The Organisation for Economic Co-operation and Development (OECD) has published guidance and experimental statistics on measuring platform-mediated work—paid activities organized through a digital intermediary that matches supply and demand. Staff Gauge reads that material as a set of staff readings that must be taken before a national count enters an employment or income frame. Without those readings, “platform work” collapses into a slogan that mixes employee jobs advertised online, true multi-sided marketplace tasks, and ordinary self-employment that merely uses a website.
Definitional gates
OECD measurement notes typically distinguish work obtained through digital labour platforms from broader online activity. Criteria often include that the platform intermediates a commercial transaction, that payment is mediated or strongly organized by the platform, and that the worker’s access to clients depends on the platform’s infrastructure. Borderline cases—software freelancers who find clients on a board but invoice outside it, or drivers coordinated by apps under employment contracts—illustrate why a single binary flag is unstable across countries.
Those gates matter for secondary earnings. Platform tasks frequently sit beside a primary job. A survey that asks only about the main job’s employer will understate platform hours and pay unless it adds explicit prompts. A survey that asks whether any platform work occurred in a long reference window may capture prevalence but not intensity. OECD-style guidance therefore stresses separating status (did platform work occur?) from volume (hours and earnings), and separating labour input from the consumer-side use of platforms.
Income statistics versus employment statistics
Employment frames classify people by labour-force status and job characteristics. Income frames classify flows: wages, mixed income, transfers. Platform receipts can appear as wages when the legal relationship is employment, as mixed income when the person is self-employed, or as poorly classified residual categories when respondents cannot map app payouts onto survey concepts. Experimental OECD and national-statistical-office pilots document that classification error is common when question modules are short.
Cross-country tables compound the problem. Legal tests for employee versus contractor differ; platform markets differ in sector mix (passenger transport versus goods delivery versus online freelancing); and survey fielding modes differ. OECD compilations that publish experimental indicators usually label them as such and warn against treating early estimates as a settled international league table. Staff Gauge repeats that warning: contested or experimental figures stay labeled in the same paragraph as the number.
Limits that stay visible
Platform firms’ administrative data can measure transactions on that platform but not work off-platform. Household surveys can miss rare or stigmatized activities and may confuse unpaid digital participation with paid tasks. Time-use modules improve hour estimates when fielded, yet they are expensive and infrequent. No single source currently provides a complete staff gauge of platform-mediated secondary earnings across OECD members.
Policy debates sometimes treat platform shares as evidence that households “need” side activity. That leap is not licensed by the measurement literature. Prevalence and hours are descriptive. They do not establish motives, hardship, or recommended conduct for any reader.
Editorial stance
Staff Gauge cites OECD platform-measurement guidance when a note needs an international definitional baseline. National estimates are attributed to the producing office and marked experimental when the source does so. The desk does not translate platform statistics into income instructions, product claims, or personal forecasts.
Readers should consult OECD publications on measuring platform work and the technical annexes of participating statistical offices. Corrections: [email protected].