The Lift Line
If the deletions are wrong, a fifth of two States has lost the vote. If they are right, the Election Commission has just told us that the elections it ran on the old rolls were held on a register inflated by a fifth. There is no third reading, and both are indictments.
Why This Editorial Matters for Your Exam
This is the rare editorial that hands you a complete GS2 answer structure on a live constitutional controversy: a constitutional body, a statutory process, a documented failure mode, a Supreme Court precedent that speaks directly to it, and a transparency remedy. The Election Commission appears in the syllabus under constitutional bodies, but examiners increasingly reach it through accountability and transparency, which is exactly where this piece sits.
GS Paper 2: Constitutional bodies, their powers, functions and responsibilities; salient features of the Representation of People’s Act; transparency and accountability in governance; the electoral process and electoral reform.
Background and Context
An electoral roll is revised in one of two ways. A summary revision is the annual housekeeping exercise: claims and objections are invited, Forms 6, 7 and 8 are processed, and the roll is updated at the margins. An intensive revision is the rarer and far heavier exercise in which enumerators visit households afresh and the roll is substantially rebuilt. The Special Intensive Revision (SIR) now under way is of the second kind, and it has been extended State by State after being run first in Bihar.
The statutory architecture is straightforward. Part III of the Representation of the People Act, 1950 governs the preparation of rolls. Section 21 provides for their preparation and revision, Section 22 for correction of entries and Section 23 for inclusion of names. The Registration of Electors Rules, 1960 supply the machinery, with Rule 21A dealing with deletion. The officer who actually decides is the Electoral Registration Officer (ERO), appointed under Section 13B, with an appeal to the District Magistrate and a further appeal to the Chief Electoral Officer under Section 24.
Above all of this sits Article 324, which vests the superintendence, direction and control of the preparation of electoral rolls in the Election Commission of India. Article 325 forbids exclusion from the roll on grounds of religion, race, caste or sex, and Article 326 makes adult suffrage the basis of the franchise. The provision that makes the roll decisive is Section 62 of the Representation of the People Act, 1951: no person whose name is not on the roll may vote. Deletion from the roll is therefore not an inconvenience. In law, it is the removal of the right.
The Analysis
The scale is the first fact, and it is not a marginal one. The revision has struck out close to 22 per cent of electors in Telangana and 19.5 per cent in Karnataka, among the highest anywhere in the country. The concentration is sharper still: five constituencies in Bengaluru have lost more than half their electors, and nine of Hyderabad’s fifteen have seen deletions above 40 per cent. These are not the numbers of a roll being tidied. They are the numbers of a roll being rebuilt.
The second fact is that the migration explanation does not survive its own logic. Deletion is justified where the elector has left. But an elector who moves from one part of Karnataka to another does not disappear from the Karnataka roll; her entry is relocated within it. For the deletion to be sound she must have left the State altogether. That means the deletions assert out-migration on a scale neither survey data nor the government’s own population projections support, and they assert it most strongly for Bengaluru and Hyderabad, among the largest net recipients of in-migration in India. The direction of the claim runs against the direction of the demography.
The third fact is the closed loop. If the new roll is accurate, the old roll was inflated by roughly a fifth and the Commission was conducting elections on it. If the old roll was sound, the new one is a mass disenfranchisement. The Commission is the only body positioned to settle which, and it has declined to supply the instrument that would let anyone else check. The elector-to-population (EP) ratio is the standard diagnostic for under-enrolment, its publication is mandatory, and it has not been published for any State during this revision.
The Commission has an answer on the ratio, and it should be stated. Chief Election Commissioner Gyanesh Kumar said publicly in February 2026 that the elector-to-population ratio has not been calculated because no fresh census has been conducted. That is an explanation for the absence, not a justification for it, because the same denominator problem did not prevent publication during earlier revisions, and because projections have been used for every other administrative purpose in the interim.
