How the bureaucratic deletion of 4-5 lakh Muslim voters reveals the economic cost of democratic backsliding

On 27 January 2026, Assam Chief Minister Himanta Biswa Sarma stood before reporters and said something that should have stopped the nation. "Himanta Biswa Sarma and the BJP are directly against Miyas," he declared, using the derogatory term for Bengali-speaking Muslims in Assam. "We are saying it openly; we are not hiding it. We are ensuring that they cannot vote in Assam."

He was not speaking metaphorically. He was describing, in precise administrative terms, how his government was using the Special Revision (SR) of electoral rolls to systematically disenfranchise Muslim voters. He predicted that under the upcoming Special Intensive Revision (SIR), "four to five lakh Miya votes will have to be deleted in Assam." He admitted that he had directed BJP party workers to file complaints and objections during the voter list revision process. He said his responsibility was to 'make suffer' and 'give trouble' to the Miya community to force them out of Assam.

A sitting chief minister, in a democracy with universal adult franchise enshrined in its Constitution, openly admits to using state machinery to delete hundreds of thousands of citizens from voter rolls based on their religious and linguistic identity. The Election Commission of India, the constitutional body responsible for free and fair elections, has remained silent. No notice has been issued. No inquiry has been ordered. No action has been taken.

On January 30, 2026, human rights activist and former civil servant Harsh Mander filed a police complaint against Sarma at Hauz Khas Police Station in Delhi under Sections 196, 197, 299, 302, and 353 of the Bharatiya Nyaya Sanhita for promoting enmity, outraging religious feelings, criminal intimidation, and abusing official position. Retired Allahabad High Court Chief Justice Govind Mathur released a statement saying that Sarma's utterances violated basic constitutional principles and that "as Chief Minister, Mr Sarma has taken an oath to uphold the Constitution, and his words carry the authority of the state."

What Himanta Biswa Sarma is doing in Assam is not a policy disagreement or political rhetoric. It is the administrative blueprint for how democracies die. Not through coups or constitutional amendments, but through the quiet manipulation of electoral rolls, the weaponisation of bureaucratic forms, and the systematic exclusion of citizens from the franchise. He is doing it openly, with impunity, while the institutions responsible for protecting electoral integrity remain silent.

The consequences go far beyond electoral outcomes. When you delete 4-5 lakh voters from specific communities, you are not just changing who wins elections. You are determining which areas receive infrastructure investment, which neighbourhoods get public services, which populations can access credit, and which communities face permanent economic marginalisation. This is how political exclusion becomes economic exclusion, compounding over time until the gap is irreversible.

Form 7: How You Delete a Citizen

The mechanics of voter deletion are bureaucratically mundane. Under the Representation of the People Act, 1950, the Election Commission periodically revises electoral rolls. Citizens can file three types of applications: Form 6 to add a new voter, Form 8 to correct details or update address, and Form 7 to object to someone's inclusion in the electoral roll or request deletion.

Form 7 is the weapon. It allows any person to file an objection against another person's inclusion in the voter list on grounds including: the person has died, the person has shifted residence, the person is a duplicate entry, or the person is not ordinarily resident. Once an objection is filed, the person whose name is challenged receives a notice and must appear before the Electoral Registration Officer (ERO) with documents proving their residency and eligibility.

The Election Commission announced a Special Revision of Assam's electoral rolls on 17 November 2025, with 1 January 2026 as the qualifying date. The timeline was tight: the draft electoral roll was published on 27 December 2025; the claims and objections period ran from 27 December 2025 to 22 January 2026; objections were disposed of by 2 February 2026; the final electoral roll was published on 10 February 2026.

During this period, reports emerged of mass weaponisation of Form 7. In Morigaon district, Faridud Eunush and his siblings received notices saying they had "shifted" from their address, even though the rest of their family remained on the electoral roll at the same address. When they appeared at the hearing with documents demonstrating continuous residence, the ERO refused to disclose who had filed the objection against them. Eunush was informed that he would need to file an RTI application to obtain the complainant's name.

Across Muslim-majority constituencies, voters received notices claiming they were dead or had shifted. Many discovered that their names had been challenged only when they checked the draft electoral roll online. Some found that signatures of Booth Level Officers (BLOs) on verification documents had been forged. Congress and opposition parties submitted a joint memorandum to the Chief Electoral Officer on 25 January, alleging large-scale irregularities, including fake objections, wrongful deletions, and political interference.

