English only · Odia translation in progress

Chapter 4: The Patronage Equilibrium

On the morning of June 14, 2024 — two days after Mohan Charan Majhi was sworn in as Odisha’s first tribal Chief Minister — the Government of Odisha issued transfer orders for 50 IAS officers. District Collectors in 15 of 30 districts were replaced. The Principal Secretary to the Chief Minister was changed. The Home Secretary, the Agriculture Secretary, the Industries Secretary, the Tribal Development Secretary — all reassigned. Officers perceived as close to V.K. Pandian or the BJD leadership received new postings. Some were moved to what the bureaucratic ecosystem euphemistically calls “waiting” — positions with no operational authority, the administrative equivalent of a holding pen. Over the next six months, more than 200 IAS and OAS officers would be transferred or reassigned, the most significant bureaucratic reshuffling in two decades.

This was not exceptional. It was exactly what the system was designed to do.

When the BJD took power in 2000, a comparable wave had rolled through the same buildings. When the BJD-BJP alliance broke in 2009 and Naveen Patnaik consolidated control, another wave. When V.K. Pandian was elevated to cabinet rank in October 2023, another. When Pandian was given authority over the 5T Secretariat in 2019, another. The bureaucratic apparatus of Odisha has been reshuffled so many times, with such regularity, that the reshuffling itself has become the system. The 2024 wave was notable only for its scale — 200 officers in six months — not its nature.

The question Chapter 2 posed was: what does the broken operating system look like at ground level? The question Chapter 3 posed was: what made OSDMA the exception? This chapter asks the question that sits underneath both: why does the system stay broken? Not how it breaks — the transfer-posting regime, the scheme proliferation, the monitoring-without-learning, the vacancy burden — but why nobody fixes it. The answer is not incompetence. It is not ignorance. It is not cultural. It is structural, and it is best understood through the lens of game theory.

What a Nash Equilibrium Actually Is

In 1950, John Nash proved a theorem that would earn him the Nobel Prize four decades later. He demonstrated that in any finite game where players choose strategies simultaneously, there exists at least one combination of strategies where no player can improve their outcome by unilaterally changing their own strategy. This combination is a Nash equilibrium. It is not necessarily the best outcome for anyone. It is not necessarily the socially optimal outcome. It is simply the outcome from which no individual player has an incentive to deviate, given what everyone else is doing.

The power of the concept is in the word “unilaterally.” In a Nash equilibrium, every player might agree that a different outcome would be better for everyone. But getting there requires coordinated action — multiple players changing strategy simultaneously. And each player, looking at their own position, concludes that changing alone would make them worse off. So nobody moves. The suboptimal outcome persists, not because anyone wants it, but because nobody can improve on it alone.

The classic illustration is the prisoner’s dilemma: two criminals arrested separately, each offered leniency for betraying the other. Both would be better off if both stayed silent. But each, reasoning individually, betrays — because if the other stays silent, betraying produces a better outcome, and if the other betrays, betraying produces a less bad outcome. The Nash equilibrium is mutual betrayal. The optimal outcome is mutual cooperation. The equilibrium prevents the optimum.

Odisha’s institutional weakness is a Nash equilibrium with four players: the politician, the bureaucrat, the contractor, and the voter. Each is acting rationally given what the others are doing. The result is institutional hollowness that everyone recognises and nobody fixes.

The Naveen Model

To understand the equilibrium, start with the 24-year experiment that produced and stabilised it.

Naveen Patnaik entered Odisha politics in 1997 after his father Biju Patnaik’s death. He was 51, Delhi-raised, Doon School-educated, a published author and socialite who had never fought an election and spoke no Odia publicly. He won the 1997 Aska by-election on name recognition alone. He formed the BJD in December 1997 — named after his father, carrying no ideology beyond Odia regional identity and anti-Congressism. He became Chief Minister on March 5, 2000, five months after the super cyclone, and held the post for 24 years, 3 months, and 7 days.

The electoral record is the first piece of the equilibrium puzzle:

YearAssembly Seats (of 147)BJD Vote ShareLok Sabha Seats (of 21)
200068 (alliance: NDA 106)29.40%—
200461 (alliance: NDA 99)27.36%11
2009103 (solo)38.86%14
2014117 (solo)43.35%20
2019112 (of 146)~44%12
202451 (lost)~33%1

Two features of this trajectory matter for the institutional argument.

First: the 2009 break. When Patnaik split with the BJP after the Kandhamal communal violence — 100 people killed, 50,000 displaced, 300 churches destroyed — the calculation was not principled but strategic. Going solo would consolidate minority and liberal votes. The gamble paid off spectacularly: BJD jumped from 61 alliance seats to 103 solo seats, the largest single-party tally in Odisha’s history at that point. The lesson Patnaik’s system absorbed was that electoral success depended not on a party machinery but on the leader’s image and the welfare architecture’s reach. The BJD had no cadre structure, no booth-level organisation, no internal democracy, no ideological programme. It was a personality vehicle. This is not a criticism. It is a structural description with institutional consequences.

Second: the 2014 peak. In the year the Modi wave swept India, when every regional party from Trinamool to TDP to AIADMK lost ground nationally, BJD won 117 assembly seats and 20 of 21 Lok Sabha seats. Odisha was the only major state where the ruling regional party not only survived but strengthened. Odia voters split: Modi for PM, Naveen for CM. This split-voting pattern was the most dramatic evidence that the welfare-based patronage architecture had created beneficiary-government links strong enough to withstand the most powerful national political current in decades.

The governance architecture that produced these results rested on a specific design: centralise all significant decision-making in the Chief Minister’s Office at Naveen Niwas — Patnaik’s private residence on Aerodrome Road in Bhubaneswar. Route transfers, postings, scheme design, industrial clearances, and electoral strategy through the residence rather than through normal institutional channels. Convert ministers into implementers rather than policy-makers. Make cabinet meetings perfunctory. Let the IAS cadre govern while the leader provides the clean brand above politics. District collectors, not party workers, become the interface between government and citizens.

This architecture produced two outcomes simultaneously, and understanding the simultaneity is critical to understanding the equilibrium.

Outcome one: stability and predictability. Investors, bureaucrats, and international agencies knew that decisions came from one node. The governance system was remarkably orderly by Indian state standards. The fiscal discipline was genuine — Odisha maintained revenue surpluses, kept debt-to-GSDP below 16 percent, built a Budget Stabilisation Fund exceeding Rs 20,890 crore. The bureaucracy functioned as a quasi-professional apparatus because the leader was not extracting from it in the crude ways that characterise patronage systems in some other Indian states.

