The government’s brand-new gig economy data is already outdated

Being counted counts, but sometimes counting is harder than you’d expect it to be. That seems to be true for the Bureau of Labor Statistics (BLS), which this morning released its first “Contingent Workforce Survey” (CWS) in 13 years. It’s a welcome new data set but, I believe, a sorely inadequate one.


To be sure, what the BLS is trying to do is admirable. Based on a growing body of research, more people are choosing to freelance by capitalizing on the digital transformation of work, in many cases generating new opportunities for themselves and their communities. Unfortunately, the new CWS isn’t likely a complete snapshot of these changes. For starters, the government’s research samples the rapidly evolving freelance workforce only at wide intervals. Aside from one study in 2001 and another in 2005, every other public report from the Bureau of Labor Statistics is from the 1980s and ’90s. Contingent workers just don’t seem to be top of mind for federal researchers.

Second, the latest study’s findings are already more than a year old now. As the analyst Mary Meeker points out in her annual Internet Trends report, released last week, work is changing rapidly, with technology driving high growth in on-demand jobs. Any changes that might have occurred in roughly the past 12 to 15 months aren’t likely captured in the CWS report.

But perhaps most worrying is that its methodology probably undercounts the size of the freelance economy. In designing this year’s survey, BLS researchers used benchmarking terms like “primary job,” a concept that doesn’t always reflect how people approach and perform various kinds of independent work anymore. It’s shortsighted to use an outmoded measure just because it makes for a handy comparison with historical data; worse, it may not give researchers the insight they’ll need to design better future studies.

Other benchmarking tactics are troubling for different reasons. The CWS is framed around a one-week snapshot of labor activity, which doesn’t provide an adequate timeframe to capture how freelancers are working today. Research that my own company, Upwork, released last fall in partnership with the Freelancers Union, found that independent workers aren’t necessarily working every day or even every week; that income volatility was actually a top concern that freelancers voiced. Analyzing a single week of labor activity every 13 years is unlikely to tell you that.

What’s more, the government found that just under 7% of American workers are freelancers. By contrast, our research estimates that workforce at 36% of American workers.


Still, the CWS adds useful context to the assumptions economists, entrepreneurs, and ordinary workers have otherwise had to make about the freelance workforce. Even if it took the BLS 13 years, it’s a good thing to have this data in hand. The more we know about Americans’ new ways of working, the better we can write policies and develop products and services to support them.

At the moment, we’re largely swinging in the dark. In the short term, without an accurate count or understanding of the freelance economy, we can’t tax work effectively. Longer-term, we can’t be sure that we’re investing in the right types of training, designing innovative benefits programs, or even planning cities for the future of work. Unlike any private entity, the government has the key to unlocking a more complete picture of how the labor force is changing, by tapping into national-level data from many departments and institutions. Those insights could have a tremendous impact on policymaking, infrastructure, and private-sector innovation.

To begin with, the government should run robust studies like the CWS more often. Upwork and the Freelancers Union commission our own survey annually because the rapid evolution of the contingent workforce justifies this frequency. By contrast, the government studies the unemployment rate even more frequently, on a monthly basis. That regularity may be a heavy lift for the BLS, but checking in every year seems not only reasonable, but also crucial to supporting such a significant segment of the U.S. economy.

In the meantime, the government should leverage the data-based advantage it already has. Rather than rely simply on a limited one-week snapshot of labor activity, the government should supplement the CWS’s findings with other data, asking broader questions to tease out the motivations and impacts of changes independent workers respond and adapt to over the course of a year. Aggregating BLS and IRS data, for instance, seems an obvious place to start.

The federal government is great at keeping tabs on various aspects of the U.S. economy and labor force, but particularly where independent workers are concerned, it’s not very good at putting it all together. That’s a shame for everybody, and for freelancers in particular.


Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s

Up ↑

%d bloggers like this: