Ridership and Service Since 1976 (Updated)

At the 501 Queen Forum last week, I and others talked about the declining service and ridership in the Queen Street corridor.  This post reviews the published statistics from 1976 to 2005, the latest information available so far.

Streetcar Ridership and Mileage 1976 to 2005

These data are taken from the annual Service Plan and related documents.  The most recent counts are on the TTC’s website.

Updated December 11:  A consolidated count has been added for the Queen services (501, 502 and 503) to show the ridership and mileage in the three routes serving this corridor. 

Like so much TTC data, there are several caveats to reading these charts:

  • Originally, the data were reported for the year before the date it was published.  More recently, the time period spans years with the most recent data being for 2005-2006.  These are shown as “2005” in the charts.
  • All values for riding and mileage are rounded to hundreds.
  • For 2005-2006, mileage is reported in kilometres.  Many values change slightly from the previous year when stated as miles indicating that the TTC converted its data on the unrounded values, and then rounded the results.
  • Riding counts are taken once per year, at best, and reflect the conditions on the date of the count.

Various historical notes:

  • From 1991 onward, data for 503 Kingston Road Tripper was consolidated with 502 Downtowner.
  • From 1991-1996, data for 509 Harbourfront includes only the original route on Queen’s Quay.  The 510 Spadina car began operation in 1997.
  • From 1994 onward, the 507 Long Branch service was amalgamated with 501 Queen.


With the exception of Harbourfront/Spadina, riding has been falling on all routes over the long term.  However, counts are taken infrequently with the value for 501 Queen unchanged since 2001.  The TTC claims that it has been adjusting its service to match current demand, but if so, one wonders why measurements of that demand have not shown up in the published counts.

A riding drop on 501 Queen in 1991 corresponds with the replacement of CLRVs by ALRVs (also visible in the drop in service mileage).

The line for 501 Queen merges with that of 506 Carlton because the riding for both lines has been reported as 41,200 since 2002.


Whether you prefer Imperial or Metric units, the story is the same — the amount of service, measured as vehicle mileage, has dropped on almost all routes.

501 Queen shows a big drop in 1991 with the change from CLRV to ALRV operation, and a subsequent rise in 1995 following amalgamation with the 507 Long Branch car.

Service cuts in 1991 are visible on 506 Carlton, 505 Dundas, 512 St. Clair and 507 Long Branch.  Despite the consolidation of data for 502 and 503 services to Kingston Road, there was no change in the reported mileage indicating a substantial service cut.


This value shows the ratio of riding to service level.

No surprise here — the 510 Spadina car has the highest ratio of riding to service.  This reflects the strong bidirectional demand on the line and the short trip lengths for most passengers.

On 501 Queen, the density value went up in 1991 with the introduction of ALRVs, but this fell again in later years.  Note that the same riding count was reported for this route for 1991 through 1994 and the impact of the service change cannot be determined at the annual level.  Similarly, the same count was used for 1995 to 1997.

The real “chicken and egg” question here, of course, comes from the consistency of many of these values over two decades.  Does ridership decline because there is less service (thereby preserving the ratio), or is service cut in response to falling ridership?

Transit networks can fall into a hopeless decline because these two factors are closely linked.  Bad service drives away riding thereby “justifying” even worse service.  When widened headways compound with ineffective line management, actual service quality drops more than the raw mileage figures show.

Yes, there have been economic and demographic changes over the years in Toronto that affected transit ridership.  However, the TTC’s cutbacks have cemented the loss of riding and undermined the system’s attractiveness and capacity in response to growing travel demand.

19 thoughts on “Ridership and Service Since 1976 (Updated)

  1. Excellent analysis Steve. Given the coarse resolution of the data, the impact of transit priority signals, ineffective though they are, cannot be seen.

    I recall TTC claiming that such prioritization had saved them 2 cars on the 506 College route, but I don’t recall any time savings mentioned.

    I know that there is no such prioritization on Lakeshore Blvd, though we have very poor 501 service that could definitely use any improvement available. My request for signal prioritization on Lakeshore was met with the comment “ridership doesn’t justify the investment”. Maybe not, but the poor service certainly does!

    This entire situation cries out for a Streetcar Ridership Growth Strategy.

    Steve: The saddest part about the “savings” from Transit Priority was that they were used to reduce operating costs rather then improve service on the streets where they were implemented. Savings from streetcar routes downtown were ploughed into improved bus service in the suburbs.

