The TTC’s proposed 2021 Service Plan will go before the TTC Board at its meeting on Tuesday, December 15, 2021.

In September 2020, I wrote about the draft Service Plan as it was presented for public consultation. The final version has been amended in parts and contains more detail than the draft version.

The effects of the pandemic will run through service plans for many years, but the TTC’s approach is to maintain a relatively high level of service in support both of social distancing and of the large number of essential trips that are taken by transit. The Service Plan is silent on how this will funded, and that will no doubt be a topic of the 2021 Budget to be discussed at a special meeting on December 21.

Although the TTC plans to operate at close to 100 per cent of its pre-pandemic service level, the actual distribution of service may not be the same. Service will not return to its former state with recognizable peak periods and heavy demand to the core area as long as business and educational travel is replaced by work/study from home arrangements. Until the effect of the just-announced availability of a vaccine works its way through society, we will not know when and how various activities will return to “normal”.

Demand on the system had been growing through 2020 after a trough in the spring, but that growth was blunted recently by the renewed lockdown in Toronto. The strongest return of demand has been on the bus network which serves more areas where trips are taken by transit to jobs where work-at-home is not practical. In some locations demand is already over 50 per cent of former levels and, coupled with a desire for less-crowded vehicles, service at pre-pandemic levels is required.

The TTC anticipates that system-wide demand will return to 50 per cent of former levels by the end of 2021, and this implies higher values particularly on the bus network. It will not be possible to provide social distancing beyond whatever can be achieved by fielding as many of the buses the TTC already owns. One advantage of the flattened demand curve is that the fleet can achieve better utilization with demand spread out over more service hours.

The Service Plan notes that a common complaint during the consultation process was uneven vehicle spacing and bunching problems. This is extensively documented in many articles on this site and the problem continues to this day. The only “solution” proposed by the Plan is changes to schedules so that they reflect actual operating conditions, and there is no mention of the need for much better service management.

There is a lot to cover in the overall plan, and the remainder of this article will address major topics. Some items, notably those related to customer service issues at stops, are not covered here but I will turn to them as and when specific proposals appear. Interested readers should refer to the full document.

Ridership Trends

Ridership on the TTC is running overall at about 30 per cent of prepandemic levels, but the proportion varies by mode. These values have fallen slightly from earlier in the fall before the current lockdown took effect.

The change is very different across the city and by mode. The lowest demand, relatively speaking, is in the core area at 25 per cent of former levels on the surface routes. The proportion rises in the outer part of the city where many jobs do not lend themselves to work-from-home.

The map for subway demand echoes the surface map, in part, but with some important distinctions. Although the Downsview area has the highest surface ridership at 57 percent of former levels, this is not reflected on the subway at 21 per cent. Although many subway trips in this area have been replaced by work/study at home, demand for bus service remains strong. A similar situation applies to the Sheppard subway corridor.

The chart below has appeared in previous reports, but this version is revised to reflect the recent lockdown and uncertainty about the pace of recovery in 2021.

There is a wide variety of crowding levels on buses by time of day. The chart below shows crowding levels based on automatic passenger counter data on the bus network by time of day. Note that this is relative to a standard of 50 percent of normal levels (25 for a standard bus versus the normal 50). The ranges show that there are routes at and beyond the target crowding while others fall well below and could be candidates for trimmed service.

An important issue with these values is that they are averages for routes and do not reflect the range of values bus-by-bus due to bunches and gaps in service. On average, a route’s numbers might look good, but actual values could be worse because some buses are overcrowded.

Through 2020, the TTC has operated a lot of “demand responsive” service using spare vehicles on a “run as directed” (RAD) basis to deal with local crowding problems. However, due to technical limitations of the vehicle tracking system, it is difficult to know exactly where and when these buses operated. As scheduled service levels build up in 2021, the proportion of RAD buses will fall.

