The Sharing Economy’s Dirty Laundry
Sharing economy companies like Uber and Airbnb aren't helping local economies — they're just helping themselves.
Silicon Valley “sharing economy” darlings Uber and Airbnb raised mad money in 2015. Dwarfing other “unicorns” (startups whose valuation reaches a billion dollars), Airbnb is now valued at $25 billion — rivaling the largest US hotel chains — while the market puts Uber’s value at $65 billion, similar to major car companies. Last year alone Airbnb raised $1.6 billion to push its total funding over the $2 billion mark, and Uber raised almost $5 billion, for a total of more than $6.5 billion.
The two companies were busy with more than raising cash however. In 2015 they intensified their lobbying, PR, and customer mobilization efforts to create a regulatory landscape that embraces their business needs. One of the most potent arrows in the companies’ lobbying quiver is that legislators and the public can be swayed by the apparent inevitability of a technology-driven future: “Don’t be left behind!” is a clarion call that few can resist.
In just two examples: last July Uber quashed Mayor Bill de Blasio’s plan to cap the number of cars on New York City streets, and in October, Airbnb fought off Proposition F (an initiative to restrict short-term housing rentals) in San Francisco, out-advertising its opponents by a ratio of 100 to 1.
Mobilizing cash and mobilizing votes go hand in hand: investors only profit if the regulatory landscape is changed, not just to be tech friendly, but to support the specific business models that companies like Uber and Airbnb have put in place.
But the sharing economy also exemplifies another corporate maneuvering tactic: both Uber and Airbnb remain private companies, and neither is in a rush to go public. By postponing their initial public offerings (IPOs), the companies give themselves maximum flexibility: they don’t have to please shareholders or report short-term profits. They publish no prospectus, there is no independent audit, and we can’t see their accounts.
This tactic isn’t new, but in today’s financialized economy it creates a perfect storm of bad incentives. Investors are looking for an “exit” (a successful IPO) so they can cash in; fortunes will be made or lost depending on rewriting laws around the world; and at the same time, the companies operate as what Frank Pasquale calls “black boxes” because they don’t have to submit public, auditable business reports.
Staying private is particularly appealing in technology sectors where competition is intense and expectations are high. Take a company like Theranos, a privately held health care company. It gained investment based on its innovative blood-testing technology, but was immediately under the gun to prove that it was a game changer.
When its new technology ran into problems it simply covered them up, resorting to traditional methods of blood testing instead to sustain the image of success. Ashley Madison is another example; it created thousands of fake accounts to make it seem like men were meeting women on their site. And in hindsight, companies like WorldCom and Enron were just early examples of a common practice.
Simply put, there are huge rewards for companies that can fake it until they make it, and bankruptcy for those who air their dirty laundry honestly. And tech start-ups are the biggest fakers of all.
Data Wars
Sharing-economy platforms are built on a combination of algorithms and “big data.” Every ride, every rental, every click is recorded; along with ratings, payments, and other data. Airbnb hosts are measured by the time it takes for them to respond to requests, Uber drivers by the routes they take.
But algorithms and big data are not only a central part of the companies’ operations, they have also become PR weapons in battles for legislative change. We don’t need your old rules any more, companies say: our information can deliver new levels of efficiency, convenience, and safety.
Uber and Airbnb have been sharing techniques, learning from each other about how to use their data as a tool of public relations.
The simplest technique is the selective release of data, chosen to highlight the positive side of their business. In 2014 Uber made a splash claiming that its drivers in New York were making the remarkable sum of $90,000 per year. The story ran in newspapers across the US. Reporter Alison Griswold went on a fruitless search for this mythical “Uber unicorn,” and the Uber salary claim has been picked apart over time. Yet a year later the $90,000 number was still being presented by the company as fact.
