the impact of covid 19 on airbnb case study

Travel restrictions in place, cancellations increased and therefore occupancy down. Hosts are suffering from minimal income from their properties and Airbnb is suffering from a lack of commission from these bookings. This case study looks at how the COVID-19 pandemic is impacting Airbnb and assesses the company's response.

Key Highlights

• Airbnb is unique in that the travelers are not their only customers. Hosts use the Airbnb platform to advertise properties, and benefit from the awareness Airbnb has in the market. Different to Online Travel Agencies (OTA's) such as Booking.com or TripAdvisor, the hosts are in most cases individual people who are renting out their own homes, and therefore do not have the bargaining power, cash reserves or brand image that hotels would do on other OTA's.
• Hosts are welcoming a dramatic drop in guest numbers, and in turn not receiving any income from their properties. For hosts who rely upon Airbnb for their income, it poses a worry on being able to make mortgage payments, pay bills and survive themselves during the pandemic. Airbnb's current free cancellation period up until May 31st for bookings made on or before March 14th, mirrors that offered by hotels, Airbnb's indirect competitor. However, the hosts have to offer the refunds on this personally, and unlike hotels do not have the cash reserves and ability to do so.
• The scale and extent of Airbnb's handling of the COVID-19 pandemic should be carefully thought about, as each move they make will make a large difference to how Airbnb will operate after the height of the pandemic is over. Detrimental stories that have emerged in the press such as hosts offering COVID-19 Retreats' in the UK despite national lockdown rules and the backlash of troubles in obtaining refunds for stays and experiences, could leave a bad image of the brand in the future.
• This report provides insight into how COVID-19 is impacting Airbnb and looks at the affects the pandemic is having on Airbnb's relationship with both guests and hosts.
• It also analyzes the company's response to the current crisis.

Key report benefits:

• Gain an overview of the current global COVID-19 situation
• Understand the impact that COVID-19 is having on the lodging industry
• Assess the impact on Airbnb
• Understand what the future may hold for Airbnb

Key Topics Covered:

Current COVID-19 Overview

• Airbnb Overview
• Impacts on Airbnb
• Airbnb's Response
• SWOT Analysis
• Airbnb post-COVID-19

Companies Mentioned

For more information about this report visit https://www.researchandmarkets.com/r/h93voz

About ResearchAndMarkets.com

ResearchAndMarkets.com is the world's leading source for international market research reports and market data. We provide you with the latest data on international and regional markets, key industries, the top companies, new products and the latest trends.

ResearchAndMarkets.com Laura Wood, Senior Press Manager [email protected] For E.S.T Office Hours Call 1-917-300-0470 For U.S./CAN Toll Free Call 1-800-526-8630 For GMT Office Hours Call +353-1-416-8900

• Follow UNSW on LinkedIn
• Follow UNSW on Instagram
• Follow UNSW on Facebook
• Follow UNSW on WeChat
• Follow UNSW on TikTok

Disrupting the disruption: COVID-19 reverses the Airbnb effect

Sydney rents fell during the pandemic as investors swapped short-term holiday lettings for traditional rentals, UNSW research shows.

Media contact

Ben Knight UNSW Media & Content (02) 9065 4915 [email protected]

The meteoric rise of Airbnb across cities has disrupted rental housing markets worldwide as property owners took advantage of a new avenue for investment returns. But the tables might have turned. 

A recent UNSW City Futures Research Centre report assessing how Airbnb activity and the rental market has changed during the COVID-19 shows that it’s now renters who could be benefiting from a decline in Airbnb activity due to the pandemic.

The study,  Airbnb during COVID-19 and what this tells us about Airbnb’s Impact on Rental Prices , by professor of urban science Christopher Pettit and postgraduate researcher William Thackway , found that weekly rents declined in proportion to reduced Airbnb activity, as Airbnb landlords converted their properties to long-term rentals – at cut-price rates.

Airbnb during COVID-19

The researchers used a comprehensive record of Airbnb listings and rental sales data to find the supply of long-term rentals increased during the pandemic in historical Airbnb hotspots such as Bondi, Manly and the CBD.  Meanwhile, rental prices fell proportionately with Airbnb listings, up to 7.1% in the most active Airbnb neighbourhoods.

“Since the pandemic, with border closures and city lockdowns, particularly between March and May where Airbnb wasn’t in operation, there’s been subdued [Airbnb] activity, ”  Mr Thackway says.  “ We saw the reverse of what had been happening for the last 10 years, which is that many Airbnb’s were converted to long-term rentals, presumably by landlords who are now seeking a more stable income source, particularly given that Airbnb wasn’t even operating for a few months.” 

Airbnb traditionally receives a far higher daily rate than long-term rentals due to short-term tourism demand, which has previously motivated many landlords to invest in short-term rentals, Mr Thackway says.

“While there are other actors at play, there were ultimately fewer long-term rentals in the market because of Airbnb. The reduced supply of long-term rentals, without any difference in the demand, meant that overall rental prices have, until this point, been rising,” he says. 

“Now, when you get Airbnb’s converted back to long term rentals, there’s a new influx of supply to the rental market, and there has been a corresponding reduction in rental prices, and that’s been observed for almost all active Airbnb areas.”

According to the research,  Sydney had over 23,000 active Airbnb listings at its height. The Airbnb density measure used in the study also found that the proportion of Airbnb’s was as high as 25% in some areas such as Bondi.

“It’s not necessarily saying that those are active all the time, but it’s saying that 25% of the houses in the area listed are being booked out at some point during the year,” Mr Thackway says. 

Thinking long-term

While Airbnb’s effect on the rental market may have been reversed in the city, the researchers suggest this hasn’t been the case for coastal and regional areas.

“We suspect that regional areas are experiencing increases in tourist activity, mostly associated with domestic tourism, particularly urbanites wanting to get out of the city and holiday regionally,” Mr Thackway says.

“So, we would likely see those regional tourism hotspots having experienced an overall increase in Airbnb activity, and possibly rental prices, which would contrast with urban areas where Airbnb activity and tourist activity generally has fallen quite dramatically.”

While the temporary reversal of Airbnb activity on the housing market is a timely win for renters amid the housing affordability crisis, the researchers say it’s likely to revert to business as usual once international tourism returns.

“In terms of house prices and rents, Sydney is among the most unaffordable cities globally,” Mr Thackway says. “Reduced Airbnb and subsequently, reduced rental prices can only be a good thing for renters, and ultimately, for Sydney, because it is such an exclusive market and keeps tending towards that way.”

In NSW, there’s currently a 180-day cap on the number of days an Airbnb can be listed, while local governments and council can also exert additional control within the cap. But further regulations of short-term letting could be timely, Mr Thackway says.

“If the government wants to restrict its impact on renters, then tougher regulation, specifically on commercial Airbnb’s that permanently take away supply, would be necessary.”

More From Forbes

How airbnb beat the covid-19 virus.

• Share to Facebook
• Share to Twitter
• Share to Linkedin

POLAND - 2020/05/04: In this photo illustration an Airbnb logo seen displayed on a smartphone. ... [+] (Photo Illustration by Filip Radwanski/SOPA Images/LightRocket via Getty Images)

Going into 2020, Airbnb was growing at a rapid pace and aggressively expanding into new categories. So what could possibly go wrong?

Well, within a couple months, the Covid-19 pandemic would shutdown the travel industry. And suddenly Airbnb was facing an existential crisis. 

By April, the gross bookings for nights and experiences plunged by 72% on a year-over-year basis. In fact, from March to April, there were more cancellations than bookings!

While the situation was certainly dire, the management team wasted little time in pursuing a major restructuring.  The result was that—over time—the business started to improve. By June, there was actually a 1% increase in gross bookings.

Now this is not to say that the business is no longer under pressure. The fact remains that revenues are still well off the levels of 2019. But then again, Airbnb has weathered the pandemic relatively well compared to other major travel operators, whether hotel chains or online marketplaces. 

So why is this so? What did Airbnb do? Let’s take a look:

Cost Cutting : The mantra of Silicon Valley is high growth. But this philosophy makes no sense when the market is experiencing huge problems. 

In the situation of Airbnb, management initiated a layoff of 25% of the workforce. There was also a steep reduction in discretionary and capital expenditures, a slashing of executive salaries and a suspension of all facilities build-outs.

Best Tax Software Of 2022

Best tax software for the self-employed of 2022, income tax calculator: estimate your taxes.

Airbnb CEO and founder, Brian Chesky, set forth this plan in a letter that was sent to employees and published on the company blog . He did not mince words, saying “I have to share some very sad news.” He would go on to be thoughtful, transparent and clear about the “hard truths.”

Chesky also set forth a variety of principles about how the restructuring would be handled:  

• Map all reductions to our future business strategy and the capabilities we will need.
• Do as much as we can for those who are impacted.
• Be unwavering in our commitment to diversity.
• Optimize for 1:1 communication for those impacted.
• Wait to communicate any decisions until all details are landed — transparency of only partial information can make matters worse.

Focus On The Core Business : In the shareholder letter in the S-1, the founders noted: “When the pandemic hit, we knew we couldn’t pursue everything that we used to. We chose to focus on what is most unique about Airbnb—our core business of hosting. We got back to our roots and back to what is truly special about Airbnb—the everyday people who host their homes and offer experiences. We scaled back investments that did not directly support the core of our host community.”

This move proved critical. As the pandemic deepened, people were looking more at local stays—and this played to the advantage of Airbnb. It was one of the biggest factors in the turnaround of the business. 

Creatively-led : When there are severe constraints and limitations, this can lead to even more inspiration. This was certainly the case with the Airbnb team. For example, when the in-person Experiences segment was suspended, this led to the creation of Online Experiences. It turned out to be an extremely popular offering. 

Trust : This is the heart of any successful business. But during times of crisis, it can be tempting to make short-term financial decisions that ultimately undermine trust. 

An example of this quandary for Airbnb was about the spike in cancellations. Keep in mind that many were non-refundable. But to bolster the marketplace, Airbnb used more than $1 billion of its funding to provide refunds. There was also a commitment of up to $250 million for those hosts that were impacted by the cancellations. 

Long-Term Optimism : During the early days of the pandemic, it would have been easy to lose confidence in the vision of Airbnb. But the founders did not let this happen. They knew that the long-term prospects looked bright. 

According to the founders: “A crisis brings you clarity about what is truly important. You become thankful for not only what you have in your life, but for who you have in your life. We are thankful for everyone who stuck by us during our darkest hours.”

Tom ( @ttaulli ) is an advisor/board member to startups and the author of Artificial Intelligence Basics: A Non-Technical Introduction and The Robotic Process Automation Handbook: A Guide to Implementing RPA Systems . He also has developed various online courses, such as for the COBOL and Python programming languages.

• Editorial Standards
• Reprints & Permissions

• International edition
• Australia edition
• Europe edition

Disrupting the disruptors: how Covid-19 will shake up Airbnb

Airbnb created an industry and changed the face of many neighbourhoods. Now it’s facing the challenge of the coronavirus

A irbnb was built on the premise of bringing the world closer together. Tourists could travel like locals, while locals could cash in on their desirable neighbourhood properties by letting those visitors in. Last year the company was estimated to be worth more than US$30bn . It is scheduled to go public in 2020. Then came the Covid-19 pandemic.

Travel is suspended. Australians are almost entirely confined to their homes. Now the once heralded disruptor is seeing a collapse in bookings . The hosts who have become reliant on income-generating properties to pay their bills are being bled dry by a lack of business, and already-suspicious neighbours are up in arms over the potential that short-term renters may spread the virus.

In the decade before the pandemic, Airbnb became “this very attractive thing”, says Chris Martin, a senior research fellow at the University of New South Wales’s City Futures Research Centre. The platform, and others like it, fundamentally changed not only the travel industry but housing and rental markets. A home is no longer just a home, Martin says. “[The company] tapped into the idea that a person’s house can also be viewed as an asset that has capacity for generating income.”

This shift “massively upped the scale of [short-term rental]”, which in turn put pressure on residential rental markets. The changes have resulted in lawsuits in the US and a petition by 10 cities to the European parliament in 2019. There are more than seven million Airbnb listings spread across 100,000 cities – with an estimated 346,581 listings between July 2016 and February 2019 in Australia alone .

‘The whole industry has fallen through’

Stephen Colman, who was an Airbnb host and ran an Airbnb management business up until last year, says “the whole industry has fallen through”. Many hosts are either pulling out of Airbnb to find cheaper long-term tenants or have been offering “14-day isolation suites”.

The former, Colman says, will come with a long wait and big rent cuts, as everyone else rushes to do the same. The latter is even more fraught. This week the Greens state MP for Ballina, Tamara Smith, called on online holiday booking websites such as Airbnb and Stayz in the Byron Bay region to stop marketing regional properties as ideal places for self-isolation, saying it wasn’t fair on regional hospitals and communities.

“There are real concerns around cleaning stuff,” Colman says, “How many hours between checkout and check-in … [Hosts] have seen a fightback from body corporates who are just outright cancelling the key fobs for anyone who they believe is doing Airbnb. They’ve just gone ‘we’re going to take this hard line to protect the safety of our long-term tenants’.”

Jane Hearn, deputy chair of the Owners Corporation Network, has argued that opening Airbnb properties for quarantine “increases the viral load on apartment owners and tenants”. Resident advocacy groups such as We Live Here and Neighbours Not Strangers were already lobbying against Airbnb on behalf of local communities who were sick of rowdy travellers in their apartment complexes. In a letter to New South Wales premier Gladys Berejiklian on 1 April, Neighbours Not Strangers called for an immediate ban on all short-term rentals, due to the risks from Covid-19.

Airbnb does not support reservation requests from users who are showing symptoms, or those who are awaiting test results. Last week the company instituted a ban on any listings that “reference Covid-19, coronavirus or quarantine” and listings which “incentivise bookings through Covid-19-related discounts, stocks of limited resources, or the highlighting of quarantine-friendly listing attributes”. The company’s updated instructions for cleaning and hygiene recommend hosts stock their properties with “a few extras” like “antibacterial hand sanitizer, disposable gloves and wipes, hand soap, paper towels, tissues” … and toilet paper.

The government has now mandated that all international arrivals must complete their quarantine at designated pre-booked hotels . But, before that, some Australians were happy turning to Airbnb. Comedian Alice Fraser is currently in a “good value” Bondi Airbnb after returning from London. She saw Airbnb as “the responsible option”: “[It’s there] for people who don’t have favours to call in, or family who happen to have a massive home that can be segmented into parts.”

‘We lost everything’

Lisa Porgazian and her husband have listed their three Gold Coast apartments on Airbnb for the past four years. Now the properties are empty, and the mortgage payments will come out of the couple’s superannuation. “We were relying on this for our income, as well as our retirement plan,” she says, distraught. “Now that’s completely died in the arse.”

Porgazian, a 46-year-old former IT contractor, has been managing her property portfolio full-time. She’s unable to work in many other jobs due to rheumatoid arthritis. “I’m earning zero. My husband’s earning zero. And we’ve still got these mortgages to pay.”

With the spread of Covid-19, a downturn in business was inevitable. But many Airbnb hosts were shocked at how quickly it came. The company introduced a policy earlier this month allowing all guests who booked prior to 14 March (and were checking in no later than 31 May) to cancel existing bookings for free. Porgazian says this left hosts holding the cheque.

“We lost everything straight away … Everything is cancelled, basically until Christmas.”

Susan Wheeldon, Airbnb’s country manager for Australia and New Zealand, said offering these free cancellations “wasn’t an easy decision”, but it was one made with “public health considerations” front of mind. “The primary consideration for us was protecting the wellbeing of the community.”

‘There’s not enough tenants to fill these places’

Travis Lipshus, a real estate agent in Byron Bay, thinks this chaos for Airbnb hosts could result in cheaper long-term rent for locals. He’s getting flooded with properties from Airbnb hosts who now need permanent tenants. But with so many of the town’s hospitality staff and backpackers currently out of work “there’s not enough tenants to fill these places”.

If rents did lower, it would be a massive relief. It’s notoriously difficult for locals to find affordable rentals in Byron Bay, as 17.6% of properties are listed as holiday accommodation. “Airbnb should be banned up here,” Lipshus says. “ The cost of living is insane. I’ve lived in all sorts of places here, and it’s not uncommon to pay at least 50% of your wage in rent.”

Martin describes the current regulation around Airbnb as “really quite liberal”. He says at present, it “didn’t seem to fit well with highly local impacts. Local councils [should] have a strategy and a plan around short-term letting – especially around limits.”

For now, he is not so certain we’ll see a drop in rents. “Landlords may still withhold properties and leave them vacant instead,” he says. “For whatever reason, that seems to be a surprisingly common scenario in the high-pressure [locations].”

Wheeldon says that Airbnb “has not seen a material drop in the overall number of listings on our platform. While the Covid-19 crisis has significantly disrupted the tourism industry and wider economy, we know that travel is resilient in the long term and will ultimately recover.”

But will the hosts recover?

“We know our hosts are doing it tough,” Wheeldon says. The peak body representing Airbnb and other short term rental companies has sought urgent support from the federal government for hosts in the form of mortgage payment deferrals, among other measures. For many, that may not be enough.

“A lot of people have over-leveraged themselves with these properties,” Lipshus says. “The upper-middle class are probably going to be fine. But the middle class – the ones taking risks, trying to get up that class ladder – they’re going to be pretty fucked.”

Lisa Porgazian knows what people are saying. “The criticism that we’re getting is ‘well you shouldn’t have a business if you can’t pay for it’, but who ever predicted something like this?”

• Coronavirus
• Australia travel blog
• Australia holidays
• New South Wales holidays
• Australian lifestyle

Most viewed

Advertisement

Supported by

The Future of Airbnb

Home-sharing’s challenges aren’t only about social distancing and hygiene. Overtourism, racial bias, fee transparency and controlling the party crowd are also in the mix.

• Share full article

By Elaine Glusac

In the travel wreckage caused by the pandemic, home-sharing has emerged as battered, but with a steady pulse, as rental houses became social-distancing refuges for the travel-starved.

Home rentals have outperformed hotels in 27 global markets since the onset of Covid-19, according to a report by the hotel benchmarking firm STR and the short-term rental analysts AirDNA. As leisure travel ticked up this summer, average daily rates were higher for rentals in July 2020 versus July 2019 in the United States — from about $300 to $323 — thanks to the popularity of larger homes.

Still, global restrictions have squeezed every aspect of the travel industry, including vacation rentals. Across home-sharing platforms, according to STR and AirDNA, occupancy fell by almost half between mid-March and the end of June to between roughly 33 and 36 percent, depending on the size of the rental (hotels by comparison fell to an average of 17.5 percent occupancy).

The biggest player in the short-term rental market, with more than 7 million listings in over 220 countries, is Airbnb . Over the years, its rampant growth and lack of transparency have made it a target for everything from charges of fueling overtourism and turning formerly residential neighborhoods into tourist zones to enabling raucous parties despite complaints and virus-related restrictions on gatherings.

After laying off a quarter of its work force in the spring, Airbnb jettisoned some new ventures, including forays into transportation and entertainment, and hunkered down to focus on its core strength, lodging, even as its valuation fell from a high of $31 billion to, recently, $18 billion, according to The Wall Street Journal.

Now, as Airbnb prepares to go public , we talked to Airbnb’s co-founder and chief executive, Brian Chesky, along with other industry experts, about some of the company’s challenges and the ways it is changing travel.

“People want to travel, they just don’t want to get on airplanes,” Mr. Chesky said. “They don’t want to go for business. They don’t want to stay in the really big cities as prevalently as they used to. They don’t want to be in crowded hotel districts.” But, he said, “they do want to get out of the house. And so we think demand is going to be strong in the future. I’m very optimistic, actually, about the industry.”

Lifestyle vs. vacation

Airbnb has touted privacy and guests’ control over their environment — including having your own kitchen in lieu of patronizing restaurants — as safeguards during the pandemic. It instituted new cleaning guidelines and indicated in late August that more than 1 million listings had earned the “ Enhanced Clean ” certification, which involves training in new guidelines that detail how and what to wash and sanitize. The procedures recommend 45 minutes of cleaning per room. Some listings guarantee a 72-hour vacancy window before check in.

The company says its offerings are aligned with the way people are traveling now, in family and friend groups to less populated destinations. Over Labor Day weekend, 30 percent of its bookings — double the previous year — were in remote areas, though classic vacation spots like Hilton Head Island, S.C., and Palm Springs, Calif., were among the most popular. Urban bookings remain down.

“We’re seeing a little blurring between traveling and living,” Mr. Chesky said. “Before the pandemic, you lived somewhere 50, 51 weeks of the year, and if you were so fortunate, you’d go on your once-or-twice-a-year vacation. Now the pandemic is changing how people want to work, travel and live.” Remote school and work unbind families from their homes. “People are living differently and people want to live anywhere,” he added.

Whether travel truly turns into nomadism remains to be seen, though the average length of stay since May 1 increased 58 percent to more than four days, and fall bookings are stronger than usual, according to AirDNA.

Overtourism, rising rents and housing shortages

Cities around the world, from Barcelona to Vancouver, are looking to curb Airbnb and other short-term rental companies, which many blame for hollowing out neighborhoods as real estate managers took long-term leases and listed them as more lucrative short-term rentals.

“You can earn more renting out apartments and houses on Airbnb than renting to locals,” said David Wachsmuth, an associate professor in the School of Urban Planning at McGill University in Montreal. “What’s happened on their platform is that actual home-sharing is a fraction of the activity. It’s dominated by commercial interests.”

Research published in the Harvard Business Review found that as listings rise in a city, so do rents. Analyses by the Economic Policy Institute , a nonpartisan think tank, found the costs to local communities of having Airbnb listings, including rising housing prices and shrinking availability, likely outweigh the benefits.

