Skip to content

Canadian Urbanism Uncovered

Commuting snapshots across the Spacing map


Read more articles by

Source: Statistics Canada

By Emily Richardson — cross-posted from Spacing Atlantic

Despite dramatic differences in population, density, infrastructure, and growth, there is remarkable consistency between commuting patterns in Montreal, Ottawa, Toronto, and Halifax, particularly when it comes to travelling by car. And incidentally, when it comes to getting us out of them, we seem to find buses and bike lanes unconvincing. A closer look at our most recent census data raises some surprising – and some predictable – findings about the way we get to work and how preferences change as our cities grow.

First a few words on sources and statistics: All data in this article, with the exception of bike lane information, is based on the 2006 census of Halifax, Toronto, Ottawa, and Montreal‘s census metropolitan areas with trends from the 2001 census. I will be the first to concede two important shortcomings in the data. First, neither the number of cities nor the number of data points within each city is sufficient for any analysis to be statistically significant (in other words, proper analysis requires more, and more robust, data to hold up to scrutiny). Second, the data is a static snapshot in time, and it lacks any context that might explain why the upcoming 2011 census might paint a vastly different picture.

But despite these drawbacks, the census data does highlight some consistencies between cities and concerns about the economic, social, and environmental implications of our commuting habits. It remains to be seen whether erratic fuel prices, transit-pass tax incentives, and growing bike-lane networks over the past four years will meaningfully influence our commuting habits by 2011. In the meantime, comments and observations are welcomed in response to this anecdotal food for thought.

With this disclaimer out of the way, there are numerous ways of charting commuting data in Halifax, Toronto, Montreal, and Ottawa; this discussion will highlight three. Foremost, although data can be spun, twisted, and manipulated to support nearly every position, this information isn’t so malleable that it can be finessed away from our shared affinity for personal-vehicle commuting. Perhaps it’s because the bus doesn’t come through our neighbourhood, or because the subway presents an intense and twice-daily invasion of our personal space, or because cycling means taking our lives in our hands. Regardless of the reason, the vast majority leave the house car keys in hand.

Observation #1: We all love our cars.

No self-respecting urbanite envisions car-centric urban utopias, but the vast majority – around 70% – of commuters in each city travel to work by car (by themselves or with others). Interestingly, for example, Halifax’s density of 68 people per square-km to Toronto’s 866 might suggest that Toronto would be better designed to accommodate the higher population, but some in Toronto commute from famously far-flung distances, and Halifax’s sizable rural population is not adequately served by transit to entice commuters from their cars.

In other words, density doesn’t appear to be a factor, and yet…

Observation #2: Increases in density reduce the number of single-occupant vehicles, but don’t necessarily get people out of their cars.

Note: Montreal is missing its “Other” category figure because the increase in density (0.83%) proportional to the increase in “Other” commuting (22.9%) was so small that the change charted beyond the boundaries of the graph. Recall, though, that the number of commuters using “other” modes is so small that changes in driving, public-transit use, and cycling/walking likely involve many more people and therefore have a far greater impact overall.

The good news is that single-occupancy vehicles are reduced when density increases; the bad news is that we seem fastidiously iron-fisted when it comes to letting go of car commuting. Halifax is the only of the four cities whose commuters are less likely to walk or cycle as density increases; instead, we switch overwhelmingly to public transit. The inverse relationship between Montrealers’ propensities to take the metro as opposed to walk or cycle is probably explained, at least in part, by its impressive 502-kilometre network of bike lanes.

However, bike lanes themselves don’t appear to be overly persuasive in enticing commuters to join a work-bound peloton…

Observation #3: The availability of other modes of transportation doesn’t necessarily mean people use them.

Taking a closer look at the relationship between metres of bike lanes per square-km in Halifax, Toronto, Ottawa, and Montreal compared to the percentage of commuters who walk or ride their bikes (see links for sources), it would seem that providing bike lanes isn’t necessarily correlated to increasing the number of people who use them.

I would have loved to write about the kilometres of bike lanes per square-km; sadly only Montreal would have been greater than one. And to be fair, although Halifax and Toronto have somewhat comparable total areas (5,495.62 and 5,903.63 square-km, respectively), Toronto has vastly more bike-lane-suitable stretches of road, suggesting that its 90 km of bike lanes lags disproportionately to Halifax’s 70. Montreal’s bike network is far and away the most extensive of the four cities, but Montreal also has the greatest number of drivers of the four cities (by a 0.25% margin over Toronto). And even though a demarcated lane is a victory for most advocates of multi-modal transportation, in and of themselves, they do not appear make cycling easier. Ottawa, for example, with its 150 km of bike lanes, laments a number of concerns with on-road lanes, such as lanes ending in the middle of two merging roads.