Outcome two: fragility and dependency. The entire architecture depended on the functioning of a single node. When that node degraded — through Patnaik’s advancing age (he was 77 at the time of the 2024 defeat), through the controversies surrounding his increasing dependence on V.K. Pandian, through the simple passage of 24 years of incumbency — there was no institutional fallback. The party had no internal mechanisms because the party was the leader. The bureaucracy had no autonomous capacity because the bureaucracy reported to the leader.

V.K. Pandian’s trajectory is the clearest illustration of how centralisation simultaneously enabled governance and prevented institutionalisation. Pandian — a 2000-batch IAS officer from Tamil Nadu, allocated to the Odisha cadre after marrying a batchmate — served as Private Secretary to the CM from 2011, gradually accumulating influence as the sole gatekeeper. In 2019, he was appointed Secretary of the 5T Initiative. In October 2023, he took voluntary retirement from the IAS and was appointed Chairman of 5T and Nabin Odisha with cabinet rank. In November 2023, he formally joined the BJD. The 5T framework — Transparency, Technology, Teamwork, Time, Transformation — gave Pandian formal authority over the state’s entire administrative apparatus. He reviewed government files, toured districts with CM-level authority, used state aircraft, recommended election candidates, conducted review meetings with district collectors and senior bureaucrats, and became the person through whom every significant decision was routed.

The 5T system was not without achievement. Over 4,500 government high schools received infrastructure upgrades. Mo Sarkar — the random citizen feedback mechanism where government officers were called by a centralised call centre to assess service quality — produced genuine accountability pressure in some departments. The digital dashboards tracking file movement and scheme delivery were probably the most sophisticated real-time governance monitoring system attempted by any Indian state government.

But every one of these achievements was routed through the 5T Secretariat, which was Pandian, which was a single person accountable to a single leader. When the 5T node collapsed after the 2024 election, the system had no fallback. The BJP government dismantled 5T within weeks. Mo Sarkar was discontinued. The school transformations continued under different branding because the physical infrastructure was built, but the governance innovation — the accountability mechanism, the cross-departmental coordination, the real-time monitoring — disappeared because it was not institutionalised. It was personalised. And personalised systems die with the personality.

The paradox within the Naveen model is that it simultaneously built OSDMA and prevented the OSDMA model from spreading. The same centralisation that enabled rapid decision-making during cyclone response — clear command authority, bypassed departmental silos, directly accountable leadership — also prevented the development of autonomous institutional capacity in every other domain. OSDMA worked because it was designed with institutional autonomy: professional leadership, clear protocols, performance metrics tied to lives saved rather than political approval. The rest of the governance system worked through political dependency: every significant decision routed through Naveen Niwas, every officer’s career determined by accommodation to the centre. The Naveen model could build one OSDMA. It could not build ten, because ten autonomous institutions would have distributed the authority that the model concentrated. The centralisation that enabled the exception was the same centralisation that prevented the exception from becoming the norm.

The Naveen model created the conditions for the patronage equilibrium by demonstrating that centralised, personality-dependent governance could produce genuine welfare outcomes and win elections repeatedly. The model’s success made it the template. Its structural fragility was invisible until it broke.

Welfare as Patronage

The welfare architecture was not decorative. It was the transmission mechanism through which mineral revenue was converted into electoral support. Understanding the specific schemes is necessary because the specifics reveal the patronage logic.

KALIA (Krushak Assistance for Livelihood and Income Augmentation): Launched December 2018 — five months before the May 2019 general election. Rs 10,000 per year to small and marginal farmers, Rs 12,500 per year to landless agricultural households. The scheme covered 65.64 lakh beneficiaries with a total budget exceeding Rs 10,000 crore. Its design innovation — covering tenant farmers and sharecroppers, not just landowners — was substantively superior to PM-KISAN, launched two months later by the central government but restricted to landowners. In a state where 93 percent of holdings are small and marginal and substantial farming is done by tenants without title, KALIA reached people PM-KISAN could not.

The CAG found Rs 782 crore distributed to 12.72 lakh ineligible beneficiaries — government employees, income tax payers, individuals owning four-wheelers. An 18 percent error rate. The ineligible beneficiaries were not random errors. Beneficiary lists were generated at the gram panchayat level with input from sarpanch, ward members, and block functionaries — all actors with political affiliations. The 12.72 lakh inclusions and the corresponding exclusions operated within an ostensibly universal scheme, using the administrative machinery of local governance to sort supporters from non-supporters. This is the discretionary power premium in action: the scheme design appears universal, the implementation reality is selective, and the selectivity is the patronage.

BSKY (Biju Swasthya Kalyan Yojana): Smart health cards providing cashless treatment up to Rs 5 lakh per year per family, Rs 10 lakh for women. Coverage: approximately 96.5 lakh families — 81 percent of the state’s population. Over 1.1 crore patients used the scheme. BSKY created a direct link between the state government and the individual beneficiary: the card bore the government’s name, the treatment bore no out-of-pocket cost (in theory), and the emotional register of health — relief from the fear of catastrophic medical expense — is among the most powerful political adhesives available. CSDS/Lokniti data from Odisha elections indicates that BSKY cardholders showed 18-22 percentage points higher support for BJD than non-cardholders. The scheme did not merely buy votes. It created a constituency whose rational fear of losing the card — of what happens if the next government discontinues coverage — made supporting the incumbent the default strategy.

Mission Shakti: From 41,475 self-help groups in 2001 to over 600,000 SHGs with roughly 70 lakh women members. One in three adult women in the state belonged to a group. Loan recovery hit 96 percent. Recognised by the UN Capital Development Fund as India’s largest women’s institutional network. Mission Shakti was simultaneously genuine empowerment (income, credit access, collective identity) and parallel political infrastructure (a network of 70 lakh women organised into groups that the state government had created, funded, and maintained relationships with). The BJP government kept the infrastructure intact after 2024 — rebranding its cafes and bazaars but not dismantling the SHG network. The surest sign that a welfare programme delivers real value is that the opposition retains it.

Mamata: Rs 5,000 conditional cash transfer for pregnant and lactating women. Small amount, enormous emotional resonance. Peer-reviewed evaluation in the Journal of Nutrition (2022) found it reduced child wasting by 7 percentage points — though the equity finding was critical: the reduction was driven primarily by wealthier households, with the poorest quintile seeing far smaller gains. Even a scheme that works creates stratified patronage.

