14,575 People Voted in Tughlakabad. The Draft Rolls Say Fewer Than That Now Exist.

When the Rule Says "Verify," and the Practice Produces Impossible Numbers Article 324

gives the Election Commission a clear mandate: superintendence, direction and control of the preparation of electoral rolls, to keep them accurate and current. Section 21 of the Representation of the People Act, 1950, lets the ECI order a special revision "for reasons recorded by it." Neither provision anticipates what Delhi's draft rolls now show - a scenario where verification has produced a roll smaller than the number of people who provably, recently, voted.

The rule as written treats deletion as correction: remove the dead, the duplicate, the shifted. The rule as practiced, in Tughlakabad and 23 other Delhi constituencies, has removed more people than any plausible correction should.

For Tughlakabad's draft roll to legitimately fall to 1,00,386 - 14,575 below the 1,14,961 who voted there in 2025 - one of three things would have to be true: that constituency saw enormous outward migration in seventeen months, that thousands of 2025 voters have since declared themselves ineligible or that the SIR's verification standard is systematically over-deleting people who remain, in fact, legitimate residents.

There is no independent evidence for the first two. That leaves the third. This is where the Supreme Court's May 2026 ruling in Association for Democratic Reforms v. ECI becomes relevant - not because it forbids this outcome, but because it upheld SIR's validity while explicitly emphasising procedural safeguards: prior intimation and a genuine claims-and-objections period.

The numbers here suggest that safeguard is not functioning as a check in practice. A "buffer" of just 2.6% over actual turnout - the tightest margin among all Phase 3 states - is not what a careful, error-correcting revision produces.

It is what an aggressive, poorly audited one produces, with the burden of proving continued residence falling entirely on voters who may lack the specific documentation the SIR demands, disproportionately affecting migrant workers and the urban poor.

The exam-relevant insight is this: constitutional silence isn't the only source of ambiguity in electoral administration - sometimes a clear mandate ("keep the rolls accurate") collides with an execution standard so aggressive it produces a number no legitimate correction process should.

Article 324's power to revise rolls was never meant to be judged only by whether it followed procedure on paper; Delhi's 24 anomalous constituencies are the evidence that procedure and outcome can diverge sharply and that "the rolls are being purified" and "the electorate is quietly shrinking" can, for a period, describe the exact same set of numbers.