David Spiegelhalter and Anthony Masters 

Covid has played havoc with interpreting economic trends

Caution is needed when extrapolating from statistics in times that are far from normal
  
  

An Iceland supermarket advertising for staff in London.
An Iceland supermarket advertising for staff in London. Photograph: Andy Rain/EPA

Some economic effects of the coronavirus are obvious. Pre-pandemic, the share of retail sales conducted online had taken around eight years to go from 10% to 20%, but then in nine months shot up to 36% in January 2021.

But perhaps more surprising was a recent dramatic headline from Bloomberg: “UK Wage Growth Hits a Record as Vacancies Pass 1 Million”. This was based on estimates from the Office for National Statistics’ monthly wages and salaries survey that average regular pay had increased by 7.4% over the past year. So why should we be cautious in interpreting this huge rise?

As the virus spread, restrictions arose and economic activity slowed. GDP dipped drastically but by June 2021 had recovered to 2% below pre-pandemic levels and was increasing fast, at 1% per month. The headline change for wages compares April-June 2021 with the dip a year before. This lower starting point inevitably leads to larger relative increases – so-called “base effects”. Similarly, job adverts are currently 30% higher than pre-pandemic, but the change from a year ago is even more startling – they have more than doubled.

Another important issue behind the growth in average wages is that the people employed may have changed – known as a “compositional effect”. If those on lower pay lose their jobs, the remaining average pay rises. As the ONS explains, if a short person leaves a room, the average height increases, even though nobody has got taller.

Also back in June 2020, there were about 6.8 million employees on the HMRC furlough scheme, but this had dropped to 1.9 million by June 2021. The ONS has ways of trying to remove effects and judges that the true underlying increase in wages is around half the headline figure, although these unprecedented times mean we cannot be precise.

As people have got back into shops, internet purchases have dropped to around 26% of all retail sales, but this is still well above pre-pandemic levels and could be a permanent shift. Some measures are more easily sensed through our own experiences.

• David Spiegelhalter is chair of the Winton Centre for Risk and Evidence Communication at Cambridge. Anthony Masters is statistical ambassador for the Royal Statistical Society

 

Leave a Comment

Required fields are marked *

*

*