Back to Blog

Correlation, Causation, and the Overton Window That Only Shrinks

If you spend enough time around superficially intellectual people, you will have heard the phrase. Someone points at a pattern, someone else says “correlation doesn’t imply causation,” and the conversation ends there, usually with a small, satisfied smile. When I picture the moment, there is a specific face attached to it: the smug pseudo-scientist, chin slightly raised, having just deployed the one piece of statistics they remember from a first-year course as a conversation-ending device rather than a conversation-starting one.

What bothers me is not the sentence. The sentence is true. What bothers me is that people only ever remember the first half of it, and that this is not an isolated case. It happens to aphorisms constantly, and it is almost never innocent.

“The customer is always right” is missing its second clause: in matters of taste. The original point was narrow (do not argue with someone about what they like) and got flattened into a blank check for entitlement. “Blood is thicker than water” is reputed to descend from “the blood of the covenant is thicker than the water of the womb,” which, if true, means the proverb originally argued the opposite of what everyone now uses it for: chosen loyalty over accident of birth, not the other way around. “Curiosity killed the cat” drops “but satisfaction brought it back,” which turns a warning into a complete thought about risk and reward instead of a blunt instrument for shutting down questions. “Money is the root of all evil” quietly deletes “the love of,” a difference Paul actually cared about making. In every case, the second half is where the nuance lives, and the second half is exactly the part that gets dropped when people want to use the phrase as a stop sign rather than a signpost.

I think “correlation doesn’t imply causation” belongs on this list, and I want to use this article to make the case for its missing half.

What people are actually doing when they say it

My honest belief about the motivation behind this specific phrase, deployed in the specific reflexive way it usually is, is that the person saying it has something to protect. Not always consciously.

Sometimes it is closer to a Freudian slip than a statistical objection: a phrase produced under mild pressure that reveals more about the speaker’s discomfort than about the data.

Here is the pattern I mean. Every time you use the lab equipment, something breaks the next day. You mention this. The response: “correlation doesn’t imply causation.” Technically correct.

It is also, functionally, an attempt to make sure nobody keeps pulling on that thread. The phrase is not being used to advance an inquiry into what is actually going on. It is being used to close one down.

It marks a boundary: this direction of thought is uncomfortable to me, and I would like you to stop walking in it.

Tabea (@tabeatheunicorn) and I have talked about this a lot, especially in the context of our alma mater, because a university is precisely the place where this move should never work. A university is supposed to be a sandbox for ideas. As a researcher you cannot actually do very much to the world directly. Mostly you can write a paper, and maybe someone reads it, and maybe they don’t. Given how limited the actual leverage is, putting artificial shackles on your own thinking, inside the one institution explicitly built to remove them, seems like a strange thing to volunteer for.

So: should someone be allowed to think outside the current boundaries of the status quo? Probably.

I want to approach that question from two directions at once, because I think both of them point at the same underlying structure.

First angle: the shrinking window is a design property, not an accident

The setup

Imagine a world with one rule: say anything you want within a certain boundary, and the instant you say something outside it, you are eliminated from the conversation permanently. The boundary itself is set democratically: whatever more than half the population currently finds acceptable defines the edge. To make this tractable, assume outrageousness can be projected onto a single axis. (If the space is actually multidimensional, assume everyone uses something like a norm to collapse it onto one axis anyway. It does not change the conclusion.)

At first glance this looks fair. Democratic input, majority rule, the boundary reflects genuine consensus.

Why it can only shrink

Here is the catch. In a system where speaking outside the line gets you instantly and permanently removed, the window can only shrink or stay the same. It can never expand. Nobody is ever rewarded for testing the edge, because testing the edge is fatal. Growth is structurally impossible.

Will it actually stay the same, though? Probably not, and here is the mechanism, step by step:

1. Model people’s expressions as roughly Gaussian along that one axis. Assume people mostly stick to their own strategy of speech, occupying roughly the same bin over time.
2. They do adjust, though, based on their personality and their neighboring bins. Someone easily moved by social pressure drifts toward the mean. Someone who is not sometimes drifts further out, looking for a niche where the conversation is still interesting.
3. If they drift too far into the tail, they get eliminated. But elimination is not instant. There is a delay: something has to be said first, and the consequence arrives after a lag.
4. So the people in the neighboring bin do not see their neighbor vanish the moment they speak. They notice only after the delay has passed.
5. When they do notice, they know their now-absent neighbor was sitting near the edge. Conversations are messy, and it is genuinely hard to know in real time exactly which sentence is the one that gets you removed. The safe strategy is to hedge inward, to move a little closer to the center, just in case.

Multiply that inference across a whole population making the same calculation repeatedly, and you get a standard deviation that compresses over time. The tails get pruned faster than the center refills them, because refilling the tail requires someone brave (or reckless) enough to occupy it again, right after the population was just handed a fresh, painful reason not to.

The Overton window, under this model, does not stay put. It contracts. The only thing that could counteract this is a supply of what I’ll call martyric psychopaths: people sufficiently indifferent to elimination that they keep re-occupying the tail regardless of the cost. Absent that supply, the window is a one-way ratchet, and it only ratchets in one direction.

Second angle: the tunnel effect and the cruelty of a fixed boundary

Here is the second angle, and I will be upfront that I do not have an airtight argument for it. Treat it as a thought experiment and decide for yourself how far to trust it.

