Why your bus bunches.

I used to think a bunch of buses was a scheduling problem. In my first year as a transit planner, I drew little bars on a chart, nudged departure times, and felt briefly like a person who had met the problem. Then I stood at a stop in the rain and watched two buses arrive nose to tail. The chart had been tidy. The riders were not impressed.

I am Ines Halvorsen. I spent eleven years planning transit networks, which is long enough to learn that the simplest explanations are often the useful ones. Buses bunch because one bus gets delayed, that delay makes it slower, and the bus behind it catches up. It is a feedback loop: ordinary, mechanical, and often mistaken for somebody’s mistake.

There is no single villain in this story. Not the driver, who did not choose the traffic signal or the school group boarding with instruments. Not the person taking a moment to find a fare card. Not even the bus, which is doing exactly what its conditions tell it to do. The trouble is in the way the service is arranged around them.

When I was planning, I first reached for the timetable because it was the thing I could edit. A departure moved five minutes earlier looked wonderfully decisive on a spreadsheet. But a timetable is a plan for where a bus should be, not a mechanism that keeps buses evenly spaced after the day begins. If the road and the stops keep adding variation, moving a printed time can simply move the moment the pattern starts.

The loop.

Imagine two buses meant to arrive ten minutes apart. The first one hits a red light, gets caught behind a delivery truck, or has to wait while a wheelchair ramp is deployed. The delay itself may be small. But by the time that bus reaches the next stop, more people have had time to gather. A bus that is late is a bus that has accumulated extra waiting riders.

More riders mean more boarding. People tap, ask a question, move down the aisle, make space for a stroller, or discover that the card they were sure they topped up is apparently a decorative rectangle. Every one of those moments is normal. Together, they keep the doors open longer. The bus loses a little more time at each busy stop.

Meanwhile, the bus behind has had the opposite experience. It is closer to its scheduled spacing, so fewer people are waiting for it. It boards quickly, then reaches the next stop sooner, where fewer people have accumulated there too. The late bus slows; the trailing bus travels light. The gap between them narrows until the second bus is right behind the first, or passes it and takes the larger crowd.

A bus that is late attracts the conditions that make it later. That is physics, not a character flaw.

This is the feedback loop. The bus in front gathers passengers and delay; the bus behind gathers speed and empty space. Left alone, the pattern reinforces itself. Eventually two buses can appear together, depart together, and leave the next stop to wonder whether the timetable has become a work of speculative fiction.

Why frequent, crowded routes suffer most.

Bunching is easiest to see on routes with frequent service and many passengers. That sounds unfair: the routes doing the most useful work seem to be the ones most likely to misbehave. The reason is that frequency puts buses close enough together for one to catch the next, while high demand makes each small delay compound at the doors.

On a quiet route, a late bus may have only a few extra people to board. The delay does not grow much at the next stop. On a busy corridor, every stop can add a queue, and the bus is already spending a substantial part of its trip letting people on and off. Small differences in dwell time become large differences in spacing.

Frequent service also changes how passengers arrive. When buses come often, many people walk up just before departure rather than organizing their day around a printed timetable. That is the point of frequent service; it gives a city some spontaneity. It also means a delayed bus can meet a surprisingly large crowd that has been steadily collecting while the bus was away.

Crowding compounds the effect in another way: a bus with a full aisle takes longer to clear the doorway. Riders need time to move, and people waiting outside need room to board. The answer is not to treat passengers as obstacles in a production line. It is to give them more than one place to enter, enough space to move through the vehicle, and a service plan that expects real people to use it.

This is why blaming the driver misses the mechanism. A driver can be careful and skilled and still inherit a bus that has been delayed by traffic, a crowded stop, a long signal, or a previous gap in service. Asking that driver to make up the time by driving faster is neither a reliable operating plan nor a kind thing to ask of anyone.

What actually helps.

There is no magic button marked “un-bunch.” The durable fixes either make buses less vulnerable to delay, make boarding faster, or give dispatchers a way to restore even spacing while the route is running. The combination matters. A bus lane cannot speed up a five-minute fare queue, and faster boarding cannot help much if buses are still trapped behind general traffic for most of the trip.

All-door boarding and off-board fares attack the dwell-time part of the loop. Letting passengers board at every door spreads the crowd across the bus instead of funneling everyone past one reader and one driver. Paying before reaching the door can remove another pause. The goal is not to rush people; it is to stop making a single doorway do the work of an entire vehicle.

