Why Are There So Many Empty Buses?

Transit advocates spend a lot of time explaining why we need more and more transit service.  Riders spend a lot of time complaining about sardine-can loading on vehicles.  Politicians spend a lot of time complaining about “inefficient” transit operations, most likely the empty buses they saw while driving hither and yon across the city.

Where do those buses come from?

Leaving aside the obvious cases of buses enroute to or from a garage, there is a basic fact that may seem rather strange:  most buses will not be full most of the time.  This flies in the fact of those who demand efficiencies and treat half-empty buses (or worse) as a sign that transit riders live in coddled luxury compared to the hapless drivers stuck on the 401.

Transit routes have many different types of loading patterns, and they may have different patterns at different times of the day and different days of the week. 

Here are some examples:

  • Some routes have very highly directional loading — heavy one way, light the other — and they have only one major focus, usually at their destination.  This is a classic commuting route.  In the morning, we pick up people enroute to the subway, dump a jam-packed load off at the station and drive almost empty back out to get another load if there is still a “rush hour” by the time we get back to the outer end of the line.  The pattern reverses in the evening.
  • Some routes are bidirectional — they are lucky enough to have at least two major destinations, and they attract riding in both directions going to them.
  • Still others have many minor destinations and a few major ones.  They have a lot of turnover along the line with passengers constantly getting on and off.

In the first and worst case, the bus will not even reach a standing load until well along its peak-direction trip, and it will be a lonely, near-empty trip on the return journey.  The riding count will be one full load, plus a few returns, say 80 people, and the space utilization will be somewhere between 25 and 30 percent.  (The inbound trip started empty and ended full for an average of 50 percent, but the outbound trip carried almost nobody dragging the average down to 25.)

In the second case, we get some reverse loading, but if they are all going to one place, we still only manage an average utilization of around 50 percent, and a rider count of two full loads.

In the third case, we have many destinations and a fairly balanced flow with ons and offs along the route.  Obviously the bus has to be empty sometime, but something else interesting happens here.  If we have a lot of turnover, the total riding count can easily exceed twice the capacity of the vehicle, and the bus may never be completely full.  This is an ideal world, but sadly few of our routes behave like this.  Talking about “efficiency” gets tricky here — are we measuring the number of riders we carry per bus hour, or the average occupancy of the bus? 

The next factor is the length of a route.  Assuming that population (and hence demand) is uniformly distributed along a route, there is going to be a fixed number of riders per route kilometre per hour.  If the route falls into our first category (gradual buildup to a major destination), then obviously a long route has more time to build up a full load than a short one.  However, the short one costs less per trip because it doesn’t go as far. 

Indeed, if we run the same headway on both routes, the short one will always have a lower average occupancy than the long one even though the cost per rider may actually be lower.  If we cut the headway to get better “efficiency”, we will lose riders because service frequency (a major component of convenience) goes down.  Somewhere there is an appropriate balance, but the bean counters care only for dollars, not riders.

[Careful readers of my posts may think that I am on a single-minded mission to save the Forest Hill bus.  That is not true.  I have been on it, I think, possibly twice in my entire life.  Its disappearance would pain me far more in theory than in practice.]

The point in all of this is that transit service, by definition, has to have some space free on vehicles in order for riders to board.  It is pointless to attempt a full load all of the time in both directions.

What is needed is a recognition that most routes will have partly empty buses at least some of the time, and that packed loads are not a mark of success and cannot be achieved on some routes.  We need to look at the resources needed to carry riders and we also need to recognize that parts of the network will always be less “efficient” than others. That’s what networks are all about.