The model

Model the Overton window as an infinitely deep potential well, and the distribution of things people actually say as an electron sitting in the well’s ground state. The naive read: since the well is infinitely deep, nobody’s expression ever ends up outside it.

That is not actually true, and the reason it is not true is instructive.

The quantum mechanical analogy is the tunnel effect. The probability of finding the electron outside the well is strongly suppressed the further out you go, but it never goes to exactly zero, no matter how deep or how sharp the walls are.

Translate that back: even with a completely fixed, unmoving boundary of acceptable speech, there will always be some nonzero density of people saying things over the line. Not because the boundary moved. Because the underlying distribution has a tail that never fully vanishes, no matter how steep you make the walls.

Why that makes elimination a cruelty, not a filter

Given that this seems true regardless of where you draw the line, the actual question becomes: is it wise to eliminate the people who show up in that tail?

If the tail is a structural, permanent feature of any population large enough to matter, and you eliminate everyone who falls into it as a matter of policy, you have not designed a system with an edge case. You have designed a system with a permanent, self-renewing supply of victims.

That is a design property, not a bug you can patch out. It is a form of cruelty built into the architecture, aimed at people who were never doing anything other than what the distribution guarantees someone will eventually do.

The presupposition hiding underneath both thought experiments

Both of these thought experiments share an implicit premise I want to name directly: nothing outside the currently known can ever be useful or good. That is the assumption doing all the quiet work.

If you actually believed that everything worth knowing is already inside the window, elimination at the boundary would be costless. You would not be losing anything by pruning the tails, because the tails, by definition, contain nothing of value.

I do not believe that premise. Showing up to a university is itself an implicit rejection of it. Attending one means you believe there is something out there, currently unknown, worth discovering, worth the risk of moving toward. That is the whole justification for the institution’s existence.

Not moving forward is not a neutral, stable resting state. It is the precondition for falling behind, because everyone else’s window is either expanding or contracting too, and a system that structurally cannot expand is, relative to one that can, already losing.

Refining the sentence

So here is my refinement of the phrase this whole article is built around: not every correlation implies causation, but every correlation deserves to be investigated.

Judea Pearl makes a version of this argument with far more rigor than I can bring to it here. When two variables move together, there are only a small number of structural explanations available:

1. X causes Y.
2. Y causes X (reverse causation, easy to miss if you assume the arrow only points one way).
3. A third variable Z causes both (a confounder).
4. Selection bias or measurement artifact (the correlation is a property of how you looked, not of what is actually there).
5. Coincidence (with enough variables and enough tests, some correlations are guaranteed to appear by chance alone, see Tyler Vigen’s Spurious Correlations for the reductio ad absurdum of this).

The phrase “correlation doesn’t imply causation,” said and left there, correctly rules out jumping straight to option one. What it does not do, and what the smug delivery conveniently obscures, is rule out options two through five.

So how do you actually test for which of the five you are looking at, in real life, without a physics lab and a controlled universe? A few concrete moves, roughly in order of how much they cost you:

• Check temporal order first. If Y consistently precedes X, X cannot be causing Y. This kills more bad hypotheses than people expect, and it costs nothing but a careful look at your own data.
• Look for a plausible mechanism. Not proof, but a candidate story for how X would produce Y, mechanically. If you cannot even sketch one, that is evidence the correlation is closer to option 4 or 5 than 1 or 2.
• Ask whether the effect scales with the cause. A genuine causal relationship usually shows some kind of dose-response pattern: more X, more Y, in a way that tracks. A spurious correlation usually does not survive this kind of stress test.
• Look for a confounder deliberately, rather than waiting for one to be pointed out to you. What third thing could plausibly move both variables at once? If you can name a candidate, control for it statistically or by design and see if the correlation survives.
• Find or construct a natural experiment. Something in the world occasionally varies X for reasons that have nothing to do with Y (an instrumental variable, a policy change, a discontinuity). If the correlation still holds when X moves for an unrelated reason, that is real evidence for causation, not just correlation.
• Run the actual intervention if you can. Change X directly and observe whether Y moves. This is the randomized controlled trial, the gold standard, and it is expensive and often impossible outside a lab, which is exactly why the four steps above matter: they are what you do when you cannot afford this one.
• Replicate. If the pattern only shows up once, in one dataset, treat option 5 as the leading hypothesis until someone finds it again independently.

None of this is exotic. It is closer to the Bradford Hill criteria used in epidemiology than to anything requiring a PhD in causal inference. The point is that it is work, and the phrase “correlation doesn’t imply causation,” used as a mic drop, is specifically the sentence people reach for when they want credit for rigor without doing any of it.

What I actually want from you

So the next time you hear the phrase deployed as a full stop rather than a comma, I would like you to think about this article and do the exploratory thing instead. Ask which of the five categories you are actually looking at. Check the temporal order. Look for the mechanism. Go find the confounder before someone else has to hand it to you.

There is no instant death penalty for speech in most rooms you will find yourself in, university or otherwise. But if the shrinking-window argument above is right, that only stays true if people keep using the full width of the window rather than quietly retreating toward the middle every time someone drops a conversation-ending aphorism on them. The window does not maintain itself. It has to be used, fully and repeatedly, or it contracts on its own. That is not a metaphor I am fond of. It is, as far as I can tell, the actual mechanism.

If you want to push back on any of this, come find me on Discord.