Shopping online became the norm during the coronavirus pandemic, but for some sectors, the sharp shift towards online purchases may be more lasting.

Despite restrictions easing and people returning to the shops, there are some things we’re buying online more, notably in the categories of “clothing”, “electrical household appliances”, and “sports equipment, games and toys”.

Inflation and the cost-of-living are also causing people to reduce their delayable expenditure and how much food they are buying.

Different sources of spending data give us various insights, we are going to use three sources for the analysis in this article.

For context, let’s start with what’s happening to our spending overall.

Online spending high but falling

Overall, we’re still spending substantially more online than before the coronavirus pandemic. In May 2022, seasonally adjusted internet sales accounted for 26.6% of all official retail sales, compared with 19.7% in February 2020.

Although online shopping has been a growing behaviour over several years, the pandemic appears to have accelerated the trend.

However, in recent months, we have started shifting back towards shopping in store, although spending online remains high.

Online shopping falling back towards pre-pandemic trend

Percentage of retail sales made online, by seasonally adjusted value

Embed code

Data for proportion of retail sales made online, RSI (XLSX, 19.4KB)

Notes:
  1. The blue shaded areas refer to periods when restrictions across the UK were in effect. Restrictions varied across UK nations and display is limited to three main periods. Further details can be found in our About the data section below.

This trend is also borne out in an experimental real-time source, Revolut transactions data. Among Revolut debit card users, the proportion of transactions that are made online also saw peaks during lockdowns and has been falling in recent months. This data source shows online spending is now at around 38%, relative to 62% in store.

Revolut users are typically younger and more metropolitan, which may account for some of the differences in figures when we compare this data with the official retail sales index and other sources, as younger people are more likely to shop online. Since the pandemic began, there may also be more in-store transactions recorded through card use as a result of people switching to cashless transactions.

Online spending has dropped in recent months

Revolut spending data, non-seasonally adjusted

Embed code

Data for proportion of retail sales made online, Revolut (XLSX, 35.7KB)

Notes:
  1. Highlighted periods of national restrictions are defined in our About the data section.

Online shopping for sports equipment, and other hobbies

As well as affecting pounds spent, the pandemic appears to have changed our habits when it comes to buying more online – particularly in some sectors.

Some of the changes in sector spending may be owing to people taking up hobbies, well-being activities or lifestyle changes during lockdowns.

Gardening items, such as flowers, plants, seeds, fertilisers, and pets and pet food have also shifted online, and we are buying more of them than in 2019. However, this category is a small component of our official retail data and that means a small change will cause the trend to move around more than it would in larger categories.

The same shift to online is true of “sports equipment, games and toys”, which still has spending higher than before the pandemic.

Online increase in leisure and gardening related spending

Retail sales data, showing proportion of sales made online and quantity bought for selected sub-sector categories compared with 2019

Embed code

Notes:
  1. Highlighted periods of national restrictions are defined in our About the data section.

Data for online spending for leisure and gardening related sub-categories and volumes (XLSX, 32.6KB)

Household goods have shifted more online compared with pre-pandemic, around 22% of sales in the past 6 months in this category have been online. Spending on household goods is likely linked to the strong housing market as people refurbish and improve their homes.

Household goods spending has moved more online

Retail sales data, showing proportion of sales made online and quantity bought for selected sub-sector categories compared with 2019

Embed code

Notes:
  1. Highlighted periods of national restrictions are defined in our About the data section.

Data for online spending for household goods related sub-categories and volumes (XLSX, 32.5KB)

In recent months, people have cut back on food shopping as a result of the cost-of-living crisis. When asked about their shopping habits in the past two weeks, around half (49%) of adults reported that they were buying less food when food shopping and 48% of adults reported they had to spend more than usual to get what they normally buy, according to our Opinions and Lifestyle Survey. Although the amount we are buying in food stores has dropped, the percentage of food spending online is still higher than before the coronavirus pandemic.

Clothing has seen an acceleration towards online spending, although it is not clear how permanent these trends will be in future. Retail volumes in May 2022 were only just above what they were in 2019, so that, despite large drops during restrictions, there has not been a rebound in how many clothes we are buying.

Clothes shopping moves online but volumes only just above 2019 levels

Retail sales data, showing proportion of sales made online and quantity bought for selected sub-sector categories compared with 2019

Embed code

Notes:
  1. Highlighted periods of national restrictions are defined in our About the data section.

Data for online spending for food and clothing sub-categories, compared with volumes (XLSX, 32.3KB)

Sales of alcohol and tobacco products remained high throughout much of 2021 and 2022, both in terms of spending (value) and quantity bought (volume).

Delayed purchases and rising costs

Retail spending, in pounds through checkouts, has largely recovered compared with pre-pandemic levels, but the cost of living is having an impact on transactions and how much we are buying in volumes.

This means that, while overall the amount we are spending is now above what it was pre-pandemic, this is largely owing to rising prices rather than the number of purchases.

In May 2022, our retail spending was 13% higher than in February 2020, but retail volumes were only up by 2.6%.

