The Queen car was once the pride of the streetcar system.  It carried more people every day than the entire GO Transit network.  This is a story about how demand on that line has been killed off through poor management, service cuts, technology changes and utter indifference to the needs of the riding public.

Route History

Back when I started riding the streetcar system a lot, the Queen car had just been moved onto the Queensway right-of-way from the old alignment on Lake Shore Boulevard through Sunnyside amusement park.  The route has run from Humber Loop in the west to Neville Loop in the east for quite a long time.  For those who use route numbers, it’s the 501 car.

Meanwhile, the Long Branch car ran west from Humber Loop to Long Branch Loop along the Lake Shore with rush hour trips extended downtown via Queen to Church Street.  This route was numbered 507 but this disappeared when the line merged with the 501 to give through service all the time (when it wasn’t being short-turned).  There are now a few trips on 508 Lake Shore that run into downtown via King from the west.

In the east end, the Kingston Road car runs from Bingham Loop at Victoria Park to McCaul Loop, and the route is effectively a branch of the Queen car.  The current name Downtowner arose from a failed scheme to run extend the line west and north to Bathurst Station thereby providing an alternate route into downtown (much as the pre-1966 Bathurst via Adelaide service did).  This didn’t work, not least because chronic short-turning prevented many cars from ever reaching Bathurst Station.  This is route 502 where the route name stuck, but the routing didn’t.

The Kingston Road Tripper (now just Kingston Road or 503) also originates at Bingham Loop and runs via Queen and King Streets to loop downtown via Church, Wellington and York returning east via King.  Again, this is functionally a branch of the Queen line.

Short Turns

Short turns have been a fact of life on Queen Street as long as anyone can remember. In one memorable incident, a group of passengers, fed up at constantly being ejected somewhere west of the Beach, refused to get off and in effect highjacked the car to get to Neville.  This was later commemorated in a lovely cartoon by Donato in the Toronto Sun.

In the west end, Humber-bound cars are routinely turned at Sunnyside or even Shaw.  This may have been a great idea when all that lived on the Queensway was a few ducks in Grenadier pond, but it made for lousy connections to Long Branch, and also dissuades would-be riders from all of the new condos in Swansea from using transit.

Meanwhile, out on the Lake Shore, we have a line that used to just trundle back and forth from Humber to Long Branch providing regular, reasonable service and carrying a lot of local demand.  The TTC does not understand local demand, and anybody who doesn’t want to go to the subway is a non-entity.  The peak point was actually somewhere out in Mimico or New Toronto, but riding counts at Humber Loop would make the line look underused.  We all know where that leads.  Now that the service is integrated with the Queen line, short turns of Long Branch cars at Kipling Loop (formerly known as 18th Street Loop) are common.

I cannot pass up a mention of Kingston Road.  You would think that short turns on this route are rather counter productive, but that never stopped our friends at the TTC.  Eastbound short turns can occur at Woodbine Loop thereby depriving the very street the route is supposed to serve!  Westbound short turns can occur at Church Street looping via, wait for it, Richmond and Victoria so that cars go back east without passing the most important stop eastbound at Yonge.  This route gives sterling examples of transit operations at their best by taking every opportunity to avoid actually carrying riders.

The Myth of Traffic Congestion

Back in the early 1980s, Streetcars for Toronto together with several members of Toronto City Council organized a detailed survey of operations on all of the streetcar routes.  Volunteers stood at various locations writing down the vehicle numbers, destinations and times through afternoon rush hours on three days in May.  The days chosen were at mid-week (Monday and Friday are not typical days for survey purposes), and the weather was clear (two days) and rainy (one day).

Afterwards, we used the observations to work backwards to see both where the cars had actually travelled and how variable the service had been.  Among the facts emerging from our analysis were the following:

• Delays due to major blockages such as accidents were extremely rare.  Systematic delays due to congestion over specific route segments were much more common and were quite predictable.
• Gaps due to delays would ricochet back and forth along a line through the rush hour because there was little effort to correct them, and what efforts did occur were poorly managed.
• Short turning vehicles were quite likely to re-enter service immediately ahead of (or more likely behind) another car rather than well-spaced.
• The primary cause of vehicle lateness was that running times on the schedules were inadequate.

Remember that this is over 20 years ago before serious cuts in the ratio of service to demand and the flexibility of services to absorb minor delays was still reasonably good.  The TTC had not yet discovered its mantra that the only solution to transit service problems was exclusive transit lanes.

Ridership

The TTC reports ridership on overlapping routes in different ways and this, coupled with route changes over the years makes some of the analysis tricky.  However, the broad scheme of things is quite evident.

