None of them knew what, exactly, was causing these symptoms. But they all shared a singular grievance: a dull aural hum had crept into their lives, which growled or roared depending on the time of day, rattling their windows and rendering them unable to sleep. The hum, local law enforcement had learned, was emanating from a Bitcoin mining facility that had recently moved into the area—and was exceeding legal noise ordinances on a daily basis.
When it comes to environmental harm, where should the burden of proof stand? How much evidence do we need before we pursue regulation, or pull back on usage? In America, we like the idea of “innocent until proven guilty,” but it doesn’t necessarily make sense to apply this logic beyond humans to novel chemicals (e.g. xylitol, PFAS) and industrial processes (e.g. Bitcoin mining, fracking).
The lesson it took me many years to learn after reading Silent Spring is that anecdotal evidence serves as an early warning, and that the cost of collecting “enough evidence” means that much more harm happens in the time you *know* something is very probably true but you can’t yet *decisively* prove it. Proving causality is hard, time-consuming, and expensive — especially given cuts to funding scientific research. This prove-the-harm approach allows companies to delay, delay, delay acting because the evidence is not 100% iron-clad.
Gas companies knew from the start that adding lead to gasoline would be harmful, but it was cheaper 🤷♀️🤷♀️🤷♀️ Generations of people have literal brain damage because of exposure to lead from burning leaded gas — because the burden of proof was on the public, to first suffer the harms and then prove they were caused by leaded air pollution.
It took decades for wolverines to be listed as a protected species under the ESA — the Trump administration claimed the data weren’t sufficient to justify their listing. There are so few wolverines that collecting data is very challenging — a classic Catch-22. In a court ruling, the overturning judge wrote:
“[T]he Service’s decision against listing the wolverine as threatened under the ESA is arbitrary and capricious. No greater level of certainty is needed to see the writing on the wall for this snow-dependent species standing squarely in the path of global climate change.” (via)
Anecdotal evidence is a faster way to identify correlations — which may or may not be causally related. Once a correlation is identified, we ought to practice greater care.
Remember the popular early pandemic phrase “abundance of caution”? Lots of people are retroactively saying that governments overreacted to COVID, forgetting that we were working with much less information than today. (I also disagree, but that’s beside the point here.) What they are really arguing is that we should assume the best case scenario rather than the worst in future situations; that we should wait for more people to die and a future virus to spread further before we take significant precautions. If you’ll recall, early in the pandemic New York City was running out of places to store dead bodies — but apparently we should have let it get worse???
The COVID pandemic is also an instructive situation on when we should have listened to anecdotal evidence faster. All signs pointed to the virus being transmitted by air — I was convinced as soon as I read the case study on the Mount Vernon choir — but the CDC wouldn’t admit to it for some time.
Ironically, conservatives are also weaponizing this mindset to go after the abortion drug Mifepristone, claiming that decades of anecdotal evidence on top of the data justifying its original approval are insufficient to demonstrate its safety.
I’ve also come to value anecdotal evidence more as I’ve recognized the foundational importance of environmental justice. Human lives are externalities from the perspective of corporate executives — so we as a society must make companies internalize these societal costs through policy. Preventing harm ought to be as important to local and state government as generating profit. Politicians today can overlook the social costs of environmental harms because we don’t have Universal Healthcare — adopting healthcare for all could make these costs internalized to government decision-making as well.
See also:
Environmental Lessons, Past and Future (on my old blog)
Re-doing the Industrial Revolution
Misadventures in leaded dishware
