In 2014, the UK government privatised its postal service after nearly 500 years of state control. Along with the Royal Mail, the government also privatised the postal address file (PAF), the dataset which contains all the postal address information of everyone in the UK, and links these to specific postcodes.
The sale was made despite calls inside and outside of government to keep it in public hands. Sir Tim Berners Lee, founder of the internet and president of the Open Data Institute, and Professor Nigel Shadbolt, the government’s open data advisor, both recommended that, rather than sell the PAF, access should be increased by making the database open to the public. These calls were ignored and the PAF data was sold together with the Royal Mail.
The data is valuable: it is used by businesses for mailer advertisements, and allows developers to link their apps and services to individual users’ postcodes, increasing their functionality. The government now has to pay to access its own data, while businesses and entrepreneurs have missed out on the opportunity to work with a vast free and open data source to develop innovative new products and services. Although the Public Administration Select Committee has since declared the sale a ‘mistake’, it remains a permanent one.
The problem extends far beyond the PAF. Much earlier privatisation, such as the 1985 Transport Act, which privatised bus services and prohibited local government from regulating bus routes and fares, has had big effects on access to data and, as a direct result, the transport system.
Privatised bus and rail services across the UK are also deregulated, meaning government has no control over how the services are run. London alone is the exception: exempted from the act, it retains the power to regulate its bus and rail services. Via Transport for London (TfL), it is consequently able to provide ample data on how people use its transport services throughout the city. Deregulated services across the rest of the country don’t release data for commercial reasons.
“At a time when there is an urgent need to improve the efficiency of public services, data analytics has huge potential”
Again, such data is incredibly valuable. From seeing how people travel, when, and where to, both government and the private sector can provide services to plug the gaps and provide new ones. Using data from TfL and transit agencies, the company Citymapper is able to tell users the best way to get across the city, via whatever mode of transport, when exactly their bus or train will arrive, and how much it will cost. While there are versions of Citymapper in Manchester and Birmingham, they don’t have the full range of features: they can’t tell a user how much their journey will cost because the data is not publicly accessible.
The measure of the problem
Beyond transport, government now outsources many other services, particularly at the local level. Since September 2008 the number of public sector employees working in local government has fallen by 28%, and spending on public services by councils has dropped by 22%. As services are outsourced, critical data that we are increasingly understanding the use of in improving services and cutting costs is lost to the private sector.
“With the advance of data analytics and machine learning tools, people are now realising some of the exciting things that can be achieved,” said Eddie Copeland, Director of Government Innovation at the innovation foundation Nesta. “At a time when there is an urgent need to improve the efficiency of public services and make them more responsive to citizens’ needs, data analytics has huge potential.”
Since Autumn 2016, Nesta’s London Office of Data Analytics (LODA) has been running a pilot project using big data to identify rogue landlords: those who rent to five or more occupants without holding a license. Licensing requires that houses for multiple occupations (HMOs) are safe and that the landlord is properly vetted. In some boroughs as few as 10% of these households are properly licensed.
“It is possible that when the original contracts were drafted, there was little understanding of what the data could be used for”
Nesta worked with 12 London boroughs to create a model to predict which HMOs might be unlicensed. By amassing multiple sets of data made available by the boroughs, such as the height and age of buildings and noise complaints, and comparing this data to the known characteristics of existing HMOs, they could put these through a machine learning algorithm to weigh the likelihood of these indicating a rogue landlord. As a horizon scanning tool, this could save local government time and money.
In this case, one borough ran into difficulty collecting its own data. “Some of the local authorities we’ve worked with in the past have encountered problems when they have been charged to access the data they require,” said Copeland. Since local governments often hire contractors to provide their services, they cede control of the data these services collect. It can be a tough job for them to get it back.
“It is possible that when the original contracts were drafted, there was little understanding of what the data could be used for, above and beyond normal reporting needs,” said Copeland. “The potential of this data has only been envisaged fairly recently.”
The ‘smart city’
While in the past, government has signed contracts that lock itself out of its own data, now some cities are playing smarter. In October 2017, Toronto and Sidewalk Labs announced a new development on the city’s quayside. Over the next few years, Sidewalk Labs will spend $50 million redeveloping the waterfront into a ‘smart city’, using a variety of Internet of Things (IoT) sensors to track the way the city’s services run in real-time and use this data to constantly improve them.
Sidewalk Labs is a subsidiary company of Alphabet, the multinational umbrella company which owns Google. The company will directly shape and help run a section of the city. Using the data people generate through the normal run of their lives to improve the way the city works and match the environment to the needs and behaviour of its citizens. This data will be open: available to the public via a set of APIs.
“If we’re looking at a future where we want to use data to improve services then it’s imperative that they can access their data”
Sidewalk has the expertise and the spending power to completely remodel Toronto’s quayside; it has the resources a city doesn’t, and the ability to lose money on an experimental venture, while most cities can’t. Yet the benefits of such a responsive approach to conventional government are clear to see. The data these APIs release is yet to be confirmed, but it raises the question of what public administrations could do with their own data.
“From the council’s point of view it’s about being aware,” said Copeland on Nesta’s pilot. “If we’re looking at a future where we want to use data to improve services then it’s imperative that they can access their data. A separate point would be to loudly and clearly tell the IT industry that it is unacceptable for them to be charging local authorities to access their own information.”
The future of urban data
There’s also the option for more radical change: governments could access the data already collected by people as they go about their everyday lives. Mobile phones collect transit data on how people get across cities far more effectively than any transport department can. Credit card data can show how people spend money, and where.
“Citizens are a major source of useful information; businesses are as well,” said Copeland. “There’s an interesting conversation to be had around what is the relationship between the public sector and the private sector. We can use this data to try to solve some of the big urban challenges that we face.”
“Public infrastructure providers do not have the same speculative interest in comprehensive data capture as the commercial tech sector”
Right now, governments do not have access to the data that people use through online services, and accessing the same data through IoT devices is expensive and complex.
“Public infrastructure providers do not have the same speculative interest in comprehensive data capture as the commercial tech sector,” said Justin O’Connor, Professor of Communications and Cultural Economy at Monash University and Mark Andrejevic, Professor and Chair, Department of Media Studies, Pomona College. “They can provide greater end-user control and transparency. They can also be reconfigured in response to public goals and priorities.”
Nesta is now proposing an entirely new solution. From early 2018, with a program named DECODE (Decentralised Citizen Owned Data Ecosystem) it will pilot two new ‘data commons’ programs in Amsterdam and Barcelona. Citizens of the two cities will be invited to give their individual data collected via the devices they own – and already handed over to commercial companies – to public projects aiming to improve city services.
Explaining the system, Copeland said “rather than hoovering up our passive data, could we put in place a tool whereby citizens are invited to donate their data to the city.”
Through such a process, city governments could perform the same kind of data analyses that Sidewalk Labs are able to do in Toronto, without the resources, allowing them to create targeted and bespoke services at a reduced cost.
“At the moment we sign up for a service, hand over all our data, and end up having no control over how it’s used. Why does this model continue to dominate? We believe it’s because there’s a notable lack of established tools offering an alternative. Part of the DECODE project is to try to create a technical infrastructure which will allow individuals more granular control over their own data.”
Getting people to share their data with the public sector is not straightforward – the DECODE pilots will provide evidence of whether it can really work.
“The aspirations to use the possibilities of smart city technologies in a way that enhances the workings of contemporary government cannot, in our view, be separated from the political work to rehabilitate the progressive role of public authorities,” said O’Connor and Andrejevic. A data commons may offer the means to do it.
(Picture credit: Flickr/iamdanw)