PIB TrackerIndia proposes to start counting unpaid care work, and gig work
From January 2027 the labour force survey will carry a separate stratum for digital platform workers and new questions on unpaid household, caregiving and volunteer work - with comments invited in fifteen days.
What happened
- The NSO proposes to modify the PLFS sampling design and schedule from January 2027, when fresh samples for 2027-2028 (Panel 6) will be drawn from an updated frame; a concept note is on the MoSPI website and comments are invited within 15 days.
- A separate Second Stage Stratum for digital platform workers is being introduced because PLFS 2026 did not capture an adequate number of them - four SSSs in rural areas and five in urban, with 14 households per First Stage Unit.
- The sampling frame is being updated for the revised Urban Frame Survey, urbanisation and de-urbanisation, and the creation, splitting or merger of districts, with follow-up updation of the rural Census 2011 frame.
- For the first three months of 2027 the sample will combine Panel 5 and Panel 6 frames with pooled monthly estimates, and relative standard errors will not be presented for those months.
- Schedule changes: a binary question in Block 5 of Schedule 0.0PL to identify households with digital platform work, and new Schedule 10.4 questions on own-use production of goods, unpaid household and caregiving activities, volunteer work and unpaid trainee work, for persons aged 15 and above on the first visit.
For Prelims
- PLFS began in 2017, replacing the five-yearly NSSO Employment and Unemployment Surveys. It reports the three headline ratios - Labour Force Participation Rate, Worker Population Ratio and Unemployment Rate - on two measures: Current Weekly Status (a seven-day reference) and usual status (a 365-day reference with principal and subsidiary activity).
- The 2025 revamp moved PLFS to monthly estimates with sub-state estimation; the sample is built as a two-stage stratified design - First Stage Units are census villages or urban blocks, Second Stage Strata group households within them so that rarer categories can be over-sampled.
- Why a stratum for platform workers: if a group is perhaps 1 per cent of households, a general sample finds too few of them for a reliable estimate. Creating a stratum and screening every household for membership is the standard fix, and the binary screening question is what makes it possible.
- Unpaid work in the statistical system: the System of National Accounts counts own-use production of goods (collecting firewood or water, subsistence farming) inside the production boundary but leaves unpaid domestic and care services outside it. The 19th and 20th ICLS resolutions define five forms of work including own-use production, volunteer and unpaid trainee work - the framework these new questions follow.
- India's existing measure of unpaid work is the Time Use Survey (2019 and 2024), which found women spending several hours a day on unpaid domestic and caregiving services against a fraction of that for men. Moving some of this into the PLFS makes it annual rather than occasional.
- Relative Standard Error is the standard error as a percentage of the estimate - the published measure of sampling reliability. Suppressing it for a transition quarter means the level can be read but its precision cannot.
- Gig and platform work in law: the Code on Social Security 2020 defines gig worker and platform worker and provides for aggregator contributions of 1-2 per cent of turnover, capped at 5 per cent of payments to such workers; e-Shram is the registration database. None of these produce a population estimate, which is why the survey change matters.
- The frame problem: rural sampling frames are built from the Census, and with the census due in 2027 the current rural frame still rests on 2011; the Urban Frame Survey is MoSPI's own rolling update of urban blocks, which is why urban coverage can be refreshed while rural cannot.
For UPSC: A methodology release that decides what India will know about its own labour market, and a rare chance to cite ICLS definitions against an Indian survey. Deploy it on employment data and its limitations, on the gig economy and social security, and on the valuation of unpaid care work.
What it is NOT: No estimate of how many digital platform workers the 2026 experiment actually found, which is the fact that would justify the redesign; no sample size, no cost, and no statement of whether the new unpaid-work questions will be published as a separate report or folded into the annual PLFS. Nothing on whether the unpaid-work data will be valued in money terms or only in participation and time, and nothing on the discontinuity the frame change will create - a series break is implied by pooling two frames, but the release does not say how it will be flagged in future releases. Fifteen days is also not justified anywhere.
For Mains
Syllabus: GS3.1 · GS3.2 · Linkage L1
Anchor
The National Statistics Office has proposed changes to the PLFS from January 2027: a separate second-stage stratum for digital platform workers, an updated sampling frame, a pooled-frame transition quarter, and new questions on own-use production, unpaid household and caregiving activities, volunteer work and unpaid trainee work - with comments invited in fifteen days.
Substantiation (data)
PLFS has run since 2017 and was revamped from January 2025 for monthly, quarterly and annual estimates with sub-state estimation; digital platform workers have been covered experimentally since January 2026, and the release states that arrangement did not capture an adequate number. The new design has four second-stage strata in rural areas and five in urban, with 14 households per First Stage Unit; the first three months of 2027 will pool Panel 5 and Panel 6 frames without publishing relative standard errors; and the unpaid-work questions will be canvassed on the first visit for persons aged 15 and above.
Position
This is the most consequential statistical decision of the year. A country that does not measure unpaid care work systematically cannot see why its female labour force participation looks the way it does, and one that cannot count platform workers cannot design social security for them - the Code on Social Security defines them without anyone knowing how many there are. Putting both into the regular labour survey, with a screening question in every household, converts two policy debates from anecdote into data.
Counterpoint
The foundations are shakier than the superstructure. The rural frame still rests on Census 2011, so in 2027 the survey will sample a country mapped fifteen years earlier - the single largest source of error in Indian household statistics, and one no amount of stratification fixes. Suppressing relative standard errors for a quarter removes the only published guide to reliability at exactly the moment the frame changes, and fifteen days is a short window to comment on a design that will govern the employment numbers for two years.
Way forward
Publish the platform-worker count from the 2026 experiment so the redesign can be judged, extend the comment period, and state plainly how the frame change will be flagged as a series break. On unpaid work, commit in advance to publishing participation rates and average time by sex, and to reconciling the PLFS numbers with the Time Use Survey rather than running two unconnected measures.
Conclusion
India is about to start counting two kinds of work it has never counted properly. The design deserves the fifteen days of scrutiny it has asked for, and more than fifteen days of it.
Deploys into: Employment data and its limitations · Gig and platform work and social security · Unpaid care work and women in the labour force · Statistical system and census frames
Ministry of Statistics & Programme Implementation · 2026-10-23 · PRID 2327153 · PIB source ↗