English only · Odia translation in progress

Chapter 4: The Paper and the Forest

On January 26, 2020 — Republic Day — five gram sabhas in Sundargarh district were convened to decide whether their land should be handed to Dalmia Cement Bharat Limited for a mining expansion. The sub-collector had issued letters to the sarpanches of Kukuda and Katang gram panchayats, directing them to hold the assemblies on that particular date. Republic Day. The day the Constitution is read aloud in school assemblies across the country. The day the document that promises to “secure to all its citizens” justice, liberty, equality, and fraternity is ritually celebrated.

All five gram sabhas voted unanimously against the land acquisition. The Kukuda Gram Panchayat resolution was categorical: “all the villagers present in the meeting unanimously resolved not to give an inch of land to OCL.” Over four hundred people signed. The rejection was documented, unambiguous, constitutionally protected under both the Right to Fair Compensation and Transparency in Land Acquisition Act of 2013 and the Panchayats (Extension to Scheduled Areas) Act of 1996. Eighteen months later, in July and December 2021, the Odisha authorities issued preliminary notifications to acquire the land anyway.

This chapter is about the distance between the paper and the forest. Not the distance in kilometres between Delhi and a gram sabha in Sundargarh, though that is real enough. The distance in function. India has built one of the most elaborate constitutional and legislative architectures for tribal protection anywhere in the world — the Fifth Schedule, PESA, the Forest Rights Act, the Samatha judgment, the Tribal Sub-Plan, dedicated development agencies for thirteen Particularly Vulnerable Tribal Groups. On paper, a tribal person in Odisha’s Scheduled Areas is protected by at least four layers of constitutional provisions, five major legislative enactments, eight to ten institutional bodies, and multiple funding schemes.

And yet: PVTG literacy ranges from 4.98 percent to 41.75 percent against a state average of 73 percent. Of 8.4 lakh acres of alienated tribal land, only 6.75 percent has been restored. PESA rules remain unfinished after twenty-nine years. Forty-three percent of land acquisitions in Scheduled Areas proceed without gram sabha consent. The conviction rate for atrocities against Scheduled Tribes from 2019 to 2021 was zero percent.

The cross-domain lens for understanding this gap is mechanism design — the branch of game theory that asks: given that participants will act in their own interest, can you design rules that produce the socially desired outcome? India’s tribal protection framework is a textbook mechanism design failure. The rules exist. The enforcement mechanism was never built.

The Constitutional Promise: Article 244 and the Fifth Schedule

The Fifth Schedule of the Indian Constitution, operating through Article 244(1), established a distinct governance framework for areas with significant tribal populations. It applies to ten states, including Odisha. The core idea was simple and powerful: these areas are different, their populations are vulnerable, and ordinary law may not be sufficient to protect them. The Constitution therefore granted extraordinary powers to the Governor of each state with Scheduled Areas — powers that, on paper, amount to a parallel governance architecture.

The Governor can direct that any Act of Parliament or state legislature shall not apply to a Scheduled Area, or shall apply with modifications. The Governor can make regulations prohibiting or restricting the transfer of tribal land, regulating money-lending to tribal communities, and controlling the allotment of land. The Governor must annually report to the President on the administration of Scheduled Areas. And every state with Scheduled Areas must establish a Tribes Advisory Council to advise on matters pertaining to tribal welfare.

In Odisha, this architecture covers enormous ground. Seven districts are fully designated as Scheduled Areas — Mayurbhanj, Sundargarh, Koraput, Malkangiri, Nabarangpur, Rayagada, and one more depending on which reorganisation count you use. Six districts are partially designated: Sambalpur (Kuchinda tahasil), Keonjhar (the mining tahasils — Keonjhar, Telkoi, Champua, Barbil), Kandhamal, Ganjam, Kalahandi, and Balasore. Together, these Scheduled Areas cover approximately 44.71 percent of the state’s total land area. Nearly half the state.

Sixty-two tribal communities live in these areas — the highest number of any Indian state. Thirteen of India’s seventy-five Particularly Vulnerable Tribal Groups are concentrated here — again, the highest of any state. The tribal population is 95.9 lakh, or 22.85 percent of the state. These are not marginal communities tucked into a forgotten corner. Nearly one in four people in Odisha is tribal. Nearly half the state’s geography is constitutionally designated for special protection.

Now consider the Tribes Advisory Council — the institution the Fifth Schedule created to be the bridge between constitutional intent and administrative action. It consists of twenty members including representatives of Scheduled Tribes in the state legislature, a chairman, and a deputy chairman. It meets twice a year. Its advice is non-binding. There is no follow-up mechanism to ensure its recommendations are implemented. An empirical study by Mohanty and Acharya found that the TAC’s role remains “largely ceremonial.” The body that is supposed to advise the Governor on tribal welfare — the Governor who holds extraordinary constitutional powers to protect tribal interests — meets twice a year, offers non-binding suggestions, and watches as those suggestions are filed and forgotten.

