PT hasn’t covered the debate over Toronto’s Gardiner East Expressway, given the amount published elsewhere. But this piece by Auckland’s Darren Davis (a participant in the SFU City Prgram’s Next Generation Transportation certificate) is worth reprinting for its insight on the impact of traffic-capacity reduction.
From Architect This City:
Guest Post: Three minutes that rule the world – Will demolishing the Gardiner East actually make traffic worse?
The decision this month on whether to demolish the Gardiner East Expressway and replace it with a surface boulevard or to rebuild it with a similar elevated structure will be a watershed moment for Toronto, akin to the decision not to have freeways in the urban core of Vancouver in the late 1970s.
There are numerous good reasons why removing the Gardiner East elevated structure is the right move for Toronto, covered previously by The Globe and Mail and in the Council for Canadian Urbanism’s open letter to Toronto City Council. If you aren’t aware of these, I strongly recommend reading them.
Also of note are two things:
- Just 3% of Downtown commuters drive on the Gardiner Expressway East, a tiny fraction of the 68% who arrive Downtown on public transit.
- The “remove” option does not reduce traffic capacity over the “hybrid” option but could reduce travel speeds in uncongested conditions (read on for why I don’t think that this will happen in real life).
However, one element of this debate that has not got much airtime is the transportation modelling that claims an additional two to three minutes travel time with the “remove” option over the “hybrid” option which retains the elevated expressway in a slightly modified form.
Toronto Mayor John Tory has stated that “I didn’t get elected to make traffic worse. And let’s be clear, removing that piece of the Gardiner will almost certainly make traffic worse.” But is it in fact true that demolishing the Gardiner Expressway East will make traffic worse as the transportation modelling claims?
The key thing to understand is that transportation modelling, which tries to predict future travel times, is the product of a bunch of assumptions which may not in fact be borne out in real life.
For example, the effect of induced traffic, where additional traffic capacity leads to additional traffic being attracted to the route, is reasonably well known but generally not factored into transportation modelling. Induced traffic happens because increased travel speeds attracts traffic that would have otherwise avoided the route at congested times; attracts people to drive where they previously used other modes and encourages trips that would not otherwise have taken place. The effect of induced traffic is to quickly nullify the benefits of adding traffic capacity.
However, what is less well known is that the reverse happens where there is either a reduction in traffic capacity (not the case in Toronto as the “remove” option retains the same traffic capacity) or speed.
This is because in this situation, four things happen:
- Some travel re-routes. The “remove” option with a widened Lakeshore Boulevard as part of the street grid simply gives more ability for traffic to re-route away from congestion over an elevated expressway structure where drivers are literally trapped until they reach their exit.
- Some travel re-times. Some people will retime trips to avoid congested travel times, such as starting and/or finishing work at less congested times.
- Some travel changes mode. Some people will be encouraged to change mode by the perceived worsening in traffic conditions.
- Some travel is avoided entirely. Some people will choose not to travel at all in the peak of the peak. For example, this could take the form of working from home or shopping on the internet instead of by car.
Transportation models only take into account the first item but not the others. The modelling is also sensitive to traffic growth assumptions and assumed mode split and trip distribution. Often transportation models assume continued growth in car travel even though per capita kilometres travelled peaked about a decade ago.
While this may all sound well and good theory, does this actually happen in practice? Read more »