What can we expect from Land Registry's new "open approach" to data?

Post: 14 January 2018

“An open approach to data is one of the three main strands of HM Land Registry’s ambition”, according to a business strategy for 2017 to 2022 published by Land Registry in late November.

Is this ambition practical, or empty rhetoric?

At the moment Land Registry is a rich source of public information about land and property in England and Wales. However none of its core datasets are available as open data. These include:

• title registers, with ownership details
  
• polygons that show the extent of registered land
  
• address-level Price Paid Data for property transactions since 1995
  
• information on Land Charges (currently being consolidated from local authorities)
  

Land Registry does publish a range of useful statistical and administrative datasets under the Open Government Licence, including transaction data and the UK House Price Index data downloads. Other Land Registry datasets, such as the recently released Commercial and Corporate Ownership Data and Overseas Companies Ownership Data, are free to use for limited purposes.

However it’s the detailed property-level datasets, currently unavailable for open re-use, that have the most potential to drive innovation and deliver analytic insights.

Land Registry has a chequered history with open data.

Price Paid Data was ostensibly open for several years, until April 2017 when Land Registry quietly updated its terms to exclude the address fields – effectively making the whole dataset unusable for most purposes without an additional paid licence.

The Information Commissioner’s Office has also cast doubt on Land Registry’s contention that residential sale prices are not personal data.

And we have the ongoing absurdity of Land Registry’s INSPIRE Index Polygons; re-usable under the Open Government Licence except for the polygons themselves.

Some of these issues are not entirely Land Registry’s fault, but in future the organisation will need to be clearer and more candid with re-users if it wants to be a trusted source of open data.

On the other hand there is some basis for optimism that Land Registry could develop a credible strategy in support of open data. Unlike Ordnance Survey, Land Registry’s business model is not dependent on commercial licensing of data.

Most of Land Registry’s income is from registration fees. Now that Land Registry has taken over responsibility for local land charges, and with a new target to achieve complete registration of all land in England and Wales by 2030, revenue-generation from registry data may be an unnecessary distraction.

There is also a growing sentiment within Government that opening up a range of land- and property-related datasets held by the public sector, including those managed by Land Registry, would help incentivise housebuilding and improve industry response to the UK’s housing crisis.

Under the best case scenario, the Geospatial Commission announced by Government in November’s Budget would negotiate open data release of both OS MasterMap (the authoritative source for land and property polygons) and a national address dataset. This would remove most of the external barriers to open publication of Land Registry’s core registry datasets.

Until then, here are a few other ideas that Land Registry could consider in support of open data:

• Release a fully open version of the Price Paid Data with UPRNs in place of the address fields. (I believe Land Registry should be able to do this under existing PSMA rules.) This would enable re-users to aggregate transactions to a range of small area geographies via ONS’s UPRN datasets (NSUL or ONSUD), and match Price Paid Data to other datasets such as Environment Agency flood data, without relying on the non-open address fields.

• Alternatively, release more detailed statistical products based on aggregation of Price Paid Data to small area geographies.

• Release the Title Number and UPRN Look Up dataset as open data. (Again I believe Land Registry should be able to do this within existing PSMA rules.) 

• Review the practice of charging £9 + VAT for flood risk indicators. The title plan itself is only £3 and the underlying flood data is already open. Land Registry should at least signpost the EA service where customers can eyeball this risk for free.
  

(For a more detailed listing of datasets held by Land Registry, see the organisation’s dataset inventory and Data.gov.uk records.)

Image credit: Cityscape by GDJ via Pixabay (CC0)