Don’t sell our postcodes! Why we should have Open Addressing in the UK
Open Data User Group Response to the Ofcom Postcode Address File (PAF) Consultation
ODUG’s response to the Ofcom PAF Consultation is published here. Our primary feedback is that this consultation simply considers Business–as-usual for the Postcode Address File (PAF) – and therefore assumes that increased revenues should be derived from the Royal Mail‘s current monopoly position – to increase Royal Mail’s profits. Please read the full response for more detail.
What we are saying here is don’t sell our postcodes. This has nothing to do with how much junk mail you get through your letterbox and the data does not include people’s names. It’s about whether we have a central, high quality, publicly available set of address data for the country – which is owned by the public, not by private organisations and available to anyone as Open Data for use and re-use.
We would welcome your views in comments below and, if you are interested in the detail, please read on …..
Some eyebrows have been raised at the Open Data User Group’s decision to highlight the need for an Open National Address Dataset as their first target. Surely addresses aren’t a problem? Write one on an envelope, it usually gets there. Pop a postcode in a sat-nav and it guides you to a place near your destination (except when it doesn’t and you end up in a field). What is the problem?
The Open Data White Paper which announced the setting up of ODUG specifically mentioned a national address dataset as one of the matters that government wants to resolve. This is because addresses and addressing have been a contentious area in the data world for more than a decade.
So what is there to argue about? As is usually the case it is mainly money; who is entitled to make money from addresses and how much. Who owns addresses and therefore the revenue? These are all publicly owned organisations in play here by the way – your taxes are funding them, public assets are being traded and there is a capped return going back into the Treasury (who appear to discount the wider economic case which they should consider if they were to refer to the ‘Green Book’ which is the economic ‘bible’ they should use to make public spending decisions for the good of the nation).
The real issue here is that as publicly funded organisations struggle to balance budgets, access to a nice little natural monopoly (theoretically there can only be ONE correct set of addresses and their locations) in the provision of data that everyone needs, particularly if it is lightly regulated, can lead to a permanent highly lucrative revenue stream.
This is good, easy, business ….. Imagine if every time someone purchased something on the internet and entered their address, or used a sat-nav, or needed to verify an address list you could charge a toll, which doesn’t need to be related tightly to your costs, that’s a real ‘nice little earner’. For this reason a long standing set of fierce arguments, colloquially known as the ‘address wars’ have raged for years between Local Authorities, who legally ‘create’ addresses, Royal Mail, who add a postcode and modify addresses to reflect the delivery network and Ordnance Survey who map where the address is; constant squabbles have persisted over who can leverage the Intellectual Property Rights (even though this is all publicly funded Crown Copyright data) and hence the revenue from addresses .
Royal Mail, because they compiled the first comprehensive list of postal delivery addresses, the PAF, and started selling it in the 1980s, claim first mover advantage and demand payment from anyone who uses or sells data that includes addresses with postcodes. Local Government have tried to assert their moral ownership of addresses, based on their statutory addressing powers and Ordnance Survey, who were first to map all postal addresses, want their slice of the cash whenever an address is placed on the map, not just their own maps – any map!
So there is an unseemly tussle over which of these organisations can put up metaphorical “toll gates” on the physical information highway and claim a payment every time an address is used.
A contrary view, which ODUG and others hold, is that definitive addresses are a natural monopoly; a core-reference dataset which is lubricant to ensuring efficiency and accuracy in record keeping, location and deliveries. We believe addresses are a true, non-rivalrous (anyone can use it without limiting other users), non-excludable (anyone can get hold of it) public good. It is in the public interest that as many organisations and individuals use correct addresses as easily as possible and the best way to ensure that is not to charge for an address at the point of use.
The internet would grind to a halt if everyone had to pay a lookup charge every time a web site address or an email address was converted to the numeric address used internally on the internet. For this reason internet addressing databases are freely shared, in real time and free at the point of use. Those who register an address pay, as do those using an internet service, but not for looking up an address.
All internet users have free access to the address data under the Domain Naming System (DNS). We believe that the same should be the case for postal and geographical addresses.
