SIR: deletions have increased in Phase 3 States/UTS

1. At a Glance

• Special Intensive Revision (SIR) of electoral rolls is a nationwide, phase-wise re-verification exercise by the Election Commission of India (ECI) to purge ineligible/duplicate/deceased entries and ensure no eligible citizen is excluded [1].
• Phase 3, covering 19 States/UTs, is showing a sharply higher deletion rate (17.1%) than Phases 1–2 combined (12.3%) — a live case study in electoral administration, data governance, and franchise rights [4].
• Relevant for Polity/Governance (GS-II) and as a current-affairs anchor for larger debates on voter list integrity, migration, and disenfranchisement risk.

2. Why in the News

• Of the 19 Phase-3 States/UTs, 17 have released draft SIR rolls (Nagaland and Tripura pending) [4].
• These 17 had 36.1 crore voters pre-SIR, falling to 29.9 crore in draft rolls — a deletion of 6.15 crore names (17.1%) [4].
• This is 4.8 percentage points higher than the 12.8% deletion rate recorded during the enumeration phase of the 13 States/UTs in Phases 1 and 2 (including Bihar) [4].
• Those 13 States/UTs had 58.9 crore voters pre-SIR, reduced to 51.7 crore in draft rolls via deletion of 7.22 crore names (12.3%) [4].

3. Background & Evolution

• ECI announced SIR in June 2025, citing "rapid urbanisation" and "frequent migration" over the preceding two decades as rationale for a fresh, intensive (not routine) roll revision [3].
• Phase 1 (Bihar): first State to undergo SIR; resulted in a net reduction of ~47 lakh names [3].
• Phase 2: covered 9 States and 3 UTs; over 50.99 crore electors enumerated, with net deletion of ~5.18 crore names (10.2%), final rolls settling at ~45.81 crore; deletions included ~66.88 lakh deceased electors, duplicates, and absentee/shifted voters [3].
• Phase 3: launched per ECI order dated 14.05.2026; enumeration began 30.05.2026 in Odisha, Mizoram, Sikkim and Manipur, later extending to 19 States/UTs including Delhi [1][2].
• Deletion trend across phases: 12.3% (Phases 1–2 enumeration) → 10.5% (final rolls) → 17.1% (Phase 3 enumeration, draft rolls) [4].

4. Core Static Facts

Item	Detail
Implementing body	Election Commission of India (ECI)
Field-level functionary	Booth Level Officers (BLOs), via house-to-house enumeration [1]
Phase 1	Bihar; net deletion ~47 lakh names [3]
Phase 2	9 States + 3 UTs; 50.99 crore electors → 45.81 crore; deletion 5.18 crore (10.2%) [3]
Phase 3	19 States/UTs; enumeration started 30.05.2026 (Odisha, Mizoram, Sikkim, Manipur first) [1]
Phase 3 draft-roll deletions	6.15 crore of 36.1 crore (17.1%), across 17 of 19 released States/UTs [4]
Phases 1–2 combined enumeration deletions	7.22 crore of 58.9 crore (12.3%) [4]
Phases 1–2 final-roll deletion rate	10.5% (down marginally from 12.3%) [4]
Pending draft rolls (Phase 3)	Nagaland, Tripura [4]
Objective stated by ECI	"No eligible citizen is left out while no ineligible person is included" [1]

5. Multi-Dimensional Analysis

Legal/Constitutional

• SIR draws on ECI's constitutional mandate under Article 324 (superintendence, direction and control of electoral rolls) and the Representation of the People Act, 1950 (roll preparation/revision provisions).
• Draft-roll deletions are provisional; affected voters can file claims/objections before final publication — a due-process safeguard cited by ECI [1].

Administrative

• Deletion criteria include deceased electors, duplicates, and absentee/shifted voters [3].
• Scale disparity: Phase 3 covers a much larger and more heterogeneous set of States/UTs in a compressed timeline, which may explain both efficiency gains and error-prone over-deletion.
• ECI has revised the SIR schedule multiple times (e.g., extending Phase timelines by a week) reflecting operational strain [1].

