Creating a civic realm: the future of design codes

Given the way design codes are being promoted it is not surprising that there was a capacity attendance of delegates from a wide range of backgrounds at a Prince's Foundation conference on the issue in February. The six speakers were carefully chosen so as to present a wide spectrum of interests - although there were no exponents of the fundamentalist anti-coding camp as represented by the RIBA's latest pronouncement that 'we oppose the application of design codes beyond the limited circumstances of masterplanning' (Building Design 10 March 2006, p6).

The UK context

Following an introduction by James Hulme, the Acting Director of Education at the Princes Foundation, Chief Executive Hank Dittmar set the scene by outlining the recent evolution of design codes in the US. The view from across the Atlantic has informed the current official British view of coding as one means by which better quality housing can be delivered more quickly under the Sustainable Communities Plan. And in this view, 'means' is the operative word - Dittmar emphasised that a code was a tool for implementation and its use per se was not an automatic guarantee of quality. Design codes therefore need a vision, often some sort of masterplan, which they then translate into built form. They also need to be distinguished from the codes established to manage a place once it is built.

The potential of design codes

The promise of coding lies in the hope of achieving efficient implementa-tion of a predictable outcome through the minimisation of 'regulatory friction'. It is hoped that a system of codes will maintain standards of quality, respond to market choice and lead to environmental sustainability - in fact nothing less that the urban design equivalent of the philosopher's stone.

At its simplest, a design code is a set of graphic instructions for the building of a place. It should be positive in its approach to the evolution of a town or city and has been described as 'form-based' since it focuses on physical form rather than land use. In this respect, codes represent another step in the shift from the planning system, established 60 years ago and mainly concerned with land use, to a system concerned with the implications of form.

Hierarchy of Public Spaces/Streets types (images provided by John Simpson & Partners)

In their emphasis design codes focus on the public realm rather than the individual building and so differ from the pattern book, that other recent transatlantic import - although it, like codes, has a long established pedigree in Britain. Design coding can be considered as responsible for the most significant British contributions to urbanism - Georgian London, Bath and Edinburgh, all of which were speculative operations.

The use of codes in Upton, Northampton

Dittmar described the use of codes on the widely publicised English Partnerships project at Upton, Northampton, the first stage of which has been completed. This was particularly interesting since it was to figure in a later presentation by one of the house builders involved. Here the code had accompanied the planning application. It was presented in three sections:

  • masterplan principles: character areas, street hierarchy, massing and height, landscape and the urban design principles which had informed the decisions taken on all of these
  • the urban code for the main and secondary squares, the different streets in the hierarchy and lanes and courtyards
  • the architectural code which covers materials, elevations and openings and issues of sustainability and delivery - although the last two must have been of major significance in driving the decisions taken in the previous two sections.

Design codes at Upton: The Site Plan (image provided by Cornhill Estates)

How could coding work in the UK?

There followed a very useful résumé of the way design coding could work in the UK, given that any transfer of methods between contexts has to take into account the legal and planning frameworks in the receiving country. In this respect the US differs in its decentralisation of planning powers, which vary widely between jurisdictions, and a tradition of making codes that different from Britain's. The US Constitution, no less, is a code that is the foundation of that country's legal system, while the UK doesn't have any constitution at all.

Perhaps the simplest application of design coding is by a public or private landowner to regulate the activity of a number of builders. This is how it has been used at Upton and Harlow. The second application is as a contractual agreement between a house builder and a local authority. This could take the form of a Section 106 agreement. Finally it could be a local authority tool for ensuring quality in the new planning framework. This could be as part of an Action Area Plan that might have a number of masterplans for different parts of the area.

These applications are all subject to questions of enforceability, whether the code is part of a supplementary planning document or as part of a section 106 agreement or through a Local Development Order (LDO). In the last case it is intended that the design code should speed up the development process by allowing conforming applications to by pass the normal procedures. However there seems to be some doubt about how compliance might be assessed and enforced.

The transect: a tool from New Urbanism

In coding, the guidance of form is directed by a concern with specifying appropriate ranges of types for a given context, including parts of a masterplan. These are types of buildings, types of street and other public spaces and landscape elements. In this concern for what might be termed 'families of types', the concept of the transect comes into play. This is a systematic way of organising or bundling appropriate groups of types and John Massengale, an architect practising in the north east United States, elaborated on the origins of the transect and described its application.

