5 key trends that have emerged from SIR exercise | Number Theory

It has now been more than eight months since the Election Commission of India (ECI) started its Special Intensive Revision (SIR) process to overhaul electoral rolls in India. The exercise started with Bihar and has now been conducted in nine states and three Union territories, accounting for 237 of India’s 543 Lok Sabha constituencies. This number excludes the 14 Lok Sabha constituencies of Assam, where the roll revision exercise is called Special Revision (SR) rather than SIR, and where the process is significantly different from that of SIR. ECI has already announced that it will begin this exercise in the remaining parts of the country from April.

February 28 was an important milestone in this journey as ECI released the final electoral roll for West Bengal with six million electors on the rolls still facing scrutiny. The list was issued after a Supreme Court order, which perhaps tried to pre-empt the disruptive impact the ongoing scrutiny could have had on the timely conduct of elections in the state, as a new assembly must be formed before May 7. West Bengal’s problem is not the only one associated with the ongoing SIR exercise. In fact, India’s biggest state, Uttar Pradesh, is still trying to finish the process and ECI has set a revised deadline of April 10 for the state.

All of this makes it a good opportunity to ask what do we know about the SIR exercise so far? Here are five key trends which can be flagged.

To be sure, the exercise is not yet complete in West Bengal either, where the process was already 116 days long on February 28. Moreover, electors not mapped to the 2002 SIR roll or with logical discrepancies were not called for hearings in the SIR exercise conducted in Bihar, a process which has started only with the SIR across 12 states/UTs that began on November 4. However, that has extended the timeline by only around 10 days in most states.

Seen on a cumulative basis, the SIR exercise deleted a net 35.4 million or 8.1% of India’s electors in 12 states and UTs where the exercise has been completed.

To be sure, this figure relates to only the changes in the SIR process. For example, Bihar added half a million electors to its rolls between the completion of the SIR exercise on September 30 and the 2025 election, who have not been included in these calculations. Moreover, the numbers will change significantly once West Bengal and Uttar Pradesh complete this exercise. As is to be expected, the extent of state-wide deletions shows a large variation.

Once this process is completed, a “draft roll” is published which includes everyone who submits their enumeration form, with a specified deadline for the final roll’s publication which will correct for errors/omissions in the draft roll as well as make new additions. In almost all states, the number of voters fell between the reference roll for SIR and the draft roll, and then increased in the final roll. Of course, there are large variations across states here too. And West Bengal has become the first state to put a large number (six million) of voters under scrutiny even after publishing its final roll. It also the only state apart from Goa, where the voter count has decreased between the draft and final roll.

To be sure, the SR exercise in Assam deleted voters only in the final roll and there was no physical verification of documents in the state. The draft roll had only marked people for deletion.

This essentially suggests that a lot of the deletions as part of the SIR exercise in Bihar were of voters who were either dead, had moved (more on this later) or were registered in two places but not necessarily voting in both of them. It is likely that the trend will continue across other states also. This is exactly why drawing a one-to-one correspondence between SIR and voter-suppression has been problematic. To be sure, as many as six million voters ending up under scrutiny in West Bengal has only fanned these concerns.

To be sure, this does not mean that migration is the only reason behind deletions. For example, the correlation described above does not hold true for the summary revision exercise in Assam. Moreover, for migration to be the only reason for deletion, past growth rate of electors would be strongly correlated with the rate of deletion in SIR, which is not necessarily the case. In addition, it is not possible to say if the trend will continue to hold in West Bengal, where the status of over 8% of electors in the final roll is not clear yet. It is also not possible to say yet if the trend holds everywhere.

Only six states – Tamil Nadu, West Bengal, Rajasthan, Gujarat, Uttar Pradesh, and Bihar – published district-wise summary of deletions at the draft stage. Among these, Rajasthan did not publish district-wise details of the final roll.