The aggregate numbers: total welfare spending in Odisha runs between Rs 10,000 and Rs 14,000 crore per year, covering KALIA, BSKY, PDS rice at Re 1 per kg, Mission Shakti, Mamata, social pensions (Madhu Babu Pension Yojana covering approximately 48 lakh beneficiaries at Rs 500-700 per month, costing approximately Rs 3,000 crore annually), and numerous smaller schemes.

Each scheme individually is defensible. Many are genuinely effective. The collective effect is the equilibrium: a welfare architecture comprehensive enough that a voter’s rational strategy is to support whichever party credibly commits to maintaining it. The BJP understood this in 2024. It did not run against welfare. It ran on better welfare — continuing BSKY under a new name (Gopabandhu Jana Arogya Yojana), restructuring KALIA into CM Kisan Yojana merged with PM-KISAN to deliver Rs 10,000 with credit split 60/40 between centre and state, and launching Subhadra Yojana (Rs 10,000 per year to women aged 18-60, budget Rs 10,000 crore for 2025-26). The first Subhadra instalment was disbursed in September 2024, three months after taking power, timed to coincide with the Nuakhai festival. The timing was as strategic as KALIA’s pre-election launch. Different party. Same logic. The equilibrium absorbed the new player.

Transfer-Posting as Currency

If welfare schemes are the output of the patronage system, transfer-posting is its operating currency. This is not metaphor. Transfer-posting is the single most valuable tool of political control in Odisha, as in most Indian states, and it is the primary mechanism through which institutional weakness is maintained.

The numbers from Chapter 2 bear repeating in this context because they acquire a different meaning when viewed through the game theory lens. The average tenure of a District Collector in Odisha is 14-18 months. The annual transfer probability for IAS officers across Indian states is approximately 53 percent — more than half of all officers in position can expect to be moved in any given year. In the first BJP reshuffle after June 2024, over 50 IAS officers were moved. Within six months, over 200 IAS and OAS officers had been transferred.

The comparison case that illuminates the structure is Ashok Khemka — a Haryana-cadre IAS officer who has been transferred 57 times in a career spanning roughly three decades. Khemka’s case is extreme but instructive: an officer who repeatedly enforced regulations against politically connected interests — he cancelled a land deal involving Robert Vadra that was later investigated by multiple agencies — was moved as punishment each time. The transfer was not a personnel management decision. It was a disciplinary tool used by successive governments of different parties, because the logic of transfer-as-discipline is party-neutral. The signal to the rest of the cadre was unmistakable: enforce rules against powerful interests and your career will be disrupted. Accommodate those interests and your career will be smooth.

The Supreme Court has attempted to address this. The Prakash Singh v. Union of India directives (2006) mandated minimum tenure guarantees for police officers. The TSR Subramanian Committee (2014) recommended a minimum two-year tenure for all IAS officers. The Supreme Court in TSR Subramanian v. Union of India (2013) directed states to establish Civil Services Boards to manage postings. Odisha constituted a Civil Services Board. It is advisory, not binding. The Chief Minister retains effective authority over all significant IAS postings. The Board’s recommendations can be — and routinely are — overridden.

Why would any political leader voluntarily surrender the power to transfer officers? This is where the game theory becomes uncomfortably precise.

A politician who controls transfers controls the bureaucracy. An officer who can be moved at any time has a powerful incentive to accommodate the politician’s preferences — whether those preferences involve directing scheme benefits to political supporters, prioritising construction contracts for connected firms, or simply not enforcing regulations that would inconvenience allies. An officer who knows they will remain in position for a guaranteed minimum term has less reason to accommodate and more reason to follow institutional mandates.

Strong institutions constrain political discretion. Weak institutions preserve it. A functional agriculture department with secure-tenure officers implementing evidence-based extension services would produce better agricultural outcomes. It would also reduce the politician’s ability to direct agricultural subsidies, procurement centres, and fertiliser dealer licences toward political supporters. A functional education department with stable leadership and outcome-based accountability would improve learning. It would also reduce the politician’s ability to influence teacher postings, textbook contracts, and school construction tenders. A functional police force with guaranteed minimum tenures would improve law enforcement. It would also reduce the politician’s ability to influence FIR registration, investigation direction, and the selective application of the law.

The discretionary power premium — the political value of maintaining institutions that are functional enough to deliver visible outputs but weak enough to allow political direction of those outputs — is the core of the equilibrium. It is not a conspiracy. It is a rational calculation made by every political leader in every Indian state, in every electoral cycle, under every party flag.

The Mining-Politics Nexus

The fiscal engine that powers the equilibrium is mineral extraction. Mining constitutes approximately 84 percent of Odisha’s non-tax revenue. The state holds 28 percent of India’s iron ore, 98 percent of its chromite, 55 percent of its bauxite. The Odisha Mining Corporation alone reported net profit of Rs 9,076 crore in 2023-24.

Before 2015, mining leases were allocated through a combination of first-come-first-served and discretionary allocation. The Shah Commission (2012) documented illegal mining worth Rs 59,000 crore in Odisha — 63 cases in Keonjhar district alone, systematic violation of environmental clearance conditions, encroachment on forest land, under-reporting of mineral production. The 2015 MMDR Amendment mandated auction-based allocation, and Odisha has been the most aggressive state in conducting auctions. The auction regime reduced discretionary allocation of leases.

But patronage is adaptive. When one channel closes, it migrates to adjacent areas. Post-2015, the discretionary opportunities shifted: land acquisition for mining infrastructure, facilitation of environmental clearances, transport permits, and — most significantly — DMF spending decisions.

The District Mineral Foundation was created to channel a portion of mining royalties back into mining-affected communities. Odisha leads India in DMF collection: approximately Rs 31,324 crore cumulative by 2025. Keonjhar alone has collected over Rs 8,840 crore — the highest of any district in India. The spending problem is where patronage re-enters. Odisha consistently spends only 50-55 percent of collected DMF funds. Rules require at least 60 percent on “high priority areas” — drinking water, healthcare, education, environment in mining-affected communities. What actually happens: funds flow toward roads, buildings, and water supply benefiting broader populations rather than the specific mining-affected communities — mostly tribal, mostly poor — that bear extraction’s social and environmental costs. The District Collector supervises; a governing council decides. Mining-affected communities rarely make spending decisions. CAG has flagged diversion of funds to non-mining areas and urban areas. DMF money, created to compensate for extraction, is itself extracted from its intended purpose through the same patronage channels.

