English only · Odia translation in progress

Chapter 5: The Export Factory

On a December morning in 2024, the placement hall at NIT Rourkela filled with the quiet electricity of ambition about to be rewarded. Three hundred and seventy-three companies had visited the campus that season — Google, Amazon, Qualcomm, AMD, DE Shaw, Texas Instruments, American Express. Students in pressed shirts and borrowed blazers waited in queues outside interview rooms that had been makeshift-constructed in classrooms and laboratory annexes. The campus, tucked into the sal forests of western Odisha, three hours from the nearest international airport, was for a few weeks each year connected by a thread of corporate attention to Bangalore, Hyderabad, Pune, Mumbai, and points further afield.

The numbers, when they came, were genuinely impressive. 1,274 job offers. 509 six-month internships. The highest package: Rs 62.44 lakh per annum, a Computer Science student whose identity remained undisclosed per convention. Eighty-five students secured CTCs above Rs 30 lakh. The average BTech salary came in at Rs 14.10 lakh. The placement rate hit 82.20 percent. NIT Rourkela celebrated. The institute’s social media channels pushed the numbers. Parents in Cuttack and Bhubaneswar forwarded screenshots. Local newspapers ran headlines. The narrative was clear: Odisha produces world-class engineers.

But there was a number the placement office did not publish, has never published, and has no institutional incentive to publish: of those 1,274 offers, how many placed graduates in Odisha?

The answer is knowable in principle. The data exists in the placement records. The institute does not release destination breakdowns. It does not need to. Anyone who has attended an engineering campus in India knows the geography of aspiration. The highest packages go to Bangalore. Then Hyderabad. Then Pune and Mumbai. Then Delhi NCR. Then, for the top percentile, the United States. The few who stay in Odisha go to TCS Bhubaneswar or Infosys Bhubaneswar — offices that function, for many ambitious graduates, as a two-year stepping stone before a transfer to a larger city.

The Rs 62.44 lakh package did not land in Rourkela. It did not land in Bhubaneswar. It landed, with near certainty, in Bangalore or Hyderabad or possibly California. And the student who accepted it — raised in Odisha, educated at Odisha’s taxpayer-subsidized premier engineering institution, trained by faculty paid from the public exchequer — walked out of the placement hall and, in the only sense that matters to an economy, walked out of the state.

This is what NIT Rourkela was built to do. Not by design, not by intent, but by the structural logic of the economy that surrounds it. The institute was established in 1961, alongside Rourkela Steel Plant, as part of Nehru’s vision of industrial India. The steel plant was supposed to transform the economy of western Odisha. The engineering college was supposed to supply it with engineers. Sixty-three years later, the steel plant has not transformed the economy of western Odisha. And the engineering college has become something Nehru never imagined: the most efficient talent extraction pipeline in the state.

This chapter is about that pipeline. Not just at NIT Rourkela, but across the entire education system — engineering colleges, medical schools, management institutes, civil service training, academic departments. The system that produces Odisha’s best and brightest, and then exports them. Not under coercion, like the dadan road. Not under semi-coercion, like the Surat corridor. Under celebration. The export factory produces human capital at state expense and ships it out to standing ovations.

In the language of investing, this is the worst possible outcome: selling an asset at cost basis when its true value is many multiples higher.

The Per-Graduate Economics

To understand the export factory, start with the numbers the way the Value Chain series started with a tonne of iron ore — by following a single unit through the production process.

A child enters Odisha’s government school system. The state runs 66 lakh students through its public education apparatus. The government spends approximately Rs 53,800 per student per year through its education budget of Rs 35,536 crore (2024-25). Over fourteen years of schooling — Class 1 through Class 12 — the state’s cumulative investment in a single student comes to roughly Rs 7.5 lakh. The family bears additional costs: uniforms, coaching, transport, books. For a student from a modest household, the family contribution might be Rs 2-3 lakh over fourteen years. Total investment to produce a Class 12 graduate from the government system: approximately Rs 10 lakh.

Now the student clears JEE and enters NIT Rourkela. The four-year BTech program charges Rs 6.4-8.2 lakh in fees. The actual cost of educating a student at an NIT is estimated at 3-5 times the fees charged, with the difference subsidized by the Government of India. Call it Rs 25-35 lakh in actual institutional costs, of which the student pays roughly a quarter. The government — central and state, through subsidies, infrastructure investment, and faculty salaries — absorbs the rest. Add the family’s contribution for hostel, coaching, and living expenses.

The full worked example for one NIT Rourkela graduate:

K-8 education (government school): ~Rs 1 lakh (government cost) + Rs 1.9 lakh (family) = Rs 2.9 lakh. Class 9-12: ~Rs 50,000 (government) + Rs 30,000 (family) = Rs 80,000. JEE coaching (private): Rs 2-5 lakh (family cost). NIT Rourkela BTech (4 years): student pays Rs 6.4-8.2 lakh, government subsidy estimated at Rs 15-30 lakh. Total investment: Rs 25-40 lakh. Government share: Rs 17-32 lakh. Family share: Rs 8-14 lakh.

Now the graduate takes a Bangalore placement at Rs 14 lakh per annum — the average, not the highest. Over a 35-year career, with normal salary progression in the IT industry, their lifetime earnings will be Rs 8-15 crore. They will pay income tax (central), GST on their consumption (destination state), property tax on their Bangalore apartment (BBMP, Karnataka), professional tax (Karnataka), school fees for their children (Karnataka’s education system), medical bills (Karnataka’s hospitals). Every rupee of the multiplier effect — the restaurants they eat at, the autorickshaw driver they hire, the domestic worker they employ, the school that hires teachers to teach their children — generates economic activity in Karnataka.

Return captured by Odisha: approximately zero. Occasional remittances to aging parents. A flight home for Durga Puja. An annual donation to the alumni association. Maybe, if sentiment is strong, a contribution to a hometown temple renovation.

The state spent Rs 17-32 lakh producing this graduate. The graduate’s lifetime economic value is Rs 8-15 crore. The state captures none of the appreciation. This is selling at cost basis.