Social/Ethical

• A jump from ~12–13% to ~17% deletion raises concerns about potential disenfranchisement of genuine voters, especially migrants, urban poor, and those unable to furnish documentary proof.
• Civil society and opposition scrutiny (as tracked in Phase 1 Bihar) centers on whether deletions are actually "excess" rather than genuine ineligibility.

Governance/Transparency

• ECI's phase-wise data disclosure (numbers before/after SIR) allows public and media cross-verification — an accountability mechanism, though final-roll reconciliation is still pending for Phase 3.

6. Recent Developments (last 12-18 months)

• June 2025: ECI announces SIR exercise citing urbanisation/migration rationale [3].
• Phase 1 (2025): Bihar SIR conducted and completed; net deletion ~47 lakh [3].
• Phase 2 (Nov 2025 onward): Launched across 9 States and 3 UTs [1].
• 14.05.2026: ECI order for Phase 3 SIR [1].
• 30.05.2026: Enumeration begins in Odisha, Mizoram, Sikkim, Manipur, later extended to all 19 Phase-3 States/UTs [1].
• 16.09.2026 (reporting date): The Hindu reports 17 of 19 Phase-3 States/UTs have released draft rolls showing 17.1% deletions, markedly higher than the 12.8%/12.3% seen in Phases 1–2 [4].

7. Prelims Hooks

• SIR stands for Special Intensive Revision of electoral rolls, conducted by the ECI [1].
• Phase 1 of SIR was conducted first in Bihar [3].
• Phase 2 covered 9 States and 3 Union Territories [1].
• Phase 3 covers 19 States/UTs, including Delhi (NCT) [1][2].
• Phase 3 enumeration formally began on 30 May 2026 in Odisha, Mizoram, Sikkim, and Manipur [1].
• As of the draft-roll stage, 17 of 19 Phase-3 States/UTs have published rolls; Nagaland and Tripura are pending [4].
• Phase 3 draft rolls show a 17.1% deletion rate (6.15 crore of 36.1 crore voters) [4].
• Phases 1–2 combined enumeration-phase deletion rate was 12.8% (per composite reporting) / 12.3% for the 13-State/UT set including Bihar [4].
• Phases 1–2 final-roll deletion rate settled at 10.5%, down from the enumeration-phase figure [4].
• Field verification for SIR deletions is carried out by Booth Level Officers (BLOs) [1].
• Phase 2 alone recorded ~66.88 lakh deceased electors removed [3].
• ECI's stated twin objective: exclude no eligible citizen, include no ineligible person [1].
• ECI has revised the SIR schedule on more than one occasion, extending timelines [1].
• Phase 2's net deletion figure was ~5.18 crore names (10.2%), bringing rolls to ~45.81 crore [3].

8. Why 17.1% Is a Draft Figure, Not a Disenfranchisement Count

• The draft roll is the mid-point of the process, not its output — deletions at draft stage are provisional; the statutory claims-and-objections window (Forms 6, 7 and 8 before the ERO/AERO) sits between draft and final publication [1][8].
• The Phases 1–2 gap quantifies the restoration effect — the same States fell from 12.3% deletions at enumeration to 10.5% in final rolls, i.e. roughly 1.8 percentage points of a 58.9-crore base were put back after objections [4]. Applied naively to Phase 3, that would shave ~1.1 crore off the 6.15 crore headline — which is why quoting 17.1% as a final loss of franchise is a factual error in an answer script.
• But restoration is opt-in, and that is the real defect — the burden of noticing a deletion, locating the draft roll, filing Form 6 and, if refused, appealing to the DM and then the CEO falls on the deleted elector [1]. A voter absent during BLO enumeration is, almost by construction, also absent during the objection window.
• The 10.5% final figure therefore measures who contested, not who was wrongly deleted — no published metric separates correctly-deleted-and-silent from wrongly-deleted-and-silent [4].