The draft electoral roll published on 27 December showed 2.52 crore voters. Officials identified 4.78 lakh deceased voters, 5.23 lakh voters who had shifted, and over 53,000 duplicate entries for deletion. But these were official deletions based on BLO verification. The Form 7 objections were something else entirely. They were targeted, systematic, and concentrated in areas where "Miya" Muslims lived.

This is voter suppression through bureaucratic harassment. You don't need to prove someone is ineligible. You just need to force them to prove they are eligible, again and again. You create a gauntlet of documentation requirements. You demand proof of residence going back decades. You schedule hearings during working hours in distant ERO offices. You reject documents on technical grounds. You file new objections after old ones are cleared. Eventually, people give up, or miss a hearing, or cannot produce a document that exists but is 50 years old and stored in a government office that has since burned down.

The beauty of this system, from Sarma's perspective, lies in its deniability. No Muslims are being deleted because they are Muslim. They are being deleted because they could not produce adequate proof of residence. The discrimination is laundered through procedural requirements. The targeting is obscured by bureaucratic language. And when hundreds of thousands of voters disappear from the rolls in Muslim-majority areas, the state can claim it was simply enforcing electoral integrity.

Except that Himanta Biswa Sarma is not bothering with deniability. He is saying the quiet part loud. He is admitting that the process is designed to target "Miyas." He is boasting about directing party workers to file objections. He is predicting specific numbers of deletions and framing it as a political achievement. And still, the Election Commission does nothing.

There is a revealing pattern here about how exclusion operates through measurement. As a development economist, I have spent years examining how counting systems determine who gets counted. Voter rolls. Census data. Ration cards. Land records. Each is presented as a neutral administrative exercise. Each systematically undercounts the most vulnerable.

The mechanisms are always the same. You establish evidentiary requirements that are difficult for low-income individuals to meet. You demand documents that require literacy, bureaucratic knowledge, and time to obtain. You schedule verification during working hours, forcing people to choose between proving their existence and earning their livelihood. You place the burden of proof on individuals rather than institutions. And then you count the people who cannot clear these hurdles as not existing.

Voter roll revision through Form 7 is this logic perfected. It is not enough to be on the voter list once. You must continually demonstrate that you deserve to remain. The required proof is not evidence of citizenship (you already proved that to get on the list initially), but rather evidence that can be weaponised through administrative discretion. Proof of residence. Proof that you have not shifted. Proof that you are not dead. All of which requires documents that officials can reject on technical grounds.

This matters because measurement systems have economic consequences. Who gets counted determines who gets resources. Census undercounts affect parliamentary seats, which in turn affect funding allocation. Ration card errors determine who has access to subsidised food. Land record gaps determine who can obtain credit. Voter-roll deletions determine who has a political voice to demand that any of this be remedied.

But there is another question worth asking: who benefits economically from these deletions? When 4-5 lakh voters from one community lose political voice, who gains? When land records become unreliable for people under citizenship verification, whose land claims become more secure? When "Miya" entrepreneurs cannot access formal credit, which businesses capture that market share? When certain neighbourhoods become politically irrelevant, which areas see increased infrastructure investment?

The political economy of exclusion is not just about who loses. It is about who wins. And in Assam, the pattern is clear. Upper-caste Hindu businesses benefit when Bengali Muslim competitors face credit discrimination. Land developers benefit when citizenship uncertainty forces asset sales at distressed prices. Politicians benefit when they can deliver resources to communities that can vote while ignoring communities that cannot. The exclusion creates economic winners, who have every incentive to maintain the system that privileges them.

The NRC Blueprint: 1.9 Million Rehearsals for Voter Deletion

To understand what Sarma is doing with electoral rolls in 2026, you need to understand what happened with the National Register of Citizens in 2019. The NRC was the dress rehearsal. The voter revision is the main event.

Assam is the only state in India that has updated its National Register of Citizens. The NRC is meant to identify Indian citizens and exclude illegal immigrants. The exercise was mandated by the Supreme Court in 2013 in response to a petition by Assam Public Works, following the Assam Accord of 1985, which promised "detection and deportation" of all persons who entered the state from Bangladesh after March 24, 1971.

To be included in the NRC, applicants had to prove that they or their ancestors were residing in Assam before March 24, 1971. The process involved submitting legacy documents, such as land records, voter lists from before 1971, birth certificates, educational certificates, and other papers establishing continuous residence. Families were required to appear in person before NRC officials for verification. Anyone declared a foreigner by the Foreigners Tribunals was automatically excluded.

