Discovery: summer 2019
Data extraction insights
Problems with data extraction
It is often impossible to extract ‘finished’ data from local plans, either during their production or even when published. This is because:
- the plan making process is cyclical rather than linear, with policy decisions in flux rather than ever truly finished
- there are no effective systems or processes in place to extract the data, so this needs to be done manually
- no one party has complete oversight of local plans in development from start to finish, and the progress of a local plan towards publication can be delayed by a wide range of external factors, which means it’s often difficult to know what stage of development a local plan is in
- there is a lack of standardisation to ensure consistent methods and formats for production of evidence and publication
- there is an objective requirement for housing numbers when the local process for determining this is, judging from initial feedback from local authorities and examiners, a very subjective one
- authorities self publish data, which affects the quality and reliability of that data - especially given many are not data experts
We also developed a prototype to extract housing numbers from local plans as part of our discovery and to test our assumptions about collecting data at source. This revealed that:
- the web addresses of published local plans and local development scheme timetables were inconsistent and unpredictable
- housing numbers were, when they could be found, expressed in a variety of ways
- it’s difficult to know which document on a local authority’s website is most up to date
Opportunities for data extraction
We can start to build a useful picture of where a plan is in its development by looking at emerging data that can be extracted at certain established milestones in the local plan lifecycle. This data is available when:
- the local development scheme is published (it proposes both a timetable for local plan publication and what general policy areas will be covered in it)
- the local plan is opened up for formal consultation with residents and stakeholders, which gives further details about areas of policy involved
- the local plan is submitted to PINs for examination
- the local plan is formally adopted by the local authority (and published)
At the third milestone (when a plan is submitted for examination according to Regulation 24), it could be possible to create a standardised executive summary of the plan containing policy areas and commitments etc, which could help examiners in their task of assessing the local plan. This could then be revised when the plan is adopted by the local authority. However, this may be beyond the remit of this discovery, which has a specific focus on data.
Our research revealed that policymakers can’t easily extract data from local plans. The format and presence of any given data varies widely and the numbers in local plans are often derived from data that doesn’t appear in the local plan in its raw format. As a result, it’s often impossible to use local plan data to provide UK-wide information about a planning concern.
These issues prevent policymakers from easily:
- tracking the progress of local plans
- using the content or decisions recorded in local plans
It became evident early in the discovery that there is a clear tension between the making of local plans as a quick and easy step-by-step process, as opposed to a political process for building legitimacy and trust for a planning vision.
With housing numbers local authorities are being asked to forecast a planning commitment in a shifting market environment while having little control over developers to enforce housing delivery targets. Unsurprisingly, this affects the quality of data.
It can take a local authority anywhere from 2 to 5 years to produce a local plan. The longer the process goes on, the higher the risk of evidence going out of date or changes in policy affecting intended outcomes.
Regarding data capability, local authorities have limited expertise of producing and publishing data and are stretched for resources in this area – although there were a number of examples of data being used successfully.
Access to reliable and accurate data about neighboring authorities was a clear blocker to cross-authority collaboration on issues such as housing delivery.
Planning Inspectorate (PINs)
The core mandate of PINs is to arrange inspections of local plans and their soundness. It’s difficult for them to know where local plans are in their production. This is in part because:
- the local development schemes that provide information about emerging local plans are irregularly updated and often not published (local authorities are required to publish them but this policy requirement isn’t enforced)
- local authorities can delay examination at any point and once plans complete examination there is little sight of when they become adopted (as a result, the collection of emerging local plan intel is a manual process)
- there is no methodology to guide the documentation submitted and therefore the examination, which means inspectors develop bespoke assessments for each Local Plan under examination
- of the increasingly complex nature of plans (for example joint plans, scope of planning policies, statutory documents) and the consequent legal considerations further increase inspectors’ workloads – as well as the duration of examinations
The systems that PINs uses to carry out their work don’t support the extraction and collection of this data. PINs collects much of the data they need from local authorities on a manual, ad hoc basis – often through desk research. Neither PINs nor the inspectors keep track of what’s in these local plans, so they aren’t able, for example, to generate data on local plans nationally without manually investigating each time.
We arrived at 3 hypotheses based on our analysis of the discovery research.
- Better information about the status and stage of local plans in development could help:
- determine the accuracy of content
- better resourcing of local plan examination
- reduce uncertainty relating to planning permissions
- support intervention
- Better visibility and accessibility of the overarching content of emerging and existing local plans could improve:
- transparency and efficiency of planning decisions
- clarity of what local authorities are planning for and the feasibility of delivery
- opportunities for successful support or intervention
- the efficiency of the local plan examination process
- Greater consistency of part or all of the local plan content and supporting evidence could improve:
- users’ ability to aggregate and analyse content as data at scale
- collaboration across authorities and at a regional level
- efficiency of the end-to-end local plan making process
- the transparency of planning decisions
- the ability of local authorities and PINS to measure the delivery of local plans
We recommend testing the first 2 hypotheses (tracking the status of local plans and establishing what is covered by them) through the following interventions:
- setting a data standard for the timetable of local plan making which will be kept and maintained by each local authority and based on the existing local development scheme. This will help build a national picture of the plan making process.
- setting a data standard for the policy areas contained in each local plan, which will help to index policy areas covered in a local plan, which will help build a national picture of planning policy
- tracking datasets needed for each planning policy (tentative), which could help us prioritise the Digital Land work, assessing which data impacts the highest number of local planning policies
We are currently preparing to investigate these interventions as part of an alpha for prototyping, building and testing.
Other possible projects include:
- developing the geospatial element of the data alongside the Geospatial Commission
- further discovery work on housing delivery and numbers, which will help us to:
- better understand the relationship between this measurement and the housing delivery test metrics
- explore how we might monitor the actual delivery of housing from systems such as Council Tax and the address register
Principles for further investigation
We suggest using the following principles across projects:
- it isn’t necessary to overhaul or reinvent the entire process of local plan making
- where possible keep data as close to its original source as possible
- avoid adding unnecessary burdens on local authorities