The fourth fact is the quality of the disclosure that did happen. In Karnataka the Chief Electoral Officer released deletion lists across multiple Google Drive links, without the old booth numbers that would let a name be traced to its previous entry, and without a gender-wise break-up. This is not a technical shortcoming. Verification of a deletion needs exactly two things: the old entry and the reason. Withholding the key that links the new lists to the old rolls makes the data unauditable while still allowing the Commission to say the data was released.
The fifth fact is the burden of proof, and it is the design flaw underneath all the others. The revision places the onus on the elector to notice she has been struck off and to file an objection. The editorial offers a candid explanation for why so few do: a voter identity card matters to most citizens once in several years, so vigilance between elections is low. During the Bihar revision the Supreme Court itself asked why political parties had filed so few objections seeking restoration of wrongly deleted names. A safeguard that depends on the alertness of the person least likely to be alert is not a safeguard.
The sixth fact is that we already know how this ends, because West Bengal has run the experiment. The Commission’s “logical discrepancy” process disenfranchised lakhs of electors, who were pushed to tribunals constituted under the Supreme Court’s aegis. Records obtained under the Right to Information Act, 2005 show that of nearly 38 lakh appeals before 19 tribunals, barely 82,000 had been decided months after the elections, and more than 90 per cent of those decisions went in favour of the elector seeking restoration. Read that success rate carefully. It is not the signature of a system making occasional random errors. A process whose decisions are reversed nine times in ten was not erring randomly. It was erring in one direction.
Data and Institutions Vault
Prelims-grade facts:
The deletions:
- Telangana: approximately 22 per cent of electors deleted
- Karnataka: approximately 19.5 per cent of electors deleted
- Five Bengaluru constituencies: more than 50 per cent of electors deleted
- Nine of Hyderabad’s 15 constituencies: deletions above 40 per cent
West Bengal aftermath:
- Appeals filed: nearly 38 lakh before 19 tribunals
- Appeals decided: barely 82,000
- Share of decided appeals favouring the elector: more than 90 per cent
Constitutional provisions:
- Article 324: superintendence, direction and control of elections and of the preparation of electoral rolls, vested in the ECI
- Article 325: no person to be excluded from the roll on grounds of religion, race, caste or sex
- Article 326: adult suffrage as the basis of elections to the Lok Sabha and State Assemblies
- Article 324(5): the CEC may be removed only in the same manner and on the same grounds as a judge of the Supreme Court; other Election Commissioners only on the CEC’s recommendation
Statutory provisions:
- Representation of the People Act, 1950: Part III on preparation of rolls; Section 13B (Electoral Registration Officer); Section 21 (preparation and revision); Section 22 (correction); Section 23 (inclusion); Section 24 (appeals to DM and CEO)
- Registration of Electors Rules, 1960: Rule 21A (deletion of names); Form 6 (inclusion), Form 7 (objection to inclusion or seeking deletion), Form 8 (correction, shifting, replacement)
- Representation of the People Act, 1951: Section 62: only a person whose name is on the roll may vote
Case law:
- Lal Babu Hussein v. Electoral Registration Officer (1995): names cannot be struck off without notice and an opportunity of hearing; an elector already on the roll enjoys a presumption in her favour
- Anoop Baranwal v. Union of India (2023): Supreme Court prescribed an appointment committee for the ECI; Parliament thereafter enacted the Chief Election Commissioner and Other Election Commissioners (Appointment, Conditions of Service and Term of Office) Act, 2023
The institution:
- Composition: Chief Election Commissioner and two Election Commissioners
- Incumbent CEC: Gyanesh Kumar, the 26th CEC, in office since 2025, sworn in on 19 February 2025
- Established: 25 January 1950, commemorated as National Voters’ Day
Key concept:
- Elector-to-Population (EP) ratio: electors as a proportion of the eligible population; the standard test for under-enrolment, and the figure not published during this revision
The Debate
The case for an intensive revision is real and should not be caricatured. Rolls that go decades without house-to-house verification accumulate the deceased, the permanently migrated and the duplicated. Every such entry is a standing opportunity for impersonation, and the Commission is under a statutory duty to maintain an accurate roll. A revision that removed only the names nobody contests would be cosmetic. Some disruption is intrinsic to the exercise, and a Commission unwilling to bear any of it would never clean a roll at all.
