Sequence is not the same as causation
The rooster crows, then the sun rises. Therefore the rooster’s crowing causes the sunrise.
Obviously absurd.
People make this exact mistake constantly with less obvious examples. The rooster version is easy to laugh at. The real-world versions slip past most people’s filters because the two events feel plausibly connected.
Post hoc ergo propter hoc means “after this, therefore because of this.”
It’s the error of assuming that because one thing happened after another, the first thing caused the second. Sequence gets mistaken for causation. Time order becomes stand-in for a mechanism. That’s a huge leap, and most of the time the leap is unjustified.
Somebody starts taking a supplement and their headaches go away.
The supplement must have worked. Or maybe the headaches were going to go away anyway. Or maybe they’ve been sleeping better. Or maybe it’s the placebo effect. Or maybe somebody finally left a stressful job the week before they started the supplement. Sequence alone doesn’t prove causation. Something else is almost always changing at the same time, and most of the time the thing that mattered isn’t the thing that got credit.
Crime went down after a new tough-on-crime mayor got elected.
The mayor must have reduced crime. Or maybe crime was already on a downward trend that started years earlier. Or maybe the economy improved. Or maybe demographic changes were underway. Or maybe the police department under the previous mayor had been building community programs that finally showed results on the new mayor’s watch. There are a dozen possible explanations. The one that takes credit is whichever one is politically convenient.
Politicians love post hoc reasoning.
They can take credit for anything that improved while they were in office and blame predecessors for anything that worsened. “The economy grew on my watch – I caused it.” “Crime rose during their administration – they caused it.” Rarely does anybody demonstrate the mechanism by which specific policies produced the observed outcome. The presidency isn’t a toggle switch that runs the whole economy. Almost nothing that happens in a four-year term can be cleanly attributed to whoever sat in the office for it.
Medical testimonials run entirely on this fallacy.
“I took this product and my symptoms improved.” The improvement is real. The testimonial is sincere. But sequence tells you almost nothing about causation. Bodies improve on their own. Conditions are cyclical. Regression to the mean means that people often try treatments when symptoms are at their worst, which is exactly when symptoms would improve anyway. So the treatment gets credit for the body doing what bodies do.
This is different from the Texas Sharpshooter fallacy from 15.19, by the way.
Texas Sharpshooter is about finding clusters in data after the fact. Post hoc is about treating sequence as cause. They can overlap, but the core move is different. Texas Sharpshooter paints the target. Post hoc assumes whatever happened next was caused by whatever happened first.
The correction for post hoc reasoning is to ask for the mechanism.
What would explain this causally? What else might have caused the observed change? Would the change have happened anyway? Has this been tested with controls that actually rule out the obvious alternatives?
Sequence is a starting point for investigation.
Not a conclusion on its own.
And when somebody claims credit for an outcome that followed their action, ask what the counterfactual would have looked like. If the thing would have happened regardless, nobody caused it. The world just did what it was already going to do.