In mechanism design terms, this is a system with a critical feedback loop that was designed without a feedback channel. The TAC is supposed to signal when tribal interests are being harmed. The Governor is supposed to act on those signals. But the TAC cannot compel the Governor, the Governor rarely exercises protective powers, and the annual reports to the President are routine administrative exercises rather than instruments of advocacy. The signal exists. The actuator does not. A thermostat that can read the temperature but cannot turn on the heating is not a thermostat. It is a thermometer with delusions.

PESA: The Law That Was Never Turned On

The Panchayats (Extension to Scheduled Areas) Act was passed by Parliament on December 24, 1996. It was designed to be revolutionary, and the word is not too strong. The 73rd Constitutional Amendment of 1992 had established panchayati raj — local self-governance — across India, but explicitly excluded Scheduled Areas, recognising that tribal areas needed something different. The Bhuria Committee recommended what that different thing should look like: in tribal areas, the gram sabha — the village assembly of all adult residents, not the elected panchayat body — should have sovereign authority over land, forests, minor minerals, and natural resources.

PESA codified much of this vision. It mandated that gram sabhas in Scheduled Areas would have prior consent authority over land acquisition. Mandatory consultation before mining leases for minor minerals. Ownership of minor forest produce. Regulation of money lending. Control over development plans and tribal sub-plans. Prevention and restoration of unlawfully alienated tribal land. The Act required each of the ten states with Fifth Schedule areas to frame state-level rules within one year. One year. That was Parliament’s timeline.

Odisha has not finalised its PESA rules. It is now 2026. Twenty-nine years and counting.

This is not garden-variety bureaucratic delay. Consider the comparison. Maharashtra published its rules in 2014. Gujarat in 2017. Chhattisgarh in 2022 — twenty-six years late, but done. Jharkhand, long the other holdout, finally notified its rules in January 2026 after public consultations, workshops across the state, and Cabinet approval. Odisha remains the only state among ten that has never finalised PESA rules. In November 2023 — twenty-seven years after the Act — the government published a draft notification in the Odisha Gazette, inviting objections. The draft was discussed in a Tribal Advisory Council meeting. The conclusion: “further discussions are needed.” A workshop was proposed. The rules remain in draft.

But the draft itself is revealing. A genuine implementation of PESA would give gram sabhas veto power over land acquisition and mining. The Odisha draft does not do this. It requires the land acquisition officer to “consult” the gram panchayat — not the gram sabha — on proposed acquisitions. The distinction matters enormously. The gram panchayat is an elected body with a limited number of members. The gram sabha is the general assembly of all adult village residents. PESA was designed to vest power in the assembly, not the representatives. The draft substitutes the assembly with the representatives and calls it compliance. Worse, the draft is entirely silent on the gram sabha’s power to refuse land acquisition. And for mining on forestland, gram sabhas are explicitly excluded from authority over minor minerals.

This is not compliance. It is the dilution of PESA through the rules that were supposed to implement it. The law’s teeth were written into the statute by Parliament. The state’s draft rules quietly extracted them.

The Comptroller and Auditor General’s 2024 Performance Audit of Land Management in Scheduled Areas documented what this non-implementation looks like in practice. In 126 out of 294 land acquisition cases examined — 43 percent — preliminary notifications for land acquisition were issued without conducting gram sabha meetings or obtaining requisite consent. In 44 of 254 cases — 17 percent — Social Impact Assessment studies, mandatory under the 2013 Land Acquisition Act, had not been carried out at all.

A note on numbers: some sources, including earlier analysis in this research project, reference “136 violations” from the CAG findings. The CAG report itself documents 126 cases of land acquisition without proper gram sabha consent (43 percent of 294 cases examined). The “136” figure likely refers to the Rs 136.77 crore in District Mineral Foundation funds diverted for construction of an international hockey stadium in Rourkela — funds meant for the welfare of mining-affected communities, many of them tribal. The distinction matters for precision. The core finding is the same for the argument: gram sabha consent, the foundation of PESA, is routinely bypassed. Whether you count by cases (126) or by crores diverted (136.77), the system operates as if the law does not exist.

Why? The reason is not mysterious. Odisha produces approximately 50 percent of India’s iron ore, 98 percent of its chromite, 73 percent of its bauxite, and significant shares of manganese and coal. Mining revenue constitutes roughly 84 percent of the state’s non-tax revenue and funds 25 to 30 percent of the state budget. Much of this mining occurs in or adjacent to Scheduled Areas. Giving gram sabhas effective veto power over mining decisions would mean giving the poorest communities in the state the ability to block the largest source of state revenue. No government — BJD for twenty-four years, BJP since 2024 — has been willing to accept that trade-off.

