a blog by Owen Boswarva

17 Jan

Alasdair Rae (@undertheraedar), a lecturer at the University of Sheffield, has questioned why spatial data for England’s Green Belt is not available as “open data”.

Mr Rae has obtained Green Belt data from the Department for Communities and Local Government (CLG) for research purposes. However to re-use that data more publicly he would have to license it on commercial terms from Landmark Information Group (quoted price £35,000 + VAT).

This is similar to but separate from the problem that Andy Wightman (@andywightman) wrote about in the Guardian Datablog in September. Mr Wightman was trying to get greenspace data for Scotland, and in that case the sticking point was Ordnance Survey licensing.

I completely agree that spatial data for the Green Belt should be freely available; particularly so at a time when the National Planning Policy Framework is under review and housing pressures are high on the public agenda. Green Belt boundaries are a prime example of the type of national “information infrastructure” that Government should be unlocking for general re-use as open data.

However there are some special problems associated with the Green Belt data, and I’ve set out my interpretation of them below.

Green belts in England

A declaration of interest: I did some work for Landmark Information Group last year under contract. That work was unrelated to the Green Belt data, and this post contains only publicly available information.

Designation and review of Green Belt boundaries are the responsibility of local planning authorities, through the development plan process (subject to a national policy framework set out in Planning Policy Guidance 2). This means that anybody who wants to produce a Green Belt boundary data set for the whole of England first has to obtain Green Belt data from each of the relevant local authorities.

CLG is the Government department responsible for the national planning framework, and publishes annual statistics on the Green Belt. You would think they would therefore maintain their own national Green Belt boundary set, but apparently not. This is from a CLG report released in April 2011:

"The Department originally captured its own Green Belt data from local authority paper records. This activity was then outsourced to a private sector supplier. Considerable effort was required from Departmental staff to quality assure the data provided by the supplier. Given resource constraints and technological advancements by local authorities, many of whom are now able to produce digital map data of Green Belt boundaries, in 2007 it was concluded that a more cost effective approach would be for the Department to build and maintain a new Green Belt dataset from the increasingly available local authority digital data, thus also improving quality assurance."

That private sector supplier was Landmark. According to a House of Commons Written Answer from 2007 the previous Government funded the production of a national Green Belt boundary set by Landmark, based on digitisation of paper maps, at a cost of £32,606.25. CLG then cancelled the contract, citing the increased availability of digital data from the local authorities.

When paper maps are digitised the intellectual property rights in the paper maps themselves are unaffected. However the party that digitises them can claim new rights in the digitised versions, particularly if they then combine those versions into a larger product or database.

My educated guess is that currently the only complete national Green Belt boundary set held by CLG is the 2007 version produced for them by Landmark. This would mean either that some local authorities still do not provide CLG with digital data or that CLG have not yet collated that information into a new national boundary set of their own.

My further assumption is that under the original contract with Landmark CLG did not retain full intellectual property rights to the digitised 2007 boundaries, preventing them from licensing that version for commercial re-use themselves (and from releasing it as open data now). That left Landmark free to continue developing their own Green Belt Constraints product from the local authority source maps.

Landmark says it is “the only company that legally licenses Green Belt data for commercial business use”. That doesn’t stop other companies from making similar arrangements to obtain and commercialise the local authority data.(Exclusivity wouldn’t stand up to scrutiny under competition law or the Re-use of Public Sector Information Regulations.) However I doubt there is enough demand in the property information market for a competing Green Belt product to emerge.

While it’s clear that Landmark has subsequently continued to update their own product for commercial licensing, it’s less clear how often the product is updated. Green Belt boundaries do not change frequently. There is a version of Landmark’s Green Belt data used on DeFRA’s MAGIC website, dated July 2009.

As a business Landmark are free to set their own price for licensing of their Green Belt product. This will partially reflect the cost of procuring, digitising and collating the local authority data. However the lack of any competing product on the market puts Landmark in rather a strong position.

None of the above suggests unusual commercial behaviour. However the practical effect of the pricing barrier is that Green Belt boundaries are not widely available for re-use by the public, or by small businesses and the third sector.  

Coordination of a national Green Belt boundary set by the CLG, and its release as open data, would remedy that problem. As steps towards that objective:

  • CLG should tell us what proportion of local authorities now provide Green Belt data in a digital format, i.e. how close CLG is to having the material for a complete national boundary set.
  • CLG should seek to persuade the remaining local authorities to provide the data to them in a digital format.
  • Individual local authorities should release any existing digital versions of their Green Belt maps as open data, at least until the national boundary set is complete.

This leaves the question of why CLG didn’t strike a deal in the first place that gave them full rights to the Green Belt data captured by Landmark on its behalf in 2007, particularly if CLG staff had to “quality assure” that data. As a matter of general good practice, this is exactly the sort of complication I hope policymakers will watch out for in the future whenever maintenance of public information assets is outsourced to the private sector.