Restaurant spending per capitaPedestrian traffic fatalities
It turns out that the more money Americans spend on restaurant meals, the more of them get hit by cars, a relationship so tightly wound that a statistician looking at the correlation coefficient would have to sit down for a moment. One might have thought these two phenomena operated in entirely separate domains of human experience—one involving pleasant appetizers, the other involving considerably less pleasant meetings with traffic—but the data suggests they are locked in a cosmic embrace, moving together across twenty-one years like dance partners who have never met but cannot seem to separate.
The real culprit, almost certainly, is that both restaurant spending and pedestrian fatalities rise and fall with population and economic activity. When more people are around and more money is circulating, more of them eat out (restaurants being what humans do when they have disposable income and somewhere to go), and more of them are also walking the streets where cars happen to be. Add in the fact that urbanization intensified during this period—cities grew by roughly 20 million people between 2002 and 2022—and you have two entirely reasonable outcomes moving in synchrony, like synchronized swimmers who happen to be performing in the same pool but have never rehearsed together.
We humans are pattern-recognition machines who have somehow survived long enough to build statistics websites, which means we are very good at spotting correlations and very bad at resisting the urge to announce them. The restaurant-death relationship teaches us nothing except that prosperity and urban density travel together, a fact that would have been perfectly obvious if we had simply thought about it for thirty seconds instead of admiring the remarkable coefficient. But we didn't, and here we are, marveling at the coincidence of our own prosperity and peril.
As an Amazon Associate, getspurious.com earns from qualifying purchases. Learn more.
Want to learn more about why correlations like “Restaurant spending per capita” vs “Pedestrian traffic fatalities” don't prove causation? Read our guide to statistical thinking.