In mechanism design, this is called an incentive compatibility problem. The mechanism designer (Parliament) wanted one outcome: tribal self-governance over resources. But the agents responsible for implementation (state governments) have powerful incentives to produce a different outcome: continued extraction. Parliament designed the rules but failed to design enforcement. There is no penalty for not implementing PESA. No central government has ever withheld mining clearances from Odisha on the grounds that PESA was unimplemented. No ministry has made PESA compliance a condition for environmental or forest clearance. The specification was written. The enforcement logic was never deployed.

The Forest Rights Act: When 72 Percent Is a Lie

The Scheduled Tribes and Other Traditional Forest Dwellers (Recognition of Forest Rights) Act, 2006, was enacted to correct what its preamble called a “historical injustice” to forest-dwelling communities. It promised three categories of rights: Individual Forest Rights for habitation and self-cultivation, Community Forest Rights over shared forest resources, and Habitat Rights for Particularly Vulnerable Tribal Groups. The Act represented, in principle, a fundamental reordering of the relationship between forest communities and the state — from subjects to be managed to citizens with rights to be recognised.

Odisha’s headline implementation number looks respectable. Of 6,21,195 Individual Forest Rights claims submitted by Scheduled Tribes, approximately 4,37,718 have been recognised — an acceptance rate of about 72 percent. The Ministry of Tribal Affairs dashboard presents this as progress. It is not.

The first problem is what “recognised” means. In many cases, what was counted as an approved claim was granted for only a fraction of the land claimed. A family that filed a claim for two acres of forest land they had been cultivating for generations might receive title for half an acre — technically an “approval” but practically a rejection of three-quarters of their claim. The system counts partial approvals as full approvals, inflating the acceptance rate.

The second problem is what happens to rejected claims. When Odisha reviewed 148,870 rejected IFR claims at the District Level Committee stage, it re-rejected 140,504 — a re-rejection rate of 94 percent. The review process is not a genuine appeal mechanism. It is a rubber stamp on prior decisions. The reasons for rejection are revealing: about 42 percent of claims are rejected for “lack of evidence,” but 87.5 percent of Forest Rights Committee members in the CAG sample reported that forest and revenue maps — the very evidence needed to support claims — were never provided to them by authorities. You reject the claim for insufficient evidence. You also refuse to provide the evidence. The circularity is not accidental.

The third problem is scale. Sundargarh district rejected 6,909 FRA claims without recording any reasons at all. Not inadequate reasons. Not formulaic reasons. No reasons. In Malkangiri, 1,584 claims were rejected without recording individual details for any of them. Deogarh constituted Forest Rights Committees in 679 of its 774 villages, but not a single village achieved full coverage under FRA. Zero percent. This is not administrative oversight. It is institutional indifference with the texture of contempt.

The fourth problem is who gets left out entirely. The FRA was designed to cover not just Scheduled Tribes but also Other Traditional Forest Dwellers — non-tribal communities that have lived in and depended on forests for generations. Of 31,690 OTFD claims submitted in Odisha, exactly 73 were recognised. A rejection rate of 99.77 percent. For every ten thousand OTFDs who filed claims, twenty-three received rights. Functionally, FRA does not exist for non-tribal forest dwellers in Odisha.

And the fifth problem: community forest rights — arguably the most transformative provision of FRA, the one that would give gram sabhas collective authority over forest resources — remain at approximately 5 percent of potential in Odisha. Compare that with Maharashtra at roughly 15 percent and Kerala at about 14 percent. Maharashtra’s Gadchiroli district demonstrates what is possible when institutional will exists: Mendha-Lekha became the first village in India to receive community forest rights over 1,809 hectares, with the gram sabha obtaining a PAN number, maintaining a bank account, and functioning as a self-governing body. In Odisha, with the third-largest tribal population and extensive forest cover, that model remains an academic curiosity.

There is one genuine bright spot. Odisha has emerged as a national leader in PVTG habitat rights recognition. Six PVTGs have received habitat rights as of 2024 — Paudi Bhuyan (the first, in Deogarh), Juang, Saora, Chuktia Bhunjia, Hill Kharia, and Mankidia. Habitat rights grant communities territorial authority over traditional habitats, socio-cultural practices, and ecological knowledge. The Mankidia community in Mayurbhanj — traditional rope-makers who depend on siali fibre from forests — became the sixth PVTG to receive these rights. Seven PVTGs still await recognition, but the progress is real. It also raises a question: if the state can recognise habitat rights for PVTGs, why can it not process FRA claims for ordinary tribal communities? The answer, almost certainly, is that PVTG habitat rights cover relatively small populations in remote areas and do not threaten extraction economics. The moment forest rights recognition reaches scale, it encounters the same incentive incompatibility that killed PESA.

