The Rise of Invisible Work
Last week, in the face of growing scrutiny from the city of New York, the five-year-old accommodation-sharing company Airbnb released some data designed to make its case as a good neighbor – even an economic catalyst – in the five boroughs.
In the previous year, according to numbers crunched by the consulting firm HR&A Advisors, Airbnb had helped generate $632 million in economic activity throughout town, supporting 4,580 "jobs." Its hosts – individuals often held up as direct competition to major hoteliers – were making on average $7,530 a year renting out their homes. And the visitors they welcomed stayed longer (6.4 nights on average) than the typical New York tourist, and spent more money in the process ($880 at New York businesses).
This was the sixth local impact study Airbnb has produced as part of its campaign to move beyond personal anecdotes about struggling renters to hard economic data. Specifically, that $632 million speaks to the question of Airbnb’s footprint in New York City: Is it siphoning customers from Manhattan’s Marriotts, or creating new customers – and economic activity – that never existed before? (Another data point: 82 percent of Airbnb’s New York listings are outside of the tourist epicenter of Midtown Manhattan.)
But there’s also a much larger question lurking behind early data of this sort that's starting to trickle out of "sharing economy" companies. Exactly what is this entire industry doing to the economy? When people share and rent things instead of buying them, does that mean we need to produce less stuff, requiring fewer jobs, ultimately creating less economic growth?
So far, the sharing economy’s impact has been largely unseen because we (and the Bureau of Labor Statistics) are used to counting employment in whole jobs, or part-time jobs, not something-I-do-on-the-side-while-I-freelance jobs. Currently, companies like Airbnb, and Etsy, and Sidecar enable tens or hundreds of thousands of people who are even further down the food chain than “small businesses.” They’re micro-entrepreneurs doing something so nontraditional we don’t even know how to measure it.
"It's like an invisible economy," says Molly Turner, Airbnb’s director of public policy. And companies inside it need to get real sophisticated, real fast in figuring out how to measure this invisible economy, she says, because for many people this looks like an important part of the future of work. It could even be a big component in our understanding of why America's jobs numbers look so lethargic.
"That's not to say Airbnb advocates for counting this as work, or as a 'business,'" Turner says. "But it's a productive economic activity that has a really important impact."
So how do you measure that impact when the standard definitions of "work" and "jobs" don't even apply?
"The overarching question here is 'how much economic growth will be induced by the sharing economy?'" says Arun Sundararajan, a professor at the NYU Stern School of Business who has been studying the question. "Contained in that is the possibility that the sharing economy may not lead to economic growth."
• • • • •
It’s notable that a company like Airbnb, at this stage in its evolution, is trying to put out data like this at all. Look back at Google circa 2003, Sundararajan says. The year before Google went public, the company was interested in trying to size up how large its market would be. It was interested in articulating its revenue model for the public. It was not all that interested in painting a picture of its broader impact on the world.
“Most entrepreneurs and firms are not out sharing their data, and I would even use the term, trying to justify their existence,” says Lesa Mitchell, the vice president of innovation and networks at the Ewing Marion Kauffman Foundation, which advocates for entrepreneurship. “You don’t see that normally. They just do their stuff.”
But that’s not really an option for Airbnb or Uber, companies founded on tech platforms that are creating new, more efficient ways of doing what other people are already doing (and what regulators are already regulating). There’s a creative destruction element to the sharing economy that theoretically threatens hotels or cab companies, or even the auto manufacturers who used to build and sell cars to 25-year-olds who’d now rather use Zipcar instead.
Back in August, Business Insider published a story warning that the sharing economy and the rise of renters could have “catastrophic ripple effects.” A market strategy firm called the ConvergEX Group had been circulating a memo outlining how demand for new cars and houses would drop, with a potential effect on the consumer psyche “similar to that of the Depression.”
That argument also neatly nests within wider fears that technology is now destroying jobs faster than it is creating them.
The innovation behind many sharing economy companies, however, is that they are making it more efficient to hail a cab, or rent a hotel, or order a wedding cake topper. On the supply side, they’re making it more efficient to provide these services or sell these products. And an increase in efficiency generally leads to an increase in economic productivity, which, generally speaking, translates to economic growth.
Give people a more efficient, convenient, cost-effective way of getting something, and typically the demand for that thing actually goes up. “If something becomes better,” Sundararajan says, “people want more of it, not less.”
In other words, an expansion of consumption possibilities leads to an increase in consumption. This is essentially a story of economic progress, Sundararajan says: new technology leads to more efficiency, which leads to an expansion in production and consumption possibilities. There are substantially more hotel rooms in America today than there were 50 years ago, because people have accumulated more leisure time, more income, more need for business travel.
On the other hand, there are substantially fewer farmers than there were in America in 1900. But as farming became more efficient, people who once were farmers found other things to do instead (conveniently right around the same time as a massive expansion in secondary education).
“So to think about things in a really static way – we have a certain amount of stuff, and if people are using it more efficiently, we need less stuff – it’s sort of myopic,” Sundararajan says.
• • • • •
The sharing economy, however, is not exactly like economic disruptions that came before it. More complex technology typically demands more complex skills. And, in the past, major technological change in the economy has been accompanied by a shift toward higher-skilled work. Initially, there’s a shortage of those high-skilled people. Their wages go up, while wages go down for the people who are steeped in the old technology.
The sharing economy is fundamentally premised on new technology, and it's creating new jobs exactly like this for the developers and programmers on the back end of Etsy’s platform or SideCar’s app. But that’s not the most interesting part of this story.
