Chapter 1: The Numbers and the Names

---

On May 1, 2020, the first Shramik Special train pulled out of a Gujarat railway station carrying migrant workers back to their home states. Within days, the Odisha government had set up a portal for stranded migrants to register for return. By June, the trains were running round the clock. By July 7, 2020, exactly 3,58,401 people had arrived in Odisha by Shramik Special train alone. Another 1,85,504 had come by bus, private vehicle, and on foot. On the state registration portal, approximately ten lakh people — one million — had entered their names, declaring they wanted to come home.

The number that should unsettle you is not the ten lakh. It is the gap between ten lakh and everything the government thought it knew about how many had left.

The Census of India, 2011 — the last census conducted, since the 2021 exercise was postponed due to the pandemic and has not been completed as of 2026 — recorded 12,71,121 interstate out-migrants from Odisha. That is the official number. It sits in government documents, gets cited in policy papers, and forms the statistical basis for planning. The census is now fifteen years old. When one million people registered on a single emergency portal in a matter of weeks, and 36% of returning train passengers came from Gujarat alone — a state that according to the census accounted for 18% of Odisha’s out-migrants, roughly 2.3 lakh — the distance between the official record and reality became impossible to ignore.

The honest answer to “how many Odias live outside Odisha” is: nobody knows. The estimates range from two million to five million. The range itself is the finding. A state of forty-six million people cannot say, within a factor of two, how many of its citizens have left.

This chapter is about the numbers — what we know, what we estimate, what we cannot count, and why the inability to count matters. It is also about the names: which districts send the most people, where they go, what kind of work they do when they get there, and why they left in the first place. Before examining the human cost of migration, or the economics of remittances, or the identity questions of diaspora, the first task is simpler and harder: establish the scale.

---

What the Census Says

The 2011 Census remains the only comprehensive dataset on migration in India. Its migration tables record place of last residence, reason for migration, and duration. For Odisha, the headline number is 12.71 lakh interstate out-migrants. To put that in context: Odisha’s total population in 2011 was 4.197 crore. The interstate out-migrants represent about 3% of the population.

But the census captures a specific kind of movement. It records where a person was enumerated versus where they were born or previously resided. It captures people who have moved and settled elsewhere for a period long enough to be counted at the new location. What it misses — systematically, by design — is the vast category of seasonal, circular, and short-term migration. The brick kiln worker from Balangir who spends six months in Andhra Pradesh and six months at home is, in a census year, either at home (and therefore not a migrant) or at the kiln (and possibly uncounted because of the transient nature of the housing, the informality of the arrangement, the fact that nobody at a kiln site is filling out census forms). The powerloom worker in Surat who has been there for a decade but returns to Ganjam for festivals may or may not be captured, depending on when the enumerator visits.

The census is a photograph, not a film. Migration from Odisha is a moving target.

Still, even the frozen photograph reveals a structure. Of those 12.71 lakh interstate out-migrants, 58% of males cited “work and employment” as the reason for moving — more than double the all-India average of 24%. Female work-related out-migration was 6.11%, triple the national average of 2%. The net interstate out-migration was 2.8 lakh — meaning significantly more people left than arrived. And Odisha, along with West Bengal, Bihar, and Jharkhand, accounted for 24.4% of all interstate migrants in India. The eastern states were, and remain, the labor reservoir.

---

What the Surveys Add

The next layer comes from the National Sample Survey and the Periodic Labour Force Survey. The PLFS 2020-21 estimated approximately 8.51 lakh laborers from Odisha migrating annually for work. This is a flow number, not a stock number — it measures annual movement, not total population living outside. It also likely undercounts, because the PLFS relies on household surveys. If the entire household has migrated, there is no one at the source to report it.

The Centre for Migration and Labour Solutions, which has tracked seasonal laborers from Odisha since 2008, recorded a steady increase in the numbers it could identify: 87,000 in 2008, rising to 1.05 lakh in 2012, 1.35 lakh in 2014, 1.45 lakh the year after. These are tracked numbers — workers who appeared on an NGO’s radar. The organization estimates 3 to 5 lakh laborers migrate annually to different states, and this is explicitly described as conservative, excluding semi-permanent and professional migration.

The most recent and rigorous study is the Odisha Migration Survey, conducted in 2023 by researchers from IIT Hyderabad, based on 15,000 representative households across all 30 districts. The OMS estimated 1.7 million interstate migrants originating from Odisha. This is significantly higher than the census figure but still based on household surveys, which inherently undercount households that have migrated entirely.

