English only · Odia translation in progress

Chapter 6: The Missing Institution

In 2019, a steel company with operations across four countries began evaluating Odisha for a downstream processing facility — flat steel products for the automobile sector, the kind of value addition that the SeeUtkal Value Chain series documented as Odisha’s most consequential economic absence. The company had already established a presence in Gujarat, where the journey from initial inquiry to commissioned production had taken 22 months. The GIDC estate at Sanand had pre-developed land with power, water, and road connectivity. iNDEXTb had assigned a relationship manager who coordinated 26 approvals through the single-window mechanism. The Gujarat facility was producing by month 18 of a 22-month timeline.

In Odisha, the company submitted its expression of interest through the GO-SWIFT portal in January 2019. By April 2019, a site had been identified near Kalinganagar — geographically ideal, within the steel belt, close to raw material sources, and within trucking distance of Paradip port. By August 2019, the environmental impact assessment process had begun. By March 2020, COVID intervened. The process restarted in late 2020. By mid-2021, land acquisition was still in progress — not because land was unavailable but because the conversion from agricultural to industrial use required sequential approvals from the Revenue Department, the Forest Department (a portion of the identified land abutted reserve forest), and the Pollution Control Board, each operating on independent timelines with no institutional mechanism for synchronisation. By 2022, three years after the initial inquiry, the company had received environmental clearance but was still negotiating power supply arrangements with GRIDCO. The facility’s groundbreaking occurred in late 2023. Production was projected for 2025 — six years after the initial inquiry, against Gujarat’s 22 months.

The company’s India head, in an industry forum in Mumbai, offered a diagnosis that was more precise than he perhaps intended: “In Gujarat, the system works for you. In Odisha, you work against the system.”

This is not a story about bad people or bad intentions. The IPICOL officers who processed the Odisha application were competent. The GO-SWIFT portal was functional. The Industrial Policy Resolution of 2022 offered generous fiscal incentives. The Chief Minister’s office was supportive. But support, competence, and incentives are not the same thing as an institution. An institution is a machine that converts intention into outcome reliably, repeatedly, and independently of which specific human beings happen to be operating it at any given moment. Gujarat has that machine for industrial development. Odisha does not.

The previous five chapters diagnosed the problem. Chapter 1 established that OSDMA proves institutional capacity exists within Odisha’s bureaucracy. Chapter 2 documented what the broken operating system looks like at ground level — the transfer-posting regime, the scheme proliferation, the monitoring-without-learning. Chapter 3 decomposed OSDMA’s success into seven factors and assessed which are reproducible. Chapter 4 showed that institutional weakness is not an accident but a stable Nash equilibrium that serves identifiable interests. Chapter 5 examined what functional institutional architecture looks like elsewhere — Tamil Nadu’s bureaucratic compounding, Kerala’s decentralised governance, Gujarat’s investment machinery, Singapore’s EDB, South Korea’s EPB, Botswana’s resource management.

This chapter attempts something different. It asks: given the constraints that actually exist — the federal system, the IAS cadre, the patronage equilibrium, the fiscal position, the 17% urbanisation, the caste and tribal complexity — what would functional institutions look like in the specific domains where Odisha’s institutional absence is most consequential? Not a policy prescription. A design exercise. The difference matters. A policy prescription says “do this.” A design exercise says “if you wanted this outcome, here is the architecture that could produce it, given the materials you actually have.”

The cross-domain lens is software architecture. OSDMA is a service that the operating system runs reliably. The OS is missing other services — for industrial development, for education, for urban planning, for agriculture. The task is to design those missing services with the same rigour that a systems architect would bring to specifying a new OS module: what inputs does it need, what outputs must it produce, what interfaces does it require with other services, what failure modes must it handle, and what are the minimum system requirements for it to run?

The Design Problem

Before designing anything, the nature of the problem must be precisely specified. The problem is not a shortage of policy. Odisha has an Industrial Policy Resolution revised every few years. It has an education policy aligned with NEP 2020. It has an urban development framework. It has agricultural schemes covering everything from seeds to soil health cards. The policies exist. The institutions that would convert those policies into outcomes do not.

Lant Pritchett’s concept of isomorphic mimicry — borrowed from biology, where harmless species evolve to resemble dangerous ones — captures the situation with uncomfortable precision. Odisha’s industrial promotion apparatus looks like Gujarat’s. There is IPICOL (corresponding to iNDEXTb), IDCO (corresponding to GIDC), GO-SWIFT (corresponding to Gujarat’s single-window clearance). The organisational charts are similar. The mandates overlap. But the institutional performance diverges radically. Gujarat’s MoU-to-production conversion rate is roughly 40-50%. Odisha’s historical rate is 25-35%. The gap is not explained by the presence or absence of agencies. It is explained by whether those agencies function as institutions or as institutional mimics.

Andrews, Pritchett, and Woolcock’s PDIA framework — Problem-Driven Iterative Adaptation — offers the design methodology. The conventional approach to institutional reform starts with a model (Singapore’s EDB, Gujarat’s GIDC) and asks how to transplant it. PDIA starts with a specific problem and asks what institutional capacity would solve it. The model approach produces isomorphic mimicry — agencies that look like their templates but do not function like them. The problem approach produces institutions that may not resemble any existing template but that solve the problem they were designed to solve.

OSDMA is itself a PDIA success story, though nobody called it that at the time. The problem was specific: people die in cyclones because nobody evacuates them, nobody warns them, and nobody shelters them. The solution was not imported wholesale from any other country’s disaster management system. It was built iteratively over two decades, starting with the immediate post-1999 crisis response and evolving through each subsequent cyclone event. The early warning system was adapted to Odisha’s specific coastal geography. The shelter network was located based on Odisha’s specific population distribution. The volunteer architecture was integrated with Odisha’s specific community structures. The institution does not look like Singapore’s National Environment Agency or Japan’s JMA or Bangladesh’s CPP. It looks like OSDMA — an institution shaped by the problem it was designed to solve and the context in which it operates.

What makes institutional design different from scheme design? A scheme has a budget, guidelines, beneficiaries, and a timeline. An institution has autonomy, metrics, learning loops, and permanence. A scheme distributes resources. An institution builds capacity. A scheme ends when the budget is spent. An institution exists until the problem is solved or the context changes. The KALIA scheme distributed Rs 10,000 to farmer families. An agricultural development institution would build the capacity to raise rice yields from 1,600-1,800 kg/ha to 3,000+ kg/ha — not through a single intervention but through sustained, adaptive, context-specific institutional work on extension, irrigation, input supply, market access, and cold chain infrastructure, compounding over decades the way OSDMA compounded its cyclone response capacity.

The design question for each domain is: what is the minimum institutional architecture that could produce OSDMA-quality outcomes, given the constraints that actually exist?