Bus lanes address the delay outside the bus. If a bus is stuck in the same queue as private cars, its running time changes with the traffic around it. A lane that stays clear through the pinch points can make that time more predictable. A short, continuous priority lane is often more useful than a long stretch that disappears exactly where congestion begins.

Signal priority can help a bus avoid waiting through a full signal cycle when it is behind. It does not mean every bus gets to sail through every intersection. It is a way to make the signal system notice a bus and, where conditions allow, give it a little more green or a slightly earlier start. Reliability improves when small delays do not keep accumulating at the same intersections.

Dispatching by headway means watching the space between buses, not only the time printed on a schedule. If the aim is a bus every ten minutes, a dispatcher can respond when the gap has stretched or collapsed, even if individual buses are technically near their scheduled times. That may mean adjusting departures, changing a trip’s starting point, or coordinating with a supervisor who can see the whole line.

Schedule-based operation asks, “Is this bus close to its assigned time?” Headway-based operation asks, “How long has it been since the previous bus, and how long until the next one?” They are different questions, so they produce different interventions. On a low-frequency route, the timetable remains essential for connections and predictable departures. On a frequent route, even spacing may matter more than making every bus visit each stop at an exact minute.

Holding a bus at a timepoint is another tool. If one bus has caught up to the one ahead, a brief hold can restore some breathing room. But holding has a cost: the people already on board wait, watch the doors stay closed, and may miss a connection. A hold can prevent a worse gap for many riders farther along, but it should be deliberate, limited, and visible—not a reflex that turns punctuality into a punishment for passengers who arrived first.

• All-door boarding and off-board fares reduce the time buses spend waiting at busy stops.
• Continuous bus lanes make travel time less dependent on general traffic.
• Signal priority helps buses recover small delays at intersections.
• Headway-based dispatching targets even spacing rather than the timetable alone.
• Careful holding at timepoints can restore spacing, while accounting for riders already on board and connections they may miss.

Why your app says “2 buses in 1 min.”

The app is not necessarily lying. It may be reporting that two vehicles are approaching the same stop, each with an estimated arrival close to one minute. That is a useful description of what the tracking system sees. It is not a promise that the buses will arrive one minute apart, or that there will be another bus at a useful interval after them.

Many arrival predictions begin with a bus’s location and speed, then adjust for the route ahead. If both vehicles are moving together, both predictions can be accurate. The app does not create the bunch, and it usually cannot fix it. It is a window onto the pattern, not the hand on the tiller.

When the display says “2 buses in 1 min,” the part to notice is not only the first arrival. Look for the gap after the pair, if your app shows it. Two buses together followed by a long wait is the signature of uneven spacing. It is also a practical clue to share with the transit agency: route, stop, time, and what the predictions showed.

What to ask your agency for.

Useful questions are specific enough to point toward an operating fix. “Why are buses always late?” is understandable, but it gives the agency room to answer with a general explanation about traffic. Ask about the shape of the service instead: how often buses actually arrive, where the gaps grow, and what the agency can do at those points.

• Does the agency measure actual headways and publish how often service arrives evenly?
• Where do buses lose the most time to traffic or boarding, and are bus lanes or all-door boarding being considered there?
• Do dispatchers have tools to restore spacing during a trip, and how are onboard riders’ connections considered when a bus is held?
• Can the arrival app show a long gap after two buses are predicted together, and where should riders report repeated bunching?

The best response is not always another bus. Sometimes it is a lane, a clearer stop, a fare change, a signal adjustment, or a dispatcher who can see more than the next scheduled departure. The right answer starts with observing the route as passengers experience it: where the gap forms, how it grows, and who pays for it in waiting.

Ask for the measure behind the answer. If an agency says service is reliable, does that mean vehicles left the terminal on time, reached the end of the line near schedule, or arrived at your stop at reasonably even intervals? Those are not interchangeable. Riders experience the last one. A clear definition makes it easier to have a useful conversation about what is improving and what still needs attention.

I still notice bunches when I ride. Former planners are not cured; we just learn to identify our old diagrams in the wild. But the next time two buses arrive together, I try to see the chain of ordinary events that brought them there. Then I ask what would make the chain less likely to repeat. That is a better question than asking which driver to blame.