But mining is only the most visible patronage stream. The equilibrium runs on multiple currencies simultaneously.

Sand mining: Odisha’s rivers — Mahanadi, Brahmani, Baitarani, Subarnarekha — produce construction sand with an estimated legal extraction of 40-50 million cubic metres annually. Illegal extraction is estimated at 1.5 to 2 times the legal volume. Legal sand generates Rs 500-800 crore annually for the state exchequer; the illegal economy is estimated at Rs 1,000-3,000 crore. Sand mining leases require political connections. MLAs, block-level politicians, and local strongmen control access to river stretches. Illegal operations require the complicity of Revenue Department officials, police, and transport officials. The National Green Tribunal has repeatedly passed orders against illegal sand mining in Odisha with limited effect. In 2023, the Odisha High Court ordered a CBI investigation into illegal sand mining in several districts after state police investigations were alleged to be inadequate. The architecture is the same as mineral mining but operating at a more granular, more localised level — the equilibrium functioning at the district and block scale rather than the state scale.

Liquor licensing: Excise revenue is the second-largest source of the state’s own tax revenue after GST — approximately Rs 6,500-7,000 crore in 2023-24, with targets of Rs 7,800 crore for 2025-26. Liquor retail licence allocation is a highly discretionary process. Who gets a licence and where is a political decision. Country liquor manufacturing licences are controlled. Illicit liquor operations require political and police complicity. Periodic hooch tragedies — deaths in Mayurbhanj (2023), recurrent incidents in Ganjam — expose the gap between regulatory mandate and ground reality. The BJP government restructured excise policy in 2024-25, claiming reduced political interference. The revenue targets increased. The underlying patronage logic — discretionary allocation of valuable licences — remained structurally intact.

Government recruitment: In a state where private sector employment is limited, government jobs are the most sought-after career path. The Odisha Public Service Commission (OPSC) and Odisha Staff Selection Commission (OSSC) conduct recruitment that has been plagued by paper leaks, irregularities, and allegations of political influence. The 2022 junior clerk exam was cancelled after paper leak allegations. OPSC medical officer recruitment in 2021 faced High Court intervention. Approximately 16,000 teacher vacancies persist across the state (UDISE+ 2023-24), with the recruitment process slow and allegations of political influence on postings consistent. The state employs a large number of contractual workers — siksha sahayaks, anganwadi workers — at significantly lower wages than regular employees, creating a two-tier workforce where regular employment generates political loyalty and contractual workers lack the same voice.

Each of these patronage streams — mineral leases, sand mining, liquor licences, government recruitment, construction contracts — operates within its own domain but follows the same structural logic: discretionary allocation of valuable resources through political channels. The equilibrium is not a single game but a portfolio of interlocking games, each reinforcing the others, each making unilateral deviation more costly.

The Welfare-as-Substitute Logic

The equilibrium does not merely produce welfare as electoral currency. It produces welfare as a substitute for institutional development — and the substitution is the mechanism that locks the trap.

Consider agriculture. KALIA provides Rs 10,000 per year to farmers. The institutional alternative — functional agricultural extension services (one officer per 800 farmers instead of the current 1:1,500), reliable irrigation (35-40 percent coverage versus Punjab’s 98 percent), cold chain infrastructure (7-8 lakh MT capacity against a requirement of 15-20 lakh MT), efficient market linkages (52 regulated mandis for 64.5 lakh farm households) — would cost more in the short term, take longer to build, and produce benefits harder to attribute to a specific political actor. Rs 10,000 in a farmer’s bank account is visible, immediate, and stamped with the government’s name. An extension officer who improves the farmer’s yield by 500 kg per hectare over three years is invisible, delayed, and attributable to weather as much as policy.

Consider health. BSKY covers 81 percent of the population for hospital treatment. The institutional alternative — adequate primary health centres with staff, medicines, and equipment (doctor-population ratio of 1:1,800 versus the WHO recommendation of 1:1,000; approximately 50 percent specialist vacancies in Community Health Centres; 1,254 PHCs many operating below capacity) — would reduce the need for BSKY by preventing the diseases and complications that BSKY treats. BSKY’s design — cashless treatment at empanelled private hospitals — acknowledges the failure of public health infrastructure by routing patients to the private sector. Every rupee spent on BSKY treatment at a private hospital is a rupee not invested in making the public system functional. The scheme does not fix the institution. It routes around it.

The logic extends to the judiciary. Odisha has approximately 10-11 lakh pending cases across all courts, with judicial vacancies running at 25-30 percent of sanctioned positions. Average disposal time: 3-5 years for civil cases, 2-4 years for criminal. The non-functioning judiciary pushes dispute resolution into informal, politically-mediated channels — the MLA who resolves a land dispute, the party worker who mediates a family conflict, the contractor who “arranges” a favourable outcome. Judicial dysfunction does not merely result from the patronage equilibrium. It reinforces it, by making the political intermediary indispensable for a function the institution should be performing.

The substitution compounds over time. As institutions weaken, the welfare programmes that substitute for them become more politically valuable, which increases the incentive to maintain the welfare programmes, which reduces the incentive to repair the institutions, which makes the welfare programmes more essential. The cycle is self-reinforcing. It is Mancur Olson’s logic of collective action applied at scale: institutional reform is a public good whose benefits are diffuse (everyone benefits from a functional agriculture department) while its costs are concentrated (specific politicians lose discretionary power). Welfare distribution is a private good whose benefits are concentrated (the beneficiary receives the transfer) while its costs are diffuse (funded from general mining revenue). The political system naturally favours private goods over public goods, welfare over institutions, transfers over transformation.

The BJP Transition: New Faces, Same Logic

The 2024 election produced the most consequential political change in Odisha in a generation. BJP won 78 of 147 assembly seats and 20 of 21 Lok Sabha seats. The BJD collapsed from 112 seats to 51 in the assembly, from 12 to 1 in the Lok Sabha. The factors were multiple: 24 years of anti-incumbency, the Pandian identity crisis (the “Odia Asmita” campaign targeting his Tamil origin in a state whose founding myth was linguistic identity), simultaneous elections collapsing the split-voting pattern, central welfare schemes constructing a parallel patronage infrastructure through PM Awas Yojana and Ujjwala, and a decade of booth-level RSS organisational groundwork reaching critical mass.