Now multiply. NIT Rourkela graduates roughly 1,200 students per year. IIT Bhubaneswar adds 300. KIIT sends out 5,000. SOA another 1,900. VSSUT contributes 700. The smaller engineering colleges produce another 15,000-25,000. Conservative total: 25,000-35,000 engineering graduates annually from Odisha. Of the top-tier graduates — the NIT, IIT, KIIT premium placement cohort — 85-95 percent leave the state. Even among the lower-tier graduates, a significant fraction migrates for IT jobs in Bangalore, Hyderabad, Pune, and Chennai.

The annual human capital export from Odisha’s engineering pipeline alone runs into the thousands. The cumulative fiscal transfer — education costs borne by Odisha, economic value captured by destination states — amounts to thousands of crores per graduating cohort. Over a decade, the number is measured in lakh crores. And this is just engineering. The medical graduates, the management graduates, the civil service candidates, the academics — each add their own stream to the river flowing out of the state.

In the Value Chain series, we traced a tonne of iron ore that left Keonjhar at Rs 4,200 and became a car panel in Tamil Nadu worth Rs 80,000-90,000 — an 18-20x multiplication of value, with Odisha capturing only the first step. The per-graduate economics follow the same structure, but the multiplication is worse. A graduate produced at Rs 25-40 lakh generates Rs 8-15 crore in lifetime value — a 20-60x multiplication. And unlike the iron ore, which at least generates royalties at the mine gate, the graduate generates no formal fiscal return to the state after departure.

The iron ore has a royalty rate. The human being has none.

The Company Roster Tells the Story

NIT Rourkela does not publish destination breakdowns. Neither does IIT Bhubaneswar, KIIT, SOA, or any other major institution. What they do publish is company names. The company names are the story.

NIT Rourkela’s top recruiters in 2024-25: Apple (Cupertino/Hyderabad), Google (Bangalore/Hyderabad), Microsoft (Hyderabad/Bangalore), Amazon (Bangalore/Hyderabad), Goldman Sachs (Bangalore), DE Shaw (Hyderabad), Qualcomm (Hyderabad/Bangalore), Texas Instruments (Bangalore), American Express (Gurgaon). None of these companies have significant operations in Odisha. The highest-paying offers — the ones that make the headlines, the ones that parents forward on WhatsApp, the ones that NIT Rourkela’s PR team pushes to local newspapers — go overwhelmingly to cities that are not in Odisha.

IIT Bhubaneswar’s roster tells the same story: Google, Oracle, Microsoft, D.E. Shaw, Zomato, MathWorks, Accenture Japan. The institute placed 254 BTech students at a 90.07 percent rate in 2025, with the highest package hitting Rs 67.6 lakh — the highest ever for the institute. Every company in the top salary bracket is headquartered outside Odisha. The CSE department’s 92.4 percent placement rate (61 of 66 students) is functionally a 92.4 percent export rate.

KIIT, the private behemoth with 5,585 offers from 700-plus companies in 2024, feeds the same pipeline. Amazon, DE Shaw, McKinsey, Salesforce, Dell Technologies, Deloitte. The placement brochure is a catalog of companies whose nearest offices to Bhubaneswar are in other states. SOA’s 2,218 offers from 210 companies follow the pattern. XIM Bhubaneswar’s 100 percent MBA placement rate, with the highest package at Rs 30 lakh, sends graduates into BFSI (25 percent of placements), IT (23 percent), and consulting (24 percent) — sectors whose high-salary jobs are concentrated in Mumbai, Bangalore, Delhi, and Hyderabad.

Even VSSUT Sambalpur, the lower-tier data point in western Odisha where the median salary is Rs 6.38 lakh — modest by national standards — sends its top placements to Amazon, JP Morgan, UBS, Maruti Suzuki, Cummins, all headquartered outside the state.

The pattern is total. It operates across the quality spectrum. At the top tier (NIT, IIT, KIIT best placements), the export rate is 90-95 percent. At the middle tier (SOA, VSSUT, C.V. Raman), it is lower but still substantial. At the lower tier — the dozens of small private engineering colleges scattered across the state — graduates either find employment in Odisha at rock-bottom salaries or join the general migration stream to Bangalore, Hyderabad, and Pune as entry-level IT workers, hoping to climb from there.

The entire engineering education system, from the flagship to the forgotten, operates as an export-processing zone. Raw material enters at one end (children), is processed over 16-18 years (schooling plus college), undergoes quality sorting (JEE, NEET, campus placement), and exits the state as a finished product. The product never returns. The royalty is never paid.

Six Lakh Odias in Bangalore

The destination of this pipeline has a name, a latitude, and a population estimate that has become canonical: approximately six lakh Odias lived in Bangalore as of 2016, making it the largest concentration of Odia professionals outside Odisha. The number has almost certainly grown in the decade since, but nobody has updated the count. Nobody in Odisha’s government has commissioned a survey. Nobody at the state’s universities has studied where their graduates end up. The absence of measurement is itself a data point — a state that does not count what it is losing has no basis for designing interventions to stop the loss.

The growth trajectory is striking. From approximately 20 Odia families in 1968 to 600,000 people in less than five decades. The acceleration came with the IT boom of the late 2000s, as Bangalore cemented its position as India’s technology capital. The Odia community there is organized and visible: Odia Samaj Bengaluru runs social and cultural events; the Odisha Puja Committee coordinates Durga Puja celebrations; Juhar Parivar Bangalore’s 25th Nuakhai Mahotsav drew over 10,000 attendees at KTPO, Whitefield.

Six lakh Odias. In a city that is the center of an IT industry generating Rs 4.09 lakh crore in exports. A city with over 40 percent of India’s 1,750-plus global capability centers, employing more than 8 lakh professionals. If the Odia community in Bangalore constitutes even 10 percent of the city’s tech workforce, that is a material economic force — producing value, paying taxes, buying apartments, starting companies, mentoring juniors, building networks. All of it in Karnataka.

Now set this against Odisha’s own IT sector. Bhubaneswar has over 300 IT companies, collectively employing nearly 35,000 professionals. That is the total. The entire Bhubaneswar IT ecosystem is smaller than the Odia community in Bangalore alone. Odisha’s total software exports: Rs 12,905 crore in 2023-24. Karnataka’s: Rs 4.09 lakh crore. The ratio is 32:1. Not a gap. A chasm. And a significant portion of Karnataka’s IT workforce consists of engineers who were educated in states like Odisha.

