In response to the Government’s request for ideas for the UK’s Digital Strategy, I offer the following recommendations for Section 2 – Transforming Government.
Public sector digital transformation fundamentally depends on making smarter use of data. Therefore:
1 – Data.gov.uk should be converted into a data marketplace where different creators and consumers of data can buy, sell, request or freely exchange data.
The UK’s approach to open data has a major shortfall: the flow of information is in just one direction. Businesses, charities and citizens can receive data from government. But there is no official, standardised and automated mechanism to provide data to government. This is a missed opportunity when much of the most valuable information that could help reform public services lies outside government (consider the data that retailers, electronic payments companies or mobile phone operators have about the way citizens interact with cities).
The solution is to create an online data marketplace that connects organisations and individuals that have useful data with those that want it. Similar to how sites like Odesk.com or TaskRabbit.com connect those offering their skills for hire with those that need jobs done, the data marketplace would bring together creators and consumers of data.
The site would have two types of listings:
Creators: Individuals or organisations with datasets could advertise them on the site, specifying the quality, format, range and scope of the data, together with the price (if any) they would charge for providing them. Government could continue to use the same marketplace to provide its open data free of charge.
Consumers: Those needing data – whether central government, a local authority, a business, charity or individual – could advertise their data needs. Creators could then see that request and publicly bid to provide the data, specifying how they would source it, its quality and so on. Consumers would state whether they wanted to make the data open or if it was solely for their private use.
The marketplace would help unlock, correctly price, and stimulate demand for data held by citizens, businesses and other organisations both within and outside the public sector. It could be used to crowdsource facts, figures, images, map coordinates, text – anything that can be collected as data. Instead of simply publishing its own datasets, government could outline the problems it was trying to solve, leaving the market to find the best data for addressing them.
2 – As a pre-condition of receiving devolved powers, city regions wishing to benefit from City Devolution should be required to establish an Office of Data Analytics (ODA), led by a Chief Analytics Officer, reporting directly to the mayor.
The new generation of elected city mayors will need the tools to match their new powers. Many of those tools will involve harnessing data. Cities will need to be able to join up, analyse and act upon data from across their whole city regions to identify the scale and location of the problems they seek to tackle and the service demand they aim to meet. Data will be needed to reform public services. And data will be required to monitor – and to report back to central government – the results achieved with devolved powers.
Currently, most cities lack any such capabilities.
In 2016, the Government should therefore ensure that an integral part of its negotiations with each city region concerns how they will put in place an ODA – similar to the Mayor’s Office of Data Analytics pioneered in New York. Each ODA should be tasked with helping the mayor and the city region’s public sector bodies bring together and analyse data to deliver regional economic growth and local public sector reform.
3 – Offices of Data Analytics should collaborate with local university departments, FE colleges, the Open Data Institute and the Digital Economy, Transport and Future Cities Catapults to set up Data Analytics 101 training courses to educate public sector workers – and especially public sector leaders – about how to use data.
If government is serious about ‘dogfooding’ its own data, it is not enough to have a dedicated data team – every public sector worker needs a basic competency in data analytics. ODAs could be the catalyst for spreading those skills throughout the sector.
4 – As part of the devolution process, central government should establish a dedicated team responsible for “data devolution”.
Cities and city regions should be able to request access to datasets held by central government that support them in managing their devolved powers. For example, Bristol successfully negotiated access to higher level census information to better understand the city’s population, as well as data from DWP on local job seekers. If necessary, No. 10 should use its convening power to identify which datasets held by central government could help deliver reform of local public services.
5 – A national Office of Data Responsibility (ODR) should be established.
The ODR would be an independent body that: a) provides common legal guidance on data sharing across the public sector based on current legislation; b) independently reviews novel ideas for using data and helps share examples of best practice; and c) gives independent auditing and accreditation of public sector data privacy and data ethics policies.
Many of the big opportunities for increasing the efficiency and effectiveness of public services – joining up services across local authority boundaries, intelligently coordinating activity between different public sector partners, predicting and preventing problems from occurring – require making smarter use of shared data.
It is therefore problematic that public sector bodies each commission their own independent legal advice, frequently resulting in each organisation having a different interpretation of the same data protection legislation. Data sharing will be limited by the most cautious organisation in any group. Though valuable, the advice provided by the ICO tends to be very high-level and does not easily enable public sector practitioners to know whether specific initiatives they wish to pursue are viable.
6 – The Government Chief Data Officer should work with representatives from local government, central government departments, other public sector bodies and industry to define – and continuously update – open standards for data for the entire public sector.