Mohan Charan Majhi — a Santal tribal, four-time MLA from Keonjhar, born in 1969 in Raikala village — was the first tribal Chief Minister in Odisha’s history. His selection was strategic: tribal representation in a state where 22.85 percent of the population is Scheduled Tribe, mining belt credibility from Keonjhar (the state’s largest iron ore producer), and RSS organisational loyalty. The signal was clear: this government represents the Odisha that the BJD left behind.

What followed was structurally indistinguishable from a BJD government operating the same machinery.

Welfare architecture: retained and expanded. BSKY was renamed Gopabandhu Jana Arogya Yojana in February 2025 — same smart health cards, expanded coverage to Rs 6 lakh for urban areas, integrated with Ayushman Bharat. KALIA was restructured into CM Kisan Yojana: Rs 4,000 from the state plus Rs 6,000 from PM-KISAN equals Rs 10,000, with 45.77 lakh small farmers and 19.12 lakh landless farmers enrolled. PDS rice at Re 1 per kg: continued. Mission Shakti’s SHG infrastructure: retained, rebranded as Subhadra Shakti Cafes and Bazaars. And Subhadra Yojana was added on top: Rs 10,000 per year direct benefit transfer to women aged 18-60, budget Rs 10,000 crore for 2025-26, disbursement timed to festival dates for maximum emotional impact.

Transfer-posting regime: replicated. Over 200 officers transferred in six months. The officers changed. The mechanism did not. Civil Services Board: still advisory. Chief Minister’s authority over postings: intact.

5T/Mo Sarkar: dismantled. This is the one significant discontinuity, and it is instructive. The BJP government did not replace 5T with an equivalent governance reform framework. It simply removed the centralised monitoring layer that the BJD had used. The school transformations continued because the physical infrastructure was built. Mo Sarkar — the citizen feedback mechanism that created genuine accountability pressure — was discontinued. The governance innovation, being personalised to the Pandian-led architecture, was not retained because retaining it would have validated the previous regime’s approach. Political considerations overrode institutional logic. The new government preferred no accountability mechanism to the previous government’s accountability mechanism.

Mining policy: continued. DMF framework: unchanged. Auction-based lease allocation: continued. Mineral revenue: still the fiscal foundation. Make in Odisha/Utkarsh Odisha 2025 announced Rs 16.73 lakh crore in investment intentions — the same announcement economy with larger numbers.

Central-state alignment: new variable. For the first time in Odisha’s post-independence history, the same party controlled both the central and state governments. This alignment produced faster central scheme clearances, coordinated mining policy, and accelerated infrastructure announcements. It also eliminated the federalism-based incentive to build independent state capacity. When the centre and state are aligned, the state can substitute central transfers for institutional development. The “double engine” model may deepen the patronage equilibrium rather than disrupting it.

The structural conclusion is stark: the most consequential political change in a generation has produced, in institutional terms, minimal structural change. New branding. Same architecture. The game absorbed the new player.

Why the Equilibrium Is Stable

Now apply the Nash equilibrium framework formally. There are four players. Each has a dominant strategy — a strategy that produces the best outcome regardless of what the other players do. The combination of dominant strategies is the equilibrium.

Player 1: The Politician. The politician’s objective is to win elections and maintain power. Two strategies are available: invest in strong institutions (which constrain discretionary power but improve long-term outcomes) or maintain weak institutions (which preserve discretionary power and enable welfare-based patronage).

Strong institutions would mean: guaranteed officer tenures that prevent transfer-as-patronage, transparent procurement that eliminates contractor commissions (widely reported at 10-30 percent of contract value), merit-based scheme beneficiary selection that removes political intermediation, and independent regulatory bodies that enforce rules regardless of political preferences. Each of these reduces the politician’s ability to reward supporters and punish opponents.

Weak institutions preserve all of these levers. And the welfare architecture — funded by mining revenue that flows regardless of institutional quality — provides a visible, attributable mechanism for generating electoral support. CSDS/Lokniti data shows welfare scheme beneficiaries support the incumbent at 15-20 percentage points higher than non-beneficiaries. The rational calculation: discretionary power over weak institutions plus welfare distribution produces more electoral return than strong institutions that deliver better long-term outcomes but reduce political control.

The politician’s dominant strategy: maintain institutional weakness and distribute welfare. This holds under both BJD and BJP. It holds under Naveen Patnaik and Mohan Majhi. The strategy is not party-specific. It is structurally determined.

Player 2: The Bureaucrat. The bureaucrat’s objective is career advancement and posting comfort. Two strategies: perform against institutional mandates (which may conflict with political preferences and trigger transfers) or accommodate political preferences (which ensures career stability and access to desirable postings).

Under the current equilibrium, accommodation is dominant. A Block Development Officer who prioritises MGNREGA implementation according to rules over an MLA’s preferred contractor may find themselves transferred within weeks. A District Collector who enforces environmental regulations against a politically connected mining operation may find themselves moved to a “punishment posting” — a remote tribal district like Malkangiri or Nabarangpur, far from the career-advancement circuits of Bhubaneswar. The informal hierarchy of desirable postings is well understood: at the top, Chief Secretary, Finance, Revenue, Home, OMC Managing Director, Collector of Khordha; at the bottom, Secretary of Tourism or Odia Language and Literature, Collector of remote districts, positions in training institutions without executive authority. The distance between top and bottom is the disciplinary range within which the transfer mechanism operates.

With a 53 percent annual transfer probability and average postings of 14-18 months, the rational bureaucrat learns to accommodate. Not because they lack competence or integrity — the OSDMA experience proves the cadre contains officers of the highest calibre — but because the incentive structure punishes institutional fidelity and rewards political accommodation. The bureaucrat who accommodates gets posted to Finance or Revenue. The bureaucrat who resists gets posted to the training institute. Ashok Khemka’s 57 transfers are the extreme case. The lesson every IAS officer absorbs from his example is not about courage. It is about cost.

Player 3: The Contractor. State government expenditure on roads, bridges, buildings, and irrigation works runs to Rs 15,000-20,000 crore annually. The contractor’s objective is to win contracts and receive payment. Two strategies: compete on merit through transparent procurement (which requires investment in capacity, quality, and efficiency) or compete on political connection (which requires investment in relationships with politicians and bureaucrats).

