Nearly three in ten Indians are migrants by the government’s own definition. That is the finding of “Migration in India, 2020-2021,” released by the Ministry of Statistics and Programme Implementation (MoSPI) on 14 June 2022, based on the Periodic Labour Force Survey (PLFS) conducted from July 2020 to June 2021. It remains the most recent national picture of who moves, where, and why.
The survey was large by design: 12,562 first-stage units, made up of 6,930 villages and 5,632 urban blocks, covering 1,00,344 households (55,389 rural and 44,955 urban). Enumerators reached 4,10,818 individuals, of whom 1,13,998 were classified as migrants. Yet the fieldwork itself sat inside the pandemic.
PLFS operations were suspended in March 2020 and again in April 2021, a disruption the report discloses and one worth stating plainly rather than glossing over.
That caveat matters because there is nothing more recent to replace it with. A follow-up answer given in the Rajya Sabha by the Minister of State (Independent Charge) for Statistics and Programme Implementation, released by the Press Information Bureau on 2 February 2026, confirms that the most recent migration data available is still PLFS 2020-21 alongside the Multiple Indicator Survey of the same period.
The same release carries a piece of forward-looking news: the National Statistics Office has proposed a dedicated Survey on Migration, to run from July 2026 to June 2027. Until it reports, every claim about who India’s internal migrants are rests on a five-year-old snapshot taken while the country’s mobility itself was frozen and then restarted in fits and starts.
The country nobody has fully counted
India has never run a survey whose sole purpose is to measure migration nationally. What we have instead are estimates drawn from broader labour and consumption surveys, where migration is one module among many.
The PLFS 2020-21 round is the best of these so far, and its headline number is striking: a migration rate of 28.9 per cent for all persons, rural and urban combined, according to MoSPI’s June 2022 release (PIB, 14 June 2022).
Put plainly, almost one in three people counted in that survey were living somewhere other than where they had spent the previous six months or more of their life. That is a workforce, a school-going population, and a set of household budgets in motion on a scale that ordinary civic planning rarely accounts for.
A country that cannot say with confidence how many of its people moved this year, or where they went, cannot easily plan housing, water supply, ration distribution, or school seats for them either.
What “migrant” means in the data
The size of the number depends entirely on the definition behind it, so it is worth stating the three terms the report uses before reading any further figures.
- Usual Place of Residence (UPR): the village or town where a person has lived continuously for at least six months.
- Migrant: any household member whose last usual place of residence, at any point in their life, differs from where they are now enumerated.
- Migration rate: the share of migrants within a given category of persons, expressed as a percentage.
Under this definition, migration status is a lifetime marker, not a recent event. Someone who moved once, decades ago, and has lived in the same place ever since is still a migrant in this count. That single choice shapes everything that follows, starting with the gap between how men and women appear in the data.
| Category | Rural % | Urban % | Rural + Urban % |
|---|---|---|---|
| Male | 5.9 | 22.5 | 10.7 |
| Female | 48.0 | 47.8 | 47.9 |
| All persons | 26.5 | 34.9 | 28.9 |
Female migration runs at 47.9 per cent against 10.7 per cent for men, MoSPI’s PLFS 2020-21 data shows, a gap so wide that it demands an explanation before any other conclusion is drawn from the headline rate.
Marriage, not employment, drives the headline number
The explanation is straightforward once the reasons for migration are laid out. Marriage accounts for 86.8 per cent of female migration recorded in PLFS 2020-21, against just 6.2 per cent of male migration, according to MoSPI.
A woman who relocates to her husband’s village or town after marriage is, by the survey’s own rules, a migrant. In a country where marriage overwhelmingly still means the bride moving, that single life event explains most of the female migration rate on its own.
This is not a flaw in the survey. It is an accurate reflection of a real social pattern, correctly counted. But it means the 47.9 per cent female migration rate, and by extension much of the overall 28.9 per cent figure, is not a measure of women entering the labour market by moving. It is largely a measure of where women live after marriage, which is a different question with different policy implications.
That distinction matters when the same figure gets read alongside women’s actual presence in paid work. Nearly half of Indian women are technically migrants, yet India’s female labour force participation trails even Pakistan’s, a gap showing that movement and employment are not the same thing for women in the way they often are for men.
