Post Hoc
Also known as: ph, post-hoc-ergo-propter-hoc
Assuming that because one thing followed another, the first caused the second.
Share: also:
In plain terms
Post hoc ergo propter hoc is Latin for "after this, therefore because of this." The fallacy treats sequence as causation: B happened after A, so A must have caused B.
Sequence is necessary for causation but nowhere near sufficient. Two events can follow each other through coincidence, through a shared third cause, or because the second would have happened no matter what came before.
Why it's fallacious
Causation requires more than timing. It requires a mechanism (some plausible way A could produce B) and usually some comparison case (what happened in situations where A didn't occur). Post hoc reasoning skips both. The before-and-after is taken as the whole story, and everything else that might explain the outcome is quietly ignored.
The fallacy is everywhere. Human brains are wired to find patterns, and the temporal pattern of "this, then that" is almost automatic. Noticing the pull toward it is the first step in not being ruled by it.
Canonical example
"I started taking elderberry syrup in October. I didn't get a cold all winter. The syrup works."
The cold-free winter followed the syrup, but so did a lot of other things. The person might have washed their hands more. They might have avoided a sick coworker. The flu season might have been mild. They might have gotten colds in other years without syrup anyway. One winter, from one person, tells us almost nothing about whether the syrup is causally doing anything.
Counter-example (not a fallacy)
"Infant mortality in the city dropped 22% in the two years after the water system was overhauled. Follow-up studies traced the reduction to lower rates of waterborne illness, which matched the specific contaminants the overhaul removed. The timing is part of the argument, but the mechanism is separately shown."
This isn't post hoc. The claim of causation is supported not just by the timing but by a mechanism (contaminant removal) and by follow-up investigation that tied the drop to the specific change. Sequence plus mechanism plus evidence of the mechanism operating is how causation is established. Sequence alone is just noticing.
How to fix it
If you've been linked here, check whether your case for cause rests on more than the timing. The fix is to add at least one of two things: a mechanism (some plausible way A could produce B) and a comparison (what happened in similar situations where A didn't occur). Either one alone is much stronger than sequence by itself. If you can offer both, you have a credible causal argument. If you can offer neither, you're noticing a pattern in one data point, which is interesting but not yet evidence.
If you're on the receiving end, two questions usually settle it. "What else was happening at the same time?" and "How would the first thing actually have caused the second?" Most post hoc claims don't survive the first; the ones that do still need the second.