“The problems of overtourism were in the making for a long time,” said Makarand Mody, an assistant professor of marketing in the School of Hospitality Administration at Boston University. “Airbnb came along and made it worse. It was seen as one evil that needs to be sorted out, but there are much deeper societal and economic issues. Airbnb is just the supply side. But demand has increased so much.”

By 2019, the rise of the middle class globally contributed to expanding tourism above the rate of worldwide economic growth for nine years in a row, according to the World Travel & Tourism Council. In Airbnb, many travelers found affordable accommodations that allowed them to stay in neighborhoods rather than business centers.

Now that the pandemic is the ultimate overtourism disrupter, Mr. Chesky believes travel has been redistributed in a lasting way to places beyond bucket-list capitals. “It’s kind of redeemed our vision,” he said. “What I would love is to be able to help spread out travel to as many communities as possible rather than over-concentrating them in any one place.”

“My speculation is that the world does not quickly snap back to the way it was,” he added. “I don’t think travel will ever, ever look like it did in January. The world can’t change so dramatically like it has and then one of the industries that’s been hit hardest just looks exactly like it did before.”

Communities aim to ensure that. Last summer, Oahu enacted a law to restrict rentals without permits on the Hawaiian island, enforced with fines. In Europe , cities like Lisbon and Dublin are buying back leases or forcing landlords into long-term rentals in an effort to ensure that when tourism rebounds it won’t overwhelm them again.

Enforcement remains thorny, and Airbnb has been accused of looking the other way when it comes to illegal listings. Last year, Los Angeles limited rentals to owner-occupied properties registered with the city, though many illegal units remain on the site, according to the Los Angeles Times .

In response, Airbnb just launched a new City Portal that it says will allow governments to more easily identify listings that don’t comply with local regulations, such as unregistered listings.

Before the launch, the company shared the new tool with San Francisco’s Office of Short-Term Rentals . “They’re pretty positive about it and hopeful this will definitely improve their ability to get bad actors off the platform,” said Jeffrey Cretan, a spokesman for the city’s mayor.

Perhaps because of these scofflaws, Airbnb says it has not lost significant listings. According to AllTheRooms Analytics, among popular cities in Europe, only Rome and Lisbon have shed listings, about 2,000 each. In Lisbon, the crackdown still leaves just above 14,500 listings, the same figure as in January 2019, but down from the peak in July 2019.

The effect of more regulations may show up in the future, posing a threat to a robust portfolio. “For a platform like Airbnb, they’re not just worried about the demand side, but the supply side,” Mr. Mody, of Boston University, said, noting the travel freeze may convince hosts to put their units in the long-term rental market, shrinking the platform, and worrying potential investors. “When you’re living on venture capital, profitability is not as important as growth,” Mr. Mody added. “Shareholders will be a lot less patient.”

The music stops at party houses

During the pandemic, Host Compliance , which tracks legal compliance among short-term rentals for 350 cities and counties in the United States, said noise complaints about so-called “party houses” tripled.

“A lot of people have been at home for a long time and they have to let some steam off and can’t jump on a plane to go to Europe or Cancún to party so they are renting out short-term rentals in driving distance from their homes,” said Ulrik Binzer, the founder and general manager of Host Compliance.

Often, these rentals are in residential neighborhoods, triggering noise complaints and health concerns about large gatherings.

In Miami Beach, short-term rentals were closed this summer, though those within condo and apartment buildings were allowed to reopen, with capacity limits, in August. That month, the city of Los Angeles cut the power on a house (not an Airbnb property) rented by prominent TikTok stars during a large party.

In August, Airbnb pulled the plug, too, announcing a global ban on party houses , defined as those that persistently generate complaints from neighbors. The company says 73 percent of listings already ban parties, though hosts often allow small gatherings like baby showers and birthday parties. Occupancy is now limited to 16 people.

Airbnb imposed a similar restriction in Canada earlier this year after a party in Toronto ended in three shooting deaths, according to BBC News .

“We want to do everything we can do to preserve the character of the communities and not allow these parties to get out of hand,” Mr. Chesky said.

It’s too soon, say observers, to know if the ban is working.

“The issue with Airbnb party houses is enforcement,” Mr. Binzer said. “It’s a little like having the fox watch the henhouse.”

The $52 rental with a $125 cleaning fee

On Sept. 14, a Twitter user wrote , “Found a cheap @Airbnb for 52 dollars. Cleaning fee for 1 night, 125. Nonsense.”

It’s a typical complaint about the platform, which lists attractive nightly rates, but buries the fees until users begin booking. Cleaning and service fees can be modest — zero to $25, say — or add $450 to a booking, reflecting a mix of mandatory and optional host-applied fees. Sometimes there are additional occupancy taxes. And in some countries, Airbnb applies a Value Added Tax on its service fees.

Under Airbnb’s pricing structure , hosts pay the company 3 percent of the booking subtotal, which includes the nightly rate plus any cleaning fee and fees for additional guests. Most guests are charged a service fee of less than 14.2 percent of the booking subtotal, which goes to Airbnb. (If hosts elect to cover the fee entirely, they normally pay Airbnb 14 to 16 percent of the subtotal.)

Because of their variability and lack of transparency, fees are the latest financial facet users have fixated on after the company created its extenuating circumstances policy during the pandemic. It said that travelers who had reservations made on or before March 14 could cancel and not be subject to cancellation fees, even if, in their rental agreement, they were in the penalty period. The policy has been extended several times, now to Oct. 31. (While most guests were happy with the resolution, many hosts were not and Airbnb later apologized to hosts for not consulting them).

Airbnb said it aims to introduce a redesign of price displays this year. “We’re trying to partner with hosts to create clear standards and change the search line, so if someone has higher cleaning fees, that affects their placement” in search results, Mr. Chesky said. “We’ve heard from travelers that they want a simpler way for us to show more of the price up front.”

Reckoning with racial bias

Four years before George Floyd was killed by police in Minneapolis, igniting this summer’s protests for social justice, the emergence of the hashtag # AirbnbWhileBlack called attention to a spate of racist incidents that users said happened at rental homes. Some Black renters were reported by neighbors as thieves. Others were subject to abuse by racist hosts rejecting their bookings. Complaints by Muslim, transgender and other groups followed.

Airbnb worked to purge discrimination from its platform by hiding guest’s profile pictures until a booking is confirmed; hiring anti-discrimination specialists to audit the platform; and creating a reporting channel to identify listings not complying with its nondiscrimination policy . The company said it has removed 1.3 million offenders.

This month, Airbnb plans to launch Project Lighthouse , a research initiative in the United States that aims to measure bias through perception based on names and photos, to determine where and when bias happens on the platform, from booking through reviews.

According to the company, the study has been in the works for two years in partnership with the racial justice organization Color of Change , with input from several social justice nonprofits, including Asian Americans Advancing Justice and the National Association for the Advancement of Colored People .

“It’s really hard to change what you can’t measure,” Mr. Chesky said. “Then hopefully we will use this data to continue to evolve our platform and reduce the bias.”

Its tech focus — on the platform rather than the in-person experience — won’t address incidents of in-person bias. Through its existing Open Doors policy , Airbnb offers to find a guest an alternative place to stay if they feel they have been discriminated against by a host.

“In a departure from its peers in Big Tech who pass off structural problems on the behavior of individual users, with Project Lighthouse, Airbnb is attempting to take responsibility for how tech platforms create the opportunity for harm at scale,” wrote Jade Magnus Ogunnaike, senior campaigns director at Color Of Change, in an email.

Experiences in the digital age

Airbnb doesn’t just rent lodgings. Through its Airbnb Experiences branch, it offers classes in mole making with an Indigenous cook in Mexico City, a music and cultural tour of Havana with a D.J. and walks among penguins with a conservationist in South Africa.

During the pandemic, many of its Experiences went virtual . Now, via Zoom, armchair travelers can visit an animal rescue farm in Connecticut, follow a plague doctor through Prague and sit in on a songwriting session in Nashville.

After Airbnb’s layoffs, many wondered whether Airbnb Experiences, long rumored to be losing money, would be shelved, too. In January , it had 50,000 Experiences in 1,000 cities. During the pandemic, the division was shut down, and later transitioned, with a fraction of its offerings, online. Today, it has 700 virtual Experiences generating $2 million in bookings over the past five months. In-person Experiences have resumed in more than 70 countries with restrictions on group sizes, though the company declined to say how many Experiences are available in person and how much money they are making.

“I would be surprised if they drop it completely,” Mr. Mody, of Boston University, said. “They don’t want to be just a home rental company. Travel is about experiencing the destination in its entirety and they want to play a role in that.”

The company said it stands by Experiences, even waiving its take — which is normally 20 percent — for its Social Impact Experiences , which include playing with shelter cats in Osaka, Japan ($25) and learning beat-making with an organization devoted to teaching underserved youth ($75).

“Experiences was hit hard by social distancing,” Mr. Chesky said, maintaining that the online transition has been successful. “In a world where there’s not a lot of things to do, we think there’s a window for Airbnb Experiences,” he said.

Follow New York Times Travel on Instagram , Twitter and Facebook . And sign up for our weekly Travel Dispatch newsletter to receive expert tips on traveling smarter and inspiration for your next vacation.

• Skip to primary navigation
• Skip to main content

UT Urban Information Lab

The Urban Information Lab uses emerging information technologies to better understand, measure, plan, and develop our urban environments.

April 7, 2024 , Filed Under: Projects

Cities reshaped by Airbnb: A case study in New York City, Chicago, and Los Angeles

• The Director
• News & Events
• Digital twin
• Miscellaneous

Become a member today, It's free!

Remember me

Send my password

Case study on the impact of covid-19 on airbnb - researchandmarkets.com.

The "The impact of COVID-19 on Airbnb: Case Study" report has been added to ResearchAndMarkets.com's offering.

Travel restrictions in place, cancellations increased and therefore occupancy down. Hosts are suffering from minimal income from their properties and Airbnb is suffering from a lack of commission from these bookings. This case study looks at how the COVID-19 pandemic is impacting Airbnb and assesses the company's response.

Key Highlights

• Airbnb is unique in that the travelers are not their only customers. Hosts use the Airbnb platform to advertise properties, and benefit from the awareness Airbnb has in the market. Different to Online Travel Agencies (OTA's) such as Booking.com or TripAdvisor, the hosts are in most cases individual people who are renting out their own homes, and therefore do not have the bargaining power, cash reserves or brand image that hotels would do on other OTA's.
• Hosts are welcoming a dramatic drop in guest numbers, and in turn not receiving any income from their properties. For hosts who rely upon Airbnb for their income, it poses a worry on being able to make mortgage payments, pay bills and survive themselves during the pandemic. Airbnb's current free cancellation period up until May 31st for bookings made on or before March 14th, mirrors that offered by hotels, Airbnb's indirect competitor. However, the hosts have to offer the refunds on this personally, and unlike hotels do not have the cash reserves and ability to do so.
• The scale and extent of Airbnb's handling of the COVID-19 pandemic should be carefully thought about, as each move they make will make a large difference to how Airbnb will operate after the height of the pandemic is over. Detrimental stories that have emerged in the press such as hosts offering COVID-19 Retreats' in the UK despite national lockdown rules and the backlash of troubles in obtaining refunds for stays and experiences, could leave a bad image of the brand in the future.
• This report provides insight into how COVID-19 is impacting Airbnb and looks at the affects the pandemic is having on Airbnb's relationship with both guests and hosts.
• It also analyzes the company's response to the current crisis.

Key report benefits:

• Gain an overview of the current global COVID-19 situation
• Understand the impact that COVID-19 is having on the lodging industry
• Assess the impact on Airbnb
• Understand what the future may hold for Airbnb

Key Topics Covered:

Current COVID-19 Overview

• Airbnb Overview
• Impacts on Airbnb
• Airbnb's Response
• SWOT Analysis
• Airbnb post-COVID-19

Companies Mentioned

For more information about this report visit https://www.researchandmarkets.com/r/h93voz

About ResearchAndMarkets.com

ResearchAndMarkets.com is the world's leading source for international market research reports and market data. We provide you with the latest data on international and regional markets, key industries, the top companies, new products and the latest trends.

View source version on businesswire.com: https://www.businesswire.com/news/home/20200616005562/en/

ResearchAndMarkets.com Laura Wood, Senior Press Manager [email protected] For E.S.T Office Hours Call 1-917-300-0470 For U.S./CAN Toll Free Call 1-800-526-8630 For GMT Office Hours Call +353-1-416-8900

Related News

@ the bell: the market sell-off continues, @ the bell: market drop and overseas tensions hurt tsx, recent u.s. press releases, update: new aurinia presentation details at the 2024 bloom burton & co. healthcare investor..., coursera receives industry-first authorized instructional platform designation from the american..., philip morris international to host webcast of 2024 first-quarter results, featured news links, this gambling tech stock is future-proofing the world’s casinos and has no direct competition, q precious & battery metals contracts first class drilling for quebec projects, why the tonopah gold project is one to watch, discover your next investment at this free online conference, targeting nevada and ontario for the critical minerals supply chain, asset sale allows biotech company to refocus on the high value dermatological market.

Get the latest news and updates from Stockhouse on social media

Follow STOCKHOUSE Today

• Open access
• Published: 12 April 2024

Estimated public health impact of concurrent mask mandate and vaccinate-or-test requirement in Illinois, October to December 2021

• François M. Castonguay 1 , 2 , 4 ,
• Arti Barnes 3 ,
• Seonghye Jeon 1 , 2 ,
• Jane Fornoff 3 ,
• Bishwa B. Adhikari 1 , 2 ,
• Leah S. Fischer 1 , 2 ,
• Bradford Greening Jr. 1 , 2 ,
• Adebola O. Hassan 3 ,
• Emily B. Kahn 1 , 2 ,
• Gloria J. Kang 1 , 2 ,
• Judy Kauerauf 3 ,
• Sarah Patrick 3 ,
• Sameer Vohra 3 &
• Martin I. Meltzer 1 , 2  

BMC Public Health volume  24 , Article number:  1013 ( 2024 ) Cite this article

201 Accesses

2 Altmetric

Metrics details

Facing a surge of COVID-19 cases in late August 2021, the U.S. state of Illinois re-enacted its COVID-19 mask mandate for the general public and issued a requirement for workers in certain professions to be vaccinated against COVID-19 or undergo weekly testing. The mask mandate required any individual, regardless of their vaccination status, to wear a well-fitting mask in an indoor setting.

We used Illinois Department of Public Health’s COVID-19 confirmed case and vaccination data and investigated scenarios where masking and vaccination would have been reduced to mimic what would have happened had the mask mandate or vaccine requirement not been put in place. The study examined a range of potential reductions in masking and vaccination mimicking potential scenarios had the mask mandate or vaccine requirement not been enacted. We estimated COVID-19 cases and hospitalizations averted by changes in masking and vaccination during the period covering October 20 to December 20, 2021.

We find that the announcement and implementation of a mask mandate are likely to correlate with a strong protective effect at reducing COVID-19 burden and the announcement of a vaccinate-or-test requirement among frontline professionals is likely to correlate with a more modest protective effect at reducing COVID-19 burden. In our most conservative scenario, we estimated that from the period of October 20 to December 20, 2021, the mask mandate likely prevented approximately 58,000 cases and 1,175 hospitalizations, while the vaccinate-or-test requirement may have prevented at most approximately 24,000 cases and 475 hospitalizations.

Our results indicate that mask mandates and vaccine-or-test requirements are vital in mitigating the burden of COVID-19 during surges of the virus.

Peer Review reports

Introduction

Public health mandates or requirements are laws put into place to promote healthy behaviors that mitigate disease burden. The mechanisms through which these policies interact with public health can vary substantially. For instance, a mask mandate aims to reduce disease transmission and a vaccinate-or-test requirement aims to reduce exposure to infectious cases through testing and also reduce the incidence of severe disease and mortality through vaccination. Several studies have estimated the impact of face mask mandates [ 1 , 2 , 3 , 4 ] and vaccination requirements [ 5 , 6 ] on COVID-19 transmission.

Facing a surge in COVID-19 cases during the fall of 2021 due to the emergence of the Delta variant [ 7 ], the Governor of Illinois issued an executive order on August 26, 2021 [ 8 ] that required workers in certain frontline professions ( e.g. , healthcare workers, school personnel, higher education personnel, and state owned or operated congregate facilities) to get vaccinated against COVID-19 (i.e., at least receive the first dose of a two-dose COVID-19 vaccine series), or undergo, at a minimum, weekly testing for COVID-19 if they remained unvaccinated by September 19, 2021 [ 9 ]. At the latest, individuals had to receive their second dose of a two-dose COVID-19 vaccine series 30 days after the September 19 th deadline [ 9 ]. There was also a mandate that required face mask use for any individual in a public indoor setting, beginning August 30, 2021, regardless of vaccination status. The August 26, 2021 executive order [ 8 ] was issued while the rate of cases continued to rise steeply, despite an earlier mask mandate for schools, daycares, and long-term care facilities enacted on August 4, 2021 [ 10 ]. The Delta variant represented more than 99% of sampled strains in Illinois at that point [ 7 ] and healthcare capacity was strained.

In this paper, we provide estimates of cases and hospitalizations averted by the concomitant mask mandate and vaccinate-or-test requirement in Illinois between October 20 and December 20, 2021. We estimate separately the impact of increases in masking and vaccination. We adjust for a wide range of compliance levels by estimating the impact for various levels of mask-wearing pre- and post-mandate, and for various potential levels of vaccination uptake in the front-line workforce that would follow the announcement of the vaccinate-or-test requirement.

We estimated the impact of increases in masking and vaccination on COVID-19 incidence in Illinois from October 20 to December 20, 2021. Previous reports have shown that use of face masks reduces SARS-CoV-2 transmission [ 11 , 12 , 13 ]. We simplified estimating the impact of mask wearing, hereby referred to as mask effectiveness, by using (i) an average estimate of mask efficacy and (ii) an average percent of population compliant with correct mask wearing (see Technical Appendix for more details). Footnote 1 For vaccines, the effectiveness of the vaccinate-or-test requirement depends on the difference between the percentage of population that complied with the executive order to vaccinate-or-test, and the percentage that were already vaccinated. This research activity was reviewed by the Centers for Disease Control and Prevention (CDC) and was conducted consistent with applicable federal law and CDC policy. Footnote 2

We used CDC’s COVIDTracer modeling tool [ 14 ] to build an epidemic curve that mimicked the observed one in Illinois over the two-month period (October 20 - December 20, 2021). Footnote 3 Similar to other studies [ 15 , 16 , 17 , 18 ] (Castonguay FM, Borah BF, Jeon S, Rainisch G, Kelso P, Adhikari BB, Daltry DJ, Fischer LS, Greening Jr B, Kahn EB, Kahn GJ, Meltzer MI: The Public Health Impact of COVID-19 Variants of Concern on the Effectiveness of Contact Tracing in Vermont, United States, unpublished), the two-month duration balances the need for sufficient time to pass after the start of the mandate to allow for an adequate assessment of the impact of the interventions being studied. Simultaneously, it aims to limit the potential for unknown confounding factors that may alter the impact of the interventions. We assumed that the effectiveness of interventions remained constant over the two-month study period. Those vaccinated following the vaccinate-or-test requirement should have achieved full vaccine-induced immunity Footnote 4 by mid-October 2021, which matches the start of our analytic timeframe. Footnote 5 COVIDTracer uses a compartmental Susceptible–Exposed–Infectious–Recovered (SEIR) mathematical model [ 19 ]. A user enters location-specific COVID-19 case counts, vaccination levels, a set of parameters describing COVID-19 epidemiology ( e.g. , basic reproduction number), and estimates of the effectiveness of the interventions (Table 1 ) (see Technical Appendix for details and Appendix Table A4 for a list of Illinois-specific inputs).

Impact of mask mandate

To model the impact of mandate-induced increased mask effectiveness, we first inputted a selected value from the range provided in Table 1 into COVIDTracer. The pre-mandate mask effectiveness range was 3.6%–16.8% and post-mandate mask effectiveness range of 6.1%–23.3%. For baseline analysis, we used post-mandate mask effectiveness of 14.2%, assuming 20% mask efficacy and 71% compliance. As it is very difficult to measure the degree of compliance with effective mask wearing in any population, we therefore constructed 24 scenarios of combinations of pre-and post-mandate mask effectiveness. We avoided over-estimating the impact of the mask mandate by using pre-mandate mask effectiveness values of less than 20% and only one post-mandate mask effectiveness value of over 20% (see Sensitivity Analyses).

We then “fitted” the curve of cumulative cases modeled by COVIDTracer to the jurisdiction’s reported cases by altering the percentage reduction in transmission ascribed to vaccine and various Non-Pharmaceutical Interventions (NPIs). The estimated percentage reduction in transmission that minimized the difference ( i.e., minimized the mean squared error) between the fitted and reported cumulative case curves is the estimate of the effectiveness of non-CICT NPIs (see Appendix for further details). Note that, because there were no measurements of the degree of under-reporting, we had to use the reported cases without any adjustments for possible under-reporting. Correcting for under-reporting may well have increased the estimates of cases and hospitalizations averted, for both mandates. Finally, we simulated what would have occurred without the mask mandate by re-setting the impact of mask effectiveness to one of 2 pre-mandate effectiveness levels (7.2% and 11.2%). The resulting plots are the number of cases that would have occurred without the mask mandate. The difference between the hypothetical plots of cases without mandate-induced increases in mask effectiveness and the plot of the reported cases (which includes the impact of mask mandate) are the cases averted due to the mask mandate. By “fitting the curve” (finding the best match between the SEIR model and the observed data), this methodology prevents over- or under-estimating the combined impact of all interventions.

