How the ILO frames digital platform employment for statistics

The International Labour Organization has issued guidance aimed at national statistical offices that need to enumerate work organised through digital labour platforms. The documents do not replace a country’s labour-force survey. They set out definitions, boundaries and recommended distinctions so that offices can add questions or modules without collapsing platform-mediated work into unrelated categories, and so that totals published in one country can be compared more carefully with totals published in another.

Pay Census records that guidance as an instrument note: a statement of how an international statistical body asks offices to think when they design the form, not a single global headcount of platform workers.

What the guidance asks offices to separate

Digital platform employment, in the ILO framing, is work obtained and organised through a digital platform that intermediates between the person performing the work and the client or requester. The guidance presses offices to keep that intermediation visible in the questionnaire — to ask whether a platform played that role — rather than relying only on traditional status-in-employment categories that were built for other purposes.

Offices are also asked to distinguish forms of platform work that differ in location and task type, such as location-based services performed in physical space and online web-based tasks performed remotely. Mixing those cells into one undifferentiated “gig” total hides differences in hours, pay arrangements and regulatory exposure that labour statistics are meant to illuminate. The guidance treats those distinctions as measurement choices that belong on the form, not as optional colour for a press release.

Boundaries with related concepts

Not all remote work is platform work. Not all self-employment is platform-mediated. Not all contingent or temporary employment involves a digital intermediary. The ILO materials stress these boundaries because household surveys already classify people by employment status, contract type and occupation; adding platform items without clear definitions risks double-counting, undercounting or re-labelling ordinary employees who merely use software at work.

Reference periods and job units remain a national-survey design choice, but the guidance reminds offices that platform activity is often secondary, intermittent or multi-platform. A module that asks only about the main job in one week will miss much of what platform markets generate, even if the questions about intermediation are well worded. Conversely, a module that asks about any platform activity over a longer window will not be comparable to a main-job contingency share without explicit qualification.

What the guidelines are not

The ILO guidance is not a claim about how many people worldwide perform platform work, and it is not a recommendation that every country adopt the same questionnaire verbatim. It is a framework for definitions and boundaries. Implementation still depends on sample design, language testing, interviewer training and the parent survey’s existing concepts. Countries that follow the spirit of the guidance may still publish numbers that differ because their reference weeks, sampling frames or modules differ.

The guidance also does not settle legal classification disputes — whether a given platform relationship is employment or self-employment under national law. Statistical categories and legal categories can share vocabulary without sharing tests. The ILO materials are written for statistical measurement; courts and labour inspectorates apply different instruments.

Why the guidelines sit in this series

Household supplements such as the Contingent Worker Supplement and research programmes that compare surveys with administrative data answer questions inside particular national systems. The ILO guidelines sit one layer above: they describe how an office should name digital platform employment when it builds or revises an instrument. Without that vocabulary, cross-country tables and secondary write-ups tend to treat incompatible forms as if they measured the same object.

Read with the other forms on this site, the guidance reinforces a single editorial rule. Before comparing two counts of platform or online side income, name the definition, the unit, the reference period and the population each form reached. The ILO documents exist so that those names can be standardised enough to be useful — and kept distinct enough that the resulting totals are not mistaken for one another.