On August 31, 2019, the final NRC was published. Out of 3.3 crore applicants, 1.9 million (19 lakh) were excluded. The results shocked everyone. Families were split. Children whose parents were included were excluded. War veterans who had served in the Indian Army were sent to detention camps before courts freed them. According to Himanta Biswa Sarma's claims in a March 2024 interview, those excluded included 7 lakh Muslims, 5 lakh Bengali Hindus, 2 lakh Assamese Hindus, and 1.5 lakh Gorkhas. His breakdown totals only 15.5 lakh, leaving 3.5 lakh unaccounted for.

The BJP, which had championed the NRC, was furious with the results. Sarma immediately called for fresh re-verification, claiming that many Indian citizens who had migrated from Bangladesh as refugees before 1971 were wrongly excluded. The state NRC coordinator appointed after the 2019 publication, HD Sarma, refused to issue rejection slips to the excluded, claiming the list was not final and contained "anomalies."

Then came the Citizenship (Amendment) Act, 2019. The CAA fast-tracks Indian citizenship for undocumented non-Muslim migrants from Pakistan, Bangladesh, and Afghanistan who arrived in India before 31 December 2014. In March 2024, when the CAA rules were finally notified after years of delay, Sarma announced that the Assam government would withdraw "foreigner" cases against non-Muslims under the CAA. Foreigners Tribunals were instructed not to pursue cases against persons belonging to the six specified communities (Hindu, Sikh, Buddhist, Jain, Parsi, Christian) who entered Assam before 31 December 2014.

The pattern became clear: the NRC excluded 1.9 million people. The CAA provides a path to citizenship for non-Muslims among the excluded. Muslims among the excluded lack such a path. They remain in limbo, vulnerable to detention, statelessness, or deportation.

But there was another consequence of the NRC. During the exercise, the biometrics of 2.7 million people were frozen, preventing them from getting Aadhaar cards and accessing welfare benefits. This occurred extralegally, without mention in Standard Operating Procedures, and was discovered only two years later, when individuals attempted to access government services. Even those who made it onto the final NRC after being excluded from the draft list found themselves unable to access welfare because their biometric details remained frozen.

The NRC taught the Assam government several lessons. First, bureaucratic processes can be weaponised to create fear and force compliance. Second, documentation requirements can be manipulated to exclude specific communities. Third, the veneer of legal procedure provides cover for discrimination. Fourth, institutions intended to provide oversight (the Supreme Court, the Election Commission, and the civil administration) will either actively facilitate the process or passively permit it. Fifth, electoral arithmetic trumps constitutional rights: protecting Hindu Bengalis while targeting Muslim Bengalis serves the BJP's vote bank strategy.

Most importantly, the NRC taught them that you can do all this openly and suffer no consequences. The NRC excluded 1.9 million people based on a process widely documented as discriminatory. No one was held accountable. No officials faced sanctions. The Supreme Court that monitored the exercise took no action. The process is still officially "not final" even though it has shaped lives, destroyed families, and created a population in bureaucratic limbo for over six years.

The voter roll revision is the NRC playbook applied to the electoral franchise. You don't need to prove someone is a non-citizen to delete them from the voter list. You just need to create enough bureaucratic obstacles that they cannot clear. And unlike the NRC, which affected citizenship status, voter deletion has immediate electoral consequences. Delete 4-5 lakh voters from Muslim-majority constituencies, and you alter electoral outcomes. You don't need to win more votes. You just need to ensure your opponents have fewer voters.

But the NRC also taught us something else: the economic cost of verification regimes. The exercise costs more than Rs 1600 crore in public funds. When 1.9 million people do not know if they are citizens, when 2.7 million cannot access Aadhaar and therefore welfare benefits, when families spend years gathering documents and attending hearings instead of working, the economy contracts.

This reveals a fundamental aspect of how institutional breakdown operates. The same administrative machinery that selectively enforces citizenship also selectively enforces property rights, contract law, and regulatory compliance. If you are marked as a potential non-citizen, banks are reluctant to extend credit. Buyers are reluctant to purchase your property. Employers are reluctant to hire you formally. Government officials are reluctant to process your applications for licenses, permits, and registrations.

The discrimination is not just political. It is embedded in every transaction that requires state verification. Every mortgage application. Every business registration. Every land title transfer. Every welfare enrollment. When the state treats your citizenship as provisional, markets treat your economic participation as provisional. And the cost of operating under perpetual suspicion is enormous. You pay higher interest rates because you are riskier. You accept lower prices because buyers discount the legal uncertainty. You stay in the informal sector because formal sector participation requires documents you may not be able to produce next year.