    This is the same sort of attitude that showed up in the original claims that the St. Clair right-of-way would save money by allowing the same service to be provided with fewer cars. The idea that they might actually operate service more frequently was completely ignored giving the project one of its many well-deserved black eyes.


  2. Clearly, if we want to reverse this trend of declining ridership, the TTC needs to buy more streetcars than are necessary to simply replace the existing fleet, even excluding additional streetcars required for new lines. Assuming that we have no choice but to buy ALRV-length streetcars (few manufacturers, and no major manufacturers, make anything like a CLRV), I would suggest buying slightly more of the new streetcars than we have of the current streetcars today. This does not include additional cars needed for new lines. This would allow similar frequencies to today but with more capacity; however, Queen and Kingston Road (which have unacceptably low frequencies right now) would have improved frequencies. This would almost certainly require building a new carhouse (too bad we closed Wychwood) but it would greatly improve streetcar service in Toronto.


  3. Does the dropoff of the 506 in 1999 have to do with the closing of Maple Leaf Gardens and moving of the Leafs to the ACC?

    I remember in the summer of 2006, I boarded a very crowded Spadina rushhour car with my visiting sister. We were right at the front and somehow a question of service entered the conversation. The driver mentioned “we’re a couple minutes behind schedule”. To which I thought all the streetcar signs have FS on them, yet the driver knows when s/he is supposed to show up. Shouldn’t the passengers be afforded that knowledge?

    Then I remembered – this was the TTC we’re talking about. Apropos, on Saturday night around 1245am, I was waiting for a streetcar northbound on Spadina. Waited a fair while – but saw 3 in a row southbound. How can the CIS supervisor mess it up that badly?


  4. Thread after thread, I have a hard time swallowing the line that this site is about transit advocacy and not about streetcar railfanism. I know it’s your blog, but the 501 and 504 are getting way too much screen time here.

    Steve: If making service run properly isn’t “transit advocacy”, I don’t know what is. We have a huge credibility problem in Toronto with people who live in neighbourhoods where taking transit rather than driving should be a no-brainer. However, we are converting these folk to die-hard car drivers because we can’t provide decent transit service.

    How can we possibly advocate improved transit in the outer 416 and the 905 when we can’t get it right in the most easily-served part of the city?

    The TTC uses every excuse they can think of for the appalling condition of service to avoid taking responsibility for it themself. Exposing just how hollow their position is, and showing what can (and cannot) be done is an essential part of looking at all of these routes.

    I suppose that when I turn my attention to Dufferin, Bathurst, Don Mills and Finch, I will be accused of running a bus fans’ site. The shame. The shame.


  5. The point I was trying to make is that a disproportionate amount of attention is being given to streetcar routes here, when service on the TTC’s bus routes are just as bad. We never hear talk of the crappy service on Eglinton, Finch, or Islington. How about a CIS analysis of Eglinton?

    Queen and King represent a very small percentage of the TTC’s total overall ridership numbers, yet these two routes seem to get all the attention, over and over again.

    Since the TTC reads this blog, and since they’ve known about the issues on King and Queen for the last 25 years and have done nothing, it’s a waste of time and energy to keep on beating this dead horse with them.

    That energy would be better spent lobbying the City, GTTA, and the province for a sweeping regime-change at the TTC. That’s what’s needed here.

    Steve: First off, I have every intention of turning to the bus routes. I don’t have CIS data for the whole network, but do have data for Dufferin, Bathurst, Finch East and West, Don Mills and Victoria Park.

    When I started all of this, my intention was to establish that the way the TTC runs service in general has horrible problems. I have already looked at Dufferin and Bathurst on a preliminary basis, and it ain’t pretty.

    The King and Queen routes get all the attention partly because that’s where I started and partly because they are two of the heaviest routes downtown. Understanding why they don’t work is essential to the system as a whole. Queen in particular has a lot of attention here because it is the subject of very strong complaints from riders. The service is much more screwed up there than on other streetcar routes.

    If you don’t like it, you can spend all of your spare time combing through inconsistent data, building programs and analysing the fine details. It’s taken a lot of my time, and I am not getting paid for this.

    As for the dead horse, yes TTC management needs some serious housecleaning, but getting rid of managers won’t change the culture at the operating level where a laissez-faire attitude, coupled with the perception that the TTC is perfect, combined to absolve everyone of giving good service the attention it deserved. The work I have done makes it impossible for the TTC to ignore the problems any more.

    The TTC gets away with claims it can do nothing because very few people have the information to counter their mythology. Indeed, I don’t think the TTC really understands how its own system works. When CIS was first installed, there were promises of analytical tools to understand operations, but they were never built due to budget cuts and a total lack of interest. I shouldn’t have to be doing this, but I am.

