Sarah Holder is a staff writer at CityLab covering local policy, affordable housing, labor, and technology.
A study on a subset of the gig economy—yes, another one—out of JP Morgan Chase suggests the share of people using online platforms to find work is growing, but hours and earnings are far from regular.
Assessing the size of the gig economy is like looking into a double-sided mirror: Look in it one way, and it’s bigger than ever. Stare at the other, and you might have to squint.
This spring ushered in studies that seemed to indicate the latter. In June, the Bureau of Labor Statistics reported that only 6.9 percent of the total workforce counts as a contingent worker (loosely defined as workers who may have a long-term employer, but are compensated by the individual gig; or self-employed contract workers). Another analysis from the Economic Policy Center found that Uber drivers (mainstays of the online gig economy) only accounted for about half a percent of total employment nationwide.
But a new study out of the JP Morgan Chase and Co. Institute attempts to analyze a more granular sector of the gig economy: People who find and are paid for their labor through online platforms. What they found is that the number of workers using apps to get gigs is increasing. It’s just that most of them still aren’t gigging that often.
Instead of relying on self-reported surveys, tax information, or data from gig companies themselves—which don’t always capture the full scope of online platform users—the bank analyzed deposits made to 39 million Chase checking accounts over 5.5 years. Of that pool (which is large but not comprehensive: It only captures users in states where Chase has retail branches, about half of the 50 states.), 2.3 million people had earned a total of $38 million from the online platform economy—for JP Morgan’s purposes, a sample of 128 apps that “connect suppliers directly with demanders,” “mediate payment,” and “empower participants to enter and leave the market whenever they want.”
The study didn’t identify the apps by name, but they fell into four broad categories: Those used for transportation (ride-share innovators like Uber and Lyft, but also delivery apps like Postmates), non-transport work (most everything else, from home repairs to writing Tinder bios), selling (sites like eBay and Etsy), and leasing (home-sharing apps like Airbnb).
First thing’s first: By these metrics, the share of U.S. residents participating in the online-driven gig economy seems to be growing. In the first quarter of 2013, only 0.3 percent of people had participated, a number that jumped to 1.6 percent in the first quarter of 2018. By March 2018, 4.5 percent of families had participated at least once over the previous year, the study claims. Transportation workers were, unsurprisingly, the largest cohort.
But Fiona Greig, the lead author of the study, also found that in general, this deepening pool of gig workers are not gig working consistently at all. In no sector did the majority of workers participate in any online platform for more than a quarter of the year. In the selling and leasing marketplaces, around 70 percent of participants worked only one to three months out of the 12; and in the transportation sector (the biggest cohort), almost 60 percent participate for only three or fewer months per year.
The study didn’t parse whether people are jumping from sector to sector (giving up on Lyft after a couple of months to try dog-walking instead, for example), but Greig says that 20 percent of transport gig workers in their study used more than one transportation-focused platform each month. They just weren’t using any platform for that long.
The barriers to entry are so low, both from a skills and car-procurement standpoint, and integrating delivery or ride-share trips into your day is so easy, it makes sense that workers could dabble as occasional transporters, and that more are doing so today.
But perhaps the most alarming finding is that over the last five years, monthly earnings have dipped across the entire online platform economy, especially for transportation industry workers, for whom earnings have dropped 53 percent since 2013. This decline could be interpreted as a harbinger of doom for this type of work, suggesting that as the supply of drivers grows (and remember, it’s growing), they’ll all be earning less.
Uber pushed back on that explanation Monday, publishing a challenge to the JP Morgan report that highlighted the key distinction between monthly earnings, which the study measures, and hourly wages, which the study doesn’t: “If the share of our partners who drive only occasionally has increased over time, as it has, it stands to reason that the average of every driver’s monthly (or, for that matter, weekly or yearly) earnings would decrease,” Libby Mishkin, a senior economist at the ride-sharing platform, wrote.
In a previous Uber-led analysis, published with an NYU economist, the company found that average hourly wages have remained fairly stable, even as prices fell. (The report called the equilibrium “largely inefficient.”) Another, which compared the wages of drivers across 20 cities, found that over a 16-month period, “there was virtually no difference between people driving for many hours each week and those who drove only a few hours per week,” because when fares fell, more people used the app.
Greig said she “completely agree[s] with their assessment.” The JP Morgan study only measures the total amount each worker gets from a platform each month. “We don’t observe directly whether people are working more or less in terms of quantity of time—we don’t observe quantity of time at all—nor do we observe whether they’re getting paid per trip or per hour,” she said. “We don’t observe what’s happened to wages; and we don’t observe what’s happened to hours worked.”
There were other limitations to the data—namely, that it can’t be used to make conclusions about the entire gig economy ecosystem. “Instead, it might reflect the migration of previously existing forms of contingent work,” like what the BLS measured, “onto online platforms,” it reads. “In fact, a recent survey of how Gig Economy workers are paid shows that an increasing share is being paid through software platforms rather than directly from customers.”
But the fact that monthly earnings have dropped so precipitously, especially in the transportation sector, is still significant, Greig says. (And, she notes, 35 non-Uber platforms, some of which aren’t ride-sharing platforms at all, were analyzed. Few of them have produced their own evidence of stable, hourly wages.)
If monthly earnings aren’t falling because trip prices and hourly wages have, in fact, dipped in a saturated marketplace, then why? It could be that, as the data shows, people are participating on a more sporadic basis over time, therefore earning less each month—or that, as the unemployment rate shrinks, they simply have fewer hours to offer.
Uber isn’t worried. On the contrary, the company hopes the study will tamp down on the fear that 9-5 desk jobs will soon be eclipsed by gigs. "This study should put any notions about platform work becoming the predominant form of work anytime soon to rest,” said Mishkin. “Though more people are clearly signing up to earn with Uber, it's primarily a form of supplementary work for most drivers."