88. Testing design codes in England – 21 lessons

Following the publication of the National Model Design Code (NMDC) last year, the Government launched a six month pilot programme to test the preparation of design codes by local authority-led teams across England.  The significance of this initiative was reinforced following the inclusion in the Levelling-up and Regeneration Bill of a requirement for local planning authorities to adopt authority-wide design codes as an integral part of their development plan and “as a framework, for which subsequent detailed design codes can come forward, prepared for specific areas or sites” (para.46 explanatory notes).

Personally, I have had a long-term interest in design codes, having conducted an evaluation of a previous design codes pilot programme for the then New Labour government, eventually writing a good practice guide on the back of our learning.  Since then I have been monitoring the spread and mainstreaming of design codes across the country and more recently have even been advocating for their use following the findings of the 2020 Housing Design Audit which showed design codes to be a particularly effective mechanism for delivering design quality, when used as a site-specific tool.

Given this history, it was with delight and some degree of déjà vu that I was asked by DLUHC and PAS to lead the monitoring and evaluation of the new pilot programme which ran between April and September 2021 and involved sixteen local authorities spread across fifteen pilot teams sharing fourteen awards of £50,000 to assist them in their work.  Our report was published hot on the heels of a second phase of pilots being announced (this time 25 with much longer testing programme).  The research has much to inform the new round of work, and indeed the work of any local authority or developer considering the use of design coding as a means to deliver better design quality in their area.

The monitoring and evaluation involved a two stage interview process and analysis of the outputs produced by the pilot teams.  It covered four phases of coding – inputs, process, outputs and ‘anticipated’ impacts – with teams all starting from different places as regards relevant work already done, following different planned trajectories and drawing from different pre-existing experiences of using design codes (or not).

So what did we find?  We can break the key lessons down into 21 findings spanning the four coding phases:

Inputs into coding

Looking, first, at the inputs to coding, we start with three important findings:

1. Reinforcing design quality expectations across the scales

First, design codes were seen by all as a tangible demonstration of commitment to design quality, and came in three types:

• Site-specific design coding was used to optimise the responsiveness of development to local conditions and character
• Area-based coding was used to capture multiple smaller sites within an area of uniform character (although typically as a pre-cursor to site-specific codes)
• Authority-wide coding was seen as a means to tackle common authority-wide design problems through the production of more generic design guides.

Most of the pilots looked to code at the site-specific scale as this was where they felt the real problems lay and where they could have most impact most quickly.

2. Vision defining and vision delivery tools

Interestingly, the pilots often disrupted the usual design coding process in which codes are seen as ‘vision delivery’ tools, being prepared following and in order to implement a masterplan.  Instead, by preparing design codes in advance of such an agreed site-based vision, design codes were placed in a new ‘vision defining’ role.  This tended to result in more strategic design codes, focussing on key design principles rather than on detailed parameters for implementation.

3. It takes time, skills, resources and leadership

It goes without saying that sorting out the resources, skills, capacity, and organisational barriers to preparing design codes is critical.  Place Alliance research has revealed the challenges local authorities face in this regard, but it is difficult to underestimate the vehemence with which the pilot teams drove home these messages, namely that to move beyond the usual process of waiting until developers are in place then being led by them, requires that authorities have access to design skills and capacity.  For its part the Levelling-up Bill envisages that leadership on codes can come from developers as part of the planning application process, as well as from within authorities and even from neighbourhood planning groups (para.46 explanatory notes).

The next two findings are largely self-explanatory:

4. Getting those hooks into policy

Providing policy hooks in the local plan was seen as vital to give design codes status …

5. Getting those hooks into highways

… and somehow bringing highways authorities onboard with local design aspirations was believed to be key if coding was to be successful; although often the pilots met resistance in doing this from highways authorities wedded to their own standards.

