How has coronavirus affected pubs, cafes and restaurants?
The restaurant industry was effectively brought to a halt by lockdown. Since re-opening, there have been modest signs of recovery, but social distancing and widespread working from home continue to limit progress – and the sector is typically hit hard by recessions.
Lockdown orders between 20 March 2020 and 4 July 2020 closed all pubs, cafes and restaurants in the UK. During this time, the entire industry more or less came to a complete halt. Some restaurants remained open for delivery and takeaway, but the industry was a shadow of its pre-Covid-19 self.
Since then, we have seen modest signs of recovery, but the industry remains vulnerable. Recently announced social distancing measures to address the resurgence in Covid-19 infections seem largely focused on restricting the operation and use of pubs and restaurants.
The sector is important. Prior to Covid-19, the average UK household spent £27.40 per week, 5% of total expenditure, on food and drink away from the home (Office for National Statistics, ONS, 2020a).
Pubs, cafes and restaurants directly employ 4% of the UK workforce. They also employ a disproportionate number of workers with low qualifications – over 50% of workers have a GCSE or less – who will have few other job opportunities. In addition, pubs, cafes and restaurants are part of the amenities that make neighbourhoods and local high streets better.
Figure 1: Average hours worked 2020 versus 2019
Source: Calculated using microdata from the Quarterly Labour Force Survey, 2019 and 2020 waves. Restaurant defined by standardised industry codes 5610 and 5630 (Restaurant and mobile food service activities and Beverage serving activities). Vertical lines show: the first reported Covid-19 UK cases; the announced closure of restaurants, pubs and cafes; and the week of re-opening for restaurants, pubs and cafes.
Has the restaurant industry recovered from lockdown?
The restaurant industry has been hit particularly hard by the effects of lockdown. The accommodation and food service industry experienced the largest decrease in working hours of any industry (ONS, 2020b). During lockdown, hours worked for restaurant employees were down 65% over 2019, compared with a 24% decrease for all other workers. Despite the lifting of lockdown restrictions, hours for restaurant employees in July remained 52% lower than 2019 hours, compared with 12% lower for all other industries (see Figure 1).
Evidence from consumer expenditure data shows that UK restaurant spending by households in April and May 2020 was 30% of its total in the same months of 2019 (Surico et al, 2020).
As of mid-September, 18.3% of accommodation and food businesses had temporarily ceased trading, an improvement from 22.5% at the end of lockdown and the high of 81.2% in April. But an average of 27.4% of UK accommodation and food businesses employees remain furloughed, compared with 9.4% for the all UK industries. Nearly half (45.9%) of operating accommodation and food businesses report September turnover down significantly compared with a normal year; and 30% are not confident that they will survive the next three months (ONS, 2020c).
Post-lockdown: Did the Eat Out to Help Out scheme help?
In a May 2020 YouGov survey, between 57% and 63% of respondents stated that they would feel uncomfortable about visiting restaurants, pubs and coffee shops when lockdown restrictions were lifted. To address this post-lockdown hit to the industry, on 4 August 2020, the UK government implemented a major initiative to help the restaurant industry: the Eat Out to Help Out scheme. During the month of August, dinners in participating restaurants received a 50% discount on food items, up to £10, purchased in restaurants, pubs and cafes.
Figure 2: Restaurant dining relative to 2019
a) All of the UK
b) London only
Source: Calculated using data collected by Open Table. Each point shows the percent difference in seated diners for 2020 compared to the same day of week/week of year in 2019. Vertical lines show: the day of re-opening for restaurants, pubs and cafes; First day of Eat Out to Help Out; final day of Eat Out to Help Out.
The re-opening of restaurants on 4 July was followed by a gradual return to eating out across the UK. But by the end of that month, dining in restaurants was still below 2019 levels (see Figure 2a).
The Eat Out to Help Out scheme resulted in a significant increase in in-restaurant dining for the UK. For restaurants that had re-opened, Monday-Wednesday dining increased 50-100% over 2019 levels. Notably, on Thursday to Sunday, when the scheme discount did not apply, there was also an increase in restaurant dining over 2019 levels.
