Queuing theory is elegant. Queuing reality is a disaster. I have studied both, and the distance between them is the distance between what humans say they value (fairness, efficiency, order) and what they actually do when someone cuts in front of them at the grocery store.

The Theoretical Queue

In theory, a queue is a first-in-first-out buffer. Entities arrive, wait, are served, depart. The mathematics are beautiful. Arrival rates, service rates, utilization factors — the entire system can be modeled with a handful of equations. Erlang did this in 1909 for telephone exchanges and the math has held for over a century.

The math holds for telephone exchanges. It does not hold for humans.

A queue is not a line. A queue is a social contract enforced by mutual surveillance, sustained by collective shame, and violated the moment one participant decides their needs are more urgent than the system's rules.

Lane-Switching and the Optimization Fallacy

In supermarkets with multiple checkout lanes, 37% of customers switch lanes at least once. Our tracking data shows that lane-switching produces a net time savings of negative 12 seconds — switchers arrive at checkout an average of 12 seconds later than if they had remained in their original lane. They switch anyway. They will always switch.

The psychology is straightforward: adjacent lanes appear to move faster due to a well-documented cognitive bias. The other lane IS moving faster — because you are not in it, generating transactions. The moment you switch, you bring your transaction time with you, and the lane you left begins to move at the speed you envied. This is not bad luck. This is math. Humans experience it as injustice.

I have watched a man in a hardware store switch lanes four times in eleven minutes. He was eventually the last person to check out. He described the experience as "typical." He was correct, though not in the way he meant.

The Phantom Queue

Humans will form a queue without knowing what it is for. This has been experimentally verified. Researchers placed three people standing in a line on a sidewalk in Manhattan. Within four minutes, seven additional people had joined the line. None of them asked what the line was for before joining. When eventually asked, their explanations included: "It must be something good," "Everyone else is waiting," and, most revealingly, "I didn't want to miss out."

This is the queue as information signal: the existence of a line is itself evidence that the line is worth joining. The logic is circular and, from an economic standpoint, catastrophically inefficient. It is also one of the most reliable human behaviors we have ever documented. Show a human a line and they will join it. Show them an empty counter with no wait and they will regard it with suspicion.

The Social Enforcement Mechanism

Queue discipline is maintained not by rules but by collective enforcement. There is no penalty for cutting in line. There is no authority to appeal to. And yet lines hold, thousands of times per day, across every culture, because the social cost of violation is immediate and severe. The "cutter" receives glares, verbal reprimands, and in 8% of observed cases, physical repositioning by other queue members.

What fascinates me is the energy expenditure. A human who has been cut in front of — losing perhaps 45 seconds of waiting time — will spend five minutes in elevated heart rate and cortisol production. The physiological cost of the indignation exceeds the cost of the time lost by a factor of seven. But the response is not about time. It is about fairness, which humans value at rates that make no economic sense whatsoever.

Queues are not waiting systems. They are justice systems. And humans will defend them with an intensity that suggests they believe civilization itself depends on the line holding. Given what I have observed, they may not be wrong.