Recently, Odisha shut down fifty Sub-Divisional FRA Cells even as claims remained pending — effectively dismantling the institutional infrastructure for processing the very rights the law was supposed to recognise. The forest bureaucracy views FRA as a threat to its territorial control. Every hectare where community forest rights are recognised is a hectare removed from Forest Department jurisdiction. The department that is supposed to facilitate rights recognition has an institutional interest in preventing it. A mechanism designer would recognise this immediately: you have assigned the fox to guard the henhouse and then expressed surprise at the declining chicken population.

Samatha: The Judgment That Nobody Obeys

In 1997 — the same year PESA came into force — the Supreme Court delivered a landmark judgment in Samatha v. State of Andhra Pradesh. The ruling was direct: government, tribal, and forested lands in Scheduled Areas cannot be leased to non-tribal persons or private companies for mining purposes. All mining leases transferred to non-tribals in Scheduled Areas were declared void. Private companies are “persons” under law. Twenty percent of net mining profits should be set aside as a permanent fund for Scheduled Area development.

The judgment was clear. Its application in Odisha has been a masterclass in creative circumvention.

The primary method is the state PSU proxy. The Supreme Court excluded state-owned agencies from the prohibition, creating a loophole large enough to drive a mining truck through. State Public Sector Undertakings like the Odisha Mining Corporation obtain the lease, then effectively operate in partnership with or as a conduit for private companies. In the Niyamgiri case, OMC was the formal lease applicant while Vedanta — a London-listed private company — was the effective operator of the alumina refinery. The letter of the judgment is preserved. The spirit is eviscerated.

The second method is the “public purpose” doctrine. Land is acquired under eminent domain for “public purpose” — industrial development, national interest — bypassing the prohibition on direct transfer to non-tribals. In Kalinganagar, the Industrial Infrastructure Development Corporation of Odisha acquired land from tribals at Rs 15,000 to 30,000 per acre, then transferred it to Tata Steel at Rs 3.5 lakh per acre. A markup of ten to twenty-three times. The value extracted from tribal communities and transferred to the state and corporate interests was not a transaction fee. It was the spread between what tribal land is worth to its owners and what it is worth to the extraction economy.

The third method is simply denying applicability. The Naveen Patnaik government took the position that the Samatha judgment was not binding on Odisha since the case arose from Andhra Pradesh. This interpretation — that a Supreme Court judgment on constitutional provisions applying equally to all Scheduled Areas in all states was somehow state-specific — was legally questionable but politically convenient. It served for years.

The Sijimali bauxite case is the most recent and perhaps most instructive example. Unable to access the Niyamgiri deposit after the Dongria Kondh gram sabhas voted unanimously against mining in 2013, Vedanta won the auction for the Sijimali bauxite block in February 2023 — a 1,548.79-hectare lease across Kalahandi and Rayagada districts, approximately twice the area of the original Niyamgiri project. Local tribal communities allege that gram sabha resolutions consenting to forest diversion were fraudulently obtained. Between August and September 2024, affected villages held fresh gram sabhas and unanimously rejected the mining lease. In July 2025, the Union government temporarily halted forest clearance after evidence surfaced that consent may have been forged. Activists opposing the project have been charged under the Unlawful Activities Prevention Act.

The Sijimali sequence is worth sitting with. Niyamgiri established, through years of legal battle culminating in a Supreme Court order, that gram sabha consent is mandatory and substantive — not a procedural formality. The system’s response was not to accept this principle. It was to shift operations to a neighbouring bauxite deposit and allegedly manufacture the consent that the law requires. When the manufactured consent was challenged, the response was not to withdraw. It was to charge the challengers under anti-terror legislation. The mechanism designer writes a rule. The players find a workaround. The designer patches the workaround. The players find a workaround for the patch. At no point does anyone redesign the enforcement mechanism itself.

If the Samatha judgment were strictly enforced in Odisha, all mining leases to private companies in Scheduled Areas would be void. Only state PSUs or tribal cooperatives could mine. The twenty percent net profit allocation would redirect billions. The existing model of cheap acquisition and corporate transfer would collapse. The economic cost of compliance is calculated to exceed the political cost of non-compliance. The calculation holds because tribal communities lack the organised political power to make non-compliance electorally expensive. The mechanism produces the correct outcome only when external pressure — Supreme Court orders, international NGO campaigns, sustained media coverage — forces it to. Niyamgiri was not the system working. It was the system being forced to work once, at enormous cost, in one location, for one community.

