ODUG calls on the government to deliver an Open National Address Dataset under the Open Government License

[The views expressed in this blog are those of the Open Data User Group]

Addresses and postcodes are essential in our society. Yesterday I used or provided my address to at least five organisations or people – giving my property’s Land Registry details to a prospective tenant, informing a removals company where I will need to move from and to, enabling a bank to perform a credit check on me, arranging a grocery delivery and arranging a visit from a gas service engineer. Suppose none of these people could accurately identify where my address was located? Or if they had a vague idea what area I was located in but could not easily find my home? Amazingly this is all that open free-of-charge publicly available address data currently allows.Imagine if in an emergency the emergency services did not know exactly where to go, or if the location of a potential natural disaster, like a river flooding, could only roughly be estimated. Does this sound implausible? Maybe it is less implausible than you would like to imagine.

The lack of an Open National Address Dataset has, over the last 20 years, undermined the Census, the effectiveness of emergency services, conveyancing and planning, reliable in-vehicle navigation and many other government and private sector applications.

I suspect that many of you reading this blog will be overly familiar with the colloquially named ‘address wars’ which have raged for years.  However, most citizens would be surprised to hear that their address and postcode, once compiled by publicly owned bodies into a database, becomes a commercial asset which is exploited, through a restrictive and cumbersome licensing regime, to generate what are fairly small levels of profit for public sector organisations.

Since its formation in June this year the Open Data User Group (ODUG) has received strong representations from the Open Data Community that the UK Postcode Address File (PAF) should be made free to use and freely available under the Open Government License (OGL). So we have decided to pick up the baton on national address data. We are today publishing a paper calling for the publication and release of the Postcode Address File (PAF) from Royal Mail and of Ordnance Survey AddressBase Plus, the product that delivers the National Address Gazetteer (NAG), all as open data under the Open Government License.Why is this so important? Because we know the benefits that releasing an Open National Address Dataset will bring to the UK will include improvements in innovation, growth, transparency and efficiency.

How can this be done? Well actually it’s very easy. We have been hampered in our research by a lack of transparency in the business models and operating costs of the various public organisations involved in national address data.  But, undeterred, we are taking this debate to a different level.

We’ve analysed the top line figure in the published accounts of the Royal Mail, Ordnance Survey and GeoPlace and established that these publicly owned and funded organisations can well afford, within their current operating and funding models, to release this data for free. Ordnance Survey makes a tidy 30% return on the address products it supplies through GeoPlace, on revenues of £9.6m - well above the targets set for it by HM Treasury. It reported an overall operating surplus of £32m in 2011-12.  Local Authorities and other public sector organisations already get Ordnance Survey address products free, funded through the £55m per anum Public Sector Mapping Agreement (PSMA). Why shouldn’t the rest of the economy benefit from addresses as well?

Meanwhile Royal Mail has a licensing regime in place for the Postcode Address File (PAF) which is cumbersome for potential users of this data and also costly to run and administer. The organisation spends £24.5m in operating costs to maintain the PAF, making only £2.6m profit its licensing activities. The revenue this activity delivers is only 0.3% of overall revenues. We think Royal Mail can well afford to release this data for free, considering that it is a £1.5bn organisation. In my personal view the Royal Mail should focus on its building its core business, recently returned to profitability, and give up its’ aspirations to further commercialise the monopoly position it has over a national asset which is based on crown copyright data.Also, if the Royal Mail is to be privatised we certainly would not want to see the Postcode Address File belonging to a non-publicly owned organisation. The Dutch made this mistake and then had to fight a lengthy EU court case and eventually buy back their ‘key-register’ of addresses and buildings, including postcodes, which they immediately made available for re-use without licensing conditions.

But it’s not all about the price tag. In our paper we have set out the various classes of benefit an Open National Address Dataset will deliver. Many of these benefits are difficult to quantify but we have some strong indicators of the real potential value of this data:In 2010 the Danish published a report indicating that the direct financial benefits of their free-of-charge address data will be around EUR 14 million, 70% of which will accrue to the private sector, with costs of distribution at only EUR 0.2 million.

Our own PAF Advisory Board’s 2011 study into the Economic Value of PAF estimated the value of PAF data to the UK economy at between £992m and £1.38bn per annum.Meanwhile compiling an adequate list of dwellings (National Address Register) for the 2011 Census cost £12m as a one-off activity.There would also be vast cost savings to be realised, not to mention service efficiencies, if all public sector bodies operated from a single National Address Register.The Open Data User Group believes that it is only by releasing a definitive National Address Dataset as an open data product that inefficiencies and lost opportunities will be overcome. The use of correct addresses will be maximised in the national interest. It believes that releasing addresses as open data will help individuals grow their own businesses, and also benefit the wider economy.

The potential for £1.38bn of annual economic value balanced against the few £m of profit currently realised by the Royal Mail and Ordnance Survey would appear, to me, to be a no-brainer economic decision!

