English only · Odia translation in progress

Chapter 1: The 17 Percent

Pull up a satellite image of India at night. The kind NASA releases every few years, where city lights trace the skeleton of an economy against the dark mass of the subcontinent. Start in the south. Tamil Nadu glows like a circuit board — Chennai anchoring a corridor of light that runs through Coimbatore, Madurai, Tiruchirappalli, Salem. Five urban agglomerations above 500,000 people, each burning bright enough to be visible from orbit. Swing northwest to Gujarat. Ahmedabad, Surat, Vadodara, Rajkot — four cities above a million, strung along highways like nodes in a network, each with its own distinct signature: diamonds, textiles, petrochemicals, auto components. The western coast is a continuous ribbon of light from Mumbai through Gujarat’s industrial belt.

Now look at Odisha. The coast is there, 480 kilometers of it, but the light is thin. A faint cluster where Bhubaneswar and Cuttack sit twenty-five kilometers apart — the twin cities, the only significant urban concentration in a state of forty-six million people. Below that, the dimmer glow of Berhampur. To the northwest, the isolated pinpoint of Rourkela, a steel plant dropped into tribal country in 1955, still burning alone after seven decades. Sambalpur, Balasore, Puri — scattered specks, none bright enough to be called a city in any state with a functional urban system.

Then the darkness. Western Odisha — Kalahandi, Bolangir, Nuapada, Koraput, Malkangiri — is nearly invisible from space. This is where a third of the state’s population lives. This is where the sardars come in August with cash advances, where the Shramik Special trains fill up every October, where the brick kiln corridors begin. The darkness is not a metaphor. It is an economic fact rendered in photons. Where there are no cities, there are no lights. Where there are no lights, there is no concentrated economic activity. Where there is no concentrated economic activity, there are no jobs. And where there are no jobs, people leave.

The number that explains the darkness is 16.68 percent.

That was Odisha’s urbanization rate in the Census of 2011 — the last census conducted, since the 2021 exercise was postponed and remains incomplete as of 2026. Roughly seven million out of forty-two million people lived in places the Census classified as urban. By 2025, based on population projections from the National Commission on Population and continued slow growth trends, the estimated rate is approximately 17 to 18 percent.

Seventeen percent. For every six people in Odisha, five live in villages.

This chapter is about that number — what it means, how it got this way, and why it persists. The explanation is not a single cause. It is a stack of absences, layered over two centuries, each one making the next one worse. Think of it as a platform problem, the kind that software engineers understand intuitively: you cannot build applications without an operating system. You cannot run a functioning urban economy without the institutional, commercial, and infrastructural platform that makes urban life possible. Odisha never built the platform. The 17 percent is what happens when you try to run a modern economy on bare metal.

The Number

Start with what 16.68 percent actually means in comparative terms, because the number by itself is abstract. Context makes it visceral.

Tamil Nadu: 48.4 percent urban. Kerala: 47.7 percent. Maharashtra: 45.2 percent. Gujarat: 42.6 percent. Karnataka: 38.7 percent. The all-India average: 31.2 percent. Even West Bengal, with its own history of economic stagnation: 31.9 percent. Jharkhand, carved from the same mineral-rich, feudal-legacy belt of eastern India: 24.1 percent. Chhattisgarh, another peer state: 23.2 percent.

Odisha: 16.7 percent. Ranked 31st among Indian states and Union Territories. Only Bihar at 11.3 percent and Assam at 14.1 percent are worse among major states. The gap with the national average is not a gap — it is a chasm. India was 31.2 percent urban in 2011. Odisha was barely half that. And the chasm is widening, not narrowing: the difference between Odisha and India was about 11 percentage points in 1941, about 12 points in 1981, and 14.5 points in 2011. Seventy years, and the state fell further behind.

The absolute numbers make the structural shallowness concrete. Odisha has no million-plus city. Not one. Bhubaneswar’s Urban Agglomeration reached roughly 886,000 in 2011 — it fell short. The combined Bhubaneswar-Cuttack metropolitan area reached approximately 1.7 million in 2011 and an estimated 1.86 million by 2018. That twin-city agglomeration is the only significant urban concentration in the entire state. After that, the drop-off is brutal: Berhampur at 356,000, Rourkela at 311,000, then nothing above 200,000 except Puri, which is a temple town, not an economic hub.

Compare this with Tamil Nadu’s five urban agglomerations exceeding 500,000 — Chennai, Coimbatore, Madurai, Tiruchirappalli, Salem — each with a distinct economic engine: IT and automobiles, textiles and pump manufacturing, trade and education, heavy engineering, steel and manganese. Or Gujarat’s four cities above a million — Ahmedabad, Surat, Vadodara, Rajkot — each a self-sustaining economic ecosystem. These states did not just have more urban people. They had urban systems — hierarchies of cities at different scales, each generating its own gravity, each pulling in migrants, each creating the conditions for the next tier of towns to grow.

Odisha has a hierarchy that looks like a cliff. Bhubaneswar-Cuttack at the top, a massive gap, then a handful of medium towns, then nothing. The “missing middle” of the urban system — the Tier 2 and Tier 3 cities that should exist between the capital and the villages — was never built. There is no city in the 500,000 to 1,000,000 range. There is no equivalent of Coimbatore, which grew from 80,000 in 1901 to 2.15 million in 2011 on the back of textiles, engineering, and manufacturing. There is no equivalent of Surat, which grew from a trading town to a 4.5-million-person diamond-and-textile powerhouse entirely through private industry. There is no equivalent of Pune, which escaped Mumbai’s shadow to become an independent IT-and-auto hub.

Odisha’s urban system is not merely small. It is structurally shallow. And structural shallowness is a different kind of problem from small size. You cannot fix it by growing the one city you have. You need to build the entire hierarchy — and the hierarchy requires an operating system that does not exist.