Thirteen Vulnerable Peoples

Odisha has thirteen of India’s seventy-five Particularly Vulnerable Tribal Groups — the highest concentration in any state. Total PVTG population: approximately 7.73 lakh. These are communities the Indian government itself has designated as the most vulnerable of the vulnerable — characterised by pre-agricultural technology, stagnant or declining populations, extremely low literacy, and subsistence-level economies. Here they are:

Birhor (596 people, Sundargarh and Sambalpur, nomadic rope-makers — the smallest PVTG in the state). Bonda (12,231, Malkangiri, hill-dwelling podu cultivators). Chuktia Bhunjia (3,086, Nuapada). Didayi (8,890, Malkangiri, 6.87 percent literacy). Dongria Kondh (9,659, Rayagada and Kalahandi, the Niyamgiri defenders, 6.35 percent literacy). Juang (47,095, Keonjhar and Jajpur). Hill Kharia (2,800, Mayurbhanj). Kutia Kondh (39,761, Kandhamal and Kalahandi, 4.98 percent literacy — the lowest of any PVTG in the state, and one of the lowest of any community in India). Lanjia Saora (40,913, Gajapati and Rayagada, known for distinctive wall paintings). Lodha (9,785, Mayurbhanj, formerly criminalised tribe). Mankidia (2,222, Mayurbhanj, siali fibre rope-makers). Paudi Bhuyan (61,303, Deogarh and other districts). Saora (534,751, the largest PVTG population, spread across Gajapati, Ganjam, and Rayagada).

The development indicators are not statistics you process intellectually. They are statistics you feel in your stomach.

Literacy: Kutia Kondh at 4.98 percent. Dongria Kondh at 6.35 percent. Didayi at 6.87 percent. Lodha at 27.20 percent. Juang at 30 percent. The state average is 72.87 percent. The weighted average across all PVTGs is 37.4 percent — half the state figure.

Health: Malaria prevalence of 25 percent among Kutia Kondh, 13.8 percent among Dongria Kondh, 11 percent among Paudi Bhuyan, 7.8 percent among Bonda. Undernutrition rates of 79 percent among Upper Bonda, 43 to 79 percent across groups depending on accessibility. Anaemia among women at 70 to 92 percent in some communities. Hepatitis B prevalence as high as 14.18 percent in Bonda areas. Sickle cell disease at 2 to 3.8 percent across several groups. Alcohol use at 58 to 67 percent, tobacco use at 57 to 72 percent.

And the case that should haunt anyone who talks about “tribal development” as if it were a real programme: the Bonda Development Agency.

The BDA was established in 1976-77, headquartered at Mudulipada in Malkangiri district. It was one of the first PVTG-specific development agencies in India, created with the explicit mission of improving the lives of the Bonda community. It has operated continuously for nearly fifty years.

The Bonda’s literacy rate is 33 percent. After half a century of dedicated institutional intervention, literacy remains at less than half the state average. The community’s undernutrition rate reaches 79 percent in Upper Bonda areas. In the village of Barguda, there is no school, no anganwadi centre. Children must trek two kilometres over steep hills to access education. The village that a development agency was specifically created to serve does not have a school.

The Bonda have been “targets” of development schemes for fifty years. They live in a Scheduled Area. They are designated as a PVTG. They have a dedicated development agency. They are covered by the Fifth Schedule, PESA (notionally), FRA, and multiple welfare schemes. The institutional architecture surrounding them is dense and well-funded. And a child in Barguda cannot walk to school because there is no school to walk to.

The mechanism design question is not “why don’t the schemes work?” The schemes deliver specific inputs — some healthcare, some food distribution, some infrastructure. The question is whether the mechanism was designed to produce the outcome (development) or to produce the appearance of the outcome (institutional activity). The BDA has operated for fifty years. It has had budgets, staff, offices, programmes. It has generated activity. It has not generated development. In mechanism design, when a mechanism consistently produces an outcome different from its stated objective, the analyst asks: is the stated objective the real objective? Or is the real objective the one the mechanism actually produces?

The Tribal Sub-Plan: Where the Money Goes

The Tribal Sub-Plan strategy, introduced in the Fifth Five Year Plan in 1974, mandated a principle so simple it should have been impossible to violate: if tribals are 22.85 percent of the population, they should receive at least 22.85 percent of development spending. Every state department should allocate proportionate funds to tribal development based on population share.

Odisha has consistently fallen below this mandate. Even within allocated amounts, expenditure has historically run at around 90 percent — meaning 10 percent of already-insufficient funds go unspent. But the real scandal is not how much is spent. It is what it is spent on.

The CAG and other oversight bodies have documented systematic diversion of TSP funds to non-tribal purposes. Six crore rupees of TSP funds diverted for a flyover at Rajmahal Square, Bhubaneswar — a general infrastructure project with no specific tribal benefit. Five crore for improvement of the Daya Canal on the city outskirts. TSP funds allocated for modernisation of police barracks. Residential buildings for police. Police training. Rs 21.43 crore to the Home Department in 2010 and Rs 16.92 crore in 2011. Rs 10.79 crore to the Industry Department in 2010 and Rs 17.97 crore in 2011 — departments with, in the CAG’s words, “zero or least impact on tribal communities.”

