Election Commission’s Data on Bengal SIR Appeals Is Designed to Confuse. Here’s How.

This is the second in a series of articles on the tribunal hearings on Special Intensive Revision (SIR) of electoral rolls in West Bengal.

Read part one, here, part three, here, and part four, here.

Kolkata: The Special Intensive Revision (SIR) of West Bengal’s electoral roll marked “27,16,393 people as not eligible” in the Supreme Court-mandated judicial review. Every one of those names was struck off the roll and published in a list of deleted electors.

In March 2026, the Supreme Court directed the Commission to set up Appellate Tribunals so that anyone struck off could appeal and ask to be put back on the voter list.

The Election Commission says 75,443 of the 27,16,393 (or 2.78%) electors struck off West Bengal’s roll have been put back by the Appellate Tribunals. The figure comes from a 28-page Right to Information reply issued on August 17 to Isha Khan Choudhury, the Congress MP for Dakshin Malda.

Read closely, the reply does not support that figure. Nor does it support several other data points provided in the same 28 pages. The Wire raises the many questions raised by the Election Commission’s own data.

1. One word is being used to mean two opposite things

The tribunals in West Bengal were expected to decide on the restoration of electors who lost their voting rights after being marked “not eligible” during the SIR. The ECI, though, says it has received two kinds of appeals: (1) Inclusion requests, from those struck off the lists; and (2) Exclusion requests, demands from people to delete some names of those on the list.

The first is someone defending their own vote. The second is an individual or an authority seeking to end somebody else’s right to vote.

The Commission has added both together and published one number. Its own footnote to the table says so, recording that the total covers appeals against exclusion from the roll and objections against inclusion in it. Nowhere in the 28 pages of the RTI reply are separate figures given for either.

The same obfuscation is clear when looking at electors who have been included.

When a tribunal decides in the elector’s favour, the EC has marked that under one category – even though that could either mean that a voter has been included based on their own appeal application, or that a voter has not been struck off despite someone else’s objection application!

So when the EC’s reply says 75,443 people were ‘included’, this does not mean that 75,443 people were put back on the voter list. Some of these people were on the list all along. How many? Only the EC knows.

2. Six constituencies ‘put back’ more people than they removed?

If every restoration were a struck-off elector getting her name back, no constituency could restore more people than it had seen removed in the first place. But six did – confirming that the EC’s data is combining categories in a way that makes the data difficult to truly analyse.

In Manbazar in Purulia, 71 electors were recorded as ‘not eligible’. The tribunal has restored 525 people there. Amta in Howrah had 393 ‘not eligible’, but restored 1,495. Sonarpur Uttar in South 24 Parganas found 4,270 ‘not eligible’ and restored 7,022. Kashipur, Para and Binpur show the same pattern, on a smaller scale. Across the six constituencies, a total of 4,874 restorations are in excess over the number of people who had been found ‘not eligible’ to begin with.

These six constituencies clearly establish that the ‘restoration’ column is not counting only restorations.

3. EC’s reply gives two different numbers of appeals for the same districts

The RTI reply provides data constituency-wise as well as district-wise. But the two sets do not show comparable numbers for appeals.

Under the head of Malda in the district-wise table there are 5,31,149 appeals. But all the assembly-level numbers for the district given in the assembly-wise table add up to 6,85,341 appeals. Similarly, Murshidabad has recorded 7,47,305 appeals in the district-wise table, but assembly-level appeals show 4,27,358 filings.

Kalimpong district consists of a single constituency. But it shows 3,638 appeals in the district-wise table and 44,698 appeals in Kalimpong constituency in the assembly-wise table.

The other numbers, such as those covering electors placed under adjudication, reconcile between their two versions, when classified as districts or constituencies. But appeals are not reconciliable.

4. The tribunals’ own published lists do not match the reply

West Bengal voted in two phases. Before each phase, the tribunals published lists naming electors who had been cleared and could vote. For the first phase, the tribunals examined 138 appeals and allowed 136. For the second, the figure was 1,468, on a list closed on April 27. That is 1,604 people named in total.

Set those lists against the reply, which records restorations as of August 7. In 11 constituencies the list of voters cleared by tribunals to vote, contain more people than the RTI reply says were restored. For example, in Chowrangee, the tribunal cleared 322 people to vote just before the phase two of the polling. But the RTI reply records just 26.

5. One district logged 25 filings for every person it removed

Bankura found 6,533 electors not eligible to vote across its 12 constituencies and recorded 1,61,434 appeal filings. That is close to 25 filings for every person it removed. Onda found 114 ‘not eligible’ but recorded 10,489 appeal filings.

The appeal totals appear to be carrying multiple objections rather than adhering to the principle of one deletion-one objection. The 38,10,620 figure cannot be read as a measure of how many people sought their vote back.

What is not clear

1. How many electors struck off the rolls have appealed, because appeals and objections are counted as one category.
2. How many struck off voters have been put back, because restorations and successful defences against objections are counted as one. Six constituencies make this crystal clear.
3. How many appeals each district has received, because the reply gives two different answers.
4. How many electors were restored in time to vote, beyond the 1,604 the EC has named, because neither the RTI reply nor the lists put a date against individual orders.

What can be stated is that at least 7,21,036 names on the rolls found ineligible did not appeal and were struck off by the judicial officer. We reported this in part 1 of this series. But data provided by the Commission is not helping us call out exactly how many of the people it removed have been put back on the rolls.

The Supreme Court has directed the Commission to file a category-wise account of the appeals before the tribunals, separating those brought by electors seeking their own names back from those brought by authorities or third parties against someone’s inclusion. If the EC complies, that would answer the first two questions of how many electors struck off the rolls have appealed and how many have been put back on the rolls by the EC. It would not touch upon the third question of how many appeals each district has received.

Sources and method: Appellate Tribunal Appeal Report and Individuals under Adjudication Report as on 07.08.2026, with district and constituency-wise elector data, furnished by the Office of the Chief Electoral Officer, West Bengal under the Right to Information Act vide Memo No. 8551/Home (Elec) dated 17.08.2026, on an application by Isha Khan Choudhury, MP for Dakshin Malda. Constituency-wise figures for electors found not eligible were compiled from the Commission’s published deleted-elector lists.

This story has been updated and republished at 5:48 pm on Tuesday, September 1. 

The Wire is now on WhatsApp. Follow our channel for sharp analysis and opinions on the latest developments.