00Three years on from lockdown: has the pandemic changed the way we shop?

Three years on from the start of lockdown, the legacy of the pandemic and its impact on how and what people shop, across the UK’s largest urban areas, is still visible. But while some of the behaviour changes have persisted, none have sparked a fundamental shift. Shopping is one of a number of areas where radical change was predicted but hasn’t been borne out. Despite early predictions, people have returned to bricks and mortar stores, which is good news for high streets up and down the country. But the challenges many of them faced three years ago largely remain.


The Covid-19 pandemic that hit the UK- and the rest of the world- exactly three years ago was a large-scale forced social experiment. How and what we consume, in particular, went through a radical shift. As shops, offices, cafes, pubs, and restaurants momentarily closed their doors, the response has, naturally and by and large, been a massive shift online.

Concerns rapidly emerged that people would never go back to the high street.1 Why would they return to a physical store after experiencing the ease of virtual shopping and online deliveries? And, perhaps more importantly, what would that mean for the future of British high streets and city centres?

This short briefing explores whether and how, three years on from the start of lockdown, these predictions have proven right over time. It uses consumer spending data up until the end of 2022 across the UK’s 62 largest urban areas to track the changing nature of consumption and the geography of online spend.

The state of play in 2019: across UK cities, more than a quarter of spend was made online

Even before the pandemic hit, online shopping was already a deep-rooted habit among many British households. By the end of 2019, 30 pence in every pound spent was spent online.2

There was variation between places (Table 1). London and Cambridge were the online capitals of the UK, with 34 per cent of all spend made online. By contrast, in Newport and Preston, it was 28 and 27 per cent respectively – lower relatively speaking, but still more than one in four pounds.

Several sociodemographic factors, like age and income level, likely played a role in this. Younger people and those on the higher end of the income spectrum, for example, are more likely to make purchases online.3 This may explain, at least in part, why affluent cities like London and Brighton rank first, and why those with younger demographics like Cambridge, Oxford and Coventry also had a higher online share than elsewhere (Table 1).

Table 1: The share of spend made online in 2019 varied between places

City Online share – 2019 City Online share – 2019
London 34 Norwich 29
Cambridge 34 Stoke 28
Liverpool 33 Swindon 28
Brighton 33 Warrington 28
Hull 32 Exeter 28
Coventry 32 Burnley 28
Dundee 32 Telford 28
Oxford 32 Gloucester 28
Southend 31 Newport 28
Birmingham 31 Preston 27

Source: Beauclair, 2023

This also varied between sectors. Averaged across all cities, sectors like fashion and retail already had a much larger online presence than food and drinks and groceries (Box 1). But a ‘zoom in’ on two of them, fashion and groceries, shows a very different geography of online spending (Figure 1).

The share of all spend made online in the fashion sector was much higher in cities and large towns in the North of the country – places like Hull and Wigan – than elsewhere. Of the ten places with the highest share of fashion spend made online, seven were in the North of England.

By contrast, the opposite North-South divide was true for groceries shopping, with cities like Cambridge and Brighton having a much higher (respectively 13 and 12 per cent) online share than places like Middlesbrough and Hull (between 5 and 6 per cent). Of the ten places with the highest share of groceries spend made online, nine were in the Greater South East.

Figure 1: The geography of online spend varied between sectors before the pandemic

Source: Beauclair, 2023

There are a few reasons why this could be the case. Firstly, online groceries deliveries are more likely to be for large and costly baskets that are more affordable for those on higher incomes, while day-to-day store purchases for food might be a better option for those on lower incomes. Secondly, the quality of the offer on the high street for non-essentials like fashion is likely to be higher in more affluent cities and towns, particularly for high-end brands, while the market for cheaper clothes is likely to be larger online.

Box 1: The share of spend made online varies between sectors

In 2019, not all sectors had the same online presence. Consumer spending data across ten categories shows how this played out (Figure 2). Entertainment and travel had the highest online share- mostly driven by the expansion of online booking systems- followed by more traditional high street services like retail and fashion (between 40 and 50 per cent of all spend). By contrast, sectors like hospitality (restaurants, pubs and cafes) and grocery had a much smaller online share, with more than nine in ten purchases still made offline.