Under the current equilibrium, political connection is dominant. Tender specifications can be designed to favour specific contractors through pre-qualification criteria. Large projects can be split into smaller ones to avoid higher scrutiny thresholds. Cartelisation — where groups of contractors rotate bids, each winning a share — is well documented. Quality compromise through sub-standard materials with inspector complicity produces short-term profit. Payment delays weaponised against politically unaligned contractors force them into patronage relationships. The commission system — 10-30 percent of contract value flowing back to political actors — is the tax that the patronage equilibrium extracts from every public works project.

CAG audit reports on Odisha’s Works Department, Rural Development Department, and Water Resources Department have repeatedly flagged cost overruns, quality deficiencies, and procedural violations. Roads that deteriorate within months of construction. Irrigation projects incomplete years after deadlines. The Pradhan Mantri Gram Sadak Yojana — a centrally funded rural roads programme with formal selection criteria based on population thresholds — has been flagged by CAG for instances of roads built to ineligible habitations while eligible ones waited. The formal criteria exist. The actual sequencing of which roads are built first is influenced by political considerations. Even a rule-bound central programme, when implemented through the state’s patronage-driven machinery, produces patronage-shaped outcomes.

The contractor who builds to specification and refuses to pay commissions loses contracts to the contractor who builds cheaply and pays. The competitive advantage is not competence. It is connection.

Player 4: The Voter. The voter’s objective is to maximise personal benefit from the political system. Two strategies: demand institutional reform (which produces diffuse, long-term benefits that are difficult to attribute to any specific politician) or demand direct benefits (which produce immediate, tangible, attributable returns).

Research on Indian voter behaviour — Kanchan Chandra’s work on ethnic parties, World Bank studies on political economy of public spending, CSDS/Lokniti election data — consistently finds that voters reward visible direct transfers (cash, subsidised rice, health cards), physical infrastructure (roads, buildings), and personal intervention (MLA resolving individual problems). Voters do not systematically reward institutional quality: better courts, more professional bureaucracy, transparent procurement. The political incentive therefore systematically favours visible transfers over institutional reform.

The 147 assembly constituencies create 147 centres of demand for localised, visible spending. Each MLA’s electoral survival depends on demonstrating visible local delivery. The MLA Local Area Development Fund — Rs 3 crore per year per MLA, Rs 441 crore total across 147 constituencies — is spent on community halls, drainage, streetlights, park development. These projects are politically valuable precisely because they are visible, attributable, and delivered within the MLA’s direct control. State-wide institutional reform — strengthening the judiciary, professionalising the bureaucracy, reforming land records — is invisible to voters in any individual constituency.

The voter’s dominant strategy: demand direct benefits and visible infrastructure. This is not irrational. In a system where institutions are unreliable, the rational citizen maximises returns through the channels that actually deliver: the MLA’s personal intervention, the welfare scheme’s direct transfer, the contractor’s willingness to build the road that serves the voter’s village. Institutional reform is a public good whose benefits are diffuse and uncertain. Direct benefits are a private good whose benefits are immediate and specific.

The depth of this rationality is visible in the electoral data. In 2019, Odisha voters demonstrated the most sophisticated form of strategic patronage-seeking in Indian electoral history: split voting. They voted BJD for the state assembly — protecting the state welfare architecture — and BJP for the Lok Sabha — accessing the central welfare architecture. BJD won 112 assembly seats; BJP won 12 of 21 Lok Sabha seats. The same electorate, on the same day, in the same polling booth, made two different calculations about which party would maximise benefits at which level. This is not the behaviour of uninformed voters. It is the behaviour of voters who understand the patronage equilibrium with precision and optimise within it. When simultaneous elections in 2024 collapsed the split-voting pattern by making the choice binary — one ticket, one party, two levels — the equilibrium shifted not because voters changed their preferences but because the electoral structure removed the option of dual optimisation.

The Equilibrium

Each player’s dominant strategy:

  • Politician: maintain weak institutions + distribute welfare
  • Bureaucrat: accommodate political preferences
  • Contractor: compete on political connection
  • Voter: demand direct benefits

No player can improve their outcome by unilaterally changing strategy. If the politician strengthens institutions while the voter continues to demand direct benefits, the politician loses elections. If the bureaucrat enforces institutional mandates while the politician continues to control transfers, the bureaucrat loses career stability. If the contractor competes on merit while the system continues to reward connection, the contractor loses contracts. If the voter demands institutional reform while the system continues to deliver through patronage channels, the voter loses access to the benefits that actually reach them.

The equilibrium is stable because each player’s strategy is a best response to every other player’s strategy. This is the textbook definition. It produces, collectively, a governance system that everyone inside it recognises as suboptimal — and that nobody inside it can individually improve.

The welfare spending confirms the stability. Rs 10,000-14,000 crore per year — covering KALIA/CM Kisan, BSKY/GJAY, PDS, Mission Shakti, Mamata, Subhadra, pensions — is funded by mining revenue that constitutes 84 percent of non-tax revenue. The fiscal loop closes: extraction generates revenue, revenue funds welfare, welfare generates electoral support, electoral support ensures extraction continues. The politician who breaks the loop — by diverting welfare spending to institutional reform, by imposing processing mandates that reduce near-term mining revenue, by strengthening institutions that constrain discretionary power — faces electoral punishment. The politician who maintains the loop faces electoral reward. The equilibrium selects for politicians who maintain it.

OSDMA as Equilibrium-Breaker

If the equilibrium is stable, how did OSDMA break it?

The answer is that it did not break the equilibrium. It operated in a domain where the equilibrium’s logic produced a different outcome — because the payoff matrix was different.

On October 29, 1999, approximately 10,000 people died. The political cost of institutional failure in disaster management became catastrophically, immediately, and visibly high. Critically, the cost was borne by the politician: the Giridhar Gamang government fell. Naveen Patnaik’s government was elected on the implicit promise that the next cyclone would produce a different result. The payoff matrix for disaster management was rewritten by 10,000 deaths.

In the new matrix, the politician’s dominant strategy shifted. Maintaining weak disaster management institutions now carried an unacceptable political cost: the next cyclone would kill thousands, and the government would be held responsible. The electoral penalty for failure exceeded the discretionary power premium from maintaining institutional weakness. For the first time, the incentive structure aligned political interest with institutional strength.