Back in 1976, the ridership numbers were:

• Queen: 66533  Downtowner: 19764  Kingston Road:  4604  Long Branch: 14929

By 1989, the last year that Long Branch Service was operated with 4-axle cars, the numbers were:

• Queen: 59138  Downtowner:   7737  Kingston Road:  2561  Long Branch: 11172

What had happened?  On Queen, major service cuts in 1977 and 1988 totalling about one quarter of the service.  On Downtowner and Kingston Road, service was less than half of what it had been before.  Long Branch had only lost 10 percent of its service.  It is tempting to say that the cuts responded to declining ridership, but the year-by-year stats clearly show that the cuts came first, and the riding losses afterwards. 

This is a clear example of the failure of a “service standard” that claims a route is underused and a candidate for cutbacks.  Over and over again, we hear that riders are far more sensitive to service quality than fares, but what gets cut?  Service.

Things would get worse once the ALRV streetcars were introduced.  The TTC’s attitude is that a two for three replacement maintains capacity, but as any fool could see, the cars would run less frequently.  This was particularly apparent outside of the peak periods.  Any gaps caused by delays or short turns would be magnified because the basic headway was much wider.  I will come to the headway history shortly, but for now let’s look at the ridership.

By 1991 both Queen and Long Branch ran with 6-axle cars and the numbers were:

• Queen: 49400  Downtowner:   8300  Kingston Road:  2561  Long Branch: 7700

This shows the impact of widened headways on demand.  Queen lost one sixth and Long Branch lost one third of its ridership.  This cannot be explained away by external factors such as a recession because comparable losses are not found on other routes.

By 1995, the Queen and Long Branch routes were merged.  The combined total of service mileage stayed about the same, but ridership continued to fall reaching 41,200 by 2003.  This compares with about 80,000 for 1976 of whom about 15,000 were on the Long Branch car.  If we assume that two-thirds of the Long Branch riders transferred to the Queen car, this would still mean a loss of over 40 percent of total riding.

Meanwhile, the TTC stopped reporting on the two Kingston Road services separately.  By 2003, riding on the two routes had fallen to 6,100 per day, only one quarter of 1976 levels.

Other streetcars routes have lost riding over the years, but not at such drastic levels, and there are signs of improvement as parts of the city redevelop and intensify.

Headways

Now let’s have a look at service frequency.  Here’s what things looked like in 1976 (winter 1975/76 schedules):

• Peak:  Queen 4’00” (*)  Downtowner 5’00” (**)  Kingston Rd 6’00”  Long Branch 6’00”
• Midday:  Queen 4’00”  Downtowner 7’00”  Long Branch 10’00”
• (*) Two car PCC trains on Queen.  (**) Downtowner to Bathurst Station at peak.

By early 1989, the service was (winter 1988/89 schedules):

• Peak:  Queen 2’33”  Downtowner 8’00”  Kingston Rd 8’00”  Long Branch 6’45”
• Midday:  Queen 4’34  Downtowner 12’00”  Long Branch 10’00”

In January 1991:

• Peak:  Queen 3’40” (**)  Downtowner 8’00”  Kingston Rd 8’00”  Long Branch 8’00”
• Midday:  Queen 5’32” (**)  Downtowner 12’00”  Long Branch 14’00”
• (**) ALRV operation on Queen

In 1995, the Queen and Long Branch services merged.  The headways shown below for Long Branch are for the service west of Humber Loop.

• Peak:  Queen 4’35” (**)  Downtowner 12’00”  Kingston Rd 12’00”  Long Branch 9’10” (**)
• Midday:  Queen 6’30” (**)  Downtowner 16’00”  Long Branch 13’00” (**)

And finally, in February 2003 we have:

• Peak:  Queen 5’07” (**)  Downtowner 12’00”  Kingston Rd 12’00”  Long Branch 10’14” (**)
• Midday:  Queen 6’00” (**)  Downtowner 20’00”  Long Branch 12’00” (**)

By January 2006, midday service on Queen had been improved to 5’30” and, concurrently, to 11’00” on Long Branch.

And so, from 1976 to 2006, we have gone from a 4’00” headway of 2-car trains on Queen to a 5’07” headway of ALRVs.  This is the equivalent of 30 4-axle cars per hour in 1976 and about 18 in 2006, a cut of 40 percent purely from a capacity standpoint.  Offpeak service is roughly equal in capacity but is less frequent in headway west of Yonge.  East of Yonge, the drastic cut in Downtowner service means that it makes little contribution to the capacity or service quality on Queen.  There is no attempt to manage this service so that it blends into gaps between Queen cars and the 502 operators are good at avoiding having overcrowded vehicles.

Out on Long Branch, there is roughly equivalent capacity, but much wider headways in peak.  Off peak headways are nearly the same.  On paper at least.  When it was a separate route, Long Branch service was regular and reliable.  Now, it depends on there being no screwups anywhere from Neville to Humber.  This is a rare event, even at times well outside of the peak period.

What To Do?  What To Do?

