I provide an assessment based on the criteria published in the “4545 IFG – Showing your workings v8b” report. Overall Evidence Level: 0.2 (3/15) Section: Diagnosis – Why something is proposed Evidence Level: 0 Reason: This section did not present any evidence regarding ‘What policy makers know’ about the potential benefits and/or drawbacks of ‘digital government’, nor did it present anything regarding the way in which evidence for/against the ‘digital government‘ agenda has been assessed. Instead this section starts with a series of slogans that summarize the ‘vision’ of the digital transformation agenda. The ‘what we do’ and ‘how we do it’ paragraphs are reasonably clear, in the same was as typical a marketing pitch from a consultancy company might be. Section: Proposal (Action/Plans) – What the chosen intervention is Evidence level: 0 Reason: This section is very briefly presented through two links to previously published reports and one statement to the effect that the government intends to refresh its agenda based on a review of the progress so far. If the intention in this document is to present the Government Digital Strategy review as the proposed intervention, then the evidence has been presented. If however we are meant to infer the proposed intervention from the linked documents on the ‘Departmental Digital Strategies’ and the ‘Approach to Assisted Digital’ then this digital government has failed to present us with any information regarding the chosen intervention, the evidence this choice is based on, how it was assessed, what other options there were, or how any unaddressed issues will be dealt with. For all of these we must search through the linked documents to find the answers ourselves. Section: Implementation – How the intervention will be rolled out Evidence level: 1 Reason: The chosen implementations are identified as: User Research, Digital Inclusion, Common Technology Services, Government as a Platform, and Data. The overall explanation for these choices for delivering the intervention is given by the marketing slogans “we focused on users, saved billions” and by restating general vague targets that were set in the Government Digital Strategy “simpler, clearer, faster”. No evidence is presented for why these choices for implementation were made. On ‘User Research’, a brief explanation of what it means is given ending with the vague statement “[we do it] so we can make better services for them”. On ‘Digital Inclusion’, there is a brief list of partner organizations and a reference to the Digital Inclusion Charter [not linked] and the government aspirational targets, but no clear statement of how digital inclusion is being implemented, not any evidence concerning decision making processes. On ‘Common Technology Services’, a bullet-list of actions is provided. The evidence for the choice of implementation and decision process is summarised by stating that this is a continuation of the successful “Cabinet Office Technology Transformation programme”, though no evidence is given regarding how successful that programme was. On ‘Government as a Platform’, the general concept is explained and various services that were built with this concept as listed. No evidence regarding the decision process is given, nor indication if there are alternatives that could have been chosen. For the example of ‘GOV.UK’ as money savings figure is presented as evidence of success. For ‘GOV.UK Verify’ no evidence is given concerning the rate of adoption by citizens. For ‘Digital Marketplace’ expected saving as stated without supporting evidence. On ‘Data’, general statements are made regarding the aim and aspirations from the use of data but no actual methods of implementation are presented, not any evidence to support the feasibility of achieving the stated aims. Section: Value for Money – Cost/benefits to show why the policy is worth doing Evidence level: 0 Reason: In this section a number of services are listed with indications of the amount of money that is being claimed to have been saved. There are links to website of those services but no clear documentation or links that give evidence about the monetary savings. There is also no indication regarding increases/decreases or maintain of levels of service. To evaluate if something is ‘value for money’ it is not enough to know that there was a monetary benefit, we also need evidence that there was no prohibitive cost to the level of service. Section: Testing and Evaluation – How to judge if the policy worked Evidence level: 2 Reason: A brief summary is provided of the testing and evaluation methods that are used to maintain and monitor the standard of digital services. Links are provided to the tools, like the ‘performance platform’ that measures user feedback. Unfortunately, no information is provided regarding the decision making process that lead to this choice of testing-evaluation.