The reason this is so important now is that Royal Mail may soon be put up for sale, in full or in part, and ministers will need to decide whether to allow them to take the national Postcode Address File (PAF) into private ownership. When the Dutch government sold their postal address file with their equivalent of Royal Mail to a private owner they soon came to regret their decision and had to fight in the courts to buy their national address file back at a reasonable price. We do not want the same error to be made here in the U.K.
It looks very likely that ministers will take a final decision on the future of the Postcode Address File before the Summer Recess.
ODUG believes that the decision will come down to one of three choices:
Option 1 -‘Business-as-usual’
PAF ownership is rolled in with the sale of Royal Mail and becomes a private asset lightly regulated, as at present, by Ofcom (the regulator for Royal Mail). The current Ofcom review assumes Business as usual and proposes that in this case PAF licensing restrictions should be made simpler – we think they should be removed. Looking at the business-as-usual scenario Ofcom also thinks that the cap on the profits Royal Mail is allowed to make on PAF should be lifted. This is because their current remit is to allow Royal Mail to make as much (short-term) profit as possible – regardless of the wider economic benefits, or dis-benefits of this approach in the medium to long term. Ofcom views Royal Mail through a lens which does not acknowledge privately or publicly that the relatively small profit Royal Mail currently gets from PAF should not be material to the returns one might expect from a huge organisation whose core business is in delivering post! This option will give Royal Mail a permanent (or, at least, difficult to reverse) right to toll all transactions that involve an address and essentially privatises a natural, publicly funded, monopoly dataset to the detriment of us all.
Option 2 - ‘Monopoly rents elsewhere’
The PAF maintenance monopoly could somehow be transferred to GeoPlace LLP (the commercial partnership between the Local Government Association and Ordnance Survey), creating a consolidated monopoly for commercial exploitation by a body/bodies which, at present, are not fully publicly accountable and which could follow Royal Mail into private hands (although we believe this is less likely).
Option 3 - An Open National Address Dataset
PAF maintenance gets rolled into the GeoPlace National Address Gazetteer (Royal Mail) contributing as a regulatory obligation in support of it Universal Postal Service provision or, put more simply, the requirement it has to deliver post to us all. A new business model is produced for GeoPlace (or, following a public procurement, another contractor) to maintain the National Address Gazetteer including PAF as a free at the point of use. An Open Data product is delivered under the Open Government License. This can be funded out of the government’s existing financial commitments to addressing from funds currently committed, long-term, via the Public Sector Mapping Agreement and the very recently announced Public Sector PAF Licence. These funds could be made up with a small charge on address registration fees which are already levied by many individual Local Authorities for naming and numbering.
ODUG believes that Option 3 best serves the national interest and understands that this view is shared by the Open Data Institute, the Demographics User Group and many others. ODUG is concerned that the Shareholder Executive (the agency that holds Royal Mail shares on behalf of the public) and the Treasury, appear to be lobbying hard for Option 1 because they see this as giving immediate shareholder return in the price paid for Royal Mail. Reversing Option 1 – if this was the government’s decision would be extremely hard. The Public Data Group (which groups together Trading Funds including Ordnance Survey) is likely to favour Option 2.
We strongly believe that Option 3 is in the national interest and consistent with the National Information Framework produced by another government advisory body, APPSI (the Advisory Panel on Public Sector Information) and the overriding principles set out for the re-use of Public Sector Information.
We also do not believe that bundling in ownership of PAF with Royal Mail will make a material difference to the sale price of the business. On the contrary, a potential investor is likely to view the persistent debate about the ownership and future of PAF as a risk factor which will deter them from investing in the Royal Mail. Offering PAF as part of the deal will not solve their concerns if they fully understand the wider issues at stake.
This is why our response to Ofcom asks them to evaluate and consider our preferred option, which is for PAF to be turned into Open Data now. This can be followed by consolidating address management so that the nation has a single, properly maintained, fit for purpose, definitive Open National Address dataset in the future which is free for all to use to promote the growth of the economy and to support national efficiency.
The Ofcom consultation remains open until 21st of March so there is an opportunity to express your views there as well. (http://stakeholders.ofcom.org.uk/consultations/postcode-address-file/)
10 comments
Comment by StuMitch posted on
I think, like most people, that this is an immensely sensible idea.
