Almost every ridership study and projection, almost every major infrastructure and scheme looks at peak period, core focused commuting patterns. Huge efforts go into multi-decade plans to deal with growth, but much of these plans obsess on getting people to and from work downtown.
This leaves millions of trips unaccounted for outside of downtown and outside of traditional commuting hours. In the pandemic era, with that “normal” commuting load stripped away, we now see the importance of the local bus service – much of which does not serve downtown – where ridership remains strong. The subway may be quiet and GO Transit almost deserted, but the bus routes carry well, in some cases too well for comfort.
A major group of transit travellers is students, and in particular those at post-secondary institutions who can face much longer trips over a variety of hours than their younger colleagues in secondary and elementary schools. Many are now learning from home, but when they return to in-person classes, their effect on the transit system should be remembered.
In fall 2019, ten universities and colleges, Metrolinx, the City of Toronto and other organizations launched research into post-secondary student travel patterns. Preliminary results from this study have now been published and these are available on the StudentMoveTO website.
This survey, necessarily, reflects “before times”. It tells a lot about student travel patterns and shows how they differ from the classic transit model. This is interesting not just for students, but as an example of a group and their travel demands that are poorly understood and poorly served by core-oriented transit thinking.
Moreover, as a group who, relative to many others, are less likely to use or have access to cars, they show the problems of a so-called regional transit system that provides much less service at the edges than in the network’s core, Toronto.
The survey cast a wide net across the member institutions.
Collectively these institutions represent over 300,000 students and their travel. With a response rate averaging 6 per cent, there were over 18,000 responses with varying participation rates from each college and university.
An important caveat is that those who chose to participate do not necessarily represent the population as a whole, but as with any survey, one works with the data available. Cross-checks on these samples would be interesting (such as using the postal codes of all students to verify their geographic distribution), but that is work for another day.
Of the students who participated, about 40 per cent filled out a travel diary, and this gives us information about the types of trips they made.
It should be no surprise that trips related to education accounted for only 36 percent of total journeys. People do have other things to do with their lives, especially in the much more social pre-pandemic era.
A considerable chunk of the trip pie, 18 percent, consists of work-related trips as many students also have jobs. This might be a secondary demand pattern overall, but it is part of many students’ lives and commuting burden.
The overall location of students and the institutions they attend is shown in the map below.
Some people commute a very long distance to school in much the same way as many commute to work. Post-secondary institutions, however, are not clustered in a few nodes such as King & Bay Streets, but are scattered around the region.
The map above does not show campuses of other institutions that did not participate in the survey. Those located outside of the south-central part of the GTHA have commuting issues of their own.
The trip lengths and times vary immensely over the region, and those who use regional transit face the longest journeys averaging an hour and a half, one way. Note that this includes all legs of a journey including access to a trunk route such as GO or the subway, a connection that can be difficult in some locations and times of day.
Forty percent of students have travel times under half an hour reflecting a preference for living close to their campus if possible. The data are not subdivided by institution and campus and so we cannot see how much these numbers vary between, for example, Ryerson University at Dundas Station and Centennial College in eastern Scarborough.
As noted above, less than half of all trips were education-related. The mode share for the non-school trips is quite different from the school trips. Transit gets 60 per cent of education trips, but only 39 per cent of travel overall (implying an even lower value for the non-school trips).
Travel is skewed to early days in the week and falls considerably, especially for commutes to and from campus between Mondays and Fridays. What is unclear is whether this reflects institutional scheduling practices, or attempts by students to pick courses that cluster their time at campus and reduce their commuting load, or a combination of these.
Student travel varies over the course of the day. While it is true that many trips occur within the traditional peak period, a large number lie outside this period, a time when transit may not offer a convenient alternative making long trips even more tedious.
The distribution of home-to-campus trip lengths by origin is shown in the map below. Note that these are not necessarily times to a downtown Toronto campus, but averages based on where students in each part of the region are going. For example, there is a clear pattern around eastern Toronto and Durham, as well as in Hamilton reflecting trips to campuses in these areas.
Updated October 15, 2020: I have been advised by one of the authors that the figure “39 minutes” in the chart’s legend below should read “49 minutes”.
