The Siren Song of Regional Fare Integration

The Toronto Region Board of Trade recently published a discussion paper on the subject of regional transit integration focused on fare structure and the barriers it creates to regional travel.

See: Erasing the Invisible Line: Integrating the Toronto Region’s Transit Networks

This is to be the first of four papers with others to follow on subjects such as increasing utilization of the regional rail network and improving service on the local bus networks around the GTHA. No publication dates have been announced for the remainder of this series.

The absence of those papers leaves the first one on fare integration out on its own missing some of the context that drives choices of what might make an appropriate “solution”. In particular the key roles of both GO Transit and local transit systems, or at least as the Board of Trade might see them, inform the proposal of a new fare structure, but more as background. Are these assumptions valid and does a new tariff based on them actually stand up to scrutiny?

Schemes to unify the regional fare structure have floated around the GTA for years. Lots of ink was spilled on reports, models and consultation. Nothing much has actually happened, or at least that’s the impression one might get, in part because different players have different goals.

Metrolinx is absolutely wedded to a zone fare system because that is how their fare collection technology works. They speak of it as “fare by distance”, but their zonal structure contains many inequities because it evolved piecemeal along with their network. Long trips are cheaper than short ones, measured in cost/km, both to discourage short-haul riding and to give greater incentives to long-haul commuters to switch from their cars to GO’s trains. Relatively recently, GO introduced reduced short-haul fares so that it could attract more short trips, but the tariff as a whole remains a patchwork.

When Metrolinx first proposed a distance-based regional fare strategy, it had an added wrinkle with a premium fare for “rapid transit” which meant anything on rails on its own right-of-way including the subway. Any trip longer than 10km (slightly above the average trip length on the TTC) would cost more than it does today, and drawing 10km circles around various centres easily shows who would pay more to travel. This had the effect of preserving GO Transit’s revenue stream, while raising the cost of subway travel for longer-than-average journeys.

This was in aid of a “zero sum” solution where the cost of lower fares for riders crossing the 905-416 boundary would be recouped from higher fares within Toronto. Metrolinx showed only a few sample fares to illustrate changes, but neglected to present a thorough review of the effect on TTC riders who are by far the majority of transit users in the GTHA.

In time, Metrolinx, or at least some members of its Board, came to realize that this was not a viable solution, and that any new fare structure would require added subsidy to avoid penalizing one group of riders to reduce fares for others. Alas, nothing official ever came of this.

The regional transit agencies were not sitting still, however, and the now-universal fare model is based not on distance travelled, but on the elapsed time for one or more trips, in effect a limited duration pass. Not only is this scheme easy to understand and administer, it removes a long-standing penalty against riders who took multiple short trips, typically to run errands or stop off in a longer journey just as one would do as a motorist.

Even Toronto, after much foot-dragging, embraced the two-hour transfer when it became politically beneficial. What was once portrayed as an unaffordable fare giveaway morphed into a modest-cost change that greatly simplified fares and improved system convenience. The only remaining gap in this arrangement is the lack of reciprocity across the 416-905 boundary so that a two hour fare can buy rides inside and outside of Toronto.

The odd man out remains GO Transit, a regional, long-haul carrier, an operator of fast trains where two hours would take a rider a far greater distance than on a bus, streetcar or subway. There will always be a conflict between seeing GO as a “rapid transit” line serving local demand as opposed to “commuter rail”. Just to complicate things, GO buses fall somewhere in between because they operate limited stop service with much more comfortable accommodation than, say, the Dufferin bus.

GO faces an additional problem with a penny-pinching master at Queen’s Park for whom spending more money on transit operations (as opposed to capital construction) is not a priority. Even the GO-TTC co-fare was eliminated although it remains in place for GO-905 travel. It is ludicrous that a “first mile” trip in the 905 gets a co-fare subsidy, but not one in Toronto.

First Principles

I spoke with the Erasing the Invisible Line’s principle author, Jonathan English, to explore the rationale behind the proposed regional fare structure.

Certain themes, outlined at the outset of the report, are key:

  • There is no appetite for one gigantic agency. Local autonomy and control over transit are essential. Indeed, an amalgamation would bring years of bureaucratic, financial and political wrangling that could well be interrupted by a change in government(s) and a new transit orthodoxy.
  • Any fare system should have “mode neutrality” so that it does not reward or penalize riders whose trips are best served by one mode versus another, especially considering that Toronto’s system was designed with this intent.
  • GO rail services and the subway should not be viewed as premium modes, but integrated in the TTC and the local networks beyond the 416.
  • There should be minimal unnecessary disruption in any change leaving Toronto’s flat fare intact, but without unreasonable subsidy of long trip commuters on GO.

These are noble goals, but it is not clear that they can all be achieved at the same time.

There are two fundamental assumptions here:

First: Past Board of Trade proposals doomed their credibility by treating the “governance” problem as one that could only be solved with centralization. There is a classic flaw of thinking that if only the “right people” were in charge, all our problems would be solved. I am sure that the Board’s idea of the “right people” and mine would be very different.

The concept of a grand, unified agency as a way to “fix” transit has been replaced by the European concept of a Verbund, an association of agencies serving a common territory. This puts the focus on finding ways to operate and improve transit rather than playing musical chairs with transit management and politicians.

Second: There is a clear need for more revenue to make a “better” fare structure, whatever it might be, work. This leads to questions of how much new subsidy would be required, who would pay, and how reliable a partner they would be in any arrangement.

The “zero sum” solution with frozen subsidies is nowhere to be seen nor is the idea that if only riders paid “their fare share”, there might be no need for transit subsidies at all.

Chatting with English, I argued strongly that no matter how attractive the fares might be, it is the quantity and quality of service that determine the attractiveness of transit. Service is a key factor, English agrees, and he has written about the quality and importance of suburban TTC service in the Globe & Mail. Unfortunately, the “service” volume in the Board of Trade’s planned tetralogy will be the last to be published even though its topic is key to making any transit network actually useful for riders.