I would suggest there is a stronger business case though, or one that you've mentioned, but needs strengthening, and that is: the true cost of badly addressed mail. Difficult to calculate, but potentially worthwhile to estimate. The cost of incorrect or incomplete addresses must be greater than that of the value of the data to Royal Mail. If Royal Mail is truly a public service, even post-privatisation, then I would say that this data set needs to be Open.
Comment by Ian AD posted on
It should be recognised that if the outcome is that the current post code system continues to be closed and run by a monopoly then it will set exactly the right conditions for 2 kids in a garage to invent and implement an new better postcode which is based on open access.
The ordnance survey has suffered immensely from this closed commercial model. Not long ago the OS was a world leader in maps and their data was the foundation of most mapping. However they are now almost invisible, few people use their products compared to the vast usage of Google tools and open streetmap data.
By making postcodes open it will make everything run more smoothly, if they are not open then in a few years time they will be obsolete.
Comment by exstat posted on
The issue is not whether postcodes are open - anyone can look up a postcode on the Royal Mail's Postcode Finder for free - but whether they are available in bulk, along with the various geographical data associated with them - are available free without significant licensing restrictions. The link (or lack of it) with the data produced by authorities responsible for addresses is also a major issue. "Addresses" go beyond mail delivery points, while some postcodes are non-geographical (e g for PO Box numbers, I think) so the relationship is not a nice simple many-to-one thing.
Two kids in a garage might invent something but how they would tie it to individual address I do not know (I know - let's invent a "grid square"). Then they would get bored and the thing would get out-of-date but still be praised as brilliant because it was open! (I think this could become a more general problem, as "nice ideas" get left on the internet with nothing to warn that they are five years out of date. There are many dedicated open data users out there, such as OpenStreetMap, for which this will never be the case, but "open" means open to the undedicated as well.) And of course thgeir verison wouldm have no real world use external to itself. Postcodes were invented to help Royal Mail do its job more efficiently. Its role as a cornerstone of geographical analysis is sernediptiy really, but its real world use means it HAS to be maintained and will not become obsolete unless something better comes up that meets Royal Mail's internal needs.
Comment by sfgreenwood posted on
The raw postcode database should be open to provide the greatest benefit to the greatest number of people and has to be secured soon to ensure that it doesn't get handed over to whoever buys the Royal Mail when it is inevitably privatised. This wouldn't prevent the Royal Mail or its successors from providing lookup services but would take away its exclusivity.
Comment by iaincollins posted on
The continued attempts have the offical data as commercial pay-for data are harmful to business large and small, by restructing their ability to innovate.
I've worked on projects internally at large UK companies and on open projects for fun and have repeatedly been hamped in both by the effort required to cobble together a "best effort" set of data from the very limited and near-uselessly formatted data currently (& seemingly begrugingly) made avalible for general use and commercial open platforms from providers like Google and Yahoo!.
Postcode and full address data in a complete, up-to-date and easily consumable format (e.g. using common geographic coordinates - not the anachronistic OS grid) should be available to everyone, without any barriers.
As noted this isn't about names and addresses or spam, it's just addresses and postcodes - use for everyone from food delivery companies, to GPS manufactuers to companies that sell goods and services online or over the phone, or indeed any company that has any form of delivery offering. It includes groups as diverse as estate agents, local service providers such as locksmiths & glaziers and a myriad of new types of online businesses.
Even those business that don't consume this data directly, depend on tools and services that do (whether that's a GPS, their phone or a website to lookup where they are going).
Thousands of missed (and ultimately repeated and costly) deliveries and appointments every day could be avoided if we all had access to an open and authoritative set of address data.
It's incalcuably harmful to innovation and prosperity for something so fundamental not to be open and freely avalible to all.
Comment by kristina posted on
I am working with similar data in Sweden and find this subject very interesting and am trying to compare with the situation we have in Sweden. I have a question: does PAF contain any kind of geographic coordinates, or is it simply a database of existing combinations of postcodes/cities/streets/street numbers etc?
Comment by Paul Malyon posted on
Hi Kristina,
PAF doesn't contain any geolocations but other files from the Ordnance Survey (such as AddressBase or CodePoint Open do).
Comment by michaelg posted on
I don't understand why we are wasting time consulting on common sense. Access to the PAF database should be made available free of charge. Please don't let RM fool you with arguments such as security, data protection, and misuse - these are all designed to help RM keep control of the PAF database for commercial gains. I hope the government backs "Option 3" before RM is sold to a private organisation.