Sensitivity analyses: mask mandate

Mask wearing compliance depends on several locality factors [ 32 , 33 ]. As noted earlier, it is very difficult to measure the degree of compliance in large populations. We therefore constructed 24 scenarios of combinations of pre-and post-mandate mask effectiveness (Appendix Table A1 ).

Impact of vaccinate-or-test requirement

To estimate the impact of the vaccinate-or-test requirement, we followed the same process as described above for masking (Table 1 ; see the Appendix for further details). The number of cases averted by the vaccinate-or-test requirement depended on (i) the baseline, pre-requirement, vaccination coverage, and (ii) the increase in vaccination coverage attributable to the requirement. We obtained an estimate of 929,370 frontline workers in Illinois who were potentially affected by the vaccinate-or-test requirement (see Appendix Table A2 ). The National Healthcare Safety Network (NHSN) reported that, for the period analyzed, vaccination coverage of the staff working in certified Centers for Medicare & Medicaid Services (CMS-certified) nursing homes in Illinois increased from 64.8% to 75.9%––an 11.1 percentage point increase in coverage (see Appendix Table A3 ). We used this 11.1 percentage point increase (equivalent to 103,160 additional persons vaccinated) as a base case scenario for analyzing the impact of the vaccinate-or-test requirement.

Sensitivity analyses: vaccinate-or-test requirement

We do not know what proportion of the 11.1 percentage point increased coverage was due to the vaccinate-or-test requirement. Other factors, such as intent to be vaccinated regardless of the requirement or other requirements/mandates ( e.g. , the federal mandate for CMS facilities announced during the study period), could have contributed to the increase. To address this uncertainty, we evaluated the impact of an arbitrary assumption that only half of the recorded increase in vaccine coverage could be attributable to the mandate (i.e., a 5.6 percentage point increase, equivalent to 51,580 additional persons vaccinated). Note that the potential indirect impact that the vaccinate-or-test requirement could have had on the general population [ 6 ] is not accounted for, which may have increased the overall impact of the requirement.

In Fig. 1 , we present the plot of the cases assuming the post-mandate mask effectiveness of 14.2% (calculated assuming 20% mask efficacy and 71% compliance; represented by the solid black line in Fig.  1 ). This plot is then compared to the hypothetical plots of increased cases, assuming no mask mandate and a continuation of pre-mask effectiveness of either 7.2% or 11.2% (dotted and dashed lines). The cumulative difference between the dotted or dashed plotted lines and the solid line plot is the estimate of additional cases averted due to the mask mandate.

Fitted epidemic curves of COVID-19 case counts showing the impact of the mask mandate

Notes: Fitted epidemic curve of observed COVID-19 case counts, assuming a post-mandate mask effectiveness of 14.2%, and simulated epidemic curves assuming no mandate and continuation of pre-mandate mask effectiveness of either 7.2% or 11.2% (all three for the October 20 – December 20, 2021 period). The solid line is Illinois’s observed (fitted) cumulative COVID-19 case counts, and the broken lines are the simulated curves illustrating the cumulative total COVID-19 cases for the various scenarios that might have occurred if the mask mandate had not been enacted and mask efficacy was 20%. The differences between the solid and broken lines show the benefits of the mask mandate with greater divergence between the solid and broken lines indicating a greater impact. All results assume that the effects of nonpharmaceutical interventions —including masks—were constant over the two months shown

To calculate a lower-bound estimate of cases averted, we assumed that the mask mandate increased mask effectiveness from 3.6% to 6.1%. This resulted in an estimate of 149,817 additional cases and 3,028 additional hospitalizations averted due to the mask mandate (Table 2 ). We calculated an upper-bound estimate by increasing the mask effectiveness for pre- and post-mandate to 10.2% and 21.3%, respectively. This resulted in an upper estimate of 1,820,764 additional cases and 36,801 additional hospitalizations averted due to the mask mandate (Table 2 ). The remainder of the results in Table 1 show the estimates of cases and hospitalizations averted from another 10 scenarios. The results from all 24 scenarios of combinations of pre-and post-mandate mask effectiveness are presented in Appendix Table A1 .

Assuming that the vaccinate-or-test requirement resulted in an 11.1 percentage point increase in persons vaccinated, we estimated that 23,593 cases and 477 hospitalizations were averted (Table 3 and Fig.  2 ).

Fitted epidemic curves of COVID-19 case counts showing the impact of the vaccinate-or-test requirement

Notes: Fitted epidemic curve of observed COVID-19 case counts and of two assumed increases in vaccination coverage attributable to the announcement of the vaccinate-or-test requirement (for October 20 – December 20, 2021). These represent approximately 51,580 and 103,160 individuals vaccinated because of the vaccinate-or-test requirement. The solid line is Illinois’s observed cumulative COVID-19 case counts, and the dashed and dotted lines are the simulated curves illustrating the cumulative total cases for scenarios where there would have been a lower vaccine uptake without the vaccinate-or-test requirement. The differences between the solid and dashed or dotted lines show the number of cases averted by the vaccinate-or-test requirement. All results assume that the effects of other nonpharmaceutical interventions (NPIs) were constant over the two months analyzed

Sensitivity analyses: Vaccinate-or-test Requirement

When we assumed that only half of the increase was attributable to the vaccinate-or-test requirement, an estimated 11,571 cases and 234 hospitalizations were averted by the vaccinate-or-test requirement (Table 3 and Fig.  2 ).

We estimated that increases in masking following the announcement of the mask mandate may have averted at least 58,000 cases in Illinois. The vaccinate-or-test requirement among frontline workers averted up to 24,000 cases during the period studied. The assumed post-mandate mask effectiveness (6.1% - 21.1%) was the most influential variable assessing the impact of mask mandates.

Our results provide data-driven evidence that can inform decision-making regarding public health interventions during times of surge. While there have been concerns regarding the negative impact of such mitigation measures [ 34 ], mask mandates are impactful [ 35 ] and vaccination requirements have demonstrated the ability to strengthen vaccination intentions across racial and ethnic groups, and even those who may be resistant [ 36 ]. Public adherence to these practices due to requirements as opposed to free choice is a more complicated debate. Several studies have demonstrated that while vaccine mandates may result in some vaccine hesitancy, they have been associated with improved vaccination rates [ 37 , 38 ]. Hospital staff vaccination reports have found that many employees chose vaccination over resignation [ 37 ].

Our study has limitations. Several factors made it difficult to use a direct causal identification methodology, such as difference-in-differences. These factors included the absence of a credible control group and several confounding factors due to the important period between the announcement of the intervention and when compliance was required. We had to make several assumptions because the precise impact that the concomitant mask mandate and vaccinate-or-test requirement in Illinois had on COVID-19 burden depended largely on several unobserved factors, namely mask quality, level of mask-wearing pre- and post-mandate, and the proportion of vaccine uptake attributable to mandate. Further, while we estimated the impact of increases in masking and vaccination separately, there may have been unaccounted synergistic effects from combining both interventions [ 39 ]. We also assumed that the impact of face masks and vaccination remained constant over the period of study. To reduce the potential impact of such assumption, we limit our study period to two months. We do not account for partial immunity ( e.g., if an individual received their first vaccine shot during the study period), and hence assume individuals are either fully susceptible (because the individual was never vaccinated or never infected, or because immunity acquired through vaccination or prior infection was more than 180 days ago [ 29 ] and is no longer protective) or fully immune (due to prior infection or vaccination) during the two-month study period. By doing so, we may underestimate the impact of the policies (e.g., because the first dose of a two dose COVID-19 vaccine series may still provide some protection [ 40 ] which would increase the impact of the vaccinate-or-test requirement) or overestimate the impact of the policies (e.g., because those protected by the first dose of a two dose COVID-19 vaccine series would not have been protected by masks), with the overall direction of the bias being uncertain. There is also the possibility that the vaccinate-or-test requirement for frontline workers may have had an impact on the general population, as the requirement may have signaled the importance of vaccination for individuals not directly covered by the vaccinate-or-test requirement [ 6 ]. Finally, we assumed that there are no differences in disease transmission that are attributable to age, location, or occupation—in other words, every individual in the population is assumed to have the same risk of catching COVID-19 and is assumed to behave in the same way as any other individual. Those affected by the vaccinate-or-test requirement were frontline professionals who could have had potentially very different mixing patterns compared to the general population.

During the two-month study period, almost 2,000 hospitalizations were averted according to our model. Had these hospitalizations occurred, they would have had a significant impact on an already strained healthcare system. These findings can help control viral transmission of diseases other than COVID-19 at both hospital and community levels, and will help refine future decisions on the timing and scale of such public health measures should we find ourselves again in a similar healthcare crisis.

Availability of data and materials

No datasets were generated or analysed during the current study.

Mask effectiveness is the product of (i) mask efficacy and (ii) mask compliance. Mask efficacy is defined as the extent to which masks reduce the output and uptake of the virus droplets/aerosols (range 10%-30%, see Ueki et al. [ 11 ]) and mask compliance is defined as the percentage of the population properly wearing masks.

See e.g., 45 C.F.R. part 46, 21 C.F.R. part 56; 42 U.S.C. §241(d); 5 U.S.C. §552a; 44 U.S.C. §3501 et seq.

This modeling tool has been used by several studies to estimate the impact of CICT [ 15 , 16 , 17 , 18 ] (Castonguay FM, Borah BF, Jeon S, Rainisch G, Kelso P, Adhikari BB, Daltry DJ, Fischer LS, Greening Jr B, Kahn EB, Kahn GJ, Meltzer MI: The Public Health Impact of COVID-19 Variants of Concern on the Effectiveness of Contact Tracing in Vermont, United States, unpublished) along with instructions provided for replicating this analysis and model use [ 15 , 16 ].

We defined fully vaccinated as either having received two doses of the monovalent mRNA BNT162b2 (Pfizer-BioNTech, Comirnaty) or monovalent mRNA mRNA-1273 (Moderna, Spikevax) COVID-19 vaccine, or one dose of the single-dose adenovirus vector-based Ad26.COV.S (Janssen [Johnson & Johnson]) COVID-19 vaccine [ 48 ].

Recall that the vaccinate-or-test requirement required individuals to get their first dose of a two-dose COVID-19 vaccine series, or undergo, at a minimum, weekly testing for COVID-19 if they remained unvaccinated by September 19, 2021 [ 9 ]. CDC recommended at least three weeks between two doses for Pfizer-BioNTech and 28 days between two doses for Moderna any two vaccine [ 49 ]. To comply with the requirement, individuals had to receive their second dose of a two-dose COVID-19 vaccine series on October 19, 2021, at the latest.

Abaluck J, Kwong LH, Styczynski A, Haque A, Kabir MA, Bates-Jefferys E, Crawford E, Benjamin-Chung J, Raihan S, Rahman S, and others. Impact of community masking on COVID-19: a cluster-randomized trial in Bangladesh,. Science . 2022;375(6577):eabi9069.

Article   CAS   PubMed   PubMed Central   Google Scholar  

Bundgaard H, Bundgaard JS, Raaschou-Pedersen DET, von Buchwald C, Todsen T, Norsk JB, Pries-Heje MM, Vissing CR, Nielsen PB, Winslow UC, et al. Effectiveness of adding a mask recommendation to other public health measures to prevent SARS-CoV-2 infection in Danish mask wearers: a randomized controlled trial. Annals Internal Med. 2021;174(3):335–43.

Article   Google Scholar  

Howard J, Huang A, Li Z, Tufekci Z, Zdimal V, Van Der Westhuizen HM, Von Delft A, Price A, Fridman L, Tang LH, et al. An evidence review of face masks against COVID-19. Proc Nat Acad Sci. 2021;118(4):e2014564118.

Eikenberry SE, Mancuso M, Iboi E, Phan T, Eikenberry K, Kuang Y, Kostelich E, Gumel AB. To mask or not to mask: Modeling the potential for face mask use by the general public to curtail the COVID-19 pandemic. Infectious disease modelling. 2020;5:293–308.

Article   PubMed   PubMed Central   Google Scholar  

Mohammed I, Nauman A, Paul P, Ganesan S, Chen KH, Jalil SMS, Jaouni SH, Kawas H, Khan WA, Vattoth AL, et al. The efficacy and effectiveness of the COVID-19 vaccines in reducing infection, severity, hospitalization, and mortality: A systematic review. Hum Vaccin  immunother. 2022;18(1):2027160.

Howard-Williams M, Soelaeman RH, Fischer LS, McCord R, Davison R, Dunphy C. Association Between State-Issued COVID-19 Vaccine Mandates and Vaccine Administration Rates in 12 US States and the District of Columbia. JAMA Health Forum. 2022;3(10):e223810–e223810.

E. B. Hodcroft, "CoVariants: SARS-CoV-2 Mutations and Variants of Interest,". Available: https://covariants.org/per-country?region=United+States&country=Illinois . Accessed 29 Aug. 2023.

JB Pritzker, Governor, "Executive Order 2021-20 (COVID-19 EXECUTIVE ORDER NO. 87)," Governor JB Pritzker © 2023 State of Illinois, August 26, 2021. [online] https://www.illinois.gov/government/executive-orders/executive-order.executive-order-number-20.2021.html .

JB Pritzker, Governor, "Executive Order 2021-22 (COVID-19 EXECUTIVE ORDER NO. 88)," Governor JB Pritzker © 2023 State of Illinois, September 03, 2021. [online] https://www.illinois.gov/government/executive-orders/executive-order.executive-order-number-22.2021.html .

JB Pritzker, Governor, "Executive Order Number 18 (COVID-19 EXECUTIVE ORDER NO. 85)," Governor JB Pritzker © 2023 State of Illinois, August 04, 2021. [online] https://www.illinois.gov/government/executive-orders/executive-order.executive-order-number-18.2021.html .

Ueki H, Furusawa Y, Iwatsuki-Horimoto K, Imai M, Kabata H, Nishimura H, Kawaoka Y. Effectiveness of face masks in preventing airborne transmission of SARS-CoV-2. MSphere. 2020;5(5):e00637-20.

Cheng Y, Ma N, Witt C, Rapp S, Wild PS, Andreae MO, Poschl U, Su H. Face masks effectively limit the probability of SARS-CoV-2 transmission. Science. 2021;372(6549):1439–43.

Li Y, Liang M, Gao L, Ahmed MA, Uy JP, Cheng C, Zhou Q, Sun C. Face masks to prevent transmission of COVID-19: A systematic review and meta-analysis. American journal of infection control. 2021;49(7):900–6.

Article   CAS   PubMed   Google Scholar  

U.S. Centers for Disease Control and Prevention , "COVIDTracer and COVIDTracer Advanced," 19 January 2021. Available: https://www.cdc.gov/coronavirus/2019-ncov/php/contact-tracing/COVIDTracerTools.html . Accessed 5 Apr. 2023.

Jeon S, Rainisch G, Lash RR, Moonan PK, Oeltmann JE, Greening B, Adhikari BB, Meltzer MI, et al. Estimates of cases and hospitalizations averted by COVID-19 case investigation and contact tracing in 14 health jurisdictions in the United States,. J Public Health Manag Pract. 2022;28(1):16–24.

Article   PubMed   Google Scholar  

Rainisch G, Jeon S, Pappas D, Spencer K, Fischer LS, Adhikari B, Taylor M, Greening B, Moonan P, Oeltmann J, Kahn EB, Washington ML, Meltzer MI. Estimated COVID-19 Cases and Hospitalizations Averted by Case Investigation and Contact Tracing in the US. JAMA Network Open. 2022;5(3):e224042–e224042.

Jeon S, Rainisch G, Harris A-M, Shinabery J, Iqbal M, Pallavaram A, Hilton S, Karki S, Moonan PK, Oeltmann JE, Meltzer MI. Estimated Cases Averted by COVID-19 Digital Exposure Notification, Pennsylvania, USA, November 8, 2020–January 2, 2021. Emerging Infectious Diseases. 2023;29(2):426.

S. Jeon, L. Watson-Lewis, G. Rainisch, C.-C. Chiu, F. M. Castonguay, L. S. Fischer, P. K. Moonan, J. E. Oeltmann, B. B. Adhikari, H. Lawman and M. I. Meltzer, "Adapting COVID-19 Contact Tracing Protocols to Accommodate Resource Constraints, Philadelphia, Pennsylvania, USA, 2021," Emerging Infectious Diseases, vol. 30, no. 2, 2024.

IHME COVID-19 Forecasting Team. Modeling COVID-19 scenarios for the United States,. Nat Med. 2021;27(1):94–105.

Wang Y, Chen R, Hu F, Lan Y, Yang Z, Zhan C, Shi J, Deng XaJM, Zhong S, et al. Transmission, viral kinetics and clinical characteristics of the emergent SARS-CoV-2 Delta VOC in Guangzhou. EClinicalMedicine. 2021;40:101129.

Li, B, Deng A, Li K, Hu Y, Li Z, Shi Y, Xiong Q, Liu Z, Guo Q, Zou L, et al. Viral infection and transmission in a large, well-traced outbreak caused by the SARS-CoV-2 Delta variant,. Nat Commun. 2022;13(1):1–9.

Google Scholar  

Y. Liu and J. Rocklov, "The reproductive number of the Delta variant of SARS-CoV-2 is far higher compared to the ancestral SARS-CoV-2 virus," Journal of travel medicine, vol. 28, no. 7, 2021.

Alimohamadi Y, Sepandi M, Esmaeilzadeh F. Estimate of the Basic Reproduction Number for Delta variant of SARS-CoV-2: A Systematic Review and Meta-analysis. Journal of Biostatistics and Epidemiology. 2022;8(1):1–7.

He X, Lau EH, Wu P, Deng X, Wang J, Hao X, Lau YC, Wong JY, Guan Y, Tan X, et al. Temporal dynamics in viral shedding and transmissibility of COVID-19. Nat Med. 2020;26(5):672–5.

Ferretti L, Wymant C, Kendall M, Zhao L, Nurtay A, Abeler-Dörner L, Parker M, Bonsall D, Fraser C. "Quantifying SARS-CoV-2 transmission suggests epidemic control with digital contact tracing," Sci. 2020;368(6491).

U.S. Centers for Disease Control and Prevention, "COVID-19 Pandemic Planning Scenarios," United States Government, 19 March 2021. Online. Available: https://www.cdc.gov/coronavirus/2019-ncov/hcp/planning-scenarios.html#five-scenarios . Accessed 26 Oct. 2022.

U.S. Centers for Disease Control and Prevention, "COVID-19 Vaccinations in the United States, County,". [online] https://data.cdc.gov/Vaccinations/COVID-19-Vaccinations-in-the-United-States-County/8xkx-amqh . Accessed 8 Mar 2022.

Lopez Bernal J, Andrews N, Gower C, Gallagher E, Simmons R, Thelwall S, Stowe J, Tessier E, Groves N, Dabrera G, et al. "Effectiveness of Covid-19 vaccines against the B. 1.617. 2 (Delta) variant,". N Eng J Med. 2021;385(7):585–94.

Levin EG, Lustig Y, Cohen C, Fluss R, Indenbaum V, Amit S, Doolman R, Asraf K, Mendelson E, Ziv A, et al. "Waning immune humoral response to BNT162b2 Covid-19 vaccine over 6 months,". N Eng J Med. 2021;385(24):e84.

U.S. Bureau of Labor Statistics, "Occupational Employment and Wage Statistics," United States Government, May 2020. Available: https://www.bls.gov/oes/2020/may/oes_il.htm#00-0000 . Accessed 10 Mar. 2022.

U.S. Centers for Disease Control and Prevention, "Weekly HCP & Resident COVID-19 Vaccination," National Healthcare Safety Network (NHSN), 17 April 2023. Available: https://www.cdc.gov/nhsn/ltc/weekly-covid-vac/index.html#anchor_21687 . Accessed 22 Feb. 2022.

Kahane LH. Politicizing the mask: Political, economic and demographic factors affecting mask wearing behavior in the USA. Eastern economic journal. 2021;47(2):163–83.

Illinois, Capitol News, "CAPITOL RECAP: Ruling on legal challenge to school mask mandates could come soon," Herald-Whig, 31 01 2022.

Scheid JL, Lupien SP, Ford GS, West SL. Commentary: physiological and psychological impact of face mask usage during the COVID-19 pandemic. International journal of environmental research and public health. 2020;17(18):6655.

Hansen N-JH, Mano RC. Mask mandates save lives. Journal of Health Economics. 2023;88: 102721.

Albarracin D, Jung H, Song W, Tan A, Fishman J. Rather than inducing psychological reactance, requiring vaccination strengthens intentions to vaccinate in US populations. Scientific Reports. 2021;11(1):20796.

Lewandowsky S, Holford D, Schmid P. "Public policy and conspiracies: The case of mandates," Curr Opin Psychol. 2022;47:101427.

de Figueiredo A, Larson HJ, Reicher SD. The potential impact of vaccine passports on inclination to accept COVID-19 vaccinations in the United Kingdom: Evidence from a large cross-sectional survey and modeling study. EClinicalMedicine. 2021;40: 101109.

Brussow H, Zuber S. Can a combination of vaccination and face mask wearing contain the COVID-19 pandemic? Microbial Biotechnology. 2022;15(3):721–37.

Mazzoni A, Di Lauria N, Maggi L, Salvati L, Vanni A, Capone M, Lamacchia G, Mantengoli E, Spinicci M, Zammarchi L, et al. "First-dose mRNA vaccination is sufficient to reactivate immunological memory to SARS-CoV-2 in subjects who have recovered from COVID-19," J Clin Investig.2021;131(12):e149150.  https://doi.org/10.1172/JCI149150 .