A high migration rate for women coexists with a low work-participation rate. The two numbers are measuring different parts of the same lives.
For men, the picture inverts. Marriage accounts for only 6.2 per cent of male migration. The largest single reason for men is the search for employment or better employment, at 22.8 per cent, according to the same MoSPI release, confirmed again in the 2 February 2026 Rajya Sabha answer (PIB, 2 February 2026).
Marriage moves women. Work moves men. Both are counted in the same national migration rate, and conflating them flattens two very different stories into one number.
The minority that moves for work
Once marriage migration is set aside, labour migration turns out to be a minority of all migration recorded in PLFS 2020-21, even though it is the part that keeps construction sites, factories, kitchens, delivery networks and farms running. The full breakdown by reason, split by sex, makes this clear.
| Reason | Male % | Female % | All persons % |
|---|---|---|---|
| In search of employment or better employment | 22.8 | 0.6 | 4.8 |
| For employment or work (job, business, transfer, proximity) | 20.1 | 0.7 | 4.4 |
| Loss of job, closure of unit, lack of opportunities | 6.7 | 0.4 | 1.6 |
| Migration of a parent or earning member of family | 17.5 | 7.3 | 9.2 |
| To pursue studies | 4.7 | 0.6 | 1.4 |
| Marriage | 6.2 | 86.8 | 71.6 |
| Housing problems | 4.8 | 0.8 | 1.5 |
| Natural disaster (drought, flood, tsunami) | 0.6 | 0.1 | 0.2 |
Taken together, the three explicitly employment-related rows in the report, search for employment, employment or work, and job loss, add up to 10.8 per cent of all migrants. That is our own arithmetic on MoSPI’s published rows.
That is the closest the PLFS gets to a direct headcount of India’s labour migrants inside the broader 28.9 per cent migration rate, and it is a fraction of the total, even though it is the group whose movement is most tightly bound to the economy’s day-to-day functioning.
The employment-search rate among men, 22.8 per cent, is also a quiet signal about the domestic job market itself. Movement in search of work, rather than movement into a confirmed job, implies a search process, which in turn implies scarcity at the point of origin.
That scarcity shows up elsewhere too. The same competitive pressure that pushes a young graduate to move cities looking for any opening is visible in accounts of PhD holders applying for peon posts, where credentials have outpaced the number of matching jobs in one place, and migration becomes one of the few available responses.
None of this means labour migrants are a small or marginal group in absolute terms. It means that when policy discussions use the headline 28.9 per cent figure to talk about migrant workers, they are describing a population that is mostly something else: mostly marriage-linked movement, and movement tied to a family’s primary earner relocating, which is 9.2 per cent of all migrants per MoSPI.
The workers who move for jobs are a distinct, smaller and specifically economic subset within that larger number.
Where India’s migrants come from
Origins matter as much as reasons. Among all migrants recorded in PLFS 2020-21, 73.4 per cent had last lived in a rural area, 25.9 per cent in an urban area, and 0.7 per cent had come from another country, according to MoSPI’s June 2022 report. India’s migration story is still overwhelmingly a movement out of rural places, whatever the eventual destination.
| Category | From rural areas % | From urban areas % | From other countries % |
|---|---|---|---|
| All persons | 73.4 | 25.9 | 0.7 |
| Male | 50.0 | 47.0 | 2.9 |
| Female | 78.8 | 21.0 | 0.2 |
The male and female columns diverge in an instructive way. Men’s origins split almost evenly between rural (50.0 per cent) and urban (47.0 per cent) areas, with a small but notable 2.9 per cent arriving from other countries, MoSPI’s data shows.
Women’s origins remain heavily rural, at 78.8 per cent, which is consistent with the marriage pattern already described: rural brides moving to another rural or urban household, rather than urban-to-urban relocation driven by careers.
For men, the near-even rural-urban split hints at a two-stage pattern the PLFS snapshot cannot fully resolve on its own. Some men move directly from village to city for work, while others move between cities, or from smaller towns to larger ones, chasing better-paying versions of jobs they may already hold. Both patterns are folded into the same national figure.