Evidence Check Web Forum: Digital Government 29th January 2016 Key Points • The Government should follow a strategy of privacy by design. • Reassurances must be given about the security of personal information. • Proper evidence must be presented to show that any new digital service will provide real benefits. Response The document, although very clear about the potential benefits that a more digital government can bring, lacks any focus on the privacy of citizens. Only the need “to retain citizen trust with a clear approach to what should and should not be done with these powerful tools” is mentioned in the section on safeguards. No information has been provided on the impact “Digital Government” will have on the privacy of citizens. In the modern world personal information is being stored in an increasing number of places. The potential for misuse; both accidental and malicious is therefore greatly increased. In 2015 there were a number of high profile hacks, some resulting in massive data theft, others just highlighting the need to keep personal information properly secured. Ashley Madison, the US Office of Personnel Management and TalkTalk all suffered serious hacks and as a result the security of information has shifted in the public’s mind. The plans to make government more digital will almost certainly require the collection and storage of more data . It is therefore important that information is provided on how this personal data will be kept secure. The starting point for any government project involving use of personal information should be to adopt a “Privacy by Design” approach. Privacy should be seen as a central plank of any project involving the retention and use of personal data. The Information Commissioner’s Office (ICO) has warned that too often the issue is “bolted on as an afterthought or ignored altogether”[1]. The benefits of prioritising privacy early on are clear. Any potential problems can be identified either before roll-out or at an early stage meaning that they are far less costly to rectify. Our reports on data protection have consistently shown a large proportion of data breaches occur through mistakes or employees not fully understanding their responsibilities. A clear demonstration that privacy is a core focus will increase public confidence in any project, increasing its potential effectiveness. A key part of this approach is the completion of a thorough Privacy Impact Assessment (PIA) before a project commences. Whilst it is expected that this process will be completed anyway it would be beneficial for any PIA to be made public, this will increase both public understanding of what is being attempted and what it may mean for their data. There is also value in using the PIA as a retrospective tool to ensure that privacy considerations were met throughout, this will assist with any audit of projects and encourage best practice going forward. Whilst the evidence presented mentions the intended results of the increased digitisation of government there needs to be a concerted effort to prove that this will yield beneficial results. All too often ideas that sound new and innovative are rushed into place without proper groundwork being done to show that they will actually have a positive impact. It should be the case that before any plan is rolled out the government publishes the business case and any other supporting documents to show why the move is necessary as well as the intended impact it will have, consultation with a broad range of stakeholders including the general public should always be considered. This, alongside the publication of a full PIA will help citizens judge for themselves the benefits of the scheme and will increase transparency and clarity of any initiative. As we all become digital citizens and are encouraged to be digital by default it is critical that engagement and explanation of a scheme is conducted with people and that no one is unclear on what the scheme may entail. It is also important that any project focuses on providing a service for the citizen not on asking the citizen to provide the service to the State. As noted the evidence presented outlines clearly why the Government wants this increased digitisation, it does however lack detail on what they think needs to be done to protect the privacy of the citizens they are trying to help. [1] Information Commissioner’s Office , Privacy by design: https://ico.org.uk/for-organisations/guide-to-data-protection/privacy-by-design/

2 of 2 These include: • Public trust: adopt Policy Exchange’s proposed Code for Responsible Analytics to create a political framework to enable smarter services for cross Government data sharing and enable consumers and businesses to make informed choices about data use • Collaboration: Government as a Platform must be built as a platform for citizens, public sector organisations and commercial profit/not-for-profit organisations to collaborate and communicate effectively, efficiently and securely • Technology: build a platform for Government applications that scales, is simple to use for developers, and enables services to be created quickly and efficiently using cloud platforms The idea for a Code for Responsible Analytics was first put forward by Policy Exchange, in its 2012 report “The big data opportunity” supported by EMC, as a means of enabling innovative use of data and analytics in the public sector. Its key tenets are: • Put outcomes before capabilities. Data and analytics capabilities should always be acquired on the basis of a clear and openly communicated public policy justification. Such capabilities should never be accumulated for their own sake, and when redundant be surrendered promptly. Curiosity alone should never be a good enough reason. • Respect the spirit of the right to privacy. Data and analytics should not be used to infer personal or intimate information about citizens (including, but not limited to, family status, sexual orientation, and political activity). Where this data is needed for public policy reasons, consent should be sought explicitly. • Fail in the lab, not in real life. A sandbox environment and synthetic data should be used to test all major big data initiatives – after which they will be subject to intense scrutiny and peer review. Initiatives that are deemed to overstep the mark on ethics or privacy will be dropped. Only those that stand up to close scrutiny by Ministers should be implemented. Beyond these principles, the Code for Responsible Analytics could also classify data use into different categories, providing citizens with an easy and transparent way to understand the potential ways their data could be used, and exercise control over this. The six levels of Creative Commons licenses (https://creativecommons.org/licenses/) provide a simple, standardised way for individuals and businesses to grant copyright permissions to creative work online. This could provide a useful model, though some thought and consultation may be needed to assess exactly how this could work. The public and private sectors will clearly want to use data analytics for different reasons. Nevertheless EMC believes that a Code for Responsible Analytics in the public sector would also enable more organisations in the private sector to use public sector data in their analytics programmes. In addition to enabling more businesses to innovate using public data sets, this would also provide a framework for reassuring consumers that their data was being used responsibly by businesses and being kept safe and secure. We therefore believe the Government should actively support the development of a Code for Responsible Analytics, and raise awareness of it among consumers, as part of a National Big Data Programme. This would provide considerable benefits by supporting the transition to digital government and unlocking new opportunities for businesses and consumers to drive economic growth, and develop innovative products and services. About EMC EMC is a global technology leader, employing around 70,000 people worldwide, with a mission to help organisations store, manage, secure and interpret information. We operate as a federation of businesses – EMC Information Infrastructure, VMware, Pivotal, RSA, VCE and Virtustream – to offer agile solutions that harness opportunities for transformational change. This requires a mix of sales, support, development, engineering and data science teams. EMC Computer Systems (UK) Limited was incorporated in the UK in 1986 and now has around 1,800 staff here and 3,000 in Ireland, across 11 sites.

