Between April 13 and September 10, West Bengal’s appellate tribunals rejected the appeals of 6,444 voters who had been struck off the electoral roll during the judicial review conducted in the state. Civic Research Group, an independent group of researchers, analysts and field contributors in West Bengal, shared the details of every one of those decisions with The Wire, which analysed them constituency by constituency.By a name-based estimate, four in five of the voters turned away – 80.2% – have Muslim names. In the 129 constituencies where these rejections happened, minorities make up about 22% of the electoral roll. In Hooghly, Muslim names account for 83% of rejections in seats where minorities are 15% of voters. In Khanakul alone, 831 appeals were rejected, 96.5% of them from voters with Muslim names.But the rejections are only a part of the story. The tribunals hear two kinds of appeals. Around 22 lakh were from voters struck off in judicial review who wanted to be reinstated. Around 16 lakh were filed by the Election Commission to remove voters the judicial officers had cleared.Of the 38.1 lakh appeals filed across the state, a little over 1 lakh had been decided in the appeals data. Nine in ten of those were restored to the roll. In Malda, where 6.85 lakh appeals were filed, just 0.15% had been decided. For most people who appealed, the question is not whether they won or lost, but whether anyone has heard them yet.The tribunals also look very different from one district to the next. The bench hearing Purba Bardhaman rejected 85% of the cases it decided. The benches for North 24 Parganas, Nadia and Dakshin Dinajpur rejected none.The 80.2% figure needs careful reading. It describes whose appeals for inclusion were rejected among those marked “not eligible” by judicial officers, not the 16 lakh appeals for exclusion by the ECI. Without data on the religious identity of every appellant, the data cannot reveal that the tribunals treated one community differently from another. The Wire sets out what would settle that question.How the tribunals work, and what the data isDuring the SIR, about 60 lakh voters in West Bengal were placed under judicial adjudication for “logical discrepancies”. Judicial officers deleted 27.16 lakh of them. The appellate tribunals, set up under the Supreme Court’s directions in Writ Petition (Civil) 1089/2025, are where those voters can challenge their removal.There are 19 tribunals, each headed by a retired high court judge, and each district’s appeals go to a named bench. All 19 were sitting in Kolkata, so voters from Malda or Cooch Behar had to travel hundreds of kilometres to be heard. The CEO also said that anyone cleared on appeal would only be re-enrolled for future elections.For this report, the voter-level list of 6,444 people whose appeals were rejected by the appellate tribunal has been analysed. Each record has the voter’s EPIC number, name in Bengali script, gender, part number and the time of the decision.The document does not record religion or caste. That information is estimated, both from the voters’ names and by using markers common to each community in Bengali. This works well for religion in most cases. For caste, the estimation is rougher because many surnames are shared across castes. Every religion and caste figure in this report is an estimate, not an official record.Who was rejectedOf the 6,444 voters whose appeals were rejected, 5,169 have Muslim names and 1,221 have Hindu names. Fifty-four could not be classified. In district after district, Muslim names make up a far larger share of the rejections than Muslims’ share in the local electorate.The pattern is sharpest at seat level. Five constituencies account for 44% of all rejections. In Khanakul (831 rejections), 96.5% are Muslim names, in a seat where minorities are about 20% of the roll. In Bolpur (808) the figure is 88%, against a minority share of 30%.In Raiganj (659) it is 92%, against 19% minorities on the rolls. In Sabang (234) it is 88%, against under 7%. Of the 51 seats with at least 20 rejections, 32 had 80% or more Muslim names among the rejections.This does not by itself mean the tribunals were harder on Muslim appellants. If most of the people who appealed were Muslim, most of the people rejected would be too. The adjudication deletions that led to these appeals were the heaviest in Muslim-majority districts, so that is plausible.What the numbers do show is that the burden of losing an appeal has fallen overwhelmingly on Muslim voters, including in districts where they are a small minority.The exceptions, and the womenDakshin Dinajpur runs the other way. Of its 215 rejections, 98% are Hindu names, and about 70% carry surnames usually counted as Scheduled Caste. Balurghat (63 rejections) had no Muslim names at all. Gangarampur (57), Tapan (39) and Kushmandi (25) were all over 90% Hindu.A handful of other seats also lean Hindu, like Cooch Behar Uttar (74% of 35), Uttarpara (76% of 59), Kharagpur Sadar (74% of 31) and Chunchura (62% of 85).Across the state, 445 of the 1,221 rejected Hindu names carry surnames counted as Scheduled Caste, 192 as general caste, 138 as Other Backward Classes (OBC) and 37 as Scheduled Tribe. The rest could not be placed. The Scheduled Caste figure is likely to be on the high side, since the classifier counts common surnames such as Das, Mondal, Biswas and Sardar as Scheduled Caste, although they are shared with other castes.Also read: In Bengal, SIR-Affected Voters Are Concentrated in Muslim-Majority, Migrant-Heavy and Competitive SeatsWomen make up 52.7% of those rejected, similar to the trend witnessed during the Absentee Shifted Deceased and Duplicate (ASDD) and judicial review phases. Appeals of 3,396 women, 3,047 men and one voter of the third