Introduction
India’s Special Intensive Revision (SIR) of electoral rolls is often described through a single number or the names removed. Yet a deletion from a draft roll, a net reduction in a final roll and an unresolved verification case are not the same event. Across three phases beginning in June 2025, the Election Commission of India (ECI) initiated revision exercises covering roughly 7.89 crore electors in Bihar, 51 crore across 12 further jurisdictions and 36.74 crore across 19 more. These populations were recorded on different dates and they are not a single, contemporaneous electorate. The published figures reveal substantial state-to-state differences but their meaning depends on when and what was counted.
Special Intensive Revision: Authority to Revise, Duty to Account
Article 324 entrusts the ECI with the supervision and control of elections, while Article 326 grounds elections to the legislatures in adult suffrage. Section 21(3) of the Representation of the People Act, 1950 provides for special revision of electoral rolls. In Association for Democratic Reforms v. Election Commission of India (27 May 2026), the Supreme Court upheld Bihar’s SIR framework, while recognising procedural safeguards and distinguishing electoral eligibility inquiries from conclusive citizenship determinations. The judgment settles the Commission’s authority to undertake such a revision; it does not establish that every subsequent omission is correct. That question requires examining the roll and the route for correction.
Phase I: Bihar Shows Why Stages Matter
Bihar began with approximately 7.89 crore registered electors in June 2025. When its draft roll appeared on 1 August, it contained approximately 7.24 crore names, around 65 lakh fewer than the starting roll. The final publication on 30 September recorded approximately 7.42 crore electors. Between draft and final publication, about 21.53 lakh names were added and another 3.66 lakh removed. The final roll therefore stood roughly 47 lakh below the original, a net contraction of about 6%.
These are three different measures. The 65-lakh draft gap describes the first published result of enumeration and the 47-lakh net reduction describes the endpoint after corrections. Neither is a verified count of ineligible citizens, nor a measure of wrongful exclusion. An elector removed at draft stage and later restored appears in the former number but not necessarily the latter. Conversely, a final net change conceals the larger volume of separate additions and deletions occurring during the process.
Bihar thus supplies the measuring rule for later phases: compare like stages, identify the denominator and disclose intervening corrections.
Phase II: One Exercise, Markedly Different Results
The second phase was ordered in October 2025 for 9 states and 3 Union Territories, beginning with approximately 50.997 crore electors. Even before draft publication, administrative indicators needed care. An ECI bulletin counted enumeration forms as digitised even when electors were recorded as absent, shifted, deceased or duplicated. Consequently, a near-complete digitisation rate was not the percentage of electors successfully retained.
The final-roll contrasts were striking. Uttar Pradesh’s electorate fell from about 15.44 crore to 13.40 crore, a reduction of approximately 13.2%. Gujarat moved from 5.08 crore to 4.40 crore, about 13.4%. Tamil Nadu from 6.41 crore to 5.67 crore, about 11.6%. Kerala’s February final publication, by contrast, recorded 2.695 crore against a starting 2.785 crore, a decline of only 3.2%. Chhattisgarh contracted by approximately 11.8%, while Madhya Pradesh recorded a reduction of about 6%. Even the smaller territories diverged. Lakshadweep’s final electorate was roughly 0.4% below its starting roll, against approximately 16.9% in the Andaman and Nicobar Islands. The contrast matters all the more because the electorates differ so greatly in size.
A larger percentage does not prove a more accurate revision, nor does a smaller one prove a less effective exercise. The figures could reflect different registration histories, movement of populations, field verification and claims processing. Without comparable breakdowns of the reasons for exclusion and successful challenges, the rankings tell us the scale of change, not its correctness.
Rajasthan illustrates the importance of scope. Its final comparison covers 199 Assembly constituencies because the Anta exercise was deferred. Within that comparable population, approximately 12.91 lakh additions and 2.43 lakh further deletions after draft publication produced a net increase of roughly 10.49 lakh. Combining that final roll with an earlier all-state baseline would manufacture an inaccurate rate. Goa raises a different denominator problem. Its final roll was 2.53% smaller than its draft but approximately 10.8% smaller than its pre-SIR roll. Neither percentage is inherently wrong and neither is meaningful without identifying the comparison.
