The earlier post identified three negative impacts of filtered permeability:
- Filtered permeability breaks the street network and pushes traffic away from “residential streets” on to the surrounding roads which increases congestion, air pollution, and journey times.
- By describing these areas as “residential streets” it implies that no one lives anywhere else. Often the streets that traffic is filtered towards are actually high streets with higher densities and a wider mix of uses. There are consequently more people on them, and therefore more people exposed to increased traffic and air pollution.
- The streets where traffic has been restricted are already quiet, and experience further reductions in natural surveillance, reducing personal safety (and the perception of safety).
These points were made on the basis of street network analysis, and some further investigations in to the distribution of land use and density in and around de Beauvoir.
A further consideration in the earlier study is that the resulting distribution of traffic volumes (and air pollution) is socio-economically distorted: higher density, multi-dwelling, buildings are more exposed to traffic, while lower density houses are less exposed.
The drive for filtered permeability is to encourage cycling, and this is positive for many reasons. However, cities are complex systems that work across many scales simultaneously, and are formed from different layers of infrastructure, public transport, land use, density, and people, all of which need to be considered in an integrated way.
The last fifty years of urban planning, and specifically the dominance of traffic engineering, shows how negative outcomes for the whole city arise by working with one infrastructure system in isolation from the others. Despite the positive objective, the risk of using filtered permeability to try to deliver improved cycling infrastructure is that it will create unintended negative outcomes across the wider area.
Hackney council currently has a consultation running to extend a filtered permeability scheme further north into Stoke Newington. This post uses spatial network analysis to test the impact of these new proposals.
Already the proposed (and implemented) scheme has proved divisive with comments on social media reflecting some of the complexities Hackney must respond to. Some focus on the negative impacts of breaking the street grid and reducing the number of routes through the area, while some identify funding cuts to local services as a problem.
Cities are made up of many physical and non-physical systems that interact with each other to shape what happens. However, the spatial network is hugely influential on the outcomes in a city, and is essentially locked-in from the moment land ownerships are defined.
It is true that removing funding for services won’t help to reduce levels of anti-social behavior or crime, however blocking these streets will create the spatial conditions where this type of activity has been found to happen, regardless of what social services programmes are running.
Justification for further changes
Before starting any analysis of proposals, the consistency and logic of the justification is worth looking at. The set of policies the scheme aims to deliver questions how well suited they are to the urban characteristics of the area:
“The Transport Strategy includes a number of actions that are relevant to this scheme:
• LN15: Filtered Streets – Reduces motor traffic on residential streets by use of road closures to create safer walking and cycling conditions.
• C33: Area Based Filtered Permeability Reviews – Undertakes area wide traffic reviews to stop rat running and continue rollout of filtered permeability schemes.”
This suggests that people only live on “residential streets”, however as mentioned above, the streets that traffic is encouraged on to also have residential uses (in the case of de Beauvoir at higher densities). They have mixed uses (including schools and doctors) too, which means more pedestrians across a wider period of the day.
So if filtered permeability works as intended, it is likely to reduce traffic in some of the (lower density) mainly residential areas. By doing this it is arguable that restricting traffic to what are effectively high streets will expose more people to more traffic.
The description of the scheme contains contradictions that question whether the current implementation of filtered permeability has worked at all.
While apparently succeeding in: “reducing traffic on the cycle route and improving air quality and road safety near the schools” at the same time the introduction notes; “it is clear that some roads, particularly Walford Road, Brighton Road and Nevill Road, are now suffering from increased traffic and aggressive driving”.
It also mentions that; “Residents have raised concerns about the volume of traffic and the difficulty to negotiate certain roads, especially Walford Road and Brighton Road, as well as concerns about safety of children crossing these roads, traffic congestion and anti-social behaviour as a result of driver confrontations.”
By mentioning these as outcomes of the earlier scheme, it’s questionable whether it should be referred to as “successful”. Especially when set in the highly emotive language used in the scheme description:
“Walford Road and Brighton Road are not just places to park vehicles or drive, walk and cycle on. They are the places where we live our lives and Hackney’s children play. We want to reclaim Hackney’s streets from motor traffic congestion and transform them into the most attractive and live-able neighbourhoods in London.”