The Service Budget

The Service Budget is based on vehicle hours because most operating cost (driver wages) is related to hours of operation on surface routes. The lion’s share of this is for the bus network and it has the smallest change from the 2020 budget.

The breakdown of the changes shows how various factors affect the totals.

Bus service sees a new increase in hours over 2020 while streetcar and subway services are reduced. The streetcar cut is partly offset by reducing the amount of bus service on streetcar routes. Other reductions arise from efficiency improvements as well as the absence of a leap day in the calendar.

From a financial budget view, there will be a reduction of about $20.5 million year-over-year, but there are important caveats to this.

• Most of these savings are one-time effects. They provide some elbow room in a difficult budget year, but they will not be reproduced in future budgets.
• The cost of LRT construction buses is covered by Metrolinx and the reduction in hours on that account does not translate to a saving for the TTC.
• Budget to budget comparisons do not necessarily translate to “as spent” numbers because of in-year changes.
• Eventually, demand will build up in the core area, and the saving achieved with streetcar and subway service cuts will disappear creating pressure on budgets for 2022 and beyond. This will be only partly offset by revenue from higher ridership.

The Fleet Plan

Plans for fleet requirements vary substantially by mode over the next five years. Total vehicles on the bus network will be essentially static, although this is in part due to the replacement of some bus service by other modes:

• The streetcar system will (eventually) return to 100 per cent streetcar operation.
• Opening of Line 5 Eglinton in 2022 and Line 6 Finch in 2023.

The buildup of streetcar service in the short term reflects a return to full fleet availability after major repairs and retrofits from manufacturing defects. At 168 cars in 2022, this represents a spare factor of 21.4 per cent, close to a typical industry value of 20 per cent. Beyond that point, the increase depends on purchase of more streetcars and expansion of carhouse space that are not yet fully funded. Any new cars would go to service improvements on the existing system as expansion projects will probably not come on stream within this five year window.

Service build-up on Line 1 Yonge-University-Spadina reflects closer headways possible with the automatic train control system. Note that two factors are at work here:

• ATC allows trains to run closer together and so more trains/hour can be scheduled on the line.
• ATC allows trains to safely run faster reducing the number of trains needed to provide service.

These factors combine to allow a greater capacity improvement than the change in peak train requirements would imply on its own.

Line 2 Bloor-Danforth will not see ATC operation over its full length until late in the 2020s and when there is a new ATC-capable fleet. The timing of these projects is uncertain due to funding issues with the overall capital budget.

Line 3 Scarborough is currently running with only four of its seven 4-car sets in peak service. Five are scheduled, but they are not reliably available. A report on whether to attempt life extension of the SRT until the planned Scarborough subway opening is expected to come to the TTC Board early in 2021.

That will be a difficult decision either way due to the cost and technical challenges of running the SRT possibly until 2030 as opposed to the service implications and political embarrassment of replacing the route with buses for several years.

Line 4 Sheppard will see no service improvement in the next five years.

Express Buses

Service on much of the express bus network will be restored in 2021, and new routes will be added:

• 938 Highland Creek (See eastern Scarborough route changes below)
• 968 Warden Express (Warden Station to Steeles)
• 943 Kennedy Express (Kennedy Station to Steeles)

Express service will be added during some periods on four existing routes.

The service plan includes a review of eight new express services with a “scorecard” for each based on prepandemic conditions. These are consolidated for convenience on one page on a spreadsheet. The colour coding here reflects the scheme shown in the service plan. Some aspects of this are worth noting:

• Some services were overcrowded based on prepandemic standards.
• The balance between crowding on local and express services indicates that the relative proportion of these services should be changed.
• The speed savings hoped for with express operations have not always been achieved.
• Change in ridership varies considerably from route to route. Although latend demand is a factor, there is also the question of whether there was any actual service improvement overall, or merely a reallocation of buses between local and express operation.

This analysis is a useful way to look at the success (or not) of express vs local service, and it should be produced for all routes, not just those recently implemented.