This is a difficult problem, at least as challenging as fare levels, because the quantity of service – frequency, hours of service, coverage – in much of the area outside Toronto is nowhere near as good as inside. Having a cheaper fare to cross Steeles Avenue or Etobicoke Creek is little value if the “connecting” bus service is infrequent or non-existent. This applies whether there is a forced transfer at the boundary or if the connection occurs at a regional terminal such as Finch or Kipling Station.

“Better service” usually translates into demands for “more buses” and possibly a new garage. But those buses are useless without operating subsidies so that the fleet can actually show up for riders rather than sitting in a shiny new depot. “Accessibility” to transit is not just a question of a few ramps and blue seats, but of service that riders can count on to arrive promptly.

The Board of Trade Proposal

The Board begins by mapping transit usage in the GTA. The report notes that the rate of transit usage in the 905 is lower than within Toronto “in part as a result of fare boundaries”. The other part, of course, is that there is a strong disincentive to use transit in the 905 because of service levels, route patterns and development patterns. Making travel between the 905 and 416 cheaper will help riders making that type of journey, assuming that service exists that they can use, but it will do nothing to improve transit’s attractiveness for travel within the 905.

Presenting the problem with a strong focus on Steeles Avenue, Etobicoke Creek and the Rouge River is a very Toronto-centric view of travel patterns and needs. This is not to say that the cross-border fares are “fair”, but there are many more trips in the GTA that transit does not serve as a visit to any suburban arterial or expressway will demonstrate.

To put it another way, if your obsession about transit is that you pay two fares, you might focus on the tariff rather than any other factor in “solving” the transit problem.

Within Toronto, there is a problem that populations who are more transit captive are concentrated in the parts of the city where housing is cheaper, but where transit is less attractive. The map below shows Toronto’s “Neighbourhood Improvement Areas” (or NIAs) but significantly it does not show where people living in these areas go to work or school, nor does it show comparable areas and trip patterns for the 905.

An underlying reason the Board of Trade would like to see a new tariff is to encourage use of the GO rail network for trips within Toronto. I will get into that in more detail later, but for now the chart below shows the problem as they see it. A hypothetical rider travels from Steeles & Markham Road to Union Station. By TTC, this is a single fare but it takes an hour and a half (peak period, local bus service used). In theory, the rider could transfer at Milliken GO station and get downtown 40 minutes faster, but at an extra cost of $7.62. The fares shown do not include any concession allowances, pass-based fares or loyalty discounts.

Both Agincourt and Milliken GO stations are far enough from Union that they are not part of GO’s reduced fare zone for “short” trips which only kicks in at Kennedy and points closer to Union. Therefore, the fare comparison here emphasizes how GO’s cost is disproportionately high for some trips even within Toronto.

Not shown is the GO service level today. The presumed travel time saving requires that a rider make a good connection to a rail line that is much less frequent than the TTC (hourly service most of the time). GO plans to have frequent service on this line some day, although exactly how much depends on which plan you believe. This is also, of course, SmartTrack territory, and we still have no idea how much GO will charge the City to provide ST service at TTC fares. That operating subsidy would almost certainly compete with the local TTC services for funding.

With any fare scheme, it is possible to construct trips that show the tariff in a bad light. By contrast, if the GO-TTC co-fare were still in place (a $1.50 discount) and if the $3.70 short-trip GO fare at Kennedy were charged at Milliken, the total fare via GO would be $5.45 or half of the example cited in the map. [$3.25 for TTC plus $3.70 for GO less $1.50 co-fare]

Any attempt to concoct a tariff that treats GO trains and local transit equally implies that the fare should be the same no matter which mode one uses to travel from “A” to “B”. However, preserving Toronto’s single fare requires a large zone through which any rider can travel for a flat fare rather than the more finely-grained zones needed to preserve GO’s quasi-fare-by-distance scheme.

The Board of Trade proposes a zone system where travel between adjacent zones would be free, but travel through more than two zones would attract an extra fare. Toronto is carved into an inner and outer zone so that travel anywhere in the city would be at a base fare, as would travel between the outer (larger) zone and the immediately adjacent zones in the 905

The northern boundary of Zone A (Inner Toronto) is Eglinton Avenue on the dubious premise that someone should be able to travel across the city on the Crosstown and beyond the 416 boundary for one fare. However, someone who took the GO train via Union (or even the Bloor-Danforth subway) would pay a surcharge. This is complete nonsense.

Someone who crosses Steeles from York Region but stays north of Eglinton would pay only $0.50 more than today, but if they worked at Davisville (TTC headquarters, say) they would pay a lot more. Putting the core area its own zone makes a trip from Mississauga, Southern York or Durham (Zones C, E and G respectively) a three zone trip for a total fare of $5.75. This is not a huge discount compared to current bus+subway fares and certainly not an elimination of the invisible line at Steeles Avenue.

Outside of Toronto, the proposed zones split some municipalities such as York and Durham making any trips within them a single fare, but imposing a surcharge for trips that cross both an internal and external boundary.

Like Metrolinx, the Board of Trade fails by omitting before and after comparison of fares for typical trips.

This map illustrates the huge problem of attempting to consolidate GO’s fare structure strongly rooted in a radial, commuter oriented history, with a grid of municipalities and their travel patterns which are completely absent here.

This produces some odd results because the region is, more or less, rectangular, and there are more zones for east-west trips than north-south. Additional problems arise for trips that go diagonally across the zonal grid. For example, on this map a trip from Waterloo to Toronto by train skirts the northern tip of Mississauga. Should riders pay an extra fare because the CNR did not align its rail corridor with a municipal boundary?

As for the tariff itself, the rationale behind the various zone-to-zone fares is not explained, and one cannot help think that some tweaking went on to get the numbers to “come out right” in a regional revenue projection. I have split the tariff itself out of this map to display it at a resolution where the numbers are clear.