West Bengal presents an additional qualification. Its February roll was followed by supplementary adjudication and publication. It is therefore a dated result, not necessarily the settled position by October. Phase II cannot be reduced to a single league table without specifying each state’s reference date and whether further revisions remain open.
Phase III: The Biggest Gaps Are Still Draft Figures
The ECI’s May 2026 order covered 16 states and 3 Union Territories, with approximately 36.74 crore registered electors at announcement. Implementation and final-publication dates were staggered and, in several cases, extended. As of 20 October, comparisons across this phase are especially vulnerable to mixing unfinished drafts with final rolls.
Three large jurisdictions show why this matters. Delhi’s pre-draft electorate of 1,45,10,299 fell to 97,53,577 in its draft: a gap of 47.57 lakh, or 32.8%. Maharashtra’s comparable figures were 9,78,54,049 and 7,71,65,562, a gap of 2.07 crore, or 21.1%. Karnataka’s were 5,54,32,314 and 4,46,80,151, a gap of 1.08 crore, or 19.4%. These calculations use the respective state-reported pre-draft populations, which differ slightly from the ECI’s earlier planning annexure.
Those are major administrative discontinuities requiring explanation, not established final deletions. Karnataka’s final publication was scheduled for 27 October; Maharashtra’s for 16 November; Delhi’s for 4 December. Measuring those draft gaps against Bihar’s completed final reduction would place different stages on the same scale. The figures instead identify where the volume, grounds and outcomes of notices and claims deserve particular scrutiny.
Final results elsewhere in Phase III underline the problem with treating all 19 jurisdictions alike. Punjab’s October final roll contained approximately 2.01 crore electors, around 6.3% below its May baseline. Sikkim’s final roll contained 4,18,467 electors against a state-reported starting population of 4,71,018, a reduction of 11.2%. These are final net changes, unlike the three large-state draft gaps.
Completed Phase III results also reveal why post-draft scrutiny matters. Mizoram’s draft listed 8,28,906 electors. Its subsequent 39,243 additions and 32,280 deletions yielded 8,35,869 in the final roll: a net post-draft gain of only 6,963 despite more than 71,000 gross changes. Uttarakhand likewise moved from a draft of 71,33,785 to a final 71,58,949 after 1,29,872 additions and 1,04,708 deletions. A net increase of 25,164 thus masks more than 2.34 lakh decisions moving in opposite directions.
Meghalaya offers another distinction. Of its enumeration population, 2,22,786 electors were not matched to older electoral records. The state expressly treated these as cases requiring verification, not as automatic deletions. Being “unmapped” is a status in a checking process, not an independent legal finding that someone lacks the franchise.
What the Published Numbers Can and Cannot Decide
Across the phases, the strongest measurable conclusion is variation, not uniformity. Bihar’s draft to final recovery, Kerala’s relatively modest final net reduction, Rajasthan’s substantial post-draft additions and the unfinished gaps in Delhi, Maharashtra and Karnataka describe different administrative pathways. They cannot be merged into one national deletion percentage without erasing the very differences they reveal.
For meaningful public scrutiny, electoral authorities should publish standardised state-wise tables giving the starting roll, forms processed, draft omissions by recorded reason, notices issued, claims received, additions and deletions after adjudication, appeals and final totals each with dates and constituency coverage. The distinction between gross transactions and net change is indispensable.
Accurate rolls protect electoral integrity. Erroneous exclusions threaten the franchise protected by the Constitution. The published numbers establish the scale and uneven progress of SIR but cannot alone settle how many removals were justified or how many eligible electors struggled to return. The republic needs both careful revision and a publicly traceable account of its consequences.
Frequently asked questions
What is the Special Intensive Revision (SIR) of electoral rolls?
The SIR is a special revision of electoral rolls undertaken by the Election Commission of India, for which Section 21(3) of the Representation of the People Act, 1950 provides. Across three phases beginning in June 2025 it covered roughly 7.89 crore electors in Bihar, 51 crore across 12 further jurisdictions and 36.74 crore across 19 more.