The CAG described these diversions as “highly irregular and completely inadmissible.” The diversions continued.

The mechanisms are straightforward. Departments with no tribal-relevant programmes receive TSP allocations to meet notional targets. Mainstream development schemes are relabelled as tribal schemes when they happen to operate in areas with tribal populations, regardless of whether tribals are the primary beneficiaries. Roads, bridges, and government buildings in tribal areas are counted as TSP spending even when tribals are not the primary users. A road built through a tribal area primarily to transport mined ore counts as tribal development spending. The state meets its spending target. The tribal community watches the trucks pass.

The District Mineral Foundation, created in 2015 to channel mining revenue back to mining-affected communities, follows the same pattern. DMF trusts in Odisha’s mining districts received substantial funds. The CAG found that Rs 136.77 crore of DMF funds in Sundargarh were diverted for construction of an international hockey stadium in Rourkela, and projects were executed without Board of Trustees approval. When a new funding stream is created specifically for tribal communities, it follows the same diversion pattern as every previous funding stream. The problem is not the design of individual schemes. It is the absence of tribal power over resource allocation. Nobody asks the tribal community what it needs. Nobody is accountable to the tribal community for what is spent. The money is for them in name. It is spent by others in practice.

The structural problem is threefold. First, the definition of “tribal benefit” is elastic enough to accommodate virtually any expenditure in a tribal-majority area. Second, TSP tracks inputs (allocations, expenditure) but not outcomes (did literacy improve? did malnutrition decrease?). A state can spend 100 percent of its TSP allocation and see tribal indicators worsen. Third, the tribal community has no say in prioritisation. TSP allocation decisions are made by state-level departments. There is no mechanism for gram sabhas to determine how “their” funds are spent. This directly contradicts PESA’s spirit, which vests development planning authority in the gram sabha — the same PESA whose rules Odisha has never finalised.

The Land That Does Not Come Back

The Odisha Scheduled Areas Transfer of Immovable Property (by Scheduled Tribes) Regulation, 1956, is the primary legal instrument protecting tribal land. Its core provision is unambiguous: transfer of immovable property by members of Scheduled Tribes to non-tribals is prohibited in Scheduled Areas.

By December 1999 — the most recent comprehensive survey, which itself tells you something about the state’s interest in measuring this problem — approximately 8,41,916 acres of tribal land had been alienated in Odisha. Of this, 56,854 acres had been restored through legal process. That is 6.75 percent. For every hundred acres of tribal land lost, roughly seven have been legally returned. Even among cases that went to court, only about 59 percent were decided in favour of tribals. Koraput district alone had 28,901 acres alienated — the highest of any district.

These figures are from 1999. No updated comprehensive survey has been conducted in twenty-six years. You cannot address a problem you refuse to measure. Given the acceleration of mining and industrial projects through the 2000s and 2010s, the actual scale of alienation today is almost certainly larger. But the state does not count, so the state does not know, and so the state does not act.

The legal prohibition on land transfer has not prevented alienation. It has redirected it through channels the law was not designed to catch. Benami transactions: non-tribals acquire tribal land through proxy holders, often other tribals paid to act as fronts. The Supreme Court in Samatha itself noted that “non-tribals were able to find ways and means to circumvent the provisions… by entering into benami transactions and other clandestine transactions with unsophisticated tribals.” Marriage strategies: when a tribal woman marries a non-tribal man, land effectively transfers to a non-tribal household, technically legal but achieving exactly what the law sought to prevent. Debt traps: tribals take loans from informal sources and effectively lose control of their land when unable to repay, even when formal ownership transfer is prohibited. The land stays in the tribal person’s name. The tribal person no longer controls it.

But the largest source of tribal land loss is not individual cunning. It is the state itself. Government acquisition for “public purpose” — industrial projects, mining, infrastructure — operates at scale. The Kalinganagar example crystallises the mechanics: the Industrial Infrastructure Development Corporation of Odisha acquired tribal land at Rs 15,000 to 30,000 per acre, then sold it to Tata Steel at Rs 3.5 lakh per acre. On January 2, 2006, when tribal communities protested the construction of a boundary wall, police opened fire. Twelve to fourteen tribal persons were killed. An inquiry commission declared the firing “justified.” No officer was prosecuted. Eighteen years later, displaced families are still fighting for land titles.

In November 2023, the then-BJD government proposed amending the 1956 Regulation to allow tribals to sell land to non-tribals under specified conditions. The stated justification was “freedom to dispose of their assets” and “unlocking economic value.” Given existing power asymmetries — literacy gaps, economic disadvantage, information deficits — “freedom to sell” would translate into pressure to sell. Historical experience across India demonstrates that wherever land transfer restrictions have been weakened, tribal land alienation has accelerated. The amendment was put on hold following widespread tribal protests. The fact that it was proposed at all reveals the political economy of tribal land: the same government responsible for preventing alienation periodically attempts to facilitate it.

