Viewpoint

Statistics underly much of our economic and political decision-making. The range and detail of available information is growing at a substantial rate.

DOI: https://www.doi.org/10.53289/JCGL3263

The future of official statistics is already here

Volume 22, Issue 10 - July 2021

Professor Sir Ian Diamond

Professor Sir Ian Diamond

Professor Sir Ian Diamond FBA FRSE FAcSS DL is the National Statistician. He is the principal adviser on official statistics to the UK Statistics Authority and the Government. He is Head of the Government Statistical Service and the Government Analysis Function. He is also a member of the UK Statistics Authority Board as Chief Executive and Permanent Secretary. Sir Ian is a Fellow of the British Academy, the Royal Society of Edinburgh and the Academy of Social Sciences.

Providing the public, Government and all who make important decisions with independent, robust, and timely statistics has been the remit of the Office for National Statistics (ONS) for years, but the past 12 months have shown just how much our statistical system can really do.

Around this time last year we took an urgent call. The UK needed quick, reliable data on the scale of COVID-19 infections across the country. Without that information it would have been impossible accurately to track the progress of the virus within and among communities in the UK.

We took on that challenge and continue to monitor the status of more than 400,000 survey participants, with the latest data showing a hugely encouraging growth in antibody immunity at older ages, largely from the country’s enormously successful vaccination programme.

A huge challenge

The speed and scale required for the initial operation of our COVID-19 Infection Survey was a huge challenge that required collaboration, innovation, and an incredible amount of hard work from many of our office and field staff. Thanks to their efforts, we now produce reliable weekly data on infections, antibodies and more.

It remains a vivid demonstration of how fast data can inform important decisions, but the level of infection was not the only challenge facing the UK at the peak of the pandemic. Decision-makers also needed indicators about the state of the economy and how people were feeling about restrictions on their freedoms, all in virtually real time.

Meeting those demands required multiple actions: first, we ramped up our regular social opinions survey to a weekly schedule, providing information about the mood of the nation, adherence to social distancing rules and expectations for the future.

Second, to measure how much people were travelling we used information from, for example, traffic sensors, as well as anonymised mobility data from Google, generating a fast estimate of how strictly lockdown restrictions were being observed in different areas of the country.

Third, throughout the pandemic we have continually increased the scope of a series of new faster economic indicators to include more information from card transactions, automated tracking systems on cargo ships and prices data scraped from the internet, all novel data sources that could deliver critical insight faster than ever.

Innovation

This was, of course, innovation born of necessity. The need to socially distance removed our ability to gather information face-to-face from households and at ports of entry. Data gathered by these means were still fundamental to our statistics on vital issues like employment, population change and crime. Indeed, there were some who questioned the ability of the statistical system to produce robust figures under such constrained circumstances.

Such doubts were confounded by the agility with which the statistical system pivoted to new ways of working. Our key economic indicators – including employment, prices, retail sales and the public finances – have all been produced with minimal discontinuity. It was vitally important that these long-running statistical series were maintained. The Labour Force Survey provides uniquely rich detail on the scale and nature of employment but its production takes time. For an immediate view of the impact of the data on jobs we could draw on latest PAYE tax data from HMRC which vividly demonstrated Covid’s shock effect as hundreds of thousands of people disappeared from employer payrolls.

Thus, as the traditional economic surveys have continued to give a reliable view of the economy through the rear-view mirror, these new, faster data sources are helping to give an indication of the road to recovery following COVID-19.

Traffic sensors, as well as anonymised mobility data from Google, generated a fast estimate of how strictly lockdown restrictions were being observed.

Communicating clearly

Politicians and the public have followed our data on deaths, infections, vaccinations and more in briefings, news conferences and reports throughout the crisis. We have strived to communicate clearly, showing the limitations of our data, the uncertainty that comes with it and the nuances needed to understand the full picture where other data was involved.

The televised No 10 briefings and the Gov.UK Covid 19 dashboard have been unprecedented examples of public data used to explain important policy decisions. The gathering and presentation of indicators from multiple sources – case rates, hospitalisations, infections, tests, vaccinations and, most regrettably, deaths – have been the result of intense cross-Departmental collaboration and development.

The value of the UK Statistics Authority as an impartial protector of standards, whose interventions have significantly informed the improvements to the public presentation of statistics, has also been demonstrated.

In terms of public attitudes towards data, I would say the past year has not been so much a step change as a rocket boost, carrying expectations and understanding of the statistical system to heights we never imagined, but welcomed wholeheartedly.

As a result of the pandemic, the public have become consumers of statistics like never before, so communicating to them – on a broad range of issues – is something we must do more than ever now. We know there is more to do to make statistics fully accessible and we are committed to improving the way we present our work with different  formats, interactive tools and concise, well-articulated analysis.

While our expanded programme of antibody testing will monitor the effectiveness of the vaccine rollout – and we are continuing to investigate radical new data sources – we plan to go even further.

We want to use our experience gained during the pandemic to help the country tackle some of the big issues we face. Some have been exacerbated or brought to the surface by COVID-19, others are yet to emerge.

The drive to a ‘net zero’ economy and the ‘levelling up’ agenda are examples of the highly complex, long term strategic projects which come with very substantial data requirements, including the need for a detailed understanding of communities, our economy and society.

If we can truly understand and accurately measure the whole of the UK, actively collaborating with experts from other Government Departments, academia and elsewhere, we can begin to build towards those goals. The highly successful 2021 Census in England, Wales and Northern Ireland, is a brilliant starting point for that, but good quality, inclusive data should sit at the foundation of everything Government does.

To that end we are driving efforts for better, more joined-up analysis of data across Government through hosting a new Integrated Data Programme which will, with strong ethical and privacy controls, enable linked data from multiple sources to impact on policy to improve people’s lives. This programme builds on our successful work with the Economic and Social Research Council to develop the Administrative Data Research UK network and with Health Data Research UK to develop a public health data asset.

As a result of the pandemic, the public have become consumers of statistics like never before, so communicating to them – on a broad range of issues – is something we must do more than ever.

Joined-up analysis

Some examples of the power of joined-up analysis just from the pandemic include analysis of deaths among ethnic or religious groups which was enabled by linking death registration with census data from 2011. This showed the increased risks experienced by people from some ethnic groups; as well as work to identify the characteristics and symptoms of people who have been infected with the virus, the prevalence of Long COVID, the effectiveness of vaccines and more besides.

Our increasing ability to link data in order to produce these kinds of new insights promises a seismic shift in how Government can manage the operation of our essential public services to provide better services for citizens as well as in ensuring greater value in how taxpayers’ money is spent.

The ONS and the wider Government Statistical Service stands ready to use our expertise and impartiality to drive these changes and unlock the power of our data for the public good – now and in the decades to come.