The process of coding

Moving on to the process of coding we start with a seemingly obvious point:

6. Some places are more challenging to code, and all places are complex

There is no single one-size-fits-all coding process.  In this regard, the good news was that pilot teams had been able to successfully adapt the process recommended in the NMDC to local circumstances.  Nevertheless, some locations are inherently more complex to code than others given different delivery constraints, for example in low market value areas.  There is a need to be realistic whilst remaining positive and ambitious about design quality.

7. Defining area-types may not be necessary, characterisation is

Coming to a particularly important finding, the pilots demonstrated that the use of area-types (the division of whole local authorities into areas of distinct character as recommended in the NMDC) may not always be necessary, notably in relation to coding conducted for areas of unified or negative quality, for site-specific coding, or in relation to authority-wide guides dealing with generic principles.  Most pilots passed over this aspect of the NMDC and even authorities planning authority-wide coding tended to opt for more flexible guidance that covered their entire area and therefore avoided creating area-types.  The use of area-types at this scale clearly needs more testing.  This, however, should not obscure the need for character analysis, conducted at the most appropriate scale, typically that of the site or neighbourhood.

8. Understanding viability is key to coding

Turning to the critical input from developers, viability represents a major constraint on the mix of uses that can be supported and the mix of housing typologies the market will support, and handling developer pressure on these issues was a key concern amongst pilots.  The key lesson here is that it is better to engage developers early in the coding process and in tangible ways relating to their actual investments, rather than on abstract principles.

The next two findings deal with engaging communities:

9. Engagement is a journey

The pilots showed the value of early engagement with communities, but also that this is a time-consuming process during which trust is gradually built with communities that have often been intrinsically opposed to development.  It goes beyond simply asking what people like or dislike and at its best is a journey of education (in both directions) – from analysis, to vision, to coding and testing of design codes.

10. Moving beyond passive engagement, perhaps by mixing methods

Over-reliance on single forms of passive engagement tended to lead to lower response rates and to more basic (less informed) responses concerning community preferences.  In this respect, combining traditional and technological means of interactive engagement around issues of genuine public interest – vision-making rather technical or generic concerns – seemed to facilitate a wider and more inclusive engagement.

11. Staged coding, the fundamentals first and detailed design later

Finally, in this section, the pilots were clear that there is a hierarchy from the fundamental design qualities relating to the form, layout and use of new development that need to be prioritised early as they impact on viability, to those that are nice to influence but can be worried about later.  Some argued for a staged process, where site-specific codes follow on from authority-wide or area coding, or where detailed codes for different phases of a development build upon the principles contained in a more strategic overarching code for a large site.

Outputs from coding

Turning now to the outputs from coding, these issues span the content of codes to how they are expressed:

12. Character areas / area types can be complex and overlap

First, returning to area types or (at a smaller scale) character areas, to be useful analysis needs to reflect the fine-grained complexity, variation and constraints that characterise many urban areas.  This means that area types may overlap and mix, and so rather than seeing them as self-contained and bounded entities, more sophisticated approaches may be required to capture the different overlapping layers of character. Again, this needs further testing.

13. Balancing certainty with flexibility and creativity

Next, another seemingly obvious point relating to the need for a balance between prescription and flexibility depending on what is being coded and the context.  Issues seen as critical such as heights, quantum (density), uses, parking, dimensions for bin access, and access for pedestrians and cyclists tended to be more rigidly coded whilst aesthetic issues were treated with greater flexibility, particularly where variety and the creative interpretation of context was favoured as a design outcome.

14. Prioritise character not style

Local character was also seen as a fundamental concern, particularly for councillors.  Almost all the pilots struggled with the notion of beauty and few found it useful in either their analysis, engagement or coding.  Instead, codes tended to prioritise tangible issues such as landscape, density, height and building line as the enduring qualities of places that, It was argued, define character.

Next, two issues that we can deal with quickly:

15. Code for ‘process’ as well as ‘product’

First, it is possible to code for desirable and rigorous design process as well as for desirable design product, for example requirements that give development managers the confidence to ask that sites should be subject to character analysis and community engagement.

