Questions and answers about
the economy.

Zoomshock: how is working from home affecting cities and suburbs?

The switch to remote working during the pandemic means that much more of our spending happens close to home rather than in city centres, shifting the geography of demand for locally consumed services. Longer-term consequences will depend on whether workers go back to the office.

For many people, Covid-19 has led to an unprecedented shift in where they work – from the office to their homes. This has inevitably changed where people spend their money too. Workers who frequented cafes and shops near their offices before the pandemic will have found alternatives near home. Other spending – such as on gym memberships, haircuts and after-work drinks – has been curtailed entirely at various points over the past year due to lockdown measures.

By changing the geography of where we work, working from home has consequences for the geography of demand for locally consumed services – something we call the ‘zoomshock’ after the now widely used video conferencing software. Its immediate economic impact has been large and the long-term consequences could be too: much depends on how many of us return to the office once the pandemic is over.

How have working patterns changed?

Currently, 36% of working adults are working at home exclusively, while 47% are travelling to work, either exclusively or in combination with working from home (Office for National Statistics, ONS, 2021). As nearly nine in ten UK workers lived and worked in different neighbourhoods before the pandemic, this change in location has resulted in a substantial shift in people’s commuting patterns. Figure 1 shows the drastic reduction in journeys taken on the national railway network, the London underground and buses since March 2020.

Figure 1: Use of transport modes (percentage of an equivalent day or week)

Source: Department for Transport

Urban areas are particularly affected by these changes. The City of London, for example, has only around 8,000 residents but, prior to the outbreak of Covid-19, received over half a million commuters each day.

The huge decline in commuters to the capital and other cities around the UK following the switch to working from home has led to a reduction in demand for services near workers’ offices. In practice, we should expect a corresponding increase in demand for these services where (former) commuters live. A whole range of businesses is likely to be affected by this zoomshock – from cafes and sandwich shops to hairdressers, gyms and drycleaners.

The long-term impact will be determined, in large part, by whether workers return to their offices after the pandemic. Even a moderate permanent shift to working from home may result in a substantial relocation of businesses to where former commuters live, which would also bring the potential for business closures and redundancies. In the meantime, these local businesses, which are predominantly located in city centres and other areas with a high concentration of offices, may need support to avoid job losses.

How many jobs can be done from home?

If the scale of the zoomshock will be determined by future work patterns, workers’ ability to do their jobs from home over the longer term will be an important factor. The consequences for unemployment and output for businesses serving commuters are unlikely to be severe if working from home is viable for only a small number of workers. Certainly, the effects would be minor compared with other aspects of the pandemic, such as lost output, substantial increases in government debt and the closure of many firms.

Studies in the United States and Germany have found that a substantial share of jobs can be done from home: 37% in the case of US jobs (Dingel and Neiman, 2020; Alipour et al, 2020). Nevertheless, the ability to do so is not uniform. Fewer people are able to undertake their work remotely in lower income countries, in part due to the higher share and nature of self-employment in those countries (Gottlieb et al, 2020).

Even in the United States and the UK, there are inequalities – for example, relative to their male counterparts in the same occupation and industry, women on average do fewer of the tasks that their job involves from home (Adams-Prassl et al, 2020). Further, when both partners work from home, some evidence suggests that childcare is shared evenly, but other housework tasks fall mostly on women (Del Boca et al, 2020).

How much does the decline in commuting matter?

The overall impact of working from home on businesses depends not only on how many are now doing it, but also on where they live relative to where they (ordinarily) work. Consequently, the size of zoomshock is not the same in all parts of the country.

The zoomshock is small if the number of people working from home in an area is about the same as the number no longer commuting, even though the behaviour of many workers has changed. This is because the reduction in people commuting in is balanced by the decrease in numbers commuting out. A largely residential suburb, where formerly many commuted to a city centre, experiences a large positive zoomshock, while the city centre suffers a large negative zoomshock.

These observations lead to a definition of the zoomshock in an area as its net inflow of commuters whose occupation permits them to work from home. The net inflow of an area is the difference between on the one hand, those who reside in the area and whose job can be done from home, and on the other hand, those whose place of work is in the area and whose job can be done from home. In some areas, there is a net outflow of people, as shown by the areas marked in red in Figure 2.

Calculating this difference gives the zoomshock for a given area, based on the jobs it is possible to do from home (De Fraja et al, 2021; Dingel and Neiman, 2020). This is done at the MSOA level for all of England, Scotland and Wales. MSOAs (‘middle layer super output areas’) are a census partition of the country, designed so that each has an approximate population of 9,000. Having a consistent population is useful as it means that the zoomshock can be directly compared between neighbourhoods.

Figure 2 shows the zoomshock in each of the 983 MSOAs in Greater London on the left, and each of the 32 boroughs on the right. The zoomshock is measured as net inflow of workers per 100 workers, as the boroughs are administrative subdivisions with very different populations.

One expects boroughs in the city centre to witness an exodus of commuters and suburbs to teem with remote workers – and this precisely what the map on the right shows: red areas have a negative zoomshock (a net outflow); blue areas a positive zoomshock (a net inflow).

Figure 2: Zoomshock for Greater London, by MSOA and borough

Note: These maps show quantiles of the zoomshock by MSOA (left) and local authority (right) in Greater London. Quantile boundaries are calculated separately for positive and negative shocks to represent the 25th, 50th, 75th, and 95th percentiles for Great Britain in the left-hand side map and the 25th, 50th, 75th, and 95th percentiles for Great Britain in the right-hand side map.
Source: Authors' calculation based on data from the ONS Annual Survey of Hours and Earnings, 2017, 2018, 2019

How will the geographical shift affect restaurants, gyms and local retail outlets?