Among other problems here, note that the ratio between monthly passes and the single fare varies considerably. The zone A-B pair (inside Toronto) is 48 based on the existing TTC multiple. Outside of Toronto, passes are priced at $125 or $150 against base fares ranging from $3.25 to $3.75 with multiples from 42 to 38.

English (as do many other advocates for cheaper TTC fares) feels that the 48 multiple for TTC passes is out of line. However, the basic issue has always been that there is much more off-peak riding in Toronto simply because the TTC has so much more service during that period compared with the rest of the GTA. A pass priced strictly at 20 round trip commutes per month would represent a considerable revenue loss. Indeed, it took heroic budget fights to get the multiple below 50.

The multiple actually used to be close to 40 until the transit tax credit was eliminated. That credit might have been a boon to commuters looking to reduce their costs, but it brought no new money to transit coffers.

Another anomaly here is that the multiple for the extra zone premium is less than 40 ranging from 28 to 35. This perpetuates the long-commute discount already found in GO fares.

The fares shown are for “adult” riders who do not have any discount due to local policies such as senior, student, child or low-income fares.

There is a huge complexity built into all of this, not to mention the incentive for long debates to gerrymander the map to suit local interests. That could make the discussion of new station names on the Crosstown seem positively simple by comparison. (Metrolinx famously devoted an entire Board meeting to one station with two competing names, and in the end opted for a hyphenated solution.) This scheme would replace one invisible line with a multiplicity of others.

Why, one might ask, are time based fares, a standard already used across the GTA and built into Presto, not on the table? The reason, of course, is that this conflicts with integrating GO’s fares into the scheme. The tail is wagging the dog here. Of particular note, there appears to be no recognition that time based fares should survive so that, for example, a trip south then north across Steeles in a brief period would not attract another fare.

Tap-on, tap-off would be required across the system to capture destination information and calculate an appropriate fare. Claims that other systems do this routinely are misleading on a few counts.

First, and most importantly, the existing local transit systems are not set up for this. Successful implementation should include a powerful incentive to changing everyone’s fare-payment behaviour, but this is a change for administrative, not rider convenience.

Second, some systems have a mixture of flat and zone fares, and this eliminates the need to tap off for some rides. London, UK, is often cited for its distance based fares, but these only apply to the rail networks, not to the bus system.

Third, the hassles of getting people to tap off are well known to GO who provide the option within Presto for a “standard” destination where one does not have to tap off. This works because most GO trips occur between the “home” and “work” stations of a typical commuter. When GO Transit gets rid of that option, then we can talk about the “convenience” of tapping off.

If you start out with the premise that all fares must be unified including those of the commuter railway, then all manner of distortions and complexities are inevitable in the local transit systems. On its face, this looks like a “reasonable” arrangement, but actual implementation would bring many headaches both operational and political.

Looking Into the Details

Mode Neutrality: The current network is neutral as to mode except for commuter rail and a handful of now-suspended premium express bus routes. Whether you cross the city by bus, streetcar or subway, the fare is identical. What is not, however, is the service level either within Toronto itself, or across the region.

Subways are typically over-served off-peak as a matter of policy regardless of demand, and because the marginal cost is lower than for other modes. (Subways have a large fixed cost even if trains run infrequently.) By contrast, surface routes are scheduled based on demand (and budget) with a maximum permitted headway in Toronto of 30 minutes. Service can be considerably less reliable than the advertised schedule thanks to bunching, gaps and a laissez-faire approach to service management.

Bus services outside of Toronto are reasonably frequent in some areas, but not in others. Their span of service may be oriented to peak commuting, but not necessarily be good for off peak and counterpeak travel.

One might speak of “equity” in fares, but a bus that runs every half hour is little more than a line on a map, not a service. This is one big reason for the demand for subways – not only are they faster, they will have guaranteed frequent service from 6 am to 1:30 am unlike the bus and streetcar routes.

GO Rail corridors are under-used: This is a strange position for the Board of Trade to take considering that they have hosted presentations by Metrolinx CEO Phil Verster about the marvels of GO’s expansion plans, and have also been quite warm about John Tory’s SmartTrack scheme, ephemeral though it may be.

Planned service levels on GO can be hard to pin down because different numbers are cited in different circumstances. The current round of public consultation on GO expansion, as described in a Metrolinx Blog post of November 27 enthuses:

Get a glimpse at the latest design of a transit system that will deliver 15-minute two-way, all day service on core segments of the GO rail system.

However, tables in the GO Business Case Analysis for the same project cite more frequent service, even during the off-peak, on some corridors. The Stouffville line, for example, would see 8 trains/hour each way if one believes that report.

Still other service levels have been cited in reports for SmartTrack with “subway-like” service of 10 trains or better per hour, at least if the demand projections for the service are to be believed. Currently, with the Mayor no longer running for election on that scheme, the proposed service level is every 15 minutes which just happens to be what GO claims they will run anyhow.

Jonathan English argues that there are many rail corridors around the world with very frequent service, and that by analogy more could be squeezed out of the GO rail network. This begs two questions. First, can more service and riders be carried on GO even beyond their current plans (whatever they actually are). Second, what are the implications of concentrating very frequent service on corridors that meet at Union Station.

An important distinction here is not just the number of trains per hour, but what these trains might look like. There is a big difference between a locomotive-hauled 12-car train of GO bilevels and, say, a five car train of electric multiple units. (EMUs are essentially the same technology as subway cars and streetcars, but with a different body and performance characteristics). Running a five minute headway (12 trains/hour) does not necessarily mean running the huge trains we see today. This is a tradeoff between service frequency, convenience and operational constraints. The choice affects both infrastructure planning and future network capacity.

All of these claims are not just a question of shifting political and spending priorities. As GO rebuilds its physical network, certain assumptions are built into the track and signalling design that are driven by the service plan. Track must exist for the proposed service including local and express trains. Signal systems must be capable of handling very frequent service with trains close to each. Power supply for electrification must be sufficiently robust to support service at the planned level.

