Inside the new Periodic Labour Force Survey

Jyoti Thakur and Dhruv Pratap Singh

India's struggle with timely labour statistics dates back decades. The National Sample Survey Office's (NSSO's) Employment and Unemployment Surveys operated on thick and thin rounds, with comprehensive data collected only roughly every five years — a significant inadequacy for a large, fast-changing labour market. The Periodic Labour Force Survey (PLFS), launched in 2017, addressed this by providing annual estimates for rural and urban areas and quarterly estimates for urban areas, replacing the episodic framework with continuous measurement. However, PLFS itself retained limitations in frequency, geographical granularity, and international comparability. The 2025 redesign directly addresses these gaps.

The PLFS, revamped from January 2025, pursues three objectives: monthly estimates of key labour indicators — worker population ratio (WPR), labour force participation rate (LFPR), and unemployment rate — at the all-India level for both rural and urban areas; extended quarterly coverage to rural areas for the first time, enabling combined national estimates; and continued annual estimates in both usual and current weekly status on a calendar year basis.

Sampling methodology has been substantially strengthened. The first stage units (FSUs) surveyed have increased from 12,800 to 22,594, while households per FSU have risen from eight to twelve, yielding a total sample of approximately 270,472 households — a 2.65 times increase. Critically, the district now replaces the NSS region as the basic stratum, making PLFS estimates representative at the district level and significantly improving the granularity for sub-state labour market analysis.

The rotational panel design has been restructured significantly. Previously, urban households were visited once per quarter with no revisit provision for rural households. Under the new design, all selected households are visited four times across four consecutive months, ensuring 75% matching between consecutive months and 50% between successive quarters. The selection method has shifted from PPS with replacement to simple random sampling, with sub-stratification introduced in urban areas.

Beyond sampling, several additional items have been incorporated into the first visit schedule, including five education-related questions on class completed and years of schooling; a question on the nature of the vocational training certifying body; two questions on land possessed and leased out; and four questions on household income from rent, interest, pension, and remittances. These additions make the new PLFS schedule considerably richer — effectively combining labour force measurement with household income, asset, and education profiling in a single instrument.

Recognising that the July-June framework made India's labour statistics difficult to compare with international databases — all of which operate on calendar year cycles — and to provide a meaningful transition between the old and new series, MoSPI has released a back series of PLFS estimates on a calendar year basis. The back series begins in 2021 and not earlier for two reasons. First, 2020 was severely disrupted by Covid-19, making a clean, fully canvassed calendar year impossible. Second, the PLFS schedule was expanded around this period to incorporate a consumption block — household usual monthly consumer expenditure collected through five questions mirroring recent NSS rounds — making pre-2021 unit-level data structurally incompatible with later years.

A critical distinction must be noted. The calendar year back series through 2024 was constructed using the old sample design re-cut to a calendar year window; January 2025 marks the first data collected entirely under the new design. When multiple design elements change simultaneously — sample size, stratification, selection method, panel structure, and schedule content — differences in estimates may reflect genuine labour market shifts, seasonal effects, or design effects, making clean separation difficult. MoSPI itself has cautioned that post-January 2025 results cannot be straightforwardly compared with earlier publications.

India now has better labour market data than it has ever had. The revamp of PLFS comes at a particularly consequential moment — the four new labour codes covering wages, industrial relations, social security, and occupational safety have been brought into force in November 2025, reshaping hiring practices, formalisation incentives, and social protection coverage across the economy. Whether these reforms are delivering — whether formalisation is increasing, whether wage floors are binding, whether social security coverage is expanding — requires high-frequency, district-level data. India's five trillion-dollar ambition and the Viksit Bharat vision both rest on assumptions about who is working, in what conditions, and with what protection. For too long, the data arrived too late and at too aggregated a level to be of real policy use. That constraint has now been lifted. The question is whether policymakers and researchers will use what they now have with the rigour the moment demands.

(Jyoti is an associate fellow, and Dhruv is a research analyst at NCAER, New Delhi)

Disclaimer: The views expressed above are the author's own. They do not necessarily reflect the views of DH.