ODUG progress on a National Address Dataset

ODUG progress on a National Address Dataset

ODUG is continuing to work with stakeholders across the Open Data community to strengthen, with improved evidence and examples, the business case for an Open National Address Dataset. Two strong themes are emerging from our ongoing work:

  1. The complex Postcode Address File (PAF) licensing regime is a higher barrier to the wider use of the PAF than the actual licensing costs.

  2. There is a general lack of understanding about the importance of this data as core-reference data.

For this reason, whilst we are working to improve the detailed benefits case for the release of a National Address Dataset as Open Data under an Open Government License (which means that anyone can use and re-use this data) ODUG is asking for comment on the second recommendation made in the paper we published in November 2012 - on the future of the Postcode Address File – namely that in the event of the Royal Mail being privatised the Postcode Address File should not be placed into private ownership.

The short paper we are publishing today captures international examples which illustrate some very good reasons why national addressing data – a natural monopoly – should be owned and maintained as a public asset – so we can ensure its reliability and availability for all to use. It is simply wrong to put a single organisation, particularly one owned by private shareholders, in a position where they can command monopoly rents from an essential national asset. A natural monopoly should not be privatised!

There is very little in our society which is not underpinned by location information. Our lives are governed by time and place – address and location data is the DNA of our society. So you do not need to be an open data user or an expert in addressing data to support the proposition that our national postcode address file should not pass into private ownership. Before you make your comments please consider the following:

  • Addressing is fundamental core reference data – which underpins our society

  • This data is, by implication, a natural monopoly – the public needs to own this data

  • The data underpins property ownership, national security and the emergency services.

  • The data is essential to (amongst other things) public service delivery, many products, all physical travel infrastructure, the utilities we all depend on, all deliveries and many leisure activities.

  • It has been a widely contested dataset over many years, and is the most requested dataset here on data.gov.uk – because of its importance and wide reach.

Please let us know …. what do YOU think? …. as an individual? …. as a business? …. as an organisation? We will welcome your comments below:

Comments

National Address Dataset

The problem with the PAF and postcodes is that they are designed for the operational purposes of the Royal Mail and not as persistent identifiers of properties. This is something done by a national property gazetteer such as http://www.nlpg.org.uk/nlpg/link.htm?nwid=3988 and

 Flag as offensive 

PAF and National Address Datasets

Absolutely right - which is why the ODUG National Address Dataset Paper also calls for AddressBase and NAG to be made Open Data under OGL But it is important to remember that these are underpinned by the Postcode Address File. So we need PAF too! Heather

 Flag as offensive 

UPRN

Although the post above focuses on PAF, the original ODUG case from November also calls for open data release of OS's AddressBase product. AddressBase incorporates data from the national property gazetteer as well as PAF, and includes a persistent unique identifier (the UPRN).

Edit: Disregard; this was a response to the comment below but I see Heather has responded while I was composing.

 Flag as offensive 

Postcode Address File (PAF)

This is an important subject, and you've done a critical service to not just the open data community, but all those who use UK address data in any way. As you rightly point out, this is a natural monopoly, and by not making this open the Royal Mail is stifling innovation, and, essentially, adding a tax on users. There's also a significant danger that this monopoly could easily pass into the private sector, if and when the Royal Mail is sold, and here the Dutch and Danish examples are very useful -- thanks for surfacing this.

Of course, as other commenters have pointed out, the ultimate goal must be to have an open address register database, but this is an import first step towards this. Great work.

 Flag as offensive 

Supporting this all the way!

I am completely in support of this proposal.  PAF licensing is a nightmare for Local Authorities; it is overly complex, completely unnecessary, leads to a lot of wasted time and worst of all it prevents Local Authorities from being able to work efficiently.  If Royal Mail is privatised, and PAF is included as part of this, the situation will inevitably get worse.

An Open National Address Dataset can only be a good thing.  I can’t think of one good (or justified) reason for our national Address data to be privatised and not openly shared.  We are now in an age where open data and sharing information is proven to benefit all; privatising PAF would be taking a step backwards.

 Flag as offensive 

Sustainable..

Although not ideal for many reasons, the PAF is a core dataset which should form the basis of an open National Address Dataset. What is required is a definitive address register that also includes non-postal addresses and is maintained at source which would therefore require input from Local Government agencies. While our focus is naturally on the dataset itself, more important in my mind is the process by which it is maintained as a sustainable national resource. 

