In the summer of 1894, a young Mohandas Gandhi stood before the Natal Assembly in South Africa to protest a law designed to strip Indians of their voting rights through procedural sleight of hand. The British colonial administration had found an elegant solution to a political problem: don’t ban voting outright simply make it administratively impossible. One hundred and thirty years later, the Bharatiya Janata Party appears to have learned from that colonial playbook.
The 2026 West Bengal Assembly elections have produced a verdict that will be argued over in courtrooms and classrooms for a generation. The BJP has captured power in a state that had resisted it for over six decades, winning 207 of 293 declared seats and reducing Mamata Banerjee’s Trinamool Congress to a rump of 80. But the scale of the victory is inseparable from a question that haunts every number in those results: how many of the voters who might have changed that outcome had already been quietly, procedurally, administratively erased?
To understand what happened in West Bengal, one must first understand the Special Intensive Revision, or SIR a mechanism so mundane in its bureaucratic framing that its political implications are easy to miss at first glance.
The SIR was first deployed in Bihar in June 2025 before being expanded to nine states and three Union Territories, including West Bengal. On paper, its purpose was hygienic: clean up electoral rolls, remove duplicate entries, weed out those who had died or relocated. No democracy should object to an accurate voter list. The problem was not the stated purpose. It was the method, the timing, and the outcome.
Under the SIR, Booth Level Officers local election officials conducted house-to-house verification requiring citizens to re-establish their voting eligibility with documentary proof within extremely tight deadlines. Failure to produce the right documents meant deletion. For the first time since India’s first general election in 1951-52, the burden of proof was reversed: citizens had to prove their right to vote, rather than the state having to prove grounds for removing them. The presumption of democratic belonging was abolished by administrative decree.
The numbers are staggering. Approximately 9 million voter entries were removed from West Bengal’s rolls 12% of an electorate of 76 million in what The Diplomat described as “the biggest revision in any state in India’s electoral history.” Of those, roughly 6 million were classified as Absent, Shifted, Dead or Duplicate the ASDD category and struck off entirely. A further 2.7 million were funnelled into adjudication proceedings over documentation errors, their democratic futures placed in the hands of tribunals deciding whether their existence on paper matched their existence in fact.
Electoral roll revisions are not inherently sinister. The sinister question is: who gets deleted? In West Bengal, the answer appears to follow a pattern that is difficult to attribute to coincidence. Multiple independent analyses and field reports documented unusually high deletion and adjudication rates in Muslim-majority localities. In several constituencies, the proportion of Muslims flagged for scrutiny reportedly far exceeded their demographic share. In some regions, villages with mixed populations saw deletions concentrated almost entirely among Muslim residents sharing the same polling booth. As The Diplomat documented in ground-level reporting from Metiabruz one of Kolkata’s largest garment manufacturing hubs Muslim families found their names deleted while their neighbours’ remained intact, despite generations of documented residence.
“Paanch peeri of my family have been living here and this is our bastu bheeta,” five generations in a family home said one woman, Masooda Bibi, who found herself deleted while her husband was not. Her story was not exceptional. It was representative.
The Dalit community faced analogous vulnerabilities. Poor archival records, inconsistent transliteration of names across linguistic traditions, migration histories and uneven access to digital infrastructure all disproportionately affect already marginalised populations. When documentation becomes the price of citizenship, those who have historically been denied access to documentation pay the highest price. As one analytical report put it: “Exclusion becomes depersonalized. No longer visibly political. Simply procedural. And procedural exclusion is often harder to challenge because it presents itself as administrative necessity rather than political intent.”
The BJP defended the exercise as targeting illegal Bangladeshi migrants. The framing was politically potent but analytically convenient: it allowed the party to present the systematic reduction of Muslim and minority voter rolls as national security policy rather than electoral engineering.