This is what development economists mean by institutional quality. It is not just about having laws on the books. It concerns whether enforcement is predictable, whether rights are secure, and whether individuals can make long-term economic decisions without fearing that administrative discretion will undermine their plans. Assam's NRC did not just exclude 1.9 million people from citizenship verification. It created an environment in which millions more live in economic uncertainty, making it impossible to invest, impossible to plan, and impossible to build.

And now Sarma wants to do it again, this time with voter rolls.

"Miyas" and the Language of Dehumanisation

The term "Miya" is itself a marker of how discrimination gets encoded in language. Literally meaning "gentleman" in Urdu and Bengali, the term in Assam has become a slur. It refers to Bengali-speaking Muslims, many of whose families have lived in Assam for generations, descended from peasants who migrated during British colonial land settlement policies or who fled East Pakistan during Partition and the 1971 Bangladesh Liberation War.

In 2016, a group of young poets reclaimed the term with "Miya Poetry," writing verses about their experiences of discrimination, eviction, and marginalisation. In 2019, the Assam Police filed FIRs against some of these poets for "trying to create hatred and differences between communities." The government's message was clear: you can call them "Miyas" as a slur, but they cannot embrace it as an identity.

Himanta Biswa Sarma uses the term deliberately. Not "Bengali-speaking Muslims" or "citizens of Bengali heritage," but "Miyas." Not "voters whose eligibility is under review," but "Miya votes will have to be deleted." The language dehumanises. It creates a category of people who are neither fully citizens nor fully belonging, always suspect, always requiring proof.

This matters because bureaucratic violence requires linguistic preparation. Before you can delete someone from a voter list, you need to create a category in which deletion seems reasonable. Before you can target a community, you need to mark them as different, as outsiders, as problems. The term "Miya" does this work. It signals that these are not "real" Assamese. That their citizenship is conditional. That their voting rights are subject to verification in ways that other communities' rights are not.

Sarma has made similar statements before. In July 2023, he commented on high vegetable prices, saying that "Miya vendors charge more in Guwahati compared to Assamese vendors." He emphasised clearing footpaths and urged Assamese people to initiate businesses. In December 2025, he made remarks on population growth, urging Hindu families not to limit themselves to one or two children and suggesting that, where possible, they should have three children, while stating that "family sizes among Muslims should remain limited."

The pattern is consistent: Muslims are framed as demographic threats, economic burdens, and political liabilities. The state's responsibility is to control, limit, and exclude them. When confronted with criticism, Sarma defends himself by citing Supreme Court observations about the Muslim population increase in certain districts, saying: "The only difference is that they (SC) have mentioned the word 'Muslim', and I use 'Miya' so that the Assamese Muslims do not take offence."

This is gaslighting as governance. The Supreme Court does not authorise discrimination. Data on demographic change does not justify voter suppression. And using "Miya" instead of "Muslim" does not make targeting less discriminatory. It makes it more specific. It allows Sarma to claim he is not targeting all Muslims, just this particular subset, as if precision in exclusion is somehow more acceptable than broad discrimination.

The Silence of Institutions

The most striking aspect of this entire affair is what has not happened. The Election Commission, which has constitutional authority under Article 324 to oversee free and fair elections, has not issued a notice. It has not sought an explanation from the Assam government. It has not investigated complaints of systematic misuse of Form 7. It has not commented on a chief minister openly admitting to engineering voter deletions.

This is the same Election Commission that can announce election dates, transfer government officials, impose the model code of conduct, ban exit polls, seize cash, and disqualify candidates for procedural violations. This body recently announced a nationwide Special Intensive Revision of electoral rolls in 12 states, citing concerns about voters with "dual documents" and the need for comprehensive physical verification.

But when Assam's chief minister admits on camera to using the revision process to delete hundreds of thousands of voters from a specific religious community, the Election Commission has nothing to say. This is not an oversight failure. This is institutional complicity through silence.

The Supreme Court, which monitors the NRC process and recently stayed the UGC equity regulations for being "vague" and having "dangerous consequences," has not taken suo motu cognisance of a state chief minister openly declaring intent to disenfranchise citizens. Justice Govind Mathur, a retired jurist, can release statements condemning Sarma's remarks as violations of constitutional principles. But the Supreme Court as an institution, which has not hesitated to intervene in matters from government advertising to cricket administration, remains quiet on a sitting chief minister weaponising voter rolls.