gender were rejected by the tribunal. That is a little higher than their roughly 49% share of the final roll. The skew comes from Muslim names, where 54% of those rejected are women. Among Hindu names it is 48%. In Sabang, 63% of those rejected were women, and in Bolpur 56%. Raiganj went the other way, at 39%.How the rejections came, and how fastThe first rejections were recorded on April 13 this year. In that first month, they were split almost evenly between Hindu and Muslim names. From May onwards, Muslim names made up between 70% and 90% of each month’s rejections. May was the busiest month, with 1,599 objections.The decisions were often recorded within minutes of each other. In the same seat on the same day, the typical gap between one rejection and the next is two-and-a-half minutes, and almost one in five came within a minute of the previous one. On July 3, the tribunal recorded 85 rejections from Khanakul between 12:41 pm and 6:10 pm, about one every four minutes. Khanakul accounts for six of the ten seat-days with 40 or more rejections.These timestamps show when a decision was recorded, not how long a hearing lasted. Decisions may have been entered in batches after hearings, and the restorations the same bench made on those days are not on our list. Even so, the pace raises a fair question about how much time each appeal received.The backlogThe rejections are a small part of what has happened to the people who appealed. Most of them are still waiting. The appeals data records 38.1 lakh appeals filed across the state, and 82,782 decided, about 2.2%. Of those decided, 75,443 (91%) were restored to the roll and 7,339 (9%) were rejected.The districts where the most appeals were filed are the ones that have heard the fewest. Malda has 6.85 lakh appeals and has decided 1,027 of them. Murshidabad has decided 0.43% of 4.27 lakh, and North 24 Parganas 0.39% of 3.41 lakh. Ten districts, holding two-thirds of all appeals, have decided less than 1%. By contrast, Purulia has decided 22% of its appeals, Hooghly 12.5% and Howrah 10.5%.This is why Malda, Nadia and North 24 Parganas barely appear in the rejection list. Voters there are not winning their appeals. Most simply have not been heard.The Election Commission’s appeals may also explain why, in some districts, appeals far outnumber the voters deleted there – like by about 24 times in Bankura and 18 times in Kalimpong. The commission has not clarified how its appeals are spread across districts.However, speaking to The Wire, Justice Indrajit Chatterjee, presiding over Appellate Tribunal no 11, confirmed that all the 2.54 lakh appeals against inclusion in his tribunal for 10 seats in Murshidabad were filed by the Additional CEO.Same law, very different oddsEach district’s appeals go to one tribunal, so the district figures amount to a bench-by-bench comparison. The odds of losing an appeal vary enormously.These figures need careful reading. Because the data mixes voters’ appeals with the Election Commission’s, a high share could mean a bench turned down many voters who wanted to be reinstated, or that it agreed with many of the Commission’s requests to remove voters the judicial officers had cleared.A low share could mean the reverse. Without knowing who filed each appeal, the figures show how outcomes differ between districts, not how strict each bench was.For example, in Purba Bardhaman, Bhatar is left out of the chart because its figures could not be verified. The appeals data records 1,064 decided appeals there that left the voter off the roll, out of 1,252, which would be 85%, far higher than any other district.Also read: The Mess That SIR Has Created Is Beyond Repair. It Must be Scrapped.The ECI’s affidavit records 1,064 decided appeals there that left the voter off the roll, out of 1,252, which would be 85%, far higher than any other district. But a scrutiny of the voters who appealed for inclusion shows only 15 rejections from the whole district. Most of the difference lies in one seat. The appeals data shows 852 such outcomes in Bhatar, yet not a single excluded voter who have appealed for inclusion was rejected.This means all those excluded from the voter list were cleared by the judicial officers, but are now out of the roll after the ECI’s appeal. Bhatar is also unusual within its own district. It has 906 decided appeals, while most other seats in Purba Bardhaman have decided fewer than 20.Purba Medinipur, Birbhum and Jhargram rejected around a third. At the other end, North 24 Parganas, Nadia, Dakshin Dinajpur and Alipurduar rejected none, and South 24 Parganas rejected 24 out of more than 20,000.Some of this gap may reflect the cases rather than the judges. The same bench hears Paschim Medinipur and Jhargram, yet it rejected 10% in one and 30% in the other. In Purba Bardhaman’s 1,064 rejections, 852 are from a single seat, Bhatar.Districts that have decided only a few hundred cases may also have started with particular kinds of appeals. Still, a voter in Purba Bardhaman and a voter in South 24 Parganas going through the same process have had very different chances of getting back on the roll.But did the tribunals rule differently for Muslim and Hindu appellants? Answering this question would need, for every decided appeal, the voter’s religion and who filed the appeal. The 6,444 voters whose own appeals were rejected are 80% Muslim by name.If the voters whose own appeals succeeded are also around 80% Muslim, the rejections simply mirror who appealed. If the Muslims are much fewer, it would mean that Muslim appellants were being turned down at a higher rate. A voter-level list of all decided appeals, marking who filed each one, would settle it.For our full coverage on the SIR and the issues it has brought to citizens, click here.