It turns out that as Americans dug less coal out of the ground between 2002 and 2022, fewer of them got hit by cars, which suggests either that coal dust was making us reckless jaywalkers or that the universe maintains a strict conservation law whereby every ton of bituminous we stop extracting must be balanced by a pedestrian who doesn't get flattened. The correlation is so tight it feels like a cosmic joke written by someone with a very specific grievance against pattern-seeking statisticians.
What's actually happening is far more mundane and therefore oddly fascinating: both trends ride the same economic wave. Coal production crashed during the 2008 financial crisis and never really recovered as natural gas became cheaper and renewables got subsidized, while pedestrian fatalities actually declined thanks to better road design, smartphone awareness (paradoxically), and aging infrastructure in coal-dependent regions where populations were aging or leaving anyway. A coal mining town that loses 40 percent of its working-age population between 2005 and 2015 will naturally report fewer pedestrians getting hit by cars, simply because there are fewer pedestrians, full stop. The real third variable here is probably economic decline in specific regions, and the fact that we can draw a line through twenty-one points so tight it almost looks intentional is really just testimony to how thoroughly interconnected—and how thoroughly declining—coal country became.
This is what happens when you let statistics loose in a world full of correlations but relatively short on causation: two completely independent systems bump along together for twenty years, creating the phantom sensation that one is dragging the other around by the nose. We are, it turns out, pattern-recognition machines first and careful thinkers second, which explains both why Spurious exists and why you found yourself nodding along just now. Coal and pedestrians: bound by the chains of economic history, not physics.
As an Amazon Associate, getspurious.com earns from qualifying purchases. Learn more.
Want to learn more about why correlations like “US coal production” vs “Pedestrian traffic fatalities” don't prove causation? Read our guide to statistical thinking.