“eBay’s impact hasn’t been on the thousands of tech jobs it created for eBay,” Sundararajan says, “but on the hundreds of thousands of sellers it created.”
That’s where the real economic impact here lies, and it’s not actually clear if all of those people – Uber drivers, Etsy sellers, Airbnb hosts – need more complex skills than what was required of them a decade ago. If you sell furniture on Etsy that you built with a Makerbot 3D printer that you keep in your living room, your skills probably have grown more advanced.
But, for the most part, the sharing economy is not creating new machines that people must learn to use to produce more stuff. It’s creating new marketplaces to access familiar things in better ways.
"There’s something a little different going on with the sharing economy," Sundararajan says. And he doesn’t yet know the broader implication of this.
He’s wary of all the estimates he sees of the size of the sharing economy, or any projections for its future growth, because we also don’t know how quickly this new kind of consumption will become “normal.” This is a question of culture, not public policy. Everyone knows how to behave when you walk into a hotel. But most of us are still a little uncertain about what’s supposed to happen when we arrive in someone’s else apartment to rent a room for the night. That’s a sign that an idea like Airbnb hasn’t yet become legitimized. And the growth of the whole sharing economy – and its economic ripple effect – depends on the speed at which that happens.
Sundararajan also suspects that many of these platforms are quietly operating as “finishing schools” for tentative entrepreneurs, another effect that so far hasn't been measured. You can test out selling some of your crafts on Etsy, for example, without quitting your job and cashing in your 401K to open a brick-and-mortar store. Whole new companies will start this way, he predicts. And that would have an impact on the economy, too, that would be inarguably positive.
“I don’t know how it can’t be,” Mitchell says, “simply because in using things like Daily Grommet, Etsy, these platform technologies that are out there, that’s allowing people in small towns that have been impacted by business closures to actually make and sell things online to customers they’re never going to meet or see.”
• • • • •
Sundararajan chuckles whenever New York City gets wound up about a new tech company relocating to town.
“Google comes, hundreds of tech jobs are created, and there’s a lot of hoopla about these things,” he says. “Meanwhile, Etsy is quietly creating massive amounts of employment, and they’re not counted as jobs.”
Pull back and consider this network of uncounted jobs, sort-of-income, and tentative entrepreneurs all across the country, and a larger possibility emerges: Maybe the economy is doing marginally better than we think. Those stagnant jobs numbers that come out of the Bureau of Labor Statistics the first Friday of every month don’t account for much of this activity.
Earlier this summer, the MIT Technology Review published a startling article based on the work of MIT’s Erik Brynjolfsson and Andrew McAfee showing not only that technology was destroying jobs, but that productivity in America had recently become decoupled from employment. Traditionally, the two rise in tandem. As productivity increases, jobs increase, too. But amid the massive new wave of automation, starting around 2000, this has ceased to be true.
That chart, from McAfee, shows the widening gap between productivity and employment in America (the pink arrow inserted is ours). This is deeply troubling to economists. But it’s possible – and this is what Sundararajan and others are hoping to measure – that the gap could contain all these people we are currently not counting. That's not to say that these unconventional workers fill up that entire gap, but they could be an important part of the picture.
And if that turns out to be true, it should change not only how we think of people like the part-time Airbnb host, but also how we create climates that enable them.
Update: A couple of readers have asked why elements of the sharing economy wouldn't be captured in existing government measures like the BLS Current Population Survey (or household survey), so here's a deeper dive into the weeds:
Take a person who was completely unemployed and now does some kind of work as a sharing economy supplier, maybe driving an Uber car, or providing services on TaskRabbit. If that person reports this new work for profit in a response on a BLS survey, they would be counted as an additional job. However, Sundararajan adds, unemployment (or employment) numbers reported by the BLS don't capture the additional "employment" or "work" generated by underemployed people who were already working at least an hour a week (like a software contractor who now also does Lyft on the side).
Additionally, that data isn't good at reflecting people who dip into the sharing economy while holding regular full-time jobs (a lawyer who rents out a room on Airbnb, a furniture supplier who creates and sells his own work on Etsy on the side). It can't capture the effect when a previously existing small business owner now receives substantially more employment thanks to Etsy. It also can't tell us anything about one of Sundararajan's key questions: Are these micro-entrepreneurship platforms enabling people to create more traditional new businesses that wouldn't have existed without the launching pad of the sharing economy?
Part of the challenge also lies in how people who do these activities think of them. Sundararajan worries that survey questions that ask about your "job," your "main job," your "other job," or your "business" may lead people in the sharing economy to under-report what they're up to simply because they don't think of their activities in those terms. A knitter on Etsy thinks she's selling the products of her "hobby." A driver on SideCar thinks he's taking gas money to drop people off while he's already on his way across town to his part-time job. And an Airbnb host doesn't quantify how many hours she spent last week "hosting" tourists from Chicago. Airbnb in particular leverages excess space, not excess time, making it difficult to measure that "work" in hours.
The underlying point here isn't necessarily a critique of official statistics. Sundararajan believes the only way you measure the impact on the economy of this kind of technological change is to go to the source of the technology itself: these companies. Only they can answer how many sellers use their platforms, or how much money the average vendor is earning, or how many people are served by their products, enabling ripple effects throughout the economy. Which brings us back to our starting point: It's time for companies in the sharing economy to begin rounding up their own data.