The opposition parties have filed complaints and submitted memorandums. Beyond that, there has been no sustained national outcry, no continuous session disruptions, and no mass mobilisations demanding Election Commission action. Perhaps because the victims are "Miyas" in Assam, far from national media attention, easy to ignore. Perhaps because other states have their own versions of this exclusion and do not want attention on the mechanics. Perhaps because when discrimination is bureaucratized, it becomes harder to generate outrage than when it is violent.

The media, with few exceptions, has treated this as a regional political story rather than a national constitutional crisis. "Assam CM's controversial remarks" rather than "Sitting Chief Minister admits to systematically disenfranchising religious minority." The framing matters. One is politics as usual. The other is democratic backsliding.

What 4-5 Lakh Deleted Voters Means

Numbers abstract suffering. So let us be concrete about what Himanta Biswa Sarma is engineering. He is predicting that 4-5 lakh voters will be deleted. That is 400,000 to 500,000 people who will lose the right to vote.

Assam has 126 assembly constituencies. An average constituency has roughly 2 lakh voters. Deleting 4-5 lakh voters means eliminating the voting population equivalent to 2-3 entire constituencies. But these deletions will not be spread evenly. They will be concentrated in constituencies with significant "Miya" Muslim populations.

Consider the electoral impact. In the 2021 Assam Assembly elections, the BJP-led alliance won 75 seats, and the opposition won 50. The margin of victory in many constituencies was narrow. Delete a few thousand voters from select constituencies, and you can flip results. You don't need to convince anyone to vote for you. You need to prevent your opponents' voters from voting at all.

This is not hypothetical. This is happening. Forms 7 have been filed. Notices have been sent. Hearings have been conducted. The final electoral roll is scheduled for publication on February 10, 2026. Assam Assembly elections are expected to be held before May 2026. The deletions will take effect in time for the election, and because the process is framed as routine electoral roll revision, it will be nearly impossible to challenge individual deletions on grounds of discrimination.

But there is another consequence that gets less attention: the political economy of public goods provision. When 4-5 lakh voters disappear from certain constituencies, what happens to infrastructure spending in those areas? What happens to welfare programs, development budgets, and allocation of public resources?

Electoral democracy, for all its flaws, provides a crude mechanism of accountability. Politicians compete for votes by delivering public goods to their constituents. Roads get built. Schools get funded. Health centres get staffed. Water connections get extended. Not because politicians are altruistic, but because voters can punish non-delivery.

But this mechanism only works if you can vote. If your name disappears from the voter list, you become politically irrelevant. A politician has no electoral incentive to deliver services to you. You cannot reward good performance or punish bad performance. You have no leverage.

Multiply this across hundreds of thousands of people concentrated in specific areas, and you get systematic under-provision of public goods. Constituencies with large populations of deleted voters will experience declining infrastructure investment. Development funds will flow to areas where voters can actually vote. The pattern will not be random. It will track the communities being targeted for exclusion.

This is how political exclusion translates into economic exclusion. Not just through direct discrimination, but through the resource allocation decisions that follow from changing electoral arithmetic. When Sarma deletes 4-5 lakh "Miya" voters, he is not just changing who wins elections. He is changing which areas receive government investment, which communities get access to public services, and which neighbourhoods see infrastructure development.

These effects compound over time, and the economic divergence grows. Areas with political voice accumulate public goods. Areas without political voice fall further behind. The initial exclusion from voter rolls becomes a permanent exclusion from development. Twenty years from now, researchers will point to infrastructure gaps, income disparities, and educational differences between communities and attribute them to "cultural factors" or "historical disadvantage," without noting that the divergence began when one community lost the right to vote.

Each deletion is a person. Someone who has voted before. Someone who may have voted in every election since adulthood. Someone whose grandparents were born in Assam. Someone with land records, ration cards, Aadhaar, PAN card, driving license, and bank accounts. But someone whose name was challenged by an anonymous objector using Form 7. Someone who received a notice asking them to prove they still live at the address where they have always lived. Someone who had to take time off work, travel to an ERO office, stand in queues, and submit documents. Someone whose documents were found insufficient for reasons not clearly explained. An individual whose name was removed from the voter list. Someone who, when they go to vote in the next election, will be turned away. Someone who will be told they are not on the list. Someone who will become a non-person in the democracy to which they belong.

Multiply this by 400,000 to 500,000, and you have what Himanta Biswa Sarma calls a political achievement. What constitutional scholars would call mass disenfranchisement. What historians might recognise as the administrative mechanics of how democracies become ethnocracies.