 Flag as offensive 

Let the PAF Free if the UK wants to be a Global Competitor

As an individual, I completely agree.  It is completely unacceptable that when trying to find my way to a meeting, two of the biggest mapping groups in the world, advised completely different, and one completely incorrect route (a longer walk by 17 minutes) to get to a building for a meeting in Brixton.  These problems are all based around mapping inadequacies in the UK.

As a businessmen leading the Market Development in the UK and Europe for Honest Buildings, a network for owners, occupiers and service providers, with a building profile for every building with an address...the PAF is absolutely essential for our business.  As a start-up, we do not have the funds to allocate to the very high licensing fees for this information, and our mapping and accuracy of building profiles for addresses in the UK is unfortunately not as good as it is across our other operations in the US.  Why?  Because all the mapping APIs we can link into have all sorts of difficulty even as much larger organisations...because of postcode discrepancy and the PAF.

This space has been made incredibly difficult to navigate, especially as a start-up, when there is not an open and complete dataset available for postcodes and addreses.

From an all around perspective...if the UK seriously wants to compete on a global level in the data/technology/mobile/mapping space...the PAF and other national assets, absolutely and unquestionably must be made open and available to the public.  It's as simple as that.

Nick

@nicholaskatz

 Flag as offensive 

Some thoughts on progress with the ODUG proposal

The ODUG case for an Open National Address Dataset strikes most people involved in addressing, and in particular those in local government, as something blindingly obvious, given the debacle around addressing that has taken place over the last 20+ years.  We have been subject to the three big players fighting over IPR and revenues (although only two actually have a direct part in the creation of address!).  Recent steps between Local Government and the Ordnance Survey have proven the case that given enough pressure being applied from above, collaboration is possible and I believe that AddressBase represents a massive step forward.

If a public sector PAF licence is finally agreed, it seems likely that the current drive to make the UPRN the ubiquitous reference for properties will gather momentum.  Already there has been significant uptake of AddressBase and as application and system vendors adapt the products accordingly, this already is increasing.  The future looks bright for government but what happens when government interacts with the private and commercial sectors which don't have access to the PSL?  This has been raised by Unifi Scotland in terms of property transactions.  This suggests that the anticipated PSL, whilst helping government, is not the final answer.

However, it does appear that ODUG needs to be very clear about the proposal as it seems to waver between PAF and a National Address Dataset.  These are very different things PAF is only a list of postal delivery points, formatted in a way which supports efficient delivery of mail.  A National Address Dataset is far more that this and is a gazetteer of property with their locations.  This has been pointed out elsewhere and Heather has responded to these comments but it needs to clearer as what the extent of the proposal actually covers.  It seems pedantic but makes a big difference in determining who are friends or foes.

The other obstacle that appears is the main criticism levelled at the ODUG proposal by opponents is that it is built largely upon assertions with no hard figures to support them.  Let's be very clear about this - the hard figures do exist but are guarded by the almost impregnable defences of commercial confidentiality and non disclosure agreements.  If the true costs of maintaining addressing were opened up and the current inefficient practises exposed, the true cost of making addressing open could surely be calculated.  The question for ODUG is how to break down these defences?

As ever, these are my personal views.

 Flag as offensive 

Addressing is infrastructure!

Just as no-one would seriously entertain letting a single private operator take control of our road-network, armed forces or police, so it should be with addressing. Our address data represents both a natural monopoly and a truly elemental piece of our national digital infrastructure. Could it be any more obvious?

 Flag as offensive 

National Address Dataset

Despite PAF’s limitations, postcodes are now an integral part of the process of delivery for virtually all goods and services provided by both public and private sectors in the UK. Given PAF’s great importance and the fact that the Royal Mail has already proved not to be a benevolent custodian, as evidenced by the revenue margins they have sought, it would seem extraordinarily rash to allow the dataset to drift yet further into the private sector.

If data is charged for then inevitably the licensing conditions will be significantly more complex than where it is freely available. A revenue-based model also hinders co-operation with local authorities who initiate the addressing process and maintain address data separately. There should be a single integrated process of address collection and maintenance as part of a “national information infrastructure” and the resultant data should be available to all – for free.