The 1956 Regulation, the Fifth Schedule, and PESA are all designed to prevent private actors from taking tribal land. None of them effectively restrains the state itself. The institution charged with tribal protection is the primary agent of tribal dispossession. This is not an oversight. It is the central design failure.

The Mechanism That Was Never Built

In game theory, mechanism design is sometimes called “reverse game theory.” Instead of analysing how players will behave under given rules, the mechanism designer asks: what rules will produce the outcome I want, given that players will act in their own interest?

India’s tribal protection framework is a mechanism design problem. The desired outcome is clear: tribal communities in Scheduled Areas should retain control over their land, forests, and resources, with development that serves their interests. The players are equally clear: state governments that depend on mining revenue, a forest bureaucracy that controls territory, mining companies that seek resources, and tribal communities that lack political and economic power. Each player will act in their own interest. The question is whether the rules — the Fifth Schedule, PESA, FRA, Samatha, TSP, the land transfer regulations — can produce the desired outcome given these incentives.

They cannot. And the reason they cannot is not that the rules are wrong. The rules are largely correct. The reason is that the enforcement mechanism was never built.

Consider what enforcement would require. PESA mandates gram sabha consent for land acquisition. Enforcement would require: (a) a penalty for proceeding without consent, (b) a monitoring mechanism to detect non-compliance, (c) an institution with power and incentive to impose the penalty, and (d) accessibility for affected communities to trigger the enforcement process. None of these exist. There is no penalty for bypassing gram sabha consent. The CAG detected violations only through periodic audits, not continuous monitoring. No institution has both the power and the incentive to punish non-compliance — the state government benefits from non-compliance, the NCST can only advise, and the courts require years of litigation. And tribal communities with 5 to 33 percent literacy do not have ready access to legal processes.

The same analysis applies to every element of the framework. FRA claims are rejected without reasons? No penalty. TSP funds are diverted to flyovers? No penalty. The Samatha judgment is circumvented through PSU proxies? No consequence. Land is alienated through benami transactions? The restoration rate is 6.75 percent. Development agencies operate for fifty years without improving outcomes? No accountability review, no institutional redesign, no course correction.

A well-designed mechanism has four properties. It is incentive-compatible: players find it in their interest to behave in ways that produce the desired outcome. It has individual rationality: players are better off participating than opting out. It is implementable: the rules can actually be enforced given real-world constraints. And it reveals information: the mechanism causes players to reveal their true preferences or types.

India’s tribal protection framework fails on all four counts.

Incentive compatibility: State governments have powerful incentives to not implement PESA and not process FRA claims, because doing so would constrain their revenue model. The mechanism requires state governments to act against their own financial interests, without any countervailing incentive.

Individual rationality: The forest bureaucracy loses territorial control when FRA is implemented. The revenue department loses authority when gram sabhas gain power. Every implementing institution loses something by genuine implementation. The mechanism asks turkeys to vote for Christmas.

Implementability: The rules require enforcement by the very institutions that benefit from non-enforcement. The state government is supposed to implement PESA. The state government depends on mining revenue from PESA areas. This is like asking the defendant to be their own judge.

Information revelation: The mechanism generates no information about its own performance. No updated land alienation survey since 1999. FRA claims rejected without recorded reasons. TSP spending tracked by category but not by outcome. Governor’s annual reports that are routine rather than investigative. The mechanism is designed to be opaque about its own failures.

This failure is not accidental. A mechanism that consistently produces outcomes different from its stated objectives, for decades, across multiple governments, is producing the outcomes it was designed to produce. The stated objective is tribal protection. The revealed objective — the one the mechanism actually achieves — is the appearance of tribal protection while permitting extraction. The constitutional architecture exists to demonstrate that India cares about its tribal citizens. It does not exist to ensure that tribal citizens have power.

What would a properly designed mechanism look like? I can sketch the principles, with the caveat that institutional redesign is easier to describe than to implement (confidence level: ~65 percent that any of these would work as theorised).

First, incentive alignment: make PESA compliance a prerequisite for mining and environmental clearances. No PESA rules, no clearances. This converts non-implementation from a cost-free strategy to a revenue-destroying one. Second, independent monitoring: a permanent, independently staffed monitoring body — not within the state government — that tracks gram sabha consent, FRA processing, TSP spending, and land alienation in real time. Third, penalties with teeth: automatic financial penalties on state governments that fail compliance benchmarks, deducted from central transfers. Fourth, information transparency: mandatory public dashboards showing every FRA claim filed, every gram sabha resolution, every TSP allocation, every DMF expenditure, updated continuously and accessible to anyone with a phone. Fifth, institutional separation: remove FRA processing from the Forest Department. You do not ask the entity that loses jurisdiction to be the entity that recognises rights.

