More than one-fourth of the voters in 10 Karnataka assembly constituencies represented by Muslim MLAs have been deleted from the electoral rolls during the Special Intensive Revision (SIR), according to a Vartha Bharati analysis of data provided by political research consultancy Sankalp. The consultancy’s research is based on booth-wise deletion data published by the Election Commission of India (ECI).The analysis found that 27.59% of the total voters in these 10 constituencies were deleted during the SIR, considerably higher than Karnataka’s overall deletion rate of 19.45%.The 10 constituencies together had 9,62,359 Muslim voters on the 2025 electoral roll. Of these, 2,46,759 names were deleted, translating into a deletion rate of 25.64%. Among voters from other communities in these constituencies, the deletion rate stood at 28.79%.The constituencies include Chamrajpet, represented by B.Z. Zameer Ahmed Khan (1,07,722 total voters deleted); Shivajinagar, represented by Rizwan Arshad (81,230); Narasimharaja, represented by Tanveer Sait (99,367); Gulbarga Uttar, represented by Kaneez Fatima (1,05,257); Ramanagaram, represented by H.A. Iqbal Hussain (28,088); and Mangalore, represented by U.T. Khader Fareed (16,989).The other four constituencies represented by Muslim MLAs are Shantinagar, represented by N.A. Haris; Bidar, represented by Rahim Khan; Belgaum Uttar, represented by Asif (Raju) Sait; and Shiggaon, represented by Yasir Ahmed Khan Pathan.Of the 10 constituencies, nine were won in the 2023 assembly elections, while Shiggaon was won by Congress candidate Yasir Ahmed Khan Pathan in the November 2024 by-election.Over 22 lakh voters deleted from Congress strong booths. Photo: Vartha Bharati.The analysis also found that more than 22.35 lakh voters were deleted from electoral booths where the Congress secured 60% or more of the votes in the 2024 general elections.According to the data, 22,35,155 deleted names belonged to booths categorised as Congress strong booths, accounting for 20.74% of the total 1,07,76,645 deletions covered in the booth-wise analysis across Karnataka.For the analysis, a Congress strong booth is defined as an electoral booth where the party secured 60% or more of the votes.A constituency-wise examination further shows that in 39 of Karnataka’s 224 assembly constituencies, more than 35% of the deleted names belonged to these Congress strong booths.These 39 constituencies together account for 19,85,732 deleted names, representing 18.4% of all deletions covered in the booth-wise analysis. Of these 39 constituencies, 26 were won by the Congress in the 2023 assembly elections, 11 by the BJP and two by the JD(S).The analysis also found that these 39 constituencies have an average Muslim share of 22% in their electoral rolls, compared with 12.9% across all 224 constituencies in Karnataka.The findings are based on booth-wise lists of deleted voters and the reasons for their removal, published by the chief electoral officer (CEO), Karnataka. The data examined covers 99.67% of the total deletions in the published booth-wise records. A total of 19,846 deleted names from the overall ECI figure could not be placed in the constituency-wise analysis.The ECI’s published figures put the total number of voters deleted during the SIR in Karnataka at 1,07,96,339.Muslim voters overrepresented in deletions marked ‘absent’ and ‘already enrolled’Analysis of the reasons recorded for voter deletions during the SIR shows that Muslim voters were overly represented in categories such as ‘absent’ and ‘already enrolled’ that depend on enumerator’s assessment.Muslims accounted for 12.9% of the state’s voters on the 2025 electoral roll but made up 17.1% of those deleted under the ‘absent’ category and 20.3% of those removed as ‘already enrolled’. Of the 15,15,088 voters deleted for being absent, 2,58,770 were Muslims. Similarly, 1,43,940 of the 7,08,666 voters removed under the ‘already enrolled’ category were Muslims. In other words, one in five voters deleted as already enrolled elsewhere was Muslim, even though Muslims accounted for only about one in eight voters on the original roll.The pattern was different for the ‘dead’ category. A total of 2,00,277 Muslim voters were deleted under this ground, accounting for 12.2% of all deletions. Unlike ‘absent’ and ‘already enrolled’, which are determined through an enumerator’s assessment during verification, death is a ground that can be verified against official records such as death registers. The fact that Muslim voters were overrepresented in the first two