This is the fiscal externality in its purest form. The concept, in economics, describes a situation where one jurisdiction bears the cost of producing a public good while another jurisdiction captures the benefit. It is usually applied to infrastructure — a state builds a highway, but the trucks that use it pay taxes elsewhere. The Odia brain drain is a fiscal externality of a different, more devastating kind. The public good is an educated person. The cost is borne by Odisha’s exchequer. The benefit is captured by Bangalore’s tax base, Bangalore’s consumer economy, Bangalore’s real estate market, Bangalore’s innovation ecosystem.

Nobody sends a check back to NIT Rourkela for services rendered. The alumni contribute through nostalgia, occasional donations, and cultural association memberships. The institution that produced them receives no share of the economic value their education creates.

If the six lakh Odias in Bangalore have an average household income of Rs 15 lakh per annum — conservative for IT professionals — the total community income is approximately Rs 90,000 crore annually. Even at modest consumption-to-income ratios, tens of thousands of crores in consumer spending generate state-level tax revenue and economic multipliers for Karnataka that Odisha never sees. Compare this with Odisha’s entire education budget of Rs 35,536 crore. The Odia community in a single city generates more annual economic output than Odisha spends on educating its entire population.

The Medical Drain: Where the Loss Becomes Physical

The engineering drain is an economic abstraction. The medical drain is not. When a doctor leaves, the harm is measurable in bodies.

Odisha has 15 government medical colleges (including AIIMS Bhubaneswar) offering approximately 1,725 MBBS seats annually. The three legacy institutions — SCB Medical College in Cuttack (founded 1944), MKCG Medical College in Berhampur, and VIMSAR in Burla — form the backbone. AIIMS Bhubaneswar, established in 2012, was supposed to be the crown jewel.

No systematic study tracks where Odisha’s medical graduates end up. But the structural evidence is clear enough to describe the pipeline without needing precise percentages. Post-graduation — the MD, MS, DM, or MCh degree that transforms a generalist into a specialist — increasingly requires clearing NEET-PG and competing for seats at AIIMS Delhi, JIPMER Puducherry, CMC Vellore, or private colleges in Karnataka and Maharashtra. Once trained as a specialist in Bangalore, the incentive to return to a district hospital in Odisha that lacks an MRI machine is approximately zero.

The salary differential is stark. Specialist salaries in Bangalore or Delhi run 3-5 times what comparable specialists earn at Odisha’s district hospitals. A senior ophthalmologist at a corporate hospital in Bangalore might earn Rs 40-60 lakh per annum. The same ophthalmologist, if they returned to practice at a district hospital in Bolangir or Kalahandi, would earn Rs 8-15 lakh. The Rs 50 lakh bond penalty that government medical college students must sign — agreeing to three years of government hospital service or else paying the fine — is easily recoverable within one to two years of metro practice for a specialist. The bond is a retention mechanism based on coercion rather than incentive. And coercion, unlike incentive, does not build loyalty.

The consequence maps directly onto Odisha’s healthcare statistics. India’s national doctor-patient ratio stands at approximately 1:1,456, itself below the WHO recommendation of 1:1,000. In rural areas, the ratio collapses to 1:11,082. Nearly two-thirds of the health workforce is concentrated in urban areas. Odisha is classified among “less developed states reflecting acute shortage of health workforce” alongside Bihar, Jharkhand, and Rajasthan. A study of 18 rural health centers in a single Odisha district found shortages of 43 doctors and 15 nurses.

The vicious cycle is the same as in the IT sector, but its costs are measured in lives rather than rupees. Hospitals are under-equipped, so doctors leave. Because doctors leave, hospitals cannot upgrade — because a cardiac catheterization lab without a cardiologist is just expensive furniture. Because hospitals cannot upgrade, patients travel to Hyderabad or Vellore for treatment, draining healthcare spending from the state. Because revenue drains, equipment procurement stalls, and more doctors leave. The cycle feeds itself.

At the national level, over 60,000 doctors and 75,000 engineers leave India annually. 45 percent of Indian medical undergraduates plan to pursue residency abroad. Odisha’s medical graduates participate in this national pattern, but they do not need to leave the country to exit the state. Chennai, Bangalore, Hyderabad, and Delhi are far enough.

The per-graduate economics are even more dramatic for medical education. Government MBBS education in Odisha costs the student Rs 2-3 lakh total for five years. The actual cost — faculty salaries, hospital infrastructure, clinical training, equipment — is many multiples higher. A specialist trained through government subsidy from primary school through medical school through post-graduation represents a total public investment of Rs 25-40 lakh. Their lifetime earnings as a metro specialist will be Rs 5-15 crore. The entire value appreciation is captured by the destination.

The iron ore analogy from the Value Chain series is imperfect for human capital in one specific way. A tonne of iron ore that leaves Odisha as raw material is at least still available in the earth. The deposit can be mined tomorrow. The next tonne is waiting. But a trained cardiologist who leaves for Apollo Hospital in Hyderabad does not leave a replacement behind. There is no equivalent of the ore body. The 1,725 MBBS seats are a flow, not a stock, and each graduate who leaves creates a gap that takes five to ten years to fill — if it gets filled at all.

In 2025, an exam to recruit 45 District Judges in Odisha produced zero qualifying candidates. Out of 83 judicial officers and 283 lawyers who appeared, not one met the qualifying standard. This is not merely a vacancy. It is a recruitment failure so complete that it suggests the talent pool for certain senior professional positions in Odisha has thinned past the point of institutional function. The medical drain is moving in the same direction. It is not there yet. But the trajectory is clear.

The Civil Services: When the Administrators Leave

The bureaucratic drain is subtler than the IT or medical drain, but potentially more damaging, because it affects the state’s capacity to govern itself.