The most damning evidence is arithmetical. In the Supreme Court, senior TMC advocate Kalyan Banerjee presented a calculation that is difficult to dismiss: in at least 31 Assembly seats, the BJP’s margin of victory was lower than the number of names removed from the rolls during the SIR process. In one constituency alone, a TMC candidate lost by 862 votes while more than 5,000 names had been struck off. The overall vote difference between BJP and TMC across the state stood at roughly 3.2 million while appeals against deletions exceeded 3.5 million.
In other words, the margin of the BJP’s triumph is smaller than the pool of people who were removed from the electorate before a single vote was cast.
The Supreme Court has since opened the door for fresh legal challenges, with a bench led by Chief Justice Surya Kant hearing petitions that could yet unravel the electoral arithmetic. Mamata Banerjee has confirmed she will personally take the case forward, refusing to resign as Chief Minister in protest. TMC offices across the state played Rabindranath Tagore’s songs on the day of the new government’s inauguration a gesture of mourning that was also a declaration of intent.
The Election Commission of India was once among the most admired democratic institutions in the world a rare example, in a region of fragile democracies, of a truly independent electoral authority. That reputation has not survived the West Bengal exercise with its integrity intact.
Al Jazeera’s analysis described grounds to suspect the verdict was “stolen with the assistance of the state machinery, including the ECI, although it is constitutionally mandated to function as an impartial body.” The Wire went further, arguing that BJP’s victory “scripted by subverting universal adult franchise testifies to the charge hurled at India since 2014 that it has become an authentic example of electoral autocracy.” The opacity surrounding the SIR process combined with slow adjudication mechanisms, difficult-to-search voter lists, and widespread confusion among affected citizens generated the perception, if not the proof, of institutional capture.
The BJP countered every allegation by pointing to the record 93% voter turnout as proof of democratic health. This argument deserves scrutiny. High turnout in an environment of mass deletions does not indicate democratic vitality it indicates democratic desperation. Communities that feared their names would disappear permanently if they did not vote turned out in record numbers not to express political preference, but to defend the basic fact of their existence. That is not democracy functioning well. That is democracy functioning under duress.
In How Democracies Die, Steven Levitsky and Daniel Ziblatt wrote that democratic backsliding today begins at the ballot box not through coups or tanks, but through the gradual hollowing out of the institutions meant to protect political equality. The West Bengal case is a case study in exactly this process.
No law was openly broken. No election was visibly stolen at gunpoint. The SIR was conducted under judicial supervision, the Supreme Court itself directing the Calcutta High Court to deploy 150 district judges to assist the exercise. The BJP can say, truthfully, that it won an election that courts were watching. What the courts were watching, however, was the adjudication process not the original design of a system that placed the burden of proof on the citizen, not the state.
When a 35-year veteran of the Central Reserve Police Force finds his name missing from electoral rolls as documented in the course of legal proceedings the argument that SIR was purely a targeted tool against illegal immigrants collapses. When the citizens who are erased include the very people who spent their careers in service to the Indian state, the exercise reveals itself as something broader than border security. It reveals itself as political engineering wrapped in administrative language.
The Bengal verdict has redrawn India’s political map. The BJP now controls most of India’s most populous and politically significant states. West Bengal was the last major holdout and its fall, through whatever combination of genuine electoral swing, anti-incumbency against fifteen years of TMC rule, and systematic voter disenfranchisement, closes a chapter.
But the legal battle is only beginning. The Supreme Court challenges, the pending repolling in constituencies where EVM tampering was documented, the appeals of 3.5 million deleted voters all of these will drag through India’s judiciary for years. The question is whether the courts can restore what the administrative machinery has taken away, and whether the political will exists to let them try.
Democracy is not merely a counting exercise. It is a compact between citizens and the state: a guarantee that every person’s voice carries equal weight, regardless of their name, their faith, or the legibility of their paperwork. When that compact is broken by the very institutions charged with protecting it, the damage is not repaired by a single court ruling. It accumulates in communities that no longer trust the system, in voters who will think twice next time about whether their name is still on the list, in the slow corrosion of the belief that elections mean what they say they mean.
