A couple of weeks back the Design Council Newsletter re-published some work from 2014, looking at how Hackney had become one of the most liveable boroughs in London. This is evidenced by an increase in people cycling to work and a decline in those driving – both around or over 10%.
One of the reasons set out for this change in behaviour is the success of policies implemented over the last 10 years, which include re-designing residential streets. In particular the use of “filtered permeability” – using bollards to make some streets only accessible to pedestrians and cyclists – is described as part of this achievement.
Increasing active travel share and decreasing private car use is very positive, however, closing streets immediately raises questions around the wider impacts.
In a very quick summary there are three major issues with filtered permeability:
- Fragmented networks increase congestion and lengthen car trips
- Increased likelihood of air pollution in higher density, mixed use areas
- Reduced levels of natural surveillance impact personal safety
These points are expanded in the post below:
Fragmented networks increase congestion and lengthen car trips
De Beauvoir Town was designed as a fairly regular grid with diagonal routes in three of its four corners. The east-west routes are generally well-connected to the wider network – to the west they extend to Islington, to the east they reach further into Hackney. These connections mean that the original grid worked well to allow wider scale movement around and through the area.
By breaking the network in a few places, it effectively prevents some of these connections being used at the wider scale. What this does to the network is to split it in to culs-de-sac. This creates a number of fragmented areas. Movement through these areas may reduce, but because wider scale movement is now restricted to a smaller selection of routes, it is likely to increase on the unbroken connections – as shown in the models below:
So not only is this likely to create more traffic on the surrounding routes, but because the people who drive out of de Beauvoir are now restricted in the way they can exit the area, it also makes their journies by car longer.
While lower levels of traffic in the area itself may make better conditions for cycling and walking, because it increases traffic on the surrounding areas, it may actually make it more difficult to cross the surrounding roads – this is called severance.
Increased likelihood of air pollution in higher density, mixed use areas
On one level it may be fair to penalise journies by car; making them longer and slower should provide an incentive to use active or public transport, however the question is how it impacts the surrounding area.
One implication of the term “residential streets”, is that there are no residential uses on the surrounding streets, however this is not the case. In fact, the images above show that the built density is higher on the surrounding areas (top row). This is confirmed in the maps further down.
Increasing traffic on these routes adds to congestion, and with cars stationary for longer could create more air pollution. These areas are not just higher density, they are also mixed use, with shops, cafes, pubs, pharmacies, clinics, and community centres located on them. Increasing traffic on these streets actually brings more people in to contact with air pollution, including moe vulnerable people.
While overall this principle stands, it needs a slightly more nuanced analysis. The increase in movement on the higher density streets tends to be lower than the decreases on the low density streets. This leads to the question of whether the overall reduction offsets any overall increases.
By relating potential movement to the number of people exposed to it, it helps build this fuller picture. The analysis (below right) still shows that around the edges of De Beauvoir Town more people are exposed to more movement by car.
Reduced levels of natural surveillance impact personal safety
While restricting access to cars generally makes things better for cyclists, there is one point worth discussing in more detail: even a person in a car is a person, and this creates a level of natural surveillance.
Around five to ten years ago, De Beauvoir Road/North Church Road had a reputation for somewhere that was dangerous to cycle through at night; there were reports that people cycling on their own had been knocked off their bikes and attacked whilst their bikes were stolen.
More recently this has shifted further north towards Culford Road. As movement through the area is restricted, the number of people passing through during the evening (even by car), has reduced further. This makes areas that had a reputation for being dangerous worse: it’s less likely someone will pass through and put off an attacker, less likely someone will see what is happening and call the police, and less likely that someone will find a cyclist on the floor having been attacked.
An increase in people walking and cycling has to be positive, however in this case, filtered permeability creates unintended outcomes.
At the scale of De Beauvoir Town this may have a fairly small impact, but rolling out the same approach across the borough, the risk of increased congestion and air pollution on a few key routes is increased.
These key routes are the spaces of higher density and mixed use, meaning more people will be exposed to traffic and air pollution.
Assumptions and Limitations
There are some limitations and assumptions to this model;
- It hasn’t been calibrated against movement data, nor weighted by land use/density (this would likely further amplify the impact on the edges. It shows potential movement only as a result of street configuration, however in other parts of London this has been shown to correlate well with both pedestrian and vehicle movement.
- The only area tested was De Beauvoir Town. There have also been similar changes made across the wider borough. If these areas had been modelled it is likely the impact on the key routes would increase further.
Streetview images extracted from Google