The Airbnb effect
Stories by Carolyn Said | Design by Michael Grant | Development and graphics by Aaron Williams, Maegan Clawges and John Blanchard | Photography by Paul Chinn, Liz Hafalia, Jason Henry, Leah Millis, Amy Osborne, Lea Suzuki and Sam Wolson | Edited by Ben Muessig | Copyedited by Carolyn Grannan
Both Airbnb and its San Francisco critics are correct. Most of the booming vacation-rental company’s local listings are only occasionally rented to travelers, as Airbnb says. However, at least 350 entire properties listed on Airbnb — and hundreds more listed on competing sites HomeAway/VRBO and FlipKey — appear to be full-time vacation rentals, bolstering claims by activists that the services remove scarce housing from the city’s limited inventory.
Those are the among the findings from a data dive into the three companies’ websites.
It’s not the thousands of illegal hotels that critics allege have sprung up thanks to Airbnb and its rivals. But in a city wracked by a housing crisis, where a typical year sees just 2,000 new units added, a few hundred units off the market makes a significant dent.
Using houses as hotels ranks among San Francisco’s most-contentious issues, debated fiercely by lawmakers, residents, advocates and corporations, and now likely headed for a ballot-box showdown.
Opponents say short-term rentals are so lucrative that greedy landlords and tenants illegally divert precious housing stock to the practice. Proponents say services like Airbnb help regular people afford to stay in San Francisco, while forging international friendships.
For decades, a seldom-enforced law banned vacation rentals in San Francisco. A new law enacted on Feb. 1 legalized such rentals, with conditions: hosts must be full-time residents, whole-property rentals are capped at 90 days a year, and all hosts must register with the city. But so far only a fraction of hosts — about 700 — have registered, while the city itself admits it struggles to catch scofflaws. Mayor Ed Lee this month created a new office to streamline registration and pursue violations.
At the core of this dispute is a simple question that is devilishly difficult to answer: How do the impromptu inns affect the city’s housing market?
To investigate, The Chronicle commissioned data-extraction companies Connotate Inc. and Import.io to harvest San Francisco information from the Airbnb, HomeAway and FlipKey websites on May 19. Connotate also extracted Airbnb listing data for The Chronicle on the same date in 2014, allowing for insights about year-over-year changes. Here’s a summary of the findings:
Total listings and cost
Amid all the controversy, the number of Airbnb listings in San Francisco grew 13.8 percent in the year, hitting 5,459, despite a significant portion of properties dropping out. The $202 average nightly cost among all property types — entire homes, private rooms and shared rooms — was up $19 (10.9 percent) from a year ago.
Entire homes accounted for 3,264 of the 2015 listings, or 59.8 percent; private rooms numbered 1,969 or 36 percent and shared rooms tallied 226 or 4.1 percent. Prices for entire homes rose 13.3 percent to $255; private rooms edged up 6 percent to $123 and shared rooms ballooned 55 percent to $124. That increase was propelled by some new hacker hostels — shared houses that pack young workers into bunk beds.
Length and frequency of rentals
Almost two-thirds — 64 percent — of all listings had 10 or fewer reviews over the past year, supporting the idea that they were rented only infrequently.
However, the remaining listings racked up a significant chunk of guest reviews — a telling, albeit conservative indicator of usage as not all guests leave reviews.
Properties listed on Airbnb both last year and this year averaged 22 reviews each. Among entire homes, 352 showed 26 or more reviews in the 52-week period. That means new guests checked in at least every other week and likely more often.
An additional 349 new or returning listings for entire homes also had 26 or more reviews, attained during an indeterminate time period.
Where are they?
Vacation rentals are spread throughout San Francisco, in contrast to the city’s 34,000 hotel rooms, which are heavily clustered near Moscone Center, Union Square and Fisherman’s Wharf.
