Examine the challenges to the comparability and reliability of India's labour statistics after the 2025 changes to the PLFS methodology.

The National Statistics Office (NSO) under MoSPI runs the Periodic Labour Force Survey (PLFS). It revamped the survey's sample design in January 2025 so that it could release monthly all-India estimates using the Current Weekly Status (CWS) [1]. The data now comes out more often, but this has created new problems in comparing and trusting it.

Challenges to comparability

• Break in the series: the design changed. The district is now the basic stratum, each sample block has 12 households, and the sample is about 2.65 times larger. MoSPI therefore advises caution when comparing results from 2025 onwards with estimates up to December 2024 [1].
• Change in reporting period: annual reports used to cover July–June and now cover the calendar year. As a result, the seven earlier annual reports do not line up neatly with the new series [1].
• Short series: the monthly bulletins began in April 2025. September 2026 is only the 18th bulletin [2], so only two Septembers can be compared and there is no seasonally adjusted series yet.
• Different reference periods: monthly CWS uses the last 7 days, while annual usual status (ps+ss) uses the last 365 days [1]. Mixing the two leads to wrong conclusions.

Challenges to reliability

• Seasonal swings: MoSPI linked the fall in rural female WPR in June 2025 to fewer unpaid farm helpers and to summer heat [3]. In September 2026, rural female LFPR fell 0.5 pp from the previous month but rose 1.0 pp from a year earlier [2].
• No view of job quality: the bulletin reports only LFPR, WPR and UR [2]. Unpaid family helpers are counted as workers, so a rise in paid jobs and a rise in unpaid farm help look the same in the data.
• Old sampling frame: villages are still drawn from Census 2011, along with Urban Frame Survey blocks [1]. Areas that have urbanised since 2011 may be under-represented.
• Thin monthly samples: a single month's sample is much smaller than a full year's, so small changes may be noise. For example, urban UR moved from 6.8% to 6.9% [2].

Way forward

• Publish a bridge series that links estimates from before and after 2025.
• Release the share of unpaid helpers every month alongside WPR [3], and add seasonally adjusted figures once about three years of data exist.
• Update the sampling frame after the next Census and publish standard errors with every bulletin.

The 2025 revamp has made the PLFS larger, more frequent and more detailed. However, the breaks in design, reporting period and sampling frame mean that single months must be read with care. With bridge series, more openness about unpaid work and a fresh sampling frame, the PLFS can become a reliable basis for evidence-based employment policy and for tracking SDG 8 (decent work).