A new article in the New York Times blares the headline that many have been waiting for: “Airbnb Releases Trove of New York City Home-Sharing Data.” Why is Airbnb data so important? At this point it’s well-established that Airbnb has been invaded by professional real estate operatives, who are exploiting Airbnb as a loophole that allows them to evade local housing laws forbidding rentals for fewer than 30 days.
Those laws have been on the books for many years to prevent the housing stock from being rented out to tourists instead of local residents. Airbnb is aware of this professional invasion, yet has done nothing to rein it in. After all, with one flick of the computer mouse, Airbnb could “evict the evictors” – kick off its platform the hosts that are renting out multiple properties (some Airbnb “hosts” control dozens of properties).
Cities trying to enforce their housing and rental laws have been nearly powerless to do so. With thousands of hosts engaged in illegal and quasi-legal activity, it’s almost impossible for existing levels of city staffing to locate these hosts, assess how many nights they are renting their property, how much they are charging per night, how much hotel tax they should be paying, and other details. Airbnb is also aware of the enforcement nightmare that it has unleashed. Even in the handful of its 34,000 cities where Airbnb has agreed to pay taxes, like in San Francisco, without the host data city regulators can’t verify that Airbnb is paying the correct amount it owes. They have to take the company’s word for it.
So one of the primary demands of regulators, housing activists and neighborhood advocates trying to cope with the Airbnb disruption has been for the company to turn over data to local agencies, so that it can be used for enforcement purposes. Despite its claim to being a “sharing economy” company, Airbnb has steadfastly refused to turn over the data in city after city. The only city where Airbnb previously turned over some data was in New York, after Attorney General Eric Schneiderman subpoenaed it and went to court. Airbnb typically has hidden behind the excuse that it would be an invasion of its hosts’ privacy to turn over the data. Never mind that hosts turning their properties into Airbnb hotels has been an enormous invasion of neighborhood privacy – that’s the type of privacy that doesn’t go ka-ching at the cash register, and has never interested Airbnb.
So Airbnb finally turning over data, voluntarily, without the legal hammer of a subpoena, shows progress. Or does it? Airbnb has a history of unilaterally releasing selective data that paints a picture of the company and its service in a favorable light. So is this newly released stockpile of data just more whitewashing?
Unfortunately, the data reported in the New York Times has the whiff of “funny numbers,” including in this case a rather dramatic “correction” that suddenly appeared on the New York Times website after the article was published. Let’s parse through the numbers a bit.
Initially, the article reported “some 93 percent of revenue earned by active hosts in New York City who share their entire home comes from people who have only one or two rental listings on the platform.” But that data point wasn’t granular enough to reveal how many of those "whole homes" are with hosts who actually live there, and are just renting the whole home when they happen to be away -- or how many are "absentee owners" who bought a property as an investment to let out on Airbnb.
A huge part of the professional invasion in New York City, San Francisco, Los Angeles and elsewhere has been from investors buying up whole homes and never living there, only renting on Airbnb. Those kinds of absentee owners remove housing stock and create a permanent hotel within a neighborhood that was never zoned for such hotel activity. So how much of that 93 percent figure involves "whole home" rentals by absentee owners? The data as reported by the New York Times doesn't tell us.
But then, a funny thing happened. Later in the day the New York Times issued a correction, and that specific sentence was changed, quite dramatically to this:
“From November 2014 until November 2015, some 75 percent of revenue earned by active hosts in New York City who share their entire home came from people who have only one or two rental listings on the platform. Over 2015 to 2016, Airbnb projects that number will rise to 93 percent.”
So the “93 percent” number is based on an Airbnb projection. The real number is only 75 percent. Data analyst Tom Slee, who has done numerous analyses of Airbnb’s business model by “scraping” data from the website, reacted by saying, “A drop from 93% to 75% is significant. That's almost a four-fold increase in the proportion that comes from three-or-more listers. All of a sudden the Airbnb numbers look much more like those collected by myself and other external investigators, which Airbnb routinely say are inaccurate.”
