Reducing invalid planning applications: a service pattern for digital planning submissions
This alpha does not follow-on from a single discovery project, but is the result of insights that have become clear during other projects by the partners and their collaborators. These include an automated digital planning guide, a digital planning submissions service and a back office planning case management system.
We now have a clear picture of the extent to which validation work and invalid applications represent a common problem for all planning authorities (see Research Overview here). We can also clearly see how the paradigm shift from documents to data has the potential to reduce and eventually even eliminate this problem. That is the main hypothesis.
The second hypothesis relates not to the solution itself, but to making that solution useful across multiple councils.
– Planning is a very content-intensive service. It includes legislation, policy, guidance and case law, at both national, regional and local scale. Some of this is common to all councils, and some is unique.
– Planning content is constantly changing. It is not realistic to rely on every council having a full-time team of web developers to customise and update all content.
– Much of the knowledge required is in the heads of planning officers, most of whom do not write code, and may be sceptical of digitising services unless they can see and control the content directly.
All this points towards the need for a planning submission service pattern that:
- Uses structured data wherever possible, instead of documents.
- Can be customised by planning officers using ‘low-code’ dynamic forms platforms.
- Is structured in a modular way (eg separating national, regional and local content), allowing updates to be collaboratively shared using a ‘git’ system.
Many of these are already features of our previous projects, but have not yet been applied to the problem of invalid planning applications.
Key research and design questions for this alpha:
- What dynamic form components are needed to cover most types of planning information?
- What are the data schemas that can capture most of the required information for most planning submissions?
- Can we design a core service ‘super pattern’ for planning submissions that covers most types of project in most contexts, allowing for the differing local information requirements? This design pattern must be able to be altered, evolved and extended.
- How can we ensure that such a content-heavy digital service is legible and navigable for users?
Every year, councils receive thousands of planning submissions, either directly via email / post or indirectly via planning portal. Each submission must be manually checked to ensure it is ‘valid’. This means checking that no required documents are missing, incomplete or incorrectly formatted.
Our understanding, based on conversations with 40+ councils, is that typically 50% or more of all applications are ‘invalid’, meaning the application needs to be resubmitted. This is hugely frustrating and costly for applicants, and also represents a major burden on councils.
Given the significance of the problem, many people are surprised to learn how mundane and avoidable most of the reasons for invalidity are. (Please refer to data on reasons for invalid applications, attached)
So far, approaches to tackling this problem have included:
– Publishing separate checklists, guides and calculators on websites. These make some, but little difference, since they rely on users wading through even more documents.
– Web forms with field validations that recognise whether a piece of information has been provided or not These are powerful, but are limited in scope because they are largely non-dynamic (one-size-fits-all), and in reality the required information will depend on the project
– Machine-learning software that automatically reads documents to check them. This is impressive, but is a very complex, laborious way around the problem, with limited reliability.
However, there is a simpler approach. The web allows us to shift from sharing information within electronic documents to sharing it directly as structured data, readable by both humans and machines.
So, wherever possible, instead of asking users to upload documents, they can complete dynamic web forms that ask them to provide key pieces of machine-readable information directly, often within predefined schemas. Using dynamic forms, users need only be asked for information that is relevant for their project, based on the information they have already given.
So instead of developing technology to check documents for errors, we can avoid the need for many documents in the first place.
For example, instead of asking applicants for a Location Plan, they can add the property boundary onto a digital map as location data.
Similarly, instead of asking users to work out the correct fee for themselves, it can be calculated automatically based on the information they have provided.
This same approach can apply to all submission forms, many planning reports, and perhaps – one day – even to scheme drawings too.
The benefits can be broadly divided into two categories.
Savings to applicants and their agents
We do not have any precise data for this, however we can make some informed estimates.
If we conservatively estimate that validation typically costs 1 day of an agent’s labour, at a £400 cost, then for the 447,934 planning applications received in England alone last year, this project could drive savings of around £179m. This will especially benefit homeowners and SMEs.
There are also indirect cost savings that are inestimable, such as the costs of project delays, disputes and lost opportunity. On average, an invalid application causes a delay of 34 days. Around 10% are withdrawn entirely.
Savings to councils
The average authority in England receives 1400 planning applications per year. (In reality this does not include some smaller types of submission, so the actual number is higher. For example, Lambeth Council receives 2100 applications per year, but around 4900 submissions.)
Each submission has to be manually validated, and typically around 50% have to be returned and re-validated, in some cases up to 9 times.
The time to validate, report, return and revalidate submissions varies, but we can estimate that validation takes, on average, 5 hours per submission (the PAS benchmark is 4). It is not unrealistic to imagine this could be reduced to 2 hours. At a typical hourly cost of £50 per hour (including overheads), then that would free-up at least £210,000 of resources per year for the average planning authority.
Across the whole of England* this represents an annual saving of £67m. Allowing for all submissions, the actual figures may be double this.
The shift from documents to data will not only reduce the burden of validation. It will also allow some automated policy compliance assessment. A huge proportion of the content of PDF planning submissions is, in reality, superfluous (eg a council’s own policies regurgitated back at them), so reducing this assessment overhead could reduce officer workload by over 5 or more hours on average per submission (£112m+ nationwide).
Again, this excludes other potential indirect economic benefits, such as increased revenue, reduced failure demand and communication overheads, improved user satisfaction and trust, training and staff retention, or cost savings from collaboration.
*We have used England as a benchmark based on the availability of statistics. But we would expect it to be adapted by councils throughout the UK.
Our objective – of course – is to keep the project as lean as possible for everyone involved. Since we feel we have a good understanding of what the project is trying to achieve, and the nature of the problem it is seeking to address, it should be possible to move quickly into design, research and development, then iteratively use the outputs to identify issues.
The partners will each assign a project leader as a point of contact (POC) for the project. Fortnightly 1 hour show-and-tells will be held. These will be held at the offices of one of the partners, and any partner representative or invitee may attend in person, but with an expectation that most will join by video call. There will also be a slack channel for partners where project updates can be shared, issues discussed and questions asked.
Testing workshops and 1-to-1 sessions with officers may also be held at partners’ offices.
As soon as possible, each partner council will have its own test instance of a platform running the service, for interested stakeholders to explore, test and improve or flag issues.
We do not anticipate the need for formal training as part of this alpha, or require explicit work from any other organisation.
However, we will seek to draw on the expertise and guidance of others in the many related departments and teams with knowledge of for example, the national requirements regarding planning information and validation, digital planning, data standards for planning and digital government services, such as the Local Digital Collaboration Unit, Digital Land, fellow-travellers such as Connected Places Catapult, and other suppliers. We may also invite the advice of relevant bodies such as the Planning Inspectorate, the Planning Advisory Service, the GLA and the RTPI.
We would also welcome all and any support in disseminating outputs to councils and their suppliers throughout the UK.
All and any interested parties from such organisations will be invited to join project show-and-tells.