OSDMA for Industrial Development

The problem, specified precisely

Odisha’s industrial problem is not the absence of investment interest. The Utkarsh Odisha conclave in 2025 attracted Rs 16.73 lakh crore in investment intentions across 593 projects and 20 sectors. The problem is conversion. Historically, 25-35% of MoU value materialises as actual production on the ground. For every Rs 100 crore of investment intention announced at a conclave, Rs 25-35 crore becomes an operational factory. The rest dissipates — into changed market conditions, into approvals that take longer than business timelines allow, into infrastructure gaps that increase project costs beyond viability, into the absence of a supplier ecosystem that makes the investor’s logistics uncompetitive.

Gujarat’s conversion rate of 40-50% is not explained by better natural resources (Odisha’s mineral endowment is richer), by better human capital (IAS officers are drawn from the same cadre), or by more generous fiscal incentives (Odisha’s IPR 2022 is competitive). It is explained by institutional architecture — the machinery that converts an investor’s intention into an operational factory.

The design

Structure: An autonomous industrial development authority — call it the Odisha Industrial Development Authority (OIDA) — operating outside the Industries Department. Not a department, not a corporation, not a portal. An authority with a statutory mandate, operational autonomy, and clear performance accountability.

Why outside the department? Because the Industries Department operates within the standard bureaucratic architecture: the Secretary rotates every 12-18 months, the department manages 40+ schemes simultaneously, its staff are generalist IAS and OAS officers who may have been in the Health Department last year and the Education Department next year. The institutional capacity for industrial development — sector-specific knowledge, investor relationship management, ecosystem building, long-term planning — cannot be built within a structure designed for routine administration.

OSDMA sits within the Revenue and Disaster Management Department but operates with effective autonomy during crises. The OIDA would need analogous structural autonomy: a governing board with a fixed-tenure chairman (minimum five years), a professional staff recruited on contract at competitive salaries (not IAS rotation), and a clear mandate that cannot be diluted by adding unrelated responsibilities.

Mandate: Convert investment intentions into operational production. Not “promote industry” (too vague) or “facilitate clearances” (too narrow). The mandate must be end-to-end: from the moment an investor expresses interest to the moment the factory produces its first output. The metric is not MoUs signed (input) but production commenced (output) and value added within the state (outcome).

The Singapore EDB principle: proactive, not reactive. IPICOL and GO-SWIFT are reactive systems — they process applications that arrive. The EDB is a proactive system — it identifies target sectors, researches global companies in those sectors, and approaches them directly with customised proposals. The OIDA would need a small team of sector specialists (not generalists) who understand the global automobile components supply chain, or the electronics manufacturing value chain, or the food processing industry’s specific requirements, and who can approach specific companies with pre-packaged proposals: “Here is a 50-acre plot in Kalinganagar with three-phase power, water, and rail connectivity. Here is the clearance timeline — 90 days for these 12 approvals. Here are three existing companies in the same estate who can confirm operational experience. Here is a dedicated relationship manager who will be your single point of contact for the next three years.”

The GIDC principle: pre-developed estates. IDCO manages 95+ industrial estates, but “manage” is generous — many estates lack maintained roads, reliable power, adequate drainage, or water supply. The OIDA would need authority over a small number of strategic industrial estates — perhaps five, co-located with Odisha’s existing industrial clusters (Kalinganagar for steel and metals, Angul-Jharsuguda for aluminium and power, Paradip for petrochemicals, Bhubaneswar periphery for IT and electronics, and a new estate for food processing near the agricultural hinterland of the Mahanadi delta). These five estates would be developed to plug-and-play standard: an investor could occupy a plot and begin construction without building basic infrastructure. The model is GIDC’s 239 estates, but Odisha need not start with 239. It needs five that work, from which the model can scale.

The Tamil Nadu TIDCO principle: equity participation. TIDCO’s joint venture model — taking 1-26% equity stakes in industrial ventures — aligns state and investor interests. When the state has equity in the factory, it has a financial incentive to ensure the factory succeeds: clearances move faster, power supply is prioritised, labour disputes are resolved. The OIDA would need a small capital fund — perhaps seeded from DMF collections (Rs 3,000-5,000 crore annually) — to take minority equity stakes in anchor investments, particularly in value-addition industries (flat steel, aluminium extrusion, food processing) where the state’s interest is not just the investment itself but the ecosystem it catalyses.

The OSDMA principle: measurable outcomes. OSDMA’s metric — cyclone deaths — is binary and uncheateable. The OIDA needs an equivalent. Proposal: track four metrics publicly, updated quarterly. (1) MoU-to-production conversion rate, measured not at announcement but at production commencement. (2) Time from investor inquiry to production start, measured against a publicly stated target (say, 18 months for a standard manufacturing project). (3) Jobs created — not announced, but verified through EPFO registrations. (4) Value added within the state — the rupee value of processing done in Odisha rather than exported as raw material.

The OSDMA principle: leadership continuity. OSDMA’s founding leadership served long enough to embed institutional culture. The OIDA’s chairman would need a fixed five-year term, renewable once, with removal only for cause — not at the pleasure of the Chief Minister. This is the single most difficult design requirement, because it directly conflicts with the patronage equilibrium described in Chapter 4. The politician’s power over transfers is the currency of the equilibrium. Exempting the OIDA chairman from that currency requires either crisis-level political pressure (which does not exist for industrial development as it did for cyclone mortality) or a political entrepreneur willing to trade short-term patronage for long-term developmental returns. Chapter 7 examines the conditions under which such entrepreneurs emerge.

What exists and what is missing

Odisha already has the pieces: IPICOL for investment promotion, IDCO for land and infrastructure, GO-SWIFT for clearances. What is missing is institutional integration — a single authority with end-to-end responsibility, operational autonomy, sector-specific expertise, and outcome accountability. The pieces exist as fragments of different departments, each with its own reporting line, its own budget cycle, its own leadership rotation. The OIDA design does not require creating new pieces. It requires assembling the existing pieces under a single institutional roof with the design features that make OSDMA work: clear mandate, autonomy, metrics, leadership continuity, and learning loops.

Confidence assessment: 20-25% probability that an OIDA with these design features is created within the next decade. The political economy constraints — the patronage equilibrium’s hostility to autonomous institutions, the IAS cadre’s resistance to fixed-tenure positions outside its rotation, and the absence of a crisis catalyst comparable to the 1999 cyclone — make this a structurally unlikely outcome. But the absence of crisis does not make it impossible. It makes it harder. The question is whether political will can substitute for political urgency — whether a Chief Minister can choose to build institutional capacity for industrial development the way Patnaik’s government built institutional capacity for disaster management, not because 10,000 people died but because the economic opportunity cost of not doing so is visible enough to justify the political investment.

