The UK's first national lockdown from March 2020 and its immediate aftermath saw a massive shift in consumer habits that was initially mandated but then lingered as shops and restaurants opened but risks from the virus remained.
A new study from the universities of Cambridge and Newcastle used data from the ONS to compare retail, hospitality and online sales in the UK between March and August 2020 with average figures for the same months for the years 2010-2019.
Researchers took an approach normally used to estimate cumulative excess deaths to try and measure the impact of the COVID-19 shock on sales of UK retailers and restaurants.
They say their economic models suggest that shops predominantly selling food, such as supermarkets, saw a 5-10% bump in sales in lockdown, adding up to an additional £4 billion in earnings over "business as usual" expectations.
This is "consistent with large-scale stockpiling", they say, as people prepared for an indefinite future of home-cooked meals.
With many shops shut and people stuck indoors, online sales experienced a major boost, peaking at around a third higher than business-as-usual estimates during the first lockdown - an increase that amounts to an additional £4 billion.
Non-food high street shops, those selling everything from books to clothes, saw sales evaporate during the first lockdown when they had to shut, costing around £20 billion in turnover. Sales returned to normal once national lockdown lifted.
The shortfall for bars, pubs and restaurants was "dramatic", say researchers, with the first UK lockdown causing sales to fall as much as 90% below the business-as-usual level, equating to around a £25 billion revenue loss.
Hospitality sales saw some recovery post-lockdown, as government schemes such as 'Eat Out to Help Out' kicked in, but were still 25% below estimated business-as-usual revenues by the end of summer.
Writing in the journal Global Food Security, researchers say they found no evidence of a post-lockdown fall in food-shop sales as people used up their stockpiles, or an "overshoot" on the high street due to "pent-up demand" during lockdown.
"Lockdown restrictions led to behaviour changes in consumers and retailers that caused huge fluctuations in sales," said Dr Shaun Larcom from the University of Cambridge, who co-authored the study with his Cambridge colleague Dr Po-Wen She and Dr Luca Panzone from the Newcastle University.
"Shopping frequency dramatically reduced, and footfall vanished from many commercial areas, with people going online or using local outlets within residential areas when they had to shop."
"Consequences of lockdown, such as long queues outside supermarkets, led to 'forced experimentation'. Consumers had to explore new purchasing methods," said Larcom, from Cambridge's Department of Land Economy.
"Many people shopped online for the first time. They also bought directly from wholesalers or even farms, and trialled different types of home cooking. When people are forced to experiment, it can lead to behaviour changes that last well beyond the life of a crisis."
The researcher say that, while online sales peaked during lockdown, they remained above pre-lockdown levels in August 2020, which they suggest may be early signs of a more permanent "structural change" in shopping habits.
Recent media reports suggest that the UK Treasury is considering a one-off tax for online retailers who saw profits boosted by the lockdowns.
"Understanding the monetary impact of the pandemic is important to gauge the magnitude of the damage, and can help government design policies to assist these sectors," said Panzone from the University of Newcastle.
"Food services and non-food retailers lost a huge share of their yearly business, compared to food stores and online retailers that actually gained from lockdown. One-size-fits-all policy approaches across retail won't work," he said.
In February 2020, stores primarily selling food had sales figures almost identical to business-as-usual (BAU) estimates produced by the researchers' econometric models: £12.6 billion. Sales for March ran at £17.5 billion - around 10% higher than the £16 billion BAU estimates - but had returned to BAU levels by July.
For online retail, sales sharply diverged from BAU estimates by May - £5.3 billion against a predicted value of £4.1 billion (+29%) - and peaked in June at £6.8 billion compared to £5 billion BAU estimate (+36%). While online sales then started to fall, they were still above BAU estimates by the end of summer.
Non-food shops had February sales figures almost equal to their BAU estimates: £11.6 and £11.9 billion respectively. Actual sales tumbled as the pandemic took hold, with an April nadir of £5.9 billion compared to BAU estimates of £13 billion (-54.6%). Sales then started to recover, and by August only just lagged BAU estimates.
Sales in "food and beverage serving services" suffered most in terms of lost revenue. In February, turnover was £5.7 billion, just shy of the £6 billion BAU estimate. By March this had slumped to £4.3 billion against a prediction of £6.7 billion.
April sales for bars, pubs and restaurants were just £0.7 billion compared to a BAU estimate of £6.7 billion: an approximate shortfall of 90%. While this gap shrank it remained startling. Even with the 'Eat Out to Help Out' scheme, August sales were £5.2 billion compared to a BAU estimate of £7 billon (-25%).
Global Food Security