The Zamindari System and the Missing Merchant

To understand why the operating system was never built, you have to go back to 1793. That is when Lord Cornwallis introduced the Permanent Settlement in Bengal, Bihar, and Odisha — a land revenue system that converted existing revenue agents into hereditary proprietors of land. In Odisha, the system covered nearly 70 percent of the area across six districts: Cuttack, Puri, Balasore, Ganjam, Koraput, and Sambalpur. Through a proclamation of September 15, 1804, the British turned chaudhuris, kanungoes, and mukadammas into zamindars with permanent, heritable rights to extract rent from the land.

The economic logic of the Permanent Settlement was straightforward: fix the government’s revenue demand in perpetuity, give the zamindar the right to collect whatever the land could bear above that amount, and let the difference be the zamindar’s incentive to improve the land. In theory, this should have created an investing landlord class. In practice, in Odisha, it created an extracting one. Zamindars collected rent. They did not invest in towns. They had no incentive to — their wealth came from land, not commerce. Unlike Bengal, where the Permanent Settlement produced a class of absentee urban landlords concentrated in Calcutta — merchants and bankers who at least resided in cities and created urban demand — Odisha’s zamindars largely stayed tied to their estates. When they did live in towns, they contributed nothing resembling a commercial bourgeoisie.

But the zamindari system, extractive as it was, is only half the story. The other half is more important and harder to see, because it is about something that was absent rather than something that was present.

Odisha had no mercantile caste.

This is the structural fact that explains more about the 17 percent than any policy, any colonial decision, any post-independence failure. Every Indian state that urbanized successfully had a community whose identity, social networks, accumulated capital, and generational knowledge were organized around trade. Odisha did not.

Gujarat had the Banias — an estimated 2.7 million strong in the state, overrepresented in urban centers, historically specialized in wholesale trade of grains, spices, and textiles. The Bania networks were not just commercial. They were institutional. Caste-based connections provided trust in an era before contract law. They provided informal credit when no banks existed. They provided market information — which towns were buying, which routes were safe, where prices were moving. They provided coordinated risk-sharing, where a group of traders would underwrite each other’s ventures. These networks were the operating system on which Gujarat’s textile industry, salt trade, shipbuilding, and merchant banking ran. The cities of Ahmedabad, Surat, Vadodara, and Rajkot did not emerge because the government planned them. They emerged because generations of merchants, trading through caste networks that functioned as institutional infrastructure, concentrated economic activity in nodes that grew into cities.

Tamil Nadu had the Nattukottai Chettiars — the Nagarathars — a mercantile community from Pudukkottai and Sivaganga districts that expanded from inland cotton and rice trade in the 17th century to money-lending and banking across colonial Southeast Asia by the 18th. Their accumulated wealth built the Chettinad mansions, each taking years to complete, employing carpenters, masons, tile-makers, and artisans across rural Tamil Nadu. The Chettiars were not the only mercantile community in Tamil Nadu — there were also the Nattukottai Vellalars, the Komati traders, the Chettiar banking houses in Madras — but the point is that Tamil Nadu had multiple commercial communities whose activity created the demand for urban infrastructure: warehouses, banks, transport nodes, markets, worker housing.

Bengal had the bhadralok bourgeoisie — professionals, intellectuals, and rentier landlords concentrated in Calcutta who, for all their extractive tendencies, created an urban cultural and commercial ecosystem that made Calcutta the “second city of the British Empire.”

Odisha had Brahmins. It had Karans, the scribe-administrator caste. It had Khandayats, the landed military cultivators. All three were tied to agrarian and administrative roles. None of them were merchants. None of them had generational knowledge of trade, networks organized around commerce, or accumulated commercial capital. Towns in Odisha existed as administrative seats or temple centers — Cuttack was the capital, Puri was the deity’s home — not as commercial hubs.

This is where the software metaphor stops being decorative and becomes diagnostic. In computing, a platform provides four things: hardware abstraction (making physical resources accessible through a standard interface), an API layer (connecting producers to consumers), a developer ecosystem (skilled people who know how to build on the platform), and network effects (more users attract more services, which attract more users). A mercantile caste — in Gujarat, Tamil Nadu, Bengal, or anywhere else where commerce drove urbanization — functioned as exactly this kind of platform.

The Banias provided hardware abstraction: they turned raw resources (cotton, salt, grain, minerals) into tradeable commodities by standardizing quality, establishing weights and measures, and creating market conventions. They provided the API layer: trade routes, credit networks, and commission agencies that connected producers in villages to markets in cities. They provided the developer ecosystem: each generation trained the next in accounting, trade, risk assessment, and network management. And they generated network effects: a market town with ten Bania firms attracted more suppliers, more buyers, more services, and more Bania firms — a self-reinforcing cycle that turned trading posts into cities.

Odisha had the hardware — land, rivers, minerals, a 480-kilometer coastline, location on the eastern trade route that had connected Kalinga to Southeast Asia for two millennia. What it lacked was the operating system: the institutional layer that turns resources into concentrated economic activity. Without a commercial class to abstract the resources, connect them to markets, train the next generation of traders, and generate network effects, the hardware sat unused. Towns did not form around trade because there were no traders. Markets did not grow because there were no market-makers. Cities did not emerge because there was no commercial force pulling people into dense settlements.

The zamindari system made this worse by ensuring that the dominant economic institution — land ownership — was organized around extraction, not investment. But even without zamindari, the absence of a mercantile platform would have constrained urbanization. You can see this in the princely states, which had different land systems but the same urban outcome: no cities.

