Correction: The original version of this article stated that the “Income and Transit Use” paper was the work of Steere Davies Gleave (SDG). This was an assumption on my part – that it was a continuation of their previous work. I have been advised by SDG that this paper is not their work, but that of Metrolinx staff. All references to SDG in connection with this paper have been modified appropriately. My apologies to SDG for mis-attributing work to them.
Updated: This article was updated on February 19 at 6:45 pm to include comments on the things Metrolinx should also be studying, but omitted in their review of incomes and transit use. Scroll down to the end to see the update.
In two previous articles, I have examined the February 2017 update to the Metrolinx Board by staff on Regional Fare Integration, and a June 2016 background study by Steere Davies Gleave [SDG] on fare integration concepts.
This article reviews another June 2016 study by Metrolinx Staff on income equity: GTHA Fare Integration: Income and Transit Use
The context for this study, nominally, is to determine whether a new fare scheme will affect low-income households.
In reviewing potential modifications to the transit fare system across the Greater Toronto and Hamilton Area (GTHA), the social equity implications of transit fare policy must be considered. Lower-income households rely more on transit for their mobility, are more sensitive to the fare they pay for their transit trips than higher-income households, and, as a result, fare policy choices may impact them more. [p. 1]
However, the selective examination of effects by Metrolinx staff focuses on the benefits of a lower fare for “short” trips while playing down the effect on “long” ones.
For the purpose of the analysis, Metrolinx looked at a fine-grained version of census data, “dissemination areas”, where each element contains less than 1,000 people.
[these …] typically exhibit greater homogeneity in the household incomes of their residents than larger geographic units. [p. 2]
Each of these areas would lie within one geographic section of travel surveys (the Transportation Tomorrow Survey which, at the time of writing would have been based on 2011 data), and the transit usage for each dissemination area was taken from the corresponding TTS area’s results. Census data on income was used to assign each census area to one of ten income ranges, and through this to map transportation patterns to incomes.
Note that there was no adjustment to reflect the availability of transit in any of the census areas, and the results merge data across the region. The income groupings are based on dividing a population of 6.5 million into roughly equal groups of 650,000. “Equivalent income” is a value derived from a combination of household income and household size.
The actual distribution of income shows a familiar pattern with higher incomes along the Yonge Street corridor and in some parts of the 905, notably those well-served by GO Transit.
It is also no surprise that the higher income groups travel more by automobile for the simple reason that they can afford to do so.
Moreover, given their geographic distribution, lower income riders tend more so to use surface transit modes rather than the subway or GO. This reflects the type of service available both where they live and where they are travelling (work, school, etc.).
The distance travelled by mode shows the effect of location, available transit service, speed of mode and type of trip. Except for the highest income group, there is a gradual increase in average length of trip beginning at 10km and peaking at just over 20km. What this chart does not show is the number of trips included in each mode’s total. For example, the lowest income riders also go the furthest on “GO Transit + municipal”, but this does not mean that low income riders are taking a huge number of long GO trips.
The next chart shows the distribution of trips by length. Trips under 7km (the screenline used by Metrolinx for a local trip or a zone size) are highest for the lowest income group at 45%, but they are also substantial (36%) for the highest income group. This could be due to the availability of subway service in areas where incomes are higher, and therefore short transit convenience trips are attractive. Where incomes are lower, a short trip, especially one off peak, incurs a substantial penalty compared to auto travel is this is an option.
There is no subdivision of data to indicate geographic variations in origin such as those who live in central Toronto, those who are in the “outer suburbs” (former suburbs of “Metro”) and those who are somewhere outside of the “416” completely. Variations in the availability and type of transit are important considerations.
The next part of the analysis is very revealing about the outlook of the study’s authors. In the first chart, we see the proportion of trips involving “Planning District 1”, or downtown Toronto.
Again it is no surprise that lower income group tend to be travelling outside of PD1.
