Strategic Urban Design

Roger Evans
Creating urban capital

Roger Evans, urban designer, architect and town planner, explains what makes good places and shapes urban design ideas


Over the next ten years around three million new homes will be built in the UK together with supporting uses such as convenience shops and services, schools and workplaces. At current typical housing densities around 90 square miles of land will be needed to accommodate this - roughly the size of Surrey. We need to ensure that not only is this volume of development fit for purpose, but that it bequeaths an urban form that will serve future generations.

The idea of natural capital is mainstream in town planning and perhaps we now need to be more aware of creating ‘urban capital’ - urban areas that have good urban form and which will be an asset for generations to come. So what is ‘good urban form’ and how can we ensure it is achieved?


Good urban form achieves social, economic and environmental objectives through the spatial arrangement of streets and buildings within the landscape. Buildings might be replaced over time, but street patterns and building lines can last for centuries. Urban form is the shape of our human habitat, and whether it is good or bad can be measured against six criteria:

1. Development in the right location
The location of new development will influence whether people can get about by walking or cycling, or whether they will be dependent on cars. Parking standards for car-dependent developments mean that their parking footprints will be the same as the housing footprint, unless housing densities rise to justify multi-storey or below ground parking solutions.

Where new housing is not adjacent to employment opportunities or served by rail, the connecting roads will need to cater for commuter traffic. Bus services are rarely frequent enough to rely on, unless new development is an extension of an existing urban area with a good standard of existing provision. Furthermore, the towns to which people will travel from these car-dependent developments will need to provide additional parking too. This dominance of cars brings pollution, noise and safety concerns with the likely result that residents will walk and cycle less, or encourage their children to do so. Development that is planned as growth to sustainable towns can avoid these problems. New settlements need to be of considerable size to be self-sufficient, so that they minimise the need for travel just to meet people’s daily needs.

2. Hefted to the landscape
Historic towns and city quarters were settled and grew incrementally, either in a planned manner or as plot-by-plot development along lanes and streets, or a combination of both. ‘Design’ decisions about new development were taken by people who could walk around their neighbourhoods along routes that followed the underlying topography in the most efficient way, who knew the local micro-climate and so could avoid frost pockets and shelter from prevailing cold winds.

3. Connections to existing streets and paths
Good urban form is well connected, so new streets, footpaths and cycle ways should join seamlessly onto the surrounding movement network, and ideally in all directions. These connections will generate the street layout and hierarchy of streets - from main streets to mews lanes - within the development. Development that forms a large, inward-looking pod in the countryside, or is the result of a bypass creating a barrier to the surrounding landscape, is unlikely to have this connectivity and is more likely to have a hard-to-read and poorly connected internal street pattern.

4. Rules of assembly
The form of development needs to respond to human characteristics if it is to work well for people. For example, we walk at about 2-3mph and therefore need to be able to meet our needs within 10 minutes’ walk, or about a half mile. We construct mental maps of street patterns, and have an expectation of where the main public buildings will be and the direction of the river; in fact if we find ourselves in any town that has grown gradually, we can usually find our way around instinctively.

We have a range of senses that have determined the ideal scale of public places. We can discern body gestures up to 150m away (the dimensions of sports stadia are determined by this), at 25m we can recognise a friend’s face, at 12m facial expressions can be read, and at 6m the subtleties of speech and gesture can be conveyed. Spatial dimensions up to 6m can appear intimate, while 30m is probably the maximum distance within which a shared awareness between people is likely. The public realm is common ground in both a geographic and spiritual sense: if urban space is to promote healthy social contact, then its scale is critical.

5. Settlement history
The history of a town or quarter should inform the planning decisions that we make today. Yet streets that developed over centuries to accommodate people on foot and horse-drawn wagons were seen as unfit for modern use by the second half of the 20th century. For example, the proposed destruction of Glasgow’s entire city centre in 1945, to be replaced by development in the Modernist style and including a motorway interchange, was typical of many towns and cities throughout the UK. We now value such walkable places. The archaeology of a town may be displayed in the local museum, but it should also be evident in the streets today; these are anchors to the past.

6. Adaptability
New places need time to settle into themselves. This means that we should deliberately include the capacity for change and adaptation, and avoid the temptation to fix every detail at the outset and for all time. The following approaches can encourage adaptability:

  • Allowing flexible building uses along main streets and around public squares - to enable changes from, for example, residential development to small shops or services, and back to residential use.
  • Including a narrow area in front of buildings - to accommodate adaptations to buildings.
  • Building extendable structures - to enable an increase in floorspace within limits.
  • Providing wide pedestrian pavements in central areas - to use for outdoor extensions to cafes etc.

Some of our most prized city centres were once modest residential areas, but the buildings and streets are both robust and flexible, enabling intensification and changes of use to take place yet with much of the street scene intact; these are known as long-life and loose-fit places. How many housing estates built today have that  capability for change?


A range of problems have been raised in recent studies. Transport for New Homes’ report Garden Villages and Garden Towns: Visions and Reality (2020) included research which found that most of the current programme of Garden Villages and Garden Towns have been planned in the wrong locations, far from town centres and rail stations; they lack local facilities and their streets are designed around car use. The study Location of Development (2018) commissioned by the Royal Town Planning Institute found that only a fifth of new housing units surveyed were within walking distance of a public transport node, such as a railway station or tram stop.

