SIR: Glaring oddities in Maharashtra’s deletion patterns

Why in the News

The draft roll published on 31 August under the Special Intensive Revision (SIR) of electoral rolls has dropped 2.07 crore electors from Maharashtra’s 288 Assembly Constituencies (ACs), which is 21.14 per cent of the 9.79 crore electors carried on the rolls before the revision. That single State figure is larger than the entire electorate of Chhattisgarh, and larger than the 2.04 crore deletions recorded during the same revision in Uttar Pradesh, a State with 13.4 crore electors. The Election Commission of India (EC) has not released the Elector to Population (EP) ratio for any State during the revision, though its own Manual on Electoral Rolls requires that disclosure during and after every revision. The contest is therefore not over whether a roll needed cleaning but over whether a deletion of one in five electors, justified largely by two blanket reasons, can be checked at all while the one ratio designed to check it is withheld.

What is the Special Intensive Revision of electoral rolls?

1. Rebuilding the roll by fresh enumeration: An intensive revision rebuilds the electoral roll through a fresh house to house enumeration rather than adding and removing names against the existing roll, so every elector has to be re accounted for.
2. The enumeration phase and the draft roll: The enumeration phase collects forms from electors, and the draft roll published at the end of it shows who survived enumeration and who was deleted, with a reason recorded against each deletion.

What is the Elector to Population ratio?

1. What the ratio measures: The Elector to Population ratio expresses enrolled electors as a share of the population old enough to vote in the same area, so it converts a raw roll count into a figure that can be compared across constituencies of different sizes.
2. The benchmark it is read against: Official projections for the 18 plus age group give an expected number of electors for an area, which supplies a comparison derived independently of the roll being tested.
3. What a low ratio signals: A roll sitting far below the ratio implied by those projections is evidence of under enrolment rather than of a cleaner roll, since removing duplicate and deceased entries does not pull the total below the eligible population.
4. The disclosure obligation attached to it: The Election Commission’s Manual on Electoral Rolls requires the ratio to be disclosed during and after every revision, which makes it the audit built into the revision rather than an external check on it.

How large is the deletion, measured against the roll itself?

1. Names deleted and names retained: 2,06,88,487 names were deleted from a pre revision roll of 9,78,54,049, leaving 7,71,65,562 electors in the draft roll.
2. Comparison with Uttar Pradesh: The deletion exceeds Uttar Pradesh’s, and the draft rolls of Maharashtra, Karnataka and Delhi combined hold fewer electors than Uttar Pradesh alone.
3. The gap against projected population: Set against a projected voter population of 9,65,00,000, the draft roll is short by 1,93,34,438 electors, about 20 per cent.
4. The Technical Group on Population Projections: That population estimate is the 18 plus age group projection from the Government of India’s Technical Group on Population Projections, not an independent construction.
5. Deletion size against shortfall size: The shortfall against projected population is roughly the same size as the deletion, so the revision has removed almost exactly the number of electors that a demographic estimate says the roll should have retained.

Where are the deletions concentrated?

1. Urban concentration of the deletions: About 75 per cent of all deletions occurred in the 147 constituencies classified as highly or moderately urban.
2. The clustering by district: Almost all of the heaviest deleting constituencies lie in and around Mumbai, Pune, Thane and Nashik.
3. Constituencies at the extremes: 58 ACs lost at least a third of their names and 25 ACs lost 40 per cent or more, with Bhiwandi East in Thane the highest at 49.1 per cent.
4. Night time light classification of constituencies: Constituencies were sorted into urban and rural using night time light intensity from the Socio economic High resolution Rural Urban Geographic Platform for India (SHRUG), a public geographic database maintained by the Development Data Lab, read off 2023 satellite data.
5. Akole, the lowest deleting constituency: Akole in Ahmednagar recorded the lowest share of deletions against its pre revision roll.

What do the recorded reasons for deletion show?

1. Permanently Shifted and Untraceable or Absent: Permanently Shifted and Untraceable or Absent were applied as blanket categories across large numbers of names.
2. Kalyan Rural in Thane: Kalyan Rural in Thane lost 2,52,247 names, 46 per cent of its roll, of which 2.39 lakh, or 96.5 per cent, were marked shifted or absent, against only 4,882 marked Deceased.
3. The 85 constituency cluster: In 85 ACs shifted and absent accounted for between 75 per cent and 96.5 per cent of all deletions. Those 85 constituencies held 3.21 crore electors before the revision and lost 1.17 crore, of which 97.9 lakh were marked shifted or absent.
4. The Manual’s bar on deleting an absentee: The EC’s own manual states that an absentee voter cannot be deleted, because that elector continues to be an ordinary resident of the place where enrolled.
5. The Chief Electoral Officer’s definition of Absent: An explanatory note issued by the Chief Electoral Officer of Maharashtra defines Absent as meaning that the elector was unavailable, which is a condition of the enumerator’s visit rather than a finding about residence.

