State of Uttar Pradesh & Ors. vs. Atul Kumar Dwivedi & Ors.
Background
Uttar Pradesh Police invited applications for over 3,000 posts of Sub-Inspector, Platoon Commander and Fire Officer, drawing more than 6.3 lakh (630,000) applicants. Because such a huge number of candidates could not sit a single exam, the written test was held in 29 different batches over 12 days with different question papers each time. To ensure fairness across batches of varying difficulty, the recruitment board applied a "normalization"/"scaling" formula to adjust raw scores, resulting in some candidates who scored below 50% in raw marks still qualifying via their normalized score, and vice versa. Unsuccessful candidates challenged this before the Allahabad High Court, which ruled that normalization could only be used to rank candidates for the merit list, not to decide basic eligibility (the 50% cut-off), effectively favouring candidates with higher raw marks; the State and selected candidates appealed to the Supreme Court.
Decision Breakdown
The Supreme Court reversed the High Court and upheld the recruitment board's approach of applying normalization right from the eligibility stage. It reasoned that when an exam is split into many sittings with different question papers, comparing raw marks alone is inherently unfair, since papers can differ in difficulty; normalization corrects for this and must be applied consistently at both the eligibility (50% cut-off) stage and the final merit-list stage, otherwise anomalies would arise where a genuinely higher-normalized-scoring candidate could be excluded while a lower one is included. The Court reviewed its earlier precedents on "scaling" versus "moderation" and held that this case's unprecedented scale (630,000 candidates across 29 batches) justified normalization as the fairer and legally sound method, and that courts should be slow to second-guess an expert recruitment body's chosen evaluation methodology absent bad faith. The 28 February 2019 result was restored, with a suggestion that the State consider some relief (such as age relaxation) for candidates who had qualified on raw marks but were ultimately disqualified after normalization.
Lesson Learnt
When public examinations must be conducted across multiple sittings with different question papers, using a statistical normalization method to compare candidates fairly is legally valid, even to determine baseline eligibility, and courts will generally defer to an expert body's chosen evaluation methodology unless it is arbitrary or made in bad faith.