    If you think that the GTTA or Queen’s Park could do any better, well you have your choice. GO Transit can’t make its trains run on time and cares only about commuters. The GTTA only wants to hear good news stories about how they’re making the world a better place, and Queen’s Park wants as little to do with day-to-day operations as possible. Don’t expect miracles from either of them.


  6. To tie in the bus/streetcar questions of levels of service, how much did the reduction of service on the cross-town streetcar lines effect the decline in service on the bus routes that crossed them, most notably 6 BAY (and even the elmination of service on 19 CHURCH, although it’s proximity to Yonge St. can explain it), but also 75 SHERBOURNE, 65 PARLIAMENT and 94 WELLESLEY. It seems to me that recent condo development along a number of these routes would spur an increase of service once again.

    Steve: The 19 Church bus was only carrying about 1,000 riders a day in the early 1980s, and the general service cuts of the early 90s finished it off. 75 Sherbourne and 65 Parliament have never been heavy routes, although Sherbourne does decent business as a connector to the Bloor subway. People can walk to an east-west route, or even straight over to Yonge faster than waiting for a north-south bus and transferring to an east-west line.

    Riding on Sherbourne peaked at just under 10,000 in the late 1980s, but fell steeply through the 1990s and is now at around 5,000 per day. Service by about a third over the same period.

    On Parliament, there has been a long slow decline in riding and service going back to the mid 1970s. Riding is down from around 5,000 per day to just under 3,000 while service has been cut by about a third.

    94 Wellesley is not really dependent on the streetcar system, but acts as a feeder/distributor for the three subway stations it serves. Riding is down from around 19,000 per day to just under 11,000 now. The big drop started with the early 1990s service cuts and, again, service today is about a third less than it was when riding was at its peak.

    The problem with many routes is that the TTC went through two rounds of cutting to the bone to squeeze more out of the system. However, the riders per vehicle mile declined on all three of the routes discussed here showing that the service cuts were counterproductive. The lower service quality has been in place for a very long time, and it will not be easy to woo back the lost riders.


  7. “However, the riders per vehicle mile declined on all three of the routes discussed here showing that the service cuts were counterproductive.”

    Steve, would that also hold true for the number of bus feeder routes south of Bloor-Danforth? Especially those in the east end between Jones and Main? I seem to recall much more frequent, and frequently used, service on these routes before reduction of service on the Streetcar lines they cross?

    Is this all also an example of how urban bus service suffers in favor of service in the burbs? (although these days, who can tell the difference.

    Steve: Bus service and ridership declined over most of the system, and I will be writing about this tonight. The only growth came in northeastern Scarborough where new construction caused population growth.


  8. To expand on the late 1980s, I was living at Sherbourne and Wellesely. I often used the Sherbourne bus and Bay trolley coach as an alternative to the overcrowded Yonge subway. Bay took me right to work at First Canadian Place; Sherbourne got me to King where I could catch a King car.

    With the big ridership drops of the early 1990s, the Yonge subway became habitable again, so riding on the parallel routes dropped off.

    The more mysterious thing, to my mind, is the dropoff of ridership on the Wellesley bus. In the late 1980s, there would be jam-packed bus after jam-packed bus. Some of the Sherbourne ridership may have been due to the fact that you couldn’t get on the Wellesley bus to go west, so might as well go south.

    The Wellesley bus had a lot of student ridership; Jarvis Collegiate and U of T were the main student groups east of Yonge. I don’t know if that’s dropped off. Maybe the demographics of the St. Jamestown apartment buildings have changed? Maybe the apartment residents have switched to cars?

    The interesting thing about the Wellesley bus in the mid ’80s was the pernicious short turn where eastbound buses looped via Parliament, Prospect, and either Rose or Ontario. (I have no idea how the old New Looks of the day made it through these narrow streets and tight corners.)

    The number of buses assigned to 94C, Castle Frank Station to Wellesley Station, is probably as good a measure of peak ridership on 94 as any. It did strike me as one of the awfulest routes for a driver: back and forth with crowded streets, high school students, apartment dwellers, and the tricky turns in and out of both stations — Wellesley has lots of pedestrian traffic; Castle Frank has high-speed vehicles on Bloor.


  9. I too hear from riders about other routes (BTW, is something happening to reliability on the 63 Ossington over the last three weeks?), but the discussion has to start somewhere.

    I took part in the “Fix the 501” forum last week knowing that the Queen service is but an example, and suspect that the hundred or so attendees there had other routes in mind too.