See the papers related to this blog:

http://data.gov.uk/library/the-case-for-an-open-national-address-dataset-november-2012

http://data.gov.uk/library/annexes-to-the-case-for-an-open-national-address-dataset-november-2012

http://data.gov.uk/library/pafab-consultation-odug-response-30-oct-2012

Heather SavoryChair, ODUG

Comments

Link broken?

Hi, I think the first of the three links above might be broken - I'm getting an "access denied" error from the URL http://data.gov.uk/library/the-case-for-an-open-national-address-dataset-november-2012 Thanks!

 Flag as offensive 

Links

These are now updated.

 Flag as offensive 

Superb report

Brilliant report - well written, cogent arguments and sensible conclusions. It is about time that the wastefulness of competing public sector address databases is sorted out.

 Flag as offensive 

Great work by ODUG

Congratulations to the Open Data User Group on picking address files as its first target, and swiftly producing excellent reports on both the National Address Gazetteer and PAF. It’s been a scandal for more than a decade that public bodies have sought to exploit their own monopolies, rather than co-operating to produce a definitive file of addresses and postcodes that can be used freely by everyone.Francis Maude described this as “unfinished business” more than a year ago, and it is to be hoped that ODUG’s pressure will finally convince those in BIS and the Treasury who continue to defend the status quo.

Keith Dugmore, Demographics User Group

 Flag as offensive 

Why the obsession with PAF??

Why the obsession with PAF?? It is only a list of mail delivery points and bares no resemblance to the "official" LA defined street name or number. If you want a definitive address that is free to use then bin PAF completely and use the OS AddressBase product which is delivered by Geoplace which in turn is updated & maintained weekly/monthly by every LA in Britain (who afterall are the legally defined street naming and numbering authority). Whilst we still have 2 different and completing (even if they were free) sources of national address data there will never be any clarity or standardisation for commercial/public bodies and we will never achieve any of the benefits that you are striving for.

 Flag as offensive 

PAF is required...

Hi Will

PAF is required as part of all of this as without PAF, you have no Postcodes. PAF is the only place where a postcode can be created. If there was no reliance on Postcodes then yes we could bin PAF but as it underpins so much it's impossible to do without it. Whatever national address product that is created will have to be a merger of PAF, AddressPoint and the NLPG - This is what AddressBase is. By October 2014 there will technically only be the one addressing product, and that is AddressBase.

Steve

 Flag as offensive 

Addressing

Very well argued. Terrilel reflection on our society that the  obvious need so much effort and time spent on it.

 Flag as offensive 

Support for the case

Having read the detailed papers released by ODUG I would like to thank and congratulate the group for articulating the minefield of issues around addressing in the UK superbly. It is clear that a lot of work has gone into the production of this paper, and given that ODUG has only existed for a short time, this is a very impressive turnaround. Credit too for picking the single most useful dataset which could be made Open.

I am in the fortunate (?) position of representing both a creator of address data (a local authority) and a group of users (PSMA members), and so have different angles of insight into this debate. Despite this, I see no conflict with any of the arguments your paper makes from either side – indeed I hope that the representative groups will provide further support and evidence for the case of open address data.

As a creator of address data through a local authority LLPG, I would love to see the public finally realise the full benefits of the high quality address and geo-location data they fund. For too long it has been restricted by a) the complex licensing and IPR framework of Royal Mail, OS and government departments (LGA subsidiaries) and b) a desire to profiteer from the work of local authorities which I find bizarre.

Your paper refers throughout to the costs of OS and Geoplace in compiling the NAG, but these are incidental when set alongside the ongoing cost borne by over 400 local authorities in maintaining LLPG’s (the constituent parts of the NAG product which Geoplace markets). A conservative estimate of the costs to my local authority of maintaining our LLPG is around £50k p/a – I don’t know how typical this is in relation to size, but I would be very surprised if the aggregate cost across local government were less than £20 million. All Geoplace does is compile this data and license/sell it.

There has been no financial return to local authorities for contributing this data – we do it because the internal efficiencies from creating data once and using many times across the organisation offset the costs. These internal benefits remain significantly constrained by the lack of a definitive national address dataset, and I will provide further evidence of this in response to your paper.

If I have one criticism of the paper it is that, although your recommendations are clear, references to ‘Open PAF’ and the NAG intimate that the release of these products separately would suffice – I would argue that it would not. To deliver the benefits your paper describes, a single, definitive database is the only suitable output, with the following responsibilities:

  1. Address definition and creation in database – Local Authorities
  2. Appending a postcode to created address (without variation of address)– Royal Mail
  3. Geo-referencing the address – Local Authorities using PSMA licensed data (not Ordnance Survey as indicated in the section of your paper which estimates a cost model)
  4. Compiling and distributing a national dataset – Geoplace, using existing hub infrastructure. I would also recommend an independent review of current and planned developments to this infrastructure to ensure they are delivered at an absolute minimum cost if they are to be subsidised from any of the proposed methods (e.g. PSMA settlement, levy at point of change)

A couple of final suggestions:

  1. Do Royal Mail even need to retain an AMU at all? Couldn’t local authorities apply the rather basic logical of postcode creation at source?
  2. The geo-reference of the NAG is very useful, but it is largely based on a point depicting a part of the building concerned. Perhaps the Land Registry, if linking their records to the open dataset, might add value by providing the extents?