And then there is the COVID evidence. When the crisis hit, approximately ten lakh people registered on a single portal saying they wanted to return. This is widely considered a massive undercount itself — many migrants lack smartphones, internet access, or awareness of the portal. The 5,43,905 who arrived by various transport modes by mid-June 2020, and the 8,53,777 who returned from just Gujarat, Tamil Nadu, and Kerala during April-September 2020, suggest a population in the millions, not hundreds of thousands.

The most honest aggregation of available evidence points to at least 20-25 lakh Odias — two to two-and-a-half million — living outside Odisha at any given time. Some researchers put the figure higher. The Odisha Society of the Americas, the Wikipedia article on the Odia diaspora, various journalistic accounts, and NGO estimates all circle around a range of two to five million. The range is embarrassingly wide. For comparison, Kerala — which has a systematic migration survey infrastructure in place since 1998 — can tell you that 2.12 million Keralaites live abroad, constituting 6% of its population and 17-18% of its workforce, and that remittances constitute 22-28% of state GDP. Odisha cannot produce any equivalent figure. The state does not know how many of its people have left.

---

Where They Go

The Census 2011 destination data shows a clear geography of Odia migration:

Gujarat took the largest share at 18.09% of interstate out-migrants — approximately 2.3 lakh. Undivided Andhra Pradesh was second at 13.89%. Chhattisgarh third at 12.93%. Maharashtra fourth at 11.43%. Karnataka at 8.29%. Jharkhand at 7.41%. West Bengal at 7.30%. Tamil Nadu and Delhi NCR were significant but not separately ranked in the top tier.

These census proportions, frozen in 2011, are certainly outdated. The COVID-19 Shramik Special data, while not a census, provides a more recent proxy. Of the train returnees to Odisha, 36% came from Gujarat (1,30,537 people), 17% from Telangana (roughly 60,000), 15% from Tamil Nadu (roughly 54,000), and the remaining 32% from other states. Gujarat’s dominance had doubled from its census share. Telangana, which did not exist as a separate state in 2011, emerged as a major destination. The map had shifted.

But the most revealing feature of Odia migration is not the state-level destination. It is the district-to-destination corridor — the specific pipelines that connect particular source communities to particular industries in particular cities.

---

The Corridors

Think of migration not as a diffuse flow but as a network of pipes. Each pipe has a specific origin, a specific destination, a specific industry, and a specific community that runs through it. The pipes were built over decades. Some are eighty years old. They are maintained by kinship, caste networks, labor contractors, and the accumulated knowledge of thousands of previous journeys. To understand Odia migration, you have to understand these pipes individually.

Ganjam to Surat. This is the largest single corridor. An estimated seven to eight lakh Odia workers — 700,000 to 800,000 people — live and work in Surat, primarily in the powerloom textile industry. Eighty percent of them are from Ganjam district. The corridor is at least eighty years old, originating in the 1940s when landless Dalit workers from Ganjam first migrated to rice mills in Myanmar and jute mills in West Bengal. When Surat’s textile industry expanded in the 1970s and 1980s — absorbing the powerloom capacity that was leaving Mumbai — workers from Ganjam followed the looms. The Ganjam-to-Surat pipeline is not seasonal. It is semi-permanent. Workers stay for years, sometimes decades. They live in designated Odia neighborhoods — Ved Road, Limbayat, Udhna, Pandesara. At Udhna junction, the suburban railway station, platform announcements are made in Odia alongside Hindi, English, and Gujarati. These workers produce roughly 90% of the polyester used in India. The corridor is so entrenched that it operates as infrastructure, not a choice.

Balangir cluster to Andhra Pradesh and Telangana brick kilns. This is the dadan corridor — the debt-bonded seasonal migration that has been documented, condemned, and legislated against for fifty years without being stopped. From Balangir, Nuapada, Kalahandi, Bargarh, Sonepur, and Boudh, an estimated 60,000 or more families — roughly two lakh individuals — migrate annually to brick kilns in Andhra, Telangana, Tamil Nadu, and Karnataka. Another 40,000 to 50,000 move to brick kilns closer to home, near Cuttack and Bhubaneswar. The migration follows a precise annual cycle tied to the monsoon and the harvest festival of Nuakhai. In August-September, labor contractors — sardars — visit villages and offer cash advances of Rs 35,000 to 60,000 per person. By October-November, the workers have boarded trains at Kantabanji or Khariar Road railway stations. They spend six months at the kilns. They return, if they return, in May-June, often still in debt, requiring another advance for the next cycle. The numbers tracked by the government — 70,142 workers in 2024, rising sharply to 94,106 in 2025 — capture only a fraction. The government knows this.