Analysis of current area
Before analysing the two proposed schemes it is worth looking at the current arrangement.
The area today includes bollards on Wordsworth Road, Truman Road and Cowper Road (all circled red). These are included in the space syntax analysis below:
This analysis shows how well connected every street is to the wider city, with red lines being well connected and blue lines being poorly connected.
At this scale of analysis (10km), red lines typically show the distribution of movement by car, however this analysis has to be caveated by mentioning that the model has not been correlated against traffic counts.
What it shows in terms of network hierarchy is that Brighton Road is significantly better connected than the surrounding streets. One of the reasons for this is that it forms one of the only links between Stoke Newington High Street and Albion Road. Along with Walford Road these are the streets the council state as suffering from more traffic and aggressive driving.
What the model also shows is that while Brighton Road remains well connected, others are much worse (Wordsworth and Pellerin Roads). This raises the issue of natural surveillance; even a person in a car is a person, which can be enough to provide a sense of safety to people walking home at night, or to act as a deterrent to a potential burglar, or drug dealer.
So while the model has not been validated, it shows spatial characteristics that are consistent with what has been observed, and indeed what the council use to justify the scheme.
By adding bollards to create filtered permeability, what was a connected grid has effectively been transformed in to a set of isolated areas and culs-de-sac. This is both a contributor to the current problems, and the wider critique of filtered permeability.
This option proposes changes around Allen and Barbauld Roads. These changes have been added to the spatial network (below left), and the spatial network analysis repeated (below centre).
To look at the specific impact of proposals an additional analysis has been included (below right). This new analysis looks at the percentage change from the current arrangement created by the proposal. Red lines record an increase in connectivity (which is likely to be matched by traffic) while blue measure a percentage decrease.
The impact of breaking Barbauld Road is that the routes out of this area are reduced, and this in turn pushes additional traffic on to Neville Road (or CS1).
This is also likely to increase traffic in many streets to the north, including all of the connections on to Stoke Newington Church Street. These streets are generally double parked and too narrow for two cars to pass.
It is therefore a possibility in this option that the aggressive behaviour around Brighton and Walford Roads identified as the need for these changes is likely to shift further north, and in this case on to CS1 itself. Trying to reduce the potential conflicts between cyclists and vehicles was identified as a reason for making these changes.
This option proposes further breaks to the network around Allen Road and Nevill Road.
Brighton and Walford Roads measure a percentage decrease in connectivity indicating that they would indeed be used less by traffic.
Barbauld Road and Albion Grove, along with the streets joining Stoke Newington Church Street all measure increases however. As in Option A, these streets are similar in nature to Brighton and Walford Roads, being too narrow for two cars to pass. It is therefore a distinct possibility that the aggressive driving cited as the reason for these additional changes will migrate further north.
At the same time as these increases, the north-south route formed by CS1 and followed by pedestrians measures a decrease. While this appears positive in terms of reducing the potential interactions between traffic and cyclists, it is again likely to create negative impacts on natural surveillance and personal safety at night.
Implementing the first phase of the scheme has already had negative impacts. Both proposal options are likely to create further negative impacts, either increasing the amount of traffic on CS1 (Option A), or in terms of reducing natural surveillance (Option B). Both will create an increase in traffic on the streets joining Stoke Newington Church Street which itself has high levels of pedestrians including a high proportion of children.
Looking at the wider urban context, the need for filtered permeability stems from the attempt to thread a cycle superhighway through a set of streets that aren’t really suitable. In this area CS1 follows an indirect route, often along streets that are too narrow for two cars to pass.
The way to resolve many of the issues created by filtered permeability would be to consider realigning CS1 to run along Stoke Newington High Street/Kingsland Road. This would require major work, including changing the gyratory to become two-way, however where this has been done in other areas it has been seen to bring benefits to users other than motorists. It would also allow the damage of the first filtered permeability scheme to be undone.
Friedrich, E, et Al. 2009, Anti-social Behaviour and Urban Configuration Using Space Syntax to Understand Spatial Patterns of Socio-environmental Disorder