In the month of September, following the end of the scheme, UK-wide restaurant dining for sampled restaurants was comparable to that of 2019 levels (see Figure 2a). While it is difficult to quantify the precise effect of the Eat Out to Help Out scheme, it appears to have potentially played a role in making diners comfortable with returning to restaurants. But there is also evidence to suggest that this may have led to a Covid-19 spike (Phillips, 2020)
It is important to emphasise that the return to normalcy is not observed everywhere. For example, in London, the impact of the Eat Out to Help Out scheme was much more modest. September restaurant use remained 25-50% below its 2019 levels (see Figure 2b).
This difference between London and the rest the UK in general highlights an important issue that the scheme did not address: Covid-19 has affected the geography of demand for restaurants, cafes and pubs. Areas like central London, with a large number of restaurants, have experienced a large change in employment activity due to increased home working.
The move to working from home increases the demand for food and beverage services in areas that are primarily residential, but decreases demand in areas that are primarily productive, such as the City of London. Consider that, of those who work in the City of London, almost 250,000 live in other London local authorities and over 100,000 live outside the greater London area (GLA, 2014). This represents a significant decrease in the demand for local London services.
What longer-term risk does the crisis pose for restaurants?
The Covid-19 crisis creates two sources of risk for these businesses. First, pubs, cafes and restaurants are social places by design; the negative impact of social distancing on these businesses is greater than in many industries. A strong negative relationship between social distancing measures and restaurant spending is found in the United States (Baker et al, 2020; Cronin and Evans, 2020). Combined with the small margins on which many restaurants operate, this makes the industry particularly vulnerable to such measures.
The second source of risk to the sector is the large impact of the crisis on jobs and incomes across the UK. When people experience a reduction in their income, eating and drinking out is one of the things on which they typically cut back. US evidence finds that unemployment caused by the 2008 recession led to more cooking at home and less eating out (Aguiar et al, 2011; Kumcu and Kaufman, 2011). UK survey data show a 25% fall in average alcohol consumption outside the home in the years after the 2008/09 recession compared with the years before (Family Food Survey, 2018).
The impact of the Covid-19 crisis on employment and incomes has been unequal, with some areas experiencing much larger effects than others (Davenport et al, 2020). Neighbourhoods in which many households experience an income reduction due to the crisis are more likely to lose restaurants than neighbourhoods in which household incomes are less affected. This means that the recovery of local pubs, cafes and restaurants is likely to be very uneven across locations.
What does evidence from economic research tell us about these risks?
Most restaurants operate on small margins and rely on high volumes to be profitable. Social distancing policies, limiting the number of people allowed in an enclosed area, will affect the ability of a restaurant to hit volumes necessary at current prices. One study finds that compensating restaurants for social distancing policy will require a 17.7% wage subsidy (Koren and Peto, 2020).
Workers in the restaurant industry will disproportionately bear the burden of social distancing policies (Mongey et al, 2020; Dingel and Neiman, 2020; Costa-Dias et al, 2020). Recent Labour Force Survey data show that relative to the broader UK working population, restaurant workers are 59% more likely to have completed no more than a GCSE education, are 10 years younger on average, are 17% more likely to be women and earn 50% less per hour.
These characteristics suggest that restaurant workers are also likely to have lower savings and less transferable skills, making them particularly vulnerable to experiencing prolonged unemployment (Mongey et al, 2020).
The restaurant industry produces a locally traded service: the demand for restaurants comes almost entirely through the local economy. This has two important implications for recovery:
- The recovery of restaurants within a local authority will depend crucially on the recovery of other industries in that same local authority. Of particular importance are high-productivity, high-wage industries that trade goods and services internationally. This is what economists call a local multiplier – when other industries recover, people have more income, which they spend on local services. For US cities, ten internationally traded jobs in a local economy create 16 locally traded jobs (Moretti, 2010).
- The speed and extent to which pubs, cafes and restaurants recover will vary considerably across different locations. There is considerable variation across local authorities in how vulnerable jobs are to lockdown (see Figure 3). A recent study finds workers across the Midlands have relatively low vulnerability to lockdown measures, while workers in the North and the South are very vulnerable (Davenport et al, 2020).
Figure 3: Geographical variation in workers' vulnerability to lockdown measures
Source: Davenport et al (2020)
The outlook for the recovery of the general UK economy is uncertain. Real-time survey data show that 50% of businesses in the UK expect volume to decrease. (Dhingra and De Lyon, 2020).
Related question: How much will lifting lockdown start to reverse the UK's economic slump?
How reliable is this evidence?
