The Office for National Statistics (ONS) has been keeping busy during the Covid-19 coronavirus pandemic, and the data collected and analysed by the ONS has been key in helping the country understand the impact of the virus.
The ONS might be best known for its census data, but the organisation’s remit is much broader than that. Frankie Kay, interim director general for data capability at the ONS, tells Computer Weekly how it has worked with academia, the government and the NHS to help the UK fight the virus.
Data science has played an important role, not just in informing the public and giving them “as much clarity and transparency as we can in terms of the pandemic itself”, she explains, but also in looking at the socio-economic impact of the pandemic.
ONS has had to spin up new surveys quicker than ever before. The questions asked have also changed during the course of the pandemic. While at the beginning of the pandemic it was more around whether people were adhering to lockdown rules, now it’s more around adherence to wearing masks and how people have been affected by the pandemic. The ONS has also created a business impact survey, looking at how businesses are faring in terms of furloughs, turnover and staff.
Table of Contents
New data sources
The ONS has also looked at and used new data sources to try to understand the impact of the coronavirus outbreak.
Early on in the pandemic, in the days of strict lockdown, it used Google Places data “to see whether people were adhering to it, or whether they were still gathering in parks and supermarkets”, says Kay.
Frankie Kay, ONS
“We’ve also used financial transaction data to look at the impact on the economy. It’s been a mixture of taking new, very targeted, quick surveys, and looking at different data sources to understand the impact of the pandemic,” she says.
The ONS also has a data science campus that is always looking for new and innovative data sources as part of its remit.
For its community infection study, ONS has been working in partnership with Oxford University, and the research is about understanding the spread and prevalence of Covid-19, both in terms of people who are currently testing positive and those who may have had the virus and have antibodies, to try to understand the prevalence across the country.
“The ONS found the easiest way to collect data to get an understanding of the prevalence of Covid in the community was through a survey,” says Kay. “We are looking to link that with other data sources, such as the NHS Test and Trace programme.”
Collaboration is key
This sort of collaboration, along with other government bodies and academia, has been key to the work the ONS has been doing throughout the pandemic.
“We’ve been working with the NHS on being able to link the data in terms of weekly deaths and the number of people in hospital with our census data so we can understand the prevalence of the virus based on ethnicity, which has been a big concern, as well as occupation,” says Kay. “We get access to their data, link it to our data and then provide new insights.”
One good thing that’s come out of the pandemic is that it’s made getting access to data easier, she adds.
“I think the importance and the severity of the situation has forced us to cut through some of the bureaucracy that was there previously,” says Kay. “We’ve definitely been able to work more quickly and get hold of the data at a far more rapid pace. People have been willing to share data because they understand the importance of it, and the speed at which we need to work.”
It has also worked with colleagues across government to build new analytical platforms.
“Particularly, we’ve worked with colleagues in the Cabinet Office and the Joint Biosecurity Centre on how we can create a new, safe platform using cloud technology. ONS has got real expertise in this area,” she says, adding that ONS is using Google Cloud Platform for some of the census data processing, as well as some of the Covid data. The organisation is also using cloud services from Amazon Web Services (AWS).
“If you think of the scale of the census, you can only do that in the cloud, so we’ve taken some of the work on surveys for the census and have been repurposing it to build those new surveys,” says Kay. “It’s taking the expertise that we’ve already built and applying that quickly to other instances, such as new surveys, or building environments for new data that was needed for Covid.”
Once the UK comes out of the pandemic, Kay says she will be keen to hold on to the improvements in the speed at which the organisation can work. “I think we’ve got to guard against going back to the old ways,” she says.
Another issue which many organisations have found challenging is the change to working from home. While this has also been a challenge for the ONS, it has mainly been a positive change.
“The IT operations team did an amazing job in getting everyone set up from home,” says Kay. “We’ve easily maintained our productivity, if not actually had a slight improvement.”
This has made the organisation realise there is a scope for changing the way it works in the future. “With moving more to an organisation where lots of people work from home, at least part of the time, we’re looking at how we can perhaps recruit from right across the UK rather than just tied to out location,” she says. “And that really helps bring together a much more diverse workforce.”
While sharing data across government has become easier during the pandemic, there are still challenges. Kay says it is key, when sharing data, to be very clear about what data is being shared, as well as what it’s being used for, and to make sure “it’s done in a safe, secure and ethical way”.
“That has to be the foundation of whatever we do in whatever circumstances. However, I would say there are cultural issues across government in terms of that, trying to sort of turn the mindset into how we can ensure we share data safely and ethically, rather than the mindset of ‘that’s scary’,” she says.
Frankie Kay, ONS
“Culturally, it’s about moving the view to saying that actually we must use data for the public good and that by not sharing it, we’re certainly not doing the good we’re able to do.”
It’s not just cultural barriers that are an issue, either. Technical barriers are prevalent, too.
“A lot of government IT systems are quite old, so simply extracting the data from them can be difficult and expensive,” says Kay. “We don’t have common standards across government for data or APIs [application programming interfaces], so exchanging data is tricky.
“The quality is sometimes not what it should be, and if you think about different uses, there are clearly operational uses of data that are very important. Another focus, therefore, is how we can try to use data to provide better joined-up services.
“There are all sorts of quality issues and a whole range of challenges which can make the practical side of things quite tricky.”
Proud of ONS staff
Overall, Kay says she is “extremely proud” of what the team at ONS has been able to achieve during the coronavirus pandemic. It’s meant huge changes for everyone, and ignited a drive to move faster than ever before.
She says pace is one of the biggest challenges the pandemic has brought, particularly as decision-makers at Number 10 Downing Street have been looking to the ONS for data to make decisions that affect the whole country.
“How do you provide that timely insight to decision-makers in time for them to make really critical decisions? It’s such a fast-moving picture, nobody’s ever seen anything like it before,” she says.
“It’s shown what we can do. I’m very proud of what we’ve achieved and I think we’ve provided real insight. Our stats are regularly used by Number 10 to help make these really difficult decisions about what we do and how we react, and that’s going to continue as we come out of lockdown and see the impact on the economy.”