Meeting the challenge of measuring the economy through the COVID-19 Pandemic

The developing coronavirus (COVID-19) pandemic has led to new challenges for us, such as how to collect information with many shops closed, businesses ceasing to trade and no interviewers to knock on doors. The situation continues to evolve, and even now, the economy is operating very differently to how it did just a matter of weeks ago. The ONS has published three new articles looking at the effects of COVID-19 on National Accounts, Prices and the Labour market. Jonathan Athow explains how we are addressing some of the known – and the new – challenges, including highlighting how we will be collecting and publishing some of our key indicators in the weeks and months ahead.

The COVID-19 pandemic has had profound impacts on our health and on the functioning of our society and economy. The job of the Office for National Statistics is to produce a comprehensive picture of what is happening across the UK. Our statistics are hugely important in informing policy makers and the public as the UK seeks to manage the pandemic and its wider effects.

The scale and complexity of the challenges for statistical organisations around the world are unprecedented. Professor Tara Sinclair of the George Washington University was recently quoted in the New York Times saying; “The shutting down of large parts of the economy was not anticipated in the construction of our economic statistics”. This summarises the issue perfectly.

The disruption to our daily lives from COVID-19 has made compiling many of our economic statistics more challenging. To outline how we are managing this, today we have published detailed articles laying out how we will continue to produce GDP, prices and labour market data in the current environment.

There are many moving parts to this, but they fall into three main categories:

  • Conceptual challenges; how do we measure the price of goods that are not available or adjust estimates for education output when children, in the main, are not at school?
  • Data collection problems; how do we keep collecting data when some companies are not trading or when we cannot send people to interview households or record prices from shops?
  • Methodological concerns; how do we adjust our raw data, given the way our economy functions has changed so significantly?

Taking the conceptual problems first, these are numerous and each needs to be addressed separately, considering the most appropriate treatment to give meaningful economic statistics. For example, how should the Coronavirus Job Retention Scheme – where employers of furloughed staff are paid 80 per cent of salaries by Government – be accounted for in GDP?

We have decided to count the scheme as a subsidy to business, netting it off the income measure of GDP, as the furloughed employees will continue to count as employed and the payments they receive from their employer as wages and salaries. We will adjust education output to reflect the number of students in school and add in learning from home. We are accounting for unavailable goods and services by using methods such as assuming their prices would have moved in line with the average movement for related goods and services. This is the simplest approach that comes as close as possible to reflecting that the supply of certain goods and services has been interrupted.

Turning to data next, the challenge comes from surveys, which underpin many of our statistics. For example, we survey businesses to measure GDP, and individuals to understand whether they are in work or unemployed. We also collect prices from shops to measure inflation. Many of these surveys have been disrupted to some degree, which is understandable – some businesses have temporarily shut down, or are working away from their offices and may not be able to respond to our surveys. For shops and services that are closed, or under restricted operation, we can no longer send people there to collect prices.

For both of these reasons, we are relying more on collection over the phone or online. Such a sudden change can result in fewer people responding and collecting less information than normal, and we are carefully monitoring the outcomes of our data collection, because they are pertinent to our third main challenge: methodology.

In general, one of the most common issues we deal with is when firms or businesses do not respond to our survey, or where data are late for other reasons. When that happens, we have to fill in gaps in our data collection, technically known as imputation. Under normal circumstances, we can do this by using historical relationships between different data sources. But those historical relationships may not hold now.

While each problem is being carefully thought through, there are other ways we are working to provide the best possible information on the UK economy.

Firstly, we have been looking to develop new data sources. We have created new surveys that can help us fill the gaps, and are using administrative data such as information from HMRC on employees being paid through ‘real time information’. We are also continuing to work with businesses to gain access to valuable economic data.

These new data sources will sometimes shed light on specific economic issues, such as how businesses are changing their employment practices. They can also help us cross-check our core economic data and inform the judgements we need to make.

Secondly, we have been drawing on expertise within the ONS, international statistical bodies and the academic community. We have skilled methodologists in the ONS who are helping us develop approaches to dealing with missing data. We also have, for example, an expert technical panel which supports us on inflation measurement, and another that provides a broader base of external economic expertise. We can also look to international guidance and practice to help us make decisions.

Lastly, we are being as transparent as possible about the issues and how we are addressing them. Today’s articles are part of that process, but it is important to remember that these are truly unprecedented times, and there is scope for more revisions than normal. We have made some changes to our publication schedule to account for the challenges we are facing, and these are included in the articles too.

As the pandemic progresses and our economy and society continues to change, we will undoubtedly need to refine our approach. We believe the information published today shows clearly how we are addressing the initial challenges, but we are keen to hear from – and work with – all of our users to ensure we can continue to measure the UK economy as well as possible.

 

Jonathan Athow is Deputy National Statistician for Economic Statistics