As of 2024-25, Odisha faces severe vacancies across all major administrative services. The numbers, disclosed by Chief Minister Mohan Majhi in the State Assembly, are these: IAS — 248 sanctioned, 42-46 vacant (17-19 percent). IPS — 195 sanctioned, 67 vacant (34 percent). IFS — 141 sanctioned, 66-70 vacant (47-50 percent). OAS — 2,248 sanctioned, 330 vacant (15 percent). OPS — 852 sanctioned, 147 vacant (17 percent). Total: 182 IAS/IPS/IFS posts vacant. Over 330 OAS positions unfilled.

These are not abstract numbers. A vacant District Collector position means a district of 1-3 million people is governed under “additional charge” — one officer managing two or three districts, each getting a fraction of needed attention. A 34 percent vacancy rate in the IPS means policing capacity is structurally degraded. A 47-50 percent vacancy in the IFS means half the forest service positions in a state that is 33 percent forested are empty.

The central deputation drain is the quiet mechanism. Of IAS officers nominally on the Odisha cadre, 26-27 serve in Delhi — central ministries, commissions, tribunals. 28 IPS officers are on central deputation. 7 IFS officers. Total: approximately 62 officers who belong to Odisha’s cadre but work elsewhere. When 27 of your IAS officers prefer to be in Delhi rather than Bhubaneswar, the preference is not random. Delhi offers better housing, better schools for children, exposure to national policy, proximity to power, and a professional environment where ambition is not constrained by the limited canvas of a single state. For the ambitious bureaucrat, Bhubaneswar can feel like a hardship posting.

The cadre allocation system compounds the problem. Under India’s cadre allocation policy, the insider-outsider ratio is maintained at 1:2 — one-third of direct recruits are “insiders” from the same state, two-thirds posted as outsiders. This means only one-third of IAS officers allocated to Odisha are Odia by origin. The rest are outsiders with no prior connection to the state, no understanding of the language, no intuitive grasp of local dynamics. The system works both ways: Odia-origin candidates who clear UPSC may be allocated to Madhya Pradesh, Assam, or Gujarat, spending their careers administering states they were not born in.

Then there is the OAS-IAS prestige gap. The Odisha Administrative Service starts at the same basic pay as IAS — Rs 56,100 per month under Level 12. But career progression diverges sharply. IAS officers can rise to Cabinet Secretary. OAS progression is constrained to state-level positions with lower ceilings. In practice, OAS is viewed as IAS’s poorer cousin. For ambitious candidates who narrowly miss IAS selection, OAS functions as a consolation prize rather than a first choice. This creates a selection effect: the most capable aspirants either clear IAS (and may be allocated elsewhere) or retry until they do. OAS gets those who settle.

The governance consequence is this: Odisha’s administrative apparatus is simultaneously understaffed, partially staffed by outsiders with no roots in the state, and partially staffed by officers who would rather be in Delhi. Institutional memory is destroyed when officers rotate through Odisha as briefly as regulations permit before seeking central deputation. Policy continuity suffers. Implementation degrades. The startup analogy is precise: a startup that cannot retain senior engineers is permanently in onboarding mode, perpetually losing the people who know how things work, perpetually training replacements who will themselves leave. The codebase becomes a mess. The product stagnates.

Odisha’s bureaucratic drain is this scenario at state scale. The state cannot build the institutions that would retain talent because the talent that would build those institutions has left to build institutions elsewhere.

The Academic Drain: When Knowledge Production Stops

If the IT drain costs Odisha economic output and the medical drain costs lives and the bureaucratic drain costs governance capacity, the academic drain costs something harder to measure but equally structural: the ability to understand yourself.

Odisha’s public universities are running on fumes. 1,404 teaching posts lie vacant across 17 state-run universities — roughly 60-70 percent of sanctioned positions. At Utkal University, the state’s oldest and most prominent, more than 55 percent of faculty positions are unfilled. 133 vacancies out of 238 sanctioned posts. The departments of Business Administration and Psychology have 10 vacancies each. Non-teaching staff: only 189 of 502 positions occupied.

The system is kept alive by 963 guest faculty members across six major universities. These guest lecturers earn Rs 20,000-25,000 per month — roughly one-fourth of what a permanent faculty member earns. Payment delays of three to six months are common. They receive no pension, no maternity leave, no medical insurance, no research grants. Yet they teach classes, examine papers, guide dissertations, and mentor students — performing the same duties as permanent faculty at a fraction of the compensation.

The last major hiring drive took place almost five years ago. Between 2017 and 2024, less than 12 percent of faculty vacancies in state universities were permanently filled. The 2024 University Amendment Act initiated a process to fill 1,353 positions, but whether this will complete before the next election cycle remains uncertain.

The degradation loop is self-reinforcing. Universities with thin faculty produce fewer high-quality researchers. Fewer researchers mean less funding, fewer publications, lower rankings. Lower rankings make it harder to attract students and faculty. The university becomes a place people pass through rather than build careers at. Students who might have stayed for graduate studies leave for JNU, IITs, or universities abroad.

Not a single Odisha public university appears in the NIRF top 100. The institutions that do appear are either private (SOA at rank 25, KIIT at rank 27) or centrally funded (NIT Rourkela at rank 34, IIT Bhubaneswar at rank 80). Utkal University, Sambalpur University, Berhampur University, Fakir Mohan University, North Orissa University — the state system that serves the vast majority of students — is invisible in national rankings.

The practical harm of academic drain is not abstract prestige. It is the inability to produce the knowledge needed to govern well. Consider what Odisha-specific research requires but struggles to produce. Mineral economics — the value chain from ore to finished product, the policy alternatives to raw ore export. This research should come from NIT Rourkela or IIT Bhubaneswar, but the researchers who could do it are at IIT Bombay or ISI Kolkata. Migration studies — the first comprehensive survey of Odisha’s migration was the Odisha Migration Survey 2023, conducted by IIT Hyderabad. Not by an Odisha institution. By an institution in another state. Tribal governance — the 62 communities across Odisha’s Scheduled Areas need policy research rooted in local understanding. The anthropologists who study them are often at JNU or Delhi University. Climate adaptation — Odisha’s specific cyclone vulnerability, heat stress patterns, and coastal erosion require localized research. The climate scientists are at IISc Bangalore or IIT Delhi. Odia language technology — NLP and AI for Odia, a small language by global standards, needs dedicated researchers. The computational linguists are at IIT Madras or IIT Kharagpur.