EMC believes there is a significant opportunity for government to make better use of technology and data to transform public services to make them more efficient, and drive better outcomes for citizens. Key to this will be making more use of big data analytics, and the Government can also ‘pump prime’ the UK big data sector by making more use of this innovative technology. EMC has collaborated with independent partners like Policy Exchange, the Royal United Services Institute, and the economic consultancy Volterra, to understand and articulate the opportunity for the public sector. These reports have found that more effective use of data can save central government billions of pounds, while enabling key services like the NHS to deliver better and more preventative treatments for patients. Sustaining universal healthcare in the UK: making better use of information (2014), http://volterra.co.uk/wp-content/uploads/2014/09/Final-EMC-Volterra-Healthcare-report-web-version.pdf, discussed current best practice with regard to the use of information in the NHS, and the benefits in terms of increased efficiency and better, more preventative treatment for patients that could be delivered by spreading this more widely across the health system. This found that £66 billion could be saved in the NHS every year. Sustaining universal healthcare in the UK also identified three key steps public sector organisations should take to in order to implement and benefit from big data. These included: i. Focussing on data security and operational resilience, to ensure that public sector organisations and the essential services they provide, are not put at risk in the event of system failure. ii. Encouraging a ‘collaboration culture’ in which public service commissioners, providers, academics, industry, and the third sector are incentivised to work together, share data, and innovate. iii. Creating the technical infrastructure to enable this data sharing and collaboration, built on open standards. RUSI: Big data for defence and security (2013), https://www.rusi.org/news/ref:N523BFE8013582/#.VcnY2PlVikp, highlighted the opportunities for big data to drive efficiency in the Ministry of Defence, improve intelligence, surveillance and reconnaissance, and reduce the risk of operational failure, and called on the MOD to set up a working group and demonstrator programme to begin exploring the opportunities. Policy Exchange: Technology Manifesto (2014), http://www.policyexchange.org.uk/publications/category/item/technology-manifesto. The third chapter of this report, on government transformation, reiterated Policy Exchange’s previous calls to action around the need for more digitisation and the use of analytics to boost efficiency and innovation, and called on government to set up an Advanced Analytics team and draw up a Code of Responsible Analytics to enable this. Policy Exchange: The Big Data Opportunity (2012), http://www.policyexchange.org.uk/images/publications/the%20big%20data%20opportunity.pdf, found that big data analytics could save the Government billions and help deliver more responsive, smarter, and more personalised services. It called on government to develop a Code for Responsible Analytics and set up an Advanced Analytics Team in the Cabinet Office to drive adoption. The report estimated the overall savings to the public sector from using big data to be £16 billion – £33 billion per annum. EMC’s most recent report, on the Future of Government Digital Services http://uk.emc.com/collateral/customer-profiles/government-digital-services-report.pdf, highlights how the public believe the government has made great strides in digitising public services in recent years, but feel more could be done to create services as good as the private sector. The report also outlines the views of citizens and businesses and their priorities for digital public services. Picking up on the Minister for the Cabinet Office’s ambition to create a “smartphone state”, the report shows clear demand for a multi-channel experience, with 50% of consumers wanting the flexibility to interact online and still speak to someone if needed. In terms of specific government services citizens want to see digitised, 55% want more online services in healthcare, and 40% want more digitisation in local government. For businesses, tax and business support was the biggest priority, with 50% wanting an online directory showing all available tax grants and breaks, and 41% wanting personalised updates to changes in regulation and legislation. The report makes a series of recommendations to help accelerate reform, based on how organisations in the public and private sector have handled digital transformation. 1 of 2

2 of 2 Nuance is a leading provider of intelligent solutions which transform interactions with digital services at high volume, enabling successful self-service, and fraud reduction. In addition to major UK government departments like HMRC and DWP, Nuance’s customers include leading banks and telecoms providers such as Barclays Bank, TalkTalk and BT. Overseas Nuance has deployed solutions for the Australian Federal Government’s tax and welfare departments, the Philippine Government Pension Service, New York City’s 311 multi-channel advice service, and California’s 511 multi-channel travel advisory service. (See http://www1.nyc.gov/311/index.page and http://www.511.org/) Conclusion A significant opportunity exists for the scope of the Government’s digital programme to be expanded to deliver public services more innovatively, across central, local, police, health, and other parts of government. By deploying Natural Language Understanding with voice based authentication, the quality of services can improve measurably while also reducing cost to serve. It also enables government to deliver flexible and intelligent digital services that all members of the public can use.