Figure 2: Fashion and retail had a higher online share than food and drinks in 2019

Source: Beauclair, 2023

The ‘lockdown effect’: an increase in online shopping across the board

Online shopping was already on an upward trend in the two decades that preceded the pandemic, driven by the expansion of internet access and the rise of e-retailers like Amazon. ONS data shows that nationally, internet sales as a percentage of total retail sales went from less than 3 per cent in 2006 to 20 per cent at the start of 2020.4 Across UK cities, just in the 12 months that preceded the pandemic, the share of spending made online had already increased by two percentage points.

This massively accelerated at the start of 2020 when restrictions came into place, shops closed and people were told to stay indoors (Figure 3).  Between February 2020 and June 2020, the online share across UK cities rose to 42 per cent – hardly a surprise given that most shops were closed to the public. It then peaked at 50 per cent in January 2021 at the start of the second lockdown, and went down again to 37 per cent by July that year, when ‘Freedom Day’ officially marked the end of restrictions.

All 62 cities have experienced this rise over that period. This varied between places, from 13 percentage points in Swansea and Cambridge to less than nine percentage points in Aberdeen and Bournemouth, with no clear geography to it.

Figure 3: The ‘lockdown’ effect on online spending averaged across UK cities

Source: Beauclair 2023

Three years on, online shopping is more prevalent across all urban areas

As restrictions were lifted and people returned to shops, cafes, pubs and restaurants, online spending went down again. But it never quite went back to pre-pandemic levels and instead settled at a higher rate – a sign of stickiness of behaviours as people became more acquainted with it.

By the end of 2022, nearly three years on from the anniversary of the first lockdown, the share of spending made online was higher than its pre-pandemic levels across of 62 cities. On average across all UK cities, there was a five percentage point increase in the online share between the third quarter of 2019 and the third quarter of 2022 (from 30 per cent of all spend to 35 per cent).

Figure 4 looks at how this is playing out across cities.

Figure 4: By December 2022, online shopping was higher than its pre-Covid levels across all 62 cities

Source: Beauclair, 2023.

There are three things to note here.

The first is that looking at the ‘pandemic increase’, as shown by the size of the purple bar on Figure 4, suggests there isn’t a great deal of disparity between places. Out of 62 cities and large towns, 49 have seen their online share rise roughly by the same amount, between 4 and 5 percentage points.

The second is that looking at the cities that sit on each end of the spectrum and have experienced the highest and lowest increase suggests there is a geography to this: of the ten places where the online share grew the most, six are in the South of the country and one in the Midlands, with cities like Cambridge, London and Brighton retaining their crown as online spending capitals. By contrast, all four Scottish cities are in the bottom ten cities, where the online share grew the least (less than four percentage points).

Finally, that the very rapid jump in the share of spend made online is a result of lockdowns (as shown by the green dot on the chart) is undeniable. But that current levels have directly and solely been pushed up by the pandemic is less certain. Although online shopping is more prevalent now than it was three years ago, long-term trends at the national level show that current levels are not far from what they would have been anyway had the pandemic not happened (see Figure 5).

Figure 5: Longer-term trends shows online spending is now back to its pre-pandemic trajectory

Source: ONS, 2023.

The recent increase in online share has been larger for sectors that were traditionally less exposed to it

The rise in online spending since 2019 is visible across all ten sectors. The charts below show how this has played out across cities in the fashion, food and drinks, general retail and grocery sectors. Two separate patterns are at play:

The first is between sectors. Before the pandemic hit, sectors like fashion and retail had a much higher online share than groceries and food and drinks (Box 1). As Figure 6 shows, rather than an acceleration of pre-pandemic trends, it’s the lagging sectors that were the least exposed to online spend pre-pandemic that have broadened their reach. This catch-up growth was highest in the food and drinks sector, where the online share was fifty per cent higher by the end of 2022, compared to its 2019 levels.

