Free Agent Nation is out there, and parts of it are growing fast.
The Bureau of Labor Statistics reported last week that 14.4 million Americans were self-employed in January. Of those, 9.2 million were unincorporated self-employed workers and another 5.2 million were incorporated.
That’s interesting, given that back in January 2000 (which is as far as the BLS tally of the incorporated self-employed goes), the number of self-employed was … 14.4 million. Since then there have been some modest ups and downs, but overall no change. And as you can see in the chart below, the long-term trend in the percentage of workers who are self-employed actually appears to be downward:
But isn’t this the age of Free Agent Nation, as Dan Pink declared back in 1997? What about “The Rise of the Supertemp” that Jody Greenstone Miller and Matt Miller reported in HBR in 2012? Or “The Third Wave of Virtual Work” described by Tammy Johns and Linda Gratton last year in HBR, which has untethered knowledge workers from offices and made independent work more practical? It is as if, to paraphrase economist Robert Solow, you can see the age of self-employment everywhere except in the self-employment statistics.
Why is this? Two reasons, mainly. One has to do with definitions — the BLS standard for self-employment isn’t the only valid one. The second is really about history. We may well be witnessing the rise of a new kind of independent worker, but there have been different kinds of independent workers in the past. Far more of them as a percentage of the workforce, in fact, than we see today or are likely to see anytime soon.
First, the definitions. The BLS gets its self-employment totals from the Current Population Survey, a.k.a. the household survey, a monthly quiz of 60,000 American households conducted by the Census Bureau (this is the same survey that generates the unemployment rate). Respondents are asked, “Last week were you employed by government, by a private company, a nonprofit organization, or were you self-employed?”
This either/or choice excludes a lot of people who are doing independent work on the side, or whose jobs are really more like gigs. A survey conducted for the past three years on behalf of MBO Partners, a provider of support services for independent workers, counts temp workers, on-call workers, and those on fixed-term contracts as “independent workers.” That gets the total to an estimated 17.7 million in 2013, up from 16 million two years before. “When you start throwing these other people in, that’s where the growth is,” says Steve King of Emergent Research, which designed the survey.
“The household survey is really good,” continues King. “I don’t think they’re missing people who are working; they’re just categorizing them using methods they developed in 1950. Changing that survey takes an act of God, because it messes up all the time series.”
So others are driven to adopt their own categories. The Freelancers Union frequently cites a number of 42 million independent workers, about a third of the workforce. It gets that from a 2006 Government Accountability Office report that said there were about 42.6 million “contingent workers,” meaning “agency temporary workers (temps), direct-hire temps, on-call workers, day laborers, contract company workers, independent contractors, self-employed workers, and standard part-time workers.”
Now, classifying all the nation’s part-timers as independent workers is quite a stretch. I’m not going to carp too much because it’s in keeping with the Freelancers Union’s noble aim of taking the lemons dealt by the labor market over the past decade-plus and turning them into artisanal lemonade. Plus, they could have rounded up to 43 million if they wanted to. But there aren’t 43 million truly “independent” workers in the U.S.; there are 43 million people who are working but aren’t doing it full-time for somebody else.
Another approach to the self-employment data paradox is to measure something completely different. Dan Pink, for example, pointed me to the Census Bureau’s annual tally of “nonemployer businesses,” which is taken from tax return data:
This seems like a realistic proxy for Pink’s Free Agent Nation, and it shows some actual growth, especially in the early years of the millennium. What it doesn’t show is the kind of explosive growth that advocates for the self-employed sometimes proclaim. MBO Partners, for example, predicted in 2011 that by 2020, “70 million people, more than 50 percent of the private workforce, will be independent.” Given that the actual research commissioned by MBO says that there are now about 17.7 million independent workers and the number has been growing at 5% a year, this was either a bold bet that the Affordable Care Act is going to drive/lure a lot more people out of full-time jobsthan the Congressional Budget Office is predicting, or silly marketing hyperbole.
Still, one perfectly valid reason why some people persist in being very bullish about the rise of independent work in the face of some pretty unbullish overall statistics is that some kinds of independent work, among them the kinds probably most relevant to HBR readers, are in fact on the rise. They’re just not the only or even the main kinds of independent work out there.