OSDMA for Education

The problem, specified precisely

Odisha’s education system does not have an output problem. It has a mismatch problem. The system produces graduates optimised for export — English-fluent, competitive-exam-ready, corporate-interview-prepared — while the state needs graduates optimised for its own development: industrial technicians for Kalinganagar’s steel plants, agricultural scientists for the Mahanadi delta’s rice productivity challenge, urban planners for the state’s desperately needed city-building, healthcare professionals for the 30-40% doctor vacancies in government hospitals, and institutional leaders who understand Odisha’s specific combination of mineral wealth, institutional weakness, and cultural depth.

The Education Odisha series documented this in exhaustive detail. The data bears condensed repetition. Sixty-eight percent faculty vacancy in state universities. 1,187 of 1,911 sanctioned university positions empty. NIT Rourkela placing 90%+ of graduates outside the state. KIIT University enrolling 14,064 undergraduates of whom 103 are from Odisha. Class 5 reading ability at 25% statewide, with a 56-percentage-point gap between Jajpur (70.6%) and Malkangiri (14%). 637 ITIs teaching manual arc welding while JSPL Angul, twenty-three kilometres away, needs submerged arc welding with radiographic quality.

The institutional absence is not in the Education Department — which is large, well-funded (Rs 42,565 crore in 2026-27, 13.7% of the total budget), and staffed — but in the coordination function. Nobody in Odisha’s institutional architecture is responsible for ensuring that the education system’s output matches the economy’s input requirements. The Education Department produces graduates. The Industries Department attracts investment. The two do not speak to each other through any institutional interface. The result is what the Education series called the missing API: two systems that cannot communicate because no interface was ever specified.

The design

Structure: An Education-Employment Coordination Authority — not a department, not a scheme, not a committee. An authority with representation from Education, Industries, Labour, and the private sector, with a mandate to align educational output with economic demand.

This is not the Tamil Nadu Skill Development Corporation (TNSDC) transplanted to Odisha. It is a response to Odisha’s specific problem: the coordination failure between education and employment that produces the export factory.

Three functions, clearly separated:

Function 1: Education-industry co-location. The authority would identify the five most consequential skill gaps between Odisha’s existing educational institutions and its existing industrial clusters, and create formal, binding partnerships between them. Not advisory committees. Binding partnerships where curriculum co-design, apprenticeship placement, and equipment co-investment are contractual obligations.

The Tamil Nadu SIPCOT-polytechnic model is the template. SIPCOT industrial parks maintain strategic linkages with educational institutions. Companies like Delta Electronics and Hyundai co-invest in training. The polytechnic graduate in Hosur has multiple employers competing for their skills because the education and employment ecosystems are physically co-located and institutionally linked.

Odisha has the raw material for co-location. The Kalinganagar steel belt (Tata Steel, JSPL, Essar) is within 100 kilometres of NIT Rourkela and multiple ITIs. The Angul-Jharsuguda aluminium and power corridor (JSPL, Vedanta, MCL) is near Sambalpur University and regional ITIs. The Paradip petrochemical complex (IOCL, PPL) is accessible from Bhubaneswar’s engineering colleges. What is missing is the institutional mechanism that brings education officials and industrial employers into the same room — not once, at a conclave, but continuously, as a permanent institutional function.

The authority would require companies receiving industrial incentives (land allotments, tax breaks, infrastructure support from IDCO) to co-invest in formal apprenticeship programmes with local ITIs and polytechnics. Germany’s dual system cannot be imported wholesale — the institutional preconditions are different — but the principle is transferable: companies that benefit from public industrial infrastructure should contribute to the human capital infrastructure that feeds their workforce.

Function 2: University autonomy with accountability. The Odisha University (Amendment) Act 2024 began the reform by shifting recruitment from OPSC to institutional committees and restoring the Senate with 37+ academic members. This is structurally sound but incomplete. The authority would push for genuine university autonomy: fixed-tenure Vice-Chancellors appointed through search committees (not political patronage), faculty recruitment at competitive national salaries, research mandates linked to Odisha’s specific challenges (mineral processing technology, cyclone-resilient agriculture, heat-adaptive urban design, tribal language preservation), and outcome accountability through published metrics — not NIRF rank (which rewards inputs) but graduate retention rate (which measures whether the university produces people who stay).

The 68% faculty vacancy is not a resource problem — Odisha’s education budget at Rs 42,565 crore is adequate. It is an institutional design problem. Positions remain vacant because the recruitment process is slow (OPSC processing times), the salaries are uncompetitive (guest lecturers at Rs 20,000-25,000 cannot attract talent that earns Rs 50,000+ elsewhere), and the academic environment is unattractive (a university with 55% vacancy and a politically appointed Vice-Chancellor does not draw ambitious researchers). The authority’s role would be to create the institutional conditions under which universities can recruit, retain, and develop faculty — not by running the universities but by removing the institutional constraints that prevent them from functioning.

Function 3: Vocational prestige-building. The stigma that attaches to vocational education in India — only 2% of the workforce has formal vocational training versus 50%+ in Germany — cannot be solved by policy. It can be solved by demonstrated outcomes. When an Angul ITI graduate earns Rs 25,000-35,000 per month immediately upon qualification, and when this is publicly visible, parents’ revealed preferences shift. The World Skill Center in Bhubaneswar, with its CNC machining and robotics labs, is the institutional prototype. The authority would scale this model to five regional centres co-located with industrial clusters, each with guaranteed placement linkages to specific employers.

The metric, as in the industrial development design, must be uncheateable: not “students enrolled” or “courses offered” but “graduates employed within six months at a wage above Rs 20,000 per month.” Publish the data. Make it politically visible. Create the competitive pressure that OSDMA’s mortality data creates for disaster management.

Design constraints specific to education

Education reform is harder than industrial development reform because the feedback loops are longer. An industrial investment produces visible output (a factory, jobs, tax revenue) within two to five years. An education reform produces visible output (better learning outcomes, higher retention, improved employability) within ten to twenty years. Politicians operate on five-year cycles. The institutional design must therefore produce some visible, measurable output within the electoral cycle while building capacity that compounds over decades.

The Odisha Adarsha Vidyalaya (OAV) model — 314 model schools with improved infrastructure, CBSE curriculum, and full-time teachers — demonstrates that visible output is possible within an electoral cycle. The authority would not replace the OAV model. It would extend it from a scheme (bounded, budget-dependent, politically branded) to an institution (permanent, autonomous, metrics-driven, surviving political transitions).

Confidence assessment: 15-20% probability of systemic education-employment coordination within a decade. The coordination failure is deeply embedded in the bureaucratic structure — the Education Department and the Industries Department report to different Secretaries, operate under different budgets, and have no institutional incentive to coordinate. The political economy favours scheme proliferation (each new scheme generates political credit) over institutional integration (which is invisible to voters). The most likely outcome is continued parallel operation of education and industry systems, with occasional project-level coordination that does not scale.