Xiang Y, Jia Y, Chen L, Guo L, Shu B, Long E. COVID-19 epidemic prediction and the impact of public health interventions: A review of COVID-19 epidemic models. Infectious Disease Modelling. 2021;6:324–42.

Chen M, Li M, Hao Y, Liu Z, Hu L, Wang L. The introduction of population migration to SEIAR for COVID-19 epidemic modeling with an efficient intervention strategy. Information Fusion. 2020;64:252–8.

Smith LE, Potts HW, Amlot R, Fear NT, Michie S, Rubin GJ. "Adherence to the test, trace, and isolate system in the UK: results from 37 nationally representative surveys,". BMJ. 2021;372(608).

Park CL, Russell BS, Fendrich M, Finkelstein-Fox L, Hutchison M, Becker J. Americans’ COVID-19 stress, coping, and adherence to CDC guidelines. Journal of general internal medicine. 2020;35(8):2296–303.

Pew Research Center. The Challenges of Contact Tracing as U.S. Battles COVID-19. Washington D.C.; 2020. [online]. https://www.pewresearch.org/internet/2020/10/30/the-challenges-of-contact-tracing-as-u-s-battles-covid-19/ .

Tian L, Li X, Qi F, Tang QY, Tang V, Liu J, Li Z, Cheng X, Li X, Shi Y, et al. Harnessing peak transmission around symptom onset for non-pharmaceutical intervention and containment of the COVID-19 pandemic. Nat Commun. 2021;12(1):1147.

United States Census Bureau, "Quick Facts," United States Government, 1 July 2022. Available: https://www.census.gov/quickfacts/IL . Accessed 4 May 2023.

Rosenblum HG, Wallace M, Godfrey M, et al. Interim recommendations from the Advisory Committee on Immunization Practices for the use of bivalent booster doses of COVID-19 vaccines—United States, October 2022. MMWR Morb Mortal Wkly Rep. 2022;71(45):1436–41.

Kriss JL, Reynolds LE, Wang A, Stokley S, Cole MM, Harris LQ, Shaw LK, Black CL, Singleton JA, Fitter DL. OVID-19 vaccine second-dose completion and interval between first and second doses among vaccinated persons—United States, December 14, 2020- February 14, 2021. Morb Mortal Wkly Rep. 2021;70(11):389.

Article   CAS   Google Scholar  

Download references

Acknowledgements

Nothing to declare.

The findings and conclusions of this report are those of the authors and do not necessarily represent the official position of the Centers for Disease Control and Prevention.

No funding to report.

Author information

Authors and affiliations.

National Center for Emerging and Zoonotic Infectious Diseases, Division of Preparedness and Emerging Infections, Centers for Disease Control and Prevention, Health Economics and Modeling Unit, U.S. Department of Health and Human Services, 7101 Avenue du Parc, Local 3180, QC H3N 1X9, Atlanta, Georgia

François M. Castonguay, Seonghye Jeon, Bishwa B. Adhikari, Leah S. Fischer, Bradford Greening Jr., Emily B. Kahn, Gloria J. Kang & Martin I. Meltzer

Contact Tracing and Innovation Section (CTIS), State Tribal Local and Territorial (STLT) Task Force, CDC COVID-19 Response; Centers for Disease Control and Prevention, Modeling Support Team, U.S. Department of Health and Human Services, Atlanta, Georgia

Illinois Department of Public Health, Springfield, IL, USA

Arti Barnes, Jane Fornoff, Adebola O. Hassan, Judy Kauerauf, Sarah Patrick & Sameer Vohra

Department of Health Management, Evaluation and Policy, University of Montreal School of Public Health, and Centre for Public Health Research – CReSP, 7101 Av du Parc, 3E Étage, Montréal, QC, H3N 1X9, Canada

François M. Castonguay

You can also search for this author in PubMed   Google Scholar

Contributions

Conceptualization: FMC, AB, SJ, JF, BBA, LSF, BGJ, AOH, EBK, GJK, JK, SP, SV, MIM.  Methodology: FMC, SJ, BBA, LSF, BGJ, EBK, GJK, MIM.  Investigation: FMC, SJ.  Visualization: FMC.  Supervision: SV, MIM.  Writing—original draft: FMC, AB, SJ.  Writing—review and editing: FMC, AB, SJ, JF, BBA, LSF, BGJ, AOH, EBK, GJK, JK, SP, SV, MIM.

Corresponding author

Correspondence to François M. Castonguay .

Ethics declarations

Ethics approval and consent to participate.

This research activity was reviewed by the Centers for Disease Control and Prevention (CDC) and was conducted consistent with applicable federal law and CDC policy. See e.g., 45 C.F.R. part 46, 21 C.F.R. part 56; 42 U.S.C. §241(d); 5 U.S.C. §552a; 44 U.S.C. §3501 et seq. The COVID-19 confirmed case and vaccination data used in this study are aggregated at the state level, meaning no personally identifiable information or data were used.

Consent for publication

Not applicable.

Competing interests

The authors declare no competing interests.

Additional information

Publisher’s note.

Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.

Supplementary Information

Supplementary material 1., rights and permissions.

Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if changes were made. The images or other third party material in this article are included in the article's Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article's Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http://creativecommons.org/licenses/by/4.0/ . The Creative Commons Public Domain Dedication waiver ( http://creativecommons.org/publicdomain/zero/1.0/ ) applies to the data made available in this article, unless otherwise stated in a credit line to the data.

Reprints and permissions

About this article

Cite this article.

Castonguay, F.M., Barnes, A., Jeon, S. et al. Estimated public health impact of concurrent mask mandate and vaccinate-or-test requirement in Illinois, October to December 2021. BMC Public Health 24 , 1013 (2024). https://doi.org/10.1186/s12889-024-18203-8

Download citation

Received : 14 November 2023

Accepted : 24 February 2024

Published : 12 April 2024

DOI : https://doi.org/10.1186/s12889-024-18203-8

Share this article

Anyone you share the following link with will be able to read this content:

Sorry, a shareable link is not currently available for this article.

Provided by the Springer Nature SharedIt content-sharing initiative

• Cases and hospitalizations averted

BMC Public Health

ISSN: 1471-2458

• Submission enquiries: [email protected]
• General enquiries: [email protected]

An official website of the United States government

The .gov means it’s official. Federal government websites often end in .gov or .mil. Before sharing sensitive information, make sure you’re on a federal government site.

The site is secure. The https:// ensures that you are connecting to the official website and that any information you provide is encrypted and transmitted securely.

• Publications
• Account settings

Preview improvements coming to the PMC website in October 2024. Learn More or Try it out now .

• Advanced Search
• Journal List
• Elsevier - PMC COVID-19 Collection

The resilience of the lodging industry during the pandemic: Hotels vs. Airbnb

Tarik dogru.

a Florida State University, Dedman College of Hospitality, Tallahassee, FL, USA

Lydia Hanks

Courtney suess.

b Texas A&M University, Department of Recreation, Park and Tourism Sciences, College Station, TX, USA

Nathan Line

Makarand mody.

c Boston University, School of Hospitality Administration, Boston, MA, USA

The adverse impact of the recent pandemic on the lodging industry has largely been based on anecdotal evidence. The extent to which different parts of this broad industry were individually affected by the COVID-19 pandemic also remains unclear. The purpose of this study is to investigate the effects of the COVID-19 pandemic in the various sectors of the lodging industry to identify patterns that may not be consistent with the idea that the entire hospitality industry was negatively affected by the pandemic. The results show that while the COVID-19 pandemic did have a generally negative effect on lodging demand, hotel room and peer-to-peer accommodation property bookings were not affected equally. Importantly, it appears that these variations were attributable, at least in part, to state-level variations in policy that made travel and hospitality services relatively more (or less) difficult for consumers to obtain. Theoretical and managerial implications are extensively discussed.

1. Introduction

Anecdotal evidence abounds regarding the negative effect of the COVID-19 pandemic on the lodging industry. Headlines such as “North American inbound tourism spend declined by 74.1% in 2020″ ( Johnson, 2021 ) and “Coronavirus Pandemic Sets Hotel Industry Back 10 Years, Report Finds” ( Lardieri, 2021 ) have led to both an academic and a practical understanding that, in a general sense, the COVID-19 pandemic negatively affected consumers’ booking overnight accommodations. However, the accommodations industry is quite complex. For example, the lodging industry comprises both traditional hotels and resorts as well as individuals’ home properties listed through companies such as Airbnb, Vrbo, HomeAway, etc. (peer-to-peer accommodations). Additionally, there are lodging sub-types within this stratification. Within the traditional hotel sector, there are six distinct lodging classes ranging from economy to luxury (STR, 2021). Likewise, amongst the peer-to-peer accommodation listings by companies such as Airbnb, available property types include entire homes, private rooms within an owner-occupied home, and shared-rooms with other guests. Thus, while the accommodations industry is often considered in aggregate when it comes to reporting purchasing decline from COVID-19, it is possible that not all of the various operating sectors were affected equally.

Many different factors exist that could have caused variation between (and within) bookings in the hotel and the peer-to-peer accommodation sectors. On one hand, consumer sentiment about safety could have potentially shifted demand away from hotels and toward the peer-to-peer accommodations due to the perception that private accommodations are more conducive to social distancing than hotels. On the other hand, consumers may have been willing to forego the relative isolation inherent to private or even shared homes in favor of hotels that typically provide a standardized and more professional approach to health and sanitation ( Verleye et al., 2021 ). In addition, it is possible that factors outside of consumer sentiment, such as state-level policy, travel restrictions, and limitations on hospitality operations, could have caused differential effects on various booking patterns throughout the lodging industry. While many states implemented relatively strict mandates in an effort to mitigate spread of COVID-19 (e.g., stay at home orders, curfews, limited business operating hours, business closures, visitor restrictions, etc.) other states imposed relatively few restrictions ( McCann, 2021 ). As such, it is likely that inter-state disparities in COVID-19-based restrictions contributed, at least in part, to variations in demand effects between and within the various operating sectors of the lodging industry.

Overall, it is evident that occupancy rates were disrupted at the industry level during the COVID-19 pandemic. However, it remains unclear how various sectors of the broad industry were affected. Accordingly, the purpose of this study is to investigate the effects of the COVID-19 pandemic in the various sectors of the lodging industry to identify patterns that may not be consistent with the idea that the entire hospitality industry was negatively affected by the pandemic. Additionally, this research seeks to identify the effects of state-level policy on the demand across various operating sectors of the lodging industry. The results from this study are expected to shed light on several issues ranging from the resilience from pandemic-related demand shocks to the effect of public-health regulations on guest booking hotel rooms and peer-to-peer accommodations.

2. Literature review

2.1. the impact of covid-19.

With the onset of the COVID-19 pandemic, travel restrictions and consumer concerns have had a significant impact on all facets of the travel and tourism industry. In an effort to quantify these effects, understand their precursors, and develop mitigating strategies, researchers have conducted a number of studies in various contexts, beginning very shortly after the onset of the pandemic. For example, Ugir & Akbiyik (2020) analyzed data from a large travel forum and found that prior to the worldwide proliferation of the virus, even while the pandemic was still localized, travelers were changing, adjusting, and canceling travel plans. Likewise, Li et al., (2020) found that as the pandemic began to spread globally, travelers were delaying travel plans and booking shorter stays than originally planned. Indeed, in 2020, a flurry of papers were published, analyzing the impact of the pandemic on various aspects of the tourism industry, such as trip type (Wen et al., 2020), travel likelihood (Yang et al., 2020), tourism consumer behavior (Zenker & Kock, 2020; Suess et al., 2022 ), and homesharing ( Dolnicar and Zare, 2020 ). Research into the impact of the pandemic on the generalized hospitality and tourism industry continued during the following two years, demonstrating the adverse effects on tourism asset management (Crespi-Cladera et al., 2021), tourism company valuations (Kocak et al., 2022), and various subsectors of the tourism and hospitality industry. Cepni et al. (2022) found that the hotel subsector of the tourism industry was hit particularly hard by the pandemic.

Several studies have been conducted over the past two years investigating the impact of COVID-19 on the hotel industry. For example, Guillet & Chu (2021) conducted a qualitative study, speaking with revenue managers at hotels in Asia, and found that hotel demand was down, regardless of revenue management strategies. Likewise, Ozdemir et al. (2021) found that demand was down across the entire hotel industry, although economy-scale hotels were more resilient than other segments. Ozdemir et al. (2022) found that demand was weaker in states where travel and gathering restrictions were the most stringent, and that this was not uniform across all hotel segments, with economy hotels being more resistant than luxury segment hotels. These studies would suggest that there is much to be learned about the ways in which the COVID-19 pandemic impacts various types of hotels differentially. To fill this gap, the present study aims to investigate how the pandemic impacted the various hotel classes, as well as comparing these results to homesharing platforms.

2.2. The effect of COVID-19 on the hotel industry

There is no doubt that COVID-19 has significantly impacted the lodging industry. In the face of the pandemic, visiting friends and family has been canceled, business trips have shifted into online videoconferencing, and leisure and vacation travel have been severely restricted. These changes have been a major disruptor, seriously affecting the traditional hotel business models ( Schulz, 2021 ). However, as the course of the pandemic has ebbed and flowed, it is difficult to discern exactly how the hotel industry has been affected. Impacts to hotel occupancy have been relatively fluid, with some phases of the pandemic producing drastic drops in travel and associated room bookings during times of high-risk and peak infection rates, followed by months of recovery fueled by demand from those pent-up in their residences, seeking to travel ( Eggleston and Lee, 2021 ). Consumer perceptions of safety, the ability and willingness to travel, and anxiety about the pandemic have all resulted in demand-side fluctuations in hotel reservations. In addition, government mandated closures and stay-at-home orders have had a profound effect. Given the number and the complexity of factors affecting hotel reservations, it is difficult to discern how COVID-19 has affected overall hotel occupancy. Thus, the following is proposed:

Research Question 1:

To what extent does the COVID-19 pandemic affect hotel occupancy?

To further illuminate the ways in which COVID-19 has impacted the hotel industry, it is important to examine whether the effects of the pandemic on hotel occupancy vary across hotel classes. For classification and comparison purposes, hotels are often divided into categories (i.e., economy, midscale, upper midscale, upscale, upper upscale, and luxury) ( STR Chain Scales, 2019 ). While there are many ways to classify hotels, these categories are widely accepted due to their use by Smith Travel Research and the STAR report, thus creating a taxonomy that is commonly understood by hoteliers and researchers alike (Miller, 2016). Ozdemir et al. (2021) have suggested that the pandemic has affected various hotel classes differently, but aside from this study, there is little understanding of how each class of hotel has been specifically impacted. Thus, research question 2 is proposed as follows:

To what extent does the COVID-19 pandemic affect the occupancy levels of various hotel classes ( i.e ., economy, midscale, upper midscale, upscale, upper upscale, and luxury)?

2.3. The effect of COVID-19 on Airbnb occupancy

In addition to affecting hotel occupancy, it is possible that the restrictions on travel have affected the peer-to-peer accommodations sectors as well. While there are many home-sharing platforms, for the purposes if this study, the effects of the pandemic on Airbnb property occupancy levels is most germane. Airbnb is the largest among the peer-to-peer accommodation platforms, and a primary competitor to the hotel industry, providing travelers with the opportunity to book a stay in an individual’s private home, either with or without them occupying it as a host. Moreover, data on the properties listed on Airbnbs’ occupancy levels is readily available. While COVID-19 restrictions, lockdowns, and canceled travel plans have certainly affected occupancy among Airbnb properties, research has suggested that the impact has not been entirely negative ( Dolnicar and Zare, 2020 ). As safety-conscious travelers opt out of staying in hotels where they must share space such as elevators, lobbies, and restaurants with other patrons, they have looked toward renting the private properties listed on Airbnb as an alternative, where they will not encounter other guests and extensive staff, thus decreasing the risk of catching the virus ( Carville and Bloomberg, 2021 ). Additionally, the shift at many companies to working at home has given rise to a new group of customers who choose to rent a property through Airbnb on a long-term basis so that they can live in and explore new and different cities and countries while working remotely ( Hines, 2021 ). Such digital nomads have generated demand for peer-to-peer accommodations in a way that could not be foreseen prior to the pandemic ( Rodriguez, 2020 ). Given that travel restrictions and health concerns significantly decreased travel overall during the pandemic, yet health-conscious guests and digital nomads generated new demand, it remains unclear how COVID-19 has impacted occupancy of Airbnb listings overall. Thus, research question 3 is proposed:

To what extent does the COVID-19 pandemic affect Airbnb occupancy?

One unique characteristic of Airbnb listings is that guests have a number of property options. First, the guest can choose to book an entire home with no host residing at the premises during their stay. Second, guests can book a private bedroom or suite within a home in which the host is living during their stay. Third, guests can also book a shared room (i.e., the room is shared with another guest) within a home ( Airbnb, 2021 ). Because each of these property types carries different implications in terms of contact with people during COVID-19, it is important to examine each one, separately. Prior research has demonstrated that these three peer-to-peer accommodation property types vary in such areas as pricing ( Voltes-Dorta and Sánchez-Medina, 2020 ) and occupancy ( Bresciani et al., 2021 ).

It is important to note that shared rooms and shared homes involve direct contact and shared common areas with others, a situation that is contrary to social distancing practices. Alternatively, entire-home rentals, with no host occupying the residence, provide the opportunity to isolate and social distance. Indeed, Bresciani et al. (2021) found that during the pandemic, travelers were reluctant to book homes shared with other guests or occupied by hosts during the pandemic, and the preference for whole home rentals was predicated on the desire for social distancing. Based on this idea, it seems logical to investigate whether the effect of COVID-19 varies across property types in Airbnb rentals. Thus, research question 4 is proposed:

Is the effect of COVID-19 on occupancy levels homogenous across Airbnb property types?

2.4. The effect of COVID-19 state restrictions

While consumer concerns about cleanliness, social distancing, and isolation while traveling certainly had an impact on the lodging industry’s occupancy levels, these are not the only factors to consider. State-level restrictions on hospitality businesses also had a significant impact on lodging demand, throughout the pandemic. Many states enforced social-distancing and ordered restrictions on business openings, hours of operation, number of patrons allowed in any given business, and private gatherings -in some cases any gathering over six people (Multistate, 2021). However, there has been vast deviation with regards to the number and type of restrictions each state has imposed during the pandemic. For example, California, New York, and Massachusetts enforced the strictest measures, maintaining them throughout periods when infection rates were very low, and vaccination was prevalent ( McCann, 2021 ). Meanwhile, states such as Florida and Texas had minimal restrictions, and only during brief periods when the infection cases were reported to be at their highest and when all other states were enforcing the public safety measures under federal pressure (Hunnicutt & Mason, 2021).

Such restrictions obviously had a significant impact on hotel occupancy rates, as they not only set a precedent of discouraging people from going out in public, but also limited the ability of those who chose to stay mobile to travel and book a hotel ( Gursoy and Chi, 2020 , Lai and Wong, 2020 ). Hotels were further affected as they became vulnerable to government restrictions on gathering size, impacting conferences, events, and group business significantly. Thus, it is important to understand how state-level restrictions impacted hotel occupancy, and Research Question 5 is thus proposed:

How has the level of state restrictions regarding COVID-19 impacted hotel occupancy levels?

As mentioned above, peer-to-peer accommodations listed on Airbnb differ from traditional hotels in terms of their guest capacity and host occupancy. Both of these factors are important in determining the impact of state COVID-19 restrictions on demand for Airbnb properties. First, because listings on Airbnb do not typically include property types that are dependent on large group business (as opposed to hotels), state restrictions regarding group size, for the most part, would not impact bookings on Airbnb. Second, because properties listed are mostly individual-owned private homes, as opposed to businesses, state mandates to close did not apply to the properties. Third, while state restrictions and social distancing reduced the number of guests permitted on hotel premises and amenity operations, these mandates, in general, did not apply to properties listed on Airbnb or significantly impact host operations for guests in their homes. Namely, almost none of the state restrictions on gatherings applied to entire-home listings, often booked by individuals or family and small groups. In light of the differential social distancing standards among hotels and peer-to-peer accommodations, it is important to investigate Research Question 6:

How has the level of state restrictions regarding COVID-19 impacted Airbnb property occupancy levels?

3. Methodology

3.1. sample and data.

The sample of this study comprises 49 out of the 50 U.S. states and the District of Columbia for the period between October 2014 and December 2020. The study covers the time period from the first availability of Airbnb property data (e.g., occupancy, supply, demand, etc.) to its latest availability (December 2020). Delaware was omitted from the study sample because there was no data available on this state. Accordingly, the study’s sample contains 3850 state-month observations.

The study’s dependent variable is the monthly occupancy rate, which is the most commonly accepted industry-metric reported on a monthly basis for both hotels and Airbnb peer-to-peer accommodation listings. Specifically, hotel room occupancy rates and Airbnb properties occupancy rates were specified as the main dependent variables of the study to examine the effects of COVID-19 pandemic on hotels and Airbnb property demand. The hotel occupancy data were provided by Smith Travel Research (STR), while data on Airbnb property occupancy were obtained from AirDNA, a well-recognized research company that collects detailed data on Airbnb listings through web crawling. Furthermore, monthly hotel occupancy rate for hotel categories, including economy class, midscale class, upper midscale class, upscale class, upper upscale class, and luxury class hotels were used to analyze the extent to which the effect of COVID-19 pandemic on hotels varies based on diverse hotel categories. Likewise, listings on Airbnb include entire-home, private-room, and shared-room property types. Therefore, monthly Airbnb occupancy rates for entire homes, private rooms, and shared room Airbnb properties were specified as additional dependent variables to investigate the effect of COVID-19 pandemic across alternative Airbnb property types. Fig. 1 presents the occupancy levels of the accommodations during the study period. The dramatic fall in occupancy levels across different accommodation properties is evident during the pandemic period.