[Confidence note: The claim that Odisha had “no mercantile caste” is an overstatement if taken literally — there were small trading communities, including Komatis in southern Odisha and some Marwari migration into Cuttack. The more precise claim is that Odisha lacked a large, indigenous mercantile community with the generational depth, network scale, and institutional thickness of Gujarat’s Banias or Tamil Nadu’s Chettiars. This more precise claim holds at high confidence, ~85 percent.]

Twenty-Six Princely States: The Fragmentation Tax

The zamindari system covered the British-administered districts. But nearly half of what became modern Odisha was something else entirely: twenty-six princely states, the Garajat, each with its own raja, its own revenue system, its own borders, and its own tiny administrative capital.

When the modern province of Odisha was formed on April 1, 1936, it unified the British districts with these twenty-six feudatory states: Mayurbhanj, Kalahandi, Sambalpur, Dhenkanal, Keonjhar, Gangapur, Patna, Sonepur, Bamra, Nayagarh, Boudh, and others. The British had categorized them into A, B, and C sections in 1937 based on administrative significance. The A category included Dhenkanal, Keonjhar, Mayurbhanj, Bamanda, Boudh, Gangapur, Patna, Kalahandi, Sonepur, Sareikela, and Nayagarh. These were feudal administrations where the raja’s authority was paramount, revenue came from land and forest, and there was no structural reason to invest in urban infrastructure.

Think of this as a fragmentation problem — the kind that computer scientists recognize immediately. If you have a single large system, you can build infrastructure that serves the whole thing: roads that connect markets, railways that link producers to ports, administrative systems that standardize regulations. But if you have twenty-six independent systems, each with its own incompatible standards, its own borders, its own rules, then every transaction that crosses a boundary pays a tax — not necessarily a monetary one, but a friction tax. Moving goods from Keonjhar to Cuttack meant crossing princely state borders where different revenue systems applied. A merchant in Dhenkanal could not easily extend credit to a buyer in Kalahandi because the legal and institutional frameworks were different. There was no common market. There was no integrated transport network. There was no reason for a city to emerge at the intersection of trade routes because there were no trade routes — just isolated administrative capitals serving miniature sovereigns.

The princely state capitals — Dhenkanal, Keonjhar, Mayurbhanj town, Kalahandi’s Bhawanipatna — remained small towns precisely because they had been designed as seats of power, not as economic hubs. A raja needed a palace, a court, a small administration, perhaps a temple. He did not need — and had no incentive to create — a market town, a manufacturing center, or a commercial district. The population of these capitals was determined by the size of the raja’s establishment, not by the pull of economic opportunity.

The contrast with Rajasthan’s princely states is instructive. Rajasthan also had dozens of princely states — but its largest ones were genuinely large: Jaipur, Jodhpur, Udaipur, Bikaner. These were not miniature sovereignties. They were substantial kingdoms with populations in the millions, territories the size of European countries, and capitals that functioned as genuine urban economies. Jaipur had a planned city laid out in 1727 by Maharaja Sawai Jai Singh II, with commercial streets, artisan quarters, and a merchant class. Jodhpur had its own trading economy connected to overland routes. These were states large enough and wealthy enough to sustain urban life.

Odisha’s princely states were, with the partial exception of Mayurbhanj, too small. Twenty-six miniature kingdoms where two or three large ones might have generated urban economies. The fragmentation was the tax. And the tax was paid in undeveloped towns.

The merger came late. On October 16, 1947, Chief Minister Harekrushna Mahatab convened a meeting in Sambalpur where he convinced the rajas of the necessity of merging with the democratic state for peace and order. The merger was completed by 1949. But incorporating these territories into Odisha added vast rural land mass without adding urban nodes. The institutional patchwork — different land records, different revenue systems, different administrative traditions — took decades to harmonize. The princely state territories of western and southern Odisha remain, to this day, the least urbanized parts of the state. The feudal past left no commercial towns, no market infrastructure, and no urban culture in these areas.

The fragmentation tax was not just historical. It compounded. Because the princely states had no urban tradition, they also had no urban institutions — no municipal governments, no land markets, no building regulations, no commercial associations. When the merger happened, these areas needed to build every piece of urban infrastructure from scratch. But the resources and political attention went to the existing urban centers — Bhubaneswar, Cuttack, the British-administered coast. Western Odisha was merged administratively but remained a separate country economically. In many ways, it still is.

Colonial Infrastructure: Extraction, Not Connection

The British built things in Odisha. They built railways. They built a lighthouse. They built administrative buildings in Cuttack. The question is not whether they built, but what they built for.

The Bengal-Nagpur Railway connected Cuttack to Calcutta and Madras by the last decade of the 19th century. The alignment was not accidental. The railway ran through the mineral belt — coal around Talcher and Jharsuguda, iron ore in Keonjhar and Sundargarh, chromite in Jajpur. A significant portion of the revenue of East Coast Railways, which serves more than 70 percent of the state today, still comes from transporting coal, iron ore, and other minerals. The railway was built to move raw materials out, not to move people between cities or to connect agricultural producers to urban markets.

This is a pattern that software engineers would recognize as a one-way data pipeline. The system was optimized for extraction: ore goes in at one end, leaves the state at the other, and the value is captured somewhere else. A two-way system — one that also brings goods, services, investment, and people back — would have created the kind of connections that build cities. But that was not what the British wanted. They wanted coal for their factories in Calcutta, iron for their shipyards, chromite for their steel alloys. The railway was the extraction pipe.

Consider the contrast with what railways did elsewhere. The Mumbai-Thane railway, India’s first, opened in 1853 and helped transform Bombay from a colonial trading post into Asia’s premier commercial city. The railways that crisscrossed the Punjab connected agricultural producers to the port of Karachi and the markets of Delhi, creating a chain of market towns — Ludhiana, Jalandhar, Amritsar — each growing because it sat on the route between producer and consumer. Even in Bengal, the railways concentrated commercial activity in Calcutta, creating a primate city that at least had an organic economic base in jute, trade, and finance.