Data standards should be set with a clear focus on achieving specific outcomes, for example delivering integrated care for the elderly. Compatibility with open standards should be highly recommended for 10 years, with a clear commitment that it will become legally mandated from 2025, allowing each organisation to phase out non-compliant systems. The aim of open standards is to reduce the technical barriers to data sharing caused when data is recorded according to different conventions and formats in different systems. If necessary, 10 Downing Street should use its convening power to bring together the relevant groups to make this work happen.
7 – An ‘app store’ (based on open standards) for front-end transactions should be created for central and local government.
The challenge with the current model of Government as a Platform is that it risks making government a monopoly supplier of its own front-end IT – a poor model for long term innovation (like Google saying it is the only organisation that can create apps for Android).
Instead, GDS should define clear standards for front-end transactions and apps (similar to how Apple defines rules for any app to be approved for its App Store) but let anyone build them. This would create a thriving marketplace for innovative solutions, enabling government and local authorities to benefit from the very best products provided by the private and third sectors.
Suppliers would benefit as instead of having to build bespoke transactions and apps for every organisation they could build them once and sell many times according to one standard. As prices became cheaper for standard solutions, this would in turn encourage local authorities to converge on common platforms, ways of working and capabilities, driving down costs still further. This would be the mechanism for helping phase out their bespoke legacy software and make technology a commodity.
8 – Councils should be able to act as trusted third parties to prove the identity of individuals for GOV.UK Verify.
For many vulnerable families and individuals who rely heavily on local public services, local authorities may know more about them than any other third party. Their local council should be able to act as an approved verifier alongside organisations such as Experian.
9 – GOV.UK Verify should be adapted to allow individuals to enable third parties to log into services on their behalf.
If GOV.UK Verify is to become the standard login mechanism for all government services, it will need to enable third party access so that it meets the needs of HMRC’s digital accounts. Many individuals and businesses will want their accountant to update and check their records on their behalf.
10 – A joint working group – the “Digital Nudge Team” – should be established between the Behavioural Insights Team (Nudge Unit) and GDS.
There is a huge opportunity for the UK to combine its work on digital government and behavioural insights to rapidly extend, accelerate and enhance the use of nudge prompts, to create what could be termed “digital nudge”. Many nudge prompts follow the acronym EAST: making a desired action Easy (removing barriers and friction), Attractive (standing out from other options), Social (using the power of peer pressure and people’s tendency to conform to social norms), and Timely (prompting an individual when they are most likely to make a decision). These are all things that technology is incredibly well suited to deliver.
Easy – Making it easy to complete a certain action is at the heart of good web and app design. Amazon tried to patent its “One Click” shopping (which negates the need to enter billing, delivery and payment details) precisely because it removes the friction of purchasing an item.
Attractive – Technology companies – and especially computer games developers – are masters at finding ways to maintain an individual’s engagement in a process by making it attractive. For instance, Zynga, creator of Farmville, dynamically change the game to keep users engaged when it looks like they may be losing interest (e.g. they log in less frequently). An algorithm monitors each user’s playing habits and – if needed – provides them with an incentive to stay, such as giving them a free token to access the next level.
Social – Technology companies make use of social norms: Amazon’s famous “Customers who bought X also bought Y” is one such example. Social nudge prompts are proven to work more effectively when they focus on a smaller peer group to whom an individual can relate, i.e. “the majority of people like you did X…”. Digital can offer more powerful and personalised social prompts by personalising content according to an individual’s peers.
Timely – Technology in general, and smartphones in particular, are ideal for delivering timely prompts. Apps use push notifications. Google Now, Apple’s Siri and Microsoft’s Cortana try to predict what a person might need based on their past behaviour or current location.
These capabilities of technology should be harnessed to full effect to improve the application of behavioural insights to government policy.
And just as digital technologies can support nudge techniques, behavioural insights can improve digital government. The Government recently announced that Amazon-style text messaging would be applied to public services. Nudge principles can dramatically improve the effectiveness of such messages. David Halpern gives the example of how the wording of SMS messages sent to job seekers had a substantial effect on how many turned up to jobs fairs. Sent a standard, impersonal text message, around one in ten turned up. However:
“adding the recipient’s name at the beginning of the text increased the proportion turning up by 5 percentage points, or 15 per cent… What if the adviser also added their own name? The number turning up rose even further, to 18 per cent. And how about if the adviser instead wrote: ‘I’ve booked you a place… Good luck!’ Now the proportion turning up rose to an impressive 27 per cent, a nearly threefold increase.” (David Halpern, Inside the Nudge Unit, p.121)
Find me on Twitter