The Mission District, with 789 listings (up from 681 last year), remains the city’s most popular Airbnb location, followed by SoMa (388 listings) and the Western Addition/NoPa (369). Bernal Heights added 58 properties to become fourth-most-popular, with 264 spaces for rent. The Bayview added eight listings (for a total of 58), and racked up 757 new guest comments in a year, pointing to rise in usage in one of the city’s lower-cost neighborhoods.
Who hosts?
While most Airbnb hosts control a single property, 205 hosts have three or more listings. These super hosts account for 4.8 percent of all hosts, but control 993 properties — 18.2 percent of Airbnb’s local listings. Some are property managers, some are hacker hostels, some are legal hotels and some are people who may have siphoned off housing stock to make more money hosting visitors.
Wrede Petersmeyer, who oversees Airbnb analytics, flew out from New York, to discuss the company’s data and defend its model. Most entire units in San Francisco are rented out fewer than 30 days a year, he said, while only 10 percent of hosts rent out a space where they do not live. Entire units were booked an average of 65 nights over the 12 months ended July 5, with a median stay of 29 nights, Airbnb said.
“Data scrapes are unfairly used to draw specific — often negative — conclusions about a small subset of our host community,” he said, “when the reality is the vast majority are middle-class families sharing the home in which they live.”
Since 60 percent of Airbnb’s local listings are entire homes, does that undercut the company’s assertion that guests and hosts are “sharing” spaces and forging friendships?
“Local, authentic hospitality doesn’t always mean the host has to be there to give you the keys,” Petersmeyer said.
Living with Airbnb
Stories by Carolyn Said | Design by Michael Grant | Development and graphics by Aaron Williams, Maegan Clawges and John Blanchard | Photography by Paul Chinn, Liz Hafalia, Jason Henry, Leah Millis, Amy Osborne, Lea Suzuki and Sam Wolson | Edited by Ben Muessig | Copyedited by Carolyn Grannan
Geoff Benjamin and his husband, Craig Persiko, often list rooms in their four-bedroom Castro condo on Airbnb to raise spare cash. Another host identified on the site as Yumi has allegedly turned two entire flats on Nob Hill into Airbnb hotels, hosting a frequent stream of temporary renters, while also listing another room for rent in the Richmond District. The tales of these hosts illustrate both sides of the vacation-rental debate.
Persiko chairs the computer science department at City College of San Francisco, while Benjamin recently launched the SoMa cabaret theatre and nightclub Oasis — a labor of love financed by Airbnb guests.
Renting to travelers “in many ways became like a small-business loan that we didn’t have to repay,” Benjamin said. “In essence, it paid my salary for the two years it took to open Oasis.”
The couple rent two to four rooms to travelers at different times. They paid for sleep-away camp for their children, Tobias, 10, and Serafina, 12, by renting out their bedrooms while they were away, for instance.
They love meeting people, and the feeling seems mutual. Airbnb guests read to their kids and eat meals with the family. Benjamin said the rooms wouldn’t work as permanent housing, because most people prefer not to stay long with a family with children.
“I’ve always wanted my children to identify as citizens of the world,” Benjamin said. “A really excited energy comes from people when they travel. We get to see the city through their eyes.”
Over on Nob Hill, Katieanne Moran and her flatmates say they are surrounded by Airbnb renters. Long-term tenants in the middle flat of a Victorian triplex, they discovered that the top floor and basement flats had been converted to full-time Airbnb use, with six individual rooms going for $100 to $130 a night.
“Rooms were being rented below us, above us and in the building all around us,” she said. “It was very disconcerting; we felt trapped.”
Their landlady, Yumi, does not live in the building, the tenants said. When they voiced concerns about noise and security, the landlady said “she could do whatever she wanted and if it caused problems, we would be evicted.”
Now they’re suing Yumi, whose legal name is Mingjing Li, for violating city ordinances, negligence and creating a private nuisance. Li and her attorney declined to comment.
“This is one of the more egregious cases, but we often see landlords try to force tenants out ... (to use) Airbnb and other hosting platforms,” said Mark Hooshmand, the tenants’ lawyer. “Money motivates people to not follow regulations.”
