Skip to main content

Contingent Workforce: BLS is Reassessing Measurement of Nontraditional Workers

GAO-19-273R Published: Jan 29, 2019. Publicly Released: Feb 28, 2019.
Jump To:

Fast Facts

Millions of U.S. workers are employed in the gig economy, holding nontraditional, part-time, and temporary jobs.

To track trends among some of those workers, the Bureau of Labor Statistics measures the number of workers in 3 categories that we collectively refer to as "nontraditional workers." These categories are:

Contingent —temporary workers

Alternative—contractors, on-call workers, or temp agency workers

Electronically-mediated—workers who get gigs through websites or apps

BLS' 2017 survey found fewer contingent workers than in 2005, raising questions about survey effectiveness. BLS is looking into improving the survey.

An alarm on a cellphone screen

An alarm on a cellphone screen

Skip to Highlights


What GAO Found

The Bureau of Labor Statistics (BLS) periodically collects data on and measures contingent, alternative, and electronically-mediated workers in the Contingent Worker Supplement (CWS) of the Current Population Survey (CPS). BLS broadly defines 1) contingent workers as those with temporary employment; 2) alternative workers as those who are independent contractors, on-call workers, temporary help agency workers, and workers provided by a contract firm; and 3) electronically-mediated workers as those who obtain employment through mobile applications (apps) and websites that both connect workers with customers and arrange payment for tasks. Other researchers refer to these groups as gig employment. For this report, GAO combines these groups comprehensively as nontraditional employment.

BLS measured contingent and alternative workers five times from 1995 through 2005. After a 12-year hiatus, BLS replicated the 2005 survey in May 2017 and found a smaller percentage of contingent workers (3.8 percent) than in the 2005 survey (4.1 percent). Several companies that rely on electronically-mediated workers, or employment found through websites and mobile apps, such as drivers for a ride-share company, have been established since 2008.

To better measure electronically-mediated employment, BLS added four questions to the May 2017 CWS. However, according to BLS, the data from these questions required recoding, i.e. manually adjusting the data, to provide an accurate measure, but acknowledged that this is still an area for additional exploration. BLS officials stated that they are conducting additional research as summarized in the table below. 

Bureau of Labor Statistics (BLS) Reported Steps to Address Data Limitations Regarding Nontraditional Workers

Contingent Worker Supplement (CWS) areas for additional research Reported steps being taken
According to officials we interviewed and documentation we reviewed: 
CWS only measures workers’ main jobs for contingent and alternative employment, thus potentially missing some nontraditional workers.
CWS only asks respondents about their work in the past week and may not capture supplemental and occasional work.
BLS is currently working with stakeholders on CWS questions that will better capture nontraditional employment, including:
contracting with the Committee on National Statistics (CNSTAT) of the National Academies of Science, Engineering, and Medicine to evaluate the CWS and other measures, and consider the factors that could drive design of a new supplement; and 
planning, as part of the contract, a CNSTAT-hosted panel of experts, a public workshop, and a report with objectives for CWS’s measurements of contingent and alternative employment workers as well as insights on the electronically-mediated economy.
CWS’s added questions regarding electronically-mediated employment resulted in a large number of false positive answers, according to BLS officials.a BLS officials said after manually adjusting the data to correct for the false positives, the questions regarding electronically-mediated employment provide an accurate measure. BLS officials said they plan to perform additional research before administering the questions again.

Sources: GAO’s interviews of Bureau of Labor Statistics (BLS) officials and review of BLS documents.

a False positives are when people responded "yes" to the questions regarding electronically-mediated employment when the description of job duties and employers was incompatible with electronically-mediated work

Why GAO Did This Study

The nature of employment for many U.S. workers is changing. BLS data show millions of workers no longer hold traditional, full-time, year-round jobs. Since 2008, several companies have been established that rely on nontraditional employment arrangements with workers who typically use electronically-mediated platforms such as websites and apps to secure work. BLS data on nontraditional workers in the May 2017 CWS estimated that the percent of contingent workers declined amid a rise in electronically-mediated platforms for securing work. This led researchers, stakeholders, and policymakers to question whether this segment of the workforce is being accurately measured.

GAO was asked to review BLS's measures of nontraditional workers and their accuracy. This report describes (1) how BLS defines and measures contingent, alternative, and electronically-mediated employment workers in the CWS and other BLS data sets; and (2) what limitations of the CWS BLS has identified and what steps BLS has taken to address them.

To address these objectives, GAO reviewed technical documentation and reports, and interviewed officials.


GAO is not making recommendations in this report.

Full Report

GAO Contacts

Office of Public Affairs


Contingent workersData collectionEmploymentIndependent contractorsLabor forceLabor statisticsTemporary employmentTemporary workersWebsitesWorkersPublic health emergencies