A major issue for GO Transit is the constraint on service through shared parts of the corridor, notably at Union Station. Metrolinx has already been clear that the design of the Ontario Line including shared stations with GO at Exhibition and East Harbour is intended to offload some traffic otherwise bound for Union.

This does not sound like a system with a lot of spare capacity sitting around needing to attract riders. The distinction, of course, is between peak and off-peak services. However, the Board advocates better use of GO to shorten peak trips and to relieve subway crowding, and that only applies to the peak period when GO capacity could be at a premium.

What Does the Service Network Look Like?

From a regional planning perspective, one might jump for joy at the concept of a subway-like service reaching out to corners of the city and beyond, but much travel in the GTA is not bound for the core or even lies along the rail corridors. What will be done for those travellers? What service and fare structure can they expect on their suburban buses routes?

If we are to have a new fare system, that alone will not attract riders to transit in sufficient numbers to dent the mode share now held, and held strongly, by auto travel.

Far too many plans show maps with no indication of the reach or quality of service they represent. A badly-needed planning tool are maps showing the accessibility of various locations at different times of the day. Does the transit network actually serve the pattern of trips it might carry?

How far does a one hour ride get you from, say, Malvern or northern Etobicoke? What is the corresponding reach from a concentration of jobs? How much of the region is an hour or less away from work? What is the effect of including access and transfer times? How do these maps change between peak periods, midday, evenings and weekends?

Most importantly, does the network serve the actual patterns of demand, particularly demand that is not bound for the core area?

A cheaper ride is little benefit if there is no bus to connect with. This may be the topic of a future Board of Trade report, but the fare proposal is silent on this issue.

A hypothetical rider, “Celestino”, in the Board’s report would welcome transit systems that “coordinate their fares”. He makes a longer medical trip than necessary to avoid the extra fare crossing into Toronto. For him, time is less valuable than money.

The real issue in a fare scheme is not just lost revenue for transit. Is travel hampered by service levels even without a fare boundary? What is the cost in restricted mobility that poor service brings?

Another hypothetical rider, “Tanisha”, talks of transit fares in terms of the hours needed just to earn a round trip from home to work. At $15, that is at least an hour’s work, not including the effect of taxes and other deductions. This is an important way to look at transit especially if one claims to make it relevant to people of low incomes.

An essential change for Metrolinx is to break the assumption that the typical riders are fairly affluent suburbanites who calmly sit on trains, beavering away on their laptops. For them, the cost of travel by GO is offset by reduced or eliminated car expenses including parking downtown, and the hassle of driving. This comparison does not exist for someone who could not afford to drive in the first place.

Any truly regional network will serve a wide variety of riders, but will not be able to charge commuter rail prices. Whether fares are changed across the board, or by targeted subsidies such as the Toronto Fair Pass, is a policy decision. Do we treat subsidies as giveaways to rich and poor alike? Do we treat mobility, the availability of low-cost transport, as a social and economic benefit both to riders and to those businesses that depend on people getting to their front door?

Before the name SmartTrack was invented, a similar scheme aimed to make outlying job centres in Markham and near the airport more accessible to would-be employees. That’s an economic development lens that saw better transit as a cost of doing business, not as a subsidy to employees or employers.

Any debate on fare structures must occur in this larger context.

The Bus and Subway Networks Are One

The Board recognizes the importance of the bus network in getting people to and from the subway system. They are one network and this should not be arbitrarily split by charging a premium for subway travel.

However, station nodes bring their own problems in that the route structure can be distorted to force feed the subway to the detriment of local non-subway trips.

Expanding the concept to GO stations brings several problems:

  • Is the local transit agency prepared to run enough service to the station to handle current and future demand?
  • Do GO stations, many of which are in industrial neighbourhoods because of the historic purpose of the rail lines, become distorting factors in a local route network?
  • Can GO remain attractive if parking is not its primary mode of access?

“Last mile” access to GO is a big challenge across the network, and it requires far more than a revised fare structure to address demand particularly for off peak and counterpeak travel.

Quarter-hourly service all day on GO rail is of little value if little or no service connects with it. Within Toronto, it is assumed that TTC service already exists at most GO stations, but the situation is much different in the 905.

Subway stations were built, for the most part, in areas of demand, and many are destinations in their own right. The bus and streetcar networks are physically integrated. GO stations, by contrast, did not have to worry about this because they could be in the middle of a field. All that was needed was lots of parking for morning commuters.

GO has two parking spaces for every three commuters on its network. As of winter 2019, there were 219,000 daily boardings on the rail network, or about 110,000 passengers allowing for round trips. There were about 75,000 parking spaces. The province is happy to pay for parking lots and structures and give away the service they represent free of charge.

This is as much a part of the fare and service argument as any debate about zones. Why should riders who arrive at GO stations by bus have to absorb the hidden value of parking in their fare? Should a parking fee be charged separately so that only those who actually use this service pay for it? This would allow the base GO fare to be reduced. The Board is silent on that subject, although they do note that parking occupies valuable land that could be developed.

The Airport: A Major Node at Many Boundaries

The Board rightly cites the airport district as one where fragmentation of services hampers transit access.

Bus routes themselves are also not always designed to connect seamlessly between agencies. Some routes end abruptly at a boundary, forcing riders to get off and switch to another route to continue along the same street. In several places, 905 routes run to a terminal in Toronto, but they cannot pick up or drop off passengers along the way when they pass through the city. […] The airport employment area, which spreads across three municipalities and transit agency service areas, plus GO Transit, is an important case study of fragmented routes and schedules. This is the second-largest employment area in the region, and its transit accessibility is significantly hampered by transit system divisions. [p. 13]

The airport is a particularly complex area to serve for several reasons. One major demand is focused on the terminals – the passengers – while another is on the many businesses that ring the airport but are certainly not “next door” the way downtown is a compact district of business, government and entertainment. Travel times differ for each group, and origins are scattered around the region.

