Some bus bunching does occur because of the phenomena described. But most occurs because of random traffic build up at bottlenecks and drivers intentionally running "dull" (behind schedule on purpose) or "sharp" (ahead of schedule on purpose) to extend the break time in their contracts.

While we have equipped all MTA buses with real time bus location GPS and communications and can immediately find the bus if terrorists attack, the unions have managed to prevent it from being used to check on time performance and driver "rollout", and LA still has significant bunching do to deliberate lack of on-time departure from initial points.

In some cases high seniority drivers will actually wait and not leave until a following section, scheduled to be run by a part timer departs, and then follow them, as it is so much easier to drive when you don't have to stop and pick up passengers during rush hour. It was so bad on certain routes that one jammed bus would be followed almost "nose to tail" by two empty buses hanging back on purpose, even though these were supposed to be scheduled ten minutes apart. So, in effect, you ended up with overcrowded thirty minute service. And, one part timer told me this happened every day and was all intentional.

Signal prioritization, integrated with real time bus tracking, and careful monitoring, have been somewhat successful in the Wilshire and Orange Line corridors in preventing this bunching with the Rapid buses. Better stop design to allow for full curb loading and better reentry into the traffic stream, longer spacing between fewer bus stops, and much wider bus doors and aisles on buses are successfully used to virtually eliminate this loading problem in much of Europe. Low floor buses also speed up loading, reducing such delay.

Rail, in general, doesn't have this problem, though. The high platform-multidoor loading, eliminated the potential problems from this high volume loading issue when we designed the light rail system.