Hosts or hoteliers?
Stories by Carolyn Said | Design by Michael Grant | Development and graphics by Aaron Williams, Maegan Clawges and John Blanchard | Photography by Paul Chinn, Liz Hafalia, Jason Henry, Leah Millis, Amy Osborne, Lea Suzuki and Sam Wolson | Edited by Ben Muessig | Copyedited by Carolyn Grannan
How many San Franciscans use Airbnb to turn their places into year-round hotels? Airbnb knows the answer: Its transaction system captures exactly how many nights each property is rented. The company could remove listings after they exceed the legal limit of 90 days — but it doesn’t want to do so. It says those bad players would simply migrate to other platforms such as Craigslist that don’t police listings.
Nor does it want to open the kimono on exactly how many days each place is rented out. And even though it now collects San Francisco’s 14 percent hotel tax — remitting more than $1 million a month — it resists requests to tell the tax collector which homes are rented and how often.
For someone looking in, guest reviews provide the best insight into how often a unit is rented. The more reviews, the more frequently a property hosts travelers. By counting how many reviews each property generated in the 365 days between our two data dives, we compiled a conservative estimate of places functioning as full-time vacation rentals.
Among whole properties — houses, condos and apartments — listed on the site on May 19 in 2014 and 2015, 352 racked up 26 or more reviews over the year. That includes 104 with a stunning 51 or more reviews.
For entire units that weren’t on the site last year, 349 had a total of 26 or more reviews. Some were new listings and some were returning properties, so we can’t say conclusively that all reviews were accumulated in a single year.
Reviews underestimate usage. Airbnb says only two-thirds of guests leave reviews — meaning a property with 30 reviews in a year may have hosted 45 unique visitors, nearly one a week. San Francisco guests stay an average of 5.5 nights, Airbnb says, so 30 reviews translates into an average of 165 nights of occupancy — and likely more, accounting for guests who don’t leave reviews.
Overall, the number of reviews for all San Francisco listings surged over the 12 months, rising by 44,842 to 105,192. Yet the majority of listings had only a handful of reviews, buttressing Airbnb’s assertion of infrequent usage.
Looking at all categories of properties and including both repeats and the new or returned listings, 3,242 or 59.3 percent, have 10 or fewer reviews over their entire time on Airbnb. Another 958 had 11 to 25 reviews, while 639 had 26 to 50 reviews, implying usage between casual and frequent. But 620 homes and rooms, or 11.4 percent of listings, amassed 51 or more reviews, implying very heavy usage.
The power of super hosts
Most Airbnb hosts control just a single listing — but 206 hosts control three or more listings. These “superhosts” account for 4.8 percent of all hosts, but run 18.2 percent of Airbnb’s San Francisco listings.
Annette Fajardo says it’s no mystery why landlords prefer to rent to short-term visitors.
“The dirty little secret is that owners do this to avoid rent control,” she said. At the same time, “rent-controlled tenants do it to make money on the units that the owners should be making money on.”
She’s in a position to know. Her SF Holiday Rentals, in business since 2002, is the city’s oldest short-term rental management site, she said. She handles 20 to 30 San Francisco listings and another seven to 10 in Pacifica, listing them on multiple websites, including Airbnb, HomeAway and its subsidiary VRBO.
Her listings focus on entire homes, many available year-round. About a third are in-law units; some are second homes; some are small buildings owned by a landlord, including her own Bernal Heights duplex (she lives elsewhere and rents both units to travelers).
She does this all legally. Her property owners now request a 30-day minimum stay, she said, exempting them from the new short-term rental law. “They didn’t want the headaches of working with the government or (officials) snooping into their business practices,” she said.
The change made her business easier, since she doesn’t have to handle turnover every few days.
Fajardo is a “super host” on both Airbnb and HomeAway, among the top 10 on both sites based on number of listings.