OSDMA for Urban Planning

The problem, specified precisely

Odisha is approximately 17% urbanised — among the lowest of India’s major states, against Tamil Nadu’s 48.4%, Gujarat’s 42.6%, and Kerala’s 47.7%. Bhubaneswar is the only city that functions as a genuine urban centre with multiple employment sectors, cultural amenity, healthcare infrastructure, and educational institutions. Cuttack retains administrative importance but has not grown into a modern urban economy. Rourkela has NIT and RSP but neither a startup ecosystem nor a cultural scene. After these three — which together account for less than 15% of the state’s population — there is nothing that could reasonably be called a city in the sense that urban economists use the term: a place where agglomeration effects produce economic multipliers, where talent density generates innovation, where the quality of urban life retains educated people who would otherwise leave.

The Urbanisation Odisha series documented this absence in detail. The Smart City Mission invested Rs 5,000 crore in Bhubaneswar’s Smart City project. The investment produced better roads, improved drainage, a bus system, and digital infrastructure. It did not produce a city-building institution. The Smart City is a scheme — bounded, project-based, externally designed — not an institution. When the project funding ends, the institutional capacity to plan, build, and manage urban growth will end with it, because that capacity was housed in a Special Purpose Vehicle (SPV) with a project lifecycle, not in a permanent governance body with an ongoing mandate.

The design

Structure: A State Urban Development Authority with a mandate not to manage existing cities but to build new urban capacity. Not a municipal corporation (which manages existing urban infrastructure) or a development authority (which typically manages land use planning for a single city). A state-level institution responsible for the strategic question: where should Odisha’s urban future be located, and what institutional architecture will make those locations viable?

The Singapore HDB principle: institution-led city-building. Singapore’s Housing and Development Board did not merely provide housing. It built a nation through housing — relocating populations from kampong villages to planned estates, mixing ethnic groups, creating community infrastructure, and transforming a collection of settlements into a functional city-state. The HDB was an institution, not a scheme. It had a permanent mandate, professional staff, long-term planning horizons, and outcome accountability.

Odisha does not need Singapore’s HDB. It needs the design principle: that city-building requires an institution with a multi-decade planning horizon, authority over land use, infrastructure investment, and housing, and accountability for measurable urban outcomes — not “projects completed” but “population absorbed into functional urban environments with employment, services, and quality of life.”

The GIDC estate model applied to urban planning. GIDC does not wait for factories to arrive and then build infrastructure around them. It pre-develops industrial estates so that factories can arrive and begin operations immediately. Urban planning requires the same logic. Instead of waiting for population to migrate to cities and then retrofitting infrastructure (the pattern that has produced Bhubaneswar’s current traffic, water, and sewage challenges), the authority would pre-develop urban infrastructure in locations identified for growth.

Two candidates are strategically obvious. Rourkela — with NIT, RSP, proximity to the Kalinganagar industrial corridor, and a natural hinterland covering western Odisha — could become a city of 1-1.5 million if it had the urban infrastructure (public transport, healthcare, education, cultural amenity) that retains educated people. Berhampur — with MKCG Medical College, a coastal location, proximity to Gopalpur port, and a natural hinterland covering southern Odisha — could serve the same function for the state’s southern region. Neither requires building from scratch. Both require institutional investment in the urban services that transform a town into a city: reliable water and sanitation, public transport, healthcare beyond a single medical college, multiple schooling options, and — critically — a cultural and social infrastructure that makes educated life livable.

The missing municipal capacity. India’s 74th Constitutional Amendment mandated urban local self-government through municipalities, just as the 73rd mandated rural self-government through panchayats. In practice, Odisha’s municipalities are weak — limited revenue authority, limited planning capacity, dependent on state transfers, and staffed by officers who rotate as frequently as their rural counterparts. The authority would not replace municipal governance. It would build the technical capacity that municipalities lack: urban planning expertise, traffic engineering, water management, waste management, and the ability to prepare and implement city development plans that extend beyond the current electoral cycle.

The institutional design must account for a constraint that Gujarat and Tamil Nadu did not face when they built their urban systems: Odisha’s 22.8% tribal population, concentrated in the western and southern districts that are also the least urbanised. Urban development in tribal areas requires institutional sensitivity to land rights (FRA, PESA), cultural preservation, and the specific governance structures that the Tribal Odisha series documented. The JAGA Mission — which provided land rights to slum dwellers in urban areas — is a model of how urban development can be made inclusive. But JAGA is a scheme. The authority would need to institutionalise the principle: urban development that works within the constitutional and cultural framework of tribal areas, not against it.

Confidence assessment: 10-15% probability of a functional state urban development authority within a decade. Urban planning operates on timescales (20-30 years) that exceed political planning horizons (5 years). The political return on urban investment is low — the beneficiaries are future residents of cities that do not yet exist, who cannot vote in the current election. The patronage equilibrium strongly favours rural welfare schemes (which reach current voters) over urban infrastructure investment (which serves future populations). The 17% urbanisation rate is both the problem and the constraint: the political constituency for urban development is small because the urban population is small, but the urban population is small because no institution has built the urban infrastructure that would attract it.

OSDMA for Agriculture

The problem, specified precisely

Odisha’s rice yield of 1,600-1,800 kg/ha is approximately 55-65% of the national average and less than half of Punjab’s 4,000-4,500 kg/ha. The agricultural extension ratio — one worker per 1,500-2,000 farming families, against the recommended 1:800-1,000 — means that the system responsible for connecting research to farmers is operating at half capacity. The cold chain gap is acute: 7-8 lakh MT of capacity against a requirement of 15-20 lakh MT. Farmer Producer Organisations number 500-700, of which the majority are non-functional or low-volume.

These are not policy gaps. KALIA distributed Rs 10,000 to farmer families. Soil health cards were distributed. Crop insurance covers a defined population. The Odisha Millet Mission — one of the state’s genuine agricultural successes — demonstrated that focused institutional intervention in a specific domain can produce results: millet area increased from 69,000 to 80,000+ hectares, procurement reached 20,000+ tonnes, and farm gate prices improved 30-50% in target areas.

The OMM’s relative success is instructive for design purposes. It worked not because it was a bigger scheme but because it had institutional features that distinguish it from the standard agricultural programme: focused scope (millets in rain-fed tribal areas, not the entire agricultural system), community-based approach (working through Mission Shakti SHGs and FPOs rather than individual targeting), value chain approach (from seed to processing to market), and a dedicated implementation cell with consistent leadership less subject to routine transfer. The OMM is, in miniature, what an agricultural development authority would look like at scale.