By contrast, the rise in sectors that already had a high online presence in 2019 has been more muted, perhaps because there was less room for growth as internet shopping was already a rooted habit. This was the case in fashion and general retail, where the online share only went up by less 20 per cent.

Figure 6: Food and Drinks and groceries have seen online rise the most since 2019

Source: Beauclair, 2023.

The second is between cities. Here too, the fashion and groceries sector show opposite trends (Figure 7):

  • Most of the rise in online share in the groceries sector has been in cities that already had the highest pre-pandemic share- places like Brighton, Crawley and Worthing. By contrast, cities like Glasgow, Hull and Middlesbrough that already had a low online take-up rate haven’t seen a boost in the online groceries sector.
  • The opposite is true for fashion: places where the sector had a high online presence pre-pandemic haven’t seen a huge boost – cities such as Dundee and Blackburn; while places like Oxford, Cambridge and York have seen a larger increase.

Figure 7: The fashion and groceries sector have experienced opposite trajectories

Source: Beauclair, 2023.

Both these points help explain why some of the most affluent cities – many of them in the South of the country – have seen their online share grow the most. And it has also meant that the pre-pandemic spatial disparities discussed in the first section have widened in the groceries sector and narrowed in the fashion sector (as Figure 6 shows).

The rise in internet shopping doesn’t necessarily happen at the expense of in-store spending

Online spending is often perceived as the biggest threat to bricks and mortar retailers. The recent closure of retail giants like Debenhams or Topshop, for whom Covid has been the final nail in the coffin, has raised even more concerns and received lots of media attention.5

But the evidence that a rise in online spend is killing physical retail, or indeed accelerating the demise of the high street, is less clear.

More online spend doesn’t necessarily mean less offline spend. While it is true that during lockdowns, internet spending went up as offline spend levels plummeted, when the economy reopened, both went up simultaneously. By December 2022, both overall and offline spending levels were at or above their 2019 levels across all 62 city centres, despite the rise in internet spending.6 Averaged across all cities, this holds true across a number of sectors, as Figure 8 shows.7

Figure 8: Despite the shift online, most offline spending has returned to pre-pandemic levels

Source: Beauclair 2023.

That online and offline shopping can grow simultaneously largely explains why more internet shopping doesn’t necessarily mean more empty shops on the high street. Before Covid, already there was no clear relationship between the health of the high street and the prevalence of online shopping.8  London and Cambridge, for instance, were the UK’s online capitals but had two of the most vibrant high streets in the country; while cities like Newport and Stoke had a much lower online share, but more than one in five units on the high street stood empty.

And recent rises in internet shopping are unlikely to have changed this: most of the shift online since 2019 has been in sectors like groceries – more likely to be a move from large out-of-town supermarket spending to online delivery and not quite the typical high street retailer.9 The rise in online deliveries and takeaways is perhaps affecting local restaurants and other hospitality businesses more directly; but it isn’t necessarily problematic because an increase in demand, even online, still requires (and sustains) a physical presence on the high street.

Instead, what really matters for the strength of the high street is the total amount of spend that is circulating in the local economy, rather than how much of it is spent online. What vibrant high streets had in common pre-pandemic – in places like London, Reading and Brighton – was that they sat within an affluent catchment area where disposable income was high, which meant that people had more money to spend, both offline and online. The determining factor was the strength of the local economy, and in particular the presence of lots of well-paid jobs that created a market for shops and restaurants to sell to.10

What’s true, though, is that online shopping might be changing the nature of spending on the high street. Take fashion, for instance, which still accounts for a majority of the retail offer in many cities and was already facing struggles before Covid. It may not have seen the largest jump in online shares since 2019, but it is still the most at risk: the offline recovery for fashion has been relatively weak, with very little change between 2021 and the end of 2022, and hasn’t compensated the massive losses made in 2020 (Figure 8). City-level data shows that by September 2022 – before the Christmas boost – offline spending on fashion was still just below its 2019 levels in a total of 24 cities.