Why does the variation between similar constituencies matter?

1. Pune’s 21 constituencies: In Pune’s 21 ACs, the share of deletions marked Absent ranged from 0.4 per cent to 88.3 per cent.
2. Nashik, Palghar, Aurangabad and Nagpur: Nashik’s 15 ACs ranged from 8.1 per cent to 86.5 per cent, Palghar’s six from 3.3 per cent to 85.5 per cent, Aurangabad’s nine from 7.3 per cent to 67.3 per cent and Nagpur’s 12 from 4.4 per cent to 59.7 per cent.
3. Maval and Hadapsar: Maval marked just 322 names, 0.4 per cent, as Absent while marking over 77,000 as shifted, and Hadapsar marked 86 per cent of its 2.98 lakh deleted voters as absent.
4. The Deceased and Duplicate shares: Across a sample of constituencies the share of deletions recorded as Deceased runs from 1.9 per cent to nearly 45 per cent, and the share recorded as Duplicate from 1.3 per cent to about 25 per cent.
5. Enumerator practice as the source of the spread: Constituencies of similar urban or rural character, inside one district, were processed under reasons applied arbitrarily and interchangeably, which points to the enumerator’s practice rather than to any underlying difference in the electors.

Challenges to the Special Intensive Revision of electoral rolls

1. Deletion is the default outcome of a re enumeration: Rebuilding a roll from scratch places the burden of proof on the elector, so anyone missed at the door is removed rather than retained. Eg. In 85 Maharashtra constituencies, shifted and absent together accounted for up to 96.5 per cent of all deletions.
   The Fix: Require a second visit and a recorded notice to the elector’s address before any name is deleted for absence, so a missed visit cannot by itself end an enrolment.
2. The categories used are not verifiable after the fact: Shifted and Untraceable record what the enumerator could not find, not a fact about the elector, so a wrongly recorded deletion leaves no trace to audit. Eg. The Chief Electoral Officer’s own note defines Absent as the elector being unavailable.
   The Fix: Publish the booth level list of deletions with the reason and the date of the enumerator’s visit, so a deletion can be contested against a record rather than against a category.
3. Withholding the mandated ratio removes the only aggregate check: The Elector to Population ratio is what tells a reader whether a revision corrected the roll or thinned it, and no substitute measure exists. Eg. The estimated ratio for Maharashtra shows the draft roll short by 1.92 crore electors.
   The Fix: Release the ratio constituency by constituency at draft publication, as the Manual on Electoral Rolls already requires, rather than after the claims and objections period closes.
4. Urban electors are structurally easier to delete: Rented accommodation, migrant work and multi storey buildings make an occupant harder to locate on a single visit than a settled rural household. Eg. About 75 per cent of Maharashtra’s deletions fell in the 147 highly or moderately urban constituencies.
   The Fix: Allow urban electors to complete enumeration through a self service digital submission verified against an existing identity record, so presence at the door is not the only route to staying on the roll.
5. The claims and objections window assumes the deleted elector knows: A person removed from the roll usually discovers it at the polling station, by which time the remedy has expired. Eg. Deletions in Maharashtra ran to 2.07 crore names, which no individual notice system currently covers.
   The Fix: Send a mandatory individual intimation by post and message to every deleted elector at the address on record, with the reason and the deadline for restoration stated.
6. Roll revision decides delimitation and reservation downstream: Electoral rolls feed constituency sizes, reserved seat calculations and future revision baselines, so an error does not stay inside one election. Eg. The Maharashtra draft roll now sits 20 per cent below the projected 18 plus population used by official population projections.
   The Fix: Freeze the revised roll as a baseline only after an independent statistical audit against the Technical Group’s population projections has been published.

Conclusion

The revision has produced a roll that is smaller than the population estimate by almost exactly the number of names it removed, and the reasons recorded for those removals swing from near zero to near universal between constituencies that are otherwise alike. What settles the dispute is not another analysis of the draft but a single disclosure the revising authority is already obliged to make. The marker to watch is whether the Elector to Population ratio is published constituency by constituency before the claims and objections period closes, since restoration after the final roll is published is a different and far harder remedy.

Matching Previous Year Question

“[2026, GS2, 10 marks] Is the right to vote a fundamental right? Discuss the position of the Election Commission of India while undertaking the revision of electoral rolls. Can it also examine the question of citizenship of voters?”