One term, many boundaries
“Gig economy” sounds like a single thing, but it is used for a wide set of arrangements. Depending on who is speaking, it may mean app-based delivery and ride-hailing, online freelance marketplaces, short-term contract work, on-call and zero-hours arrangements, self-employed professionals, or any paid activity outside a traditional permanent job.
Because there is no universally agreed boundary, a question like “how big is the gig economy?” cannot be answered with one number. Different definitions can yield very different answers from the same underlying reality, and often the disagreement between reports is a disagreement about what to count rather than about facts.
The main decisions that shape a count
Several choices decide how large the gig sector appears.
Platform-mediated or not. A narrow definition includes only work found or managed through a digital platform, such as an app that matches workers with tasks. A broad one includes any short-term, task-based or temporary work, whether or not an app is involved. The narrow version captures a relatively small, newer group. The broad version overlaps with long-established forms of casual, seasonal and freelance work.
Main income or extra income. For some people gig work is their main source of earnings. For others it is supplementary, as with someone who occasionally drives or does online tasks alongside another job. Counting everyone who did any gig activity gives a larger figure than counting those who rely on it. Extra-income activity also overlaps with what is described elsewhere as side hustles.
How often counts as often enough. Does a single task in the past year count? Regular work every week? A threshold for hours or earnings? Moving the threshold changes the total considerably, since many people do a little and few do a lot.
Employee or independent. Some platform workers are classified as independent contractors, others as employees, and the legal classification varies between places and has been contested. A definition based on employment status will include or exclude the same worker depending on how that is determined.
Time window. Asking about the last week, the last month or the last year each captures a different group. A longer window catches occasional participants and tends to produce a bigger number.
Workers or jobs. One person may do several gigs, and may hold a conventional job too. Counting workers, gigs, accounts or hours leads to different totals.
Where the data comes from
Different sources are suited to different definitions, and none sees the whole.
Official labour surveys. Regular household surveys classify people by main job and employment status. They may not ask specifically about platform or task-based work, and short-term activity beyond the main job can be missed. Some countries add special modules to capture it, with their own wording and reference periods.
Tax and administrative records. These show reported income and self-employment. They do not necessarily identify which income came from gigs, and informal or underreported earnings are likely to be missed.
Platform data. Companies know how many workers were active on their systems. That is accurate for that platform but covers only that platform. Accounts are not the same as individual people, since one person might have multiple accounts and some accounts might be inactive. Such platform data is also usually controlled by the companies, which choose what to publish.
Payment and transaction records. Data from payment providers can reveal flows of income, a different angle from surveys. They cover those using particular services, and categorisation of a payment as gig income involves judgement.
Independent and commissioned surveys. Surveys by research bodies, unions, think tanks and companies often use their own questions. Some use online panels, and survey panels leave out people who are not online or not on a panel. Sampling decisions affect who is represented, and different sampling frames lead to different pictures. Weighting can partly adjust for gaps but cannot remove them.
Why estimates disagree
Taken together, these choices explain why a report can say the sector is small while another says it is large. They may not be in conflict. One might count only those who earn their main income through an app in a given week; the other might count anyone who did any paid task in a year. Both could be accurately describing different things.
Trends are similarly tricky. A reported rise might reflect real growth, a new question being added, an improved way of detecting the activity, or a change in how people describe what they do. Comparing numbers across years is reliable only if the definition and method stayed the same.
Beyond the headline number
The size of the sector is just one question. Others often matter more for understanding, and they too depend on definitions:
- Who does the work and why, such as whether it is chosen for flexibility or taken for lack of alternatives. Surveys can ask about motives, but answers are shaped by wording and can differ for the same person at different times.
- What workers earn, which depends on whether time waiting between tasks and costs such as fuel or equipment are included. Reported hourly earnings can differ greatly depending on whether those are counted.
- How long people stay, which requires following people over time instead of taking a snapshot.
Reading a gig economy claim
A few practical questions help:
- What is being counted: people, jobs, hours or earnings?
- Is it limited to digital platforms or broader?
- Does it count main work, extra work or both?
- What time period and frequency threshold were used?
- Where did the data come from, and who is missing?
- If it describes growth, was the method consistent across years?
The aim is not to prove any figure wrong. It is to see that each number carries a set of counting rules, and that those rules are part of the finding. That perspective is also useful for the wider discussion of how claims about the future of work are put together.