The Pitfalls of “Cost Per Passenger”

My apologies for a short excursion away from the Service Plan itself, but this discussion is necessary.

The table above showing, among other things, the alleged cost per passenger of express services brings up a long-standing issue with TTC service evaluation. Various metrics have been used over nearly four decades of the evolution of service standards.

• Net loss (or cost) per passenger
• Peak demand
• Boardings per service hour

Service cost, especially off peak, is driven primarily by service hours with wages as the main component. The vehicles already exist, and some of their ongoing cost is not affected by usage.

In a flat fare system, it is impossible to allocate a “fare” to each leg of a journey. Even if there were fare zones, a rider might travel on more than one route within a zone, but only contribute one zone’s worth of revenue. The permutations get even more complex when one considers various discount schemes including passes, timed transfer privileges, and multi-zone discounts.

In practice, the TTC calculates an average fare per boarding. This is not the same thing as an average fare because most trips include at least one transfer, sometimes more, and there are many more boardings than trips. The fare per boarding is allocated equally to each leg of a journey regardless of its length or the proportion of a trip the boarding represents.

This has two major effects:

• The revenue per passenger hour (or passenger kilometre) is diluted for long trip segments, and inflated for short ones.
• More revenue is allocated to trips with more than the system average number of transfers than the rider actually paid, and less to trips that have no transfers at all.

These produce distortions in the comparative cost/passenger between routes. A few examples:

• A one-seat trip (no transfers) contributes only about half a fare because the typical journey includes one transfer. Routes with a higher proportion of one-seat rides are allocated less revenue than they deserve.
• Short routes necessarily have short trips and therefore a low operating cost per trip. Riders cannot make long journeys on them, and the calculated net cost/passenger is therefore very favourable.
• Long routes can have a mix of trip lengths but they receive only one trip’s worth of revenue even if they carry riders a long distance. This drives up the calculated net cost/passenger.
• Some routes have good turnover of riders and therefore a high trip count. Three riders each travelling 2 non-overlapped km contribute 3 trips’ worth of revenue, while one rider travelling 6 km contributes only 1 although the same resources are consumed in both cases.
• Transfer points, which can be an artificial result of route design, create new boardings on what would otherwise be a single long trip.

Of particular note in the Express Bus table above is 913 Progress which is alleged to operate on a “break even” basis. This is entirely an artifact of the methodology. This is a short route which operates as “express” only in the peak direction. This gives it a combination of strong demand over a short distance, and no penalty from running with little demand in the counter peak. Under these conditions, it is easy for a route to have good financial performance from a skewed cost and revenue allocation.

Politicians and managers who prefer simplistic measurements love a “cost per passenger” because it is something they can understand. Whether the number actually measures the true cost, or more importantly the net cost or saving possible with service changes, is quite another matter.

There were times when the TTC would argue that it does not really matter how one calculates a route’s cost because the bad routes will have terrible numbers no matter what formula is used. That is true for worst case situations where only the outliers in performance tables are targeted for cuts. However, we are now in an era where pressure will arise to trim more, and false calculations of costs and savings could harm the system.

Some routes are inherently expensive to run because they cover a lot of territory and carry many long trips. This does not mean that service on them should be cut.

Ironically, the TTC does not use a low cost/passenger as a trigger to improve service because they base such changes on vehicle crowding, a vital factor when deciding whether to add or remove service.

The TTC has not published calculated daily operating costs on a route-by-route basis since 2012 and their reappearance as part of service reviews shows a return to a potentially misleading way of looking at system performance.

The chart below shows the distribution of cost/passenger values by time of day based on prepandemic riding and service levels. Routes that are flagged are the subject of proposed service reductions or elimination.

Routes Confirmed as Permanent

Almost all route changes implemented for the Spadina subway extension to Vaughan have been confirmed as permanent parts of the system except where noted in following sections.

