Are more or fewer Americans engaged in self-employment than a decade ago? While you might think that this is a simple factual question, its answer depends on which federal government agency’s numbers you look at.
Internal Revenue Service (IRS) data indicates that self-employment is growing. The tax authority estimates that the number of self-employed Americans increased 26.4 percent between 2000 and 2011, the most recent year data is available.
By contrast, the Bureau of Labor Statistics (BLS) data indicates that self-employment is shrinking. The statistical agency responsible for measuring the condition of labor markets finds that the number of self-employed Americans decreased by 0.7 percent between 2000 and 2011.
Some might see these conflicting numbers as an illustration of Benjamin Disraeli’s famous observation that “there are three kinds of falsehood: Lies, damned lies, and statistics.” But I believe that the numbers can be reconciled if we understand the source of each.
The IRS measures self-employment by looking at whether or not a taxpayer claimed the self-employment deduction on his or her 1040. Because taxpayers must pay self-employment taxes if they have “net earnings of $400 or more as a self-employed person” – which the IRS defines as someone in business for him or herself. A self-employed person in the IRS data is anyone who has a non-trivial amount of self-employment income, regardless of their other income sources.
Take me, for example. I would be included among the self-employed in the IRS data because I earn more than $400 per year writing, even though my primary job is as a professor.
The BLS measures self-employment by conducting a monthly survey of approximately 60,000 households to identify the employment status of household members 16 years of age and older. To figure out each household member’s employment status, the BLS asks the respondent to identify whether each adult member of the household worked for pay or profit in the week of the survey.
If the household member had more than one job in the week, the respondent is asked to focus on the household member’s primary job, which is defined as the one on which they spent the most hours. Respondents are then asked (PDF) if the household member was “employed by the government, a private company, a nonprofit organization, or were they self-employed” to determine if they are self-employed.
If we return to me as the example, we can see why the IRS and BLS self-employment numbers differ. I would not be included among the self-employed in the BLS data. Because my full-time job is as a professor. I would be classified as wage employed, even though I earned more than $400 per year writing.
Understanding the source of the BLS and IRS self-employment numbers helps to tease out trends in self-employment. Fewer Americans appear to be making self-employment their full-time jobs than was the case a decade ago, but more of us are engaged in self-employment on the side.
That conclusion jibes with anecdotal evidence. Many observers have noted that rise of the Internet has made it easier for people to earn a little side income by buying and selling goods on eBay or by renting out rooms in their homes through sites like Airbnb.
Receiving Information Photo via Shutterstock
Scott: You could do plenty of statistics, right?! 😉 Have you seen the same challenge in comparing stats in other countries? What will happen when new ways of self-employment will develop, e.g., “business employment”?
I agree that many more people with “regular” jobs are also undertaking “side” jobs that would lead to reporting self-employment income to the IRS. The internet is a beautiful thing and makes this quite easy to do.
Very helpful article. Thanks so much for writing!
That explains the gap in the numbers. That’s because BLS don’t measure people who have a full time job as self employed even though they may be handling more than one jobs.
That makes a lot of sense. It seems to me, then, that the IRS statistic is really mislabeled. The fact that more people are taking on projects that lead to self-employment income may be meaningful in some contexts but without some understanding of the amounts of income reported (not just the raw numbers of individuals reporting some self-employment income). it is difficult to see any real significance to the statistic.