It wasn’t a direct lie, but it was certainly highly misleading. Uber selected the city: New York City drivers earn far more than any other city. It selected the drivers: the $90,000 number was an average for only those drivers who put in over forty hours a week (so some were driving much more). And Uber presented a gross income measure; driver expenses including insurance, gas, and car maintenance were completely omitted.
Airbnb has adopted the same technique: when faced with controversy in one of their markets, they have taken to putting out “city reports” like the one that claimed to give “quantitative evidence that New York hosts are good for the community.” The report runs a mere three hundred words, with no methodology to back up the bald (and bold) claims, and is full of authoritative-sounding factoids like “Airbnb . . . supported 950 jobs in the outer boroughs” or “82 percent of Airbnb properties are located outside of mid-Manhattan.”
The numbers are practically meaningless — there is no indication of what “supported” means, and when the New York Attorney General’s office got a look at the company’s data it revealed that 97 percent of citywide revenue came from just two of the five boroughs (Manhattan and Brooklyn).
Yet despite the flimsiness of these companies’ claims, their use of data has been immensely successful. High-powered technology executives delivering quantitative claims in an authoritative and confident tone, in polished marketing language, can go a long way to creating that picture of an irresistible sunny future. By the time the truth surfaces, the damage is done.
Another technique is to commission academics to write a paper or report (un-refereed of course), tempting them with an exclusive peek at the company’s internal data. The companies are careful not to overtly sway the researcher, but the very fact of collaboration and private access to data suggests something less than neutrality.
Airbnb has used this tactic multiple times to refute opponents’ claims about the company’s impact on affordable housing in some of its more contentious cities, like Los Angeles, and recruits big names like former White House National Economic Advisor Gene Sperling. Uber commissioned Princeton economist Alan Krueger to write about the conditions of its drivers in a report co-authored by Uber’s own Head of Policy Research Jonathan Hall. As with other company reports on driver incomes, the Krueger-Hall paper lacked any data about driver expenses, claiming that such data is not available.
On some occasions, these reports are released in full (but without supporting data sets); on other occasions the paper is kept internal and only the press release is made available; for example, a 2015 report by UBC professor Thomas Davidoff about Airbnb’s impact on housing prices was written up in the Wall Street Journal and elsewhere, but was never actually released, and the company did not answer my requests for a copy. It is difficult to argue against a report that you cannot read.
Companies like Uber and Airbnb also boost their public image by promising to share data with cities. At the beginning of 2015 Uber partnered with Boston to share some of the company’s ride data “to solve problems,” and at the end of the year Airbnb announced a Community Compact committing the company to providing reports on their activity in cities where they have a significant presence.
But sharing-economy companies are selective about the data they share. Airbnb may offer to pay taxes on behalf of its hosts in cities where tourism is contentious, but it refuses to hand over details of where its listings are located. This makes it impossible for popular tourist destinations to limit tourist rentals and balance the impact of tourism with other concerns in stressed areas of their cities. In January Uber was fined for failing to provide required data to the California Public Utilities Commission.
The data that is shared is carefully culled. In December, Airbnb made a data set on New York City public, some of which described the state of its business on November 17, 2015. An investigation by Murray Cox and myself showed that Airbnb had chosen the date carefully: the company booted over a thousand listings off the site in the lead-up to this date to ensure it painted a favorable picture. Initially Airbnb denied taking any such action, suggesting that the fluctuations could be due to Halloween or the New York marathon, but in recent days they have acknowledged the maneuver.
Yet when it comes to taking responsibility, Airbnb and Uber often hide behind their algorithms. For example, Uber has successfully created the impression that their surge-pricing policies are “simply Economics 101,” but researchers Alex Rosenblat and Luke Stark have shown that their algorithms do not simply reflect increased demand, but instead are part of Uber’s active management of the system, presenting a “mirage of a marketplace.” Surge pricing is part of Uber’s cat-and-mouse game with drivers who want to maximize their income; drivers are provided with the pricing information, but nothing about the number of riders, the expected duration of the surge, or the number of other cars heading towards the surge zone.