Thanks again for a job well done, and I hope your recommendations stimulate a long overdue ‘national debate.’

 Flag as offensive 

Bury's Comments

Definitely agree with Riley's comments about Royal Mail not varying the address, they do this all the time and add localaties that aren't even in our authority's boundary, they even change the Geographical Town (not the postal town) just because at some stage they decided a particular out part of a post code was in for example Whitefield when in fact it's in Radcliffe. This causes endless of problems. I have challengened them a few times but alwways get the reply they are allowed to do that. I would like to see the Council's Approved address used by everyone including Royal Mail as it has been created with no change apart from the postcode. Disagree though about no financial returns, suring the fact that we get the mapping data supplied and paid for by Central Governmentg is a sort of financial return.

 Flag as offensive 

A great step forward

This represents real progress.

If the public get free access to a quality product such as the National Address Gazateer, then it will be a real feather in the cap for the Open Data movement.

Well done to ODUG for highlighting the issue and moving so swiftly.

 Flag as offensive 

A compelling case for open address information

This was a very detailed, well written report which clearly identifies the reasons why a single source of address information is not only desirable, but a necessity for UK plc.  Well done ODUG.

 Flag as offensive 

Great Paper

Brilliantly argued case, that I struggle to see how the existing vested interests can reasonably repudiate.

Can I ask that in your ongoing discussion you remember to push for the inclusion of Northern Irish addressing and positional data in the comprehensive UK dataset? It is infuriating that OSNI and the Pointer dataset are not even included in the current OpenData suite. Many (nearly all) the projects I work on are for national organisations where UK wide data is needed. Currently we either have to just accept the OSNI terms and high prices or try and cobble things together from other sources that are not ideal.

Riley makes some good points and I can well imagine how the Local Authorities feel hard done by. In defence of the Royal Mail they do need to maintain a postal services friendly address register which places slightly different demands on the data and the assignation of postcodes is non-trivial as they are used in the creation of the 'postal walks' that are used to manage the posties routes, and the vagaries of naming streets and addressable objects can lead to legitamate occasions when addresses need to be re-formated. Also it would seem to me to be a missed opportunity not to make use of the tens of thousands of posties who walk the streets every day to ensure the data are accurate, up to date and comprehensive. For example understanding when an address is 'defunct' is very hard for someone like an LA or the OS to figure out, whereas a postie walks past the property every morning. I seem to recall the last time the RM did a large scale clear out of redundant addresses they found several hundred thousand addresses that were on the registers but were no longer used - largely rural I believe.

Finally it is important that the business addresses from VOA and the Scottish Valuers should be included in the discussion. Most PAF business address make use of VOA premises name and they are, frankly a bit of a mess. This is an example of where the data does evolve very quickly as businesses close, move and open. If we are going to put business names into formal addresses we need to make a better fist of it.

 Flag as offensive 

Scottish Data

The case here is compelling.

I would add that we should also lobby the Scottish Government, and the Improvement Service in Scotland, to make the One Scotland Gazetteer data available as Open Data too.

http://www.onescotlandgazetteer.org.uk/

It is built from the local address Gazetteers of all 32 Scottish Local Authorities.

Ian

PS On submitting this post I received the message that it had triggered spam fileters and the link was to a spam site - despite it being an official Scottish Government site!

 Flag as offensive 

Great

Well written indeed!

 Flag as offensive 

A worthy initiative

We at Beacon Dodsworth feel that this is a worthy initiative.

We have written a blog on the subject for interested parties.

 Flag as offensive 

Address Master Data

The paper is another useful contribution to the debate that has raged since digital mapping began in earnest.There are ways to proceed without banging our heads on institutional brick walls as we have done for the last quarter of a century! These data are collected freely by anyone who has access to a GPS enabled mobile phone and is the basis of much data collection in the developing world. There are two issues at stake. The first is open access to Address data and the second is the maintenance of the data. A single address file is unlikely to achieve what people think. The electricity sub-station or the signal box on the rail network do not have mail type addresses to BS7666 standard, yet they are still valuable locations to reach. These locations are part of business functions that will need to carry on irrespective of whether address data is free at the point of use. The postcode has done a lot of damage to locational analysis since it was introduced - re-assignment and variable numbers of proerties in a code mean that it does not satisfy the master data test. Haggett described better aggregation models in the 1960s than the postcode. Further, it is pleasing that the Post Office is moving to individual GPS co-ordinates for properties it delivers mail to, under the Pinpoint initative. So a national dataset which is based on registry principles (mine whatever data sets you can), free at the point of use, follows the Digital National Framework principles would be a valuable step forward.

 Flag as offensive