Did the Supreme Court uphold the SIR?
In Association for Democratic Reforms v. Election Commission of India (27 May 2026), the Supreme Court upheld Bihar’s SIR framework, while recognising procedural safeguards and distinguishing electoral eligibility inquiries from conclusive citizenship determinations. The judgment settles the Commission’s authority to revise; it does not establish that every omission is correct.
How many names were removed from the Bihar electoral roll?
Bihar’s draft roll of 1 August 2025 had about 7.24 crore names, around 65 lakh fewer than the starting 7.89 crore. The final roll of 30 September recorded about 7.42 crore electors after about 21.53 lakh additions and 3.66 lakh removals, a net reduction of roughly 47 lakh or about 6%.
Why can the SIR figures of different states not be compared directly?
The figures were recorded on different dates and at different stages: draft rolls, final rolls and net changes after adjudication measure different things. Delhi, Maharashtra and Karnataka, for example, show large draft gaps that are not established final deletions.
Further reading from AFPR: Petition, Not Placard: Why the Street Cannot Stand In for the Supreme Court, the Election Commission and the Constitution
References
Table of Cases
Association for Democratic Reforms v Election Commission of India 2026 INSC 564 (SC).
Table of Legislation
Constitution of India 1950.
Representation of the People Act 1950.
Bibliography
A. Official Publications and Electoral Authority Records
Akashvani News, ‘ECI Revises Maharashtra Electoral Roll Schedule; Final List on Nov 16’ (26 September 2026) https://newsonair.gov.in/eci-revises-maharashtra-electoral-roll-schedule-final-list-on-nov-16/ accessed 20 October 2026.
Chief Electoral Officer, Chhattisgarh, ‘Press Note: SIR 2025–26 Final Publication’ (21 February 2026) https://ceochhattisgarh.nic.in/assets/pdf%20files/PN21.02.2026FinalPublicationofERollChhattisgarh.pdf accessed 20 October 2026.
Chief Electoral Officer, Delhi, ‘Delhi Special Intensive Revision (SIR): Key Findings of Enumeration Phase’ (Press Note, 31 August 2026) https://ceodelhi.gov.in/PDFFolders/2026/Draft_PublicationSIR31Aug2026.pdf accessed 20 October 2026.
Chief Electoral Officer, Delhi, ‘Revision of Schedule of Special Intensive Revision 2026’ (Press Note, 26 September 2026) https://ceodelhi.gov.in/PDFFolders/2026/Revision_of_Schedule_of_SIR26Sept2026.pdf accessed 20 October 2026.
Chief Electoral Officer, Goa, ‘Press Note’ (21 February 2026) https://ceogoa.nic.in/PDF/SIR2026/press-note-Final-publication.pdf accessed 20 October 2026.
Chief Electoral Officer, Kerala, ‘Final Electorate: SIR 2026’ (21 February 2026) https://www.ceo.kerala.gov.in/uploads/sir-2026/final_electorate_sir_2026.pdf accessed 20 October 2026.
Chief Electoral Officer, Maharashtra, ‘Press Note: Draft Electoral Roll SIR 2026’ (1 September 2026) https://ceoelection.maharashtra.gov.in/Downloads/PDF/Notification/Pressnotemarathi01.09.2026.pdf accessed 20 October 2026.
Chief Electoral Officer, Meghalaya, ‘Draft Electoral Roll, 2026 Published Across Meghalaya under Special Intensive Revision (SIR), 2026’ (Press Information Bureau, 5 August 2026) https://www.pib.gov.in/PressReleasePage.aspx?PRID=2295108 accessed 20 October 2026.
Chief Electoral Officer, Mizoram, ‘Publication of Final Electoral Roll for Special Intensive Revision, 2026’ (Press Note No H.11011/06/2026-CEO, 7 September 2026) https://ceo.mizoram.gov.in/storage/press-releases/September_2026/01M1XFVV4G2XCZA15M3W40KESR.pdf accessed 20 October 2026.
