Kolkata: Nine in every ten West Bengal voters whom the Election Commission of India (ECI) is seeking to remove from the electoral rolls, even after judicial officers have found them eligible, have Muslim names. This is not only in the state’s Muslim-majority border belt, but also in Hindu-majority districts which are far from the border.Appeal records compiled by the Civic Research Group cover 16,08,175 appeals against the inclusion of voters whom judges cleared during the special intensive revision (SIR). An analysis of the names of those voters by The Wire finds that at least 14,39,797 of them – 89.5% – have Muslim names.Muslims form 27% of the state’s population according to the 2011 census, and make up about 29% of the electorate in the state.The same pattern in every kind of seatGrouping constituencies by their minority population share shows how far the voters appealed against differ from the population around them.Muslim share of populationSeatsMuslim share, weighted by electorateVoters appealed againstShare with Muslim namesShare of cleared voters appealedUnder 10%896.1%1,27,65361.7%47.9%10% to 20%7115.3%2,19,20883.8%46.5%20% to 30%5124.8%2,53,69490.3%49.2%30% to 50%3937.1%2,79,32193.8%47.8%Over 50%4160.3%7,10,65896.6%51.2%In seats where Muslims are about 6% of the population, more than six in ten voters appealed against have Muslim names, roughly ten times their share of the population. In seats where minorities are about 15%, it is more than eight in ten.In 137 constituencies, at least 90% of the voters appealed against have Muslim names. Of the 160 seats where Muslims are less than a fifth of the population, 69 have appeals in which at least 80% of those targeted have Muslim names. Across those 160 seats, 2,62,416 voters with Muslim names face appeals.Far from the borderThe gap is widest not in the border districts, but in districts where few Muslims live.DistrictMuslim share of populationVoters appealed againstShare with Muslim namesPurba Medinipur12.0%27,09192.9%Paschim Medinipur10.7%26,40583.8%Bankura7.4%15,40980.4%Jalpaiguri12.3%35,33685.1%Purba Bardhaman23.3%90,63795.9%Hooghly15.1%24,90284.4%Cooch Behar23.5%46,22092.4%Paschim Bardhaman10.9%25,87178.1%In Purba Medinipur, where Muslims are about one in eight of the population, more than nine in ten voters appealed against have Muslim names. In Egra, a constituency where Muslims are about 8% of the population, the figure is 94.9%. In Chandipur, Kanthi Dakshin and Ramnagar, all in the same district, it is between 93% and 95%. In Sagar and Gosaba, in the Sundarban, it is 97% and 95%. Here, Muslims are under 12% of the population.In Murshidabad, where Muslims are about 60% of the population, 98.2% of voters appealed against have Muslim names. The gap there is smaller only because the population is already mostly Muslim.Where the pattern breaksThere are constituencies where most voters appealed against do not have Muslim names. They are mostly seats where very few Muslims live, and where the SIR’s net caught others instead.Yet even in these seats, appeals against voters with Muslim names are out of proportion to their numbers.ConstituencyMuslim share of populationVoters appealed againstWith Muslim namesShareKrishnaganj (SC), Nadia5.7%2,4121576.5%Behala Purba, South 24 Parganas5.1%1,93224812.8%Jadavpur, South 24 Parganas1.2%88112514.2%Bidhannagar, North 24 Parganas4.6%1,72534419.9%Habibpur (ST), Malda4.6%1,85133318.0%Nayagram (ST), Jhargram1.3%66111317.1%In Krishnaganj, a seat reserved for the Scheduled Castes in Nadia’s Matua belt, voters with Muslim names make up 6.5% of those appealed against, slightly above the Muslim population share of 5.7%. Elsewhere, the gap is wide. In Jadavpur, where Muslims are about 1% of the population, 14.2% of voters appealed against have Muslim names, more than 12 times their share. In Nayagram, the figure is 17.1% against a Muslim population share of 1.3%, and in Bidhannagar and Habibpur, about four times the Muslim population share. In the hill constituencies of Darjeeling, Kurseong and Kalimpong, there was almost no match between Muslim names and deletion appeals.Also read: Bengal SIR Observer Says He Knew Nothing About CEO Office’s 16 Lakh Appeals Against Cleared VotersBroadly, wherever Muslims live in any number, they dominate the list of voters the Election Commission wants removed. Where they are almost absent, as in the Matua belt, Adivasi seats and parts of Kolkata’s suburbs, even if most voters on the list do not have Muslim names, Muslims still appear on the deletion appeals list in far greater proportion than in the population around