9. The Absentee Category Is Where the Franchise Risk Sits

• Three deletion grounds, three completely different risk profiles — deceased, duplicate, and absentee/shifted are reported as one aggregate [3]. Deceased and duplicate deletions are pure roll hygiene with near-zero franchise cost; absentee/shifted deletions are the only category where a living, eligible citizen loses a vote.
• Phase 2 is the only phase with a disaggregated number, and it is revealing — ~66.88 lakh deceased out of ~5.18 crore net deletions, i.e. barely 13% of deletions were the uncontroversial kind [3]. The remaining ~87% were duplicates and absentee/shifted — the contestable bucket.
• The untraceable rule converts absence into deletion — an elector is treated as untraceable only after a minimum of two-three BLO door-to-door attempts, with reasons recorded in the enumeration format under the Registration of Electors Rules, 1960 [1][7]. Two-three visits over a compressed enumeration window is a weak test for a circular migrant, a construction worker in a labour camp, or a student outside the district.
• The rationale and the casualty are the same population — ECI justified SIR by "rapid urbanisation" and "frequent migration" [3]. The mechanism that finds migration in the data is precisely the mechanism that deletes migrants, and nothing in the published figures distinguishes a genuine double-entry from a single voter enumerated at neither end.
• Phase 3 raises the stakes because of who it covers — 19 States/UTs including Delhi (NCT), a net in-migration destination [1][2], where shifted-voter volume is structurally higher than in the Phase 1–2 set.

10. Why Phase 3 and Phases 1–2 Are Not a Like-for-Like Comparison

• Different denominators, different composition — Phase 3's 17.1% is computed on 17 of 19 States/UTs with 36.1 crore electors; the 12.3% is a 13-unit, 58.9-crore aggregate dominated by a few large States [4]. Two aggregates with different State mixes cannot be read as a time series of ECI stringency.
• Different rules, not just different states — the exercise has been applied with varying documentary and procedural requirements across the three phases [3], so a rising rate may reflect a changed threshold rather than a dirtier roll.
• Different timelines — Bihar ran as a standalone Phase 1 [3]; Phase 3 pushed 19 units through a single window opened by the order of 14.05.2026, with ECI itself revising the schedule more than once, including for 6 States/UT [1][9]. Compressed enumeration raises the share of electors the BLO simply could not reach.
• Stage mismatch is the trap most answers fall into — the honest comparison is Phase 3 draft (17.1%) against Phases 1–2 draft (12.3%), never against the 10.5% final [4]. Phase 3's final rate is not yet knowable.

11. The Strongest Case for the ECI — and Where It Stops

• The case: electoral rolls in India accumulate error monotonically between intensive revisions — the dead are rarely reported, movers rarely apply for deletion at origin, and summary revision only adds. On that logic a high deletion rate is evidence the exercise is working, not failing; 66.88 lakh deceased electors removed in Phase 2 alone were 66.88 lakh live opportunities for impersonation [3].
• The procedural case is also real — deletion requires recorded reasons under the Registration of Electors Rules, 1960, publication of a draft, a statutory objection period, an appellate route to the DM and onward to the CEO, and Booth Level Agents nominated by political parties to verify entries at the enumeration and objection stages [1][7]. Political parties are not passive observers here; the BLA channel gives every party a statutory presence at the booth.
• Where it stops: the safeguards are all reactive. Every one of them triggers on someone noticing and contesting. None of them independently audits the deletions nobody contested — which, on a 6.15 crore base, is where the entire question lies [4].
• And the asymmetry of error is not symmetric in cost — a retained ineligible name is one fraudulent vote that still requires someone to physically impersonate; a wrongly deleted name is one citizen certainly disenfranchised for that cycle. ECI's own twin objective treats the two as equivalent [1]; they are not.

12. What Would Make the Next Phase Independently Auditable

• ECI: publish reason-wise deletion splits for every State at draft stage — the deceased/duplicate/absentee breakdown already exists internally and was released for Phase 2 [3]. Publishing it per State converts an alarming aggregate into a diagnosable one and removes the single largest source of contested interpretation.
• ECI: report the restoration rate, not just the final rate — the Phases 1–2 movement from 12.3% to 10.5% is currently inferable only by subtraction [4]. A published count of names restored on claim, and of appeals allowed by DMs and CEOs, would directly measure whether the reactive safeguards work.
• ECI: extend the Special Roll Observer mechanism to all Phase 3 States — Observers were already deployed for SIR in major States [5]; an independent supervisory layer over deletion decisions is the one safeguard that does not require the voter to act first.
• Political parties: use the BLA channel at enumeration, not at litigation — BLAs can flag discrepancies while entries are still being made [1]. Objections raised after draft publication start from a deleted entry and a shorter clock.
• Suo motu restoration for the deceased-negative cases — where a deletion is grounded only in "absent at time of visit" and no positive evidence of ineligibility exists, the default should be retention pending inquiry, consistent with ECI's own position that no deletion is made without due inquiry and a fair hearing [1][8].

