ECI’s Delhi Draft Roll ‘Logical Discrepancies’ Flag Disproportionately Affects Muslim Voters – Findings of SIR Analysis

The ECI released a draft electoral roll for Delhi as part of the SIR process. A software‑driven check for logical discrepancies flagged a far higher share of Muslim voters than their share in the roll. Key Developments Delhi’s draft roll contained 97.54 lakh electors; 13.3 lakh (13.6%) were identified as Muslims, close to the 2011 Census figure of 12.9%. Out of 19.33 lakh electors flagged for logical discrepancies, 20.3% were Muslims – a 12.8‑percentage‑point excess over their overall share. For every 1,000 Muslim electors, 318 were flagged versus 197 per 1,000 non‑Muslims, indicating a 60% higher likelihood of being marked. The five family‑size related reasons (e.g., “progeny ≥ 6”) alone accounted for 30% Muslim representation among the 6.75 lakh flagged electors. The flagging decisions were made by the centralized ERONET software , not by the on‑ground ERO s. Important Facts The eight discrepancy reasons fall into four groups: name mismatches, five family‑size/age‑gap criteria, a tiny “parent ineligible” category, and a catch‑all “no reason mentioned”. Muslims were over‑represented in the family‑size group, where the flagging rate was nearly three times higher than for non‑Muslims (152 vs. 56 per 1,000). The NFHS shows that Muslim fertility rates have fallen from 4.41 to 2.36 between 1992‑93 and 2019‑22, narrowing the gap with Hindus (1.94). Yet the software still uses raw age‑gap calculations that may be unreliable because many older citizens lack exact birth dates. Unlike Delhi, Bihar’s draft rolls showed Muslim deletions (18.4%) close to their population share (17.7%). Bihar did not employ the logical‑discrepancy flag in its SIR, highlighting a regional policy variation. UPSC Relevance Understanding the mechanics of electoral roll revision is essential for GS2 (Polity) and GS3 (Demography). The case illustrates how algorithmic tools ( fertility rate trends) intersect with electoral law, raising questions about fairness, data accuracy, and accountability of the ECI . Aspirants should note the procedural hierarchy: the ECI issues orders, the ERO implements them, but the flagging logic resides in the centralized ERONET software . This separation raises governance and transparency issues that are frequently asked in essay and interview formats. Way Forward Demand clear, publicly‑available definitions of each logical‑discrepancy reason from the ECI . Introduce an audit mechanism for the ERONET software to ensure algorithmic neutrality. Allow EROs to review and contest flagged entries before notices are issued. Incorporate demographic realities (e.g., declining fertility) into the design of discrepancy criteria to avoid indirect bias. These steps would strengthen the credibility of the electoral roll, uphold the principle of equal franchise, and align the SIR process with constitutional values.