The New Urban Transect describes the range of natural and built environments(image provided by the Prince's Foundation)

The marketing skills of the New Urbanists cannot be faulted and, like so many of their ideas, the transect has a powerful directness and simplicity which has ensured its ready adoption. It is eminently understandable and can be understood by laypersons as well as used by professionals - imagine instructing a planning committee on how to use space syntax as a tool for assessing applications.

The transect has entered the consciousness of the UK urban designer but one suspects that here it is regarded with some suspicion just because it seems so accessible and therefore must be an oversimplification of the complexity of our urban scene - as opposed to that of the US which is much newer and therefore less complex. Massengale took pains to point out that the familiar transect diagram has to be modified for each context where it is being applied. For example an urban centre (T5) in Sonoma is several thousand miles (metaphorically as well as literarily) away from a T5 in Stockholm. Each transect has to be constructed separately for each context. This must be a very salutary discipline and an excellent way of avoiding the 'one size fits all' plan which can result from the copying of policy clauses that seem to have worked for the authority next door.

The CABE pilot projects: 'a robust form of design guidance'

As a systematic testing of the applicability of coding, CABE is currently completing a pilot project that has applied design coding in a number of very different contexts. With the results due out in the next few months, Richard Simmons, Chief Executive of CABE, was able to offer an early glimpse into their findings. In the main these endorsed the views of the first speaker in that, while codes are a very useful tool for structuring a debate, they are not vision makers.

Case study and pilot locations and capacities (from Design Coding:testing its use in England, CABE, 2004)

On their own design codes will not deliver a faster development, but they can help with delivering a more certain design and development process as a 'particularly robust form of design guidance'. But since it is clear that one of the motivations for using design coding was to improve quality, is the adoption of coding likely to be self fulfilling - will those contexts, whether local authority or privately driven, that adopt coding be already predisposed to achieving a higher quality result no matter what means are chosen? Codes need to be supported by other factors including the right design skills, developers who are committed to quality, an enlightened highways authority and a 'consensus between key stakeholders concerning the vision for the site and the strategy for its implementation'. So, given a situation with all these in place, we have to ask - do we really need a code?

Mandatory or advisory elements?

The pilot experience has been resource-intensive and has revealed the need for 'robust working documents' that are user-friendly, no matter the status that was chosen for adoption. But does this not hold good for any planning document and, as for the use of codes where the priority is for clarity and certainly - shouldn't the latter qualities always be a priority?

Simmons suggests that a code needs to distinguish between mandatory requirements that must be adhered to and advisory recommendations that may be satisfied. A second distinction needs to be made between exact requirements which must be precisely established and discretionary requirements where a degree of freedom is possible.

After showing some of the work that had been done on the pilots for Hastings and Rotherham he repeated the first speaker's warning that neither the adoption or consents routes for codes had yet been established. Simmons concluded by suggesting that a common language or nomenclature needs to be developed and that the implications of Modern Methods of Construction need to inform future work.

The developer's perspective: costs and compliance

In many respects the star of the day was Kim Slowe of Cornhill Estates. Representing a family housebuilding firm constructing around 70 houses a year, he admitted that he could be much les guarded in the expression of his opinions than the director of a plc housebuilder. His firm had worked at both Poundbury and Upton. In the latter the site layout had been determined, together with the massing and mix of uses and house tenures. Cornhill had to respond to a detailed building code which had specified in great detail such elements as the brick bond (which added 60 per cent to the cost of the brickwork), roof covering and even house numbers. Compliance added 10-15 per cent to the normal building costs. An important factor in ensuring success was that the construction process was carefully monitored throughout so as to ensure compliance. This did not happen at Upton, nor is it likely to happen in those cases where design codes are administered by a local authority rather than a landowner.