In Odisha, the railway created no city. It passed through Cuttack, which was already the administrative capital, and it connected the mines to external markets. The internal network — the connections between Odisha’s own towns, between agricultural hinterlands and potential market centers — was never the priority and was never adequately built. The Talcher-Bimlagarh railway line, approved in the 1950s to connect the coal belt to the iron ore belt within Odisha, remained incomplete for over seventy years. That single unfinished line captures the entire colonial and post-colonial story of Odisha’s infrastructure: what serves extraction gets built; what serves internal integration does not.

No major colonial port city emerged. The three presidency capitals — Bombay, Calcutta, Madras — each developed as massive commercial hubs because the colonial government invested in port infrastructure, commercial institutions, and administrative machinery concentrated in one place. The resulting agglomeration effects created self-sustaining urban economies that survived independence and continued growing. Odisha had Cuttack as its capital, but Cuttack was never on this scale. It sat at the confluence of the Mahanadi and Kathajodi rivers, was vulnerable to floods, and served primarily as an administrative center.

The British constructed a harbor at False Point, north of present-day Paradip, in 1819. The False Point Lighthouse, built in 1838 — India’s oldest functioning lighthouse — guided British ships near the Mahanadi estuary. But it guided them past, not to. Paradip as a port was not developed until 1958, well after independence. It was declared India’s eighth major port only in 1966 — the first major port on the East Coast commissioned in independent India. The colonial era gave Odisha a lighthouse that watched ships sail by. It did not give the state a port city.

The net effect of colonial infrastructure was to connect Odisha to the imperial economy as a resource periphery. Raw materials flowed out. Administrative orders flowed in. But the internal connections — the roads, railways, and port facilities that would have linked Odia producers to Odia markets, that would have created the transport layer on which urban economies grow — were not built. The colonial infrastructure was an extraction API: it exposed Odisha’s resources to external consumers without providing the internal routing that builds cities.

Rice Monoculture and the Government Town

If the zamindari system explains the missing commercial class, and the princely states explain the missing urban nodes, and colonial infrastructure explains the missing transport connections, then the agricultural economy explains the missing surplus.

Cities grow when the countryside produces more food than it consumes, freeing labor and capital to move into non-agricultural activities. This is the basic urbanization equation, and it has held since the Neolithic revolution. The question for Odisha is: what kind of agricultural economy did it have, and did that economy generate the surplus that feeds urban growth?

The answer: rice monoculture, subsistence-level, monsoon-dependent, with no cash crop revolution.

Rice covers approximately 69 percent of Odisha’s cultivated area and about 63 percent of the area under food grains. This is not diverse agriculture. This is a single-crop economy. And the crop is produced at subsistence yields: 2.3 tonnes per hectare, below the national average of 2.8 tonnes. Almost 70 percent of rural households have holdings of less than one hectare. About 82 percent are small, marginal, or landless. The irrigation deficit — only 35 to 40 percent of cultivated land is irrigated — means that output is tied to the monsoon. A good monsoon means a decent harvest. A bad monsoon means hunger. There is no buffer, no diversification, no stable surplus.

The cash crop revolution that drove urbanization in other states never happened here. In Maharashtra, sugarcane created an entire tier of market towns. Two hundred and three sugar crushing factories in 2022-23 alone, concentrated in the belt of Sangli, Kolhapur, Pune, Satara, Solapur, and Ahmednagar. Sugar cooperatives became institutions of political and economic power. The towns that grew around them became significant urban centers — not because the government planned them, but because a cash crop created commercial activity that demanded warehouses, transport, labor, processing facilities, banks, and services. Each factory was the nucleus of a small urban economy.

In Gujarat, cotton drove the textile mills of Ahmedabad. Groundnut supported oil processing industries across small towns. In Kerala, spice cultivation — pepper, cardamom, rubber, tea, coconut — created a commercialized rural economy that, combined with Gulf remittance income, drove ribbon urbanization across the state.

Odisha grew rice. Subsistence rice. And when the monsoon failed, the rice failed too, and the state that would later become internationally notorious for starvation deaths in Kalahandi — a district that was, paradoxically, a net exporter of paddy even during famine, because the surplus went to traders while local people starved — was not generating the kind of agricultural surplus that builds towns.

This absence interacted with another structural feature to produce Odisha’s characteristic urban form: the government town.

When agriculture does not generate surplus, and there is no commercial class to create market-driven towns, and colonial infrastructure does not build internal connections, what does create urban settlements? The answer, in Odisha, is government. The government creates a capital, posts bureaucrats there, builds offices and housing, and the resulting demand for services creates a town. The government builds a steel plant, posts engineers and workers, builds colonies, and the resulting demand creates another town. The government establishes a university, and the resulting student population creates demand for housing, food, and retail.

Bhubaneswar is the archetype. It replaced Cuttack as Odisha’s capital on April 13, 1948, designed by German architect Otto Konigsberger, one of modern India’s first planned cities alongside Jamshedpur and Chandigarh. Its population trajectory tells the story: 16,000 in 1951, roughly 38,000 in 1961, 105,000 in 1971, 219,000 in 1981, 411,000 in 1991, 648,000 in 2001, and approximately 885,000 in 2011. Impressive growth — but driven by what? The State Secretariat. The State Assembly. The Raj Bhavan. Dozens of government departments, boards, and corporations. Government employment, direct and indirect, is the backbone of the economy. The city was subdivided into residential units, each designed with a high school, shopping centers, dispensaries, and play areas, and “while most of the units house government employees, Unit V houses the administrative buildings.” It was a city designed by and for the bureaucracy.