Short hops, not long hauls
The popular image of Indian labour migration, a worker crossing several state borders to reach a distant city, describes a real but minority pattern. According to PLFS 2020-21, 87.5 per cent of all migrants moved within the same state where they were last enumerated, MoSPI reports. Only 11.8 per cent crossed a state line, and 0.7 per cent arrived from another country.
| Category | Same state % | Another state % | Another country % |
|---|---|---|---|
| All persons | 87.5 | 11.8 | 0.7 |
| Male | 65.6 | 31.4 | 2.9 |
| Female | 92.6 | 7.2 | 0.2 |
The gap by sex is sharp. Inter-state movement runs at 31.4 per cent for men against 7.2 per cent for women, and among rural men specifically it climbs to 33.7 per cent, with 62.5 per cent staying within the same state, MoSPI’s report shows. Long-distance, cross-state migration is concentrated almost entirely among working-age men, exactly the group most likely to be moving for a job rather than a marriage.
It is also the group least visible to local administration. A man enumerated in one city whose voting registration, ration card and family records sit in a village several states away exists, in bureaucratic terms, in two places at once, and is fully served by neither.
When that same worker reaches a city, he often settles in the densest, cheapest available housing near the work itself. Life in Dharavi, India’s largest slum, is one well-documented example of the kind of dense urban settlement where migrant workers cluster because it is close to employment and within reach of their wages.
The PLFS 2020-21 round also captured a smaller and more transient group: temporary visitors, defined as people who arrived after March 2020 and stayed 15 days or more but less than six months, away from their usual place of residence.
This group made up 0.7 per cent of the population overall, 0.8 per cent for men and 0.5 per cent for women, MoSPI reports. It is a reminder that beyond the migrants formally counted, there is a further layer of short-term, circulating movement the survey only partially captures.
What gets left behind
Every migration decision has a mirror image: something left behind. The PLFS does not measure what migrants leave, only where they arrived from and why they moved, so this section is necessarily qualitative rather than another set of figures.
The view from the origin side is its own subject, and we have written about it separately in why India’s villages are emptying. This piece looks at the national numbers; that one looks at what the departure leaves behind in the districts people leave from.
For a rural household, the asset most commonly left behind is land, and the security of that land claim shapes both the decision to leave and the ability to return. Where ownership is documented, contested, or simply unclear on paper, even though the family has farmed it for generations, leaving becomes easier in one sense and harder in another.
It is easier because there is little formal paperwork tying the household physically to the place. It is harder because an insecure claim left unattended for years is more vulnerable to encroachment, dispute, or informal reallocation by the time anyone tries to return to it.
This is where India’s long-running land records and property dispute problem connects directly to migration patterns, even though the two are rarely discussed together. A worker with a clean, digitised, unambiguous land title has an asset that survives their absence and a place to return to if the city job disappears.
A worker whose family’s claim rests on decades of informal possession and inconsistent revenue records is migrating away from an asset that may not be there, in any enforceable sense, when they come back.
None of this means migrants are abandoning secure lives for insecure ones out of desperation. The PLFS data on reasons for migration shows people moving toward something, mainly employment for men and marriage for women, rather than simply fleeing. But the condition of what is left behind still shapes how reversible the move is, and land tenure is the clearest example of a structural gap sitting entirely outside the migration survey itself.
Entitlements that do not travel
The deeper structural problem is that most of India’s welfare and service entitlements are tied to a place, not to a person. A ration card is registered against a specific fair price shop. School admission usually depends on catchment or residence proof. Health scheme empanelment and enrolment records sit with a specific facility or district.
Voter rolls and many welfare registrations are anchored to a home address that a labour migrant, by definition, has left.
The consequence is not that migrants lose these entitlements outright. It is that claiming them requires either travelling back to the place of registration, which working migrants often cannot afford to do repeatedly, or navigating a new local system from scratch in the destination city, often without the paperwork that system expects.
Health is a clear example. A family’s out-of-pocket medical spending does not pause because a wage-earner has moved for work, and India’s pattern of out-of-pocket health expenditure falling and then rising again shows how exposed households remain when a scheme registered in one district offers no straightforward path to care in another.