Occupancy Levels of Hotels and Airbnb Properties.

The total monthly number of confirmed COVID-19 cases was specified as the main independent variable of the study, to investigate the effect of COVID-19 pandemic on the US accommodation sector. In addition, the total monthly number of confirmed COVID-19 deaths, monthly new-confirmed COVID-19 cases, and monthly new-confirmed COVID-19 deaths as alternative measures. The COVID-19 new-confirmed cases and death data were obtained from the Centers for Disease Control and Prevention (CDC). Fig. 2 presents the severity of the total COVID-19 cases across the US states during the study period in the form of heat maps.

Total Number of COVID-19 Cases in the US.

A number of control variables were included to account for macroeconomic factors and travel demand that might affect hotel and Airbnb property occupancy rates regardless of the COVID-19 pandemic. Based on extant studies analyzing hotel supply and demand dynamics (e.g., Canina & Carvell, 2005; Dogru et al., 2020 ; Lee & Jang; 2012; Lei & Lam, 2015; Tsai, Kang, Yeh, & Suh, 2006), five control variables were included in the models to account for the effects of macroeconomic factors and travel demand on the accommodation sector. Specifically, airport passenger arrivals, hotel room supply, Airbnb property supply, income, and unemployment rate were controlled for. The first control variable in the empirical models, which accounts for general tourism and travel demand dynamics, was the number of airport passenger arrivals to all state airports. Airport passenger arrivals data were obtained from the Bureau of Transportation Statistics. Because a change in hotel room supply and Airbnb properties supply is likely to have an effect on the occupancy rates ( Dogru et al., 2020 ), the effect of hotel room supply and Airbnb properties supply was also controlled for. Furthermore, the coincident index, a variable developed by the Federal Reserve Bank of Philadelphia to measure the monthly gross domestic product (GDP) for each state, was included in the empirical models to account for the effects of general macroeconomic dynamics on hotel and Airbnb property occupancy rates (Crone & Clayton-Matthews, 2005; Lee & Jang, 2012). The coincident index is a commonly used proxy for income when analysis consists of monthly level data frequency (see e.g., Dogru et al., 2019 ; Lee & Jang, 2012). To further account for overall macroeconomic conditions, unemployment rate was included in the empirical models.

Unemployment rate can help capture overall macroeconomic conditions, which are likely to affect the accommodation sector (Chen et al., 2011). Table 1 presents the summary statistics of the study’s dependent, independent and control variables.

Summary Statistics.

Although the overall examination of the effects of COVID-19 pandemic on the supply of hotel rooms and Airbnb properties is useful, individual states in the U.S. have had varying regulations in place regarding the COVID-19 pandemic. These regulations, in turn, have resulted in differential effects on the supplies across different states. To account for these potential differences, the data were sorted into two groups based on the state and the extent of the regulations, respectively. Specifically, each state was categorized into either a more restrictive or less restrictive group, depending on the degree of regulations as measured by the WalletHub index of “States with the Fewest COVID-19 Restrictions”. For a detailed description of the index and its methodology, readers are directed to the following reference: WalletHub’s States with the Fewest Coronavirus Restrictions at https://wallethub.com/edu/states-coronavirus-restrictions/73818/. The resulting groups were fairly even, consisting of 25 states in each.

3.2. Empirical methodology

The COVID-19 pandemic is treated as a variable intervention in time against hotel room occupancy and Airbnb properties occupancy rates in 50 states, which allows for the investigation of both before and after effects of the COVID-19 pandemic on hotel room occupancy and Airbnb properties occupancy rates. The dataset covered the pre-pandemic period between October 2014 and December 2019, where total number of COVID-19 cases take the value of zero based on the Centers for Disease Control and Prevention agency, and the post-COVID period between January 2020 and December 2020, where total number of COVID-19 cases take the value of the number of cumulative COVID-19 cases recorded every month in each state. This empirical approach better captures the effect of COVID-19 pandemic on the US accommodation sectors because the COVID-19 pandemic is treated as a variable intervention in time against the hotel room occupancy and Airbnb properties occupancy rates in 50 states. The empirical models are specified as follows:

The dependent variables are hotel room occupancy (OCC), which consists of all hotels, and economy, midscale, upper midscale, upscale, upper upscale and luxury class occupancy of hotel rooms, and Airbnb properties occupancy, entire home Airbnb occupancy, private room Airbnb occupancy and shared room Airbnb occupancy in state i at time t . The independent variable is the total number of COVID-19 cases (i.e., logCOVID) in state i at time t . X represents a set of control variables of the state i at time t that includes log airport arrivals, log hotel room supply, log Airbnb properties supply, log coincident index (income) and unemployment rate. The variable e is the error term, and ß 1 − 11 , ß k are the model parameters. All models include year and month effects to control for seasonality and time-specific economic and other conditions over time, and state-fixed effects to control for state-specific characteristics and dynamics. A logarithmic transformation was utilized to account for data-skewness in the study variables, with the exception of occupancy rates and unemployment rates. The central focus of the empirical specifications detailed above is the coefficient of logCOVID ( ß 1 − 11 ).

The panel data fixed effect regression model was employed to estimate the effect of COVID-19 on hotel and Airbnb. While the fixed effect approach would be recommended from a theoretical perspective due to the nature of data and analysis, a Hausman test to examine whether fixed effect or random effect models was conducted, and more appropriate from an empirical methodology perspective. The results from the Hausman test suggests that the panel data fixed effect regression model is a better fit for the analysis (Chi2: 223.88 Probability> Chi2 =0.00). Therefore, the panel data fixed effect regression model was utilized throughout the analyses. The errors were further clustered based on the states to account for potential autocorrelation and heteroskedasticity problems that may exist in the model estimates. Thus, empirical estimations yielded robust estimates based on robust standard errors ( Hoechle, 2007 ).

4. Empirical results

4.1. the effect of covid-19 pandemic on hotels.

This section presents the results from the investigation of the effect of COVID-19 pandemic on the US hotel industry. The analyses are carried out utilizing the panel data fixed effect regression analysis. Table 2 presents the results from the analysis of the effects of COVID-19 pandemic on all hotels and for each hotel class separately (i.e., economy, midscale, upper midscale, upscale, upper upscale, and luxury) controlling for macroeconomic factors, such as income and hotel industry specific factors, such as hotel room supply.

The Effect of COVID-19 on Hotel Occupancy.

Notes: Occupancy rate is the dependent variable. Robust t- statistics are in parentheses. a, b, and c denote 1%, 5% and 10% statistical significance levels, respectively.

The results show that hotel room occupancy is adversely affected by the COVID-19 pandemic. Specifically, a 1% increase in total number of COVID-19 cases decreases hotel occupancy rate by almost 0.01%. This result is both economically and statistically significant ( β = − 0.007 , p < 0.01 ). While the COVID-19 pandemic is expected to have an adverse impact on the hotel industry, the magnitude of this impact was not clear. Furthermore, the effect of the COVID-19 pandemic may not be uniform across demand associated with different hotel categories. Therefore, a further analysis was conducted on the effect of the COVID-19 pandemic including all hotel class categories.

The results showed that demand for all hotel classes was adversely affected by the COVID-19 pandemic at varying degrees with the exception of the midscale class hotels. The effect of the COVID-19 pandemic on midscale class hotels was not statistically significant at conventional statistical significance levels. The magnitude of the effect of the COVID-19 pandemic was larger in higher class hotels. Specifically, a 1% increase in total number of COVID-19 cases decreases hotel occupancy rate by between 0.01% and 0.02% in upscale class hotels ( β = − 0.01 , p < 0.01 ), upper upscale class hotels ( β = − 0.018 , p < 0.01 ) and luxury class hotels ( β = − 0.017 , p < 0.01 ). The results further show that economy class hotels and upper midscale class hotels showed higher resilience against the COVID-19 pandemic. That is, a 1% increase in total number of COVID-19 cases decreases hotel occupancy rate by about 0.005% in economy class hotels ( β = − 0.004 , p < 0.10 ) and upper midscale class hotels ( β = − 0.005 , p < 0.01 ). These outcomes collectively suggest that the COVID-19 pandemic has had a significantly negative effect on demand in the hotel sector, in general. However, the magnitude of the effect is not uniform across hotel class categories and that the midscale class hotels are the most resilient against the COVID-19 pandemic, followed by economy scale and upper midscale hotels, while the highest impact was observed in the upper upscale and luxury class hotels. The results further showed that the differences in magnitude were also statistically significant for all hotel classes except for upscale hotels compared to all hotels, as the COVID-19 pandemic has had a similar effect on upscale hotels compared to all hotels. The significance difference tests of the coefficients are presented as follows: all hotels vs. economy hotels ( t  = 2.54, p < 0.05); all hotels vs. midscale hotels ( t  = 2.75, p < 0.01); all hotels vs. upper midscale hotels ( t  = 1.90, p < 0.10); all hotels vs. upscale hotels ( t  = 1.62, n.s.); all hotels vs. upper upscale hotels ( t  = 7.28, p < 0.01); and all hotels vs. luxury hotels ( t  = 4.15, p < 0.01).

4.2. The effect of the COVID-19 pandemic on Airbnb

Although results of the analyses showed that the COVID-19 pandemic had an adverse impact on the hotel industry’s demand, and that the magnitude of this impact varies across hotel class categories, the accommodation sector is not limited to hotels only. Peer-to-peer accommodations listed through the company Airbnb have become a major player in the lodging industry, and the effect of the COVID-19 pandemic on booking demand for properties listed by Airbnb might not be similar to the that of the hotel industry. Therefore, we further analyzed the effects of COVID-19 pandemic on Airbnb property demand utilizing a panel data fixed-effect regression analysis.

Table 3 presents the results from the analysis of the effects of COVID-19 pandemic on both the demand for Airbnb properties, overall, and for each Airbnb property type separately (i.e., entire home, private room, and shared room) controlling for macroeconomic factors, such as income and hotel industry specific factors, such as Airbnb properties supply. Results from the analyses show that when the effect of the COVID-19 pandemic on demand for all Airbnb properties was examined, the COVID-19 pandemic did not have a statistically significant impact. Results are presented in Column 1 of Table 3 . This outcome is surprising when it is compared with that of hotel occupancy. However, the effect of the COVID-19 pandemic does not seem to be uniform across Airbnb property types. While the demand for entire home Airbnb properties does not seem to be adversely affected by the COVID-19 pandemic, the demand for private room and shared room Airbnb properties is negative. Specifically, a 1% increase in total number of COVID-19 cases decreases private room and shared room Airbnb occupancy rates by 0.011% (p < 0.01) and 0.012% (p < 0.01), respectively. The results further showed that the differences in magnitude were also statistically significant for all Airbnb properties except for entire room Airbnb properties compared to Airbnb properties, as the COVID-19 pandemic did not have any adverse impact on all Airbnb properties in general and entire room Airbnb properties in particular. The significance difference tests of the coefficients are presented as follows: all Airbnb vs. entire room Airbnb ( t  = 1.34, n.s.); all Airbnb vs. private room Airbnb ( t  = 7.07, p < 0.01); and all Airbnb vs. shared room Airbnb ( t  = 7.77, p < 0.01).

The Effect of COVID-19 on Airbnb Occupancy.

These results indicate that guests continued to book stays in entire home properties listed on Airbnb during the pandemic, while private rooms and shared rooms experienced significant decline in bookings from guests. It stands to reason that guests might feel less comfortable staying in a private room or shared room during the pandemic due to an increased likelihood of infection during their stay in properties with other people, whether it be hosts or other travelers. Therefore, guests appear to have preferred entire home Airbnb properties during pandemic to avoid social contact and reduce the risk of infection.

The analyses of the effect of COVID-19 pandemic on hotels and properties listed on Airbnb showed that the COVID-19 pandemic does not have equally adverse effects on the accommodation sector. Certainly, midscale class hotels are more resilient than the other classes of hotels, operating throughout the pandemic with little or no change to their demand. Likewise, entire home Airbnb property bookings were not affected by the pandemic. Moreover, comparisons of the effects the COVID-19 pandemic had on demand between hotels and properties listed on Airbnb revealed that peer-to-peer accommodations are more resilient to the pandemic, in general. In sum, while the effect of the COVID-19 pandemic is not uniform across the demand for different accommodation sectors, there was little or no impact overall to Airbnb properties. When the effect of the COVID-19 pandemic was examined across different property types, it appears that entire home Airbnb properties and midscale class hotels are the most resilient against the COVID-19 pandemic. Luxury class and upper upscale class hotels, followed by the private room and shared room Airbnb properties, have experienced the largest impact of COVID-19 pandemic in terms of magnitude.

4.3. The effect of COVID-19 pandemic on Hotels vs. Airbnb: the role of Restrictions

Main analyses showed that the hotel room occupancy is adversely affected by the COVID-19 pandemic, while the effect of COVID-19 pandemic on Airbnb properties occupancy was not statistically significant. However, these analyses do not take into consideration the policy responses of the government authorities across the states regarding the COVID-19 pandemic. That is, during the pandemic, state authorities have reacted differently to COVID-19 pandemic. Some states have taken strict measures, such as stay-at-home orders, curfews, business closures, and travel restrictions to cope with spread of the COVID-19 infections, while some states did not impose any restrictions in this context. Accordingly, the effect of the COVID-19 pandemic is likely to be more prevalent in states with higher restrictions than in states with little or no restrictions in place. Therefore, we further examined the effect of COVID-19 pandemic on hotel and Airbnb occupancies based on the degree of restrictions in 50 US states. Specifically, we categorized each state into more restrictive and less restrictive state groups based on their degree of regulations as measured by the WalletHub index of “States with the Fewest COVID-19 Restrictions”. Table 4 presents these results.

The Effect of COVID-19 on Hotels and Airbnb Occupancy: The Role of State Restrictions.

The results from Columns 1 and 2 of Table 4 show that the COVID-19 pandemic has had a significantly adverse impact on hotel occupancies in states with both more restrictive and less restrictive regulations. Further, the magnitude of the results demonstrate that the COVID-19 pandemic has much higher impact on hotel occupancies in more restrictive states ( β = − 0.011 , p < 0.01 ) than less restrictive states ( β = − 0.005 , p < 0.01 ). The difference in the magnitude of the impact is statistically significant ( t = 4.24 , p < 0.01 ). These findings suggest that pandemic related restrictions have had a significantly higher adverse impact on the hotel industry.

The results from the examination of the effect of COVID-19 pandemic on Airbnb occupancy yields results similar to that of the main findings. That is, the results from Columns 3 and 4 of Table 4 show that the COVID-19 pandemic does not have any adverse effects on Airbnb property occupancies in more restrictive or less restrictive states.

These results suggest that Airbnb properties continued to operate regularly regardless of the restrictions imposed to contain the COVID-19 pandemic by more restrictive states. These findings may indicate that guests preferred to stay in Airbnb properties while traveling during COVID-19 pandemic. Alternatively, privately owned Airbnb properties may not have been held to the same restrictions imposed to commercially operated hotel business. Regardless, peer-to-peer accommodations, at least those listed on Airbnb, appear to be more resilient to the COVID-19 pandemic in both more restrictive and less restrictive states.

We further investigated the effect of the COVID-19 pandemic on hotels and Airbnb properties using alternative measures to test the robustness of our initial findings. More specifically, we analyzed one-month, two-month, and three-month lagged effects of the COVID-19 pandemic on the accommodation sector, because the effect of the COVID-19 pandemic might last for few months. These results collectively confirmed the initial findings that Airbnb properties are more resilient against the COVID-19 pandemic than hotels. Due space limitation, we present these findings in detail in the Appendix.

5. Discussion and conclusion

The COVID-19 pandemic has had major adverse impacts on the overall economy, and many, if not all, industries and businesses have experienced significant challenges resulting from social distancing and public health regulations. While the COVID-19 pandemic continues to be a challenge around the world, the adverse effect of the COVID-19 pandemic on the lodging industry has largely been based on anecdotal evidence. That is, the effect of the COVID-19 pandemic on the lodging industry has not yet been quantified. Furthermore, although the COVID-19 pandemic has had an impact on the overall economy and the lodging industry, its effect might not be uniform across various segments (e.g., economy-scale, midscale, luxury-scale). That is, some hotel segments could be more resilient than others to the external shock.

In this study, the effects of the COVID-19 pandemic on all hotels and for each hotel class separately (i.e., economy, midscale, upper midscale, upscale, upper upscale, and luxury) were analyzed. Specifically, the COVID-19 pandemic was specified as a variable intervention in time against hotel room occupancy, which allows for investigation of the before and after effect of the COVID-19 pandemic on hotel room occupancy in 50 states. The results show that the COVID-19 pandemic has had a significantly negative effect on the hotel sector in general. However, the magnitude of the effect is not uniform across hotel class categories, and the midscale class hotels are the most resilient against the COVID-19 pandemic, followed by economy scale and upper midscale hotels, while the highest impact was observed in the upper upscale and luxury class hotels. The resilience of the midscale hotel can be attributed to the pricing points of these properties compared to upper scale or luxury hotel properties ( Ozdemir et al., 2021 ). During the pandemic, some people continued to travel; however, these travels were mostly business related and hence midscale hotels were preferred. Also, some midscale hotels were utilized by first responders and that provided a resilience in occupancy in this category of hotels. Furthermore, the quality of amenities compared to that of economy scale hotels may offer a better value for travelers and hence may have been preferred during the pandemic. Maintaining the cost of operation with low occupancy level can still be manageable in midscale and economy scale hotels because of the low level of employment needed compared to that of upper scale and luxury scale hotel properties ( Dogru et al., 2020 ). Therefore, midscale hotels showed higher resilience followed by economy scale hotels, while upscale and luxury scale hotels are the most affected by the pandemic.

In addition to alternative hotel segments, the lodging sector consists of peer-to-peer accommodations property listing platforms, such as Airbnb. Therefore, we further examined the effects of COVID-19 pandemic on all Airbnb properties and for each Airbnb property type separately (i.e., entire home, private room, and shared room). The results showed that the COVID-19 pandemic did not have a statistically significant impact on Airbnb properties, in general. However, the effect of the COVID-19 pandemic is not uniform across Airbnb property types. While entire home Airbnb property bookings were not adversely affected by the COVID-19 pandemic, there were significant declines of private room and shared room Airbnb property bookings. These outcomes suggest that travelers do not feel comfortable staying in a private room or shared room Airbnb properties due to likely risk of infection and interaction with other guests, however Airbnb guests continued to book in entire home Airbnb properties during the pandemic which facilitated social distancing.

Moreover, while some states pursued a relatively strict approach to public health regulations in an effort to combat the spread of the COVID-19 virus, including stay-at-home orders, curfews, business closures, and travel restrictions, other states imposed little or no restrictions ( McCann, 2021 ). Therefore, we investigated the effect of the COVID-19 pandemic on hotel and Airbnb property occupancies based on the degree of restrictions by state.

The results showed that the COVID-19 pandemic had a significantly adverse impact on hotel occupancies in both more restrictive and less restrictive states. However, the magnitude of these effects was not uniform. Specifically, the COVID-19 pandemic had a much higher impact on hotel occupancies in more restrictive states than less restrictive states, which suggests that pandemic related restrictions further intensified the adverse impacts of the COVID-19 pandemic on the hotel industry’s demand. However, the COVID-19 pandemic did not have any adverse effects on Airbnb property occupancies, regardless of the degree of restrictions across states. These results suggest that peer-to-peer accommodation properties listed on Airbnb seem immune to the state-level COVID-19 policy and restrictions either because the pandemic-related restrictions did not apply to privately-owned Airbnb properties, or because travelers preferred staying at Airbnb properties for social-distancing reasons.

5.1. Research implications

This study makes several key contributions to the limited empirical literature on the effects of the COVID-19 pandemic on the lodging industry. First, much of the extant studies in this context are descriptive in nature, and the impact of the COVID-19 pandemic on the lodging industry has largely been based on anecdotal evidence ( Lardieri, 2021 ). In a recent study, Ozdemir et al. (2021) showed that US hotels have experienced significant drops in demand during the early stages of the pandemic; however, the extent to which COVID-19 pandemic affects the lodging industry was not empirically investigated. In the present study, we examined and quantified the effect of the COVID-19 pandemic on the lodging industry across states. Second, while the COVID-19 pandemic might appear to have an adverse effect on the entire industry, the hotel industry is quite complex, with six distinct categories ranging from economy to luxury. As such, the COVID-19 pandemic might not have a uniform effect on these various hotel classes. However, the extent to which the COVID-19 pandemic affects the occupancy levels of various hotel classes has not been previously investigated. Thus, the effect of the COVID-19 pandemic on the economy, midscale, upper midscale, upscale, upper upscale, and luxury scale hotel classes was analyzed.

Next, the lodging industry comprises both traditional hotels and peer-to-peer accommodation properties listed for booking by companies such as Airbnb. It is possible that Airbnb property bookings were not equally affected by the COVID-19 pandemic. In a recent study, Dolnicar and Zare (2020) posed the question of whether COVID-19 has disrupted the disruptor (Airbnb). Following this, the extent to which the COVID-19 pandemic affected Airbnb property bookings, including all, entire home, private rooms, and shared room properties in the US was investigated. In so doing, this key question posed by Dolnicar and Zare (2020) was answered, in that results showed that Airbnb demand was not adversely impacted by the COVID-19 pandemic.