This was a problem for the Union-Pearson Express which approaches the airport from the southeast where there is a concentration of air passengers in central Toronto. However, most who fly from Pearson Airport do not come from downtown. There is a similar problem for workers who originate all around the airport, but who in addition do not necessarily work right where the UPX drops them off. They have a “last mile” problem.

This is a service and network design issue far more than a question of fares. A special employee fare was implemented on UPX because its original pricetag, aimed at affluent tourists, was impractical for day-to-day travel. However, that was simply a bandaid for a specific problem and did not address the wider question of how to get to and from the airport and the businesses around it.

SmartTrack was intended to serve the district south of the airport, but suffered from an unworkable design with a mainline railway on Eglinton Avenue. (Opponents of a surface LRT don’t know what they missed!) This morphed into a western extension of Line 5 Crosstown, but even that will have a last mile problem because the area is designed for access by car.

Untangling this knot does not require a formal, unified agency, but better coordination and cooperation between the relevant local agencies, together with a willingness to make this work. This will not be simple because the airport district is the size of downtown Toronto, but with a much more dispersed potential market. There is no “King & Bay” there, nor is there a pre-existing network of subways and rail lines serving the airport from many directions.

Some form of fare integration will make network planning simpler, but the airport’s problems are much more complex than fares alone can solve.

Modernizing Presto

The Board notes that Presto has its problems in part due to the complex pre-existing fare systems in the region. They neglect to mention that a major factor is the age of Presto’s technology and a desire by Metrolinx to get as much out of the original investment as possible. As things stand, any fare change has to be programmed into every reader in the GTA, and there are limits of the reader architecture on the complexity of programs they can accept. This will change in the next iteration of Presto, expected in the mid 2020s, when all of the logic involved with fare calculations will move to a central system and the readers’ sole job will be to keep track of cards presented. This will shift any system and tariff updates to a central location where changes, especially ad hoc promotions or discounts, will be easy to implement. This will do a lot for Presto’s flexibility and it does not require a harmonized fare system.

It is important that cause and effect be clearly understood here. Fare integration may be a nice idea, but improving Presto does not demand that integration occur as a pre-requisite.

The Presto system has faced considerable implementation challenges, in no small part because of the extraordinary complexity of the Toronto Region’s fragmented fare system. Significant progress has recently been made but even more can be done to harmonize everything from passes to discounted fares. When policies differ between agencies, each variation has to be programmed into Presto at considerable cost. A single, fully harmonized system would greatly improve the efficiency of Presto while making transit much simpler for riders. [p. 13]

What Would a Transit Federation Look Like?

The term “federation” is important because it implies the continued independent existence of its parts. There is no point in having a powerless talking shop whose sole purpose is to rubber stamp provincial decisions, nor should local decisions about transit quality be left to a regional agency beyond, possibly, setting minimum standards.

There will always be two “elephants in the room”.

One is the TTC which is not just a large, well established agency but an important part of the regional transportation network. The same cannot be said of others who might sit at the same table. Even GO Transit carries only a fraction of TTC’s ridership and, as its lacklustre covid-era riding stats show, is overwhelmingly dependent on core-bound commuting traffic.

The other is the province who can throw their weight around by the simple fact that they control both the funding and the legislative framework for any transit systems.

Metrolinx was created as a regional agency whose work focused on planning for the region’s future. GO Transit was not folded in until later, and the marriage between operations, planning and capital project management has not been an easy one.

It is a common story that the “Metrolinx I” with a board containing many regional politicians could not get anything done because of bickering. The real problem for Queen’s Park was that as a political board, it could not be controlled. Municipal priorities do not necessarily include making the Premier and Minister of Transportation look good. “Metrolinx II” is a board that conducts almost all significant business in private, and its members can be counted on to sing from the provincial song book.

Any federation must be a real working body, a transparent co-operative effort to improve transit through an open relationship with riders. We do not need another secretive, arrogant, we-know-what’s-best-for-you organization like Metrolinx.

The Board of Trade proposes that scheduling and public information be co-ordinated through this federation, although I cannot help noting that customer amenities such as a consolidated information portal could easily already be in the Metrolinx mandate. Scheduling – ensuring that connecting services actually connect – depend on funding issues and budget decisions.

For example, there might be a rule requiring “clock face” headways on all services that run less often than every 10 minutes. What happens, in practice, is that a 15 minute headway is stretched to 17, a 20 to a 23, or a 30 to 33, simply to give drivers a bit more time without adding another bus. Budgets are preserved, but riders have no chance of predicting when a bus will show up (even if it is on time) and scheduled connections are impossible.

Many potential changes to the tariff will require provincial agreement and funding. For example, restoring the GO-TTC co-fare or extending the short-trip GO fare to all stations within Toronto can be done today by the Metrolinx Board. The governance problem lies in the foot-dragging and doctrinaire politics at Queen’s Park, not in the absence of a transit federation.

If any regional transit network is to be credible, service will have to be at an attractive level to be an alternative to driving. That means more buses and all of their associated costs. If collectively the municipal and provincial governments do not see this as a priority, a federation will be little more than a chance for a quarterly coffee-and-a-bun among its members with no real progress.

12 thoughts on “The Siren Song of Regional Fare Integration

  1. I don’t see any incompatibility between zone-based fares and time-based fares. All the time-based fares in the GTA are currently limited to certain zones, making them compatible with a zone-based system. GO might have issues with a zone-based system but that’s more due to the fact that their fares are inherently more expensive. I think the only people who would suffer from a zone-based system are those people who currently cross 3 zones from the 905 to 416 but then happen to work right where the bus drops them off instead of taking a TTC bus/subway further onward.

    Steve: They are not incompatible, but if we’re talking about getting rid of boundaries, the entire thesis of the Board of Trade’s proposal, you don’t use zones. GO will have to remain separate to some extent, but how separate depends a lot on the amount of subsidy available for GO+local travel. A related problem is whether we should consider GO buses as “local” services particularly considering the Metrolinx proposal for a grid of frequent bus services in the 905.

