Speedy data is great for forecasters. But we shouldn’t ignore what we can’t measure

BY Douglas McWilliams   /  9 February 2021

The coronavirus pandemic has been a challenge for economic forecasters. While we at Cebr have probably fared better than most other forecasters (our March 2020 forecasts revision predicted global output to decline by 4.0 per cent for the year at a time when many of our rivals were still forecasting growth – the latest data suggest the actual figure will be quite close to our expectation at -3.5 per cent), the past year with its numerous lockdowns, government support schemes and other restrictions has made short term prediction more difficult than I can ever remember.

Never short of a pithy phrase, Charles Goodhart once described short-term economic forecasting as half knowing where you are and half having a good sense of the autocorrelations in the system. Econometrics is a useful tool to help address the second part. But, to “know where you are” you need data that is both robust and up-to-date.

When I started in economics nearly 50 years ago, effectively apprenticed to Sir Winston Churchill’s wartime economic adviser Sir Donald MacDougall in his final job as Chief Economic Adviser to the CBI, the main UK data source was the government’s statistical office, then called the CSO.

The difficulty with the CSO data was that typically they were only available some time in arrears. In the early 1970s, the earliest GDP data became available about three months after the end of the period to which it related so on average it was over four months out of date. Most monthly data releases typically appeared about eight weeks out of date. Hence the business surveys which gave information that was much more recent and often forward looking were especially useful. CBI’s monthly survey used to be released to the media on a Friday afternoon after a frantic morning where we processed the responses that had arrived in that morning’s post.

Apart from national statistics, business and household surveys (such as the YouGov Cebr surveys) have been the most important source of alternative data for economic analysts over the past decades. But with the increasing digitalisation of our lives, we all now produce huge amounts of data each day through the internet, mobile devices and electronic payment methods. During the pandemic, the ONS has moved sharply forward with the weekly publication of “faster data”, making various new data sets available to the public. The plans to do so were recommended in the Bean Report in 2015 and the data had been tentatively launched in April 2019.

But lockdown has meant that much more regular and up-to-date economic data are required. And we now have weekly figures, at most a fortnight out of date, on store footfall, use of credit cards, traffic camera data, online job adverts, online pricing and the ONS’s business survey.

This “faster data” goes some distance towards providing a more exact answer to the question of “where we are currently?”. But not all the way. Much of the data relate to the consumer side of the economy which made up £1.3 trillion of transactions in 2018, compared to total transactions of £4.5 trillion. The major part of the £3.2 trillion gap is all the business-to-business transactions that take place in an increasingly vertically disintegrated economy – where firms tend not to own or control their supply chains. And a lot of what drives GDP growth these days is not so much the consumer but business-to-business transfers that incorporate the knowledge economy (as I try to explain in The Flat White Economy). It is always tempting to ignore those things for which statistics are not available.

To understand what is really going on in the UK it is also important to get some sense of the rapidly growing knowledge economy. We should be grateful to the ONS for the considerable improvement in the data they now make available. But the trick for economic analysts is to pull in not only the ONS’s information but data from all other sources. Only in this way can we start to shine a light onto the many economic activities that are still hidden so that we may eventually have a better sense not only of where we are but also of where the economy is headed.


Douglas McWilliams is chief executive and founder of Cebr, one of the UK’s leading specialist economics consultancies.


     linkedin      Email