CAA-NRC-Voter Deletion: The Citizenship Squeeze

The CAA, NRC, and voter-roll manipulation are not distinct policies. There are three parts of a coordinated system to create a two-tier citizenship based on religion.

The NRC maintains a database of "verified" citizens and excludes those who cannot demonstrate ancestry prior to 1971. The CAA provides a safety net for non-Muslims who are excluded, offering them a path to citizenship. Muslims who are excluded have no such path. They face Foreigners Tribunals, potential detention, and the risk of statelessness.

Now add voter roll manipulation. Even Muslims who were included in the NRC or were never subjected to NRC verification face a new test: continuous residence verification through electoral roll revision. Form 7 objections filed by BJP workers force them to prove eligibility again. Those who cannot clear the evidentiary bar are deleted from voter lists.

The result is a citizenship squeeze. You create multiple verification points, each with different documentation requirements. Some people make it through one but not another. At every stage, there is discretion. EROs can accept or reject documents. BLOs can verify or challenge residence. Tribunals can determine whether someone is a citizen or a foreigner. And at every stage, there is bias. Not always explicit. Sometimes it is just that an ERO is more sceptical of a Muslim name. Alternatively, a BLO may not attempt to verify residence in a Muslim neighbourhood. Or a Tribunal member finds upper-caste Hindu testimony more credible than Muslim testimony.

Multiply these small biases across millions of verification instances, and you obtain systematic exclusion that appears as individual determinations. You get 1.9 million people excluded from the NRC. You get 4-5 lakh voters deleted from the electoral rolls. You get Bengali Hindus protected by CAA while Bengali Muslims remain vulnerable. You get an administrative apparatus of discrimination that operates through procedure rather than declaration.

And crucially, you create a population living in a state of precarity. People who do not know if their next Aadhaar update will go through. People who do not know if their name will be on the voter list when they go to vote. People who do not know if a Foreigners Tribunal notice will arrive declaring them non-citizens despite having lived in India for generations. This precarity is itself a form of control. When people are constantly proving their belonging, they have less capacity to organise, protest, or challenge the state.

But precarity also has profound economic consequences. I spent a substantial amount of time in Bangladesh's Rangpur Division studying seasonal migration and hunger. One thing you learn quickly: people facing uncertain futures make different economic decisions than people facing secure futures. They invest less in education because they do not know if they will be around to reap the returns. They do not buy land because they fear expropriation. They stay in casual labour rather than investing in skill development. They keep assets liquid rather than fixed. They prioritise survival over accumulation.

This is what economists call "investment under uncertainty." When your property rights are insecure, you underinvest in improvements. When your citizenship status is questionable, you do not start businesses that require long-term planning. When you might need to flee at short notice, you keep your wealth in forms that are portable. Gold instead of land. Cash instead of bank deposits. Livestock instead of orchards.

Migration patterns change. In Rangpur, we documented how seasonal hunger drove temporary migration to cities. But that migration was temporary because people had secure land rights back home. They could leave during the lean season and return during harvest. But when land rights become insecure, when you do not know if the state will recognise your claim when you return, migration becomes permanent. You do not come back.

This is what Sarma's voter deletion will produce. Not just political exclusion, but economic decisions shaped by insecurity. Families will sell assets while they still can. Young people will migrate permanently rather than temporarily. Entrepreneurs will close businesses rather than risk losing everything to administrative discretion. Communities that have been economically productive for generations will hollow out, not because they lack capability, but because operating under perpetual verification is economically impossible.

The long-term development cost of this precarity is staggering. Human capital flight. Currency capital flight. Entrepreneurial flight. The people with the most options leave first. Those who remain are those who cannot leave, and they survive by minimising their exposure to the state. They work informally. They do not register businesses. They avoid banks. They stay invisible. An economy built on people trying to remain invisible cannot grow.

This is not speculation. This is what we saw in Assam during the NRC. Per capita income fell by over Rs 19,000. Not because people became less productive, but because productivity requires security. You cannot run a business when you do not know if you will still be a citizen next year. You cannot invest in your children's education when you do not know if they will be allowed to finish school. You cannot plan anything when the state treats your existence as provisional.

The Question Nobody Wants to Answer

If a chief minister can openly admit to using state machinery to delete hundreds of thousands of citizens from voter rolls based on their religious and linguistic identity, and the Election Commission takes no action, what does that tell us about the state of Indian democracy?

We spend an enormous amount of time debating abstract constitutional principles. We write learned articles about secularism, federalism, and fundamental rights. We cite Supreme Court judgments about equality before the law and the prohibition of discrimination. We celebrate elections as festivals of democracy and voter turnout as civic participation.