Chief Electoral Officer, Rajasthan, SIR 2026 Final Electors Summary (As on 21.02.2026) (21 February 2026) https://election.rajasthan.gov.in/Elector%20Statics/Fact%20Sheet%2021.02.2026.pdf accessed 20 October 2026.
Chief Electoral Officer, Uttarakhand, ‘Final Electoral Roll Published in Uttarakhand after SIR 2026’ (Press Note No 63/2026, 3 October 2026) https://cdnbbsr.s3waas.gov.in/s3cbf8710b43df3f2c1553e649403426df/uploads/2026/10/202610031173939115.pdf accessed 20 October 2026.
Chief Electoral Officer, West Bengal, ‘Press Note No CEO/PN-05/2026’ (28 February 2026) https://wbceo.wb.gov.in/Downloads/News/Final%20Press%20Note%20CEO-PN-05-2026.pdf accessed 20 October 2026.
Chief Electoral Officer, West Bengal, ‘SIR 2026’ https://ceowestbengal.wb.gov.in/SIR accessed 20 October 2026.
Election Commission of India, ‘Daily Bulletin: 3:00 PM of 10th December 2025 (Graph Annexed)’ (Press Information Bureau, 10 December 2025) https://www.pib.gov.in/PressReleasePage.aspx?PRID=2201464 accessed 20 October 2026.
Election Commission of India, ‘Press Note Dated 15 March 2026 for Assam, Kerala, West Bengal, Tamil Nadu and Puducherry’ (15 March 2026) https://static.pib.gov.in/WriteReadData/specificdocs/documents/2026/mar/doc2026315824801.pdf accessed 20 October 2026.
Election Commission of India, ‘Special Intensive Revision (SIR) of Electoral Rolls in Bihar Successfully Completed’ (Press Information Bureau, 30 September 2025) https://www.pib.gov.in/PressReleasePage.aspx?PRID=2173316 accessed 20 October 2026.
Election Commission of India, ‘Special Intensive Revision (SIR) Phase-II Begins in 9 States and 3 UTs’ (Press Information Bureau, 4 November 2025) https://www.pib.gov.in/PressReleasePage.aspx?PRID=2186480 accessed 20 October 2026.
Election Commission of India, ‘Special Intensive Revision – Phase III’ (Press Information Bureau, 14 May 2026) https://www.pib.gov.in/newsite/archiveContent.aspx?lang=2®=48&relid=288845 accessed 20 October 2026.
Government of Sikkim, ‘Final Electoral Roll under SIR 2026 Published in Sikkim’ (7 September 2026) https://www.sikkim.gov.in/media/news-announcement/news-info?name=Final+Electoral+Roll+under+SIR+2026+Published+in+Sikkim accessed 20 October 2026.
B. Newspaper Articles and Media Reports
Dev A, ‘Karnataka SIR: Draft Electoral Roll Publication Deferred to August 24’ (Hindustan Times, 7 August 2026) https://www.hindustantimes.com/india-news/karnataka-sir-draft-electoral-roll-publication-deferred-to-august-24-101786112229935.html accessed 20 October 2026.
Khanna B, ‘Karnataka Has 4.46 Crore Voters, Reveals Draft Electoral Roll Data’ (The New Indian Express, 25 August 2026) https://www.newindianexpress.com/states/karnataka/2026/Aug/25/karnataka-has-446-crore-voters-reveals-draft-electoral-roll-data accessed 20 October 2026.
‘Punjab’s Final Electoral List Out, Voter Count Drops by 13.6 Lakh after SIR’ (India Today, 12 October 2026) https://www.indiatoday.in/india/story/punjab-final-electoral-roll-drops-13-lakh-after-sir-voter-count-at-2-crore-3015835-2026-10-12 accessed 20 October 2026.
Raj A, ‘How Much Did India’s Voter Rolls Shrink after SIR? State-Wise Analysis’ (Business Standard, 20 July 2026) https://www.business-standard.com/india-news/how-much-did-india-s-voter-rolls-shrink-after-sir-state-wise-analysis-126072000267_1.html accessed 20 October 2026.