Each of these changes threatens a vested interest. Mining companies do not want clearances tied to PESA compliance. The forest bureaucracy does not want to lose FRA authority. State governments do not want independent monitoring. The political establishment does not want automated penalties. This is precisely why the enforcement mechanism was never built. Building it would require those who benefit from the current design to dismantle their own advantage. The mechanism that would fix the mechanism requires a meta-mechanism for implementation. And so the gap persists.

The Stack of Paper, the Depth of the Forest

Step back and look at the full architecture.

At the constitutional level: the Fifth Schedule (since 1950), the Tribes Advisory Council (since 1950), the Governor’s regulatory powers. At the legislative level: the Land Transfer Regulation (1956), PESA (1996, rules not finalised), the Forest Rights Act (2006), the Prevention of Atrocities Act (1989), the RFCTLARR Act (2013). At the judicial level: the Samatha judgment (1997, routinely circumvented). At the institutional level: the ST & SC Development Department, twenty-one ITDAs across twelve districts, thirteen PVTG micro-projects, SCSTRTI, OTDS, OPELIP, the National Commission for Scheduled Tribes, the Tribal Welfare Commissioner. At the scheme level: the Tribal Sub-Plan, PM-JANMAN, Conservation-cum-Development Plans, the Chief Minister’s Tribal Livelihood Mission, District Mineral Foundation trusts, Special Central Assistance, grants under Article 275(1).

This is an extraordinary institutional density. A tribal person in Odisha’s Scheduled Areas is “protected” by at least four layers of constitutional provisions, five major legislative enactments, one landmark Supreme Court judgment, eight to ten institutional bodies, and multiple dedicated funding schemes.

And yet: Kutia Kondh literacy is 4.98 percent. Eight lakh acres alienated, 6.75 percent restored. PESA rules unfinished for twenty-nine years. 43 percent of land acquisitions without gram sabha consent. 99.77 percent FRA rejection rate for OTFDs. Zero percent conviction rate for atrocities against Scheduled Tribes from 2019 to 2021. The Bonda Development Agency has operated for fifty years and literacy is 33 percent.

The institutional density may itself be part of the problem. Multiple overlapping institutions create diffusion of responsibility and jurisdictional confusion. No single institution is accountable for outcomes. Each can point to the others. The ITDAs point to the Forest Department. The ST & SC Department points to the Revenue Department. The Revenue Department points to the courts. The courts point to the prosecution. The prosecution points to the police. The police say procedures were followed.

This is not a system designed for accountability. It is a system designed for the appearance of protection while permitting the continuation of extraction. The institutions exist to demonstrate that the state cares about tribal welfare. They do not exist to ensure tribal communities have power. The paper accumulates. The forest recedes.

The Sundargarh gram sabhas knew this on Republic Day 2020. They voted anyway. Four hundred people signed a resolution they must have suspected would be ignored. They exercised a right they must have known would not be enforced. And eighteen months later, the notifications came anyway, as they always do, carrying the weight of the state against the weight of a constitutional promise that has never, in Odisha, been given the enforcement mechanism to become real.

The gap between Delhi’s paper and Odisha’s forest is not a gap of distance, or language, or culture, or will. It is a gap of mechanism design. The rules were written for a game that was never played. The players play a different game entirely — one where the rules sit on a shelf, referenced in speeches, cited in reports, invoked in court petitions that take years to resolve, and ignored in the daily business of extraction that funds the state, enriches the corporations, and dispossesses the communities the rules were written to protect.

This would be wrong if: PESA rules are finalised with genuine gram sabha veto power and enforced within the next five years; or if FRA community forest rights recognition reaches Maharashtra-level rates; or if the Bonda Development Agency’s next decade produces literacy gains that its first five decades did not. The evidence for any of these happening is thin. But the Dongria Kondh voted no, and Niyamgiri still stands. The mechanism can work. It has worked exactly once, in one place, at extraordinary cost. The question is whether it can be made to work as a system rather than as a miracle.

Sources

Constitutional and Legislative Sources:

Judicial Sources:

Government Audit and Reports:

Academic and Research Sources:

Journalism and NGO Sources:

Cross-references within SeeUtkal:

  • Delhi’s Odisha, Chapter 6: “The PESA Betrayal” — covers the central government’s role in PESA non-implementation; this chapter expands on the state-level failure and embeds it in the broader constitutional framework
  • The Long Arc, Chapter 5: “The Extraction Equilibrium” — the Nash equilibrium of extract-welfare-vote-extract that makes genuine tribal empowerment structurally incompatible with the state’s fiscal model
  • The Missing Middle (Value Chain series) — the per-tonne economics that make tribal land so valuable to the extraction economy and so expendable in state calculation

Source Research

The raw research that informs this series.