Rising prices have led to changes in behaviour. In this last week, 62% of adults said they had reduced spending on non-essentials as a result of the cost-of-living crisis, according to our Opinions and Lifestyle Survey.

Looking at card transaction data, we can see other signs of how inflation is affecting our spending.

Daily payments made by credit and debit card processors to around 100 major UK retailers go through the high value CHAPS payment system. Experimental data track these payments by type of transaction, whether online or in store.

“Delayable” spending, that is, clothing and footwear, vehicle purchases and household goods, has dipped in the most recent month; but it was already well below what it was in February 2020.

The amount spent on “work-related” spending in this graph includes commuter spending on fuel, which is now considerably more expensive.

Work-related spending, including commuting fuel costs, continues to rise

Credit and debit card transactions, from CHAPS data, non-seasonally adjusted

Embed code

Notes:
  1. Highlighted periods of national restrictions are defined in our About the data section.

Data for monthly CHAPS transactions (XLSX, 19.2KB)

Since May 2020, expenditure on “staples” (food and drink, communication and utilities) has consistently remained above the February 2020 average, with seasonal peaks at Christmas.

Social spending (including travel and eating out) plummeted during the pandemic and recovered as restrictions have eased, but is now going back down again.

These trends are also borne out in more detail in Revolut data. Spending on fuel has risen continuously since the beginning of the year, as a result of wide-scale fuel price increases.

Entertainment spending, on the other hand, has dropped in recent weeks following a steady rise since July 2020. However, it is too soon to say whether this recent drop in entertainment spending is a direct result of the cost-of-living and people potentially substituting expenditure to other essential categories.

The category for entertainment consists of non-essential goods and services and includes activities such as “theatrical producers and ticket agencies”, “betting”, and “membership clubs”, including sports and gyms.

As Revolut users are typically younger and more metropolitan, it is likely that this has an impact on overall figures for each category.

Longer-term spending has returned to pre-pandemic levels

Revolut spending by sector- non-seasonally adjusted, compared with pre-coronavirus pandemic levels (February 2020)

Embed code

Notes:
  1. Highlighted periods of national restrictions are defined in our About the data section.

Data for longer-term spending trends by sector, Revolut (XLSX, 156KB)

Looking at longer term trends, spending on travel and accommodation as well as pubs, restaurants and fast food (including takeaways), was heavily impacted by coronavirus restrictions, and has recovered slowly to pre-pandemic levels. As with other sources, any recent rise in spending is likely to be a reflection of rising prices rather than purchases.

Older age groups are spending less

Analysis of more Revolut spending data suggests that people aged 55 years and older reduced spending at the start of the pandemic more than younger age groups.

This may be because those aged 55 years and over would have typically been more affected by health concerns and a larger proportion of people in this age bracket would have been advised to shield during restrictions. After this initial reduction, trends broadly matched between age bands.

However, more recently, we have seen higher growth in spending for those aged 55 years and over compared with other age groups. This could potentially indicate a build-up of involuntary savings during the restrictions period.

Revolut is a newer source of spending data and its users are typically younger and more metropolitan than the population as a whole. Analysis is done on a per-account basis to adjust for the consumer base growing over time.

Those aged over 55 years saw lower levels of spending compared to younger age groups at the start of the pandemic, but higher spending growth in recent months

Revolut spending by age, non-seasonally adjusted, compared with pre-coronavirus pandemic levels (February 2020)

Embed code

Notes:
  1. Adjustments to user count are at an aggregate level and not by age band. See our quality and methodology paper for details.
  2. Highlighted periods of national restrictions are defined in our About the data section.

Data for Revolut spending by age (XLSX, 43.4KB)

About the data

Analysis in this article is based on three different data sources.

The Retail Sales Index is a national statistic produced by the Office for National Statistics (ONS). It shows the proportion of online and in-store sales, measured in value, and the quantity bought, measured in chained volume. Retail sales data in this publication are all seasonally adjusted.

CHAPS data is an experimental data series from the Bank of England based on CHAPS payments relating to credit and debit card usage. The monthly data time includes methodological notes that users should bear in mind. The monthly CHAPS index is calculated by the ONS, rather than being an additional series that is produced and validated by the Bank of England.

Revolut data is another experimental real-time source based on aggregated transactions made using Revolut debit cards. We have published a quality and methodology paper about Revolut data which covers the strengths and limitations of this data source in further detail.

Both of the financial transaction sources are real-time indicators and not seasonally adjusted.

Revolut data will be included in our weekly Economic activity and social change in the UK, real-time indicators bulletin from 21 July 2022.

National restrictions highlighted in charts

For charts, the blue shaded areas refer to periods when restrictions across the UK were in effect. For practical display purposes, these are limited to three main periods.

In order, these were: first national lockdown in the UK (23 March 2020) to easing of restrictions with non-essential shops reopening in England (15 June 2020); second lockdown in England (5 November 2020) to lockdown being replaced with three-tier system in England (2 December 2020); third lockdown announced in Scotland and England (4 January 2021) to “stay at home” restrictions ending in England (29 March 2021).

Contact

Luke Croudace
retail.sales.enquiries@ons.gov.uk
Telephone: +44 1633 455602