Most Airbnb hosts (3,599 people, or 85 percent of hosts) control just a single listing, accounting for 66 percent of all properties. Another 434 list two properties. Many have a legitimate reason: Either they rent out two rooms in their home, or they rent out a room and also offer their entire home when they travel.
However, 205 hosts control three or more listings. These super hosts account for 4.8 percent of all hosts, but control 993 properties — 18.2 percent of Airbnb’s San Francisco listings.
Many hosts with multiple listings are either property managers like Fajardo or people running hacker hostels, packing young techies into bunk beds for cheap crash space and camaraderie.
Properties handled by managers can include ones skirting regulations, as well as those like Fajardo’s that adapted to become legal. Pillow (22 listings) and Guesthop (15 listings) are other leading concierge services.
Hacker hostels are problematic, as they may flout a variety of city rules on overcrowding, according to Scott Sanchez, San Francisco zoning administrator. If they require a 30-day minimum stay, however, they may be allowed in most districts as group houses, he said.
Airbnb’s top two super hosts both run hacker hostels. Rob, No. 1 with 30 listings, rents rooms in Tech House SF but does not live there, residents said. Gatz, No. 2, runs Looky Home with 25 listings.
Another group of people who control multiple listings seems to have turned entire homes or apartments into year-round pads for travelers. Often they have just a handful of listings in different parts of the city.
Eighty Airbnb hosts have four to six listings, controlling 374 rentals among them. Generally that’s too few for them to be property management services — and too many for them to be legitimately living at those locations. It’s a rare San Franciscan who could rent six bedrooms at his or her home.
As part of a renewed effort on enforcing vacation-rental laws, San Francisco sent violation letters this month to hosts in this category, targeting 15 hosts with 72 short-term rentals around the city. More than half are blocks of multiple units in brand-new high-rises in SoMa and Mid-Market, while others are two- to four-units buildings entirely devoted to short-term rentals, the City Planning Department said. The units were listed on various services, while some had their own rental websites.
“We want to crack down on people taking housing off the market and hurting San Francisco,” said Tony Winnicker, a senior adviser to Mayor Ed Lee. “But we also want law-abiding people to be able to do occasional (short-term rentals) because it helps many people afford to live in our city.”
Airbnb hosts come and go
Stories by Carolyn Said | Design by Michael Grant | Development and graphics by Aaron Williams, Maegan Clawges and John Blanchard | Photography by Paul Chinn, Liz Hafalia, Jason Henry, Leah Millis, Amy Osborne, Lea Suzuki and Sam Wolson | Edited by Ben Muessig | Copyedited by Carolyn Grannan
Bootstrapping a business was eating up her cash, so the San Francisco woman listed her Marina apartment on Airbnb in 2013. She rented it out about 70 nights a year for $260 a night. “I would coordinate trips for work and family, camp out with friends or do ‘Airbnb arbitrage,’ where I’d stay at a less-expensive Airbnb place,” she said. The cute one-bedroom with views of the Golden Gate Bridge was always snapped up.
“It was a self-perpetuating, ka-ching ka-ching money machine,” she said. “It was such a relief to know I could always rent it on Airbnb to subsidize my rent.” She made almost a year’s rent over two years.
But when San Francisco implemented its vacation-rental rules in February, registering her unit seemed like a hassle. She feared that the city might compel Airbnb to identify hosts and prosecute those who failed to register.
But her biggest worry — and the reason she declined to be named for this story — is that short-term renting violates her lease. She took down her listing.
“I felt like I made my hay when the sun was shining, but now it’s over,” she said.
In the year since The Chronicle’s last data dive, 2,773 listings, or 57.8 percent, dropped out of the site, although they were replaced by a larger number: 3,492 new listings for a net growth of 719 listings, or 15 percent.
Airbnb says listings often cycle in and out. For instance, people list their house before they go on vacation and take it down when they return. Indeed, many of the new listings appear to have had longer tenure than a year.