    I sense that the four TTC representatives who sat on the panel alongside Steve and James Bow understood that the issue goes far beyond Queen St. in the Beaches …

    Now the question is, how does the organization respond to the heightened degree of concern among riders and advocates?


  10. Keep up the good work Steve! I am only a Go-Rider at the moment, but shortly I will become a metro-pass holder as our office moves up to Bay/Bloor…it’s been informative to see all the issues and how they show up in the data, I’m looking forward to some more route analysis in the future!

    I know to some extent it’s probably comparing apples to oranges, but perhaps when you get done the analysis of the Toronto routes you could do some analysis of “well run” routes…perhaps from other cities so that we can compare our service with what is possible…

    I’d be interested also to hear about your “backend” system to convert the CIS data … is it automated? Would it be possible to generate a daily web-report of the whole system with “ontime” rates for different routes? How often do you get data from CIS? I’d imagine if you could automate and update the information daily it would hold people accountable pretty quickly.

    Steve: The data I have was a one-time export of CIS data for December 2006 for all streetcar routes and six of the major bus routes. My intent was never to provide daily reporting on this, but to inspire the TTC to build or acquire the analytical tools they so badly lack.

    My own “system” is a collection of scripts that first convert the CIS data into a useable format by removing the chaff (records I don’t want or which are manifestly invalid such as Queen cars showing up on Finch Avenue) and converting the mixture of intersection names and signpost ids to a co-ordinate system appropriate for each route. That version of the data generates the service charts with the diagonal lines back and forth across the page showing movement of the vehicles.

    The next major step reconstructs a “schedule” for the day’s operation as it would have appeared to match the service actually operated. This includes arrival and departure times at major reference points along the route. From this, it is easy to derive headways at the reference points, link times between them and charts showing the destination of each trip.

    The charts themself are built with Excel, and I structure the input data in such a way that, as much as possible, the charts build themselves (with the help of a few macros).

    Yes, it would be really nice if the TTC had a system that would generate these analyses automatically every day for every route and post then online for everyone to see.


  11. Further on Ed Drass’ comment.

    I used to use the 94 Wellesley back in the late ’80s to get to work at the CBC on Jarvis St. as I was then living in East York.

    (I well remember a lady who would get on at Parliament and Bloor wearing open-toed sandals–during a snow storm! brrrrrrrrrrr.!)

    The crowding was ridiculous, especially in the afternoon rush. As an alternative, I would take the 506 Carlton to Woodbine, switch to a 92 Woodbine South and then a 91 Woodbine, and get home faster than if I had waited for an empty 94 Wellesley and an equally empty subway at Castle Frank Stn!

    I seriously doubt I could do the same journey today on the 506 and beat the 94/B-D/91 combo, which is ironic, but alas just as symptomatic of the decline of options in quickly moving about that has also fallen victim to service cuts.


  12. Steve, how accurate do you think the published ridership figures are? If streetcar ridership figures have not been updated since 2001 then the significant increase in ridership over the past couple of years due to the transferable metropass has not been recorded.

    As for more service, how much more service do you think a route like St Clair needs? According to the service summary, it theoretically operates every 2’20” during the AM peak, 4′ during the mid day and 3′ during the afternoon. If it ran as scheduled, I wonder if there’s enough service on the streetcar routes to satisfy demand. Has there been a population decline in the area served by streetcars since 1976, as people have moved out to less crowded areas?

    Steve: I will add the data for 2004 and 2005 later tonight, along with information about bus route ridership and service.

    I agree that on many routes, if the service actually ran as advertised, it would carry the demand. Moreover, the reliability of the service would encourage more people to use it thereby justifying more vehicles on the route. St. Clair has been such a mess for years with construction that I suspect its latent demand has been driven away to other routes or to auto-based trips. Exactly how much service it needs is a hard call, and to some extent the question is whether we should be just accommodating the demand that is there, or showing what really can be done on our newly rebuilt line?


  13. It’s a shame you don’t work for the TTC in senior management. You have all the right ideas but no power to implement them. It must be very frustrating.


  14. Steve comments:

    –My own “system” is a collection of scripts that first convert the CIS data into a useable format by removing the chaff (records I don’t want or which are manifestly invalid such as Queen cars showing up on Finch Avenue)

    Well, sometimes I wonder if the scheduled car has been abducted by aliens. Occam’s Razor may prefer the explanation that it’s short-turning via Bathurst, Finch, Senlac, and Sheppard instead.

    Do any CIS records show any Queen cars in interplanetary space? That would indicate the “aliens” theory has some life left to it.