Finally, the COVID-19 pandemic is a major external shock to which policymakers in US states have responded differently in terms of regulations. Some states have ordered strict limitations on business openings, number of patrons allowed, hours of operation, large and even small gatherings, and social distancing (Multistate, 2021), while other states imposed almost no rules at all (Hunnicutt & Mason, 2021). This study examined the role of those different policy responses on the effect of the COVID-19 pandemic on the lodging industry. In so doing, analyses showed that while an external shock, such as the COVID-19 pandemic, might have adverse impacts on the lodging industry, policies and strategies implemented to deal with such external shocks have significant impacts on the economic actors of the society.

Accordingly, the present study contributes to the limited formative literature by analyzing and quantifying the impact of the COVID-19 pandemic on the lodging industry. Also, the current study further contributes to the extant literature through its temporal, inferential, and geographical scope. That is, the effect of the COVID-19 pandemic US hotel room and Airbnb property bookings was examined using available data during the most comprehensive timeframe. The research serves as contemporary empirical evidence of the implications of the COVID-19 pandemic on the lodging industry in the US. Results further showed a major external shock (i.e., COVID-19 pandemic) that appears to have adverse impacts to the economy, in general, and the lodging sectors, in particular, has an impact at varying degrees. While some sectors of the lodging industry are more resilient (e.g., peer-to-peer accommodation properties listed on Airbnb and midscale hotels), others are more vulnerable (e.g., upper upscale and luxury hotels). Furthermore, we showed that public health policies and regulations developed to mitigate the COVID-19 virus spread throughout the pandemic have had varying effects on its impacts. Therefore, such policy interventions should always be taken into consideration when analyzing the economic impact of external shocks, such as the COVID-19 pandemic.

5.2. Practical implications

The results of this study have significant managerial implications for the lodging industry and strategic implications for destinations and policymakers. First, although the COVID-19 pandemic is a major external shock that has had severe impacts to the overall economy and hospitality and tourism industry, the effect of the COVID-19 pandemic is not uniform across all platforms. Midscale hotels are the most resilient hotels against the COVID-19 pandemic, followed by economy-scale and upper midscale class hotels. Therefore, hotel corporations should increase midscale, economy-scale, and upper midscale class hotels in their portfolios, as they are the most resilient hotels against major external shocks.

Furthermore, the hotel industry has somewhat recognized that peer-to-peer accommodation networks are major competitors ( Dogru et al., 2020 ), and some hotel chains have even incorporated the home sharing model into their businesses. However, advertising privately-owned properties for guests is often limited to individuals, and commercial hotel companies may not be able to operate within the peer-to-peer accommodation market, in this sense. Therefore, to the extent of the limitations, hotel companies should endeavor to innovate commercially owned properties, as travelers have shifted to staying in more private properties, where they do not interact with other guests and large staff.

In addition, hotels can provide stronger value propositions than peer-to-peer accommodation operations by integrating their large-scale knowledge base, professional cleaning, hygiene standards, and technological infrastructure that enables contactless check-in and entry ( Jiang and Wen, 2020 , Kaushal and Srivastava, 2021 ). As hotel corporations have started to acquire properties and operate peer-to-peer accommodations, expanding their business model could also increase their resiliency during major external shocks, such as a pandemic, and it may also provide an opportunity to capitalize on the rapidly growing market. Offering private home and residential style property rentals, such as those advertised by individual owners on the peer-to-peer accommodation networks, would be a winning strategy for hotels, because loyal customers of major hotel corporations might not feel the need to shift their booking preferences to the properties listed on Airbnb. While peer-to-peer accommodation and home sharing business models appear to be ideal for consumers who desire to travel during the pandemic while maintaining social distancing with other travelers, incorporating experiential value propositions, such as localness, serendipity, communitas ( Mody et al., 2017 ) and a sense of “feeling at home” ( Song et al., 2021 ) through such properties would further strengthen the hotel businesses.

Furthermore, hotels could change their business models to decrease their reliance on group businesses, such as conferences, events, corporate meetings, and so on ( Hao et al., 2020 ). Similar to Airbnb’s local travel initiative, Go Near ( Jelski, 2020 ), hotels could attract customers from within or near the cities they operate. Consumer perceptions of safety, the ability and willingness to travel, and anxiety about the pandemic have all resulted in significant changes in consumer travel behavior ( Suess et. al, 2021 ). The highest standards of cleanliness, hygiene, and safety should be in the center of hotels’ marketing strategies, coupled with flexible booking and free cancellation options, and promotions that could alleviate consumers’ anxiety and restore confidence about traveling during the pandemic ( Shin et al., 2021 ). Also, the shift at many companies to working at home has given rise to a new group of customers who choose to rent Airbnb properties on a long-term basis so that they can explore new and different cities and countries while working remotely ( Hines, 2021 ). These digital nomads have generated new demand for the lodging sector, which appears to be captured by Airbnb properties ( Rodriguez, 2020 ). Hotels could offer extended stay options to these new groups of health-conscious digital nomads to create a new stream of revenues during the pandemic and thereafter.

Strict restrictions intensify the adverse impacts of the pandemic. Certainly, public health is of paramount importance during a pandemic. Specifically, mandatory business closures, limited hours of operation, curfews, stay-at-home orders, and other similar policies have had direct and significant effects on businesses. By imposing such strict policies, local and federal governments have lost significant sales, income, and other tax revenues that are essential to funding measures related to maintaining public health. While almost all US states implemented strict policy measures, some states, such as Florida and Texas, relaxed these restrictions significantly and allowed businesses to continue to operate as usual in early 2020. The benefits of these lenient policies have appeared in the results of the present study, showing that the impact of the COVID-19 pandemic to the lodging industry was lower in less-restrictive states. Policymakers in more-strictive states should reconsider their approaches to mitigating the effects of the ongoing pandemic by investigating the methods utilized in less-restrictive states to sustain their tax revenues and attract new investments in local communities from future hotel developers.

The results of the study showed that midscale hotels are the most resilient against the demand decreased related to the COVID-19 pandemic. Prior to the pandemic, Dogru et al. (2020) also showed that midscale hotels are the major hotel classes that contributes to the employment in local economies. Considering the impact that the midscale hotels have on local economies and their resilience to pandemic, local authorities should further incentivize midscale hotel development in their communities. Also, the resilience of the Airbnb peer-to-peer accommodation properties might appear to be a benefit for local communities as they bring in additional tax revenues for cities and local governments from visitor spending, especially during the COVID-19 pandemic. In fact, Airbnb properties, in general, was found to be resilient in both more-restrictive and less-restrictive states, suggesting that more restrictive states only imposed restrictions against traditional hotels or that hosts of Airbnb properties were not mandated by the same restrictions as hotels. Regardless, policymakers should consider regulating Airbnb properties in the same manner as the commercial accommodation counterparts to level the playing field. Otherwise, loose regulations and policies on privately operated properties might drive away major hotel investments to states or destinations that treat hotels and sharing economy properties similarly, where both are allowed to operate with little or no restrictions during a pandemic ( Dogru et al., 2019 ). Also, state and federal governments could initiate tax-incentives to encourage travelers to book hotels and/or Airbnb properties during pandemic to accelerate the recovery ( Ozdemir et al., 2021 , Salem et al., 2021 ).

5.3. Limitations and recommendations for future research

Although the present study makes a significant contribution to the extant literature, it has some limitations. The current study examines the effect of the COVID-19 pandemic on various sectors of the lodging industry. However, its sample is limited to the hotel categories and Airbnb property types specifically in the US. It is not clear whether the COVID-19 pandemic affected the lodging industry in other parts of the world, similarly. Future research is necessary to examine the extent to which the COVID-19 pandemic affects the lodging industry in other countries. Also, the classification of hotels in the study is based on class categories (STR). However, hotels in each class might vary in their organizational structures, such as franchised or managed, and the implications of the COVID-19 pandemic might be different. Therefore, future studies are necessary to examine these issues and compare the structures. Similarly, Airbnb properties can be compared based on management by single-unit hosts or multi-unit hosts. The examination of the impact of COVID-19 pandemic to Airbnb properties in this structural context will well serve the extant literature. The present study examined the effect of the COVID-19 pandemic on the demand for hotel and peer-to-peer accommodation properties listed on Airbnb. Finally, the impact of the COVID-19 pandemic to other sectors of the hospitality and tourism industry is not clear. Future studies should investigate the implications of the COVID-19 pandemic on restaurants, museums, and other sectors of the hospitality and tourism industry.

Declarations of interest

Acknowledgments.

We thank the anonymous reviewers for providing valuable feedback and constructive comments.

Appendix 1. 

Robustness analysis.

Our main analyses provide extensive evidence on the effects of the COVID-19 pandemic on hotels and the peer-to-peer accommodation properties listed on Airbnb. However, we further examined the effects using alternative measures of COVID-19 pandemic. While the COVID-19 pandemic might have immediate effect on the accommodation sectors, its effect could be lagged for several months. Therefore, a further examination of the lagged effects of the COVID-19 pandemic on hotels and Airbnb was warranted. Table 5 presents these findings.

The Effect of Lagged COVID-19 on Hotels and Airbnb Occupancy.

The analyses yielded results similar to that of the main findings. That is, the results show that the effect the COVID-19 pandemic on hotels continues to be significant in one-month ( β = − 0.003 , p < 0.05 ), two-month ( β = − 0.004 , p < 0.01 ), and three-month ( β = − 0.004 , p < 0.01 ) lagged total number of COVID-19 cases. While the magnitudes of the effect of the COVID-19 pandemic on hotels are slightly lower, the impact of the COVID-19 pandemic persists for at least three months. However, the effect of COVID-19 pandemic on Airbnb properties remains to be insignificant in a lagged examination of the effect of the COVID-19 pandemic. These results confirm our findings that Airbnb properties are more resilient against the COVID-19 pandemic than hotels.

In addition to the examination of the effect of lagged values of COVID-19 cases, we also analyzed the effect of COVID-19 pandemic on hotels and Airbnb properties utilizing alternative measures of COVID-19 pandemic. Specifically, we utilized total number of COVID-19 deaths, new number of COVID-19 cases, and new number of COVID-19 deaths to analyze the effect of COVID-19 pandemic on hotels and Airbnb. Table 6 presents these results.

The Effect of COVID-19 on Hotels and Airbnb Occupancy: Alternative COVID-19 Measures.

The results from the Panel A of Table 6 validates our main findings that the COVID-19 pandemic has an adverse impact on hotel occupancy rates. The magnitudes of the effect are similar to that our initial findings, suggesting that the effect of the COVID-19 pandemic on hotels does not change based on the measure of the COVID-19 pandemic. That is, regardless of the COVID-19 pandemic variable, the effect of the COVID-19 pandemic on hotels remains negative and statistically significant with similar findings in magnitude when the total number of COVID-19 deaths, new number of COVID-19 cases, or new number of COVID-19 death measures ( β = − 0.008 , p < 0.01 ) were utilized to examine the effect of COVID-19 pandemic on hotels.

These results further confirm the initial analysis findings of the effect that the COVID-19 pandemic has on peer-to-peer accommodation properties listed on Airbnb, which continues to be insignificant. That is, COVID-19 pandemic does not appear to have any adverse impacts on Airbnb properties when alternative measures of COVID-19 pandemic are utilized. These results suggest that our results are robust to alternative measures of the COVID-19 pandemic, and collectively validate our findings that Airbnb property bookings did not experience any adverse impacts from the COVID-19 pandemic and appear to be more resilient than hotels, although hotels experienced economically and statistically significant impacts due to the COVID-19 pandemic.

The anonymous reviewer suggested to examine the effect of COVID-19 pandemic on hotels and Airbnb occupancy by adjusting the total COVID-19 cases by respective state populations. These results are presented in Table 7 .

The Effect of COVID-19 on Hotels and Airbnb Occupancy: Alternative Time Periods and Measures.

The results from Panel A of Table 7 show that analyzing the pandemic period alone yields results similar to that of initial analysis. That is the effect of the COVID-19 pandemic on hotels and Airbnb occupancy remains similar regardless of the time period. Furthermore, the findings from the analysis presented in Panel B of Table 7 show coefficients similar to that of initial findings, suggesting that the state population does not have any significant impact on the effect of COVID-19 pandemic on hotels and Airbnb.

Furthermore, in a recent study Deyá-Tortella et al. (2022) argued that COVID-19 pandemic has led to significant changes in tourist behavior and hence there might be structural shifts in this context. Considering the nature of this study’s data period containing the COVID-19 pandemic, we have further tested for the existence structural breaks in the data following the procedures in Ditzen et al. (2021) . The results showed that the data contains five structural breaks. Therefore, we re-run the analysis on all hotels and all Airbnb samples considering the existence of the structural breaks following the methodology proposed by the study of Ditzen et al. (2021) . Table 8 presents these findings.

The Effect of COVID-19 on Hotels and Airbnb Occupancy: Testing for Structural Breaks.

The results yield estimates similar to that of the results from the main analysis. While the coefficient estimates changed slightly, the changes are not significant. Therefore, our estimates are valid and robust to alternative specifications. Overall, the results from the robustness tests suggest that our results are robust and collectively validate our findings.

Appendix 2. 

Notes: b and c denote 5% and 10% statistical significance levels, respectively. ns indicates not significant. All other correlations results are significant at 1% statistical significance level.

• Airbnb, n.d., Select your home type. Airbnb. 〈https://www.airbnb.com/help/article/317/select-your-home-type〉.
• Bresciani S., Ferraris A., Santoro G., Premazzi K., Quaglia R., Yahiaoui D., Viglia G. The seven lives of Airbnb. The role of accommodation types. Ann. Tour. Res. 2021; 88 [ Google Scholar ]
• Carville, O. & Bloomberg, 2021, February 25, Airbnb revenue beats estimates, showing demand amid COVID surge. Fortune. 〈https://fortune.com/2021/02/25/airbnb-revenue-beats-estimates-amid-covid-surge/〉.
• Deyá-Tortella B., Leoni V., Ramos V. COVID-led consumption displacement: A longitudinal analysis of hotel booking patterns. Int. J. Hosp. Manag. 2022; 107 [ PMC free article ] [ PubMed ] [ Google Scholar ]
• Ditzen, J., Karavias, Y., & Westerlund, J. (2021). Testing and estimating structural breaks in time series and panel data in stata. arXiv preprint arXiv:2110.14550 .
• Dogru T., Mody M., Suess C. Adding evidence to the debate: Quantifying Airbnb's disruptive impact on ten key hotel markets. Tour. Manag. 2019; 72 :27–38. [ Google Scholar ]
• Dogru T., McGinley S., Kim W.G. The effect of hotel investments on employment in the tourism, leisure and hospitality industries. Int. J. Contemp. Hosp. Manag. 2020 [ Google Scholar ]
• Dolnicar S., Zare S. COVID19 and Airbnb–Disrupting the disruptor. Ann. Tour. Res. 2020; 83 [ PMC free article ] [ PubMed ] [ Google Scholar ]
• Eggleston, K. & Lee, M., 2021, August 8, Hospitality industry moves into post-pandemic recovery mode. HotelManagement.Net. 〈https://www.hotelmanagement.net/operate/hospitality-industry-moves-post-pandemic-recovery-mode〉.
• Gursoy D., Chi C.G. Effects of COVID-19 pandemic on hospitality industry: Review of the current situations and a research agenda. J. Hosp. Mark. Manag. 2020; 29 :5. [ Google Scholar ]
• Hao F., Xiao Q., Chon K. COVID-19 and China’s hotel industry: Impacts, a disaster management framework, and post-pandemic agenda. Int. J. Hosp. Manag. 2020; 90 [ PMC free article ] [ PubMed ] [ Google Scholar ]
• Hines, M., 2021, February 9, As work and school can be done from anywhere, Airbnb sees more monthly stays. USA Today. Retreived from 〈https://www.usatoday.com/story/travel/hotels/2021/02/09/airbnb-rentals-monthly-stays-up-covid-travel-meets-virtual-life/4375982001/〉.
• Hoechle D. Robust standard errors for panel regressions with cross-sectional dependence. stata J. 2007; 7 (3):281–312. [ Google Scholar ]
• Jelski, C., 2020, Airbnb increases focus on local travel. Travel Weekly. Retrieved from 〈https://www.travelweekly.com/Travel-News/Hotel-News/Airbnb-increases-focus-on-local-travel〉.
• Jiang Y., Wen J. Effects of COVID-19 on hotel marketing and management: a perspective article. Int. J. Contemp. Hosp. Manag. 2020 [ Google Scholar ]
• Johnson, H., 2021, North American inbound tourism spend declined by 74.1% in 2020. eTurboNews. 〈https://eturbonews.com/3000768/north-american-inbound-tourism-spend-declined-by-74–1-in-2020/〉 Accessed Aug. 23, 2021.
• Kaushal V., Srivastava S. Hospitality and tourism industry amid COVID-19 pandemic: Perspectives on challenges and learnings from India. Int. J. Hosp. Manag. 2021; 92 [ PMC free article ] [ PubMed ] [ Google Scholar ]
• Lai I.K.W., Wong J.W.C. Comparing crisis management practices in the hotel industry between initial and pandemic stages of COVID-19. Int. J. Contemp. Hosp. Manag. 2020 [ Google Scholar ]
• Lardieri, A. (2021). Coronavirus Pandemic Sets Hotel Industry Back 10 Years. U.S. News and World Report . 〈https://www.usnews.com/news/national-news/articles/2021–01-27/coronavirus-pandemic-sets-hotel-industry-back-10-years-report-finds〉 Accessed Aug. 23, 2021.
• McCann, A. (2021, April 6). States with the fewest Coronavirus restrictions . WalletHub. 〈https://wallethub.com/edu/states-coronavirus-restrictions/73818〉.
• Mody M.A., Suess C., Lehto X. The accommodation experiencescape: a comparative assessment of hotels and Airbnb. Int. J. Contemp. Hosp. Manag. 2017 [ Google Scholar ]
• Ozdemir O., Dogru T., Kizildag M., Mody M., Suess C. Quantifying the economic impact of COVID-19 on the US hotel industry: Examination of hotel segments and operational structures. Tour. Manag. Perspect. 2021; 39 [ PMC free article ] [ PubMed ] [ Google Scholar ]
• Rodriguez, S., 2020, December 12, Airbnb nomads are ditching their apartments in pandemic-ridden cities and living wherever they want. 〈https://www.cnbc.com/2020/12/12/airbnb-nomads-ditch-apartments-during-covid.html〉.
• Salem I.E., Elbaz A.M., Elkhwesky Z., Ghazi K.M. The COVID-19 pandemic: The mitigating role of government and hotel support of hotel employees in Egypt. Tour. Manag. 2021; 85 [ PMC free article ] [ PubMed ] [ Google Scholar ]
• Schulz, B., 2021, August 14, I just didn’t feel like it was responsible: Growing number of travelers cancel trips as COVID-19 cases surge. USA Today. 〈https://www.pointswithacrew.com/hotels-categorize-difference-upscale-upper-upscale/〉.
• Shin H., Sharma A., Nicolau J.L., Kang J. The impact of hotel CSR for strategic philanthropy on booking behavior and hotel performance during the COVID-19 pandemic. Tour. Manag. 2021; 85 [ PMC free article ] [ PubMed ] [ Google Scholar ]
• Song S., Suess C., Mody M.A., Dogru T. Comparing the influence of substantive and communicative servicescape on healthcare traveler emotions: the moderating effect of accommodation type and interior design style. Int. J. Contemp. Hosp. Manag. 2021 [ Google Scholar ]
• STR Chain Scales, 2019, October 29, STR. Retrieved August 19, 2021, from 〈https://str.com/sites/default/files/2019–10/STR-Chain%20-Scales-20191025_0.pdf〉.
• Suess C., Maddock J., Dogru T., Mody M., Lee S. Using the Health Belief Model to examine travelers’ willingness to vaccinate and support for vaccination requirements prior to travel. Tour. Manag. 2021; 88 [ PMC free article ] [ PubMed ] [ Google Scholar ]
• Suess C., Maddock J.E., Dogru T., Mody M., Lee S. Using the Health Belief Model to examine travelers’ willingness to vaccinate and support for vaccination requirements prior to travel. Tour. Manag. 2022; 88 [ PMC free article ] [ PubMed ] [ Google Scholar ]
• Verleye K., Leroi-Werelds S., Line N., Bove L. Value proposition dynamics in response to external event triggers. J. Bus. Res. 2021 [ Google Scholar ]
• Voltes-Dorta A., Sánchez-Medina A. Drivers of Airbnb prices according to property/room type, season and location: A regression approach. J. Hosp. Tour. Manag. 2020; 45 :266–275. [ Google Scholar ]

• Research article
• Open access
• Published: 15 April 2024

COVID-19 inequalities in England: a mathematical modelling study of transmission risk and clinical vulnerability by socioeconomic status

• Lucy Goodfellow   ORCID: orcid.org/0009-0004-0434-5863 1 ,
• Edwin van Leeuwen 1 , 2 &
• Rosalind M. Eggo 1  

BMC Medicine volume  22 , Article number:  162 ( 2024 ) Cite this article

1 Altmetric

Metrics details

The COVID-19 pandemic resulted in major inequalities in infection and disease burden between areas of varying socioeconomic deprivation in many countries, including England. Areas of higher deprivation tend to have a different population structure—generally younger—which can increase viral transmission due to higher contact rates in school-going children and working-age adults. Higher deprivation is also associated with a higher presence of chronic comorbidities, which were convincingly demonstrated to be risk factors for severe COVID-19 disease. These two major factors need to be combined to better understand and quantify their relative importance in the observed COVID-19 inequalities.