 Flag as offensive 

Social surveys rely on the PAF

The PAF is used hundreds of times a year to produce samples of addresses for social reseach surveys.  If privatising Royal Mail will increase the cost of using the PAF, survey contractors will pass this cost on to funders - including the government departments, universities, charities and public bodies who commission social research surveys.  The Social Research Association therefore strongly supports the ODUG's proposal to make the PAF freely available as a public benefit.  And on research quality grounds we particularly applaud the suggestion for an enhanced national address list.  The shortcomings of the PAF listings are well known in our industry.  Survey interviewers regularly find that the addresses they are given from the PAF do not correspond to the situation on the ground, for example because listed properties have in fact been demolished, new properties have not been added, single dwellings on the PAF have been divided into flats, etc.  Nor is there an accurate identification of residential as opposed to commercial addresses.  All this makes the job of survey research more difficult, and adds to costs.  An enhanced, openly available PAF would bring significant public benefits in this area.

 Flag as offensive 

Licensing is a major barrier to increasing use

ODUG is absolutely right to highlight that the PAF licensing regime is a great barrier to increasing use of the file (and consequently blocks opportunities for increasing business efficiency, and introducing new services). I hear complaints from large companies which already use PAF that they are confused by the latest licensing options, whilst many smaller businesses that could benefit must simply walk away when they have had a look at the PAF Compliance Centre website. Any government-owned body which seeks to raise revenue by charging customers, and to extract monopoly rent from different segments of the market, is bound to create similar problems of complexity

 Flag as offensive 

Public Sector User View

As I responded to the initial ODUG paper on behalf of the PSMA User Group, i'll highlight some of those comments in support here.

Regarding the PAF licensing, it's entirely true that complexity is frequently cited by our colleagues as a barrier. Understanding where, when and how to apply PAF licenses to which end products generally takes years of experience, and this seems to have completely bypassed many of the new PSMA members. Even for those who have trodden the minefield for many years, the outcome is bemusing, with a local authority (for example) likely to require multiple PAF licenses for the same data (including data they create themselves in their LLPG).

Your point regarding core reference data is another we strongly support. Regardless of the direct costs of PAF and/or Addressing Data, we believe that the inefficiencies created in the public sector through the lack of a definitive national address dataset are likely to outweight the cost of releasing as Open Data. I hope that you are able through the ODUG to demonstrate this through your Business Case, and we would be pleased to assist if we can.

I would however echo one of the comments below regarding the clarity of the message. The benefits will not come simply from free PAF licensing, but from the product that is consequently built, maintained and released as the Open National Address Dataset. The postcode is of course a useful component of that dataset, but from my point of view that is all the PAF should contribute to the Address Dataset - the added value will come from the unique identifiers, property classifications, lifecycle status and geo-reference which are not present in PAF.

Riley

 Flag as offensive 

From the Regional Chairs Address Group

Our primary issue with the Postcode Address File is that surrounding the restrictive licencing of the postcode data. The National Gazetteers create and maintain approximately 85% of the component parts that make up an address from source and pass this information to Royal Mail to receive a postcode for new addresses. Because that postcode is then incorporated into our address data and because 13 years ago, PAF was one of the datasets originally used to create the NLPG, we are constrained in our use of all National Address data through complex and multiple licence requirements. The extent of IPR that Royal Mail holds within the National Gazetteer is debatable yet individual Local Authorities do not have the resource to challenge the status quo.

All address data gathered and maintained by local authorities is carried out with resources paid for by the local authorities themselves (and ultimately our citizens) and no revenue stream is fed back to those authorities for this work. It is essentially done to support local authority service provision, assist our emergency service partners and facilitate essential business improvement, transformation and efficiency programmes. The only part of the National Gazetteer that we cannot create and maintain ourselves is the postcode and that is the only part of the National Gazetteer data collection that is costing money.  At a time when we need to reduce our costs but maintain levels of service, it would appear ludicrous that we should continue as we are in licencing data which we originate in the first instance.

We believe that privatising PAF would continue current restrictions on the use of ‘public’ address data captured and maintained for the benefit of customers and municipal organisations and continue the current position which stifles partnership, collaboration and innovation.

Stopping the privatisation of Postcode data is an essential step on the path to releasing the National Address Gazetteer from its current constraints. 

 Flag as offensive