Maximising permeability for cars, making it easier for them to drive in all directions, will encourage people to walk and cycle.  How did such an obvious delusion come to be so widely accepted that its restatement in Manual for Streets passed almost unchallenged amongst the planning and transport professions?


Like many ideas to mislead our policymakers, this one can be traced back, at least in part to North America and a failure, when looking across the Atlantic for inspiration, to ask some basic questions.


‘Traditional neighbourhoods’ in North America were typically designed around rectilinear street grids. More recent lower density suburban developments are characterised by looping distributor roads, serving long disconnected culs de sac – similar to this country but more extreme.  Although the car is king across most of North America, the traditional neighbourhoods tend to generate lower car mileage.  Some researchers have ascribed this, at least in part, to their greater permeability.  In other studies, controlling for other factors made the form of the built environment statistically insignificant.  Either way, it seems this ‘traditional’ form of permeability exerts only a weak influence on people’s patterns of movement.


Notwithstanding these uncertainties, the New Urbanist movement has helped to spread to this country the concept that car use can be reduced by permeable street networks, which allow all road users to move unimpeded in all directions (‘unfiltered permeability’). The limited evidence available so far gives little support: car use in Poundbury is higher than both the National and West Dorset averages.  In all the debate about Poundbury and its many derivatives, few have questioned the shaky evidence on which the New Urbanist claims about transport and movement are based.


Looking again at those American studies, the comparisons disguise two countervailing forces: the traditional grid reduces journey distances on foot, but also by car.  Where all road users travel together, the car will generally emerge as the quickest and most convenient option.  A third possibility, where permeability is deliberately restricted for private motor vehicles, but maximised for walking, cycling and public transport –

‘filtered permeability’ – was absent from those studies because examples are lacking in North America (and, at the town or city-wide level, in Britain).


Not so in Europe.  I have spent the past two summers cycling to and around towns and cities with particularly low car use, interviewing planners, transport planners and observing people’s travel behaviour.  The terminology varies, but in cities such as Freiburg, Groningen and Zwolle the principle of filtered permeability is acknowledged as a key element in their success in restraining car use and promoting alternatives.  Through traffic is channelled onto a limited network of main roads.  Suburban developments are often designed as giant culs de sac for cars, while bridges, tunnels, cycle paths, bus gates – a whole panoply of short cuts provides a far more permeable network for the sustainable modes.  People use these modes – particularly cycling – because of the time and convenience advantage compared to travelling by car.


In a public transport context, the principle of filtered permeability is uncontroversial.  Its assumptions are built into the models which many of you use.  If a guided busway provides a time and convenience advantage compared to the same journey made by car, we would expect bus use to rise.  Building a new road alongside the busway would undermine the relative advantage offered to the bus, with obvious consequences for modal share.  Why then do so many planners and urban designers fail to grasp that the same principle applies to walking and cycling?


Manual for Streets argues that unfiltered permeability leads to “a more even spread of motor traffic throughout the area and so avoids the need for distributor roads”.  This is probably closer to the crux of the issue.   By multiplying opportunities for ‘rat-running’ unfiltered permeability can increase the capacity of a road network to carry traffic – and, course to emit CO2.  In other words, it is a cheaper variation on the ‘build our way out of congestion’ theme.


Distributor roads, the New Urbanists sometimes argue, are ‘anti urban’, although their masterplans still include a hierarchy of roads, however they are presented.  You won’t find any Spaghetti junctions in Freiburg or Groningen, but you will see substantial areas closed to through traffic, and wide main roads, some with tramlines down the middle, and wide, well-designed cycle routes – separating cyclists from cars and pedestrians.  Congestion is generally less of a problem because most journeys are not made by car.


Another argument relates to crime.  Again, conflicting, ambiguous research findings have been oversimplified to suggest that cycle paths, footpaths, or streets closed to through traffic lead to more crime.  In fact, some studies have suggested the opposite.  The devil, as always, is in the design detail.  The transport advantages of filtered permeability can be combined with passive surveillance, and it is worth reminding crime prevention design advisors that every street or path blocked to police cars is also blocked to getaway vehicles.


Although the authors of Manual for Streets clearly favour the unfiltered form of permeability, the wording does not preclude a filtered approach.  Other Government guidance, such as the recent Cycle Infrastructure Design Draft, is more explicitly favourable.


The built form is, of course, only one of several influences on travel behaviour.  The circumstances of every development will differ, but if we are serious about modal shift, permeability must be seen as an opportunity to differentiate between those modes we want to encourage, and those we need to restrain.



Steve Melia is a PhD researcher at the University of the West of England and represents Carfree UK on the DfT’s Round Table on Ecotowns.