Comment by guydevere posted on
Post Codes are a millstone from a past era, are fundamentally flawed & were the best solution to maximise the efficiency of a monopoly postal supplier 50 years ago. They should be sold to the highest bidder for the benefit of the tax payer before the value plummets when it’s made redundant by a 21st century location intermediary system that operates like the internet for the benefit of the consumer.
Post Codes have a fundamental flaw in that they are permanently tied to a location/building. “That’s the whole point” I hear you cry, but hear me out. What most of us actually need is a permanent identifier that is tied to a person or organisation (& not to the location) that we can use when we need to locate them. Even Royal Mail partially recognised this by having PO Box numbers that allow us to send someone something without knowing where they are located. It also allows them to move location without telling everyone their new Post Code since Royal Mail have a look up table to link the PO Box to a specific location. Only organisations who are interested in buildings regardless of the people living or working there need a building Post Code. Most consumers & organisations want a permanent means of finding a person or organisation, without needing to be informed every time they move location & change “their” Post Code. Of course when people move house they realise it’s not actually “their” Post Code at all, & it’s no different to the gas supplier’s customer account number that they assign to you for their benefit.
The internet’s location intermediary is the Domain Name Service (DNS) which looks up the data.gov.uk domain when someone sends an email to anything@data.gov.uk & finds the unique computer location (IP address) to deliver it to. If data.gov.uk moves location to another computer it doesn’t need to tell everyone who sends it email that their address has changed.
Why can’t my various types of physical locations be given a globally unique, permanent & personal identifier that I retain even when I move house in the same way as I retain my email or mobile number when I change email or mobile network supplier? If someone needs to know where my house is located they can look up the location identifier that I give them & it will respond with the appropriate location from a list of alternatives that I’ve defined based on what they want it for & how much & for how long I trust them. So if they are visiting me at my house they’ll get a lat/long that their GPS can use to navigate, if they need to deliver a parcel at a time my calendar knows I’ll be unavailable they will receive the location of my corner shop who I’ve authorised to receive goods on my behalf, if it’s a letter & my calendar knows I’m on secondment to Darwin for 3 months it will respond with the location of my scanning service who I’ve authorised to scan mail to my cloud storage (but only for the period when I’m away), & if they’re delivering a passport to me in the airport the hour before my flight leaves then they’ll get access to my dynamic location for 30mins.
The competing intermediary lookup services (like telephone directory enquiries) could provide different value added information such as estimated time to drive to the location in current traffic, a photo of the location, the access facilities such as stairs/lifts/ramps, the opening times, etc, & if you’re a historian researching how monopoly postal suppliers used to own & control consumer location data then I’m sure there’ll be a service that will provide the quaint old Post Code too.
Global Email providers like Google could set up a service to allow you to define addresses on a map (lat/long) for home, work, delivery etc so people could write your Google email address on an envelope & whichever postal supplier they used (anywhere in the world) would look up the lat/long for delivery using the same sorting machines that currently lookup Post Codes & sort the mail. If the supplier wants to add their own codes to the envelope for efficiency then that's up to them, & junk mail would finally be under consumer control.
So do you still really want to keep Post Codes alive with Option 3 or do you think consumers should "own" their location data?
So please abandon this attempt to preserve Post Codes in the public domain since it will actually stifle innovation whilst depriving the taxpayer of revenue & preventing the UK from spawning some modern solutions that the rest of the world would rapidly follow.
Comment by michaelg posted on
I still want Option 3.
Postcodes are great for fixed assets (i.e. properties) because they can't be easily moved - this is why online merchants don't deliver goods to postbox numbers. Postboxes are like personal email addresses - not tied to any physical assets, so you can move around and keep the same postbox number.
Your suggestion to have a Global Email provider controlling email-to-address lookup database is NOT a great idea nor is it innovative. Who will own the database? how would we ensure its kept up-to-date? Who will validate these addresses?
With your idea, the final destination (i.e. physical address) would still need to be verified and validated by someone or something (maybe a postcode?).
I say keep the postcode! but make it free!
Sorry G - its back to the drawing board with this one 😉 The postcode system is already and remains innovative (hats off to the original creators).