    Liked by 1 person

  2. “It is the quantity and quality of service that determine the attractiveness of transit”

    I 100% agree. If you ask people why they don’t take transit, fares are rarely the answer. The vast majority will cite transit vs. car journey time – either directly (lower vehicle speed) or indirectly (long wait / transfer times caused by poor headways).

    If anyone’s skeptical, ask someone who doesn’t take transit whether they would take transit if it was free.

    Fares aside, municipal boundaries are only problematic where there are service boundaries. MiWay and YRT both bring a bunch of services to Toronto subway stations, so those boundaries don’t create service issues. On the other hand, services on various north-south corridors don’t cross Steeles because YRT doesn’t want to pay extra for TTC to operate across the boundary. Some agencies will operate deep into other municipalities on their own dime, on the grounds services should go where their residents need to go (Brampton is the strongest example). Other agencies stick within their boundaries.

    The Board of Trade report keeps talking about “Regional Fare Integration”. Yet they surely know that transfers between 905 agencies are free and transfers between GO and 905 agencies offer a 75% discount on the 905 agency fare. We all know it’s only a problem for transfers involving the TTC. (Aside: if TTC-905 transfers became free, then the agency most affected would be… York Region Transit. They would lose about a third of their fare revenue.) The Toronto Board of Trade should be brave enough to talk about “TTC fare integration”, not “regional fare integration”

    I agree that Presto’s limitations are not the primary issue. Any fare system too complicated to Presto is too complicated to be used in the the first place.

    Quick note on GO Train signaling systems: where signals exists, they allow trains to operate with a gap of less than three minutes. Practical scheduling concerns and the limitations of Union station mean that would never happen, of course.

    Liked by 1 person

  3. How about a flat fare-for-time system. No boundaries, no zones, just time. You pay your fare and get a two or three hour “pass”. This would let you do several local hop-off, hop-on errands, or go a decent distance on a long distance trip. If you need more time, to get where you are going (say Sutton to Hamilton Airport) you pay again.

    Is there a way to make a time/distance fare system work?

    Would the average 2 hour fare be enough, or would it shortchange current TTC ridership?

    And if this system was used, then schedule adhesion would be critical to making the most of your time. TTC would need a new metric for spacing and on-time performance.

    Steve: The essential problem is that TRBOT is trying to make GO have both a local and a regional function, and the proposal is very Toronto Centric with an emphasis on faster commutes and subway relief. There is no discussion of how GO might feed other nodes such as Hamilton or the Guelph corridor, not to mention the role of a regional bus network.

    There is going to have to be some kind of surcharge for GO, but it could kick in for longer trips while leaving short trips as part of your local fare.

    Schedule adhesion? Don’t count on it. One thing Presto could do, once it is restructured with a new centralized back end, is to retroactively extend the “two hour” window for conditions such as snow storms or major service disruptions.

    Liked by 2 people

  4. Steve said: Signal systems must be capable of handling very frequent service with trains close to each.

    tompw said: where signals exists, they allow trains to operate with a gap of less than three minutes.

    Through following Steve’s blog, over the years, I have acquired an impression of signaling systems but I personally don’t know anything and am always willing to be corrected. This is my impression.

    The GO train network is a railroad company. In off hours, CN and CP run their trains on GO tracks and vice versa. In Canada, the railroad signal system is based on fixed 15 minute blocks, one train per block. This is because, trains cannot just stop at red lights, like cars can. In one round of SmartTrack, they were going to schedule two trains on one line over the same time. I asked the question, does this mean the two trains would end up in the same block? Never got an answer.

    The fixed 15 minute block system is very old and puts Canada in the dark ages, even the US has progressed, never mind Europe or Japan.

    Automatic Track Control (ATC) allows for shorter separations between trains through making the blocks dynamic. ATC requires a network control system, ATC systems on each train and along the tracks.
    GO trains are heavy railroad locomotive and coach cars. They cannot accelerate or stop as fast as EMU’s (subway like trains). There isn’t a big difference between electric to diesel locomotive performance.
    It takes a huge investment to get 3 minute separation.

    Steve: There are a few issues here. First, the mainline railways would have to outfit their locomotives with Positive Train Control to participate in any GO Transit updated signal system. Second, I don’t believe that tompw is correct in saying that GO signals can handle a three minute headway today. That may be a goal, but I’m not holding my breath to see that investment (leaving aside the much larger fleet they would need just to populate the network at that level of service). There is also the matter of combined service frequency through area east and west of Union Station. Some of that will be addressed by the proposed reconfiguration of the station and the establishment of through services to avoid the time needed to turn trains around. But a lot of this is years away, and GO/Metrolinx can be evasive when pressed for specifics, as you know.


  5. Bill R said: “In Canada, the railroad signal system is based on fixed 15 minute blocks, one train per block.”

    The block distances on GO’s lines aren’t uniform, ranging from half a mile to over three miles. In the latter case, which can be found between Kitchener and Georgetown for example, the closest possible headway would be about ten minutes, and in practice, GO hasn’t scheduled anything closer than 15 minutes there.

    Other parts of the network are different. Lakeshore West consistently has blocks about 1.5 miles long, and before the lockdown, some parts regularly saw a train every 5-6 minutes. Even shorter blocks can be found between Union and Pickering, and Union and Malton. tompw’s 3-minute headways could be seen between some Lakeshore and Stouffville trains until the lockdown, and between some Kitchener and UP trains as late as August.


  6. Bill R said: “In Canada, the railroad signal system is based on fixed 15 minute blocks, one train per block.”

    Somewhere in this blog there is an explanation that blocks are 15 minutes. The unit of time is used because a train can go farther downhill than uphill in 15 minutes, so the downhill block is longer than the uphill block. At the head of each block is the signal, whether to slow down, speed up or stop in that block. The whole purpose of blocks is to stop a train in time. The signals control the speed of the train. A shorter block means the train cannot be travelling too fast, if stop signal were to come up. The speed allowed is partly determined by the weight of the train. GO trains are much heavier than EMU’s so will not be able to travel as quickly in a block because stopping distance is a factor of speed and weight. All trains can do 60 mph, the difference is how long it takes them to stop and how quickly it takes to reach 60 mph.