Route 12D Kingston Road to UTSC does not meet the TTC service standards, but gets a bonus because it serves, albeit very infrequently, a priority neighbourhood. In 2021 it will gain midday service.

Route 175 Bluffer’s Park has performed well and will become a permanent route. The TTC evaluated weekday service and found it would meet standards, but the Service Plan does not include this.

An example of competing metrics for service is shown by the 176 Mimico GO bus. Although it has a high cost per passenger in the AM peak, it meets the TTC standard of boardings per service hour and is recommended as a permanent route. (The route is currently suspended due to very low demand on GO Transit.)

Proposed Service Changes

East Scarborough Bus Network

As proposed in the draft Service Plan, routes will be reorganized in eastern Scarborough. The plan does not include an implementation date for the revised network.

• The 905 UTSC Express route will be extended east along Ellesmere to serve the Conlins Road area replacing the existing 116A branch of the Morningside bus.
• A new 938 Highland Creek Express will operate from Scarborough Town Centre to UTSC in the peak direction (eastbound in the morning, westbound in the afternoon). Buses will continue as locals to Rouge Hill GO and over the entire route counter-peak.
• The 95 York Mills service will be extended east from Meadowvale to Kingston Road and Sheppard.
• A new route 178 Brimorton will operate from STC replacing part of the 54B Orton Park service as well as the 86D Scarborough service to Coronation Drive.
• The 54B Orton Park branch of Lawrence East will be extended east to Morningside.
• Service on 86 Scarborough will be improved in the Highland Creek area.

Wilson to Stanley Greene Service

A trial service will operate from Wilson Station to the Stanley Greene area to improve access south of Downsview Park and to provide additional service on Wilson Avenue.

The 121 Fort York-Esplanade bus has never done well on its western branch because of competition from other routes and frequent delays from events in the area. The TTC proposes to eliminate service west of University Avenue and concurrently to extend the route east and north via River Street.

121 Fort York-Esplanade Service

This will improve service in the East Don Lands and bring a north-south service to River Street which has seen a great deal of redevelopment in recent years. No new name has been proposed for this route yet.

117 Alness 107 St Regis Consolidation

These two routes were flagged as poor performers. Two options to improve their standing were presented to the community, and the preferred option is the consolidation of the routes into a large loop. This will remove the link to Pioneer Village Station but preserve the connection at Sheppard West Station.

Modified Hours of Service

Midday service will be added on:

• 12D Kingston Road to UTSC
• 119 Torbarrie

The following routes will lose their late evening service due to low ridership:

• 28 Bayview South
• 33 Forest Hill
• 167 Pharmacy North (Monday-Friday)

Proposed Changes on Hold for Review

• 101 Downsview Park is a poor performing route, but it will be maintained pending review of an improved routing to provide two-way service via Carl Hall Road rather than westbound along Sheppard where it duplicates other services.
• 162 Lawrence-Donway is a poor performing route that serves the Bridle Path. Its status will be evaluated as part of the route reorganization for opening of Line 5 Crosstown in 2022.
• Route 123F Sherway peak service to Sherway Gardens will continue, although it will be reviewed in 2022 due to ongoing changes in employment in the area.
• Route 900 Airport Express is a high-cost route because it serves relatively long point-to-point trips. Depending on ridership in 2021, it could see service changes.
• Route 903 STC Express is a parallel service to the SRT. A decision on the future of this route will be included in the review of the future of the SRT rapid transit service.
• Route 62 Mortimer was considered in the draft plan for a late evening cut. However, its ridership has not dropped during the pandemic period suggesting that it carries essential trips. It will be reviewed again as part of the 2022 plan.
• The 14x Downtown Express routes will remain suspended while demand to the core is light.

Major Construction Affecting Streetcar Services

Several construction projects will affect streetcar services in 2021. Replacement service designs have not been announced. However, it is clear that Queen Street will take the brunt of work during various periods.