Slee’s data scrape from this past September found that of the 30,268 listings in New York City, 44 percent of guest visits are to hosts with “multiple properties,” which he defined as two or more properties -- compared to the “new” data from Airbnb which found that 25 percent of listings came from hosts with three or more properties. There’s nothing inconsistent between these two findings. In fact, a year ago an investigation by New York state attorney general Eric Schneiderman found that nearly 40 percent of Airbnb’s $451 million in revenue from New York City – some $168 million – came from hosts who had at least three listings on the site.
One of those operators was a fellow by the name of Robert “Toshi” Chan, who was revealed to be the Airbnb “host” of more than 200 apartments in dozens of different buildings, known collectively as Hotel Toshi. Eventually, Toshi’s illegal operation was shut down, and he agreed to pay a $1 million settlement for not obtaining proper hotel permits or insurance. But even with 25 percent of hosts having three or more listings, how many of them have dozens or even hundreds of properties under their control? The new Airbnb data doesn’t tell us, it’s not granular enough.
These findings are consistent with what independent analyses have found about Airbnb’s business model in other cities. In San Francisco, various studies have found that a third of Airbnb rentals are controlled by people with two or more listings, and 40 percent of revenue comes from Airbnb hosts with multiple listings. The same story has held true in Los Angeles, Portland, Oregon, and elsewhere.
Let’s be clear: A huge chunk of Airbnb’s business does not come from “regular people” renting a spare room. It comes from professional landlords and property managers who rent out multiple properties, oftentimes evicting tenants and converting entire buildings into “tourist hotels.”
So meet the new data, same as the old data. The new data from Airbnb basically confirms what the critics have been saying. Nothing has really changed – despite Airbnb’s crowing to the contrary.
Another fine point: even the corrected version of the Times article reported that Airbnb data shows “the majority of New York City hosts do not have large numbers of properties to rent out.” But that point has never been in dispute. It has always been clear that that most of the host listings on the platform are from the much-mythologized "regular people," not professionals. But it’s also true that most of those regular people hosts don't rent it out very often.
Instead, the more important measurement is the total number of visits to the different types of hosts, rather than listings. The fact is, on Airbnb more and more host visits are with properties controlled by the professionals. Those “Airbnb hotels” are in operation all the time, available 365 days a year, with tourist after tourist staying in those properties inside neighborhoods where people have a reasonable expectation that they are not living next door to a hotel. That’s what zoning laws used to do, before Airbnb ripped open this loophole. It’s these properties controlled by professional real estate operatives that are the ones generating a ton of revenue for Airbnb.
And those are the ones that Airbnb refuses to evict from their platform. Why? Because this “sharing” company is heading toward an IPO and has venture capital funders who want to blow this thing open bigger than Facebook.
I emailed the New York Times reporter, Mike Isaac, asking for the back story behind his article’s correction. Did Airbnb initially report the wrong number? Or did the NYT calculate something incorrectly? The “public release” of this data was actually not so public: It was made available only by making an appointment to visit Airbnb’s New York City office. In the past, Airbnb has been known to give out only data summaries -- which of course contained their own built-in skew. Is that what Airbnb did in this case, or did they actually give to the Times some raw data for analysis? I have not heard back as of this writing.
Airbnb representative Chris Lehane (who is an insider Democratic Party player, having been a spokesperson for both President Bill Clinton and Vice President Al Gore) is now the circus handler who is supposed to smooth all of this over. He is quoted in the article saying, “The vast majority of the [Airbnb] community is doing this in the right way.” And yet that statement is unsupported by the data reported in the New York Times article. In fact, the corrected version of the data seems to corroborate what critics have been saying all along – that Airbnb has been invaded by the professional real estate operatives who are breaking laws and refusing to pay hotel taxes on their way to disrupting neighborhoods, overturning rent-controlled housing and wrecking individual tenants’ lives.
Certainly, if Airbnb would allow independent researchers to have access to its data and analyze the results, that would be a tremendous step forward. But Airbnb, as well as other sharing economy companies like Uber, continue to selectively release data, and then choose…their…words…carefully around the public release, dancing like a split wide receiver trying to stay inside the “out of bounds” line. But that’s really no different than the tobacco companies paying for their own toxicity tests. We know what that’s about, and we aren’t going to fall for it.