“What would it cost to put service back to pre-pandemic times?”

That question comes my way as riders deal with another round of service cuts, and would-be mayors vie for attention. The answer is not simple, but an unexpected statement at the recent TTC Board meeting surprised me at how low the barrier to full service was claimed to be. Responding to a question from Commissioner/Councillor Chris Moise, the TTC’s CFO stated that the cost would be $69.5 million/year.

Although hardly small change, that is a lot less than the depth of service cuts might imply. That sent me on a dive into TTC budgets and stats to validate the TTC’s claim.

I asked the TTC how they arrived at that figure and their reply is below. The important point here is that they are comparing budget-to-budget from 2022 to 2023. Originally, they had expected to restore full service in 2022 and budgeted accordingly. From that point of view, the change needed to return to 2020 levels was simply a matter of going back one year.

From the 2023 budget…page 29.

$69.5M

Net base pressure exclusive of COVID and lines 5 and 6.

This represents increases on 2022 budgeted hours which were in line with pre-pandemic service certainly comparable for this purpose.

So that’s the number Josie gave.

From Page 21 of the 2022 budget…

We basically budgeted 100 per cent of pre-pandemic service in 2022 (Q1 of streetcar and subway notwithstanding).

Email from TTC Media Relations May 9, 2023

However, this illustrates a common problem in budgeting. 2022 might have been budgeted at close to 100% of 2020 service, but that is not what was actually operated. On an actual basis, service and costs did not grow back by the second quarter. TTC avoided a $20.7 million draw from a City transit reserve and did not take the full budgeted subsidy. If 2020-level service had actually been operated, the 2022 costs and subsidies would have been higher.

That $69.5 million represents the cost to get back to 2022 budgeted levels assuming subsidy revenue higher than the TTC actually received. The 2022 budget also included funding to begin service on Lines 5 and 6, but these lines are still under construction. This gave the TTC some headroom to absorb other cost hikes such as for diesel fuel. When those new lines do open, likely in 2024, TTC’s costs will unavoidably rise and that is a separate pressure from any service restoration.

Looking at 2023 service, the TTC cited an unexpected bonus – the workforce is healthier with fewer covid-related absences than the budget assumed. The current (May 7) schedules were designed in January based on then-projected staffing and there are now has more spare operators than expected. Already, TTC back-pedalled on a deep cut in late evening subway service, a change which has low marginal cost because spare operators are already being paid.

ATU Local 113 responded to TTC’s claims about absenteeism by noting that there are many spare operators with no work to do because of a political decision to cut service.

What we do not know is the number of spare operators and the amount of service that could be added at little extra cost because operators are already on the payroll and the TTC has available vehicles. This should an issue in planning the September 2023 service which should already be underway.

The TTC has a large pool of spare buses, streetcars and subway trains thanks to service cuts and an over-large fleet even in pre-pandemic times.

See: How Many Buses Does The TTC Use?

There would be a marginal operating cost to run more vehicles (fuel, servicing, routine maintenance), but this is considerably less than the full cost of adding net new vehicles to the fleet. Moreover, with new bus deliveries the oldest buses (with their higher operating cost and lower reliability) will be retired.

Broadly speaking, there are five types of operating cost:

• Hourly: primarily wages and benefits
• Mileage: the amount of usage a vehicle gets, i.e. wear-and-tear, energy consumption
• Vehicles: routine servicing and garaging
• Infrastructure: tunnels, track, signals, station costs, garages, roadways
• Administration

Boundaries between these groups are not solid. For example, some benefits are a function of headcount (vacation pay) rather than of hours worked. Some infrastructure wear and maintenance can be affected by the level of service (well used track wears out faster than little used sidings and diversions). Vehicles running on slow, busy streets have different duty cycles (acceleration, braking, door operations) than on faster, especially express, routes.

Only with a large change up or down in service levels and network scale do all five cost groups change. This is an important part of budgetary “savings” or costs at the marginal level. Twiddling with small service changes does not, cannot, save 100-cent dollars because only the marginal costs change. Only hourly and mileage costs vary with modest changes in service levels because there is no change in the fleet, infrastructure or management.

Operator hours are the basic variable cited when comparing service levels. For situations where the average speeds lie in a narrow range, hours and mileage are equivalent. Mileage related costs differ when route characteristics or technologies do not match.

How Much Service Does The TTC Run?

This question, and its equivalent “just what is full service”, is not straightforward to answer for various reasons:

• Operators who are on standby rack up hours, but do not necessarily provide service proportional to their numbers.
• When schedules are adjusted for “service reliability” this usually means that travel times are increased, vehicles run slower and riders see less frequent service. However, the number of vehicles and operators (and hence operator or vehicle hours) does not necessarily drop and might even rise, even though vehicle mileage goes down.
• Vehicles of different sizes (10m and 13m buses, streetcars, trains) have different capacities. Service capacity depends on the combination of vehicles/hour and vehicle load. The cost base will change with the fleet mix or technology on a route.
• Trains with one person crews have half the operator hours of those with two person crews although both accumulate train mileage at the same rate.