The design

Structure: An Odisha Agricultural Development Authority (OADA) — modelled not on an existing Indian agricultural department but on the Amul/NDDB principle and the OMM’s demonstrated institutional features.

The National Dairy Development Board under Verghese Kurien transformed Indian dairying not through schemes but through institutional architecture: producer cooperatives at the village level, federated into district unions and state cooperatives, with professional management, quality control, and market linkage. The cooperative structure aligned the interests of the producer with the institution — because the institution was owned by the producers. Kurien led NDDB for over four decades, the leadership continuity that Chapter 3 identified as a critical factor in OSDMA’s success.

The OADA would adapt this architecture for Odisha’s specific agricultural context.

Extension as institution, not as scheme. The current extension system is consumed by scheme administration — form filling, beneficiary identification, verification. The extension worker’s actual function (knowledge transfer from research to farmer) occupies less than 20% of working time. The OADA would separate extension from scheme administration. Extension workers within the OADA would have a single mandate: improve farm productivity in their assigned area, measured by yield per hectare. Scheme delivery — KALIA disbursement, crop insurance enrollment, soil health card distribution — would remain with the Agriculture Department. The OADA would handle knowledge and technology; the department would handle money and compliance.

The FPO as institutional building block. Odisha’s 500-700 FPOs are mostly non-functional because they were created as scheme outputs rather than as institutions. The OADA would identify the 50-100 FPOs that show genuine collective action capacity, invest in their institutional development (professional management, market linkage, cold chain access, quality certification), and use them as nodes for scaling — the PDIA principle of positive deviance applied to agriculture. Successful FPOs would become the template for building new ones, not through top-down mandate but through demonstration and peer learning.

Cold chain as institutional infrastructure. The cold chain gap cannot be closed by government investment alone — the capital cost is too high and the operational expertise does not exist within government. The OADA would facilitate private cold chain investment by guaranteeing a minimum volume of procurement through FPOs — reducing the demand risk that deters private investors from building cold storage in areas where farmer aggregation is uncertain. This is the Amul model: the cooperative guarantees volume; the private sector provides infrastructure; the institution coordinates between them.

The metric: Rice yield per hectare at district level, published annually. Not the current agricultural dashboard’s metrics (scheme beneficiaries reached, budget utilised) but the outcome that actually matters: is the food more or less productive? The yield figure is countable, comparable across districts and across time, and resistant to gaming — you cannot fabricate rice that does not exist. It is not as binary as cyclone deaths, but it is closer to uncheateable than any of the current agricultural performance metrics.

Confidence assessment: 25-30% probability that some form of institutional agricultural authority emerges within a decade. Agriculture, unlike industrial development or urban planning, has a direct electoral constituency — 48-50% of Odisha’s workforce. The OMM’s demonstrated success provides a proof of concept. The BJP government’s integration of KALIA with PM-KISAN and the launch of CM Kisan Yojana suggests political interest in agricultural outcomes. The constraint is the same as for all the designs: building an autonomous institution requires surrendering political control over transfers, budgets, and appointments — the currency of the patronage equilibrium.

Design Constraints That Are Real

The four institutional designs above are not blueprints. They are thought experiments constrained by real conditions. The constraints are not excuses. They are design parameters — the materials and tolerances within which the architect must work. Wishing them away is not design. It is fantasy.

The federal system. Odisha is a state within the Indian Union. It cannot set its own monetary policy (Singapore), control immigration (Singapore), nationalise banking (South Korea), or negotiate independent trade agreements (Botswana with De Beers). Industrial policy must operate within the framework of central legislation — the Factories Act, the Environmental Protection Act, the Forest Conservation Act, the Companies Act, SEBI regulations, RBI norms. The “single-window clearance” that Gujarat achieves is possible because Gujarat has built institutional capacity to process multiple approvals simultaneously, not because it has fewer approvals to process. The federal constraint is binding but not disabling — it means Odisha must build better processing capacity within the existing regulatory framework, not that it cannot build processing capacity at all.

The IAS cadre. Odisha cannot hire its institutional leaders differently from any other state department. The IAS cadre allocation system, the promotion rules, the deputation norms, the pay scales — all are centrally determined. The OIDA chairman, the OADA head, the Urban Authority director — all would need to be IAS officers unless the institution is structured as a society or statutory body that permits contractual hiring. OSDMA was registered under the Societies Registration Act precisely to create this flexibility. The design implication is clear: any new institutional authority should be structured as a society or statutory body, not as a government department, to enable the hiring flexibility that institutional competence requires.

But the flexibility has limits. An autonomous authority that hires non-IAS professionals at competitive salaries will face resentment from the IAS cadre, which sees its monopoly on senior government positions as a structural entitlement. This resentment can manifest as bureaucratic obstruction — other departments refusing to cooperate with the authority, files being delayed, inter-departmental coordination breaking down. OSDMA overcomes this during cyclone emergencies because the emergency grants it command authority. An industrial development authority has no such emergency powers. The design must therefore include institutional mechanisms for inter-departmental coordination — perhaps a coordination committee chaired by the Chief Secretary, with the authority’s head as permanent member.

The patronage equilibrium. Chapter 4 documented the equilibrium in detail. The politician prefers weak institutions because weak institutions equal discretionary power. The bureaucrat prefers the current system because transfers are the currency of career management. The contractor prefers scheme-based budgets because schemes create procurement opportunities. The voter prefers direct welfare transfers because institutional capacity is invisible while KALIA money is tangible.

No institutional design can wish this equilibrium away. It can only work within it or find the specific conditions under which the equilibrium shifts. The conditions Chapter 4 identified for equilibrium shift are: crisis (which makes institutional failure politically visible and costly), external pressure (donor conditionality, central government mandates, judicial orders), intra-elite competition (rival political factions bidding for institutional reform as a competitive advantage), or a political entrepreneur willing to trade patronage for institutional capacity because they calculate that institutional outcomes will produce greater political returns than discretionary transfers.

The honest assessment is that the fourth condition — a political entrepreneur — is the most likely pathway, and it is rare. Chapter 7 examines the conditions under which such entrepreneurs emerge.

Fiscal constraints. Odisha’s fiscal position is, paradoxically, a strength rather than a constraint. NITI Aayog’s Fiscal Health Index ranked Odisha first among all Indian states in 2025, with a score of 67.8 — low debt (14-16% of GSDP, against the national average of 28-30%), surplus revenue account, and capital expenditure ratio of 18-22%. Mining revenue provides Rs 15,000-20,000 crore annually through royalties, DMF, and OMBADC. The state has money. The bottleneck is not fiscal but institutional: the bureaucratic apparatus described in Chapter 2 — rapid transfers, vacant positions, scheme proliferation, monitoring failures — is the machinery through which money must flow to become development. That machinery’s dysfunction explains the gap between fiscal health and development outcomes.