Overall transportation costs for students also vary by location and is generally higher the more remote they are from campuses. How much of this is a transit cost, and how much the cost (if any) associated with other modes is not mentioned.
There are many calls for transit fare integration as well as discounts for a variety of deserving groups. What is missing here is the “n” associated with each location on the map. Some areas might have very expensive travel costs, but what proportion of the student population do they represent? Are there offsetting considerations such as the ability to live at home that see some students so far from their campus? To what extent does this map reflect the problems of sprawl that affects all types of travel and its cost?
An important issue for students, as it is for workers and any who travel regularly, is that time and money spent on commuting is not available for other pursuits. In the case of students, participation in on-campus activities might be difficult for those who live far away and only are on site a few days a week.
This is a preliminary report, and it set me thinking about issues that it does not address in detail, in particular a more granular view of the data.
StudentMoveTO makes an anonymized version of the trip diaries available through its site, although this is clearly not the same data as that used to generate the charts above. Specifically:
- There are more rows in the supplied data than there are trips in some of the charts.
- Only the start time of journeys is included, not the end time, and so there is no simple way to calculate the duration of each trip segment.
- Locations for end points of trip segments are given only as planning districts (within Toronto) and municipal names (the wider GTHA). By comparison with the scatter on the map of home locations above, this limits the ability to correlate home location with available transit service. (These data, the census tracts and subdivisions, might have been omitted for reasons of privacy where “n” is small.)
Undeterred, I set about reviewing the home-to-campus trips and breaking them down on a more finely-grained version than in the report. (Click to expand this table.)
Although the geographic locations of the trip destinations are only given as planning districts, it is easy to map these to specific campuses of the institutions. In Scarborough, Centennial College has multiple campuses within single planning districts, and so I have had to consolidate these here. For each institution, there are some trips that do not end in a major campus location, and these are included as “Other”.
In total there are 3,624 trips logged as home-to-education out of the 19k trips reported by students. Of these, a handful of cases were multiple trips by the same student and the “Unique” column distinguishes these counts. The difference is small, but I include it because this is part of the supplied data structure.
Ten modes of transport are included, and the counts there reflect all modes used in a trip. Therefore if someone took a GO Bus to the subway and thence to their campus, that would count in both columns. The same information is expressed as percentages of total trips for each campus on the right side of the table.
Overall driving shows up in just under 20 per cent of the trips, but the proportion varies a lot from campus to campus. Also, the overall numbers are affected from the relatively high number of responses from Ryerson and UofT.
It is no surprise that for the centrally located campuses, the share of trips involving driving is under 10 per cent, but it is much higher elsewhere. There is relatively little ride sharing with only 8 per cent of trips overall, although again the proportion is higher for campuses in more core-oriented locations.
Walking has a high proportion both in central Toronto and in Hamilton with an overall value of 17 per cent. Cycling comes in much lower at 2.9 per cent. This is a case where the ability to calculate a trip length versus the mode choice would be informative.
Buses have a strong showing at 41.9 per cent overall, but much higher values where a campus is not on a subway line and buses provide the primary transit access. Streetcar numbers are very low because most campuses are not on a streetcar line, or their students do not come from the catchment area of the local network.
The subway gets about a third of all trips, but the numbers skew depending on whether one is looking at a campus with a subway connection, or one that is more dependent on buses and cars for access.
GO buses and trains have a relatively low share probably because relatively few students live far enough away from their campus to use this network.
Also of interest is a cross-tabulation of the 19,229 trip diary records by trip purposes. (Click to expand.)
Almost half of the trips, 8,235, originate at home, but of these only 3,624 are destined directly for an educational location. Overall, considerably more trips arrive at an “education” location (4,626) than leave (3,739). Similarly, there are a lot more trips leaving “home” than returning. This suggests that some trips are not reported, or they are reported under different location codes.
This is a good start on the problem of post-secondary student transportation needs, but more data are needed to reflect the constraints in transit network design, service patterns and fares on students’ travel experience. This type of analysis is also needed for other groups such as essential workers and those in the manufacturing sector, not to mention home-based travel within the suburbs for non-work or education related trips.
Too little is understood of these demands, and too much emphasis is placed on simply building a subway as the solution to a very complex set of needs.