Rourkela is another variant: a government steel plant dropped into Sundargarh district in 1955, with government housing colonies arranged around the factory, growing to 259,000 by 2011. Sambalpur: an administrative and educational center. Berhampur: the commercial center of southern Odisha, but still fundamentally a government-district-headquarters town. Even Cuttack, the former capital, derived much of its economic activity from courts, government offices, and administrative functions.

Compare these with cities built by commerce. Coimbatore in Tamil Nadu grew on textiles and pump manufacturing — private industry, not government posting. Surat grew on diamonds and textiles — private enterprise, not a bureaucratic apparatus. Rajkot in Gujarat grew on engineering and auto components. These cities grew as fast as their industries grew, which is to say, as fast as market demand grew. Government towns grow as fast as government budgets grow — which is to say, slowly, incrementally, and always limited by the fiscal capacity of the state.

This is the government-town trap. When the government is the economy, the economy grows at the pace of government spending, not at the pace of market opportunity. Government budgets in a poor state are constrained. Revenue comes from taxes on a predominantly rural, low-income population. The fiscal space for expanding government employment — the only urban-job-creation mechanism in the system — is narrow. Every government job created in Bhubaneswar depends on revenue collected from the rice farmers of Kalahandi and the forest dwellers of Koraput. The ceiling is low, and it binds.

The IT sector, which has arrived in Bhubaneswar since the 2000s, provides a glimmer of non-government economic activity. TCS Kalinga Park has capacity for 7,000 seats. Infosys, Wipro, Tech Mahindra, and Capgemini have offices. Software exports reach approximately Rs 7,500 crore. But this sector sits on top of the administrative-city base rather than transforming it. The IT campuses are typically at the city’s periphery — Infocity, Info Valley — enclaves that generate high-value employment but do not reshape the city’s fundamental character the way IT reshaped Bangalore from a military-pensioner town to a global tech hub. The reason: scale. Odisha’s Rs 7,500 crore in software exports compares to Karnataka’s Rs 5 lakh crore and above. The IT sector in Bhubaneswar is a promising overlay, not yet a transformation.

The Freight Equalization Policy: Forty-One Years of Exported Urbanization

If there is a single post-independence policy that did the most damage to Odisha’s urban trajectory, it is the Freight Equalization Policy. The FEP was adopted by the Union government in 1952 and maintained until 1993. For forty-one years — the exact period when India was urbanizing most rapidly — this policy systematically destroyed the incentive for value-addition industries to locate in Odisha.

The mechanics were simple. The government subsidized the transportation of key raw materials — coal, iron ore, bauxite, limestone, mica — so that industrialists could obtain these minerals at the same price anywhere in India, regardless of proximity to the source. If you wanted to build a steel plant, the iron ore cost the same whether you were in Jajpur (sitting on top of the deposit) or in Pune (a thousand kilometers away). If you wanted to build an aluminium smelter, the bauxite cost the same in Koraput (the source) and in Tamil Nadu (the consumer market).

The structural consequence was devastating. Odisha sits on 28 percent of India’s iron ore reserves, 98 percent of its chromite, 51 percent of its bauxite, and significant coal deposits. Under normal market conditions, proximity to raw materials is a massive locational advantage for processing industries. A steel plant near the iron ore mine saves transport costs, which translates into lower production costs, which translates into competitive advantage. The FEP erased that advantage entirely.

Why would a steel plant, an aluminium smelter, or a cement factory locate in Odisha when raw materials could be shipped to Gujarat, Maharashtra, or Tamil Nadu at flat cost? The answer: it would not. And it did not. The processing industries that Odisha’s mineral wealth should have supported located in other states — closer to consumer markets, closer to ports, closer to existing industrial ecosystems. The minerals were extracted in Odisha and processed elsewhere. The value addition — and the urban employment that comes with it — accrued to Gujarat, Maharashtra, and Tamil Nadu.

Each processing plant that located elsewhere instead of in Odisha represented a town that was never built. A steel plant does not just employ steel workers. It generates an entire ecosystem: ancillary industries (refractory bricks, industrial gases, rollers, castings), service providers (maintenance, transport, catering, housing), retail and commercial activity, schools, hospitals, entertainment. A single steel plant can be the nucleus of a town of 100,000 to 300,000 people. Over forty-one years, the FEP exported dozens of potential factory towns from Odisha to other states.

The timing is the cruelest part. The FEP operated from 1952 to 1993 — exactly the decades when the national urbanization rate doubled from 17.3 percent in 1951 to 25.7 percent in 1991. Odisha’s urbanization during the same period went from approximately 4.1 percent to 13.4 percent — a gain of 9.3 percentage points. But this modest gain masks the fundamental deformation. The urbanization that did occur was driven by government employment in Bhubaneswar and a single government steel plant in Rourkela, not by the broad-based industrial development that mineral endowment should have supported. The FEP did not merely slow Odisha’s urbanization. It exported the urbanization that Odisha’s mineral wealth should have produced.

[Confidence note: The counterfactual — what Odisha’s urbanization would have looked like without the FEP — is inherently uncertain. The claim that the FEP was the single most damaging post-independence policy for Odisha’s urbanization potential is held at ~75 percent confidence. Alternative candidates include the broader failure of land reform, the lack of irrigation investment, and the general neglect of eastern India in industrial licensing. The FEP’s impact is hard to isolate from these other factors. What can be said with higher confidence (~90 percent) is that the FEP systematically disadvantaged mineral-producing states for four decades.]

The Migration Paradox

Here is the fact that should stop everyone who thinks Odisha’s low urbanization reflects something about Odia people — some cultural preference for rural life, some lack of ambition, some inability to adapt to city living.

Two to five million Odias live outside the state, building cities elsewhere.