Each of these is a research gap that directly affects policy quality. The absence of researchers is the absence of knowledge is the absence of good policy.

And here is the meta-irony that captures the entire dynamic: Odisha’s migration crisis was not comprehensively surveyed until 2023, and the institution that conducted the survey was IIT Hyderabad. Odisha’s own universities lacked the research capacity to study Odisha’s most pressing social phenomenon. The state could not even study its own brain drain because the brain drain had hollowed out the institutions that would have conducted the study.

The Kendrapada Plumber and the IIT Graduate

One hundred thousand plumbers from a single district.

Kendrapada, on the coastal plain between Cuttack and the Bay of Bengal, has produced an estimated one lakh people working in a single trade — plumbing — spread across India and the Gulf. The villages of Pattamundai, Aul, Rajkanika, and Rajnagar blocks are the epicenter. Almost every household has at least one member in the trade. The State Institute of Plumbing Technology (SIPT) in Pattamundai is the only plumbing-dedicated training institute in India.

The earnings are not trivial. Rs 30,000 to Rs 1 lakh per month in Indian cities. Rs 50,000 to Rs 1 lakh in the Gulf. Kendrapada plumbers worked on India’s new Parliament building, on convention centers in Delhi, on Lusail Stadium in Qatar — where Argentina beat France in the 2022 World Cup final. Approximately 2,500 plumbers from Kendrapada worked on Qatar’s FIFA infrastructure.

This is skilled migration in the purest sense. A community that identified a market niche, developed training infrastructure, built a reputation through quality of work, and then exported that skill wherever demand existed. The Kendrapada plumber is not a brain drain victim. He is a freelance professional with portable skills and global demand.

But the structural logic is identical to the NIT Rourkela graduate. Kendrapada invested in producing skilled plumbers — through apprenticeship networks, through SIPT, through community knowledge transfer across generations. Qatar employed them. Delhi employed them. The skills were developed in Odisha. The infrastructure they built belongs to other places.

The distinction between the plumber and the software engineer is one of wage level and social class, not of economic structure. Both were educated in Odisha. Both work outside Odisha. Both send some money home. Both generate multiplier effects for their destination economies. Both represent human capital that Odisha developed and exported at cost basis.

The difference — and it matters for the argument — is in the nature of the loss. The plumber’s skills are portable but not institution-building. One hundred thousand plumbers in Delhi do not create a plumbing ecosystem in Delhi that attracts more plumbers. They do not start plumbing companies that IPO on the stock exchange. They do not mentor the next generation of plumbing entrepreneurs. They fix pipes, earn well, and send money home.

The NIT Rourkela graduate in Bangalore does create an ecosystem. They start companies. They mentor juniors. They invest in startups. They build the network effects that make Bangalore the self-reinforcing technology capital it has become. Each Odia engineer in Bangalore makes Bangalore more attractive to the next Odia engineer. The feedback loop works for the destination, against the source. When the Churning Fire series examined network effects and the conditions under which collective consciousness shifts, the prerequisite was local density — enough people in the same place, interacting with sufficient frequency, to generate emergent behavior. Brain drain thins that density. The engineers leave. The network never forms. The critical mass is never reached. Odisha remains permanently below the threshold where self-reinforcing dynamics would kick in.

This is why the Kendrapada plumber and the NIT Rourkela graduate, despite sharing the same structural logic, represent different magnitudes of loss. The plumber’s departure costs Odisha a salary’s worth of multiplier effects. The engineer’s departure costs Odisha a potential ecosystem. And ecosystems, once established elsewhere, are nearly impossible to relocate.

The Iron Ore Value Chain, Applied to Human Capital

The Value Chain series traced a tonne of iron ore from Keonjhar district through six stages of processing, each adding value, until it became a car panel in Tamil Nadu worth 18-20 times the mine-gate price. Odisha captured the first step — the Rs 4,200 per tonne — plus royalties amounting to roughly Rs 630 per tonne. Tamil Nadu captured the car.

The human capital value chain follows the same structure, with one critical difference that makes it worse.

StageIron OreHuman Capital
Raw materialOre in the ground (Rs 4,200/tonne)Child in government school
Initial processingBeneficiation, pelletization (2-2.5x value)K-12 education (~Rs 10 lakh investment)
Intermediate productDRI/sponge iron (Rs 25,000-32,000/tonne)Engineering/medical graduate (~Rs 25-40 lakh investment)
Finished productHot-rolled coil (Rs 48,000-55,000/tonne)Working professional (Rs 8-15 crore lifetime value)
Value capture by OdishaRoyalty (~Rs 630/tonne) + mine-gate saleApproximately nothing
Value multiplication18-22x20-60x

The analogy holds at every step. But two features make the human capital chain worse than the mineral chain.

First, the mineral at least stays in the ground until extracted. The iron ore in Keonjhar is not going anywhere. It will be there in ten years, in fifty years. Its extraction can be paced, regulated, taxed. If Odisha develops processing capacity tomorrow, the ore is available for domestic use. The graduate, once gone, does not return. There is no equivalent of the reserve estimate. Each year’s cohort that leaves is a cohort that cannot be mined later.

Second — and this is the point that the Delhi’s Odisha series developed in Chapter 1 on the Freight Equalization Policy — the mineral chain at least has a policy lever. Royalty rates can be raised. Export duties can be imposed. Processing requirements can be mandated. The central government’s Freight Equalization Policy of 1952-1993 artificially cheapened the transport of raw materials out of mineral-rich states, effectively subsidizing the export of unprocessed ore. When that policy was finally abolished, the lever existed to change it. For human capital, there is no equivalent lever. You cannot impose an export duty on a graduate. You cannot mandate that a doctor spend their career in the state that trained them (the Rs 50 lakh medical bond tries, and largely fails). You cannot require NIT Rourkela to place a minimum percentage of graduates in Odisha. Human capital moves freely in a democracy. The freedom of movement that makes this drain possible is also a constitutionally guaranteed right that no policy can constrain.