Kendrapada to the Gulf and Indian metros — the plumber corridor. Approximately one lakh people from Kendrapada district, specifically from villages in Pattamundai, Aul, Rajkanika, and Rajnagar blocks, work as plumbers across India and in Gulf countries — Oman, Qatar, Kuwait, Saudi Arabia. This is a specialized skill corridor. Kendrapada is home to the State Institute of Plumbing Technology, the only plumbing-dedicated institute in India. Almost every household in these blocks has at least one member in the plumbing trade. Workers from Kendrapada built the sanitary systems for the 2022 FIFA World Cup stadiums in Qatar — roughly 2,500 plumbers from the district worked on that project. They worked on India’s new Parliament building. They earn Rs 50,000 to one lakh per month in Gulf countries, Rs 30,000 to one lakh in Indian cities. The plumber corridor is neither dadan bondage nor professional brain drain. It sits in between — skilled, organized through informal networks, economically significant, and almost entirely undocumented in official statistics.

Coastal and southern Odisha to Kerala. A growing corridor, driven by a specific dynamic: caste discrimination in Surat. Dalit workers from Ganjam who experience caste-based exclusion in Surat’s Odia-run mess halls and loom assignments are increasingly migrating to Kerala instead. Kerala offers higher wages — it has among the highest minimum wages in India — and, according to workers who have made the switch, lower caste discrimination. Kerala’s own educated population migrates to the Gulf, creating domestic labor demand. The Ganjam-to-Kerala corridor is newer but expanding, driven not just by economics but by the desire to escape the caste hierarchy that Ganjam’s own communities have exported to Gujarat.

Various districts to Bangalore, Hyderabad, Pune — the IT and professional corridor. An estimated six lakh Odias live in Bangalore alone, a figure from 2016 that has certainly grown. The pathway is well-defined: engineering college (NIT Rourkela, KIIT, VSSUT, IIT Bhubaneswar) to campus placement to software company. NIT Rourkela’s 2024-25 placement data tells the story clearly: 1,274 job offers, highest package Rs 62.44 lakh per annum, 373 recruiters including Google, Amazon, Qualcomm, Texas Instruments. Virtually all high-package placements are in Bangalore, Hyderabad, Pune, or Mumbai. The institute trains talent that overwhelmingly leaves Odisha. Similar patterns hold at KIIT (5,585 offers, top recruiters Amazon, McKinsey, Salesforce), VSSUT Sambalpur, and IIT Bhubaneswar (90% placement rate, highest package Rs 67.6 lakh). Over 200 IT firms now operate in Bhubaneswar — TCS, Infosys, Wipro, Tech Mahindra, Capgemini — but the specialized, high-paying roles that ambitious engineers seek remain concentrated in the established tech cities.

---

The Source Districts

Not all of Odisha sends people away at the same rate. Migration concentrates in specific districts, driven by specific failures.

The state government identified eleven of its thirty districts as migration-prone in 2014. By 2024, the list had expanded to fourteen. The epicenter is western Odisha — what is known as the KBK region, after the historically underdeveloped Kalahandi, Balangir, and Koraput districts. But KBK is shorthand for a wider belt of chronic deprivation: add Nuapada, Sonepur (Subarnapur), Bargarh, Malkangiri, Nabarangpur, Rayagada, Gajapati, and the picture of the migration-prone zone becomes clearer. It overlaps almost exactly with the poverty map.

Ganjam sends the most people in absolute terms. It is the source of the Surat powerloom corridor, contributing six to seven lakh workers to Gujarat alone. But Ganjam’s migration includes an international dimension — Gulf countries, merchant marine — and a growing Kerala corridor. It is also the district that recorded the highest number of migrant worker deaths over the past decade: 59 out of 403 documented deaths, according to the state Labour Minister’s statement in the Assembly.

