Polity27 October 20265 min read
Why India's Sentencing System Is a Lottery and How a Sentencing Council Could Fix It
The Trigger and the Gap It Exposed
In Sunita Devi vs State of Bihar (2024), the Supreme Court made an observation that cut to the heart of a long-standing problem: sentencing in India had become a mere lottery. Two convicts found guilty of the same offence could walk away with vastly different punishments depending on which bench heard their case. The Court directed the Union Government to examine the feasibility of a national sentencing policy, and the Centre tasked the Law Commission of India with studying the issue and recommending a structured framework.
The timing matters. India had just replaced its colonial-era criminal codes with the Bharatiya Nyaya Sanhita (BNS), 2023 and the Bharatiya Nagarik Suraksha Sanhita (BNSS), 2023. Section 4 of the BNS catalogues the types of punishment available: death, life imprisonment, rigorous or simple imprisonment, community service, forfeiture of property, and fines. Section 258(2) of the BNSS introduced a bifurcated trial process, mandating a separate sentencing hearing after conviction where both sides can argue on factors like age, socio-economic background, prior record, and potential for reform. These are genuine improvements. Yet neither provision tells a judge how much weight to assign any of these factors, or where within a statutory range the sentence should land. The discretion remains almost entirely unguided.
What Unguided Discretion Actually Costs
The constitutional stakes are not abstract. Articles 14 and 21 of the Constitution guarantee equality before the law and require that any procedure affecting life or personal liberty be fair and reasonable. When one judge treats an offender's social standing as mitigation and another ignores it entirely, the resulting disparity is not just inconsistent — it is constitutionally suspect.
The practical costs compound the constitutional ones. Because trial courts lack uniform yardsticks, sentences are routinely challenged as either too harsh or too lenient, forcing High Courts and the Supreme Court to correct them case by case. This feeds directly into the pendency crisis. Non-custodial alternatives also go underused. The Probation of Offenders Act, 1958 has existed for over six decades, yet probation remains a marginal option in most trial courts — not because judges are unaware of it, but because no structural framework prompts them to weigh rehabilitation against deterrence in any systematic way. Research on sexual offence cases has shown that sentencing reasoning sometimes turns on legally irrelevant considerations such as a survivor's perceived chastity, while psychological harm and abuse of trust are underweighted.
The Reform Architecture Being Proposed
The solution gaining traction is a statutory Sentencing Council: an independent, multidisciplinary body that would issue guidelines rather than mandates. The Malimath Committee (2003) recommended precisely such an institution. The model has a well-tested international precedent in the UK Sentencing Council, established under the Coroners and Justice Act, 2009.
The proposed framework would operate in three stages. First, assess the seriousness of the offence by evaluating the offender's culpability and the harm caused. Second, establish an indicative starting point and a sentencing range within the statutory envelope. Third, adjust the sentence up or down based on defined aggravating and mitigating factors — a stage that draws directly on the Law Commission's 47th Report, which emphasised the role of such factors in producing fair and consistent outcomes.
The guidelines would be presumptive rather than mandatory. Judges would be required to consider them in every case but could depart from them, provided they record written reasons for doing so. This preserves judicial independence while creating an accountability trail that appellate courts can actually scrutinise.
Two structural realities complicate implementation. Criminal law sits on the Concurrent List (Entry 1) of the Seventh Schedule, while prisons fall under the State List (Entry 4). Any national sentencing framework must account for the wide variation in correctional infrastructure across states; a guideline that presupposes robust probation services will be meaningless where those services barely exist. On the data side, the National Judicial Data Grid (NJDG) offers a ready platform for building an anonymised national sentencing database, which would allow the Council to keep its guidelines empirically grounded and allow Parliament and the judiciary to monitor whether disparities are actually narrowing over time.