nEW dELHI, April 10 -- Two days after the Election Commission of India (ECI) released district-wise electoral roll data after the adjudication exercise, it also released the more disaggregated and politically salient numbers at the assembly constituency (AC) level for West Bengal. The statistics only magnify the skew in voter deletions in the adjudication exercise in comparison to the pre-adjudication Special Intensive Revision Exercise (SIR) in the state. For context, 13 states and Union territories have undergone SIR so far (although the final roll is awaited in Uttar Pradesh), and Assam underwent a Special Revision (SR) exercise. But West Bengal is the only state that underwent a special adjudication exercise after the SIR process. The state saw a net deletion of 6.2 million voters between pre-SIR to pre-adjudication exercise, and another 2.7 million have been deleted in the adjudication exercise, taking the total net deletions to 8.9 million. To be sure, the final elector count for the state before the election could still change because of new voters being added and the fate of 22,163 voters, who are still to be adjudicated in the data released. That said, how is the adjudication process different from the pre-adjudication SIR process in West Bengal? Here is what the data shows. West Bengal saw 8.1% of its pre-SIR electors being deleted in the SIR exercise before adjudication and another 3.9% of the pre-adjudication roll during the adjudication process. The degree of skewness in these two rounds of deletions, however, is very different and the adjudication deletions are significantly more concentrated in fewer ACs than pre-adjudication ones. A look at the decile-wise deletions-from bottom to top 10% across ACs-shows this clearly. The lowest 10% deletions before the adjudication process were under 3.59% (D1) and the highest 10% deletions were above 15.2% (D9), giving a ratio of 4.2 between D1 and D9. D1 and D9 for the deletions from the pre-adjudication roll to the post-adjudication roll are 0.31% and 7.93%, a ratio of 25.8. (See Chart 1) One big reason for the bigger skew in deletions during adjudications is the disproportionate deletions which have happened in ACs which are likely Muslim dominated. HT put together a list of 67 ACs which saw at least one Muslim MLA being elected in the 2011, 2016 and 2021 assembly elections in West Bengal. This would suggest a significant number of Muslim electors in the AC. Pre-2011 ACs are not comparable with the current ones because the 2008 delimitation process led to a change in AC boundaries. These 67 ACs had a share of 23% in the state's pre-SIR elector roll. They saw a 19.1% state-wise share in pre-SIR to pre-adjudication removal of voters, somewhat in line with their pre-SIR share in total electors. But they had 46% share in electors being put under adjudication and now account for 40.5% of the deletions under the adjudication process. To be sure, these ACs have not seen a drastic fall in their elector-share in the state, which continues to be 22.6% as of now. (See Chart 2) That said, the pre-adjudication deletions, or deletions post-adjudication do not seem to suggest a relation with either victory margins at the AC level in the latest (2024) election in the state or party-wise winners. (See Charts 3A, 3B)....