Balangir is the heart of dadan migration. Kantabanji, in Balangir district, functions as the largest informal labor market in western Odisha. In the village of Sargul in Muribahal block, 90% of the roughly 2,300 residents are classified as dadan workers. In Ichhapada panchayat, 80% are dadan workers. The government tracked 65,309 migrants from Balangir and Bargarh alone in 2025. In rescue operations, Balangir consistently tops the list: 626 workers rescued in 2024, 572 in 2025.

Nuapada ranks among the poorest districts in the state. It feeds the same brick kiln corridor as Balangir. It is small enough that the annual exodus empties it visibly.

Kalahandi — once internationally infamous for reports of starvation deaths — sends workers to Chhattisgarh for construction and agricultural labor in addition to the southern brick kilns. Seventy-seven percent of adults still depend on agriculture in a district where irrigation barely exists.

Koraput, Malkangiri, Rayagada, Nabarangpur — the tribal districts of southern Odisha. Nabarangpur has the highest multidimensional poverty rate in the state at 59.32%, meaning nearly six out of every ten people are multidimensionally poor. Malkangiri is second at 58.71%. These districts send workers to construction sites in Kerala, Tamil Nadu, and Andhra Pradesh, and to brick kilns. The migration from these districts is almost entirely distress-driven. There is no eighty-year pipeline here, no accumulated skill network. There is poverty, and there is a contractor offering cash.

Kendrapada is a different case — not poverty-driven in the same way, but corridor-driven. The plumbing specialization has created a self-sustaining migration system. Almost every household participates. The economics are good. But the district still loses its working-age population.

The pattern that emerges is not random. The districts with the highest migration rates are, with the exception of Ganjam (which combines poverty-driven and corridor-driven migration) and Kendrapada (skill-corridor driven), the districts with the worst irrigation, the highest poverty, the most fragile agriculture, and the least industrial development. Migration maps onto deprivation almost perfectly. The map of who leaves is a map of what was never built.

---

The Three Streams

This is the point where a single word — “migration” — becomes misleading. Odia migration is not one phenomenon. It is at least three, operating simultaneously, with different populations, different economics, different social structures, and different outcomes. Treating them as one thing produces bad analysis and worse policy.

Stream One: Dadan and seasonal migration. The participants are rural poor — predominantly from the KBK region and surrounding districts. They are overwhelmingly Dalit and Adivasi. They are recruited by sardars with cash advances that function as debt bondage. They work in brick kilns, construction sites, and agricultural labor. Working conditions are frequently illegal — 12 to 16 hour days, no rest days, confiscated identity documents, violence for non-compliance, wages withheld or manipulated through opaque accounting. The annual cycle is precise: advance in August-September, departure in October-November, six months of labor, return (if the debt is cleared) in May-June. Entire families migrate, including children. More than 20% of workers rescued from brick kilns between 2011 and 2013 were children. The economics are survival: the advance money pays for Nuakhai celebrations, repays moneylenders, sustains the family through the monsoon when there is no agricultural work. A family might receive Rs 50,000 to 60,000 as an advance. After six months of work at a kiln, after deductions for food, shelter, and the contractor’s cut, the family may still owe money. The cycle begins again.

Stream Two: The semi-permanent corridor, especially Ganjam to Surat. The participants are predominantly from Ganjam district, organized along caste lines — the Teli community (OBC) dominates the powerloom network. Unlike the dadan stream, this is not seasonal. Workers stay in Surat for years. They have addresses, even if the addresses are six-by-three-foot sleeping spaces shared with a shift partner in a room housing sixty to a hundred workers. They earn Rs 20,000 to 25,000 per month — roughly three to four times what MGNREGA pays at home. They work twelve-hour shifts, seven days a week. They send money home. The remittance from Ganjam alone is estimated at Rs 120 to 124 crore per month — roughly Rs 1,440 to 1,488 crore per year, or approximately $170-175 million. The corridor has its own social infrastructure: over 500 Odia community groups in Surat, caste-based mess halls, Odia-medium primary schools, festival celebrations. It is, in effect, a displaced Ganjam. The economics work for the individual — a powerloom worker in Surat earns more than he would farming in Ganjam. Whether the economics work for Odisha is a different question.