them.What the data shows, and what it does notThe appeals themselves do not appear to pick voters by the religious makeup of a constituency. The Election Commission appealed against between 46% and 51% of cleared voters in every group of seats, from the least to the most Muslim. Across constituencies, there is almost no relationship between a seat’s Muslim share and the share of its cleared voters appealed against.What changes is who reached this stage in the first place. The appeal-against-inclusion records cover only voters the Election Commission appealed against, not every voter a judge cleared. But a second set of records, of appeals filed by voters whom judges struck off, fills in most of the gaos.Of 27,06,057 voters deleted by judicial officers, 22,17,619, or 82%, have appealed to the tribunals. Of those appellants, 79.3% have Muslim names.An independent analysis points the same way. The SABAR Institute, which has released a public repository of West Bengal’s supplementary deletion lists classified by religion and gender, found that 70.17% of the more than 27 lakh voters deleted from the “under adjudication” list were Muslim, and 51% were women. Together, the two sets of records cover 38,25,794 of the 59,59,458 voters in these constituencies who were sent to judges, about 64%. Of them, 31,99,122 have Muslim names. Even if not one of the remaining voters had a Muslim name, that would mean at least 53.7% of everyone sent to adjudication was Muslim, nearly twice the Muslim share of the state’s population.Muslim share of populationVoters sent to judges per 1,000 electorsMinimum share of those with Muslim namesDeleted voters who appealed, share with Muslim namesCleared voters appealed against by ECI, share with Muslim namesUnder 10%3023.7%33.7%61.7%10% to 20%5242.6%64.0%83.8%20% to 30%6854.8%81.7%90.3%30% to 50%10860.3%88.5%93.8%Over 50%20865.1%95.3%96.6%In seats where Muslims are about 6% of the population, at least a quarter of those sent to judges had Muslim names, four times their share of the population. In seats where Muslims are about 15%, it is at least 43%.One comparison in these seats stands out. Among voters whom the judges struck off in seats which have less than 10% Muslim share and who then appealed, about a third have Muslim names. Among voters judges have cleared in the same seats whom the Election Commission appealed against, more than six in 10 have Muslim names. The two groups are different people, and the data cannot show the religious makeup of every cleared voter. But in Hindu-majority seats, the Election Commission’s deletion appeals fell on Muslim-named voters far more than the deletions did.That points to the stages before the appeal. Voters reached the judges only after they were flagged for “logical discrepancies” and sent to adjudication. The Commission’s own ERONET data shows that the 60 lakh voters sent to judges tracked the cases central observers had flagged, district by district, not the cases electoral registration officers had left undecided. The appeals then fell on about half of everyone the judges cleared, a pool already made up overwhelmingly of Muslim voters, in Hindu-majority districts as much as in the border belt.Who filed themIn its counter-affidavit to the Supreme Court on October 26, based on a communication from the Chief Electoral Officer of West Bengal dated October 22, the Election Commission said the appeals were filed by “officers of the Office of the CEO, West Bengal, in their capacity as authorities under the RP Act”, and that “these matters were not brought to the Commission”. In the same affidavit, it said the voters concerned have their documents on the portal and “a favourable order passed by the Ld. Judicial Officer in their favour”.The Wire has written to the office of the Bengal CEO and two additional CEOs asking for the names and designations of the officers who filed the appeals, the provision of the RP Act or the Registration of Electors Rules under which officers of the CEO’s office can appeal against voters in constituencies they do not administer, and the Supreme Court order and paragraph on which they relied. It has asked whether Electoral Registration Officers and Additional Electoral Registration Officers were consulted before appeals were filed against voters on their rolls, on what grounds and by what criteria voters were chosen, and whether each case was reviewed individually. It has also asked for a copy of the October 22 letter cited in the affidavit, and for the number of appeals and their outcomes by constituency. This report will be updated when the office responds.