We used UK Census data on health status and demography stratified by decile of the Index of Multiple Deprivation (IMD), which is a measure of socioeconomic deprivation. We calculated epidemiological impact using an age-stratified COVID-19 transmission model, which incorporated different contact patterns and clinical health profiles by decile. To separate the contribution of each factor, we considered a scenario where the clinical health profile of all deciles was at the level of the least deprived. We also considered the effectiveness of school closures and vaccination of over 65-year-olds in each decile.

In the modelled epidemics in urban areas, the most deprived decile experienced 9% more infections, 13% more clinical cases, and a 97% larger peak clinical size than the least deprived; we found similar inequalities in rural areas. Twenty-one per cent of clinical cases and 16% of deaths in England observed under the model assumptions would not occur if all deciles experienced the clinical health profile of the least deprived decile. We found that more deaths were prevented in more affluent areas during school closures and vaccination rollouts.

Conclusions

This study demonstrates that both clinical and demographic factors synergise to generate health inequalities in COVID-19, that improving the clinical health profile of populations would increase health equity, and that some interventions can increase health inequalities.

Peer Review reports

The COVID-19 pandemic disproportionately affected people in lower socioeconomic groups around the world [ 1 , 2 ]. In England, there were large disparities in COVID-19 burden between areas of different relative deprivation, measured by the Index of Multiple Deprivation (IMD). Initial reports by the UK Office for National Statistics (ONS) found that from April to July 2020, the most deprived 10% of areas in England experienced an age-standardised COVID-19-related mortality rate more than twice as high as the least deprived 10% [ 3 ]. These disparities were repeatedly observed throughout the pandemic: between June 2020 and January 2021, the age-standardised mortality rate in laboratory-confirmed cases of COVID-19 was 371.0 per 100,000 (95% confidence interval (CI) 334.2–410.7) compared to 118.0 (95% CI 97.7–141.3) in the most vs least deprived quintiles [ 4 ]. The inequality in mortality rates seen in the early pandemic exceeded that observed in previous years, indicating that there were further factors exacerbating the ‘expected’ effects of relative deprivation [ 5 ]. Even after adjusting for age, sex, region, and ethnicity, this report found worse outcomes in more deprived areas but did not adjust for the prevalence of comorbidities. Other studies have consistently confirmed an association between comorbidities and more severe COVID-19 outcomes [ 6 , 7 ].

Morbidity and the presence of underlying health conditions tend to vary greatly by socioeconomic status (SES) and are a significant risk factor for severe infection [ 8 ]. Vulnerability to more severe infection has both direct effects, including a greater risk of consequential long-term health complications and greater mortality risk, and indirect population-level effects, such as potentially increased infectiousness of symptomatic cases. In England, before the COVID-19 pandemic, life expectancy was 9.4 years longer for men in the least deprived decile than the most and 7.7 years longer for women [ 9 ]. These gaps continue to widen: female life expectancy in the most deprived decile fell by 4 weeks between 2014–2016 and 2017–2019 but rose by 11 weeks in the least deprived [ 9 , 10 ]. The prevalence of underlying health conditions that affect the quality of life is also consistently correlated with local deprivation levels: men in the most deprived decile could expect to live 18.4 years fewer in good health than those in the least deprived decile; the corresponding gap for women was 19.8 years [ 9 ].

It is well established that infectious disease burden is associated with SES [ 11 , 12 , 13 ]. This is linked to a multitude of complex and interwoven factors including, but not limited to, lack of access to healthcare, poor housing conditions, inability to avoid high-exposure settings such as crowded public places, differences in occupation type, and avoiding restrictions or testing due to mistrust of authorities [ 14 , 15 ].

Here, we use a novel transmission model to combine the differences in risk of infection and risk of severe disease infection between areas of different relative deprivation, as measured by the IMD, to quantify their relative importance in the observed COVID-19 inequalities in England. We consider underlying health conditions as a key determinant of an individual’s risk of developing a clinical case of COVID-19 and focus on the impact of IMD-specific health and age structure on infectious disease burden at the population level. To represent the health status of each decile, we used the UK Census 2021 self-reported health responses. This variable is also used by the ONS to calculate healthy life expectancy and to compare health-related well-being in subpopulations of England [ 9 ]. By making simplifying assumptions and modelling a synthetic population, we aim to produce a conceptual exploration of the interaction between underlying health and demographic structure.

We developed an age-stratified dynamic transmission model for SARS-CoV-2, which was further stratified by IMD decile, and by urban or rural classification in England. Here, we detail how the model was modified to incorporate the characteristics of each decile and geography.

IMD-specific age structure

Each epidemic was simulated on the population of a given IMD decile in either an urban or rural area, to account for the distinct underlying age structures in these areas. We used 17 age groups (0–1, 1–5, every 5 years to 75, and over 75). The mid-2020 (30 June) age-specific population of each lower layer super output area (LSOA), which is on average 1500 people, was linked via LSOA codes to their IMD decile and urban/rural classification (where urban is defined as a settlement with over 10,000 residents) [ 16 , 17 , 18 , 19 ]. We calculated the size of each age group, specific to each IMD decile and geography, and used this to determine the average age structure of each IMD- and geography-specific population, \(n=({n}_{1},\dots ,{n}_{17})\) , where \(\sum\limits_{a=1}^{17}{n}_{a}=1\) in each population. We also calculated the median age for each urban and rural IMD decile and the proportion of each IMD decile residing in urban or rural LSOAs (Additional file 1 : Section 1).

Contact matrices

To define contact between the age groups, we used age-specific social contact data for the United Kingdom (UK) for physical and conversational contacts, accessed via the socialmixr R package [ 20 , 21 ]. The contact matrices are highly age-assortative, with the highest daily contact patterns occurring between individuals in the same age group for those aged 5–19. We projected the contact patterns onto the age structure of each IMD- and geography-specific population in 2020, using the density correction method, by constructing an intrinsic connectivity matrix and scaling this matrix to match the population’s age structure [ 22 ].

The intrinsic connectivity matrix was calculated from the 2006 UK contact matrix \({M}^{2006}={\left({M}_{ij}^{2006}\right)}_{i,j=1,\dots ,17}\) and age structure \({N}^{2006}=\left({N}_{1}^{2006}, \dots ,{N}_{17}^{2006}\right)\) as follows:

The new contact matrix for a population with age group sizes \(N=\left({N}_{1},\dots ,{N}_{17}\right)\) and proportions \(n=\left({n}_{1},\dots ,{n}_{17}\right)\) had entries:

Age-specific fraction of COVID-19 cases causing clinical symptoms

We separated infections of SARS-CoV-2 as in [ 23 ], into clinical or subclinical cases. Clinical cases of COVID-19 are infections that lead to noticeable symptoms such that an individual may seek clinical care. Subclinical infections do not seek care and are assumed to be less infectious than clinical cases. We defined a population’s clinical fraction as the probability of an individual in the population developing a clinical case of COVID-19 upon infection. Here, we related an individual’s probability of being a clinical case of COVID-19 to the self-reported health status of their IMD- and age-specific population in England, as a proxy for the relative presence of comorbidities in each population, and then examined how differences in self-reported health status by IMD decile, coupled with differences in age distribution, affect the burden in each IMD decile.

To define health status, we used data from the 2021 Census, specifically the question ‘How is your health in general?’, with response options of ‘very good’, ‘good’, ‘fair’, ‘bad’, and ‘very bad’ [ 24 ]. This is provided by the Census stratified by IMD and by age. We then defined ‘health prevalence’ as the proportion of individuals reporting ‘very good’ or ‘good’ general health, stratified by the same age groups and the deciles of IMD:

To map a population’s health prevalence to clinical fraction, we used locally weighted regression (LOESS), which fits a smooth curve without any assumptions about the underlying distribution of the data, trained on age-specific health prevalence data from Census 2021 and age-specific clinical fraction values from Davies et al. [ 23 , 24 ]. Any populations with health prevalences outside of the training dataset’s range were assigned the most extreme clinical fractions found by Davies et al. [ 23 ], to avoid extrapolation outside of observed values. Health prevalence was highest in children, but children have separate risk factors for severe disease (such as smaller airways), and children under 10 have been found to be subject to a higher risk of clinical COVID-19 cases and a greater infection fatality ratio (IFR) [ 23 , 25 ] (as observed for other infections such as influenza [ 26 ]). Therefore, we fixed the clinical fraction of the 0–9 age group at 0.29, matching that found by [ 23 ].

COVID-19 transmission model

The transmission model includes a single SARS-CoV-2 variant, no existing immunity in the population, and natural history parameters drawn from the first wave of the pandemic. We considered the non-pharmaceutical intervention (NPI) of school closures and also explored the effect of vaccinating adults over the age of 65. We developed an age- and IMD-stratified deterministic compartmental model in R (version 4.3.1) (Fig. 1 c). There is no mixing between IMD deciles in the model. The aim is to demonstrate the importance of health prevalence and differences in age and social mixing in epidemic impact, rather than to reproduce the COVID-19 epidemic in England.

a Proportion of each geography-specific IMD decile in each age group. b Age- and IMD-specific health prevalence (1, most deprived decile; 10, least deprived). c Age-stratified SEIRD model, specific to IMD decile and geography. Subscript a denotes age-specificity, c clinical parameters, and s subclinical parameters

Individuals are first assumed to be susceptible ( S ) and become exposed ( E ) but not yet infectious after effective contact with an infected individual (Fig. 1 c). Each exposed individual then progresses to one of two infected states: subclinical infection ( Is ) and clinical infection, which is represented by a pre-symptomatic (but infectious) compartment ( Ip ) followed by a symptomatic compartment ( Ic ). Each individual then moves into the recovered ( R ) or dead ( D ) compartment, at which point they are assumed to no longer be infectious and to be immune to infection. This susceptible-exposed-infectious-recovered-dead (SEIRD) is an extension of [ 23 ], with the addition of a D compartment. We ran the epidemic for 365 days, which allowed the completion of each epidemic in each decile and geography. Each epidemic was run on a synthetic population of a fixed IMD decile and urban/rural geography, with no births, non-infection-related deaths, or ageing between the age groups, as the time frame of each epidemic was less than a year. The model also assumed that contact patterns remain constant throughout the epidemic.

The force of infection in age group k is given by:

where \(p\) is the probability of a contact between an infected and susceptible individual resulting in transmission of infection, \({M}_{ak}\) is the mean daily number of contacts that an individual in age group a has with individuals in age group k , and \(\xi\) is the relative infectiousness of subclinical cases. The age-specific clinical fraction is denoted by \({\pi }_{a}\) and depends on the IMD decile. Rates of transition from each disease state are given in Table 1 .

We assumed the relative subclinical infectiousness ( \(\xi\) ), to be equal to 0.5, and tested this assumption in a sensitivity analysis (see Additional file 1 : Section 12). The transmission probability during a contact was assumed to be \(p=0.06\) as in [ 23 ]. The remaining parameter estimates were taken from [ 23 ] where possible, to replicate the conditions used to derive the clinical fraction estimates. The mortality probability of subclinical infections was assumed to be 0 for all age groups ( \(a\) ). The age-specific probability of mortality of clinical cases was estimated using age-specific IFRs \(\left({\phi }_{a}\right)\) found by Verity et al. in 2020 [ 27 ] (Additional file 1 : Table S4). As the IFR is \({\phi }_{a}={\pi }_{a}{\mu }_{ca}+\left(1-{\pi }_{a}\right){\mu }_{sa}={\pi }_{a}{\mu }_{ca}\) , since \({\mu }_{sa}=0\) , the age-specific clinical mortality probabilities were estimated by:

where \({\pi }_{a}\) is the age-specific clinical fractions for the general population in [ 23 ] (Additional file 1 : Table S4).

We calculated the total infections, clinical cases, and fatalities per 1000 people, the peak number of clinical cases per 1000 people, the IFR, and the basic reproduction number ( R 0 ) for each IMD decile in urban and rural areas. We also calculated age-standardised measures of total infections, clinical cases, and fatalities within a specific geography for increased comparability. The age-standardised results were of the form:

where \({n}^{u}=\left({n}_{1}^{u},\dots ,{n}_{17}^{u}\right)\) is the standard urban population, defined as the proportion of people living in urban LSOAs who are in each age group, similarly \({n}^{r}=\left({n}_{1}^{r},\dots ,{n}_{17}^{r}\right)\) for rural areas.

R 0 in each IMD decile in urban and rural areas was calculated as the absolute value of the largest eigenvalue of the next-generation matrix N :

Counterfactual scenarios

To determine the epidemic burden attributable to the difference in underlying health status between IMD deciles, we created the counterfactual health prevalence scenario, where all deciles were assigned the age-specific health prevalence of decile 10 (the least deprived). We calculated the total clinical cases and fatalities in each IMD decile under this assumption. In order to reflect the size of each population (while each IMD decile comprises 10% of the population of England, geography-specific IMD deciles vary widely in size, see Additional file 1 : Table S1), we scaled mortality to mid-year 2020 population sizes and totalled over the 20 populations.

We also created the counterfactual scenario of constant age structure, where we held the age structure constant at the average of each geography-specific England population, independent of the IMD decile. This allowed us to determine the impact of clinical vulnerability separately from the differences in age distribution in each IMD decile. The health prevalence by age remained at the IMD-specific value.

School closures

School closures were a major NPI implemented in the UK during the pandemic, and were implemented evenly across all IMD deciles, unlike some other contact-reducing interventions. We therefore modelled school closures to determine the impact of this intervention across IMD deciles. To quantify the potential differences in the impact of school closures in different IMD deciles, we calculated the effect of school closures on R 0 and total fatalities. The social contact data used is a combination of location-specific contact matrices, defined by home, work, school, and other locations. We removed the school-specific contacts from the contact matrix (retaining contacts in home, work, and other locations), re-projected onto the 2020 age structure, and recalculated the next-generation matrix, N , and its largest eigenvalue, R 0 . While assuming that the closure of schools results in a complete subtraction of school-specific contacts may not be realistic (as some contacts would likely be replaced by social interactions in other locations [ 28 ]), the results demonstrate the maximum potential impact of school closures.

We simulated the closure of schools after a certain cumulative proportion, P , of the population developed clinical COVID-19 cases. The use of cumulative clinical cases as a threshold for implementation is reflective of using total confirmed cases as a measure of the size of an early epidemic. We assumed a value of P = 0.05 but tested different values in sensitivity analyses (Additional file 1 : Section 11).

Vaccinations

To quantify the relative impact of vaccination rollouts on populations of different levels of deprivation, we calculated the change in mortality rates in each population after vaccinating all adults over the age of 65. This correlates with the earliest vaccination programmes in England, where the first target populations were individuals of older ages. We assumed that vaccination reduced the likelihood of an individual developing a clinical case of COVID-19 upon infection but did not prevent infection. We assumed 76.5% vaccine efficacy against symptomatic infection [ 29 ] and reduced the clinical fraction of vaccinated individuals in line with this estimate. To estimate the maximum impact of vaccination, we assumed coverage in over 65s of 100%. We then calculated the change in mortality rates and the number of deaths prevented in each population. We also calculated how many vaccine doses would be given to each population.

Self-reported health prevalence is lower in more deprived areas

There was an older age structure in rural areas compared to urban and a generally younger age structure in more deprived areas (Fig. 1 a). The relationship between IMD decile and age structure was confirmed by the median age in each population (Additional file 1 : Fig. S1); rural areas have consistently higher median ages than urban areas of the same IMD decile. The median age monotonously increased with affluence in urban areas but peaked in the fourth decile for those living in rural areas.

Age-specific health prevalence was consistently lower in more deprived areas (Fig. 1 b). Forty-seven per cent of those aged 65–69 living in the most deprived decile reported living in ‘very good’ or ‘good’ health, compared to 80% of those in the least deprived decile. Those living in the most deprived decile experienced the same health prevalence (76%) at ages 40–44 as those in the least deprived decile did at ages 70–74.

Health prevalence was mapped to a clinical fraction in the age groups used in [ 23 ] as described in the ‘ Methods ’ section (Fig. 2 a). Under this assumption, all those over the age of 10 in more deprived areas had a greater likelihood of developing a clinical case of COVID-19 than in other deciles (Fig. 2 b).

Results of mapping underlying health to clinical vulnerability. a The training dataset of age-specific health prevalence and clinical fraction estimates for the general population of England over age 10, and corresponding predictive model, with linear extensions outside the domain [0.21, 0.69]. b Resulting age- and IMD-specific clinical fractions (1, most deprived decile; 10, least deprived)

Epidemic burden increases with relative deprivation

We found that total infections and clinical cases increased with deprivation (Fig. 3 a, b). In rural settings, the most deprived decile experienced 72 more crude infections per 1000 population than the least deprived decile; this inequality increased to 90 infections in urban settings. The inequalities in clinical cases were even larger: in rural areas, the most deprived decile experienced 147 more clinical cases per 1000 than the least and 130 more clinical cases in urban areas. The peak clinical epidemic size was 97% larger in urban areas of the most deprived decile than the least deprived decile under these model assumptions and 91% larger in rural areas (Fig. 3 c).

Measures of the size of a COVID-19 epidemic in each IMD decile and geography. Solid lines represent crude measures, and dashed lines represent those age-standardised by geography. The most deprived decile is decile 1, and the least is decile 10. a Total infections per 1000 population. b Total clinical cases per 1000 population. c Clinical cases per 1000 population at the clinical peak of the epidemic. d Total deaths per 1000 population. e Infection fatality ratio. f Basic reproduction number, R 0

Mortality inequalities differed between the crude and age-standardised results (Fig. 3 d). The crude total number of deaths by IMD decile and geography closely followed the median age (Additional file 1 : Fig. S1). There was a strong positive association between increasing relative deprivation (decreasing decile) and the age-standardised number of deaths (Fig. 3 d). In urban areas, 2.0 more deaths occurred per 1000 age-standardised population in the most deprived decile than the least; this inequality increased to 2.9 deaths per 1000 age-standardised population in rural areas. The IFR followed a very similar pattern to crude mortality (Fig. 3 e), likely due to a combination of the relative stability of total infections with deprivation compared to the large variation in mortality rates, and the strong relationship between median age and mortality.

R 0 was generally higher in more deprived areas (Fig. 3 f) and ranged from 2.09 in rural areas of the 7th decile to 2.71 in urban areas of the most deprived decile. R 0 was not strongly related to the median age because the lower clinical fractions in younger populations were counteracted by their higher contact rates.

Rural areas experienced fewer total infections, lower peak clinical sizes, and lower R 0 than urban areas, but more clinical cases and deaths, at all levels of deprivation. This is likely due to the older rural age structure, as older individuals had fewer daily contacts than younger individuals and so produced fewer secondary infections but were more likely to develop clinical COVID-19 if infected.

Further sensitivity analyses considering epidemiological parameters show consistent patterns of age-standardised mortality by deprivation, but a change in the pattern of crude deaths (Additional file 1 : Section 12). In particular, if subclinical cases experience a similar level of infectiousness to clinical cases, then crude deaths are more dependent on the age structure and therefore higher in less deprived areas, but in the case that subclinical cases are relatively much less infectious, more crude deaths are consistently observed in more deprived areas (Additional file 1 : Fig. S11).

Health-attributable deaths occur at all ages

Under the counterfactual health prevalence scenario, 340,532 deaths occurred, compared to 405,695 under the original assumption. Therefore, 16% of deaths, or over 65,000 fatalities, would have been prevented by achieving health prevalence equity at the level of the least deprived decile. These health-attributable deaths did not only occur in those at older ages: over 29,000 prevented deaths were in individuals aged under 65 (Additional file 1 : Fig. S6). At all ages between 30 and 70, over 20% of deaths that occurred under the original model assumptions were attributable to underlying health inequalities (Additional file 1 : Fig. S7). We similarly found 21% of clinical cases (3.8 million) to be attributable to inequalities in underlying health under the model assumptions.

Lower clinical infection and mortality rates occurred in the most deprived areas, in both urban and rural geographies in the counterfactual health prevalence scenario (Fig. 4 ). Age-standardised deaths were consistent across IMD deciles in both geographies when clinical fraction was only dependent on age (Additional file 1 : Fig. S10), as is the case in the counterfactual health prevalence scenario. This result contradicts observed mortality rates [ 3 , 4 ], providing evidence for the existence of a dependency of clinical vulnerability on IMD and more specifically underlying health. The true relationship between IMD and age-specific clinical fraction may be more complex than the assumptions made in this paper; for example, pre-existing immunity may be dependent on previous exposure to coronaviruses [ 30 ], which may be associated with SES but is not considered here.

Epidemiological burden in counterfactual scenarios. a Total clinical cases per 1000 population, in geography-specific areas of each IMD decile (1, most deprived decile; 10, least deprived), in the counterfactual health prevalence scenario, and in the counterfactual constant age structure scenario. The original model is shown for comparison in pale lines. b Total deaths per 1000 population, in geography-specific areas of each IMD decile, under the same scenarios

In the counterfactual scenario of constant age structure, we observed more clinical cases and deaths in more deprived areas (Fig. 4 ). We also considered an underlying age structure independent of IMD decile or geography and found that the most deprived decile experiences 40% higher mortality and a clinical peak 1.88 times larger than the least deprived decile (Additional file 1 : Fig. S5), demonstrating the inequality resulting from health prevalence separately from demographic differences. These results indicate that observed inequalities in clinical case numbers and mortality are the result of a complex interaction between comorbidity-related clinical vulnerability and a population’s demographic structure, the outcome of which is not necessarily consistently related to deprivation.