Jobs most likely to be negatively affected by the zoomshock are those in locally consumed services such as cafes, hair salons and gyms. Not only can these jobs not be done remotely, but they must also be provided where there is demand. They differ from an industry such as car manufacturing, which clearly cannot be done from home either, in that where cars are made is not where people drive them.

The precise location of locally consumed businesses like coffee shops, pubs and restaurants is important. A cafe in your office building is ideal, one 500 yards further away much less convenient and one 20 miles away wholly irrelevant to you.

In other words, the demand for locally consumed services literally follows workers from other sectors. This implies that the change to working from home due to the zoomshock causes a substantial mismatch between where the locally consumed businesses are and where their customers are.

Since March 2020, the zoomshock has had a major impact on businesses in certain areas – such as coffee or sandwich shops in Canary Wharf, an area with a high density of offices. Yet we must remember that others – gyms, non-essential shops, hair salons – have been closed regardless of their location during national lockdowns, making zoomshock a secondary concern.

Whether because of zoomshock or lockdown, locally consumed businesses have suffered during the pandemic. Workers in this sector make up roughly one quarter of the UK’s total labour force but many of these jobs may be at risk. Disproportionate job loss among workers in locally consumed services also stands to increase income inequality (Crossley et al, 2020). Evidence shows that this has been the case in US cities (Althoff et al, 2020).

Workers in this sector are disproportionately less educated and have lower than average earnings, though there are important exceptions such as medical professionals. Aside from carers and again medical workers, these jobs are also obviously those hit hardest by social distancing and lockdown measures (Adams-Prassl et al, 2020).

Are workers more productive at home?

The answer to this question depends in part on how working from home affects workers’ productivity. Economists have only recently begun to analyse working remotely in a systematic way. One study compared average productivity of call-centre workers randomly assigned to work from home and found the practice to increase their productivity by up to 22% (Bloom et al, 2015).

Other research finds that wage costs may also be reduced as the average call-centre worker is willing to accept 8% less compensation to work from home (Mas and Pallais, 2017). Recent work estimates that 22% of full-time work in the United States will be done from home after Covid-19, resulting in a 2.4% productivity increase (Barrero et al, 2020).

This is countered by other recent research showing that face-to-face communication also improves productivity. Perhaps then, the optimal set-up is a mix of in-person and remote interaction (Battiston et al, 2021). This is supported by economic theory, which suggests that there is an intermediate level of working from home that maximises the country’s average productivity and GDP (Behrens et al, 2021).

This logic indicates that, post-pandemic, many former commuters will return to their places of work but less frequently than before. Rather than every working day, they may, for example, commute twice a week and work from home the other three.

Will the zoomshock be permanent?

We are aware of no study that has forecast the extent to which working from home, and the relocation of economic activity it represents, will be permanent. But survey evidence suggests that both firms and workers favour some reduction in commuting post-pandemic. For example, a September 2020 YouGov survey found that 57% of UK workers want to continue to work from home some or all of the time after the pandemic – see Figure 3.

Figure 3: Working from home (current and future preferences)

Note: % of workers who were working prior to the coronavirus crisis and who also expect to still be in the workforce once the crisis is over
Source: YouGov

Likewise, firms are keen to take advantage of cost-saving opportunities made possible by having their employees working remotely. A June 2020 survey found that 44% of employers were thinking about cutting down on office space. Indeed, many large companies, including BP and Deloitte, are considering or have already announced plans to reduce or close offices as they move to flexible working.

Taken together, this suggests that while many workers will return to commuting, working from home will be much more widespread in the future. This is especially likely given that millions of employees have acquired skills (for example using Zoom, Meet, Skype and similar apps) and invested in the capital that working from home requires (from office chairs and laptop stands to improved internet connections).

The zoomshock may therefore cause substantial job losses in some geographical areas (such as city centres) but, at the same time, enable growth and new establishments in others (such as suburbs).

Are there other consequences of the zoomshock?

Beyond the impact on locally consumed services, there are other likely consequences of the zoomshock. A permanent increase in working from home will lead to a reduction in the size of offices. If many workplaces follow the examples of Dropbox and Facebook and switch to permanent ‘work-from-anywhere’ policies, then housing and road and railway networks needs will look very different from how they do now.

Large metropolitan areas as well as provincial cities and rural villages may also be transformed, for example, with new businesses being established to serve former commuters. Nevertheless, there is research showing that many will still choose to live in cities, even if they work from home, given the amenities offered (Ahlfeldt et al, 2020). Others argue that the largest cities and more rural areas will continue to flourish, but smaller cities will suffer (Delventhal et al, 2020).


One of the lasting consequences of the pandemic is an increase in the amount of time that we spend working remotely rather than in the office. This zoomshock will have permanent effects. Their size is important: aside for the consequences for the locally consumed services industry, it will impinge on the need for infrastructure, from office buildings to transport links. It is thus crucial to understand better which industries and neighbourhoods are most likely to embrace remote working to estimate long-term zoomshocks. This will help to guide a place-based recovery policy.

Where can I find out more?

Who are experts on this question?

Authors: Gianni De Fraja, Jesse Matheson and James Rockey
Photo by Chris Montgomery on Unsplash
Recent Questions
View all articles
Do you have a question surrounding any of these topics? Or are you an economist and have an answer?
Ask a Question
Submit Evidence