And critically, OSDMA’s success did not threaten any existing power centre. Mining companies did not lose from cyclone shelters. Contractors did not lose from evacuation protocols (they gained from shelter construction contracts). Bureaucrats did not lose authority — OSDMA actually expanded the bureaucracy’s operational capacity. Voters gained life-saving protection without losing welfare benefits. The equilibrium was not disrupted. A new institutional capacity was added in a domain where the equilibrium’s usual logic — weak institutions serve political interests — did not apply because the cost of weakness had become politically unbearable.

This is why OSDMA is the exception. Not because Odisha lacks institutional capacity — OSDMA proves the capacity exists. Not because Odisha lacks political will — OSDMA was built with sustained political investment over two decades. But because the conditions that made OSDMA possible — an existential crisis that rewrote the payoff matrix, a domain where institutional strength did not threaten political discretion, clear metrics (deaths) that could not be gamed, a timeline (annual cyclone season) that created recurring political accountability — have not been replicated for any other domain.

The proof of the equilibrium interpretation — rather than the simpler “incompetence” or “corruption” interpretation — is precisely that OSDMA works. If Odisha’s institutional weakness were caused by an incapable bureaucracy, OSDMA would not exist. If it were caused by a corrupt political class, OSDMA would have been hollowed out for patronage value — its shelter construction contracts diverted, its officer postings politicised, its protocols degraded. Instead, OSDMA survived 24 years of BJD rule and a change of government to BJP without institutional degradation. Cyclone Dana in October 2024 — the first major cyclone test under the new government — was handled with effectively zero deaths, 362,000 people evacuated, OSDMA protocols executed without disruption. The institution that the equilibrium has no incentive to weaken has strengthened continuously. Every other institution that the equilibrium has incentive to keep weak has remained weak. The selectivity is the evidence. The system is not incapable. It is selective in its capability, and the selection criterion is the payoff matrix of the patronage equilibrium.

For the equilibrium to break in education, the political cost of educational failure would need to become as visible, as immediate, and as politically devastating as 10,000 cyclone deaths. For it to break in agriculture, the cost of low productivity would need to trigger comparable political consequences. For it to break in industrial development, the cost of the missing manufacturing ecosystem would need to be as attributable and as electorally punishing as a failed cyclone response.

The education system exports 85-90 percent of NIT Rourkela graduates. This is a slow-motion institutional catastrophe. But it produces no bodies on television screens. The agricultural extension system operates at one officer per 1,500 farming families. This condemns millions to below-potential productivity. But the cost is distributed across decades and attributed to weather, markets, and fate rather than institutional failure. The missing industrial ecosystem means Odisha exports minerals and human capital while importing finished goods and skilled managers. But the absence of factories that were never built generates no political crisis.

Each of these failures is as costly as the 1999 cyclone in cumulative terms. None creates the political conditions for an OSDMA-like institutional response. The equilibrium prevents the crisis from materialising, because the welfare architecture provides a floor — declining poverty, expanding healthcare, converging per capita income — that makes the argument for structural change (“why risk what works?”) politically stronger than the argument for transformation (“it could work so much better”).

This is the stability trap described in the extraction equilibrium chapter of The Long Arc. The system works well enough to prevent the crisis that would force change. The patronage equilibrium is the mechanism by which this trap is maintained. Each player acts rationally. The collective outcome is an institutional landscape where one world-class institution — OSDMA — demonstrates what is possible, and everything else demonstrates what the equilibrium permits.

The Equilibrium at Ground Level

The four-player framework operates at the state level. But the equilibrium is maintained at every level of governance, all the way down to the village.

At the block level, the structural tension between the Block Development Officer — an appointed government official from the Odisha Administrative Service — and the elected Panchayat Samiti chairperson replicates the equilibrium in miniature. The BDO controls the block-level staff of multiple departments, manages MGNREGA implementation, processes beneficiary identification for welfare schemes, and has administrative authority over block-level expenditure. The elected chairperson has formal supervisory authority but no control over budget or staff. The BDO’s career depends on pleasing the state government and the local MLA, not on satisfying the elected representative. The equilibrium’s logic — political control over administrative machinery — plays out at the scale of 314 blocks.

At the panchayat level, Odisha’s 6,794 gram panchayats have formal authority over 29 subjects listed in the Eleventh Schedule of the Constitution. In practice, most functions are implemented through centrally or state-designed schemes with pre-defined guidelines, leaving panchayats with negligible discretionary authority. GP own revenue — from local taxes, fees, user charges — is typically Rs 50,000 to Rs 2,00,000 per year, making them almost entirely dependent on central and state government transfers. A gram panchayat that cannot raise its own revenue cannot set its own priorities. The fiscal dependency converts the lowest tier of democratic governance into the lowest-tier implementation agency for higher-level schemes.

Odisha’s 50 percent women’s reservation in panchayats — implemented in 2011, producing approximately 3,400 women sarpanches — is a structural reform that the equilibrium has partially absorbed. Academic studies estimate that in 40-60 percent of women-headed gram panchayats, the male family member (husband, father-in-law, or son) exercises de facto decision-making power — the “sarpanch-pati” phenomenon. The numerical representation is real. The transfer of institutional authority is incomplete. Mission Shakti’s SHG network has been documented as a pipeline for developing women’s governance confidence that partially counteracts the dynamic over time, but the timeline is generational, not electoral.

The equilibrium operates at every institutional scale. State level: Chief Minister controls transfers. District level: Collector accommodates MLA preferences. Block level: BDO accommodates both. Village level: panchayat implements schemes designed elsewhere. At each level, the pattern repeats: political control over administrative machinery, welfare delivery as substitute for institutional development, and the rational accommodation of patronage logic by every actor in the chain. This is not a single game with four players. It is the same game repeated at every scale, with the outcomes at each level reinforcing the others.

The Equilibrium’s Future

The 2024 transition tested whether a change of player could change the game. The evidence after eighteen months is that the game absorbed the new player. BJP operates the same welfare-extraction-election cycle with different branding and a second fiscal engine (central schemes). The transfer-posting regime is intact. DMF spending patterns are unchanged. The announcement economy continues. Institutional reform — Civil Services Board reform, fixed tenure legislation, outcome-based monitoring, genuine DMF governance — is absent from the policy agenda.

The game theory predicts this. A Nash equilibrium does not break because a new player enters. It breaks when the payoff matrix changes — when the costs and benefits of each strategy shift enough that the dominant strategies change.

