 13 health endpoints put annual EEA health costs at €39.5 billion in 2024, alongside roughly 169,600 attributable DALYs and 1,050 attributable deaths per year. A US analysis put annual disease costs at $5.6–63 billion. The EPA's analysis of the 2024 drinking water standards, limited to six PFAS, calculated $1.55 billion in annual avoidable health costs from preventing roughly 9,600 deaths and 30,000 serious illnesses. All of these should be considered highly conservative given they exclude thousands of compounds and look at only a dozen or so health outcomes.

Neglectedness. Toxics and pollution as a sector receive only 2% of total foundation grants, and PFAS captures only a slice of that already small figure. US philanthropic funding explicitly directed at PFAS likely sits between a few million and low tens of millions of dollars annually, predominantly grants rarely topping $1M. 

Why philanthropy specifically. There are billions of dollars in government and settlement dollars but they are restricted mostly to drinking water infrastructure. They rarely fund the industrial sources that keep producing PFAS, the contaminated farmland and food pathways, or the destruction of the concentrated PFAS waste that treatment generates. The 2024 federal drinking water standards that drive deployment of those dollars are now being partially rescinded, and philanthropic capital is insulated from these regulatory reversals in a way federal dollars are not.

On cost effectiveness. PFAS is a chemical class of thousands with diffuse, latent, multi-pathway exposure, which makes cost per outcome estimation harder than for single compound interventions. But the same complexity that makes PFAS hard to compare against other cause areas is itself the case for funding the work. Coordinated investment in exposure monitoring, biomonitoring registries, and integrated health cost modeling would let funders weigh PFAS against other cause areas. A few million dollars of analytical capital now could unlock orders of magnitude more.

Three areas where I believe philanthropy could be utilized:

1. Source destruction and prevention. Destruction technologies and upstream industrial emissions controls offer significantly higher leverage than downstream treatment but sit in early demonstration phases, with independent validation outside military contexts largely unfunded. This is early stage technical risk that public R&D and private investors are not positioned to cover.
2. LMIC exposure and burden shifting. As HIC restrictions tighten, production is scaling in China, India, and Vietnam. Dedicated LMIC PFAS funding is nearly absent, and the cost per ton of PFAS avoided is likely much lower there than what high income countries are now paying for remediation. A portfolio of early stage grants could shape the next decade of exposure trajectories at a fraction of the cost of HIC style remediation.
3. Community capacity. The most contaminated and least resourced communities are the least equipped to capture the federal application infrastructure, and the organizing, legal capacity, and technical assistance they need is funded by neither BIL nor settlements.