    There are big gaps in my understanding. You will find in this blog, mention of the Stouffville track signal system being upgraded, by Perkins but halted because of budget cuts. As practically no freight runs on it, it could be engineer knowledge of the track for speed and stops is used if the signal system was never completed?

    Steve: Someone may have talked about 15 minute blocks, but not me. The spacing possible depends on the installed signalling which could support longer or shorter headways as needed. Related questions include station spacing, the use of freight spurs, and the length of freights that might use the line, not to mention single or double tracking.


  7. The Toronto Region Board of Trade has a very long history of making excellent proposals but the problem is that politicians like John Tory don’t act on these excellent proposals. As one of many examples, consider Superlinx – an excellent Toronto Region Board of Trade idea that Tory helped derail with bogus lawsuits at the expense of us taxpayers.

    Steve: The Board of Trade has now changed its position on amalgamation of the region’s transit systems and favours a federation rather than a superagency. As for “bogus lawsuits”, how can anyone launch a lawsuit against something that was only a discussion paper. You are off base on that.


  8. Things that bugged me about the report.

    I disliked two comparisons.

    The GTHA is much more car centric than European cities like Paris and London.

    GO commuting is successful because of free parking (two parking spaces for every three commuters). Visit suburbs in York Region, Vaughan, Mississauga and you will see three cars per household. City planners in those areas never planned for walkways from doorstep to a bus stop. Instead, developers were able to maximize the land for development. The “last mile” from doorstep to bus stop is often a very long walk because of poor street layout. Arteries were not designed for bus transit. The study made no suggestions to make mass transit attractive to current car users who are still the biggest part of the commuting congestion problem.

    The authors don’t recognize that the Canadian railroad signaling system is far too rudimentary to meet rapid transit service levels of European systems or that the GO locomotives and passenger coaches are not rapid transit mode. What would be needed would be an Automatic Control Systems (ATC) signaling system, EMU trains with many doors instead of two and raised platforms. This is incompatible with freight train requirements using the tracks overnight. As Steve mentioned, the constraints of Union Station are very challenging and it is a bottleneck that must be taken into account and the report doesn’t.

    Public transit is about meeting demand with service at reasonable cost.

    This report doesn’t not work from this premise. Large demand should dictate infrastructure systems like subways, EMU’s in railway corridors or LRTs. The bus network backfill will primarily feed the infrastructure systems but as Steve points out this sometimes fails the actual service needed which may not be a route to the subway. I personally think the 905 regions have such poor suburban street layouts that the last mile to doorstep is destroys demand for public transit.

    I do feel that fare structure must be discussed simultaneously with the service design. A different implementation of SmartTrack providing 5 minute service and single fare would attract 308,000 daily trips but Tory’s 15 minute service with double fare yields 36,000 daily trips per the UofT study.

    The TTC buses running into York Region at a second fare is another example. Projects that can’t pay their way are a real dilemma and Toronto is expert at it – Sheppard subway, Vaughan extension and I predict SmartTrack and the Scarborough Subway Extension fall into this predicament. If the TTC is losing money on running buses into York region, then I would suggest the fare in York should be raised which would reduce ridership even more.

    Steve: The TTC loses money running the subway into York Region. Bus routes north of Steeles are operated under contract to York Region at no net cost to Toronto.

    As I have outlined above GO train service has limitations and cannot be considered an urban rapid transit service. There is no need for mode neutrality on GO train service, subject to whenever Canada upgrades its railroad signaling standards.

    I believe finding the “best” way to charge is a really difficult problem. The report should have done what Steve did and discuss the complications. The authors of the report are trying to push their solution where they would be of better service to define the problems and the tradeoffs of options (which they don’t seem capable of).

    As to the suggestion of administrative structure, I don’t see their proposal as convincing. What the true problem is, is a shortfall in problem solving ability to deal with the public transit requirements of the region. Steve has listed many transit challenges. He didn’t get paid to think about them, he invested his own time. Does anyone know of paid staff who work on these problems and are capable of proposing solutions? Do we have reports understanding demand flows? Do we have reports studying where all the people on the 401 are trying to get to and what kind of rapid transit route would attract 25% of them? For me, I think the 905 suburbs should think about buying strategic houses which they would tear down and use as walkways to get neighbours to bus stops quicker, to solve the last mile issue. I also would add that Toronto has a Yonge subway crisis.

    Here we can say that political interference has truly set back public transit. Mayor Rob Ford’s subway mantra Mayor Tory’s SmartTrack and Scarborough Subway Extension Both parties of the Provincial government’s failure to propose valid transit solutions.

    The report should show how current governments are not up to the task of defining the transit problems and finding adequate solutions. The biggest mistake we have been making is always proposing a solution first. Transit experts are the blind only finding one part of an elephant. I have met a lot of people of goodwill who are keen to contribute to solving our problems but we lack competent leadership.

    Steve: One generic problem with the Toronto Region Board of Trade is that they tend to be a cheering section for government, and it is unlikely we would ever see a report from them that was highly critical (except, probably, if the NDP were in power). They have often provided a platform for Metrolinx and Ministry announcements.


  9. Me: Tory helped derail with bogus lawsuits at the expense of us taxpayers.

    Steve: As for “bogus lawsuits”, how can anyone launch a lawsuit against something that was only a discussion paper. You are off base on that.

    Don’t forget that when the democratically elected provincial legislature passed a bill to reduce Toronto’s council size, how much of the taxpayers’ money did John Tory (a lawyer) waste on filing court challenge after court challenge after court challenge when it was clear all along that John Tory had no valid case? One rogue judge Edward Belobaba agreed with John Tory only to be reprimanded by a higher court calling Belobaba’s ruling in favour of John Tory “a dubious ruling that invalidates legislation duly passed by the democratically elected legislature.” John Tory may or may not be a good mayor but as this example shows, John Tory is not a very good lawyer and Edward Belobaba not a very good judge.