categories but underrepresented in the category of death raises questions about the basis and consistency of the verification process.The findings raise questions about the basis on which voters were marked absent and how many visits were made before their names were removed. It also raise questions about what documentary evidence was used to establish that a voter was already enrolled elsewhere and whether the duplicate entry was produced. The difference in the distribution of deletion grounds also raises questions about why Muslim voters were more represented in categories involving an enumerator’s assessment, while their share was lower in deletions recorded as deaths.22.67% Muslim voters from Muslim-dominated constituencies deleted. Photo: Vartha Bharati.In the 64 assembly constituencies where Muslims account for 15% or more of the electorate as per 2026 final SIR, 22.67% of Muslim voters were deleted during the SIR, compared with 20.19% of other voters. This represents a gap of 2.48 percentage points. These constituencies had 38,82,598 Muslim voters on the 2025 electoral roll, of whom 8,80,543 were deleted. In the remaining 160 constituencies, Muslim voters faced a deletion rate of 20.27%, compared with 18.79% among other voters.Muslim voter deletions in constituencies of prominent Karnataka leadersAn analysis of 25 key constituencies found varying rates of deletion among Muslim voters. In Kanakapura, represented by Chief Minister D.K. Shivakumar, 22% of Muslim voters were deleted from the previous electoral roll. In Shikaripura, represented by B.Y. Vijayendra, the figure stood at 16.5%, while Chittapur, represented by Karnataka Home Minister Priyank Kharge, recorded 25.7% deletion of Muslim voters from the previous roll.In Padmanabhanagar, represented by leader of opposition R. Ashoka, 32.5% of Muslim voters were deleted, accounting for 11,278 names. Shivajinagar, represented by Rizwan Arshad, recorded a deletion rate of 32.1% (25,673 voters). In Gandhinagar, represented by Dinesh Gundu Rao, 37.8% of Muslim voters were deleted, amounting to 6,327 names.ECI records 10.08 lakh additions against 7.85 lakh form 9 applications. Photo: Vartha Bharati.The addition of new voters to Karnataka’s electoral rolls during the SIR has raised questions, with the number of additions recorded by the ECI exceeding the number of applications filed for inclusion, according to data examined as part of this analysis.The data provided by Sankalp, which has tracked the figures from June 16, 2026, shows that the ECI recorded 10,08,038 additions against 7,85,716 form 9 applications, a difference of 2,22,322. In other words, the number of additions was around 1.3 times the number of applications recorded.The data also shows that 1,88,324 voters were deleted through claims and objections, while 8,96,519 entries were modified. The five forms tracked by Sankalp, including applications for addition, deletion, correction and shifting, accounted for 19,26,935 applications.However, the ECI has not provided a publicly available reconciliation explaining how the 10.08 lakh additions were accounted for against the 7.85 lakh form 9 applications. The disposal status of the applications has also not been published in the data examined by Sankalp.The issue is not without precedent. A similar discrepancy came under scrutiny in Uttar Pradesh, where an investigation by The Reporters’ Collective, published on October 6, 2026, found that the ECI had added 84.3 lakh voters to the final electoral roll against 71.38 lakh publicly disclosed form 9 applications, a difference of 12.92 lakh. The investigation also found that another 16.18 lakh applications, reportedly received before publication of the draft roll, were missing from the publicly available application records.The Uttar Pradesh CEO subsequently rejected the allegations, saying the 16.18 lakh applications had been collected during the enumeration exercise and that applications received before publication of the draft roll were not required to be published in form 9. The CEO maintained that the process followed ECI guidelines.In Karnataka, however, the difference between the recorded additions and form 9 applications remains unexplained in the records examined for this analysis. If the ECI has recorded 2.22 lakh more additions than the form 9 applications tracked during the revision, what is the source of these additional entries, and where are the corresponding application and verification records?This story was first published on Vartha Bharati. Read the original article here.