Self-Employment Old and New
For the first century of its existence, the United States economy was dominated by independent workers. Most of them were farmers. Others were tradespeople, professionals, hands-on service providers, and such. Even after the huge economic transformations wrought by the first half of the 20th century, the self-employed still made up more than 19% of the workforce in 1949, according to the BLS, compared with just over 10% today. (The BLS data series on self-employment starts in 1948.)
Most of this decline has been due to the continued emptying and consolidation of America’s farms (or, if you prefer, the productivity revolution in U.S. agriculture). Self-employed farmers, ranchers, hunters, fishermen, and loggers made up more than 8% of the workforce in the late 1940s. Now it’s less than 1%.
It’s not just that, though. Other long-established varieties of independent work have been declining, or at least not growing much. Mom & pop stores, definitely, but also some other very large categories that may not immediately spring to mind. Doctors, for example, who have been reacting to the increasing complexity of their business by grouping together or joining hospital staffs. Sole-practitioner lawyers who are getting out of a field that seems to be in long-term decline. And real estate agents, contractors, and others who are dependent on a boom-bust housing sector which has been mostly a bust in recent years.
Here are the occupational categories that have seen the biggest declines in self-employment since 2001:
These numbers stem from data collected by the Census Bureau in the American Community Survey, a rolling census of about three million people a year that the House of Representativesvoted in 2012 to defund (the Senate didn’t concur). The Census Bureau doesn’t publish these numbers in very user-friendly form, but Economic Modeling Specialists Intl., a subsidiary ofCareerBuilder, gets the raw data, massages it with numbers from some other government surveys, and delivers a remarkably detailed portrait of what the unincorporated self-employed are up to. On Thursday, EMSI published a nice roundup of developments since 2006. They also sent me a spreadsheet with the numbers for every occupational category going back to 2001. Here are the occupations with the biggest gains in self-employment over that period:
This doesn’t exactly offer resounding support for the thesis of a boom in independent white-collar work. The world surely needs more landscapers and maids than management analysts, and it gets them. Also, some of those construction managers and first-line supervisors from the previous chart seem to have landed less-remunerated jobs as construction laborers in this one. “Management, other,” in case you’re wondering, is a grab-bag category of managers who can’t otherwise be classified as construction managers, purchasing managers, financial managers, etc.
These numbers don’t, however, include the incorporated self-employed, who tend to skew toward higher-paid work. Also, the vagaries of the occupational categories used by the government can have a big impact. All the nation’s maids get thrown into one category; its physicians are split into at least nine. With a few days of work, I might be able to coax a clearer narrative from the data. Failing that (I may get to it later), I pulled out a few categories that grabbed my attention:
The boom in musicians and singers took me by surprise. Maybe this is one case where the Long Tail really has panned out for people. The two financial occupations both show the impact of the financial crisis, but it’s interesting (and I think heartening, from the consumer perspective) to see that the financial advisors are rebounding while the salespeople aren’t. With editors, the upward trajectory seems indicative less of a boom than a bust — the publishing industry has been shedding jobs for years, and now relies on freelancers more than it used to. But that does represent a shift to independent white-collar work, even if a lot of the people doing it would prefer a regular paycheck. The boom in independent web developers is of course exactly the kind of New Economy, Free Agent phenomenon we’ve been expecting to find. And the rise of fitness trainers and mental health counselors is indicative that there are hands-on independent service jobs on the rise that are more remunerative and presumably more pleasant than mowing lawns and vacuuming floors. (Dietitians and nutritionists and lots of different varieties of counselors, therapists, and psychologists are also seeing substantial gains.)
Sprinkled among the EMSI spreadsheet are other attractive-sounding (if not exactly giant) occupations on the rise: scientists of various kinds, human resources specialists, computer and information systems managers, technical writers, market research analysts, and so on. Free Agent Nation is out there, and parts of it are growing fast. It’s just not always easy to find, and by the looks of it still has a long ways to go before it takes over the rest of the nation.