    Steve: You made the comment about lawsuits in the context of derailing “Superlinx” which is completely separate from the issue of City Council. That body is also democratically elected and it democratically decided it wanted a new ward system with more councillors. Ford overruled that with his legislation which is as much about his desire to screw Toronto as anything. He did so while the election was already underway.

    Challenge after challenge? There was the original Superior Court challenge which Toronto won, followed by a stay and a split 3-2 ruling against the City in the Court of Appeal. A split decision hardly marks the original ruling as the work of a “rogue” judge if two of five judges at the appellate level agreed with him. The Supreme Court gave the City leave to appeal, and the hearing is scheduled for March 16, 2021.

    As for Superlinx, the Board of Trade has changed their position and it no longer supports the creation of a mega-agency.


  10. On signaling and three-minute headways… 🙂

    When I say “signaling system”, I mean the technology used, not the exact way it is implemented. Railroad signals are based on the idea of splitting up the rail corridor into “blocks”, with each block only having one train in it a a time. At the end of the block, a lineside signal tells whether you can proceed into the next block or not. The minimum headway is therefore (nominally) the time for the train to travel through each block.

    Trains (especially freight trains) take a long time to stop. Where the blocks are shorter than this distance, then this means where train A enters a block, and train B is in the block ahead, then train A must limit its speed. Consequently, the signal train A sees ‘tells’ it go slower. The same then applies to the train behind train A (which must go slow-ish) and so on.
    The information that can be conveyed by Canadian railroad signals effectively limits how short the blocks can be.

    If you end up at close to the minimum practical block size, you end up with headways of 3-5 minutes. This is the case on the Lakeshore line (I have observed gaps under 5 minutes between consecutive trains), but I know blocks elsewhere are larger.

    So, when I said “you can do three minute headways with the current signaling system”, a better statement might have been “you can do three minute headways using lineside signals and a block system, which is what GO uses”.

    This is all on a nice linear corridor. When you have a load of lines coming together to serve a multi-platform station (oh look, Union), life gets *much* more complicated. But that’s another story…


  11. Question for tompw :

    Thanks for the explanations, every bit helps.

    The GO passenger trains are more nimble than a 100 car freight train that the railroad signals were designed for.

    I’m guessing that the GO train can easily stop in a railroad block so can go at a faster speed in a block.

    Would you please explain how lineside signals let the trains run so close together. With a headway of 3 minutes, the trailing engineer has to know a lot about what the train ahead is going to do, especially if its going to slow down or stop.

    I was under the impression that GO train service was based on 15 minute service because the blocks were 15 minutes. For SmartTrack, Metrolinx had offered what sounded like 7 1/2 minute service (actually was more complicated something like scheduling 5 trains in one hour). Sounds like you might have an idea how it could be done. Could you suggest a way?

    [Bill R subsequently added another comment]

    To tompw

    My error, my laptop skipped a screen, and I missed your whole explanation on lineside signals. Your explanation is clear, you don’t have to repeat and I thank-you for taking the time.

    One thing I don’t understand. We can get smaller headways which means we can have more trains but what about speed? What is the maximum speed trains can go at 3 minute headways? The Stouffville trains are very long and I can’t imagine 3 minute headways between them at 60mph.

    Steve: Other factors here are train length (a freight can be much longer than a GO train) and braking distance. The physical distance between trains is only part of the equation. An issue on the Stouffville corridor is that Metrolinx has been evasive about service patterns including extra stops. If there is a mix of locals and expresses, but no passing tracks, and if combined service is at the limit of the signal system, you cannot have an “express”. When they talk about mixing service on one track, the scheme would be that a local and express would start out as close together as possible with the local in the rear. The express would pull progressively further ahead until it caught up to the next local train. “Caught up to” is a function of the spacing the signals permit, and if the blocks are too long, there will be little headroom for this type of operation.

    Another factor is EMU vs diesel locomotive operation. On some corridors it is likely that the express will be diesel hauled because it will originate on the outer end of a line beyond electric territory. The more sprightly EMU would pull out behind the express but gradually drop back as it made local stop. With EMUs, the time lost for deceleration, stopping, and acceleration is lower than the trains we have today, and so the penalty for stops is less severe.To put it another way, the local train can stop more times before the following express catches up to it.


  12. Being a Toronto resident who frequently visits Metro Vancouver, I can tell that Vancouver has done a reasonable job for regional transit integration. Yes, Metro Vancouver (MVRD) is somewhat smaller than the GTA, but the entire MVRD is covered by the same transit system (TransLink) divided into three zones, mostly based on the boundaries of major cities in the area. Of course there are still imperfections, such as when a two-zone fare is charged on adjacent stations on either side of a fare zone boundary, but the fact that you don’t need to pay double fares when crossing zone boundaries as you would in the GTA is the biggest factor at why Metro Vancouver’s transit ridership is a lot higher than the GTA. With the exception of the West Coast Express, there are no expensive GO Transit-style distance fares. Also, after 7:00 pm weekdays and all day weekends, the whole system is one zone. Buses are always one zone, the three zone system applies only to the SkyTrain and another separate zone system applies for the West Coast Express. I encourage you to take a look at the system/zone map and fares on the TransLink website.

    Fun fact: on weekends when only one zone is active, you can travel from Horseshoe Bay to Langley on a Compass $2.40 one-zone fare, that’s about the same distance as from Toronto Union Station to Newmarket.

    Just because the GTA is bigger doesn’t mean that Metro Vancouver-style transit integration can’t be done.

    Steve: There is nothing technical preventing the creation of a zone structure like Vancouver’s. What we are missing is the will to finance it. Queen’s Park just does not want to pay for local transit, and none of the local systems is willing to give up revenue in the name of “integration”.


Comments are closed.