This is the fourth of a series on how the adjudication process of the Special Intensive Revision in West Bengal continues in Bengal. Read part one, here, part two, here, part three, here, and part five, here.Kolkata: Samserganj in Bengal’s Murshidabad has 80,234 people waiting to find out whether they will be on the electoral rolls since the special intensive revision or SIR. The tribunal that exists to hear those appeals has decided 12 of the appeals by the 80,234 people. The constituency was decided in the April assembly polls of this year by 7,587 votes.What was done to Samserganj before the election has already been reported by The Wire. Of a draft roll of 2,36,040 electors, 1,08,400 or 45.9% of the constituency were flagged, to be placed under adjudication. This was despite the fact that only 2,459 of the 1,08,400 – under 1% – had failed to match themselves to the 2002 rolls which the special intensive revision used as its base. As much as 68.98% of the voter list was struck off.On the early morning of April 22, the day before Samserganj voted, the Election Commission published the appellate tribunal supplementary lists for all 247 of its polling stations. The tribunals had cleared one name in Samserganj. By August 7, three and a half months later, the tribunal had restored another nine people in Samserganj and refused three.Draft roll, adjudication and unmapped figures as reported by The Wire on 22 April 2026, at https://thewire.in/government/west-bengal-samserganj-sir-exclusions-muslim-voters-tmc-congress-bjp-one-thiird. Tribunal decisions from the Appellate Tribunal Appeal Report as on 07.08.2026, Office of the Chief Electoral Officer, West Bengal, RTI Memo No. 8551/Home (Elec) dated 17.08.2026. The Commission’s adjudication return records 74,775 marked not eligible, while its published list of those electors carries 71,985 names.Samserganj is the sharpest instance of a pattern that runs through 49 West Bengal constituencies, and it describes what has happened to the remedy the Supreme Court ordered when it became clear how many people the revision had removed.In those 49 seats, the number of electors struck off the rolls was larger than the gap between the winner and the runner-up. That was reported by The Wire in its analysis of West Bengal assembly election results. What has happened in the four months since is the subject of this article, and the answer is that almost nothing has.Appellate Tribunal Appeal Report and Individuals under Adjudication Report as on 07.08.2026, Office of the Chief Electoral Officer, West Bengal, RTI Memo No. 8551/Home (Elec) dated 17.08.2026. Community assigned by surname classification of the Commission’s published lists of electors marked not eligible, an estimate carrying an error rate that cannot be quantified. Migration and rural classifications are the Commission’s own.As of August 7, 2026, in the 49 constituencies, the number of unheard appeals was at 8,97,340. The Appellate Tribunals had decided 11,795 cases, which is 1.30% of the number of cases that were in front of them. This means that adjudicators in 33 of the 49 constituencies had decided less than 1% of the appeals.Farakka was won by 8,193 votes and there, 74,838 appeals are unheard. Harirampur in Dakshin Dinajpur was won by 1,986 and has 72,289 unheard appeals. Mothabari in Malda was won by 10,496 and has 58,349 unheard appeals. Raina in Purba Bardhaman was won by 834 votes and has decided only four appeals out of 33,313. In each of these seats, the number of people still waiting runs to several times the number of votes that settled the seat.Appellate Tribunal Appeal Report and Individuals under Adjudication Report as on 07.08.2026, Office of the Chief Electoral Officer, West Bengal, RTI Memo No. 8551/Home (Elec) dated 17.08.2026. Community assigned by surname classification of the Commission’s published lists of electors marked not eligible, an estimate carrying an error rate that cannot be quantified. Migration and rural classifications are the Commission’s own.Who is waitingAcross West Bengal, 70.2% of the electors made ineligible to vote by the judicial review of under adjudicated in the SIR were Muslims. Across the 49 seats where the electors under adjudication are more than the winning margin, the figure of Muslim ouster is 85.4%. The 49 seats account for 9,02,667 removals, of whom approximately 7,70,000 were Muslim electors.The concentration is sharper still inside individual constituencies. Samserganj’s published list of electors who are marked ‘not eligible’ carries 71,985 names, of which over 71,500 were Muslim, a share of 99.5%. Suti removed 37,829 Muslims, at 98.5%. Manikchak removed 23,720 Muslims at 97.8 percent. Farakka removed 36,100 Muslims at 94.5%.Appellate Tribunal Appeal Report and Individuals under Adjudication Report as on 07.08.2026, Office of the Chief Electoral Officer, West Bengal, RTI Memo No. 8551/Home (Elec) dated 17.08.2026. Community assigned by surname classification of the Commission’s published lists of electors marked not eligible, an estimate carrying an error rate that cannot be quantified. Migration and rural classifications are the Commission’s own.These are also, incidentally, the very constituencies