Many have concluded that “chasing the surge” is a mug’s game. Do the cars drivers see on their app represent actual, available Uber cars? Rosenblat and Stark suggest not: that it might be (to quote an Uber staffer) “more of a visual effect letting people know that partners are searching for fares.” How much money does Uber make from their price surges? There is no way to know.
Both companies also hide behind their rating-based reputation algorithms. Uber maintains that drivers are removed from their platform solely on the basis of bad ratings, but there are numerous reports of Uber executives firing drivers for personal and capricious reasons. And while Airbnb insists that its rating system keeps its platform trustworthy, many studies have shown that people go out of their way to give high ratings on such systems, so that they can serve to cover up customer dissatisfaction and create a false impression of quality.
As the pressure grows to deliver a successful IPO at their current stratospheric valuations, both Uber and Airbnb will feel pressure to be even more economical with the truth in their public presentations. To counter the spell that sophisticated technology and elaborately presented data seems to cast over city governments, we need to peer into the black boxes of these companies.
Some people already have, and it’s not pretty. The Information’s Amir Efrati reported that Uber lost almost $1 billion in the first half of 2015 (the story is paywalled but a summary is here), and more importantly, their current business depends not only on pushing all the risk onto individual drivers, but also on avoiding taxes (in non-US countries) through a “Double Dutch” subsidiary arrangement.
A big part of Uber’s future depends on success in China, where it’s in a desperate (and so-far losing) battle with Chinese competitor Didi Kuaidi to become the market leader; tech-press outlet Pando has repeatedly documented the flimsy ground under Uber’s claims of success in China where widespread fakery among drivers inflating the number of rides is a major problem.
Others, including both critics and academics, have done their own data collection to take an alternative look at the sharing economy. In addition to my own efforts, Murray Cox’s Inside Airbnb is a valuable source of data that has been used in many media reports. Academics are increasingly drawing on these data sources (as well as their own) to do their own investigation of the impact of the sharing economy on cities around the world.
Scholarly studies take a long time to complete, but they contribute valuable and novel perspectives. For example, a report from researchers at the Harvard Business School argues that the Airbnb platform promotes routine racial discrimination; another from Boston University shows that the ratings on Airbnb have no correlation with other quality metrics; and others argue that Uber’s surge pricing is not so transparent as it seems. Legal scholars are particularly well-positioned to untangle the complex set of issues around the sharing economy: Vanessa Katz has a wonderfully clear summary of many of the inherent legal issues.
These scholars and many others are highlighting big gaps in the sharing-economy companies’ stories, and there is every reason to believe that more dirty laundry is coming.
Data-driven critiques are part of larger organizing efforts by communities affected by the business practices of companies like Airbnb and Uber. Affordable housing advocates have taken the lead in questioning Airbnb’s impact in cities; coalitions such as Share Better in San Francisco and New York have been pursuing both lobbying and community action. Uber drivers themselves are leading the efforts to improve their working conditions (with some success, such as a decision in Seattle to let them unionize), there are active online forums at uberpeople.net and Reddit, and occasional one-day strikes and other protests are becoming more and more common.
City governments themselves have taken a lead in pushing back against Airbnb in tourist-intensive cities such as Barcelona, where the explosion of Airbnb offerings has contributed to fears of Disneyfication — a city with many attractions but no residents. The new mayor of Barcelona, Ada Colau, is a left-wing former housing activist who has taken a tough line on short-term rentals, threatening to fine Airbnb if it markets unregistered apartments.
The IPOs of Airbnb and Uber continue to be postponed, allowing the companies to operate with little transparency and great impunity. But they will happen, and when they do the debates around the sharing economy’s role in cities will only intensify. The companies will continue to use their own data to create shiny success stories, but other narratives are also emerging — narratives that challenge the vision of an inevitable Airbnb- and Uber-driven future.