The Washington PostDemocracy Dies in Darkness

Uber and Airbnb helped expand the gig economy. These researchers think they’ve measured how much.

Researchers at the Brookings Institution have proposed a new metric for measuring the “gig economy” that includes Uber drivers. (Eduardo Munoz/Reuters)

Many economists and labor experts say the “gig economy” has grown in recent years as ride-sharing services, such as Uber and Lyft, and home-sharing sites, such as Airbnb and HomeAway, have gained popularity with consumers. More people today earn part or all of their income working as contractors for these various companies, indicating that the nature of work in certain industries may be shifting away from full-time, salaried employment.

The problem, however, has been measuring the exact size of this labor pool. The gig economy is inherently fragmented as each individual works on a contract or freelance basis, and thus may have several clients and work variable hours over time. It also spans multiple industries. While tech-industry darlings such as Uber and Airbnb are relatively new, there have long been writers, house painters, tutors and Web designers, among other professionals, who are self-employed and work on a freelance basis.

But researchers Mark Muro and Ian Hathaway at the Brookings Institution suggested Thursday a new method for estimating the size of the gig economy and just how big of an impact services such as Uber, Lyft and Airbnb might be having. They used an IRS designation for “nonemployer firms,” defined as businesses with no employees that earn $1,000 or more per year, as a measure for self-employed contractors or freelancers. They then tracked the number of nonemployer firms in the transportation and lodging industries over a four-year period as Uber and Airbnb were becoming more popular in metropolitan areas across the country.

Their metric is far from perfect. For one, the nonemployer firms category includes any self-employed person with no employees who earned more than $1,000, regardless of whether they earned that money driving for Uber or mowing their neighbors’ lawns. It also only captures those individuals who declared that income to the IRS.

“This is, I think, a proxy measure that we’re very comfortable with, but it is a proxy,” said Muro, the director of policy at the Metropolitan Policy Program at Brookings and the report’s co-author. “We don’t have a specific measure of every single person using these platforms, and we are likely undercounting.”

Still, their findings may begin to answer some of the questions about the size and scale of the gig economy. For example, the report found that there were 24 million nonemployer firms in the United States in 2014, up from 15 million in 1997 and 22 million in 2007. That means the overall growth rate in nonemployer firms has been slower in recent years despite the arrival of resource-sharing apps. Nonemployer firms also remain smaller in number and economic impact than traditional payroll employment, which totaled 145 million in 2014.

In the ground transportation and lodging industries, specifically, the number of nonemployer firms grew significantly at the same time companies such as Uber and Airbnb were taking off. Again, that’s the report’s proxy for contract workers.

The number of nonemployer firms in the ground transportation industry grew at an annualized rate of 69 percent between 2010 and 2014, researchers found. The beginning year coincides with Uber’s launch in San Francisco, and the end year is the last for which data is available. During the same period, payroll employment in the ground transportation industry grew a comparatively smaller 17 percent.

Similar trends played out in the lodging industry, though on a smaller scale. During the period from 2010 to 2014, the number of nonemployer firms grew 17 percent, compared to payroll employment growth of 7 percent. The numbers are lower, at least in part, because lodging tends to require employees and has different tax reporting requirements than ground transportation, the report notes.

While the report stops short of declaring a causal relationship between the popularity of resource-sharing apps and the rise in nonemployer firms, its authors theorize that a strong connection exists between the two. Because Uber and Airbnb launched in different cities at different times, the authors said they could see these trends play out on the local level as the companies moved into new markets.

“Places that didn’t have Uber and Airbnb for very long before this time [period] don’t have very much of this activity,” Muro said.

Also unclear from the report is whether the rise in contract workers has come at the expense of payroll workers. For example, are full-time drivers losing their jobs because of gig economy drivers? The report looked at the change in payroll employment and nonemployer or “gig” employment in the top 50 U.S. metropolitan areas between 2012 and 2014 to tackle that question. The results varied greatly depending on the city.

In early-adopting San Jose, for example, “gig” employment grew 145 percent while payroll employment shrank 31 percent. In Phoenix, gig employment was up 76 percent while payroll employment was also up, 21 percent. Rochester, N.Y., was a rarity in which gig employment grew more slowly (4 percent) than payroll employment (24 percent). In all of the cities, the change in absolute numbers was quite small.

For the United States as a whole, gig employment had larger percentage growth in both the ground transportation and lodging industries. Ground transport saw 48 percent growth in gig employment compared to 6.4 percent growth in payroll employment. Lodging posted 9.7 percent growth in gig employment, compared to 3.6 percent growth in payroll employment.

“Because more places show growth on both sides, the payroll side and the gig side, on balance our measures show that thus far the gig economy in most places is creating new demand as opposed to cannibalizing a finite market,” Muro said.

“That said, there is evidence that at some point there could be saturation. In some places there has already been negative impacts on payroll. To be truthful, I don’t think we can say that’s a causal relationship, but it’s certainly theoretically a possibility,” he added.

Read more from The Washington Post’s Innovations section