Increasingly, this is an area where low income residents are few in number, and jobs they might have are greatly outnumbered by higher-paying professions. This area is served by the heart of the subway system.
Meanwhile, the group driving all discussions of fare integration are the poor souls who must pay double fares to cross a boundary. Note the huge difference in scale between this chart and the one above. The numbers below do not include GO riders who are, for the most part, bound to and from Union Station, and so this reflects mainly those who pay double fares at the 905/416 boundary. Relatively few riders come all the way into PD1 from the 905, and few attempt the crossing for areas outside the core where at least part of the trip is likely to be by subway.
Although GO customers have been omitted, any fare integration policy will include some sort of discount for a GO+TTC trip, and this will benefit all GO riders. It is also central to the implementation of SmartTrack within the integrated fare scheme. There is no estimation of the number of riders of all income classes who will benefit from this fare reduction.
A few charts reveal how different income groups use transit. These differences are almost certainly a function of availability of transit or auto as alternatives, and of the relative convenience of these modes for each type of trip. Again, there is no breakdown to show whether these patterns differ by major geographic territories.
For each type of trip, the number taken by transit drops as incomes rise, going up again slightly for the highest group.
On a time-of-day basis, proportionately more trips are taken during the peak as incomes rise. This shows how transit is less competitive as a mode for those who can afford an alternative. Even so, the overall proportion of off-peak trips shows that transit in the GTHA (taken as a whole) does not exist only to handle peak period commuters.
Although these breakdowns are all interesting, they do not address the fundamental question of how a new transit fare structure would affect riders overall especially in how it might discourage people from taking longer trips by transit due to extra cost.
As I discussed in a previous article, the 7km screen used to distinguish “short” trips by Metrolinx leaves huge parts of the city and GTHA facing trips that will fall into multiple zones, or which will exceed the point at which there is a benefit from a reduced fare. All of this is in the interest of making “short” trips across the 905/416 boundary cheaper for a small portion of the total transit population.
To some extent, inequities for short trips regardless of where they might be can be addressed in other ways including bulk or capped fare costs (passes or maximum charges per time period). These alternatives are completely ignored in this study which concentrates on the single fare model.
Updated February 19, 2017 at 6:45 pm:
In the Metrolinx study, the summation of data across the entire GTHA misses the basic fact that there are differing circumstances in each neighbourhood both within the City of Toronto and among locations beyond. When someone makes a decision to take transit or drive, this is strongly affected by their local circumstances, and in some cases the fare charged may have little to do with the decision.
Here are a few suggestions for additional data Metrolinx might like to consider:
- What is the auto ownership by household in each of the dissemination areas forming the basis for their study?
- What is the average household size and age distribution? Working adults, children, elderly, etc?
- What is the trip destination distribution by length, type (work, school, etc)? For example, are work trips typically longer than school trips? Are they more likely to be taken by transit?
- What is the transit service density available in each area for peak or off peak travel, and for regional versus local travel? In other words, how much of an option or hindrance would transit be to making journeys? Where are the “transit deserts”?
- What do the isochrones look like for major nodes, especially those not on the rapid transit network, depending on which services are available? Peak vs offpeak? Premium vs local fare? (An isochrone is a map showing how far one can travel from a point in a given amount of time.) These can be used to show the benefit, if any, of changes in a network in a manner that is more inherently understandable than abstract values such as vehicle kilometres travelled and imputed savings averaged over all trips.
Metrolinx needs to understand and to publish information showing how their simulated population of transit riders would be affected by new fare structures both for the existing network, and as it grows, particularly in the conversion of transit corridors and trips from “local” to “rapid transit” or “regional” fare premiums. This needs to consider all types of trips in the network, not just the handful affected by cross-boundary fare premiums.
Of particular importance, Metrolinx needs to distinguish between co-relation and causality. The behaviour of riders in various parts of the GTHA is a function of the services and fares available to them, not necessarily a choice that they made to favour one type of travel over another.