This means the only possible intervention is on the demand side. You cannot stop the graduate from leaving. You can only make staying a rational choice. And “making staying rational” means building the ecosystem that offers comparable salaries, comparable career trajectories, comparable quality of life, comparable spousal employment, comparable school quality, comparable urban infrastructure. This is a decades-long project. Every year it is delayed, the gap widens, because the graduates who would have built that ecosystem are building it somewhere else.

The Freight Equalization Policy destroyed mineral processing in Odisha by making raw material transport artificially cheap. Education policy — or rather, the absence of a human capital retention policy — destroys human capital processing through the same logic. The raw material (the educated graduate) is cheap to export. The destination captures all the value-addition. The source state bears the cost and captures nothing. The FEP was a deliberate central policy. The human capital drain is not deliberate — nobody designed it. But the structural outcome is identical.

Why People Don’t Return

The conversation about return always begins with sentiment and ends with arithmetic. Every Odia in Bangalore has the sentiment. Very few have the arithmetic working in their favor.

The salary gap is the entry point. A mid-level software engineer in Bangalore earns Rs 18-25 lakh. The same role in Bhubaneswar, if it exists, pays Rs 8-14 lakh. A senior engineer in Bangalore: Rs 40-80 lakh. In Bhubaneswar: Rs 15-25 lakh, and only at the handful of large IT offices that have senior roles. The gap is not primarily about cost-of-living differences. Bhubaneswar is cheaper, but not half as cheap. The gap is about ecosystem depth.

The ecosystem gap is where the arithmetic becomes decisive. Bangalore has 12,000-plus startups. Bhubaneswar has 132 tracked. Bangalore has hundreds of VC firms. Bhubaneswar has a government startup fund and a handful of angel investors. In Bangalore, if you leave one company, 50 others within a ten-kilometer radius might hire you for a raise. In Bhubaneswar, your options after leaving one of the handful of large IT offices are constrained. Career velocity — the rate at which compensation grows over time — depends on the density of the labor market. In a thin market, you plateau early. In a dense market, each job switch is a step up. Over a 35-year career, the compound effect of faster career velocity is enormous.

The quality of life gap operates on the spouse and the family, not just the professional. In dual-income households — the norm among educated urban Indians — both partners need viable careers. Bhubaneswar’s job market is thin outside IT. A software engineer married to a marketing professional, or a data scientist married to a financial analyst, faces a problem: one of them may not find a comparable role in Bhubaneswar. The dual-career constraint is one of the most powerful forces keeping families in metros, and it is almost never mentioned in policy discussions about brain drain.

Education for children matters. Professional parents optimize for school quality. Bangalore offers a wider range of CBSE, ICSE, and IB schools, plus the coaching ecosystem for JEE and NEET preparation that has become a quasi-compulsory part of upper-middle-class Indian childhood. Healthcare matters. Odisha’s healthcare bed gap exceeds 2 lakh. Diagnostic centers are short by 4,432. In rural Odisha, 54 percent choose private providers because public facilities have “inconvenient locations, lower stocks, lack of respect from providers, poorer infrastructure.”

Then there are the intangibles. Online forums — Reddit, Quora — consistently mention the “small town” feel of Bhubaneswar: social surveillance, limited cultural life compared to metros, caste consciousness in social interactions. For professionals who have experienced the anonymity and diversity of Bangalore, returning to a smaller social ecosystem can feel constraining.

And finally, network effects lock the decision in. Once a critical mass of Odias is established in Bangalore, the city becomes the default path. Juniors follow seniors. College batchmates cluster. The Odia community provides a social safety net — festivals, associations, networks. Each additional Odia professional in Bangalore makes Bangalore more attractive to the next potential migrant. This is a positive feedback loop that works against retention: the more people leave, the easier it becomes for others to follow.

COVID-19 provided the natural experiment. When lockdowns hit in 2020, approximately 600,000 migrant workers registered with Odisha’s state portal for return. Over 500,000 came back. The government mobilized 2.27 lakh beds across 7,200 isolation facilities. For a brief moment, the pipeline reversed.

The result was unambiguous. Most returned migrants went back to destination cities within months once lockdowns eased. For IT professionals, some experimented with remote work from Bhubaneswar — Bangalore salary, Bhubaneswar cost of living. But this remained the exception. When the structural conditions in Odisha did not change, even a global pandemic was insufficient to retain returning workers. The experiment confirmed: without structural reform of the local economy, reverse migration is temporary. A crisis can push people home. Only an ecosystem can keep them there.

The Wipro Promise and the Gap Between Announcement and Reality

In 2007, Wipro announced plans for a development center with 15,000 IT engineers in Bhubaneswar. The government allotted land in Infocity. By 2015, Wipro had upgraded its ambition to a facility for 28,000 staff. The current headcount, nearly two decades later, is approximately 4,000. From 28,000 promised to 4,000 delivered — a 14 percent fulfillment rate.

This is not a Wipro-specific failure. It is a structural pattern. IT companies announce large operations in tier-2 cities as a combination of genuine strategic diversification (cheaper real estate, lower attrition) and political goodwill (state governments offer land, tax breaks, and infrastructure in exchange for employment promises). The announcement generates headlines. The execution lags because the ecosystem that would support 28,000 engineers in Bhubaneswar — the housing, the schools, the spouse employment, the cultural infrastructure, the peer density, the career-progression depth — does not yet exist. Companies discover that staffing 4,000 positions is possible (through a combination of fresh graduates and mid-career returnees), but scaling to 28,000 would require creating the very ecosystem whose absence is the problem.

Bhubaneswar’s total IT employment — all companies combined — is nearly 35,000. Wipro alone promised 28,000. The gap between what was announced and what was delivered is a precise measure of the distance between aspiration and ecosystem.

And yet, the 35,000 IT jobs that do exist in Bhubaneswar matter. They represent a partial retention mechanism — the difference between zero local IT employment and some. Infosys has roughly 5,000 employees. LTIMindtree operates a 20-acre campus with 2,800 employees and plans for expansion. TCS runs a large development center. These operations are staffed partly by returnees — professionals who spent years in Bangalore or Hyderabad and came back for family reasons, aging parents, or Bangalore fatigue. The IT sector provides a critical return pathway, even if the numbers are modest relative to the outflow.