Here are two charts of the planned weekly bus hours from January 2020 to May 2023. The one on the left has the horizontal axis well above zero to emphasize the month-by-month changes. The one on the right has the axis at zero to show how the amount of change has been relatively small sitting on a core base of service.

(Note that these values do not include extra service for construction projects. Also, new schedules are implemented roughly every 6 weeks and so there are gaps in the month-based charts.)

Source: TTC Board Period Service Change Memos

Here is a chart showing the numbers for rail modes which are an order of magnitude smaller than for the bus network. This difference emphasizes where a large portion of costs lie and where big savings, if they are demanded of the TTC, will have to come.

When the TTC talks about cutting subway service to preserve operator hours for the bus network, it is important to remember that the subway does not have a large pool to begin with.

A few notes about this chart:

• Lines 1 and 4 are combined because they operate out of the same carhouses and with a common pool of operators.
• Line 1 switched to one person train operation (aka OPTO) in November 2022 between St. George and Finch stations resulting in a large drop in operator hours. Between St. George and Vaughan, this change occurred on Sundays in August 2021, and on Weekdays and Saturdays in November 2021.
• Because construction effects are not included, the streetcar hours are based on what would have operated were all routes actually running with streetcars, not buses.

The effect of construction on hours operated by surface modes is illustrated by the two charts below with buses on the left and streetcars on the right. The base bus hours are in red and the hours with construction service added are in brown. For the streetcars, the base is in purple and the effect of construction replacements in dark red.

Extra bus service is provided mainly for construction activity on the streetcar network and on new rapid transit lines.

The actual ratio of planned (before construction adjustments) service in May 2023 and January 2020 by mode is:

• Bus: 96%
• Streetcar: 91.7%
• Lines 1 & 4: 39.0%
• Line 2: 81.6%
• SRT: Nil

The bus network is operating at close to 2020 vehicle hours, but this is partly offset by “reliability” adjustments that reduce service without reducing hours. This also applies, but to a lesser extent, for streetcars.

Again, note that the big drop for Line 1 is due to OPTO conversion, and this part of the change would not be backed out in any service restoration. In March 2023, one train was removed from Line 4 during all periods, but this is a small portion of the total hours for Lines 1 and 4 overall.

The SRT capacity was limited before 2020 by equipment reliability.

A further issue in service restoration is that pre-pandemic service was already crowded on many routes. Just getting back to that level will not solve the larger issues of service adequacy and quality which have been conveniently out of sight for three years. Changes in commuting patterns will also affect the demand for service both on a route and time-of-day level. “Putting back the hours” will almost certainly not mean “putting back the same schedules”.

Service Comparisons May 2023 vs January 2020

Updated June 1, 2023: A few errors in the May 2023 headway values have been corrected.

The chart below is an update of one I published previously. It compares headways (time between buses, streetcars and trains) in May 2023 with the service operated in January 2020. There are a few caveats and notes:

• This chart compares frequency of service, not vehicle hours. It shows the change in headways in “mm:ss” format and the percentage change.
• Where 2023 and 2020 services are identical, the cell is blank.
• Colour coding:
  • Green: Service improvement
  • Pink: Service reduction
  • Blue: New service
  • Red: Discontinued service
• The TTC has already implemented some of the seasonal changes, and direct comparisons to January 2020 should bear that in mind on routes with academic or recreational demand.
• I have adjusted late night subway headways to 6 minutes in recognition of the TTC’s change, as well as including the 10 minute bus service on 503 Kingston Road which is not in the published schedules.

There is a lot of pink on this chart, and many of the percentage changes are not trivial. This is the difference between talking about overall average changes and individual cases.

Only the weekday info is shown below, but the weekends are available in this PDF containing the full chart set.

The Back of My Envelope

In trying to come up with a figure to restore service, I took a different approach from the TTC’s.

In 2023, the planned weekly service hours are about 172k compared with about 186k in 2020. This is about 92%, a number that aligns with past TTC statements. As I noted above, the actual ratio varies by mode, and the bus network is running at a higher level.

In any event, an 8.7% increase in service hours overall would be needed to get from 92% service today back to 100% service. (As I have said before, this would not all be in the same time or place because the network and demand patterns have changed over the last three years).

The Operating Budget for 2023 totals $2.32 billion (net of expansion, SRT conversion and covid costs). For the sake of argument, assume that 70% of the budget is variable costs (see the earlier discussion about components of costs) and 30% is fixed. We are only concerned with that 70% which would be $1.62 billion.

Increasing that by 8.7% would cost about $141 million annually. I believe that this is a more realistic value on an actual-to-actual basis than the $69.5 million budget-to-budget number cited by the TTC. Obviously some assumptions, such as the portion of the budget that it truly affected by service changes, are open for debate. There is also the question of the degree to which hours can be freely traded between modes such as the subway-to-bus as the TTC has argued.

A debate about service restoration and what sort of target Toronto and the TTC should aim for is a vital one. There will be competing transit costs in 2024 notably the opening of Lines 5 and 6 (yes, hope springs eternal), the effect of any fare changes, not to mention other civic portfolios. This debate should be well-informed, and it should not wait until January 2024 with staff springing a completed, undebateable budget on the city.