The Ganjam-to-Surat corridor alone accounts for 500,000 to 800,000 workers. An estimated six lakh IT professionals from Odisha work in Bangalore. Kendrapada district sends 100,000 plumbers to Gulf countries and Indian metros — they built the sanitary systems for the 2022 FIFA World Cup stadiums in Qatar. Workers from Balangir and Nuapada fill brick kilns across Andhra Pradesh and Telangana. During COVID-19, when the Shramik Special trains started running, 358,401 people arrived in Odisha by train alone, another 185,504 by bus, private vehicle, and on foot, and one million registered on a portal saying they wanted to come home.

Odias are not anti-urban. They are enthusiastic urbanizers. They just urbanize other states.

This is the migration paradox, and it connects directly to the platform metaphor. In the software world, developers do not build applications on untested, undocumented, infrastructure-free operating systems. They build on iOS and Android — platforms with tools, libraries, documentation, user bases, and distribution channels. The platform provides the conditions that make development productive. Without the platform, even talented developers cannot create viable products.

Talent goes where the platform exists. Odia workers go to Surat because Surat has the textile platform: factories, housing (however cramped), established supply chains, caste-based labor networks, wage levels that exceed what Odisha offers. Odia engineers go to Bangalore because Bangalore has the IT platform: a critical mass of tech companies, a talent pool that enables specialization, venture capital, a startup ecosystem, salary levels three to five times what Bhubaneswar offers for equivalent roles. Odia plumbers go to Qatar because Qatar has the construction platform: massive infrastructure projects, wages of Rs 50,000 to Rs 1,00,000 per month, demand for precisely the skills that Kendrapada’s training pipeline produces.

None of these people chose to leave Odisha because they dislike Odisha. They left because Odisha lacks the platform. There are no textile factories in Ganjam that could employ 700,000 workers. There is no IT ecosystem in Bhubaneswar that could absorb six lakh engineers at competitive salaries. There is no construction boom in Odisha that could employ 100,000 plumbers at Gulf-level wages.

The migration paradox creates a vicious cycle. The people who leave are, by definition, the people with the most initiative, the most willingness to take risks, the most capacity to adapt. They are the people who, if they stayed, would be the most likely to start businesses, create jobs, build institutions, and drive urban growth. Their departure depletes the very human capital that Odisha needs to build the platform. And without the platform, the next generation of ambitious Odias also leaves. The cycle is self-reinforcing.

The remittance data makes this concrete. Ganjam migrants alone send home an estimated Rs 120 crore per month — roughly Rs 1,440 crore per year. But this money flows into rural construction (houses that stand empty because the owner lives in Surat), gold, and social expenditure. It does not flow into businesses, factories, or commercial infrastructure that would create urban jobs and attract people to Odia towns. The remittance economy creates “construction without presence” — large houses in villages built by absent owners. Neither the origin nor the destination gains permanent urbanization from this pattern.

Kerala’s contrasting experience is revealing. Gulf remittances flowed into a state with high literacy, health consciousness, and demand for urban services. The result was ribbon urbanization — the entire state acquiring urban characteristics through improved housing, commercial establishments, health facilities, and educational institutions. Kerala’s urbanization jumped from 25 percent in 2001 to 47.7 percent in 2011, driven largely by census-town reclassification of settlements that had become functionally urban through remittance-funded development. But Kerala’s remittances arrived in a society that demanded urban services. Odisha’s remittances arrive in districts — Ganjam, Bolangir, Kalahandi — where literacy is lower, commercial infrastructure is thinner, and the institutional platform for converting money into urban development does not exist.

The migration paradox is the platform problem in human form. Odisha produces the people who build cities. It exports them. They build cities elsewhere. And the state that produced them remains 17 percent urban.

Census Towns: The Cities That Do Not Know They Are Cities

There is a category of settlement in India that reveals the urbanization problem from the other direction. Not the absence of cities, but the presence of cities that the governance system refuses to recognize.

In the 2011 Census, Odisha had 223 urban areas. Of these, 107 were statutory towns — governed by Urban Local Bodies such as municipal corporations, municipalities, or Notified Area Councils. The remaining 116 were census towns: settlements that met the Census of India’s criteria for “urban” — population above 5,000, population density above 400 per square kilometer, at least 75 percent of male working population in non-agricultural pursuits — but were still governed by gram panchayats. Urban by function. Rural by governance.

More than half of Odisha’s urban areas lack urban governance.

The consequences of this classification gap are practical and severe. A census town of 8,000 people, where most men work in shops, transport, or small manufacturing, is governed by a gram panchayat designed for agricultural villages. The panchayat has no mandate for urban-density service delivery: no sewage system, no solid waste management scaled for non-agricultural settlements, no land-use zoning, no building regulations designed for density. The settlement is ineligible for Smart City, AMRUT, or other urban development missions because it is not officially a city. It does not get rural development funds either, because its population is not primarily agricultural. It is trapped in a classification gap — too urban for rural programs, too rural for urban ones.

Nationally, the census town phenomenon is massive. India had 2,774 census towns in 2011, up from 1,362 in 2001 — a doubling in a single decade. The central government has asked states to convert 3,784 census towns into statutory Urban Local Bodies. But political resistance is fierce, because conversion means imposing urban taxes on populations accustomed to rural tax levels, and it means losing access to rural development funds like MGNREGA.

Odisha’s 116 census towns are the hidden urbanization — the urbanization that is happening despite the absent platform, despite the lack of industrial investment, despite everything documented in this chapter. These are places where economic activity has shifted from agriculture to non-agricultural work, where population densities have risen to urban levels, where functional urbanization has occurred organically. They represent people voting with their feet — or, more precisely, voting with their labor — for an urban life that the governance system does not acknowledge.