School closures and vaccinations prevent more deaths in less deprived areas

With school closures in place in the model, R 0 decreased for all geographies and IMD deciles but remained larger in urban than rural areas and was consistently higher in more deprived areas in both geographies (Fig. 5 a). In urban areas, R 0 was 0.38 higher in the most deprived decile than in the least; the equivalent inequality was 0.29 in rural areas. The largest reductions in R 0 occurred in the most and least deprived deciles, with the least impact in the median deciles (Fig. 5 b). This U-shaped result is likely a product of the age structure of each population, as R 0 is driven by both high daily contact patterns in young individuals and greater clinical vulnerability in older individuals (more detail in Additional file 1 : Section 11). In all IMD deciles, greater reductions in R 0 occurred in urban than rural areas, likely due to the greater proportion of school-aged children and hence larger reduction in contacts. In no scenario was R 0 reduced below 1 (Fig. 5 a), meaning that school closures were not able to halt COVID-19 transmission in any rural or urban IMD decile and could only reduce the epidemic burden under our model assumptions.

Results of implementing school closures. a R 0 in each IMD- and geography-specific population (1, most deprived decile; 10, least deprived), before (pale lines) and after school closures. b Reductions in R 0 due to school closures. c Crude (solid lines) and age-standardised by geography (dashed lines) reductions in deaths observed per 1000 population after implementing school closures at P = 0.05

By implementing school closures after 5% of the population experienced a clinical case of COVID-19 ( P = 0.05), 0.113 more crude deaths were prevented per 1000 people in the least deprived urban areas than the most deprived, with a corresponding difference of 0.073 deaths per 1000 people in rural areas (Fig. 5 c). This is likely due to a combination of more crude deaths occurring in more affluent deciles without intervention, improved health conditions, and older population structures. The deaths prevented when age-standardised by geography were approximately consistent with IMD. We also investigated the pattern of prevented mortality when changing the school closure implementation threshold (Additional file 1 : Section 11) and found that the effectiveness of school closures in less deprived areas decreased dramatically as P increased.

The reductions in crude mortality rates associated with vaccinating all over 65s were higher in less deprived urban populations, and peaked in the central deciles of rural populations, due to the age distribution of those deciles (Fig. 6 a). However, reductions in age-standardised mortality rates were consistently higher in more deprived areas; this is likely to be due to higher clinical vulnerability and therefore a greater absolute reduction in clinical fractions in vaccinated individuals.

Results of vaccinating the over 65-year-olds. a Deaths prevented per 1000 population, after vaccinating all adults over 65. b Total number of deaths prevented by vaccination in each decile (stratified by urban and rural areas). c Total number of vaccine doses given in each decile (stratified by urban and rural areas) when vaccinating all over 65s

When these mortality reductions are considered across the whole population, more deaths were generally prevented in more affluent areas, with over 25,000 deaths prevented in the least deprived deciles, compared to less than 18,000 in the most deprived decile (Fig. 6 b). Similarly, more infections were prevented in less deprived areas on a population level, with over 106,000 infections prevented in the least deprived decile compared to just over 36,000 in the most deprived (Additional file 1 : Fig. S13b). Fewer vaccination doses were given in more deprived deciles (Fig. 6 c), as a smaller proportion of individuals were over the age of 65.

We have shown that, under the assumption that vulnerability to clinical COVID-19 infection is a direct result of a population’s health prevalence, total COVID-19 infections, clinical cases, and age-standardised deaths consistently increased with relative deprivation, therefore exposing those living in the most deprived areas to a greater risk of mortality, as well as more non-fatal consequences such as hospitalisation and long COVID. The peak clinical size of the modelled COVID-19 epidemics, which describes the worst-case scenario hospitals would have to withstand, was approximately twice as large in the most deprived decile than the least deprived. We have found that 16% of the deaths observed under the assumptions of this model, or over 65,000 deaths, would be prevented if every IMD decile experienced the same age-specific health as the most affluent 10% of the country. We have also shown that school closures, which disproportionately negatively affect children’s education and well-being in more deprived areas, may also disproportionately benefit the most affluent in society in terms of epidemiological burden [ 31 , 32 ]. Vaccination programmes targeting over 65s disproportionately target and benefit the least deprived areas of the country.

This study used publicly available data and relied on simplified models of infectious disease transmission; there are hence several limitations to the study. The self-reported nature of the Census data means that there may be systematic differences in how health is reported between age groups and levels of deprivation, due to social desirability and the acceptability of self-reporting ill-health varying by demographic, cultural, and socioeconomic factors [ 9 ]. Census data and the IMD may exclude mobile communities and the over 270,000 homeless individuals in England, who are often among the most vulnerable members of society [ 33 , 34 ]. Self-reported health in 2021 may include the effects of the COVID-19 pandemic, and so preemptively confirm the inequalities that this model aims to investigate. However, the IMD-specific health prevalence in 2021 (Additional file 1 : Table S2) is very similar to that found in the 2011 UK Census (75.0% health prevalence in the most deprived decile and 86.9% in the least deprived decile) [ 35 ].

Much of the data used in calculating the IMD relate to 2015–2016 [ 16 ]. Any changes that have occurred since are therefore not accounted for in the IMD rankings, such as the wider roll-out of Universal Credit, which has been shown to have exacerbated existing inequalities and negatively impacted claimants’ well-being [ 36 , 37 ]. Health is itself a component of the IMD, potentially limiting the IMD as an exposure for studies with health outcomes; a brief analysis confirms that there are associations between domains of deprivation other than health (Additional file 1 : Fig. S14). Other studies have also confirmed the relationship between local deprivation and health outcomes when factoring out the health component of IMD [ 38 ].

The assumption of a closed population is unrealistic: apart from during the most stringent lockdowns, which are not represented by the contact patterns used in the above work, individuals will interact and transmit infection between LSOAs as well as within them. A limitation of the contact patterns used is that intrinsic contact patterns are unlikely to be consistent across all IMD deciles and urban and rural geographies. Contact patterns also drastically change in an epidemic, to an extent which depends on SES. The more affluent can more readily reduce their mobility and exposures, while many in the most deprived deciles have less control over their mobility and exposure patterns and are more likely to be in public-facing employment [ 39 ]. The ability to self-isolate may also depend on SES, for instance, through the conditions of sick pay. The assumptions of constant contact patterns were necessary due to a lack of readily available data on IMD- and age-specific contact patterns, both under NPIs and in daily life, and as a consequence this study is likely to have underestimated the socioeconomic inequalities in epidemic burden. SES-specific contact patterns should be incorporated into epidemic models to include the different contacts that for example arise from different occupational prevalences, ability to reduce mobility, household size, and classroom size. To this end, further data should be collected and made accessible for future research.

By restricting clinical fractions between 0.21 and 0.69, clinical fractions converged at the upper bound in deprived deciles over age 60 while health prevalences were still diverging, meaning that the assigned clinical fractions may underestimate the potential difference in vulnerability, and therefore epidemiological burden, between these IMD deciles. The parameters used for the model, taken from [ 23 ] and [ 27 ], contain some uncertainty which is included in the original papers but not considered in this study. We assumed that all vaccinations given in our model were given before the epidemic, instead of during, as this made mortality rates easily comparable between the scenarios. While this is unrealistic, this study does not attempt to recreate the exact COVID-19 pandemic but instead provides insight into the interaction between IMD and vaccine impact. This study has not considered vaccination of clinical risk groups, which would likely be larger in areas with lower health prevalence, or taken into account confirmed deprivation-related disparities in vaccine uptake, which are likely to exacerbate existing inequalities [ 40 , 41 , 42 ]. Further research into the impact of vaccination on these socioeconomic inequalities would improve the understanding of the interaction between comorbidities, vaccination uptake, age structure, and COVID-19 burden.

The presence of drastically worse underlying health conditions in more deprived areas of England has caused, and will continue to cause, dramatic inequalities in the burden of infectious disease. This study has quantified the potential inequalities in epidemic burden under the assumption that vulnerability to severe infection is a direct result of existing comorbidities. The most effective way to reduce the inequalities in epidemic burden caused by socioeconomic health disparities is to improve socioeconomic equity in health in England. The recommendations made by Health Equity in England: The Marmot Review 10 Years On [ 10 ], including maximising empowerment for all, improving standards of living, creating fair employment, and developing healthy communities, would reduce avoidable inequalities in health and by extension avoidable inequalities in epidemic burden.

Availability of data and materials

All analysis code and data are available at https://github.com/1035825/imd-covid .

Abbreviations

Confidence interval

Infection fatality ratio

Index of Multiple Deprivation

Lower layer super output area

Non-pharmaceutical intervention

Office for National Statistics

Susceptible-exposed-infectious-recovered-dead

Socioeconomic status

Mena GE, Martinez PP, Mahmud AS, Marquet PA, Buckee CO, Santillana M. Socioeconomic status determines COVID-19 incidence and related mortality in Santiago Chile. Science. 2021;372:eabg5298.

Article   CAS   PubMed   PubMed Central   Google Scholar  

Bilal U, Tabb LP, Barber S, Diez Roux AV. Spatial inequities in COVID-19 testing, positivity, confirmed cases, and mortality in 3 U.S. cities: an ecological study. Ann Intern Med. 2021;174:936–44.

Article   PubMed   Google Scholar  

Office for National Statistics. Deaths involving COVID-19 by local area and deprivation. https://www.ons.gov.uk/peoplepopulationandcommunity/birthsdeathsandmarriages/deaths/datasets/deathsinvolvingcovid19bylocalareaanddeprivation . Accessed 4 Oct 2023.

UK Health Security Agency. COVID-19 confirmed deaths in England (to 31 January 2021): report. GOV.UK. https://www.gov.uk/government/publications/covid-19-reported-sars-cov-2-deaths-in-england/covid-19-confirmed-deaths-in-england-report . Accessed 4 Oct 2023.

Public Health England. Disparities in the risk and outcomes of COVID-19. 2020.

Bhaskaran K, Bacon S, Evans SJ, Bates CJ, Rentsch CT, MacKenna B, et al. Factors associated with deaths due to COVID-19 versus other causes: population-based cohort analysis of UK primary care data and linked national death registrations within the OpenSAFELY platform. Lancet Reg Health - Eur. 2021;6:100109.

Article   PubMed   PubMed Central   Google Scholar  

Mason KE, Maudsley G, McHale P, Pennington A, Day J, Barr B. Age-adjusted associations between comorbidity and outcomes of COVID-19: a review of the evidence from the early stages of the pandemic. Front Public Health. 2021;9:584182.

Boglione L, Dodaro V. Impact of socioeconomic status on the clinical outcomes in hospitalised patients with SARS-CoV-2 infection: a retrospective analysis. J Public Health. 2023;31:1691–7.

Article   Google Scholar  

Office for National Statistics. Health state life expectancies by national deprivation deciles, England: 2017 to 2019. https://www.ons.gov.uk/peoplepopulationandcommunity/healthandsocialcare/healthinequalities/bulletins/healthstatelifeexpectanciesbyindexofmultipledeprivationimd/2017to2019 . Accessed 4 Oct 2023.

Marmot M. Health equity in England: the Marmot review 10 years on. BMJ. 2020;368:m693.

Rutter PD, Mytton OT, Mak M, Donaldson LJ. Socio-economic disparities in mortality due to pandemic influenza in England. Int J Public Health. 2012;57:745–50.

Ye X, Wang Y, Zou Y, Tu J, Tang W, Yu R, et al. Associations of socioeconomic status with infectious diseases mediated by lifestyle, environmental pollution and chronic comorbidities: a comprehensive evaluation based on UK Biobank. Infect Dis Poverty. 2023;12:5.

Mamelund S-E, Shelley-Egan C, Rogeberg O. The association between socioeconomic status and pandemic influenza: systematic review and meta-analysis. PLOS One. 2021;16:e0244346.

Morrissey K, Spooner F, Salter J, Shaddick G. Area level deprivation and monthly COVID-19 cases: the impact of government policy in England. Soc Sci Med. 2021;289:114413.

UK Health Security Agency. COVID-19 testing in a deprived local authority: Hackney, London. GOV.UK. https://www.gov.uk/government/case-studies/covid-19-testing-in-a-deprived-local-authority-hackney-london . Accessed 4 Oct 2023.

Ministry of Housing, Communities & Local Government. English indices of deprivation 2019. GOV.UK. https://www.gov.uk/government/statistics/english-indices-of-deprivation-2019 . Accessed 4 Oct 2023.

Office for National Statistics. Lower layer super output area population estimates (supporting information). https://www.ons.gov.uk/peoplepopulationandcommunity/populationandmigration/populationestimates/datasets/lowersuperoutputareamidyearpopulationestimates . Accessed 4 Oct 2023.

Office for National Statistics. 2011 rural/urban classification. https://www.ons.gov.uk/methodology/geography/geographicalproducts/ruralurbanclassifications/2011ruralurbanclassification . Accessed 4 Oct 2023.

Department for Environment, Food & Rural Affairs. 2011 rural urban classification lookup tables for all geographies. GOV.UK. https://www.gov.uk/government/statistics/2011-rural-urban-classification-lookup-tables-for-all-geographies . Accessed 4 Oct 2023.

Mossong J, Hens N, Jit M, Beutels P, Auranen K, Mikolajczyk R, et al. Social contacts and mixing patterns relevant to the spread of infectious diseases. PLoS Med. 2008;5:e74.

Funk S, Willem L, Gruson H. socialmixr: social mixing matrices for infectious disease modelling. https://github.com/epiforecasts/socialmixr , https://epiforecasts.io/socialmixr/ . Accessed 4 Oct 2023.

Arregui S, Aleta A, Sanz J, Moreno Y. Projecting social contact matrices to different demographic structures. PLOS Comput Biol. 2018;14:e1006638.

Davies NG, Klepac P, Liu Y, Prem K, Jit M, CMMID COVID-19 working group, et al. Age-dependent effects in the transmission and control of COVID-19 epidemics. Nat Med. 2020;26:1205-11.

Office for National Statistics. General health by age, sex and deprivation, England and Wales. 2023. https://www.ons.gov.uk/peoplepopulationandcommunity/healthandsocialcare/healthandwellbeing/datasets/generalhealthbyagesexanddeprivationenglandandwales . Accessed 4 Oct 2023.

COVID-19 Forecasting Team. Variation in the COVID-19 infection–fatality ratio by age, time, and geography during the pre-vaccine era: a systematic analysis. Lancet. 2022;399:1469–88.

Influenza (Seasonal). 2023. https://www.who.int/news-room/fact-sheets/detail/influenza-(seasonal) . Accessed 4 Oct 2023.

Verity R, Okell LC, Dorigatti I, Winskill P, Whittaker C, Imai N, et al. Estimates of the severity of coronavirus disease 2019: a model-based analysis. Lancet Infect Dis. 2020;20:669–77.

Jackson C, Mangtani P, Vynnycky E, Fielding K, Kitching A, Mohamed H, et al. School closures and student contact patterns. Emerg Infect Dis. 2011;17:245–7.

Yang Z-R, Jiang Y-W, Li F-X, Liu D, Lin T-F, Zhao Z-Y, et al. Efficacy of SARS-CoV-2 vaccines and the dose–response relationship with three major antibodies: a systematic review and meta-analysis of randomised controlled trials. Lancet Microbe. 2023;4:e236–46.

Wang G, Xiang Z, Wang W, Chen Z. Seasonal coronaviruses and SARS-CoV-2: effects of preexisting immunity during the COVID-19 pandemic. J Zhejiang Univ Sci B. 2022;23:451–60.

Lewis SJ, Munro APS, Smith GD, Pollock AM. Closing schools is not evidence based and harms children. BMJ. 2021;372:n521.

Viner R, Russell S, Saulle R, Croker H, Stansfield C, Packer J, et al. School closures during social lockdown and mental health, health behaviors, and well-being among children and adolescents during the first COVID-19 wave: a systematic review. JAMA Pediatr. 2022;176:400.

Moscrop A, Ziebland S, Bloch G, Iraola JR. If social determinants of health are so important, shouldn’t we ask patients about them? BMJ. 2020;371:m4150.

Shelter. At least 271,000 people are homeless in England today. Shelter England. 2023. https://england.shelter.org.uk/media/press_release/at_least_271000_people_are_homeless_in_england_today . Accessed 4 Oct 2023.

Office for National Statistics. General Health in England and Wales: 2011 and comparison with 2001. 2013. https://www.ons.gov.uk/peoplepopulationandcommunity/healthandsocialcare/healthandwellbeing/articles/generalhealthinenglandandwales/2013-01-30#general-health-and-area-deprivation-for-small-area-groupings . Accessed 4 Oct 2023.

Keen R, Kennedy S. Universal Credit changes from April 2016. 2016.

Cheetham M, Moffatt S, Addison M, Wiseman A. Impact of Universal Credit in North East England: a qualitative study of claimants and support staff. BMJ Open. 2019;9:e029611.

Adams J, White M. Removing the health domain from the Index of Multiple Deprivation 2004—effect on measured inequalities in census measure of health. J Public Health. 2006;28:379–83.

Article   CAS   Google Scholar  

Lee WD, Qian M, Schwanen T. The association between socioeconomic status and mobility reductions in the early stage of England’s COVID-19 epidemic. Health Place. 2021;69:102563.

Curtis HJ, Inglesby P, Morton CE, MacKenna B, Green A, Hulme W, et al. Trends and clinical characteristics of COVID-19 vaccine recipients: a federated analysis of 57.9 million patients’ primary care records in situ using OpenSAFELY. Br J Gen Pract. 2022;72:e51–62.

Dolby T, Finning K, Baker A, Fowler-Dowd L, Khunti K, Razieh C, et al. Monitoring sociodemographic inequality in COVID-19 vaccination uptake in England: a national linked data study. J Epidemiol Community Health. 2022;76:646–52.

Coronavirus and vaccination rates in adults by socio-demographic characteristic and occupation, England - Office for National Statistics. https://www.ons.gov.uk/peoplepopulationandcommunity/healthandsocialcare/healthinequalities/bulletins/coronavirusandvaccinationratesinadultsbysociodemographiccharacteristicandoccupationengland/december2020tomarch2023 . Accessed 31 Jan 2024.

Download references

Acknowledgements

Not applicable.

RME and EvL were supported by the National Institute for Health and Care Research (NIHR) Health Protection Research Unit (HPRU) in Modelling and Health Economics (NIHR200908), which is a partnership between the UK Health Security Agency (UKHSA), Imperial College London, and the London School of Hygiene & Tropical Medicine. The views expressed are those of the authors and not necessarily those of the UK Department of Health and Social Care (DHSC), NIHR, or UKHSA.

Author information

Authors and affiliations.

Centre for Mathematical Modelling of Infectious Diseases, London School of Hygiene and Tropical Medicine, London, WC14 7HT, UK

Lucy Goodfellow, Edwin van Leeuwen & Rosalind M. Eggo

Modelling and Economics Unit and NIHR Health Protection Research Unit in Modelling and Health Economics, UK Health Security Agency, London, NW9 5EQ, UK

Edwin van Leeuwen

You can also search for this author in PubMed   Google Scholar

Contributions

LG and RME conceived the study. LG developed the mathematical model and conducted the analyses. RME and EvL consulted on the analyses. LG and RME wrote the manuscript. All authors read and approved the final manuscript.

Corresponding author

Correspondence to Lucy Goodfellow .

Ethics declarations

Ethics approval and consent to participate, consent for publication, competing interests.

The authors declare that they have no competing interests.

Additional information

Publisher’s note.

Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.

Supplementary Information

Additional file 1: fig. s1..

Median ages. Table S1. Proportion of population residing in urban LSOAs. Fig. S2. Projected contact matrices and age-specific total daily contacts. Fig. S3. Age-specific health statuses. Table S2. Overall health prevalence in each IMD decile. Table S3. Age-specific health prevalence and clinical fraction estimates for the general population of England. Table S4. Age-specific infection fatality ratios, clinical fractions, and corresponding clinical mortality probabilities. Fig. S4. Infections and cumulative clinical cases over the epidemic. Fig. S5. Measures of size of a COVID-19 epidemic in each IMD decile, assuming a constant age structure. Fig. S6. Deaths occurring in each age group which are prevented in the case of underlying health equity. Fig. S8. Mortality under varying school closure implementation thresholds. Fig. S9. The reduction in R 0 after school closures, with health prevalence or age structure held constant. Fig. S10. Sensitivity analysis of the dependence of clinical fraction on underlying health varies. Fig. S11. Sensitivity analysis of relative subclinical infectiousness. Fig. S12. Sensitivity analysis of the length of the epidemiological periods. Fig. S13. Infection-related effects of vaccinating over 65 year olds. Fig. S14. Relationship between the ranks of LSOAs in each IMD domain and health prevalence.

Rights and permissions

Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if changes were made. The images or other third party material in this article are included in the article's Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article's Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http://creativecommons.org/licenses/by/4.0/ . The Creative Commons Public Domain Dedication waiver ( http://creativecommons.org/publicdomain/zero/1.0/ ) applies to the data made available in this article, unless otherwise stated in a credit line to the data.

Reprints and permissions

About this article

Cite this article.

Goodfellow, L., van Leeuwen, E. & Eggo, R.M. COVID-19 inequalities in England: a mathematical modelling study of transmission risk and clinical vulnerability by socioeconomic status. BMC Med 22 , 162 (2024). https://doi.org/10.1186/s12916-024-03387-y

Download citation

Received : 11 December 2023

Accepted : 10 April 2024

Published : 15 April 2024

DOI : https://doi.org/10.1186/s12916-024-03387-y

Share this article

Anyone you share the following link with will be able to read this content:

Sorry, a shareable link is not currently available for this article.

Provided by the Springer Nature SharedIt content-sharing initiative

• Mathematical model
• Socioeconomic inequality
• Comorbidities

BMC Medicine

ISSN: 1741-7015

• Submission enquiries: [email protected]
• General enquiries: [email protected]