This is not a story about any single ministry or scheme failing. It is a structural mismatch. The country’s major entitlement systems were built around a stable, place-based population, and a meaningful share of that population does not stay in one place for a working lifetime. A system with no route for portability will always under-serve movers, however generous the entitlement is on paper for people who stay put.
e-Shram: a database, not yet a guarantee
The most concrete attempt to address this gap so far is e-Shram, the National Database of Unorganised Workers run by the Ministry of Labour and Employment. It launched on 26 August 2021, is Aadhaar-seeded, and gives every worker who self-registers a Universal Account Number, according to a Press Information Bureau release dated 10 March 2025 (PIB, 10 March 2025).
The scale of registration is large. As on 3 March 2025, more than 30.68 crore unorganised workers had registered on the platform, of whom 53.68 per cent were women, the same release states. That is a database which, in principle, is not tied to any single district or state, making it structurally different from a ration card or a local welfare list.
A worker who registers once can, in theory, be found and reached regardless of where they subsequently move for work.
Thirteen schemes run by central ministries have been integrated or mapped with e-Shram, the Ministry of Labour and Employment reports. These include:
- PM-SVANidhi, for street vendor credit
- Pradhan Mantri Suraksha Bima Yojana (PMSBY), accident insurance
- Pradhan Mantri Jeevan Jyoti Bima Yojana (PMJJBY), life insurance
- The National Family Benefit Scheme
- MGNREGA, the rural employment guarantee
- PMAY-Gramin, the rural housing scheme
- Ayushman Bharat PM-JAY, health coverage
To make these more usable, an e-Shram One-Stop-Solution was launched on 21 October 2024, letting a registered worker see and access all mapped schemes from a single point rather than approaching each ministry separately, the March 2025 release notes.
It is fair to call this a genuine step and equally fair to be precise about what it is not. Registration on e-Shram establishes that a worker exists in a national system and is theoretically eligible for the schemes mapped to it. It does not, on its own, guarantee that a specific worker receives a specific payment, ration or hospital admission when they need it in a city far from home.
A database is the precondition for portability. It is not the same thing as portability delivered.
Measuring what has never been measured
A state cannot plan housing stock, water supply, school capacity or health infrastructure for a population it has never properly counted at the place where that population actually lives and works. This is the practical cost of relying on a snapshot embedded inside a labour survey rather than a purpose-built migration count.
The proposed Survey on Migration, to run from July 2026 to June 2027 as confirmed by MoSPI’s 2 February 2026 release, is designed to close several of the specific gaps PLFS 2020-21 could not address. According to the release, it is expected to cover:
- The extent and patterns of rural-urban and inter-state migration
- Reasons for migration, examined in more depth than a single-question module allows
- Return migration, meaning workers who move back to their place of origin, a flow the PLFS round did not separately track
- Seasonal migration, the pattern of workers moving for part of the year and returning for agricultural or festival seasons
- Employment profiles of migrants, connecting movement data directly to the kind of work migrants actually do at their destination
Each of these is a real blind spot in the current picture. Return migration and seasonal migration in particular describe circulating patterns that a single point-in-time survey, however large, cannot capture, because a worker who has already gone home by the time enumerators arrive is invisible to that round entirely.
A dedicated, purpose-built survey is the right instrument for a phenomenon a general labour survey can only glimpse in passing.
What portability would look like
None of the verified figures in this account point to a single villain or a single fix. They point to a mismatch between how India’s population actually moves and how its systems of counting and entitlement were built. Closing that mismatch is a matter of design choices, and a few of those choices are already visible in outline.
- Make core entitlements portable by default. A ration card, a health scheme registration or a school transfer record should not require a worker to return to their origin district to remain valid.
- Count people where they actually live and work, not only at their registered address, so destination cities can plan housing, water and school capacity for the population they actually host.
- Build seasonal and return migration into routine data collection, rather than treating every worker who leaves a city as having exited the workforce count entirely.
- Treat labour migration as ordinary economic behaviour, the predictable response of workers to uneven opportunity across regions, rather than as a crisis requiring emergency management each time it becomes visible.
The 28.9 per cent migration rate recorded in PLFS 2020-21 will likely look different once the proposed 2026-27 survey reports, if only because the instrument is finally built for the question. What will not change is the underlying fact the numbers already show.
Close to thirty per cent of a billion-plus population moving is not an anomaly to be managed. It is a baseline feature of how the country’s labour, marriages and households actually work, and the systems built to serve people would do better to plan for it as such, rather than being surprised by it every time it shows up in a new survey.
Photo by Dhruv Bhatt on Unsplash.