Government is spending £7bn on 1.73 billion transactions with citizens a year, at an average of £4.30 each. To fully address the cost the Government’s digital strategy should look beyond the current web-only strategy and embrace a multi-channel digital approach. The digital strategy as it stands fails to take account of the ‘smartphone state’. In the UK today two thirds of the population access the web and make calls with their smartphone. Not only does it takes little account of the different ways different groups, for example the elderly, disabled and economically disadvantaged, actually use services. Citizens involved in complex transactions around pensions, tax, benefits or licenses often want to use the telephone and speak to a human agent when they have a query about a form and have a question to ask. According to research carried out by the Corporate Executive Board, 36% of phone callers are on a company’s website while talking to a customer service representative. It is wholly unrealistic to work on the assumption that all transactions with government will move to the web but it is not wholly unrealistic to plan to drastically cut the cost of telephone transactions with digital technologies like Natural Language Understanding and Voice Biometrics. There’s a huge opportunity for organisations to create intelligent digital services which work seamlessly across multiple channels and enable self-service. The government digital strategy needs to be adapted to meet the growing demand for a digital multi-channel experience, with intelligent self-service via web and telephone. This will achieve far greater cost savings, fraud reductions and inclusivity than in the previous parliament. The advantages already seen in live deployments in other governments and the private sector are: • Simultaneous improvement in customer satisfaction and cost reductions: Typically our reference customers report a 4 to 5 point improvement in customer satisfaction, depending on how they measure this indicator • Between 50 to 90% reduction in cost to serve, for equivalent transaction types; where the comparison is between a fully automated transaction vs. a fully manual one • Improved understanding in the business operations e.g. through near real time data capture and subsequent analysis, based on real reasons why customers are interacting. This insight can be used to manage and plan, change much more effectively • Security can be improved, both protecting customers from identity theft and reducing the possibility for casual fraud. Examples of where this is being applied include: In HMRC, Natural Language Understanding for telephony has been successfully deployed on time and on budget to deliver annual savings of more than £20 million for HMRC and £5 million for citizens through shorter call times. In HMRC, if necessary the caller is then taken through security entirely in speech. Additionally, HMRC has automated c.60,000 Tax Credit Renewal applications entirely in speech, via the self-service application. http://theforum.social/Portals/28/PDFs/Conferences/2015%20Conference/2015%20Presentations/2015%20Awards%20Presentations/HMRC.pdf 100,000 ING Netherlands customers can already use the virtual assistant to check their bank balance or issue payments within the mobile app, Inge. Using voice control mode enables a conversational experience controlled by the sound of voice. As of July banking customers in the Netherlands can use their voice to make voice-activated payments, authenticated by voice biometrics technology. Swedbank is one of Sweden’s largest retail banks, serving more than 8 million private and 600,000 corporate customers in Sweden and the Baltic region. Swedbank has implemented a virtual web assistant to deliver personalised customer service to its customers online. Within the first three months, the technology successfully handled more than 30,000 customer interactions per month, and resolved 78% of queries first time (i.e., without customers having to visit a branch or phone the bank’s contact centre). Since September 2014, the Australian Tax Office (ATO, equivalent to our HMRC) has been using voice biometrics to great success. ATO recently announced that 760,000 Australian citizens are using voiceprints to authenticate their interactions with the department. This has streamlined ATO’s authentication process: calls are on average 40-45 seconds shorter, and citizens have reported an improvement in customer service. Between September 2014 and July 2015, the Australian Government successfully verified more than 580,000 transactions using voiceprints. ATO intends to ramp up to 5 million voiceprints over the next three years. 1 of 2

presumably any savings have to be offset against the cost of the implementation. Gds has an annual budget of around £50M. Is this taken into account? How are the views of users taken into account when a decision is taken to create a digital only approach, such as to car tax or license points database? Why have individual government departments recruited their own developers and digital departments in replication of a central department? Is this cost reflected in savings? How reliable have the provided services been and what are satisfaction levels compared with traditional channels? In terms of resilience, what happens if we are no longer able to use digital technology ? Have we cut off old methods completely?

I work on the Forum Team at MoneySavingExpert.com and I've been working with Lucy Denton and Lucinda Blaser on the Digital Outreach Team for a couple of years as part of that, with the aim of encouraging our site users to engage more with Parliament and show them it's not daunting and they can easily get involved. I highly commend the Digital government pdf authors for the clarity, transparency and Plain English of the report as well as the tools and dashboards being made available. What an improvement!