Evidence of the effect of lockdown on the accommodation and food service industry is based on national surveys collected in the UK and United States. There is strong evidence that as of September, the Covid-19 pandemic had had a disproportionately large effect on the restaurant industry relative to the rest of the economy (Mongey et al, 2020, Dingel and Neiman, 2020; ONS, 2020b; ONS, 2020c).
But it is difficult to infer what will be the effect of a longer-term, less severe, social distancing policy post-lockdown. This is in part because we do not know the quantitative effect that such policies will have on business profitability. Further, faced with long-term social distancing policy, it is possible that the industry will adapt through innovation.
Related question: What are the effects of coronavirus on the UK and US labour markets?
Estimates for local demand reflect pre-pandemic life (Moretti, 2010). Caution should be taken in using these estimates to inform post-pandemic policy. For example, if after Covid-19, our demand for eating and drinking outside the home permanently changes, pre-pandemic local multipliers will overstate the post-pandemic reality.
Related question: What has coronavirus taught us about working from home?
What further evidence is needed?
Restaurants, pubsand cafes are different from many UK businesses in their dependence on highly localised economic conditions. This means that granular geographical information reflecting the spatial distribution of the UK’s employment risk is required. This will allow us to identify areas in which local restaurants will struggle most to recover.
This information can inform place-based interventions to ensure that, in areas hardest hit, the negative effect of the Covid-19 crisis is not compounded by further loss of employment and local amenities. Work on this is currently in progress by Gianni De Fraja, Jesse Matheson and James Rockey.
The picture of how social distancing policy will affect the restaurant industry is incomplete. Specifically, two channels need to be understood:
- First, quantitative estimates of the effect that long-term social distancing will have on profitability (given stable demand) are required to determine how policy will change restaurant supply.
- Second, part of the value that comes with visiting a local pub or restaurant is the opportunities they provide for social interactions. It is therefore likely that social distancing policies will have a negative effect on restaurant demand.
Where can I find out more?
The large and unequal impact of Covid-19 on workers: Abi Adams-Prassl, Teodora Boneva, Marta Golin and Christopher Rauh show that the negative consequences are particularly harsh for younger workers, those with unstable employment relationships and lower labour income.
How is Covid-19 affecting businesses in the UK? Swati Dhingra and Josh De Lyon use real-time survey data and find that 50% of UK companies had lower business volume in April.
How many jobs can be done at home? Jonathan Dingel and Brent Neiman estimate that 37% of US jobs can potentially be done entirely at home.
When face-to-face interactions become an occupational hazard: Jobs in the time of Covid-19: Besart Avdiu and Gaurav Nayyar look at the relationship between how much face-to-face interaction is involved in a job and how much of a job can be done from home. Food service jobs have very high face-to-face interactions and very low home-based work.
Sectoral effects of social distancing: Jean-Noel Barrot, Basile Gassi and Julien Sauvagnat estimate how social distancing measures will affect industry-specific productivity in France.
Business disruptions from social distancing: Miklos Koren and Rita Peto model the impact of social distancing measures. Retail, hotels and restaurants, arts and entertainment, and schools are the most affected sectors. See also this VoxEU article by the same authors.
Which workers bear the burden of social distancing policies? Simon Mongey, Laura Pilossoph and Alex Weinberg find that workers in low-work-from-home jobs are more economically vulnerable, in terms of education, income and assets, than workers who can move their employment to home.
Did Eat Out to Help Out lead to Covid-19 spikes? Toby Phillips finds suggestive evidence that the government scheme may have led to observed increase in Covid-19 cases through the month of August.
Who are UK experts on this issue?
- Gianni De Fraja (University of Nottingham), Jesse Matheson (University of Sheffield) and James Rockey (University of Leicester) are working on the geographical distribution of the UK’s economic recovery in activity for the local service sector.
- The Institute for Fiscal Studies has several studies, including on workforce diversity and the role of policy in restarting the economy (Costa Dias et al, 2020) and the geographical distribution of vulnerabilities to the pandemic (Davenport et al, 2020).
- Abigail Adams-Prassl (University of Oxford, Institute for Fiscal Studies) is working on projects to examine inequality in the effects of the crisis across the labour market.
Author: Jesse Matheson
12/10/20: Article extensively updated to reflect latest data and government policies. In particular, Figure 1 was updated, and Figures 2a and 2b were added, along with the middle section on the Eat Out to Help Out scheme.
Published on: 03rd Jul 2020
Last updated on: 21st Oct 2020