From a poorly designed contact-tracing app to a biased algorithm unfairly deciding on thousands of students’ exams results: the past few months have shone a harsh light on the UK government’s shortcomings when it comes to using technology to deliver efficient public services – particularly at a time of crisis.
The problems are deep-rooted, and are made up of a cocktail of legacy IT, poor coordination, lack of skills and ultimately, failure to use quality data when making decisions. And now an Institute for Government report is calling on officials to “fix the plumbing” and focus efforts on resolving the underlying issues that have pervaded the government’s use of technology for decades.
- 3 ways SMBs use machine learning to power digital transformation
- Overnight digital transformation: Welcome to the year 2025, 60 months early
- Digital transformation: Legacy technology, lack of skills, and funding are holding back NHS plans
- After the remote working rush, here comes the CIO’s next challenge
- What is digital transformation? Everything you need to know
- Digital transformation: It’s time to invent the future we want (TechRepublic)
“If the government really wants to move into the digital world, it’s going to need to tackle some really tough challenges, particularly around how it uses and manages data,” Lewis Lloyd, who authored the report, tells ZDNet. “We want a consistency of approach that reflects the best technology that is currently available.”
SEE: Guide to Becoming a Digital Transformation Champion(TechRepublic Premium)
The researcher’s agenda matches the rhetoric that the government has maintained for a decade now. The parallels are particularly striking in the digital strategy published three years ago, which stressed the need to enable “digital by default” across government, to better serve citizens thanks to data-informed decisions and cutting-edge technologies.
Various organizations have also been set up, such as the Office for AI or the AI Council, and the industrial strategy has found resonance in other official documents such as the government transformation strategy, the UK digital strategy and, most recently, the national data strategy.
For Lloyd, however, the government’s best-intentioned initiatives are stalling; when it comes to enabling real change, political support for digital transformation is waning. “It doesn’t feel as though the government has been forced yet to focus on these foundational issues,” he says.
“You end up with this gap, where you’ve got the rhetoric on one hand, but actually with all that the government have on their plate it’s easy to brush under the carpet and wait for someone else to deal with it later on.”
One of the largest barriers to digital government, according to Lloyd’s research, lies in the failure to manage data. With a poorly handled transition from paper to electronic documents, government data from the past two decades has accumulated in no organized way, scattered across departments and ultimately impossible to search efficiently.
Lloyd found that millions of documents exist only in paper files, and are sitting in Whitehall basements, while millions of digital files are lost in unsorted folders and email inboxes across government.
Siloed data is impossible to access in order to make informed decisions for policy-making, whether to better understand the problems that need to be solved, to test the proposed solutions, or even to gather feedback once a decision has been made. Of the major programs for which government is responsible, in fact, only 8% are assessed on an ongoing basis to gauge the extent of their impact.
Using digital technologies, government could also open the door to external data – potential that has remained largely untapped so far. And yet, the rewards could be huge: in a 2017 study, Deloitte estimated that up to £130 million ($170 million) had been generated every year since Transport for London (TfL) had opened up its data to external app developers such as Citymapper, saving time for Londoners and reducing costs for TfL.
The public sector’s struggles with data are not immediately noticeable for citizens – until a crisis kicks in. In the summer, for example, while most students were still stuck at home after the COVID-19 pandemic kept schools closed, the exam regulator Ofqual announced that an algorithm would be used to predict A Level and GCSE results in the absence of physical exams.
It emerged that the system’s biased model was downgrading results in a way that affected state schools disproportionately. The government promptly backtracked, and instead of using the technology, eventually allowed teachers to predict grades.
“Concerns about data have existed for a number of years, but the A Level case is the first time we’ve really seen a real-world example that affects thousands of people in the UK in a way that is highly visible,” says Lloyd.
In another high-profile fiasco, it was recently reported that the government’s use of Microsoft’s Excel software led to up to 16,000 COVID-19 cases going initially unreported in England, as a result of a mistake.
This is another example of government significantly trailing behind in its use of digital tools, according to Lloyd. It is critical, therefore, to make sure that civil servants have access to software that isn’t “the lowest-denominator tool because that’s what everyone knows how to use”, says the researcher, but also that workers are trained to use new technologies – while carrying out change cohesively across departments to enable data to flow unimpeded across the board.
“It’s partly cultural,” says Lloyd. “People who have used Excel for many years may not realize that there are loads of other options out there. You need someone to take it upon themselves to say: ‘Look, this is a mess. The way the government is managing data is completely different across different departments and teams.’ That’s knotty and difficult, but you need someone to be coordinating, otherwise it will never be coordinated.”
Appointing a senior leader to oversee the use of data in Whitehall, in fact, is a three-year old government promise: in 2017, it was announced that a Chief Data Officer (CDO) would be hired before 2020.
Since then, the Institute for Government has renewed calls for the CDO to be appointed, with no success. Lloyd’s report stresses that the lack of leadership has only exacerbated the issues at stake, and urges government to name a CDO as soon as possible.
There is a lot to lose if a stronger digital strategy isn’t implemented soon within public services. After the series of recent failures to use technology effectively, the public’s trust has eroded. More than half of UK adults (53%) reported in a recent survey that they didn’t trust organizations that use algorithms to make decisions about them.
“There is a risk with the current government that, if you continue to double down on doing data badly, people will start pushing back even more,” says Lloyd. “If government gets it wrong too much more, you risk getting people pushing for not using technology at all.”
And yet in the current context, giving up on technology altogether certainly isn’t an option. “When you’re faced with things like pandemics or massive crises, like climate change, that are going to require coordination across the whole of government, and putting together huge amounts of information – the better you are at this stuff the better the response will be,” says Lloyd.
- Azure Synapse Analytics: A progress report
- 5 reasons AI isn’t being adopted at your organization (and how to fix it)
- InfluxData becomes the embedded IoT database for PTC, adds Azure support
- How MIT and IBM are fighting COVID-19 with AI (ZDNet YouTube)
- Google collects a frightening amount of data about you. You can find and delete it now. (CNET)
- One key to hiring your next data scientist (TechRepublic)