Stream Three: Skilled and professional migration. Engineers, doctors, IT workers, civil servants, academics, seafarers. The economics here are not survival but optimization. An NIT Rourkela graduate who takes a Rs 14 lakh per annum job in Bangalore is not fleeing poverty. They are making a rational career choice in a country where the best-paying jobs are concentrated in a handful of cities, none of which are in Odisha. Bhubaneswar’s IT sector is growing — Infocity SEZ, Info Valley SEZ, major firms present — but it cannot yet compete with Bangalore for specialized roles and career trajectories. The medical graduate who leaves for a better-equipped hospital in Chennai or Delhi, the IAS officer on central deputation (62 officers from the Odisha cadre serve outside the state), the researcher who moves to a better-funded university — each is making a locally rational choice that is collectively devastating. Odisha has 42 IAS posts vacant, 67 IPS posts vacant, 66 IFS posts vacant. The brain drain is not metaphorical. It is a measurable loss of administrative, technical, and institutional capacity.

The three streams share a root cause. Whether the push is absolute poverty or relative lack of opportunity, the underlying condition is the same: Odisha does not generate sufficient economic activity, at sufficient wage levels, across enough of its territory, to retain its own people. A brick kiln worker from Nuapada and a software engineer from Bhubaneswar are responding to the same structural failure at different points on the income spectrum.

---

The Push

Why do people leave? The question sounds simple. The answers layer on each other.

Unemployment. Odisha’s official unemployment rate was 3.9% in 2024, above the national average of 3.2%. Urban unemployment was worse at 6.2%. By the Economic Survey of 2025-26, covering April to September 2025, the rate had risen to 6.5%. These numbers mask the reality of underemployment and disguised unemployment in agriculture. A farmer who works twenty days in a kharif season and has nothing to do the rest of the year is not “unemployed” in the survey — he is “self-employed in agriculture.” Compare Odisha’s rate to Gujarat at 2.2% — among the lowest in India. The gap is not just in the headline number. It is in the nature of available work.

Wages. This is the clearest driver. Odisha’s MGNREGA wage rate is Rs 237 per day. With the supplementary scheme that the state government added in migration-prone blocks, it rises to Rs 352 per day. In Surat, a powerloom worker earns Rs 20,000 to 25,000 per month — which works out to roughly Rs 666 to Rs 833 per day for the twelve-hour, no-day-off shifts they work. The ratio is between three and four to one. A worker with a family to feed, debts to service, children to educate, and a festival to celebrate is not going to choose Rs 237 over Rs 666 because a government scheme asks him to stay home. The arithmetic is brutal and decisive. Even the dadan advance — Rs 50,000 to 60,000 upfront, against future labor — is more cash than most families in western Odisha see in a year from local sources. When the sardar arrives in August offering money, he is offering the only accessible cash in the village economy. No bank will lend to these families. No government program delivers comparable sums with comparable immediacy. The sardar’s advance is exploitative. It is also the only offer on the table.

Agricultural failure. Ninety-three percent of Odisha’s farming community are small and marginal farmers. Land fragmentation is severe. In Balangir, only approximately 3% of agricultural land has irrigation. Three percent. In a district where 77% of adults depend on agriculture, farming with 3% irrigation means farming at the mercy of the monsoon — a monsoon that arrives with increasing irregularity as climate patterns shift. Odisha’s cropping intensity was 117% in 2023, against a national average of 156% — meaning Odisha’s land produces substantially less per unit area per year than the rest of India. Foodgrain productivity was 1,798 kg per hectare versus the national 2,515 kg per hectare. The recent Economic Survey shows improvement — cropping intensity rising to 165%, foodgrain output reaching 150.5 lakh metric tonnes — but the structural deficits in irrigation, especially in western Odisha, have not been addressed at scale. Farming in Balangir or Nuapada is, for most families, a gamble against drought with no insurance except the sardar’s advance.

Poverty. The NITI Aayog Multidimensional Poverty Index, based on the National Family Health Survey 2019-21, places Odisha’s poverty at 29.35% — ninth among states. But the state average conceals catastrophic district-level variation. Nabarangpur at 59.32%. Malkangiri at 58.71%. Koraput at 51.14%. Rayagada at 48.14%. Kalahandi at 47.28%. Mayurbhanj at 44.90%. Kandhamal at 44.75%. Keonjhar at 41.78%. In the worst-affected districts, roughly half the population is multidimensionally poor — deprived across health, education, and standard of living simultaneously. Odisha’s overall poverty has improved — from 29.34% in 2015-16 to 15.68% in 2019-21 by NITI Aayog’s calculation — but in the KBK districts that generate the most migration, poverty levels remain three to four times the national average. The improvement, substantial as it is, has not reached the places that need it most.