But when the mechanism of voting itself becomes a weapon to exclude specific communities, when chief ministers boast about engineering disenfranchisement, when constitutional bodies remain silent in the face of explicit discrimination, all the constitutional philosophy in the world means nothing. Democracy is not what the Constitution promises. It is what institutions enforce. Currently, institutions are not enforcing universal franchise. They are enabling its opposite.

Himanta Biswa Sarma is not acting in defiance of institutions. He is acting with their acquiescence. He is not breaking rules. He is using rules (Form 7, electoral roll revision, verification procedures) to break people. He is not hiding his discrimination. He is advertising it because he knows there will be no consequences.

This is the question that haunts: if the Election Commission will not act when a chief minister admits to mass voter deletion, what will it take for them to act? If the Supreme Court will not intervene when voting rights are systematically targeted, what constitutional violation would merit intervention? If political opposition cannot generate sustained pressure around disenfranchisement, what issue can generate pressure?

The answer, I fear, is that we have moved past the point where institutions reliably protect rights. We are in a phase where institutions are tools to be captured and wielded by whoever holds power. The Election Commission is not an independent constitutional body protecting electoral integrity. It is an administrative arm that follows political direction. The Supreme Court is not a guarantor of constitutional values. It is a court that selectively enforces some rights while ignoring violations of others. Democracy is not a system in which every citizen has equal voting rights. It is a system in which voting rights are contingent on identity and vulnerable to bureaucratic manipulation.

This voter suppression fits a broader pattern I have documented in my work on fiscal federalism: the systematic weakening of democratic accountability at every level. The centre uses cesses and surcharges to circumvent constitutional revenue-sharing with the states. States use citizenship verification to bypass accountability to certain populations within their borders. Districts use measurement systems to exclude vulnerable groups from resource allocation. At each level, those with power find administrative mechanisms to avoid sharing it.

And the mechanism is always the same: convert political questions into technical questions. Fiscal federalism becomes a question of "optimal tax design." Citizenship becomes a question of "document verification." Voting rights become a question of "electoral roll accuracy." The technical framing obscures the political reality: powerful groups use administrative procedures to maintain their advantages while presenting them as neutral governance.

Rohith Vemula wrote that "the value of a man was reduced to his immediate identity and nearest possibility. To a vote. To a number. To a thing." Himanta Biswa Sarma has found a way to reduce people to less than even that. To a name on a list that can be deleted. To a voter whose vote can be prevented. To a citizen whose citizenship is subject to perpetual verification. To someone who exists at the state's pleasure, whose belonging is always conditional, whose rights are always revocable.

This is what democratic backsliding looks like. Not dramatic. Not sudden. Just a chief minister explaining, in bureaucratic detail, how he is deleting voters. Just an Election Commission staying silent. Just institutions that exist but do not protect. Just a democracy that looks like a democracy until you try to vote and find your name is not on the list.

What Can Be Done

On 10 February 2026, Assam will publish its final electoral roll. Hundreds of thousands of names will be missing. Some will be legitimate deletions, deceased or moved. Others will be citizens who lived at the same address for generations but could not produce documents that satisfied an ERO. Others will be people whose only crime was having a Muslim-sounding name in a state where the chief minister openly admits he is targeting their community.

When those elections happen, journalists will report voter turnout percentages. Political analysts will discuss vote share and seat distribution. Exit polls will predict winners. No one will mention the individuals who tried to vote and were told they were not on the list. Nobody will count the votes that were prevented before they could be cast. Nobody will document the systematic exclusion that happened not through violence but through voter list revision.

However, this need not be inevitable. There are concrete actions that can challenge this administrative violence:

For the Election Commission: The simplest intervention would be institutional courage. The ECI has the constitutional authority to stay the 10 February publication, order an independent audit of Form 7 complaints, mandate that EROs reveal the identity of objectors, and establish clear evidentiary standards that cannot be manipulated through administrative discretion. The fact that they have not done so is a choice, not a constraint.

For the legal community: Public interest litigation challenging the voter deletions on grounds of Article 14 (equality), Article 15 (prohibition of discrimination), and Article 326 (universal adult franchise) is urgently needed. The Supreme Court that stayed UGC equity regulations for being "vague" should be equally troubled by a voter revision process that a chief minister admits is designed to delete voters from a specific religious community.