The fiscal constraint is not the total amount of money available but the institutional capacity to spend it effectively. Across departments, Odisha consistently shows significant unspent balances — over Rs 8,000 crore in 2022-23 across all departments. Tribal Development, Rural Development, and Agriculture show the highest unspent ratios. The paradox is complete: a state ranked first for fiscal health cannot spend the money it has, because the institutional machinery for converting budgets into outcomes does not function.

Caste and tribal complexity. Odisha’s ST population (22.8%) and SC population (17.1%) create institutional requirements absent in Singapore, South Korea, or even Tamil Nadu and Gujarat. PESA mandates gram sabha consent for development in scheduled areas. FRA creates forest rights that interact with land acquisition. The constitutional and legal framework for tribal governance is extensive — and extensively violated (136 CAG violations of PESA provisions). Any institutional design that ignores these requirements will either fail legally (challenged in court) or fail legitimately (implemented without consent of affected communities). The OIDA’s industrial estates in tribal areas would need to incorporate gram sabha consultation not as a box-ticking exercise but as a genuine design input — the Ostrom principle that governance rules must be adapted to local conditions.

17% urbanisation. The low urbanisation rate constrains every other institutional design. Without cities, there is no tax base for municipal services. Without municipal services, there is no urban quality of life. Without urban quality, educated people leave. Without educated people, there is no innovation ecosystem. Without an innovation ecosystem, there is no economic diversification beyond minerals. Each constraint reinforces the others. The urban planning authority is not one design among many. It is the platform on which every other institutional design depends. Education reform without cities equals training for departure. Industrial development without cities equals mining enclaves. Agricultural modernisation without cities equals continued rural poverty with better rice yields.

What Ostrom and Scott Teach

The institutional designs above draw primarily from the OSDMA model and from comparator cases in Tamil Nadu, Gujarat, Singapore, and South Korea. But two theoretical frameworks deserve explicit application because they address the failure modes that Indian institutional design most commonly exhibits.

Elinor Ostrom’s eight design principles

Ostrom’s Nobel Prize-winning research on commons governance identified eight characteristics present in long-enduring institutional arrangements. Applied to the institutional designs proposed above:

1. Clearly defined boundaries. Each authority must have a mandate specific enough to be measureable and bounded enough to prevent mandate creep. OSDMA’s mandate is cyclone mortality reduction. The OIDA’s mandate should be investment conversion, not “industrial development” (which is everything and therefore nothing). The OADA’s mandate should be agricultural productivity, not “farmer welfare” (which overlaps with a dozen other schemes). Vague mandates produce isomorphic mimicry. Clear mandates produce accountability.

2. Congruence between rules and local conditions. Odisha’s institutions cannot be Gujarat’s transplanted eastward. The industrial development authority must account for Odisha’s specific geography (cyclone-prone coast, mineral-rich interior, tribal-dominated west), its specific bureaucratic culture (newer administrative tradition, higher transfer frequency), and its specific industrial profile (mineral-dependent, value-addition-deficient). The OADA must work with Odisha’s specific agricultural reality: rain-fed rice cultivation, fragmented holdings averaging 0.95 hectares, and a farming population where 93% are small and marginal. Ostrom’s principle is Scott’s principle stated differently: institutions that ignore local conditions will be destroyed by them.

3. Collective-choice arrangements. The people affected by the institution’s rules must participate in setting those rules. An industrial development authority that makes decisions about Kalinganagar’s future without consulting the communities of Kalinganagar will face the same legitimacy crisis that mining companies face when they bypass gram sabhas. The OADA that designs extension programmes without farmer input will produce the same disconnect between scheme design and ground reality that the current extension system exhibits. This is not stakeholder consultation as public relations. It is stakeholder participation as institutional design — the same principle that makes OSDMA’s community-level volunteer network effective.

4. Monitoring. Institutions must monitor both their own performance and the compliance of their stakeholders. OSDMA monitors cyclone events in real time and reviews its own response after each event. The OIDA must monitor investment conversion in real time and review its own facilitation performance after each project. The monitoring must be by agents accountable to the institution’s mission, not by external auditors who measure compliance with procedures. CAG audits are necessary but insufficient. They measure whether rules were followed, not whether outcomes were achieved.

5. Graduated sanctions. Institutions that interact with private actors — investors, farmers, service providers — need graduated responses to non-compliance. An investor who misses a production deadline should face graduated consequences (reduced incentives, then warning, then recovery of public investment) rather than binary outcomes (full incentive or full penalty). The current system of industrial incentives in Odisha is largely binary: incentives are granted at MoU signing and rarely recovered when the MoU is not honoured.

6. Conflict resolution. Every institutional domain generates conflicts — between investors and communities, between departments and authorities, between local interests and state priorities. The institution must have low-cost, accessible mechanisms for resolving these conflicts. OSDMA resolves operational conflicts through its command authority during emergencies. The OIDA would need an industrial dispute resolution mechanism that is faster than the courts (where cases take years) and more authoritative than departmental correspondence (which takes months).

7. Minimal recognition of rights to organise. The state must recognise the authority’s right to function autonomously. If the Chief Minister can override the authority’s decisions at will, the authority is not an institution. It is a department with a different name. This is the single most difficult Ostrom principle to implement in the Indian governance context, where political authority is constitutionally supreme and bureaucratic autonomy is culturally suspect.

8. Nested enterprises. Institutional functions at different scales require governance at different scales. The OIDA would need district-level officers who understand local conditions (Ostrom’s Principle 2) coordinated by a state-level authority that maintains consistency and enables inter-district coordination. This is the federal design principle: local autonomy within a state-level framework within a national regulatory structure. OSDMA implements this through its village-block-district-state coordination structure during cyclone response. Each level handles the problems appropriate to its scale.

James Scott’s warning

Scott’s Seeing Like a State is a warning against exactly the kind of institutional design exercise this chapter attempts. Grand schemes fail, Scott argued, when they ignore local knowledge (metis) in favour of centralised, legible, top-down plans (techne). Brasilia was designed by architects who had never lived in a spontaneous city. Soviet collective farms were designed by planners who had never farmed. Tanzanian ujamaa villages were designed by bureaucrats who had never inhabited the settlements they destroyed.

The warning applies with full force. An OIDA designed in Bhubaneswar by IAS officers who have never run a factory will exhibit the same failure modes as a Smart City designed in Delhi by consultants who have never lived in Bhubaneswar. The design must therefore incorporate local knowledge — not as ornament but as input.

Concretely: the OIDA’s industrial estate plans should be informed not just by infrastructure engineers but by the companies that will occupy them (What do you actually need? Where are your supply chain bottlenecks? What is the real reason your last project in Odisha took six years?). The OADA’s extension programmes should be informed not just by OUAT researchers but by the farmers who will adopt them (What are you actually growing? Why? What would you need to change? What has the last extension worker who visited you actually told you?). The Urban Authority’s city plans should be informed not just by urban planners but by the people who live in Rourkela and Berhampur (Why do educated young people leave? What would make them stay? What does this place lack that Bangalore provides?).