This raises the question not whether lanes are worthwhile in encouraging cycling, but what programs and policies (e.g. bike-sharing, mandatory shower facilities in new commercial buildings), infrastructure (e.g. divided bike lanes and covered racks), incentives (corporate tax breaks based on percentage of employees who commute by foot or bicycle), and disincentives (gas and congestion taxes) ought to be coupled with lanes to encourage commuting by bicycle.

And more broadly, perhaps the question is not why we take the bus, walk, or cycle so little, but rather why we drive so much. Each of these cities is served by a transit system, and although progress may be slower than many prefer, each is making efforts to increase bike-friendliness. Is it because transit planning can’t keep up with sprawl? Or is it because these cities are just too big to serve widely through alternate transportation? Is it because we would only change when we had to start paying for road congestion and air pollution? Or in the end, could it be that despite the rhetoric for more bike lanes and bus routes, we really do love to drive?

This introductory analysis would benefit greatly from the perspectives of those with more nuanced understandings of each city’s commuting patterns, why we choose what we do, and how we might change. With this, I pass this over for comments, analyses, speculation, and wild guesses.



  1. What are the other modes? I know I can’t think of any. Boat maybe? :P

  2. Those 502 kms of bike paths and lanes were’t there in 2001 when the data on commuting was collected. I have a hunch that active transportation is WAY up in this city (which is why the city built all those bike paths – its a positive feedback cycle). I am still eagerly awaiting the AMT’s 2008 report to confirm this…

  3. I’m not clear on why you say that the number of data points within each city is insufficient to be statistically significant. 1 in 5 households answers these questions on the census (the long questionnaire). The sample size for the smallest of these cities is in the order of 40,000.

    I’ve often wondered about this, so if you have more information, please share. My understanding though is that it’s safe to say, for example, that Halifax has a higher proportion of car commuters than the other cities reported here.

  4. *In 2006* Halifax had a higher proportion of car commuters than the other cities reported here.

  5. Mike — Just a point of clarification: by insufficient data points within each city, I meant rather that there aren’t enough years’ data for each city to draw any inferences about trends, not that there aren’t enough data points within each census for each city.

    And I couldn’t agree more that it will be interesting to see how active transportation changes in 2011 for each city, particularly when matched with growth in bike lanes since 2006 — matching growth in overall bike-lane distance with growth in active transportation is definitely a better indication of the success of bike-lane projects than a single point in time.

  6. According to the “Observation 2” graph, a 1% increase in density leads to a 5-6% increase in motor vehicle use in Montreal. Does the % increase in motor vehicle use use refer to the proportion of people who commute by car or a rise in the total number of vehicles? I assume it is the later, because I believe its fairly well that transit and active transportation have a higher modal share in the city centre compared to the suburbs. Certainly per-capita car ownership is lower in the denser neighbourhoods.

    Also, the article says that single-occupancy vehicles are reduced when density increases, but the graph seems to indicate that the vehicle use goes up quite significantly – does this mean that there is a massive increase in car-pooling?

    Finally, i was wondering whether the 1% increase in density in space or over time (ie are you comparing different neighborhoods or comparing places that have gotten denser over the 5-year period?).

  7. In response to Alanah:
    1. I understand the rise in the Montreal figure to be an increase in the number of people who carpool. I can’t say whether vehicle ownership changes, but the data suggest that although single-occupant-vehicle commuting falls with a 1% increase in density, carpooling increases 5.48% (vehicle ownership may rise, stagnate, or fall, but carpooling as a means of commuting seems to increase).

    2. See above. I would interpret the data not as vehicle use rising, but the people who formerly commuted as a driver (or took public transit) rising. Do you see it differently?

    3. The density figure is an overall figure for the entire Montreal census metropolitan area (unfortunately, the census doesn’t go into individual neighbourhoods, but that would be very interesting for the four cities). It reflects an increase of 846.6 people/square-km to 853.6 people/square-km, or 0.83%, the smallest increase of the four.

  8. Throw these stats out the window, BIXI makes a mockery out of them!

Leave a Reply

Your email address will not be published. Required fields are marked *