The honest assessment: Bhubaneswar’s IT sector is a necessary condition for talent retention. It is not a sufficient one. It provides the minimum viable product — a reason, however thin, for some engineers to stay or return. But an MVP is not a product. It needs iteration, scaling, and the kind of sustained investment that transforms a promising start into a self-reinforcing ecosystem. Wipro’s 14 percent fulfillment rate is the measure of how far that iteration has to go.

The Structural Architecture of Talent Extraction

Pull the camera back far enough and the architecture becomes visible. Odisha’s talent extraction operates through five parallel pipelines, each with its own dynamics, each draining a different form of capacity.

Pipeline 1: Engineering. 25,000-35,000 graduates annually. 85-95 percent of top-tier graduates leave. Destination: Bangalore, Hyderabad, Pune, Delhi, international. Loss: economic output, innovation capacity, entrepreneurial potential, network effects.

Pipeline 2: Medical. ~1,725 MBBS graduates annually, plus nursing, dental, pharmacy. Significant specialist drain post-MD/MS. Destination: corporate hospitals in metros, Gulf, international. Loss: healthcare delivery capacity, specialist availability, institutional quality at government hospitals.

Pipeline 3: Civil services. ~62 officers on central deputation. 182 IAS/IPS/IFS posts vacant. Two-thirds of IAS allocation are outsiders with no state connection. OAS as consolation prize. Destination: Delhi, other state cadres. Loss: governance capacity, policy continuity, institutional memory.

Pipeline 4: Academic. 1,404 faculty positions vacant across 17 state universities. Scholars trained in Odisha pursue careers at JNU, Delhi University, IISc. State universities invisible in national rankings. Destination: Delhi, Kolkata, Bangalore research institutions. Loss: knowledge production capacity, policy-relevant research, state-specific expertise.

Pipeline 5: Skilled trades. Kendrapada’s 100,000 plumbers. Ganjam’s 700,000 Surat migrants. The dadan road’s construction workers. Destination: Surat, Delhi, Gulf. Loss: labor power, construction skills, artisanal knowledge. Partially offset by remittances (Ganjam receives Rs 120-124 crore per month).

The five pipelines drain different things, but they share one structural feature: in every case, Odisha bears the cost of producing the human capital and the destination captures the return. The engineering pipeline exports Rs 8-15 crore in lifetime value per elite graduate. The medical pipeline exports healthcare capacity. The civil services pipeline exports governance capacity. The academic pipeline exports knowledge production. The skilled trades pipeline exports labor power.

The compounding damage is the critical point. Each pipeline weakens the state’s ability to address the others. When engineers leave, the IT ecosystem that would retain the next cohort of engineers does not develop. When doctors leave, the healthcare system that would attract professionals back does not improve. When bureaucrats prefer Delhi, the governance reforms that would build the urban ecosystem that would retain everyone else do not happen. When academics leave, the knowledge base that would inform policy does not exist. Each absence cascades into every other absence.

This is compounding in reverse. In investing, compound interest is the most powerful force — small gains, reinvested, generate exponential growth over time. The talent drain is compound interest working in the opposite direction. Small losses, compounded over decades, generate exponential degradation. Every NIT Rourkela graduate who leaves and does not return is not just one person’s departure. It is the absence of the company they might have founded, the 50 employees they might have hired, the tax revenue those employees would have generated, the school that revenue would have funded, the teachers that school would have employed, the students those teachers would have taught, the researchers among those students who might have studied the very problem that caused the first graduate to leave.

The entrepreneur who would have built the company that employed 100 people never builds it. The doctor who would have started the specialty practice that attracted other specialists never starts it. The professor who would have mentored the PhD student who would have started the research lab that attracted the grant that put the university on the map never mentors anyone. Each absence creates a cascade. Compounding works in reverse as well as forward. This is the structural architecture of a state that produces talent for other people’s use.

What Kerala Measures That Odisha Does Not

Kerala provides the sharpest comparative case — not because Kerala has solved brain drain (it has not), but because Kerala has done the one thing Odisha has not: measured the problem and built systems to manage it.

Kerala’s Gulf diaspora numbers 2.2 million people (2023 Kerala Migration Survey). Remittances: Rs 2,16,893 crore in 2023, contributing 23.2 percent of the state’s domestic product. Return migrants: 1.79 million by 2023. The Kerala Migration Survey has been conducted systematically since 1998. The state knows, with reasonable precision, how many people have left, where they have gone, what they earn, what they send back, and what would bring them home.

Odisha’s first comprehensive migration survey was conducted in 2023 — by IIT Hyderabad. Not by an Odisha institution. The survey estimated 1.7 million inter-state migrants, but it captured primarily working-class and semi-skilled migration. The skilled professional migration — IT engineers, doctors, bureaucrats, academics — is less visible in household surveys because these migrants are more likely to have relocated permanently, with their households no longer in Odisha. The true scale of skilled migration is understated by any household-based survey, and Odisha has no dedicated instrument to measure it.

The difference is institutional, not cultural. Kerala treats its diaspora as a strategic asset: it counts them, surveys them, supports them, and captures benefits through formal remittance channels and returnee programs. Kerala’s NORKA (Non-Resident Keralites’ Affairs) department provides welfare support, legal aid, and facilitation for return migrants. The state has built a political economy around its diaspora that acknowledges the reality of departure while maximizing the benefit to the source.

Odisha has an NRO (Non-Resident Odia) Cell. Its economic impact is, charitably, unmeasurable. No state-level NRI/diaspora census has been conducted. No systematic tracking exists of where Odisha’s engineering, medical, or management graduates end up geographically. No survey has asked Odia professionals outside the state what would make them return. No measurement exists of how many have returned and what their experience has been.

The absence of measurement is not passive. It is a policy choice. A state that does not know how many of its graduates leave, where they go, what they earn, and what would bring them back has no basis for intervention. The first step is not a policy. It is a survey. And that survey has not been conducted.

Telangana offers a different lesson: what aggressive IT policy can achieve. Hyderabad accounts for one-third of all IT jobs created in India. 778,121 employees in IT/ITeS, 1,500-plus companies. This was not organic. It was the result of HITEC City (established in the 1990s under Chandrababu Naidu), sustained policy consistency across governments, TS-iPASS incentives, and the network effects of reaching critical mass early. Bhubaneswar started its IT push two decades later and with less intensity.

