Per capita margarine consumptionChoking deaths on food in the US
As Americans have eaten less margarine per capita, more of them have choked to death on food, achieving a correlation of -0.995 that suggests margarine was somehow preventing choking—perhaps through its lubricating properties, perhaps through sheer willpower. The correlation is so near-perfect that it feels like a practical joke played by the universe on anyone who has ever mistaken a scatter plot for an explanation. The margarine melts, the choking rises, and the chart is immaculate.
Per capita margarine consumption has been declining for decades, falling from about 4.2 pounds per year in 2005 to under 2.5 pounds by 2021, as Americans shifted back to butter (which was nutritionally rehabilitated) and away from the trans fats that margarine once contained. Choking deaths rose as the population aged. One metric smoothly declines, the other smoothly rises, and the negative correlation is the mathematical consequence of their opposite monotonic shapes. Margarine did not prevent choking, and its decline did not cause choking. Both trends simply moved in opposite directions across the same seventeen-year window.
A correlation of -0.995 between margarine and choking deaths is a number that would make a margarine marketing department weep with joy and a statistician weep with despair. The margarine disappears from American tables, the choking increases in American hospitals, and the connection between them is the emptiest form of mathematical coincidence. I Can't Believe It's Not Causation.
As an Amazon Associate, getspurious.com earns from qualifying purchases. Learn more.
Want to learn more about why correlations like “Per capita margarine consumption” vs “Choking deaths on food in the US” don't prove causation? Read our guide to statistical thinking.