How the Gig Economy Is Measured in Labor Statistics

The gig economy—independent contractors, freelancers, platform workers, and other contingent arrangements—falls largely outside the standard unemployment rate because traditional surveys define employment as regular wage-and-salary work. Most gig workers appear as “employed” even if earnings are sporadic or supplemental, masking underemployment and income volatility.

Why the Standard Unemployment Rate Misses Gig Workers

The unemployment rate—the headline measure published monthly—counts only those without a job who are actively looking for one. Someone working 10 hours per week on a food-delivery platform while unable to find full-time work is counted as “employed,” not as underemployed or looking. Similarly, a freelancer with sporadic project income and significant gaps between jobs may report themselves as “self-employed,” removing them from the jobseeker count.

This omission is not accidental; it reflects how the Current Population Survey, the source of unemployment data, asks questions. The survey distinguishes between wage-and-salary employment (regular jobs with employers) and self-employment or contract work. Those in the latter category who want more work are often invisible because the survey’s primary question—“did you look for a job last week?"—assumes traditional employment as the baseline.

The deeper issue is conceptual. The gig economy blurs the line between employment and underemployment. A person driving for a rideshare platform may be officially “employed” even if they work fewer than 10 hours per week, cannot secure steady assignments, and rely on public benefits or a spouse’s income. Capturing this volatility requires measures beyond the simple employed/unemployed dichotomy.

Independent Contractors and Self-Employment Classification

The largest slice of the gig economy falls into census and labour categories that predate platform work: independent contractors and the self-employed. These have always been harder to measure than wage-and-salary workers. A self-employed consultant, plumber, or tradesperson might have stable, full-time income, but annual income can swing wildly. The CPS records them as employed whether they work 5 hours or 50 hours per week.

Platform workers—those using digital intermediaries like Uber, Airbnb, Upwork, or DoorDash—are typically classified as independent contractors, pushing them further from traditional employment data. The BLS estimates that roughly 10–15% of the US workforce engages in some form of gig or alternative work arrangement, but the exact proportion depends heavily on how “gig” is defined. Broader definitions (any non-traditional arrangement) yield higher shares; narrower ones (primary income from platform apps) yield lower shares.

One critical measurement challenge: a person can simultaneously be a wage-and-salary employee and a gig worker. Someone employed full-time as an accountant who drives for Uber on weekends is recorded primarily by their main job, with the gig income often absent from employment statistics. This makes it difficult to assess how reliant households are on gig income or how precarious total earnings are.

The Current Population Survey and Its Limitations

The CPS, conducted monthly by the Census Bureau for the BLS, surveys roughly 60,000 households and forms the backbone of employment statistics. It captures employment status, industry, occupation, and hours worked. For decades, it worked reasonably well for measuring traditional employment. But it was not designed for measuring contingent, episodic, or multi-platform work arrangements.

The CPS does ask about multiple jobs and reasons for part-time work (economic versus voluntary), which helps flag some underemployment. If someone reports working part-time “due to slack work” or “unable to find full-time work,” analysts can infer some degree of involuntary part-time employment. However, the CPS does not systematically ask about hours variation month-to-month, income stability, or access to benefits—all critical for understanding gig-economy precarity.

Moreover, the CPS interviews household members once per month. A gig worker with zero earnings in the survey week might be recorded as unemployed if they are looking, or as not in the labour force if they are not actively searching, even if they were working and earning in other weeks. This monthly snapshot misses the lived experience of gig workers whose activity is genuinely week-to-week.

Supplemental Measures: The Alternative Work Arrangements Survey

Recognizing these gaps, the BLS periodically conducted the Alternative Work Arrangements (AWA) survey, most recently in 2017. The AWA explicitly asks about temporary help work, contract work, on-call arrangements, and independent contracting. It found that approximately 16% of employed workers were in alternative arrangements (excluding those simultaneously employed elsewhere). This is substantially higher than earlier snapshots, reflecting the growth of platform economies and flexible staffing.

The AWA provides richer detail than the CPS on work stability, benefits access, and reasons for choosing alternative arrangements. It distinguishes between those who prefer flexibility and those forced into contingency by labour-market tightness. However, the AWA is infrequent—last conducted in 2017—and does not capture the explosive growth of platform-specific work (ride-hailing, food delivery, task marketplaces) that accelerated afterward.

Platform-Specific Data and Gig-Economy Surveys

Because traditional government surveys lag, understanding the gig economy increasingly relies on platform company data, third-party surveys, and academic research. Platforms themselves publish driver/worker counts and participation rates; academic studies use surveys of workers on specific platforms. The Bureau of Labor Statistics has begun incorporating gig-related questions into its regular surveys and has commissioned specialized research on platform work.

These sources paint a more textured picture: they reveal that gig income is often supplemental rather than primary, that earnings volatility is substantial, and that most platform workers also hold wage-and-salary jobs or rely on other income sources. They also show stark disparities in earning potential across platforms and regions, something invisible in aggregate employment numbers.

Underemployment and Income Volatility: What Is Missed

The standard unemployment rate captures joblessness but obscures underemployment and income volatility—both acute problems in gig markets. Someone working 15 hours per week on DoorDash and unable to secure additional work is not counted as unemployed, nor is there a standard measure of how many gig workers face this constraint.

The BLS publishes supplemental unemployment measures that include those “marginally attached” to the labour force (not actively seeking but wanting work), which captures some underutilized workers. But these measures still miss the gig worker working below desired hours or facing high income instability. A gig worker earning $15,000 in good months and $5,000 in slack months has a very different risk profile than someone with stable $10,000 monthly income, yet both might appear in the same employment category.

Benefits Access and the Gig-Economy Gap

One reason measurement of the gig economy matters beyond statistics: benefits eligibility depends on employment classification. Independent contractors do not qualify for employer-sponsored health insurance, 401(k) plans, unemployment insurance, or workers’ compensation in many states. They must self-fund retirement savings and manage their own benefits. This creates a wedge between headline employment growth and worker security.

If a rising share of employment growth is in gig and contingent arrangements without benefits access, aggregate employment can expand while worker economic security contracts. This possibility—visible in detailed gig-economy studies but not in headline unemployment data—has motivated calls for new measurement frameworks that separately track benefits-eligible work.

Policy Implications and Future Measurement

The gig economy’s measurement gaps have real policy consequences. When a city debates permitting more ride-hailing, or a state considers gig-worker classification, the analysis often relies on imperfect data. Labour-force participation rates, which have been puzzlingly stagnant in recent decades, might look different if gig work—particularly precarious, part-time gig work—were separately tracked and counted as pulling workers toward the margins rather than full attachment.

Going forward, the BLS and labour departments in other countries are expanding gig-specific data collection. The EU has mandated platform-work data; some US states require disclosure from ride-hailing and delivery platforms. These efforts aim to create real-time visibility into contingent-work participation, hours, and earnings—filling the gap that the CPS and AWA, operating on multi-year cycles, cannot fill alone.

See also