For opposition parties: The current complaints and memorandums are insufficient. What is needed is sustained parliamentary disruption, state-level protests, filing of individual legal challenges on behalf of deleted voters, and international attention through approaches to UN Special Rapporteurs on minority rights and free elections. The BJP won 75 seats in 2021, with the opposition winning 50. Deleting 4-5 lakh voters can flip that arithmetic. This is not just about constitutional principles. This is about electoral survival.

For civil society and documentation groups: The most important work is the unglamorous work of documentation. Every person who receives a Form 7 notice needs legal aid to challenge it. Every deletion must be documented with the person's name, address, the documentation submitted, and the grounds for rejection. Every hearing needs observers. This creates an evidentiary record that can be used in litigation, in journalism, and in historical accounting.

For affected communities, the instinct when facing state harassment is to remain invisible. But invisibility is exactly what the state wants. Communities facing voter deletion need to organise voter registration drives, create mutual support networks for documentation assistance, establish funds for legal challenges, and, most importantly, speak publicly about what is happening to them. The state relies on victims staying silent out of fear.

For journalists and researchers: Stop treating this as a regional political story. This is a national constitutional crisis that is occurring first in Assam. Document the economic consequences: which businesses are losing access to credit, which neighbourhoods are seeing declining property values, and which schools are losing enrollment as families migrate. Follow the money: who is buying the land that people under citizenship verification are selling at distressed prices? Track the public goods provision: which constituencies see infrastructure spending decline as their voter populations shrink?

For international observers: India's democratic backsliding matters globally. The world's largest democracy normalising mass disenfranchisement creates precedent for authoritarian regimes everywhere. International election-monitoring organisations, UN bodies, and bilateral partners need to make voter-roll manipulation a condition of engagement. The Economic cost of precarity extends beyond India's borders. When businesses cannot rely on stable property rights, and workers face citizenship uncertainty, foreign investment declines, trade relations become unstable, and regional economic integration becomes impossible.

For development institutions: The World Bank, Asian Development Bank, and bilateral development agencies funding projects in Assam need to ask: What happens to your infrastructure investments when the populations they serve have no political voice to demand maintenance? What happens to your education projects when students do not know if they will be allowed to complete school? Institutional quality is not just about corruption and regulatory efficiency. It is about whether people have the political rights to hold governments accountable for development outcomes.

The most important thing to understand is that this is not inevitable. Sarma is doing this openly because he believes there will be no consequences. Every day that the Election Commission stays silent proves him right. Every hearing that deletes a voter without revealing who filed the objection proves him right. Every media story that frames this as "controversial remarks" rather than "admission of systematic disenfranchisement" proves him right.

Democracy does not die in darkness. It dies in broad daylight, bureaucratically, through forms, procedures, and official notices. It dies when institutions that could act choose not to. It dies when people who could challenge it decide it is not worth the effort. It dies when journalists cover the spectacle but not the substance. It dies when economists analyse GDP growth but ignore the institutional collapse that makes growth unsustainable.

This is how democracies die. Not with coups but with Form 7. Not with constitutional amendments but with voter list revisions. Not with violence but with bureaucracy. Not dramatically but administratively. Not despite institutions but through them.

But it is also how democracies can be saved. By institutions that choose to act. By lawyers who file the cases. By journalists who tell the story. By voters who refuse to accept deletion. By communities that organise. By researchers who document. By citizens who understand that when the state comes for one community's voting rights, it is coming for everyone's democracy.

The final electoral roll will be published on February 10, 2026. There is still time. Not much. But some. The question is whether anyone will use it.

Varna is a development economist and writes at policygrounds.press

Further Reading

On the National Register of Citizens and its economic impact:

1. Final NRC list published, 19 lakh excluded (The News Minute, August 2019)

2. Is Assam's final NRC really final? (Scroll, May 2022)

3. Socio-Economic impact of NRC in Assam (Academic study, 2020)

On the Citizenship Amendment Act and its implementation:

1. Assam government to withdraw foreigner cases against non-Muslims (CJP, August 2025)

2. CAA Rules 2024 notified (SCC Online, March 2024)

On Assam's electoral roll revision:

1. Special Revision of Electoral Rolls in Assam (December 2025)

2. Political tensions rise over voter roll revision (The Sunday Guardian, January 2026)

3. Voters marked 'shifted', 'dead' in electoral rolls (The Tribune, January 2026)

On Himanta Biswa Sarma's statements:

1. In a rickshaw, if the fare is Rs 5, give them Rs 4 (The Wire, January 2026)

2. Police complaint filed over divisive remarks (The Wire, January 2026)