Scott’s metis principle is not anti-institutional. It is anti-hubris. It says: design institutions, but design them with humility. Build in feedback loops. Start small. Iterate. Scale through demonstration, not mandate. This is the PDIA principle restated in Scott’s vocabulary: the path to functional institutions is discovered through local experimentation, not prescribed from above.

The Minimum Viable Institution

The four institutional designs above are full-scale architectures. They describe what the end state could look like. But institutional development does not begin at full scale. OSDMA did not begin as a world-class disaster management authority. It began as a small team of officers, traumatised by the 1999 cyclone, trying to ensure that the next one would not kill 10,000 people. The institutional capacity that produced Fani’s 64 deaths in 2019 was built over twenty years of iterative development, learning from each cyclone event, expanding capacity after each demonstration of competence.

The question is: what is the minimum viable institution (MVI) for each domain? In software terms, the minimum viable product is the smallest version of a product that delivers enough value to justify continued development. The MVI is the smallest institutional design that could demonstrate enough results to justify expansion.

The MVI for industrial development

Not the full OIDA. A small, fixed-term investment conversion unit — perhaps 15-20 people, half of them sector specialists hired on contract, housed outside the Industries Department, with a mandate to convert a specific set of investment commitments (say, the 50 largest MoUs from the 2025 Utkarsh Odisha conclave) into operational production within 36 months.

The unit would have: a director with a three-year fixed term, authority to coordinate across departments (backed by a Chief Secretary-level coordination committee), a small discretionary budget for obstacle removal (the Rs 2 crore that unblocks a Rs 200 crore investment by funding the last kilometre of road or the missing power substation), and a public dashboard tracking each project’s progress from MoU to production.

If the unit converts 20 of 50 targeted MoUs into production within 36 months — against the historical rate that would predict 12-15 — the demonstrated improvement justifies scaling to a permanent authority. If it fails, the failure is small, contained, and informative. This is the PDIA principle: start with the problem, try small, learn, scale.

OSDMA’s early years followed this logic, even if nobody articulated it in PDIA terms. The initial post-1999 capacity was small: basic cyclone shelters, rudimentary warning systems, a small coordinating body. Each subsequent cyclone provided a test. Each test generated learning. Each learning cycle expanded capacity. By 2019, the institution that had started with emergency shelters was evacuating 1.2 million people in 48 hours.

The MVI for education-employment coordination

Not the full authority. A pilot in two industrial clusters — Kalinganagar and Angul-Jharsuguda — where the authority designs curriculum co-location partnerships between three ITIs and three industrial employers in each cluster. Six partnerships total. Each partnership specifies: the skills the employer needs, the curriculum modifications required to produce those skills, the apprenticeship positions the employer will provide, and the placement guarantee upon completion.

If after three years the ITI graduates from these partnerships show employment rates 50% higher than graduates from non-partnership ITIs, and if the employers report that the graduates meet their skill requirements without additional training, the model justifies scaling. If the partnerships fail — because the employers do not honour commitments, or the ITIs cannot modify curricula, or the institutional coordination does not work — the failure teaches what the actual constraints are, which no amount of theoretical design can anticipate.

The MVI for urban planning

Not the full authority. A Rourkela City Development Initiative — a small team with a two-year mandate to prepare a comprehensive city development plan for Rourkela’s transformation from a steel town of 500,000 to a university-industry city of 1 million. The plan would cover: public transport, water and sanitation, housing, cultural infrastructure, healthcare, and the specific institutional reforms (NIT-industry partnerships, startup incubation, cultural programming) that would make Rourkela a city that retains NIT graduates rather than exporting them.

If the plan is implementable and attracting investment within three years, it becomes the template for Berhampur and potentially for the Paradip-Cuttack corridor. If the plan fails — because the institutional capacity to implement it does not exist, or the political will to invest in a single city is insufficient, or the land and infrastructure constraints are binding — the failure reveals the specific constraints that must be addressed before urban development at scale is possible.

The MVI for agriculture

Not the full authority. An expansion of the Odisha Millet Mission model to rice — the crop that occupies 60-65% of gross cropped area and determines the agricultural productivity of the state. A Rice Productivity Mission in five districts, with OMM-style institutional features: focused scope, community-based approach through FPOs, value chain coverage from seed to market, dedicated implementation cell with consistent leadership, and a clear metric (rice yield per hectare in target districts versus control districts).

If after five years the target districts show yield improvements of 30-40% (closing the gap with the national average), the model justifies statewide scaling. The OMM precedent makes this the most realistic of the four MVIs — it builds on demonstrated institutional success rather than requiring entirely new institutional capacity.

What MVIs require

All four MVIs share three requirements that the full-scale designs also share.

First, leadership continuity. The director or head of each MVI must have a fixed term — minimum three years — that cannot be shortened by routine transfer. Without this, the MVI will be led by a succession of officers who each spend six months understanding the problem before being replaced. This is the most fundamental requirement and the most politically difficult, because it requires the Chief Minister to voluntarily surrender one instance of transfer authority.

Second, measurable outcomes. Each MVI must have a metric that is publicly visible, regularly updated, and resistant to gaming. MoU-to-production conversion rate. ITI graduate employment rate. City development plan milestones. Rice yield per hectare. The metric creates accountability that schemes do not create, because the metric is tied to the MVI’s continued existence. If the metric does not improve, the MVI has no justification for continuation. If it improves, the MVI has a demonstrated case for expansion.

Third, external scaffolding. OSDMA’s early development was supported by international technical assistance — the UNDP, the World Bank, and bilateral agencies provided expertise, technology, and funding that the state government could not supply from its own resources. Each MVI would benefit from similar external scaffolding: industry associations providing sector expertise to the investment conversion unit, international vocational training organisations providing curriculum design support to the education-employment partnerships, urban planning consultancies supporting the city development initiative, agricultural research institutions (IRRI, ICRISAT) supporting the rice productivity mission. The scaffolding provides capabilities that take years to build internally, bridging the gap between the MVI’s ambitious mandate and its initially limited capacity.

Probability

Where does this leave the institutional design question? Chapter 3 assessed which of OSDMA’s seven success factors are transferable. This chapter has attempted to design institutional architectures that embody those transferable factors. The honest probability assessment:

50-55% — Continued institutional stasis with scheme proliferation. New schemes are launched. Old schemes are rebranded. Investment conclaves produce impressive MoU numbers. The education budget grows. Urban mission funds are allocated. Agricultural productivity improves at 2-3% annually through the slow diffusion of better seeds, fertiliser, and irrigation — not through institutional innovation. OSDMA continues to work. Nothing else develops comparable capacity. The BJP government replicates the Naveen model with different personnel and different branding but the same structural logic.