Three scenarios could change the matrix. A commodity price collapse severe enough to crater mining revenue and force fiscal crisis — unlikely near-term but not impossible as green hydrogen and recycled steel challenge virgin ore demand over 15-20 years. A central policy intervention that restructures incentives — a Finance Commission formula tying devolution to institutional quality metrics, a MMDR revision rewarding processing over extraction, a mandatory minimum tenure law with judicial enforcement. Or a shift in what voters demand — a consciousness shift, documented in The Churning Fire, from patronage dependence to institutional expectation, from “give me the benefit” to “build the system that makes the benefit unnecessary.”

Each is possible. None is probable in the near term. The equilibrium’s most powerful feature is that it generates the conditions for its own persistence. The welfare floor prevents the crisis that would create political urgency. The discretionary power premium prevents the politician from surrendering control. The transfer regime prevents the bureaucrat from building institutional memory. The procurement system prevents the contractor from competing on merit. The voter’s rational demand for direct benefits prevents the political space for institutional reform.

The intra-state inequality data makes the equilibrium’s human cost concrete. As of 2019-21, Malkangiri’s poverty rate was 45.01 percent while Khordha (Bhubaneswar) was 3.95 percent — a 10:1 ratio within the same state, under the same government, after 24 years of continuous governance. The western and tribal districts — Kalahandi, Balangir, Nuapada, Koraput, Malkangiri, Nabarangpur, Rayagada, Kandhamal — have literacy rates between 40-65 percent, poverty rates between 30-45 percent MPI, and remain below the state average on virtually every development indicator. The coastal districts — Puri, Khurda, Cuttack, Ganjam — have literacy rates between 75-85 percent and poverty rates between 3-10 percent MPI. The equilibrium distributes welfare across this gap without closing it. The mining districts of the north — Keonjhar, Sundargarh, Angul — generate the revenue that funds the welfare, bear the environmental and social costs of extraction, and receive DMF funds that are spent at 50-55 percent of collection. The geography of the equilibrium is itself a map of its costs.

This is not a story about bad leaders or corrupt systems. Naveen Patnaik’s 24-year record includes genuine achievements — OSDMA, Mission Shakti, fiscal discipline, dramatic poverty reduction. Mohan Majhi’s early months show continuity of welfare delivery in a state that depends on it. The system works. Within its logic, it works remarkably well. The argument of this chapter is not that the system is corrupt, though corruption exists within it. The argument is that the system is stable, that its stability serves the interests of every player inside it, and that stability is the enemy of the institutional transformation that would allow Odisha to convert its extraordinary natural and human resources into an economy and a society that functions at its actual potential.

The 10,000 who died in 1999 broke the equilibrium for disaster management. The cost was measured in bodies. The question for every other domain is simpler and more uncomfortable: what would the equivalent of 10,000 deaths look like for education, for agriculture, for industrial development? What crisis would be severe enough to rewrite the payoff matrix? And is there a way to change the matrix without the crisis — to create the institutional conditions for transformation through design rather than disaster?

That is the question this series turns to next.

Sources

Cross-references to SeeUtkal series

Government and institutional sources

  1. Election Commission of India — constituency-wise results 2000-2024; BJD seat counts and vote shares
  2. CAG Audit Report on KALIA — Rs 782.26 crore to 12.72 lakh ineligible beneficiaries
  3. Department of Agriculture, Government of Odisha — KALIA beneficiary data, scheme design
  4. Department of Health and Family Welfare — BSKY/GJAY coverage data (96.5 lakh families)
  5. Odisha Economic Survey 2024-25 — mining as 84% of non-tax revenue
  6. DMF Odisha portal (dmf.odisha.gov.in) — Rs 31,324 crore cumulative collection, spending patterns
  7. NITI Aayog — Fiscal Health Index 2025 (Odisha ranked first), MPI 2023
  8. Government of Odisha transfer notification records, June-December 2024
  9. PRS Legislative Research — Odisha Budget Analysis 2024-25, 2025-26; scheme tracking
  10. OSDMA — cyclone response data, institutional history
  11. Shah Commission Report, 2012 — Rs 59,000 crore illegal mining, 63 cases in Keonjhar
  12. OMC Annual Report 2023-24 — Rs 9,076 crore net profit
  13. Second Administrative Reforms Commission, 7th Report — political interference in administration

Academic and research sources

  1. Nash, John. “Non-Cooperative Games.” Annals of Mathematics 54, no. 2 (1951): 286-295
  2. Acemoglu, Daron & James Robinson. Why Nations Fail. Crown, 2012 — extractive vs. inclusive institutions framework
  3. Olson, Mancur. The Logic of Collective Action. Harvard University Press, 1965 — public goods and concentrated costs
  4. Chandra, Kanchan. Why Ethnic Parties Succeed. Cambridge University Press, 2004 — voter behaviour in patronage systems
  5. Khemani, Stuti. “Political Economy Constraints on Policies.” World Bank — structural biases toward visible spending
  6. CSDS/Lokniti — National Election Study data 2014, 2019, 2024; welfare scheme beneficiary voting patterns (18-22 percentage point gap)
  7. NIPFP Study on IAS Tenure, 2012 — median tenure approximately 16 months
  8. Das, Rajat Kanti. Naveen Patnaik. Penguin Viking, 2019
  9. Banerjee, Ruben. Naveen Patnaik. Juggernaut Books
  10. Chakrabarti et al. — Mamata evaluation, Journal of Nutrition (2022)
  11. Prakash Singh v. Union of India (2006) 8 SCC 1 — police reform directives
  12. TSR Subramanian v. Union of India (2013) 14 SCC 572 — Civil Services Board directives

Journalism and reporting

  1. ThePrint, “Odisha is becoming an IAS state” (2023) — CMO-centric governance, Pandian’s role
  2. The Wire, “VK Pandian Takes Early Retirement” (2023) — 5T, cabinet rank appointment
  3. Indian Express — “Odisha bureaucratic reshuffle after new government,” July 2024; “5T: Naveen Patnaik’s governance experiment”
  4. Scroll.in — “How Odisha’s bureaucracy functioned under Naveen Patnaik,” 2024; Kandhamal violence retrospective
  5. OdishaTV — transfer tracking; “BJP government discontinues Mo Sarkar”; scheme reporting
  6. Indian News Diary — “5T Initiatives by Government of Odisha: Governance or Over-Centralisation?”
  7. Down to Earth — “M.B. Shah Commission: Odisha’s mine of scams exposed”; construction quality investigations
  8. Transparency International India — India Corruption Survey, contractor-politician nexus documentation

Source Research

The raw research that informs this series.