What does this mean for the future of the British High Street?

None of the radical change that many predicted at the onset of the pandemic has materialised. Over the past three years, internet spending has indeed become a more entrenched habit among British shoppers. But this has largely been an acceleration of pre-pandemic trends rather than a radical shift; and overall bricks and mortar spend has also bounced back.

The implications for the future of the high street, three years on from the start of lockdown, are therefore not too dissimilar to what they were on the eve of the pandemic. The first is that internet spending isn’t the biggest threat high streets up and down the country face; the lack of demand is. The second is that high streets will have to adapt the nature of their offer to changing consumption patterns and move away from an over-reliance on retail. Both are closely intertwined- and not all places face the same challenge:

Places like Oxford, Brighton, and Reading, for example, that combine high disposable income in their catchment area, and a relatively low share of commercial space allocated to retail have all the fundamentals to keep thriving. Other places like Exeter, Peterborough and Worthing rely relatively more on retail, but still in a high-wage local economy. This drives demand for local services and will support a greater diversity of amenities, provided policy enables and facilitates a flexible use of commercial space.

Cities like Blackpool, Middlesbrough and Wigan are more vulnerable. Commercial space in those cities- and many others often in the North of England- is still dominated by the most exposed sectors, like retail and fashion.11 And the relatively low affluence levels of their residents and the catchment area they sit within mean that demand levels are also relatively low. This makes it difficult for other sectors, like hospitality, to supplant fashion and retail and fill empty units.

In those places, the high street will only thrive and support a greater diversity of amenities if underpinned by higher demand for high street businesses. This has important implications for policymakers in those places: aside from concerns about the rise in internet shopping, the priority is to boost consumer spending power and address the underlying economics behind the lack of demand, by improving the strength of the economy and creating jobs that put money into people’s pockets.


  • 1 See for instance ‘Will shoppers return to the UK high street after lockdown?’ The Guardian, April 2020.
  • 2 This is higher than estimates from other sources, including the ONS. There are a few reasons for this: the first is that this dataset only includes cities, while ONS figures are country-wide. The second is that this only includes credit and debit card spending, and excludes offline cash transactions, which means that the online share is likely to be an overestimate.
  • 3 Young M, Soza-Parra J and Circella, G, (2022). The increase in online shopping during COVID-19: Who is responsible, will it last, and what does it mean for cities? Regional Science Policy & Practice.
  • 4 Other countries have seen similar trends in the last decade. In the US, for example, online retail sales accounted for 5.3 per cent of total retail sales in 2012, and this had increased to 11 per cent in 2019 (Source: US Census Bureau, 2023).
  • 5 See for instance ‘Will Coronavirus kill the high street?’ Financial Times, November 2020.
  • 6 This could in part be driven by a Christmas boost, but data for previous months also shows spending levels above their pre-pandemic baseline.
  • 7 The data has not been adjusted to inflation. This means that the quantity of items bought (volumes) volume of sales may still be down, even if spending levels are above their pre-pandemic levels, particularly in sectors like fashion.
  • 8 See for instance Quinio, V (2022), ‘How is Covid-19 impacting online shopping, and what does this mean about the future of the high street?’ London: Centre for Cities.
  • 9 The fact that an even greater share of spending on general retail is now made online is concerning for the sector but again it is more likely to be a threat for out-of-town retail parks. In 2019, general retail only accounted for 10 per cent of city centre sales (Source: Centre for Cities’ calculations from Beauclair data).
  • 10 Centre for Cities’ previous research has shown a clear negative relationship between vacancy rates on the high street and the share of high-paid, high-value jobs in the local economy. See for instance Cities Outlook 2022, London: Centre for Cities.
  • 11 See Cities Outlook 2022, London: Centre for Cities and McDonald R, Ramuni L and Tan L (2019), What’s in store? How and why cities differ for consumers, London: Centre for Cities.