Corner of Parkside and Padding Lane in Upton (provided by Cornhill Estates)

At Upton there was less control of architectural detail so that a wider variation of interpretation was possible. Slowe was critical of the way the massing and building heights, including storey heights, as specified in the code had resulted in homes which were too big for local demand. Given the small size of the average new-built home in Britain (around half the floor area of the average Danish new home) this might be something to celebrate rather than deplore. The Radio Four report on the opening of the first phase at Upton interviewed a homeowner who was especially happy about the relatively high ceilings. So, quite apart from the quality of the public spaces, the code at Upton had produced homes of a higher than average standard with respect to the private space inside the home. The implications of this in terms of buyer demand and affordability, and any impact on the creation of communities, will no doubt require further exploration.

For the small builder, a design code had the big advantage of levelling the playing field in competition with the volume house builders who usually reduced costs by deploying their standard ranges of house types. This was particularly the case when the small builder was accustomed to building high quality developments.

Codes and innovation

Slowe was concerned that codes could discourage innovation - which was the nearest we got to hearing a familiar criticism which usually emanates from architects. On the other hand he was positive about a code's ability to simplify the planning process - provided local planners had been involved in its writng. Landowners had to be realistic about the feasibility of a code and prepared to modify it if circumstances change and finally, it had to be enforced to ensure that the outcome matched the intentions.

From masterplan to code: the roles of code writer and architect

John Simpson of John Simpson and Partners used his work on the southern extension of Swindon (SDA) for 3,500 homes as a vehicle for exploring the different phases of the masterplanning and coding process.

A concern to imbue the new development with a clear sense of identity through a code was fundamental to this project. Simpson was quite clear about the division between the code writer and the architect. The code writer had to set out those parameters pertinent to the context. These were enduring and would give the expansion its public identity. The architect's task was to bring a sense of diversity and individuality to the buildings and bring to the design those 'characteristics that relate to the spirit of the age'. This is an apparent rejection of a determination control all the architectural aspects of a project, yet in some of the illustrations codes for building detail which included inter alia lintel details were included, so it was not clear how much space there was for individual interpretation. This is another issue that will no doubt require further exploration.

Swindon Southern Development Area (from Swindon Southern Development Area Master Plan Concept Document. John Simpson & Partners, 2003)

The drawings on show were seductive and certainly Simpson's work at Aylesbury appears to have successfully realised his intentions, although it was not clear how many different architects had been involved. However, one can think of other examples, notably Cambourne, where a charmingly illustrated Main Street had been hatched as big blank box supermarket sitting in a car park. Vision has to be informed by a sense of realism.

Implementing coding and the planning process

Ashley Lynch, now with EDAW but formerly a planning officer at Telford where he has been involved in implementing coding, was in unique position to tell the story as it stands today from both points of view. He provided a useful summing up of matters already touched by earlier speakers including the four options for incorporating a design code into the emerging planning process:

  • Action Area plan (AAP)
  • Supplementary Planning Guidance
  • Conditions to a planning permission
  • Local development order (LDO)

He outlined some of the difficulties that might be experienced in each case. For example an AAP can take up to two years to prepare and there is the issue of how community engagement might be incorporated into the code-making process. The SPD represents a quicker route but will enjoy less certainly in planning terms. Conditions to a planning permission are integrated into the planning system and any development would have to comply with a code but a condition could be overturned on appeal. The LDO apparently has the greatest potential as an instrument for delivering coding, but as yet it seems to be the one we know the least about.

Skills for coding: the big question

In summary it is pertinent to draw attention to one of the recurrent themes in all the presentations. Whatever framework is finally adopted for the use of design codes the issue of the availability of skills will need to be confronted. It isn't just a question of having skilled practitioners available to make the codes but (perhaps more important) suitably skilled local authority officers who will implement them.

Development control sections of planning departments are already hard pressed meeting targets and the efficacy of design codes to speed up their work, as opposed to landing them with even more, has yet to be tested.

The CABE pilot projects may give us some clues, but one suspects that testing in an actual planning department requires a time scale that is much longer than has been allowed for the outcomes of the pilots to be translated into practice.

Ivor Samuels

Additional resources

Moving From Use Based Planning to Form Based, Princes Foundation, 2006

Design Coding: testing its use in England, CABE, 2004

Swindon Southern Development Area Master plan Concept Document, John Simpson & Partners, 2003