that people leave in order to find work in the rest of India. Thirty-two of the 49 seats can be classified as out-migrant seats. These are mostly rural seats as well. This matters because the Election Commission’s remedy process requires an elector to discover that they have been removed, locate the relevant list, identify a ground nobody has informed them of, assemble documents and file an appeal which is in most cases online. A construction worker in Kerala or a migrant in Delhi is not well placed to do any of this, and the tribunal has no obligation to find these people.What the figures will not supportNone of this establishes that any seat was won wrongly. The Election Commission’s Right to Information reply of August 2026 does not record how any of these people would have voted, whether each appeal represents a distinct elector, or whether the names were in fact prevented from voting. Some will have voted and appealed afterwards, after having received summons to defend their names on the rolls.Appellate Tribunal Appeal Report and Individuals under Adjudication Report as on 07.08.2026, Office of the Chief Electoral Officer, West Bengal, RTI Memo No. 8551/Home (Elec) dated 17.08.2026. Community assigned by surname classification of the Commission’s published lists of electors marked not eligible, an estimate carrying an error rate that cannot be quantified. Migration and rural classifications are the Commission’s own.The 5,33,000 figure comes from applying the tribunals’ own success rate to the appeals they have not looked at yet. They have looked at 1.30%. Nobody knows whether the appeals still waiting resemble the ones already decided. If the tribunals worked through the easiest files first, the true rate could be lower and 5,33,000, too high. That assumption is the weakest part of our estimate, and the Commission is the only body that can replace it with a count.It also assumes the cases still waiting resemble those already decided, which may be generous, since the files reached first are not necessarily typical.The community figures are estimates. The Commission’s deleted-elector lists carry no column for community, so the shares here come from classifying surnames, a method with an error rate that cannot be quantified. The Commission’s restoration figures carry no community field either, which means nobody outside it can say what share of the 75,443 people already put back were Muslim.As this series has reported, an appeal from someone struck off is counted in the same total as an objection filed against someone still on the roll. That is why the 49 seats here are defined by removals rather than by appeals, and why every calculation above is capped at the number of people each constituency actually removed.The ceilingThe question will be asked regardless, so here is the answer. Suppose every restored Muslim elector had voted and none had voted for the BJP. This assumption is not extreme, even though the BJP won 18 constituencies where the Muslim population is above 30% due to a split in the anti-BJP vote in Muslim-majority seats.Appellate Tribunal Appeal Report and Individuals under Adjudication Report as on 07.08.2026, Office of the Chief Electoral Officer, West Bengal, RTI Memo No. 8551/Home (Elec) dated 17.08.2026. Community assigned by surname classification of the Commission’s published lists of electors marked not eligible, an estimate carrying an error rate that cannot be quantified. Migration and rural classifications are the Commission’s own.Twenty-five of the 49 seats where the margin is lesser than the appeals were won by the BJP, so to understand the impact of this, we need to look at those 25 that could change hands. Six of those 25 seats change hands on the strictest version, which shares the restored votes among the opposition parties in the proportions they actually polled and assumes only the exact number of people who voted on polling day this year. Give every vote to whoever came second and seven seats go from the BJP. Assume all voters turn out in all seats, and eight go from the BJP.A BJP tally of 207 becomes 199 at worst. The majority mark is 148. On the assumption most favourable to the argument, the government of West Bengal is unchanged.What remains is narrower and harder to answer. Nine-tenths of a million appeals sit unheard in the constituencies where the removals were already larger than the result. The people removed there were overwhelmingly Muslim and disproportionately migrant. And on the tribunals’ own record, most of them should never have been removed. The category-wise break-up which the Supreme Court has directed would settle most of it. Until that is filed, these figures mark the outer edges of what can be known. The count of unheard appeals is the most it could be, because some of them are objections rather than appeals. The projection of restorations is the most it could be, because the easiest cases might have been decided first. For our full coverage on the SIR and the issues it has brought to citizens, click here.Note: An earlier version of this piece claimed erroneously in the headline that the winning margin was more than removals in 49 seats. The mistake has been corrected in an update.