The red light and the zero-cost mind

This morning, walking to the office here in London, I found myself observing a scene that repeats every day: one pedestrian crosses on red. Nothing extraordinary, except that within two seconds another ten follow without even looking at the light. Same script at the tube: three escalators available, but everyone crowds onto the one where the first passenger stepped off. The other two remain half-empty.

It's not stupidity. It's something far more interesting.

The energy-saving brain

Every decision we make costs. Not metaphorically, it literally costs cognitive energy. Your brain consumes about 20% of your body's total energy whilst representing only 2% of its weight. It's an expensive machine to run, and it's evolved to optimise consumption.

When you're faced with a red light and an apparently clear road, your brain should perform a series of calculations: speed of cars, distance, reaction times, probability of accident. Exhausting. But if someone has already crossed and hasn't been hit, that person has just solved the decision problem for you. Free.

Following costs zero cognitive energy. Deciding autonomously costs a lot.

That's why the first person who moves triggers an immediate cascade. It's not that the other ten have verified whether it was safe, they've simply accepted the "it's safe" signal that the first transmitted through their behaviour. It's information that propagates without anyone evaluating it independently.

The savannah is still with us

There's a precise evolutionary reason for this mechanism. On the African savannah where we evolved, if a member of the group suddenly started running, you had two options: stop to rationally assess whether there was really danger, or run too and ask questions later.

Those who chose option one were occasionally right, false alarm, no danger. But that time they were wrong, they got eaten. Those who chose option two might have run unnecessarily a few extra times, but were still alive to tell the tale.

We inherited the survivors' strategy: when there's uncertainty and you see others acting, follow before understanding. It's a reflex older than the prefrontal cortex, faster than rational thought.

The problem is that this reflex worked brilliantly when dangers were real and immediate. It works much worse when the "danger" is a red light on the Strand, or when the behaviour you're imitating isn't necessarily optimal.

When everyone goes left

The dynamic becomes even more fascinating when you consider that genuine assessment from the first actor isn't even necessary. They themselves might be following a rough heuristic, but their movement communicates to others "I've assessed and it's fine", even if it isn't true.

This creates what are called information cascades, which I also discussed yesterday: the first crosses because they're in a hurry, the second because they saw the first, the third because they saw the first two (and therefore thinks "if two have assessed..."), and so on. Nobody has really assessed, but after the fifth crossing, the red light seems more like a recommendation than a prohibition.

The same mechanism explains why everyone gets on the same escalator. The first person off the train takes the nearest one. Those following see someone already on that escalator and think unconsciously "evidently that's the right one". It's not that it's more right, it's simply where someone arrived first.

The curious thing is that this behaviour persists even when it's obviously inefficient. You see the queue forming, you see the other two escalators empty, but the cognitive cost of "being the different one going against the flow" is higher than the annoyance of waiting. Even if you rationally know it would be faster to take the empty one.

The same game, different contexts

This pattern repeats identically across contexts that seem to have nothing in common.

Bank runs work exactly like this. September 2007, Northern Rock: the news that the bank had requested help from the Bank of England triggers panic. The first customers start withdrawing, not because the bank is really in crisis, but because "better safe than sorry". Others see the queues outside the branches and think "they know something I don't". Within three days, customers withdraw a billion pounds. The bank, which was perfectly solvent, collapses precisely because of the run on deposits that was meant to prevent collapse. Collective behaviour created exactly the crisis everyone feared, transforming individually rational precaution into systemic disaster.

Technology adoption follows the same script. Nobody wants to be first to buy a new product, too risky, you don't know if it'll work, if it'll be supported, if it'll become standard. But as soon as you see enough people who've already adopted it, suddenly the perceived risk plummets. It's not that the product has improved, it's that others' behaviour has solved your decision problem.

Market movements show the same dynamic in pure form. When a stock starts falling, the first seller might have real information. But those following sell because they see others selling, and interpret that behaviour as a signal that "something's wrong". Within minutes you can have a crash driven not by economic fundamentals, but by an imitation cascade.

The price of cognitive autonomy

The most interesting part of all this is how little it takes to trigger the mechanism. You don't need a charismatic leader, you don't need recognised authority, you don't even need particularly convincing behaviour. Just one person moving first is enough.

And resistance to this reflex is surprisingly costly. When everyone crosses on red and you stay put, you're not just making a different decision, you're actively fighting an evolutionary impulse millions of years old that's screaming at you "follow the group". You're spending cognitive energy to autonomously assess a situation the group has already "solved".

This explains why it's so difficult to go against the flow, even on trivial matters. It's not stubbornness or lack of personality, it's that your brain instinctively knows that cognitive autonomy is expensive, and reserves it only for when it's really worth it.

The red light? The brain decides it's not worth it. Better to save energy for more important decisions.

What stays in the mind

Next time you find yourself automatically following the flow, at the supermarket checkout, choosing a restaurant, in a group discussion, ask yourself: am I following because I've assessed, or because someone else has already "solved" the problem for me?

There's no right or wrong answer. Sometimes following is the most efficient choice, why waste energy when the group has already processed the information? But it's worth knowing when you're doing it, and when instead what you're following isn't collective wisdom, but simply whoever moved first.

The London traffic light this morning between the Strand and Arundel Street will continue to see people crossing on red. But at least now you know why they do it, and why you'll probably do it too.