The 116 census towns also suggest that Odisha’s “real” urbanization rate is higher than the official 16.68 percent. If these settlements were properly governed as urban areas, with urban infrastructure, urban services, and urban investment, they could become the nuclei of the missing Tier 3 and Tier 4 cities that the state’s urban hierarchy needs. They are seeds. But seeds planted in soil without water or sunlight do not grow into trees. Without governance recognition, without infrastructure investment, without institutional support, the census towns remain trapped — functionally urban, formally rural, and structurally neglected.

Kerala’s experience with census towns is instructive. Kerala’s dramatic jump from 25 percent to 47.7 percent urbanization between 2001 and 2011 was almost entirely driven by the reclassification of settlements as census towns. Had these settlements been ignored, only 21.9 percent of the state would have been counted as urban. The reclassification did not create new cities — it recognized existing ones. Odisha has a similar pool of functionally urban settlements waiting for recognition. The difference is that Kerala’s census towns existed within a state with high literacy, health infrastructure, and commercial activity. Odisha’s census towns exist within a state that still lacks the institutional platform for urban development.

The Platform That Was Never Built

The 17 percent is not a mystery. It is not the result of a single policy failure, a single historical accident, or a single missing institution. It is the result of every layer of the urban platform being absent simultaneously, each absence making the others worse.

A functioning urban platform — like a computing platform — provides four layers.

The first layer is hardware abstraction: land markets that make location accessible. In a computing platform, hardware abstraction means that a developer does not need to know the specific details of the processor, the memory chips, or the storage device. The operating system presents a standard interface. In an urban economy, the equivalent is a functioning land market: clear titles, efficient transaction processes, land-use regulations that allow conversion from agricultural to urban use, development authorities that provide serviced plots. Odisha’s zamindari system prevented the emergence of land markets. The zamindars controlled land as a source of rent, not as a tradeable asset. Post-independence land reform was incomplete — the Estates Abolition Act took twenty-three years to implement fully. Today, land records in peri-urban Odisha remain fragmented, contested, and poorly digitized. The BDA’s Town Planning Scheme through land pooling is a positive innovation but remains limited in scale. The first layer of the platform was never built.

The second layer is the API: transport that connects producers to markets. In computing, an API lets different software components communicate — a payment service talks to a shopping cart, a map service talks to a ride-hailing app. In an urban economy, transport infrastructure performs the same function: railways connect mines to factories, highways connect factories to markets, ports connect markets to the world. Colonial infrastructure built a one-way extraction API — minerals out, nothing back. Post-colonial infrastructure continued this pattern: the Talcher-Bimlagarh line, which would have connected the coal belt to the iron ore belt within Odisha, remained incomplete for seventy years. The internal transport network — the connections between Odia towns, between agricultural hinterlands and market centers — was never prioritized. Paradip Port was not developed until 1958 and declared a major port only in 1966. The Freight Equalization Policy made internal transport irrelevant by erasing the locational advantage of processing near the source. The second layer was never built.

The third layer is the developer ecosystem: skilled workers who know how to build on the platform. In computing, a platform’s success depends on the number and quality of developers who build applications for it. iOS succeeded because millions of developers learned Swift and Objective-C and built apps. Android succeeded because millions more learned Java and Kotlin. In an urban economy, the equivalent is a workforce with the skills to staff factories, run businesses, provide professional services, and create economic activity. Odisha’s absent mercantile caste meant there was no generational accumulation of commercial knowledge. The educational institutions — NIT Rourkela, KIIT, IIT Bhubaneswar — produce engineering talent, but the talent leaves because there is no platform to build on. The developer ecosystem exists, but it develops for other platforms. It writes code for Bangalore’s operating system, not Odisha’s. The third layer was built but immediately exported.

The fourth layer is network effects: more users attract more services, which attract more users. This is the most powerful force in platform economics and the most devastating when absent. Network effects create virtuous cycles: a city with more firms attracts more workers, which attracts more firms, which attracts more workers. Bangalore’s IT ecosystem exhibits textbook network effects — each new company that sets up there makes the city more attractive to the next company, because the talent pool is deeper, the subcontracting options are wider, the venture capital is more available. Odisha’s urban system generates the opposite: negative network effects, or vicious cycles. Because there are few firms, few workers stay, which means fewer firms come, which means fewer workers stay. The migration paradox is a network-effects problem: talent leaves because the network is too thin, and the network stays thin because talent leaves.

Without all four layers, the platform does not function. You cannot build applications without an operating system. You cannot build a functioning urban economy without land markets, transport networks, a skilled workforce, and network effects operating simultaneously. Remove any one layer and the system degrades. Remove all four and you get 17 percent.

The layered nature of the problem is why no single intervention works. Building a Smart City in Bhubaneswar addresses pieces of the first layer (land and infrastructure) but does not create the API (internal transport network), the developer ecosystem (the workforce leaves anyway), or the network effects (one city cannot generate state-wide urbanization). The IT sector addresses pieces of the third layer (skilled employment) but at a scale too small to trigger network effects. AMRUT improves infrastructure in nine cities but does not address why those cities exist as government towns rather than commercial ones.

The comparison with Jharkhand and Chhattisgarh — states from the same mineral-rich, feudal-legacy eastern Indian belt — is the sharpest diagnostic. Even these peer states are seven to eight percentage points ahead of Odisha. Jharkhand has Jamshedpur: the one town in eastern India where industry drove urbanization organically, because Tata Steel built not just a factory but an ecosystem — ancillary industries, small-scale manufacturing, a service economy, worker training, civic institutions. Jamshedpur is what happens when someone builds the platform. Rourkela, which should have been Odisha’s Jamshedpur, remained a government steel plant with government housing — a larger version of the government-town phenomenon, not a genuine industrial city. The steel plant operated as an island economy rather than generating the ecosystem that makes a city self-sustaining.

