Is the freelance economy really shrinking?

There are fewer temporary workers today than there were in 2005, according to a much anticipated report on contingent workers from the Bureau of Labor Statistics released this morning. The new data offers some insight into the number of people who rely on temporary work for income, but experts say it may not paint a full picture of the changing nature of work.

As the Aspen Institute noted in a statement shared with Fast Company, “While the Contingent Worker Supplement (CWS) data shows a slight decrease in the number of workers who rely on contingent or alternative arrangements for their main job, it does not provide information about the number of workers that engage in this type of work to supplement their income.”

BLS researchers included the freelancing survey as a supplement to the monthly Current Population Survey, which goes out to 60,000 households in the United States. For the first time the CWS, which was first instituted in 1995, includes four new questions intended to study app-based labor from the likes of Uber, Lyft, TaskRabbit, and Fiverr. The last time the CWS was conducted was 13 years ago.

Unknown labors

Since then, work has changed significantly in ways that make “nontraditional” employment increasingly difficult to quantify. A 2016 study by economists Lawrence Katz and Alan Kruger found that 94% of the 10 million jobs added to the U.S. economy were from alternative work arrangements. That’s not to say everyone is working for Uber, though; in the decade following the 2008 financial crisis, Americans have turned to a variety of work arrangements, cobbling together full-time jobs, personal businesses, and side gigs to make ends meet.

Many of them aren’t getting traditional employment benefits as a result, exposing a significant segment of the labor force to new vulnerabilities. The nature of contract work is that it is temporary. Historically, contractors have ranged from plumbers to home designers who could set their own rates for work performed over of defined time period or for a designated type or number of tasks.

With the arrival of other app-enabled “gigs,” the definition of contract work has had to evolve. In the Uber model, a middleman sets the price of the job, collects a fee on that price, and connects an independent worker with work. Because the worker is technically self-employed (a technicality that’s now at the center of lawsuits around the country), she has to cover the costs of car insurance and maintenance, as well as her own healthcare, childcare, and work leave.

In addition to traditional contractors and digitally connected giggers, there’s also an “expert economy” of highly paid freelancers, the term that a recent Civic Enterprises study gives to the pool of advertising creatives, business consultants, and similar highly educated professionals whose ranks collectively dwarf the workforces of Uber and Lyft combined. These workers typically have contracts that account for the expense of healthcare and retirement savings, still others may be taking on consulting work in addition to their full-time jobs.

Difficulties in data-gathering

All this diversity in the types of contract work makes it hard for analysts to understand the ways in which it evolves and fluctuates over time. While the CWS report does account for higher-paid independent contractors, for example, the survey doesn’t really address people who combine part-time or full-time work with independent work. (Nor does it fully account for those who take on flexible work in the widening gray area between being on-the-clock and time off in traditional employment.) In other words, the complexity of the workforce itself creates challenges in study design, which may in turn account for radically different findings.

Upwork CEO Stephane Kasriel points out in Fast Company today that, in his view, the BLS’s “methodology probably undercounts the size of the freelance economy.” For one thing, Kasriel observes, Americans might not be on the same page on what constitutes a side hustle versus a primary job, especially for people working multiple jobs. Kasriel also notes that the CWS data is already one year old and that the freelance economy may have changed significantly by now.

In recent years, Upwork, a freelance marketplace, has partnered with the Freelancers Union to release its own annual study, the latest installment of which claims there are some 57.3 million freelance workers in America. Meanwhile, cloud computing platform Freshbooks estimates that by 2020, 42 million people will be identified as self-employed. The BLS study’s figures are much lower, estimating the ranks of independent contractors today at just 10 million.

Another important hurdle to gathering data on the freelance workforce is that companies aren’t required to tell the government exactly how they’re employing people–a circumstance many employers exploit. “Companies are more sophisticated than they’ve ever been in what they know about–who’s working for them, how long, what kind of protections they have–and they don’t have to disclose any of that information,” says Bruce Reed, who coauthored the Civic Enterprises study and co-chairs the Future of Work Initiative at the Aspen Institute. If we want better data, that might need to change.

“Part of the social bargain going forward,” Reed adds, “is to figure [out] what responsibilities companies have to share useful information that’s easier for them to gather than for government [to collect].” In other words, we may need to start asking more and better questions–and of more parties–than just, ‘How big is the freelance economy?’