    Steve: There are no CIS signposts in interplanetary space and the odometer probably stops the moment the wheels leave the ground. Alien abductions may simply appear as long layovers.


  15. Steve said: “There are no CIS signposts in interplanetary space…”

    Are you certain? (Why else would they call them “Rockets”?) Seems to me that subways to The Greenbelt are but one step away from shuttles to the moon. Surely they would have planned ahead for space travel when the ‘futuristic’ CIS system was in development. Brings to mind things like the Queen Streetcar Subway Station shell and the never-connected trackage on Runnymede Road – truly visionary planning.

    I fully expect in my lifetime to be drifting in swanboats around the Rings of Saturn! A celestial experience to be brought to you by Metrolinx/Earthlinx/Astrolinx/Cosmoslinx… what’s their name again?

    Steve: There is not yet a lunar subway proposal because Vaughan and its developers have not yet annexed the moon. Of course, once they start a VIVA route to it, this will give a new meaning to … wait for it … a blue moon.


  16. Perhaps vehicles heading to the garage, instead of the plain “Not in service” should display “Abducted by aliens, destination Sirius. Extra fare required beyond the ozone layer.”

    Steve: But aliens require discretion when abducting TTC vehicles. Mind you, telling people where they are actually going will just engender disbelief, and so this may work as a backwards strategy.

    Of course, many riders will just get on without reading and ask “do you go to Broadview?”


  17. Steve said: “There is not yet a lunar subway proposal because Vaughan and its developers have not yet annexed the moon. Of course, once they start a VIVA route to it, this will give a new meaning to … wait for it … a blue moon.”

    A lunar subway will bring an entire new meaning to ‘sub-space’ (and possibly ‘rider-ship’). Sure, we’ll start with a VIVA route to build ridership. The trip to the moon will be really speedy (after the hour or so snarled in mixed traffic). However, it is likely GO Transit that The Province is eyeing for a lunar hop. There can’t be much modification required to send a hydrogen-powered train reaching for the skies. Might only take one tractor-trailer stuck on a grade crossing…..

    By the way, “lunar” is a very close relative of the words lunatic, loony, and loonie, all of which have been previously tied to the Vaughan endeavor. The Moon wouldn’t seem like such a stretch after all. Why is it that truly visionary transit arrives only ‘once in a blue moon’?

    (….and another transit discussion goes truly ‘off the rails’!)


  18. I wonder what effect the high fares may also have had on on dropping ridership over the years.

    If a rider is someone who for a variety of reasons can not plan ahead, and ends up paying cash for each TTC trip, then experiences unreliable service and crowds on the TTC, maybe that person starts to walk or bicycle to save that cash and improve the quality of their commute.

    p.s. I second the reporting of recent failure of the 63 bus in evening rush hours – 20 minute (or longer) waits and huge crowds are now a nightly occurrence. Too bad I can’t manage to bike given this snow and ice!

    Steve: The fares will have had some effect, but the combination of higher fares and poorer service is deadly. People will generally accept higher prices, within reason, provided that they feel they are getting something worthwhile. However, if they are paying dearly for poor service, they will be lost as customers. Getting them back requires much improved service because transit has become the mode that must convince them to change, rather than the one that drove them away.


  19. I have to agree about routing the Kingston Road cars up Coxwell. We could keep some trippers, but honestly the rerouting would clean things up considerably (probably even reduce the total number of cars needed).

    In the long term (with an LRT and the Malvern/Scarborough route), the service pattern I see as ideal for Kingston road is Coxwell station to Queen, up Kingston road and looping at Danforth road (about where room appears for a ROW, and a decent dividing point). Kingston Road LRVs would run from Vic Park station out Danforth, up Kingston Road to Morningside and form the Scarborough Malvern route. Kennedy would then become a short turn route, while most Eglinton trains would terminate at Kingston Road. I’d also extend the Gerrard cars to Vic Park (although probably only run half the cars that way) to better link Scarborough with the rest of the network.

    Steve: A network of lines in southern Scarborough is a long way off, but I will be a lot happier when planning for the Kingston Rd. LRT line gets integrated with overall Transit City schemes. Right now, the EA process and the line itself is something of an orphan.

    Going to Coxwell Station is quite another matter given constraints at the loop and nearby locations. I am not convinced that this is the ideal destination, but to be honest, don’t have other specific places in mind either. My main concern is that we not get fixated on where the 22A Coxwell goes today, or where the Kingston Road / Coxwell car went for decades before that.


Comments are closed.