This is why the 17 percent persists. It is not a failure of any single policy. It is not the result of bad luck or geographic disadvantage. Odisha has 480 kilometers of coastline, 28 percent of India’s iron ore, 98 percent of its chromite, 51 percent of its bauxite, fertile river deltas, a young population, and location on the eastern seaboard at a time when Asia is the center of the global economy. The hardware is there. What is missing is the operating system — the institutional, commercial, and infrastructural platform on which urban life runs.

The question for the rest of this series is whether the operating system can be built now, after two centuries of absence. Whether the IT sector, the mineral value chain, the demographic transition, the digital revolution, and the slow accumulation of institutional capacity (OSDMA as proof that Odisha can build world-class institutions when it chooses to) can, together, create the conditions for the platform to emerge. Whether the talented Odias building Surat and Bangalore and Doha can be given a reason to build Bhubaneswar and Rourkela and Berhampur instead.

The satellite image at night is a photograph of the present. The question is whether the next photograph — taken in 2036, or 2046 — will show more light.

Sources

Census and demographic data:

Census of India 2011, Data Highlights — Odisha (censusindia.gov.in)
Department of Housing & Urban Development, Government of Odisha (urban.odisha.gov.in)
Population Projections for India and States 2011-2036, National Commission on Population (nhm.gov.in)
StatisticsTimes — Odisha Population (statisticstimes.com)
Census 2011 state-wise urbanization data (sociology.institute)
List of cities in Odisha by population (Wikipedia)

Historical and institutional sources:

The Zamindari System in Odisha, International Journal of Multidisciplinary Research and Analysis (ijmra.us)
Permanent Settlement (Wikipedia)
Merger of the Princely States of Odisha (historyofodisha.in)
British Relations with Princely States of Odisha, Odisha Review (magazines.odisha.gov.in)
Orissa Tributary States (Wikipedia)
Biswamoy Pati, “Interrogating Stereotypes: Princely States in Colonial Orissa,” Studies in History, 2005

Infrastructure and economic sources:

Development of Railway Transport in Colonial Orissa, Odisha Review (magazines.odisha.gov.in)
Paradip Port history (paradipport.gov.in)
Freight Equalization Policy (Wikipedia)
Freight to Tax Equalization — The South First (thesouthfirst.com)
India’s Freight Equalization Scheme and the Long-run, Cornell University working paper (barrett.dyson.cornell.edu)
Odisha mineral resources — Department of Steel and Mines, Government of Odisha (odishaminerals.gov.in)

Agriculture sources:

Rice Farming in Odisha — Rice Knowledge Bank (rkb-odisha.in)
Status of Agriculture — Agri-Odisha (agriodisha.nic.in)
Maharashtra sugar belt — Civilsdaily (civilsdaily.com)

Urban governance and planning sources:

Bhubaneswar (Wikipedia)
Bhubaneswar at 75 — Mycitylinks (mycitylinks.in)
How Bhubaneswar’s master plan was overtaken — Question of Cities (questionofcities.org)
Bhubaneswar Development Authority (Wikipedia, bda.gov.in)
Smart City Bhubaneswar proposal (smartnet.niua.org)
Bhubaneswar Smart City 8 years — Construction World (constructionworld.in)
AMRUT cities in Odisha — Clean India Journal (cleanindiajournal.com)

Migration sources:

Odisha State Migration Profile Report — Human Dignity Foundation (humandignity.foundation)
Migration from Ganjam to Gujarat — WorkFairAndFree (workfairandfree.org)
Of Caste, Climate and Migration — The Migration Story (themigrationstory.com)
How Migration is Changing Villages — IDR (idronline.org)

Comparative analysis sources:

Tamil Nadu urbanization trends (tamilnadu.pscnotes.com)
Inclusion through modernity: Dravidian urbanisation — SAGE Journals (journals.sagepub.com)
GIDC — Gujarat Industrial Development Corporation (gidc.gujarat.gov.in)
Software industry in Karnataka (Wikipedia)
Census Towns in India — World Bank (documents1.worldbank.org)
Jaga Mission (jagamission.in; world-habitat.org)
Odisha GSDP — Economic Survey 2025-26 (pc.odisha.gov.in)

Caste and mercantile community sources:

Bania caste (Wikipedia)
Nagarathar / Nattukottai Chettiars (Wikipedia)
Banias — Encyclopedia.com

Cross-References

The Leaving (full_read/the-leaving/): The migration paradox analyzed in depth — five corridors, dadan system, remittance economics, the empty village. Chapter 1 provides the migration numbers; Chapter 3 details the Ganjam-Surat corridor; Chapter 5 covers remittance patterns.
The Long Arc (full_read/the-long-arc/): Zamindari abolition (Chapter 2), Freight Equalization Policy (Chapter 1), Hirakud and Rourkela as national projects with local costs (Chapter 3), the extraction equilibrium (Chapter 5).
Delhi’s Odisha (full_read/delhis-odisha/): Central government policies that constrained Odisha’s development — FEP (Chapter 1), railway under-investment (Chapter 4), Finance Commission devolution (Chapter 5).
Political Landscape (full_read/political-landscape/): Government as dominant employer, the administrative economy, the announcement economy.
The Churning Fire (full_read/the-churning-fire/): Institutional failure patterns, the cage of learned helplessness, consciousness shifts. The “shift from ‘we are a state that was denied’ to ‘we are a state that has not yet built’” formulation.
Value Chain (full_read/value-chain/): Mineral value chain economics — why processing did not happen locally, what it would take to build industrial ecosystems.
Environmental Odisha (full_read/environmental-odisha/): Mineral extraction infrastructure, Hirakud capacity decline, energy transition and CBAM implications for value addition.
Tribal Odisha (full_read/tribal-odisha/): Princely state legacy in tribal territories, the parallel governance systems, Niyamgiri and displacement.

Source Research

The raw research that informs this series.