The case against is not that the roll should never be revised. It is that this revision has been built to fail in one direction. Where the elector must object or lose, where the data needed to object is withheld, and where appeals succeed nine times in ten, the errors are not distributed randomly. The West Bengal appeal statistics are the strongest evidence in the entire debate, and they are evidence generated by the Commission’s own process.
The sharpest version of the criticism is institutional rather than statistical. The Election Commission has historically drawn its authority not from its powers, which are broad, but from its demonstrated impartiality. An institution once envied by wealthier democracies for bringing unprecedented numbers of citizens to the polling booth is now, in the editorial’s formulation, an obstacle to universal adult franchise. That reputational shift is a governance fact in its own right, because an electoral authority that is not believed cannot certify a result.
Diagram-in-Words
How to Think About This
The transferable frame here is the allocation of the burden of proof in a rights-extinguishing process, and it is worth far more than this single controversy.
Whenever the State runs a process that can remove an entitlement, ask three questions. Who bears the burden? What does the person bearing it need in order to discharge it? What happens by default if she does nothing? In a well-designed process the State must show cause before removal, the affected person receives individual notice with reasons, and her inaction preserves the status quo. In a badly designed process she must discover the removal herself, is denied the data needed to contest it, and loses by default.
Apply those three questions elsewhere and the pattern recognises itself at once: citizenship verification exercises, Aadhaar-based deletion of ration cards and pension rolls, land record digitisation, and beneficiary purges under welfare schemes. In every case the stated purpose is accuracy, and the actual outcome is decided entirely by the default. This is why Lal Babu Hussein travels beyond election law: it holds that the person already on the register enjoys a presumption, and the State must displace it.
For GS4, note the ethical inversion in the metric. Administrative performance measured as names removed rewards over-deletion, because a wrongly retained name is conspicuous while a wrongly deleted one is invisible unless the elector complains. When the error that is easy to see and the error that is costly to society are different errors, the metric will drift towards the visible one.
Way Forward
Publish the elector-to-population ratio, State-wise and constituency-wise, at each stage of the revision. It is mandatory, it is the only test of under-enrolment, and its absence is precisely what makes the current figures unadjudicable.
Release deletion data in machine-readable form, keyed to the previous booth and serial number, with the statutory ground for each deletion recorded. Verification requires the link between the old entry and the new absence. Without it, disclosure is nominal.
Invert the default. Deletion should follow Commission-initiated notice to the elector at her registered address and a recorded opportunity of hearing, as Lal Babu Hussein requires, rather than following her failure to object.
Audit before finalisation. A random sample of deletions in each district should be independently re-verified and the error rate published before the roll is adopted. The West Bengal appeal outcomes show exactly what such an audit would have caught.
Resource the appellate tier to the volume it will actually face. Nineteen tribunals disposing of 82,000 out of 38 lakh appeals is a remedy that exists on paper. A right restored after the election is not a right restored.
PYQ Linkage and Practice
Connects directly to standing UPSC themes on the Election Commission as a constitutional body, electoral reform, and transparency and accountability in governance, and to the recurring GS2 question of whether institutional independence is secured by appointment procedure or by disclosure practice.
Practice question: “An electoral roll revision that places the burden of objection on the elector converts a right into a claim. Critically examine, with reference to constitutional provisions and judicial precedent, the safeguards available against wrongful deletion from electoral rolls in India.” (250 words)
Sources: The Hindu, Election Commission of India
Source: Unkind Cuts: The SIR and the Vanishing Voter in Telangana and Karnataka — Ujiyari.com | Free UPSC & State PCS Editorial Analysis
