How Surjit Bhalla cut through 13 crore voter deletions to find 70 lakh unexplained

The Special Intensive Revision, or SIR, is a large-scale review of electoral rolls by the Election Commission. Its purpose is to create cleaner voter lists by checking whether listed people are alive, registered in the correct place and counted only once. This matters because the final roll determines who can vote.

During the process, names can be removed at the draft stage. The article says 13.37 crore names were deleted across three phases. These included deceased voters, relocated people, duplicate entries and people officials could not trace. Draft deletions occur before claims and appeals are heard, so they are not automatically final exclusions.

The current debate focuses on what remains after this process. Surjit Bhalla estimates that about 70 lakh eligible voters may be missing without explanation in states with final rolls. The SIR therefore involves both administrative cleaning and safeguards, including opportunities to challenge deletions before final publication.

The 13.37 crore figure counts every name removed from draft rolls across all three SIR phases. It is therefore a broad administrative total, not a count of eligible voters wrongly denied registration. The figure also includes deletions made before claims and appeals were decided.

The article lists 2.8 crore deceased voters, 6.3 crore relocated people, one crore duplicate entries and three crore untraceable people among the deletions. Bhalla then examined states where final rolls were available and compared those rolls with expected voter numbers. Across nine large states, he found a gap of about 73 lakh.

After allowing for roughly five lakh non-citizens who could be removed legitimately, Bhalla estimated around 70 lakh eligible voters remained unexplained. He said seven of every eight removals were justified. The estimate is not an official Election Commission count, and unexplained does not prove every deletion was wrongful.

A voter list must reflect current eligibility and location. Names belonging to deceased people should not remain active, while people who relocate may need registration at a different address. Duplicate entries can allow the same person to appear more than once. Untraceable entries may no longer represent voters living at the recorded location.

The article reports that the draft deletions included 2.8 crore deceased voters and 6.3 crore people who had relocated. Another one crore entries were duplicates, while three crore people could not be traced. These categories explain why the 13.37 crore headline cannot be treated as a direct count of wrongful exclusions.

The process was still not automatically error-free. The deletions were recorded before claims and appeals were heard, and Bhalla found a possible shortfall after comparing final rolls with population-based expectations. His estimate leaves about 70 lakh eligible voters unexplained, while stressing that unexplained does not prove wrongful removal.

Bhalla began by estimating India’s adult population at 100.8 crore in mid-2026. He did not assume every adult should appear on the rolls, because registration is voluntary. Using Election Commission data for people aged 18 to 29, he estimated that 7.5 per cent of adults were unregistered, plus another 1 per cent among older adults.

That produced a benchmark of about 92.2 crore names for a clean national roll. He then applied the approach state by state, using population estimates and voter registration data. In nine large states with final SIR rolls, the expected total was 53.51 crore, compared with 52.78 crore actually listed.

The difference was approximately 73 lakh. Bhalla deducted an estimated five lakh non-citizens who could have been removed legitimately, leaving around 70 lakh eligible voters whose absence remained unexplained. The estimate used information available through October 16, 2026, and was not an official Commission count.

The immediate consequence of wrongful deletion is exclusion from the final electoral roll. A person whose name is missing may not be able to cast a ballot unless the error is corrected. This is why Bhalla says the unexplained figure is not small to the people concerned, even though it is less than one adult in a hundred.

The article does not provide separate figures for poor, migrant or newly registered voters. However, it identifies relocation as the largest listed deletion category, at 6.3 crore, and examines registration among younger adults. People who move or recently seek registration may therefore be especially important cases for checking, claims and appeals.

The present reality is uncertain rather than conclusive. Bhalla estimates about 70 lakh eligible voters may be missing, but says this does not prove every deletion was wrong. The forward implication is that affected voters need the revision’s correction process to resolve unexplained omissions before rolls are treated as final.

Bhalla places the SIR debate in a longer history of imperfect electoral rolls. After the last nationwide intensive revision, the 2004 roll contained 2.8 crore more names than the estimated adult population, an excess of 4.4 per cent. In 2014, the roll was 3.1 per cent higher than the adult population.

These comparisons show that voter lists and population estimates are not identical measures. Registration is voluntary, so some adults never enrol. Lists can also contain people who have died, moved away or appear more than once. Conversely, eligible people may be absent from the roll, creating a shortfall rather than an excess.

The current revision therefore has two sides. It can remove outdated or duplicate records, but it can also leave eligible voters unexplained if checking is incomplete. Bhalla’s estimate of roughly 70 lakh is intended to measure that possible gap, not to claim that every historical or current mismatch represents fraud or error.

Maintaining a voter list means keeping millions of records current while people’s circumstances change. Registration is voluntary, so a portion of adults never enrol. At the same time, deaths, relocation and duplicate registration can make existing entries inaccurate. Officials must identify these changes without accidentally removing eligible voters.

The SIR figures show the scale of that task. Draft deletions included 2.8 crore deceased voters, 6.3 crore relocated people, one crore duplicate entries and three crore people who could not be traced. These categories require different checks, while claims and appeals must remain open for people who believe their names were removed incorrectly.

The challenge is visible in the final estimates. Bhalla calculated that nine large states should have 53.51 crore voters but listed 52.78 crore, a gap of about 73 lakh. His conclusion was roughly 70 lakh unexplained eligible voters after adjustments. The process therefore needs continuing verification, not just one-time deletion.