Tom Slee, an independent researcher and author of the forthcoming book “What’s Yours Is Mine” about on-demand companies, examined Airbnb churn worldwide, finding an exit rate of between 50 and 60 percent of listings in cities such as New York, Los Angeles, Amsterdam, Paris and Barcelona over about a year. At the same time, the entry rate of new listings always resulted in net growth, he found.
The listings that left generally had only a handful of reviews, implying infrequent usage, while those that stayed averaged reviews in the double digits, he said.
“My feeling was that it was largely people who tried it, found it wasn’t for them and left,” he said.
Looking at the Bay Area, including San Francisco, the East Bay and the Peninsula, he found a 51 percent exit rate, lower than that for the city alone. That props up the idea that San Francisco is losing more listings due to users’ fears about new regulations and landlord prohibitions on subleasing.
Flouting the law
Stories by Carolyn Said | Design by Michael Grant | Development and graphics by Aaron Williams, Maegan Clawges and John Blanchard | Photography by Paul Chinn, Liz Hafalia, Jason Henry, Leah Millis, Amy Osborne, Lea Suzuki and Sam Wolson | Edited by Ben Muessig | Copyedited by Carolyn Grannan
While Airbnb cultivates a folksy image of its hosts as middle-class locals making ends meet, HomeAway proclaims that its hosts are affluent out-of-towners renting out their pieds-a-terre. “The vast majority of our listings fall into the category that the city is trying to stop: this ‘evil’ second-home owner, who we don’t believe is evil,” said Carl Shepherd, co-founder and chief development officer of the Texas company.
Prices for its 1,001 local listings, all private homes, reflect that more upscale image, averaging $302 in San Francisco compared with Airbnb’s entire-home average of $255.
HomeAway, which owns VRBO and other brands, is up front about the fact that its business model flouts San Francisco’s new vacation-rental law. In fact, it sued the city, saying the law was tailored for Airbnb. Most HomeAway properties in San Francisco are second homes whose owners do not reside here and thus are barred by the new law from renting to tourists for periods of less than 30 days, HomeAway said in its suit claiming discrimination. A judge tossed the case on procedural grounds.
Its hosts increasingly request 30-day minimums in San Francisco, Shepherd said, although only 156 listings specified monthly rates.
People who control multiple properties — often called super hosts — are more dominant on HomeAway than on Airbnb. Almost half of the HomeAway properties — 461, or 46 percent — are in the hands of someone with multiple listings. They are controlled by 87 hosts, or 14.8 percent of the total.
Hosts with dozens of listings are clearly vacation-management companies like Pillow and RedAwning Vacation Rentals. That doesn’t prove that their listings are legal or illegal, just that they belong to other people.
HomeAway functions like a classified listing service, rather than a middleman like Airbnb. That also means that far fewer reviews appear on its website.
HomeAway said there is about a 10 to 15 percent overlap between its listings and those on Airbnb worldwide; it didn’t have a breakdown for San Francisco. It has even more overlap with FlipKey because many large property managers list on both sites, but didn’t have a specific number.
FlipKey, owned by TripAdvisor Inc., has 359 local properties with an average price of $269. Super hosts also dominate its site, with 220 properties, or 61.3 percent, in the hands of 35 hosts, or 25.7 percent of hosts.
FlipKey declined to comment.
HomeAway has a big chunk of homes — 148 — controlled by people with just two or three listings, often in different parts of the city.
Shepherd is frank about the explanation. Most, he presumes, are tenants gaming the system.
“When someone has several different listings in a city like San Francisco, those are probably all rented apartments being arbitraged by the renter,” he said. “Arbitrage means I pay $4,000 a month and rent it (short term) for $300 a night, with or without the landlord’s permission, but most likely without.”
San Francisco actually abetted such people by regulating vacation rentals, Shepherd said.
“The new law tells you exactly how to skirt it,” he said. “All you have to do is say you live there 270 days a year. You do have to lie and you do have to obfuscate. But the city is at a distinct disadvantage because they can’t catch this.”
