Last week the Bureau of Labor Statistics (BLS) concluded that just 1 percent of jobs in 2017 were enabled by online platforms, adding new data to a recent line of evidence suggesting the gig economy is not as large as many have speculated. This conclusion is the result of new questions added to the Contingent Worker Survey (CWS) that are intended to gauge the number of individuals who earn money from online platforms such as Uber, TaskRabbit, and Airbnb. While the BLS reported several issues when administering the new survey questions, the finding is consistent with other recent data and suggests that the gig workforce plays only a minor role in the economy.

Before digging into the findings and how they compare to other data, it is important to note that the BLS faced several challenges when administering these new survey questions. In particular, the BLS noted a number of false positives, meaning that many respondents indicated that their jobs were electronically mediated when, in fact, they were not. This suggests that the survey questions were unclear, confused respondents, or did not adequately communicate the intended question. This problem also relates to definitional issues that have long plagued researchers attempting to measure the gig economy. In particular, social norms have yet to clearly define a “gig” job. Evidence indicates that many who do earn income from online platforms do not consider themselves to be employed. For instance, Etsy, the online marketplace, found that 24 percent of its sellers consider themselves unemployed.

The BLS accounted for these issues by recoding the survey responses so they more accurately reflect the situation of each person interviewed. While the BLS is confident that the reported estimates are more accurate, the recoding substantially changed the survey results. Specifically, when based purely on interview responses (both correct and incorrect), the survey indicates that 5.1 million workers (3.3 percent of total employment) earned income from online platforms. After recoding the data to make survey responses more accurate, BLS concluded that just 1.6 million jobs (1 percent of total employment) were enabled by online platforms. This estimate includes those whose online platform activity was their main job, their second job, or any additional work for pay.

The survey results certainly are not perfect. The BLS finding, however, that just 1 percent of jobs are electronically mediated does fall in the range of estimates from other recent research. In particular, economists Lawrence Katz and Alan Krueger administered their own survey and estimated that those using online platforms accounted for only 0.5 percent of all workers in 2015. More recently, the JPMorgan Chase Institute analyzed bank transaction data and concluded that 1.6 percent of families earned income from online platforms in the first quarter of 2018.

The new finding is also consistent with more trusted data that indicate the gig economy is not significantly reshaping the workforce. In particular, earlier this year the BLS released the results from the CWS’s existing, well-tested questions on alternative work arrangements. Those findings indicate that the prevalence of alternative work decreased over the past decade. The portion of workers who were independent contractors, on-call workers, temporary help agency workers, or contract company workers in their main jobs declined from 10.7 percent in 2005 to 10.1 percent in 2017.

The new data also indicate that even alternative workers are not very likely to use an online platform. Of those whose main jobs were independent contractors in 2017, just 5.6 percent were electronically mediated. Similar figures were found for on-call (2.6 percent), temporary help agency (3.4 percent), and contract company (2.4 percent) workers.

Evidence on the gig economy remains flawed and inconclusive. Recently, however, new data has emerged indicating that the gig economy has not fundamentally altered the workforce. Rather, it is a small portion of the U.S. economy.