Natural disasters. Odisha is among India’s most disaster-prone states. Between 1891 and 2000, ninety-eight cyclones and severe cyclones crossed the Odisha coast. The 1999 super cyclone killed more than ten thousand people and devastated coastal Odisha. Cyclone Phailin in 2013 destroyed livelihoods. Cyclone Fani in May 2019 caused approximately $8.1 billion in damage, killed 81 people, left 60% of families in affected areas homeless, and destroyed 35% of coconut orchards that families depended on. Cyclone Amphan hit in May 2020, compounding the COVID crisis. Annual floods affect agricultural areas, particularly in the Mahanadi and Brahmani river basins. KBK districts face drought. Small and marginal farmers who lack resources for recovery face the most severe livelihood destruction. Each cyclone, each flood, each drought pushes another cohort into the migration stream. The disasters are not the root cause — the root cause is the absence of economic alternatives that would provide resilience. But the disasters are the trigger that converts latent vulnerability into actual departure.

---

The Data Gap

Here is what should concern anyone trying to make policy for this state.

There is no current census. The 2011 Census is fifteen years old. The 2021 Census was never conducted. The Ministry of Statistics has announced a National Migration Survey for July 2026 to June 2027 — the first comprehensive migration study in nearly two decades. Until it is complete, India — not just Odisha — is navigating labor and migration policy with data from a world that no longer exists.

There is no reliable aggregate of Odias living outside Odisha. The estimates range by a factor of more than two. The Census figure of 12.71 lakh was certainly an undercount even in 2011. The OMS 2023 figure of 1.7 million is based on household surveys that miss entirely departed households. The COVID evidence suggests millions. Nobody can say with precision whether the number is 2 million or 5 million.

There is no Odia-specific international diaspora count. India’s Ministry of External Affairs tracks overseas Indian populations by country but not by state of origin. Odias in the Gulf, the US, the UK, Australia, and Canada exist as a community — they have organizations, celebrate festivals, hold conventions — but their aggregate number is unrecorded.

There is no systematic remittance tracking. The best available estimate for Ganjam alone is Rs 1,440 to 1,488 crore per year flowing back from Surat. An NGO estimate from 2007 put total Odia migrant remittances at Rs 2,000 crore per year, but that figure is two decades old. Given the growth in migrant numbers and wages, a current estimate would likely be Rs 5,000 to 8,000 crore annually for all domestic remittances — but that is extrapolation, not data. Odisha is not among the states whose remittance flows are separately tracked by the Reserve Bank of India. The money comes in through bank transfers, private operators, and cash carried home during festivals. It is significant enough to transform village economies — new houses, motorcycles, micro-enterprises visible across Ganjam — but invisible in official economic statistics.

There is no data on return migration. How many migrants eventually come back to Odisha permanently? What skills do they bring? What do they do when they return? The COVID-19 lockdown was the largest unplanned experiment in return migration in Indian history — and almost no systematic follow-up study exists. We know most returned to their destination cities within months. We know some stayed. We do not know how many, or what happened to them.

The census misses seasonal migration. The PLFS underestimates because it surveys households, not individuals who have left. The OMS 2023 is the best available but still based on survey methodology with inherent limitations. The government’s own tracking of dadan workers — 94,106 registered in 2025, a 34% rise from the previous year — captures only a fraction. The Shramik Sathi app, the PAReSHRAM portal, the Mobile Migrant Resource Centre helpline — all are attempts at counting. None approach completeness.

Think of this in terms of any other domain. Imagine a company that could not say, within a factor of two, how many employees it had or where they worked. Imagine an investor who did not know the magnitude of cash flowing in and out of an account. The information deficit is not a minor inconvenience. It is a structural inability to plan. You cannot design an employment program for migrants you have not counted, track the health outcomes of workers you cannot locate, protect the labor rights of people whose employers you have not registered, or measure the economic impact of remittances you do not measure. Odisha cannot plan for migration it cannot count.

---

Recent Efforts to Count

The state government has not been entirely idle on the tracking question. In October 2024, the Mohan Charan Majhi government constituted a twenty-member Task Force on Distress Migration, led by Deputy Chief Minister K.V. Singh Deo, with members including the other Deputy CM, the Panchayati Raj Minister, the Chief Secretary, and the Development Commissioner. The task force’s objectives include assessing migration dynamics, linking welfare schemes to migrant families, and coordinating with destination states.