13. Anchors for Answers

• Data: Phase 3 draft rolls — 6.15 crore deletions from 36.1 crore electors across 17 of 19 States/UTs (17.1%), against 7.22 crore from 58.9 crore (12.3%) in Phases 1–2 enumeration [4]
• Data: Phases 1–2 deletion rate fell from 12.3% at enumeration to 10.5% in final rolls — the measurable restoration effect of the claims-and-objections stage [4]
• Data: Phase 2 removed ~66.88 lakh deceased electors out of ~5.18 crore net deletions — only ~13% of deletions were the uncontestable kind [3]
• Law/Case: Article 324 (ECI's superintendence over rolls); Article 326 (adult suffrage); Sections 16 and 19, Representation of the People Act, 1950 (disqualification and qualification for registration) [8]
• Law/Case: Registration of Electors Rules, 1960 — prescribes the enumeration format, recorded reasons for deletion, and Forms 6, 7 and 8 for claims and objections [1][7]
• Scheme: Special Roll Observers deployed by ECI for SIR in major States — the only supervisory safeguard not dependent on the voter filing first [5]
• Scheme: Booth Level Agents nominated by political parties — statutory party presence at enumeration and objection stages [1]
• Comparison: Phase 2 (9 States + 3 UTs) vs Phase 3 (19 States/UTs incl. Delhi NCT) — differing State composition, documentary requirements and timelines make the cross-phase rate comparison unsound [1][3]

14. Mains Relevance

• GS-II: Indian Polity — "Salient features of the Representation of the People's Act"; Election Commission's powers, functions; issues related to electoral reforms.
• GS-II: Governance — transparency, accountability, and citizen rights in state-run administrative exercises.
• Possible question stems: 1. Examine the constitutional and statutory basis of the Election Commission's power to conduct Special Intensive Revision of electoral rolls. Discuss the safeguards against wrongful deletion of genuine voters. 2. The rising deletion rates across successive phases of the SIR raise concerns of disenfranchisement versus roll accuracy. Critically analyse. 3. Discuss the administrative challenges in balancing electoral roll purification with protection of voting rights, in light of the recent Special Intensive Revision exercise.

15. Related Topics to Study Next

• Representation of the People Act, 1950 & 1951 — statutory backbone for roll preparation and elections.
• Article 324 & Election Commission of India — constitutional mandate and independence.
• Delimitation of Constituencies — related exercise affecting representation, also flagged as a current topic.
• NRC/NPR debates — comparative exercise on documentary proof of citizenship/residence, similar disenfranchisement concerns.
• One Nation, One Election — broader electoral reform context.
• Voter ID / EPIC and Aadhaar linkage debates — related to deduplication rationale behind SIR.
• Migration and urbanisation trends in India (Census/Economic Survey) — cited rationale for SIR.

16. Common Errors / Trap Areas

• Do not confuse SIR (Special Intensive Revision) with the routine Summary Revision of electoral rolls conducted periodically — SIR is a special, house-to-house, document-verification-based exercise.
• Do not assume Bihar was part of "Phase 3" — Bihar was Phase 1, standalone.
• Distinguish enumeration-phase/draft-roll deletion rates from final-roll deletion rates — the final figure for Phases 1–2 (10.5%) was notably lower than the enumeration-stage figure (12.3–12.8%); Phase 3's final numbers are not yet out and may similarly moderate.
• Do not conflate deletion percentages with deletion counts — Phase 3's 17.1% is a higher rate but on a smaller voter base (36.1 crore) than Phases 1–2's 58.9 crore base.
• Phase 3 covers 19 States/UTs — some sources cite "22 states" for initial announcements; verify against final ECI notification count of States/UTs actually enumerated.