Tamil Nadu demonstrates multi-city retention: Chennai as the primary IT hub, but Coimbatore emerging with 75,000 tech professionals. The state launched the “Tamil Talents Plan” — globally competitive pay, startup grants, relocation allowances for returnees. An annual conclave matches diaspora scholars with Tamil Nadu institutions. This is a state that has decided retention is a strategic priority, not a cultural hope.

What these comparators share is intentionality. The brain drain problem does not solve itself. Network effects and agglomeration economics mean that talent, once concentrated in a destination, attracts more talent to the same destination. The default trajectory, in the absence of deliberate intervention, is continued drain. Only states that treat retention as a policy priority with dedicated budgets, institutions, and measurement infrastructure have managed to slow or partially reverse the flow.

Odisha has not yet reached this level of intentionality. The state has improved — Bhubaneswar’s IT sector has grown from nothing to 35,000 jobs, the startup ecosystem is nascent but present, AIIMS and IIT provide institutional anchors. But improvement is not the same as intervention. Improvement is what happens when a state grows generally. Intervention is what happens when a state identifies a specific structural problem and builds specific institutional solutions.

The brain drain is identified. The intervention has not begun.

What the Silence Costs

The talent extraction pipeline does not make headlines. It does not produce dramatic images — no photographs of families walking on highways, no rescue operations, no legislative debates. The NIT graduate who takes a Bangalore placement does so privately. The doctor who leaves SCB Medical College does so without fanfare. The IAS officer who files for central deputation does so through bureaucratic channels. The faculty member who accepts an offer from a Delhi university simply stops showing up at Utkal.

This invisibility is precisely what makes the pipeline so damaging. The dadan road is visible, horrifying, and generates political pressure. The Surat corridor was invisible until COVID forced it into public view. But the skilled departure may never generate a crisis moment — because it operates through normal labor market mechanisms, produces outcomes that look like success (high placements, great packages, Odisha’s sons and daughters conquering Bangalore), and affects people who have the least reason to complain.

The damage is cumulative and delayed. Every campus placement season at NIT Rourkela, every IAS allocation list, every medical graduate who packs for Hyderabad, the state loses a little more of its capacity to become the kind of place where the next generation would not need to leave.

In the Leaving series, Chapter 4 identified the skilled departure as potentially the most damaging form of migration — more damaging, in terms of long-run impact on the state’s development trajectory, than the dadan stream or the Surat corridor. The argument was this: the people leaving through the skilled pipeline are the ones with the highest potential to build the industries, institutions, and ecosystems that would generate employment for the millions who leave through the other two streams. The dadan worker who builds a brick kiln in Andhra, the powerloom worker who weaves saris in Surat, and the software engineer who writes code in Bangalore are all building somewhere that is not home. The difference is that the dadan worker mostly had no choice. The engineer did. But from the source state’s perspective, the economic logic is identical: investment in human capital made here, return on investment captured there.

The dadan road is Odisha’s wound. The Surat corridor is Odisha’s scar. But the talent extraction pipeline is the reason the wound does not heal — because it drains the surgeons, the engineers, the administrators who might have built the operating theater.

The Women’s Odisha series noted the migration residual: when the educated men leave, women manage what remains. The Urbanization series identified NIT Rourkela as a brain drain engine and Bhubaneswar’s IT sector as an MVP that has not iterated. The Churning Fire series argued that network effects require local density, and brain drain thins the network below the threshold where collective capacity emerges. Delhi’s Odisha Chapter 1 traced how the Freight Equalization Policy destroyed mineral processing; education policy destroys human capital processing through the same structural logic.

Every series converges on the same point from a different angle. The export factory is not a metaphor. It is a literal description of what Odisha’s education system does. It takes raw material — children. It processes them — through schools, colleges, coaching centers, and placement seasons. It sorts them by quality — JEE, NEET, UPSC, campus recruitment. And it exports the finished product to destinations that capture the lifetime appreciation in value.

There is, I should note with the honesty that Principle 7 demands, a counterargument. Brain drain advocates argue that diaspora networks create “brain gain” — returning knowledge, investment, and connections to the source. This is true for some countries (Israel, Taiwan, Ireland have benefited from returning diaspora). It may become true for Odisha at some future point. But the conditions under which brain drain converts to brain gain are specific: the source must build an ecosystem worth returning to, the diaspora must be systematically engaged, and formal channels must exist for knowledge and capital to flow back. Odisha has none of these conditions in place. I estimate with roughly 70 percent confidence that without deliberate intervention — measurement infrastructure, retention incentives, ecosystem investment, and diaspora engagement at the level Kerala has achieved — the current trajectory of net talent export will continue for at least another decade. This estimate would be wrong if Bhubaneswar’s IT sector unexpectedly crosses the critical mass threshold (possible but, at current growth rates, unlikely before 2035), or if remote work becomes sufficiently normalized to decouple salary from geography at scale (a structural shift that COVID hinted at but did not deliver).

What remains, for now, is the fact. NIT Rourkela’s placement office will post numbers again next December. The celebration will be genuine. The parents will be proud. The students will be relieved. The companies will have secured talent at salaries that are a bargain compared to what the same talent would cost if it had to be produced in Bangalore from scratch.

And Odisha will have done, once more, what it has done for sixty-three years: invested in producing something valuable, and then watched it walk out the door.

Sources

NIT Rourkela Placements

IIT Bhubaneswar

KIIT University

SOA University

VSSUT Sambalpur

XIM Bhubaneswar

Odisha IT Sector

Karnataka IT

Odia Diaspora

Wipro Bhubaneswar

Civil Services Vacancies

OAS/OPS

Judicial Vacancies

Medical Brain Drain

Faculty Vacancies and Higher Education

Education Budget and Costs

Migration Data

Kerala Comparison

Telangana/Hyderabad IT

Salary Comparisons

Startups

Brain Drain National Data

Multiplier Effects

Ganjam Remittances

Scholarships

Tamil Nadu Comparison

Source Research

The raw research that informs this series.