16. Different audiences are often compatible

Second, audiences benefit from digestible, readable, precisely worded and attractive design codes that contain enough detail to support decision-making without being needlessly bulked out with superfluous material.  In this regard the needs of different audiences – community and professional – are strongly compatible.

Related to this is the important issue of how codes are expressed:

17. Using consistent language, graphic protocols and slimming codes down

Clear language and graphics protocols help readers to understand the relative importance of different elements within codes.  Critical issues should be expressed as ‘must’ haves, meaning they are mandatory whilst ‘should’ haves are expected not advisory and ‘could’ haves are optional.  At the site-specific scale ‘must haves’ can be beneficially brought together and reflected in a framework or regulatory plan to make their application and significance crystal clear and to reduce the volume of codes.

18. Adding weight through adoption

Finally, in this section, communities were fundamentally concerned that codes should have real teeth.  For the pilots this concerned both how they were expressed, but also how they were adopted.  Adoption can occur as informal design guidance or formally as a Supplementary Planning Document or Development Plan Document.  With each step the status of the resulting code increases but at the expense of the time, resources and risk required to get through the process and the ease with which codes can later be revised.  Different pilots weighed up these factors differently.

Anticipated impacts of coding

Turning finally to the impacts of coding, in this area impacts were anticipated rather than real given the short time span of the testing programme and the absence – yet – of any actual impacts from the processes followed.

19. Turning development management into active place shaping

Fundamentally it was believed that design codes would give development managers the tools to become active place shapers, informed by a clear vision of design expectations.  Compliance checklists, performance targets (against the code) and process guidance were all seen as helpful to encourage development managers to challenge poor schemes and evaluate proposals in a proactive, timely and objective fashion. But it was also clear that development managers will need up-skilling to take on this more proactive role.  In many circumstances they will need to be assisted by those with specialist design skills either inside or outside planning authorities, particularly in relation to those elements of design codes around which greater discretion in decision-making will be required.

20. Building in review protocols

However, design codes were not viewed as ‘be-all and end-all’ tools.  Instead, changing external circumstances and the experience of using design codes should be reflected in periodic reviews of their content.  This may involve formal review processes (every two to five years) or the adoption of the sorts of staged approach to coding already mentioned.

21. Meeting aspirations, so far

Positively, pilot teams, largely proclaimed satisfaction with the tools and / or processes that had been put in place and with their potential to deliver a more certain, streamlined and quality focussed development process.  Unanimously they expressed a desire to use design coding again – resources allowing.

Which brings us back to where we began …

Many paths to coding

The coding pilots demonstrated that there are many paths to design coding and that design codes are not a single tool or process.  These diverse ends are inevitable because there are many beginnings, with local authorities all at different stages in the development of their design governance infrastructure (the fifteen pilot teams hailed from every English region, encompassed rural to city-centre contexts, different socio-economic profiles, and relied more or less heavily on consultant inputs into the process).  Positively, however, more important than the exact form design codes took was the journey coding teams embarked on to get there, and the raised commitment to design quality which that represented.

Moving forward, there was a sense that the momentum needed to be maintained, not least by being clear in policy when design codes are expected, where and when other forms of tool might be more appropriate, and who will be responsible for producing codes in the future.

Design codes have gained a new status in the English planning system and their use looks set to grow and grow.  Like any tool, there will be good design codes and bad ones and just having a code in place is no guarantee that design quality will be delivered.  They need to be well conceived, carefully crafted and consistently executed to achieve that, and all these things take time, skills and resources to secure.  However, when done well, as other research has shown, they do have the potential to deliver superior design outcomes, and in that respect the learning from this pilot programme and the next (more ambitious) phase is to be strongly welcomed.

Matthew Carmona

Professor of Planning & Urban Design

The Bartlett School of Planning, UCL

@ProfMCarmona

April 2022