The dadan migration tracking system, which produces the 70,142 and 94,106 figures, represents a genuine effort to register workers before they leave. Awareness campaigns have been conducted in thirty migration-prone blocks. Rescue operations bring back over a thousand workers annually from exploitative conditions — 1,055 in 2024, 1,078 in 2025.

But these are inputs, not outcomes. The 94,106 registered workers in 2025 need to be placed against the actual population in motion — which, by all estimates, is several times larger. The tracked number and the actual number are not converging. The 34% rise in tracked migration between 2024 and 2025 could mean either that more people are migrating (likely) or that the tracking system is improving (also likely, but modestly). It does not mean the gap between the known and the unknown is closing.

The National Migration Survey 2026-27, when it comes, will be the most comprehensive data since the pre-pandemic period. It will cover migration rates, short-term patterns, reasons for movement, and net migration balance. For Odisha, this survey matters enormously. It may, for the first time in nearly two decades, provide a defensible number for how many Odias live outside the state, where they are, and why they left. Whether the state government will use that data to design interventions rather than cite it in speeches remains an open question.

---

What the Numbers Conceal

Every number in this chapter represents a decision. Not an abstract economic decision in the way that textbooks describe rational agents optimizing utility. A specific decision, made by a specific family, in a specific village, at a specific moment, when the available options were weighed and the conclusion reached was that leaving was less bad than staying.

Behind the 7-8 lakh in Surat’s powerloom neighborhoods, there are families that concluded agricultural income in Ganjam would not cover school fees, medical expenses, and the social obligations of a wedding or a funeral. Behind the 60,000 families who board trains at Kantabanji each winter, there are monsoons that failed and moneylenders who did not. Behind the NIT Rourkela graduate who takes the Bangalore placement, there is a calculation about career trajectory that Bhubaneswar’s growing-but-nascent tech sector cannot yet satisfy. Behind the plumber from Kendrapada who takes a contract in Qatar, there is a village where every household has one member in the trade because that is what the village learned to do when nothing else worked.

The numbers are essential. Without them, the problem has no shape. You cannot argue for irrigation in Balangir without citing 3% irrigation coverage. You cannot advocate for textile investment in Ganjam without citing the Rs 1,440 crore in remittances flowing to Surat instead of staying home. You cannot push for updated migration data without demonstrating that the fifteen-year-old census is useless for planning.

But the numbers are also dangerous in exactly the way that all aggregate statistics are dangerous: they make the phenomenon manageable. Twelve point seven one lakh is a number you can put in a table. Three lakh fifty-eight thousand four hundred and one returning on Shramik Specials is a number you can report in a press conference. Two to five million is a range you can cite in a policy document. These numbers do not sweat through twelve-hour powerloom shifts. They do not have their identity documents confiscated by a sardar. They do not sit in a quarantine camp in June wondering whether to go back to the city that nearly killed them or stay in the village that cannot feed them. They do not watch their children grow up in Surat speaking Gujarati and forgetting Odia, or leave their children with grandparents in a village with a failing school and no working-age adults.

The remaining chapters of this series will put names, mechanisms, economics, and human costs to the numbers laid out here. The dadan system. The Surat corridor. The professional departure. The remittance economy. The empty village. The diaspora identity. The conditions for return.

But the starting point is the count. Or rather, the inability to count. A state that does not know how many people have left it — that cannot say whether the number is two million or five million, that has not conducted a census in fifteen years, that tracks fewer than a hundred thousand of the hundreds of thousands who leave each year through the dadan system alone — is a state that has not yet reckoned with the scale of what is happening.

The Shramik Special trains of 2020 made the leaving visible. Three lakh fifty-eight thousand four hundred and one people on trains. Ten lakh on a portal. Images that became some of the defining photographs of India’s pandemic. For a brief moment, the country saw what Odisha’s villages had seen for decades — the departure, made visible by crisis.

The trains stopped running. The visibility faded. The migration resumed. Within months, the powerloom workers were back in Surat, the brick kiln workers were boarding trains at Kantabanji, the engineers were logging into Bangalore offices. The system had paused, not changed. The numbers, still uncounted, continued to grow.

---

Chapter 2, “The Dadan Road,” examines the labor contractor system in detail: how it works, why it persists, and what it does to the people inside it.