A Housing Design Audit for England (2020) conducted by Matthew Carmona at UCL for the CPRE, and supported by housing industry organisations and a number of environmental charities including the UDG, showed that there has been little improvement in housing design quality nationally since audits were last conducted between 2004 and 2007. However, because this improvement is from a low starting point - what CABE at the time called ‘an uncompromising and unflattering picture’ - the large majority of new housing developments are still assessed as ‘mediocre’ or ‘poor’. Three quarters of the audited projects fell into these categories. The report finds that the worst new estates lack nearby amenities such as shops, pubs and cafes. They are unconnected to the surrounding areas, with few public transport links, and do little to encourage cycling and walking.


Good urban design usually embodies three key characteristics:

  • It considers a range of spatial scales
  • It contains projects that are delivered across different time scales, and
  • It can embrace multiple land ownerships.

Spatial scales - it is important to consider the context of a whole town or sub-region, the three-dimensional form of new places, through to the architectural and engineering details of those places.
Time scales - an urban design approach co-ordinates multiple projects, each with different time frames. These might include engineering infrastructure, early landscape works, or building on different sites at different stages.
Land ownership - the real clients of urban design projects are those who will live in and use the places created, so plans need to allow for multiple land ownerships and be shaped by the landscape rather than a single land-owner’s boundaries.

In short, good urban design needs to continually work at a range of scales, across different timeframes and across ownership boundaries. This can be represented by a three-axis graph where at the origin are simple projects with a single spatial scale, a single timeframe and a single land ownership. At the other end lie projects which have resulted from a range of scales, contain multiple projects with different timeframes, and often involve land under multiple ownerships. At one end of the spectrum lie estate-building projects (which are relatively easy) and at the other town-making projects (relatively hard).


The first stage in the local planning process is a district-wide issues consultation (currently known as Regulation 18) accompanied by a ‘Call for Sites’. This is an invitation for land owners to submit a site to be considered for development. A map is compiled of the submitted sites and further iterations of the local plan process invite comments on their suitability. Land owners’ agents will argue that their clients’ sites are the best for new development, while local communities will oppose development because they know what they will get; it is a hugely expensive and time-consuming exercise with little creative thinking.

Invariably sites with the fewest objections tend to be chosen and these are often poorly connected, have few constraints, lie on the boundaries of political areas, and held in single ownership. Design considerations are deferred largely until a scheme is brought forward under the development management process.

However, this way at least 50 per cent of design decisions have already been taken - consciously or unconsciously - in fixing a location for new development. The location may rely on private car ownership; if there are few surrounding street and footpath connections, then the internal street layout is likely to be based on random choices or personal whims, rather than any reasoned connections to the wider area. With little or no existing street structure to address, such sites turn inward, and with little throughmovement, struggle to attract mixed uses or establish a thriving neighbourhood centre.

This is planning by arbitrary ownership boundaries rather than town-making principles: the wrong location, not hefted to the landscape, weakly connected, and poorly informed by local history. It is little wonder then that studies of completed strategic housing sites paint such a bleak picture. Strategic sites of upwards of 3,000 houses have populations of 7,000 people or more - the size of a small town - and require a town-making approach rather than an estate building exercise. Even at that scale its future residents will need to travel for many services.

When these isolated development pods ignore the surrounding roads and streets, those streets risk being re-engineered to support free-flowing car movement. Such road schemes, whether new roads or road improvement schemes, are not often reviewed by other designers or design review panels. Instead of becoming the high streets or boulevards at the heart of new neighbourhoods, they are too often conceived as car-dominated roads joining up development pods in a local plan.


Building towns means that strategic urban design considerations are an integral part of the local plan process, and there is nothing in the current or proposed changes to the planning system that prevent this. A strategic urban design approach to the preparation of local plans would not start with a call for sites, but would first develop a shared vision for each town and follows this process:

A. District wide issues consultation - Understand the character of the town, constraints and opportunities
The underlying local landscape, ecosystems and microclimate
The historic context - its landscape and settlement history
The shape of the town, its street pattern and how this fits the underlying topography
Architecture and the materials from which the town is made
Transport infrastructure, including rail and connections to the region.

B. District-wide options consultation - Explore options for growth
Design options for growth could then be developed as alternative scenarios through engagement with local communities. This can become the basis for the local council and community to agree the direction and form of growth, creating the pattern of districts, neighbourhoods and open spaces that make up a successful village, town or city.

C. Consultation on a draft plan – Develop an agreed vision with the community
For each town, agree with local communities a vision for the physical shape and character of each town.

D. Consultation on proposed submission plan and land availability
Ascertain the availability of land identified for development. Most land with any prospect of development has already been ‘optioned’ to developers or land speculators on the basis of its future value, and so availability is unlikely to be an issue. Where a land-owner is reluctant to make land available, a choice would need to be made between a design solution or compulsory purchase.

E. Submission of Plan, examination and adoption

F. Delivery – Design codes for principal streets and public places
If urban design is about both the design of towns and making things happen, then it must help local plans to transition from policy setting to delivery mode. Simple site-specific codes can set out the street design and the massing of buildings for the principal streets and public places in the new areas.

Yet resources in local authorities are scarce, and there is pressure to deliver growth in many regions. However, a strategic urban design dimension to local plans and site allocation policies would help to raise the quality of the built environment, reduce opposition to new development, and speed up the planning and development process. A ‘whole town’ approach would deliver good urban form that will successfully function for centuries, increasing the urban capital, as well as the natural capital, for those who come after us.

Creating urban capital


Whole Town  Landscape structure | Movement structure | Neighbourhood stucture
Neighbourhood  Local movement network | Densities and uses | Street sections
Street and plot  Architecture | Public realm


Existing market town
What goes wrong
A better way to grow
The three axes with estate building at one end and town-making at the other


Good town growth is impossible with the scatter of proposed development sites that result from the call for sites process


Alternative design approach with the same street conceived as a destination