25-30% — One or two MVIs are attempted. The investment conversion unit or the rice productivity mission (the two most politically tractable designs) is created, operates for three to five years, and produces measurable results. If successful, it creates a demonstration effect that justifies expansion. If unsuccessful, it generates learning about what the actual constraints are. This outcome requires a political entrepreneur at the CM or senior bureaucrat level who sees institutional investment as a political strategy — not a replacement for welfare but a complement to it.

10-15% — Systemic institutional development. Multiple autonomous authorities are created across industrial development, education, urban planning, and agriculture, with the design features described above: clear mandates, operational autonomy, leadership continuity, measurable outcomes, and community engagement. This outcome requires a combination of political will, bureaucratic entrepreneurship, external scaffolding, and the kind of sustained commitment across electoral cycles that Tamil Nadu’s welfare ratchet produced through competitive political dynamics. It would represent the most significant institutional innovation in Odisha since OSDMA’s creation.

5% — The operating system itself is rebuilt. Not individual applications but the underlying bureaucratic culture: tenure stability, merit-based posting, outcome accountability, professional development. This is the Chapter 8 scenario — the full operating system upgrade that would make individual institutional designs unnecessary because the default institutional quality would be sufficient. It has no precedent in Odisha’s history and would require a level of sustained political commitment to institutional reform that no Indian state has achieved.

The OSDMA question — the question that orients this entire series — applies to institutional design with particular precision. OSDMA proved that the capacity exists. The body can mount an immune response. The question is whether the immune response can be generalised — whether the specific antibodies that fight cyclones can be adapted to fight industrial absence, educational export, urban dysfunction, and agricultural stagnation.

The answer from the design exercise is: yes, in principle. The design features that made OSDMA work — clear mandate, operational autonomy, uncheateable metrics, leadership continuity, community engagement, external scaffolding, and iterative learning — are not specific to disaster management. They are general principles of institutional design. Ostrom identified them. The PDIA framework formalises them. The comparator cases in Chapter 5 confirm them.

The constraint is not design but politics. The patronage equilibrium actively resists the creation of autonomous institutions because autonomous institutions reduce discretionary power. The IAS cadre resists fixed-tenure appointments because they disrupt the rotation system that the cadre’s career structure depends on. The scheme-based budget system resists institutional permanence because institutions cannot be announced, branded, and claimed as electoral achievements the way schemes can.

The question is not whether functional institutions can be designed for Odisha. They can. This chapter has attempted to design four of them. The question is whether the political conditions exist — or can be created — under which these designs could be implemented. That is the question Chapter 7, on the builders, addresses: not what to build, but who builds it, and what conditions produce the builders.

Sources

Research and Academic

Andrews, M., Pritchett, L., and Woolcock, M. Building State Capability: Evidence, Analysis, Action. Oxford University Press, 2017.
Ostrom, E. Governing the Commons: The Evolution of Institutions for Collective Action. Cambridge University Press, 1990.
Ostrom, E. “Beyond Markets and States: Polycentric Governance of Complex Economic Systems.” American Economic Review, vol. 100, no. 3, 2010, pp. 641-672.
Scott, J.C. Seeing Like a State: How Certain Schemes to Improve the Human Condition Have Failed. Yale University Press, 1998.
Pritchett, L. “Is India a Flailing State?: Detours on the Four Lane Highway to Modernization.” HKS Working Paper RWP09-013, 2009.
Pritchett, L., Woolcock, M., and Andrews, M. “Looking Like a State: Techniques of Persistent Failure in State Capability for Implementation.” Journal of Development Studies, vol. 49, no. 1, 2013, pp. 1-18.
Rodrik, D. One Economics, Many Recipes: Globalization, Institutions, and Economic Growth. Princeton University Press, 2007.
Kapur, D. and Mehta, P.B. “Why Does the Indian State Both Fail and Succeed?” Journal of Economic Perspectives, 2020.
Fukuyama, F. State-Building: Governance and World Order in the 21st Century. Cornell University Press, 2004.
North, D.C. Institutions, Institutional Change and Economic Performance. Cambridge University Press, 1990.
Acemoglu, D. and Robinson, J.A. Why Nations Fail: The Origins of Power, Prosperity, and Poverty. Crown Publishers, 2012.

Odisha Government and Data

IPICOL / GO-SWIFT Portal — investodisha.gov.in
IDCO — Industrial estates and infrastructure data
Industrial Policy Resolutions 2001, 2007, 2015, 2022 — Government of Odisha
Utkarsh Odisha 2025: Rs 16.73 lakh crore investment intentions — Make in Odisha - Wikipedia
CAG Audit Reports on Odisha (2018-2024) — cag.gov.in
Odisha Economic Survey 2024-25, 2025-26
Odisha Budget 2026-27: Education allocation Rs 42,565 crore — Kalinga TV
NITI Aayog Fiscal Health Index 2025: Odisha ranked first — NITI Aayog
PRS India, Odisha Budget Analysis 2025-26
OSDMA official data — osdma.org
UDISE+ 2023-24, 2024-25
ASER 2022, 2024

Comparator Institutions

EDB Singapore — edb.gov.sg
GIDC Gujarat — gidc.gujarat.gov.in
iNDEXTb — Gujarat’s Industrial Extension Bureau
SIPCOT Tamil Nadu — sipcot.in
TIDCO Tamil Nadu — tidco.com
TNSDC — Tamil Nadu Skill Development Corporation
Singapore HDB — Housing and Development Board
Amul/NDDB model — National Dairy Development Board
Odisha Millet Mission — Department of Agriculture, Government of Odisha; WASSAN; FAO/IFAD publications

Prior SeeUtkal Series

The Missing Middle / Value Chain — mineral value chain economics, value-addition absence
Education Odisha — the compiler that targets the wrong machine, coordination failure
Urbanisation Odisha — 17% urbanisation, the missing urban platform
The Leaving — migration as exit from failed institutions
The Long Arc — extraction equilibrium, hollow institutions
Tribal Odisha — PESA violations, constitutional betrayal
Environmental Odisha — OSDMA and chronic climate adaptation
Women’s Odisha — Mission Shakti, institutional features of SHG networks

Earlier Chapters in This Series

Chapter 1: The Exception — OSDMA as proof of institutional capacity
Chapter 2: The Machine That Runs on Nothing — the broken operating system
Chapter 3: The Anatomy of the Exception — seven factors, transferability assessment
Chapter 4: The Patronage Equilibrium — why institutional weakness is a Nash equilibrium
Chapter 5: What Worked Elsewhere — comparator institutional architectures

Source Research

The raw research that informs this series.