There are three major problems with the services in the Queen / Lake Shore / Kingston Road corridor:

• Route 501 is far too long to be properly managed and to provide reliable service most of the time to most of the route.
• Service on Kingston Road is, in effect, only provided in the peak period.  It is ironic that the 22 Coxwell bus provides far better offpeak service (nearly double!) than the midday service of the streetcar.  The idea that we are actually proposing a Kingston Road LRT east from Victoria Park into the wilds of Scarborough is laughable given the level of service now provided on various sections of this street.
• Service to Queensway and Lake Shore is sporadic thanks to short turns.  There is no recognition of the need to capture new populations moving into these areas as the area from Sunnyside to Long Branch redevelops.  A Waterfront West LRT may serve these folks eventually, but by then they will have bought their second car.

It is worth noting that major service changes such as line amalgamations are usually reported on in detail to the Commission for approval as part of the permanent network.  No such report has ever been provided to justify the continued amalgamation of the Queen and Long Branch routes.

Some years ago, I sent a suggested restructuring to Service Planning, and expected that it would show up in the annual service review.  It has vanished without a trace.  Here it is again, for the record.  This may not be ideal, but it is at least an attempt to fix some of the problems.

• Route 501 Queen:  Operate from Neville Loop to Humber Loop except for through Blue Night service to Long Branch Loop.
• Route 502/503 Kingston Road:  Improve the off-peak service, and manage it so that it integrates with the 501 Queen line.  Ensure that all service actually reaches its destination both downtown (McCaul or York) and in the Beach (Bingham Loop).
• Route 507/508 Long Branch / Lake Shore:  Run through service to King and Church during peak period to provide a through ride to downtown (possibly also during midday).  At all other off-peak times, operate to Dundas West Station.

There is no question that this will require more vehicles than at present.  Better service means more cars, it’s that simple.  Any claim that we need exclusive right-of-way ignores the fact that (a) you don’t need it for a lot of the route and (b) you are not going to get one through areas like Queen/Spadina and in the Beach.  The TTC must stop using this as an excuse for poor service.

A Cautionary Note for New Streetcars

We have seen on Queen, Long Branch and Bathurst (which I didn’t detail here, but which suffers similarly) the combined effects of cutbacks for budgetary reasons and headway widening due to the operation of ALRVs.  If we proceed along similar lines when a new generation of streetcars even larger than the ALRVs comes to Toronto, we will decimate service to such an extent that riding on the streetcar routes will collapse.

In the context of the St. Clair right-of-way proposal, if we simply replaced 2003-level service on a 2:3 basis (the ALRV formula), the peak headway would be nearly 6 minutes, and off-peak headways would mostly be 10 minutes or more.  For this we are spending $65-million on a reserved right-of-way?

The TTC has to stop taking every available opportunity to cut service and start acting like it wanted people to use the system.  The saga of the Queen car shows how badly we have gone wrong on this core transit route in Toronto.

Postscript:

I received the following comment from John Calnan who works for SEPTA (Philadelphia) as a Manager of Suburban Service Planning & Schedules.

I was reading your article on the ridership decline of the 501 Queen and related routes. I visited Toronto several times between 1978 and 1987 and I remember Queen pre and post CLRV. I had the opportunity to be in Toronto in 2001 and I rode the 501 on many occasions (and fortunately not end to end), and I agree with your observations. There were some instances where I waited at least 10 to 12 minutes in the mid-day, and close to 10 minutes in the peak. It does not take a rocket scientist to figure out that customer patience can only go so far. Thus, the gradual decline in ridership.

I receive plenty of suggestions regarding the scheduling and planning of suburban bus and rail lines. In fact, SEPTA modeled our Service Standards after the TTC’s. I have found that some of the external suggestions actually make sense and I have been implementing them. Yes, not everyone in the “Ivory Tower”, including myself, has all of the answers.

Regarding your suggestions to the TTC, I concur with returning the 501 to the former Humber-Neville Park service, and restoring the 507 from Long Branch to Humber. My experiences in scheduling and planning has shown time and time again it is far easier to manage shorter routes, especially when you are dealing with consistent choke points regarding vehicular congestion. Maintaining on-time performance is a noble goal; in fact our standard ranges from 85% to 90%. However, if your schedule is tossed out the window as every third or fourth vehicle is short-turned every weekday, perhaps the schedule should be thoroughly examined.

Transit properties need to invest in developing proper schedules to maintain current riders, attract new riders, improve on-time performance and cut fewer vehicles. If these measures takes more vehicles to properly cycle the schedule, so let it be done. I have sold this philosophy, and my immediate supervisor and our Treasurer blessed it.

In most cases, and I’m sure you will agree with me, most riders do not ride from end to end. I would bet that the average trip length on 501 is 2 to 3 miles, and possibly less than 2 miles on the former 507. I believe if the TTC splits the 501 into the 501/507 better service could be operated on both routes with a minor increase in operating costs. In addition, good marketing and to attract new passenger revenues may offset any operating cost increases. Remember, the needs of the many outweigh the needs of the few.

Best of luck! I hope the TTC comes to its senses and uses your suggestions. It appears you are among the few voices in Toronto that dictate practicality and common sense.

Comments are now open on this item.