The Long Road from Crime to Conviction in India

At a just-concluded police conference in Bengaluru, Karnataka Chief Minister Siddaramaiah pointed out that drug manufacturing units in Mysuru and Bengaluru had been detected by teams from Maharashtra, not by local police. If other states’ officers could trace supply chains on Karnataka’s streets, he asked, what was failing in Karnataka’s own intelligence and enforcement loop?

The comment was an acknowledgement that crime control does not collapse at the point of detection, but through the chain of investigation, prosecution and trial. In that chain, Karnataka performs adequately, but not decisively.

Karnataka’s conviction rate, as per the latest NCRB Crime in India report data, sits close to the national average. Police register crimes, charge sheets are filed, but many cases lose coherence before a verdict is delivered: forensic reports arrive late, prosecutors are overworked and often understaffed, investigations are disrupted by transfers, witnesses fall out of contact. These are systemic frictions that, over time, dissolve the coherence of cases.

The conviction figures that we rely on come from the National Crime Records Bureau’s annual Crime in India reports, which are compiled from police and court returns across states and union territories.  These datasets are the only national source for state-wise outcomes, and they remain indispensable for policymaking and comparative analysis. But even as lawmakers and commissioners use these numbers to benchmark performance, the statistics have blind spots that affect how we interpret justice outcomes. For example, independent research has noted that hate-motivated incidents are effectively absent from NCRB’s tables, even where other monitoring shows significant numbers, a form of statistical omission that invisibilises entire classes of harm. Besides, under-reporting of sexual assault, domestic violence, and community hostility is well documented; convictions may rise or fall in the records, but the denominator of unreported crimes remains opaque. Civil-society researchers highlight that this erasure is not only empirical but also political, since what is not coded does not exist in the official lexicon. 

That is to say that NCRB figures lay out the visible finish line, which is the number of cases completed by courts, but they cannot reliably tell us about what never started, never reached trial, or was quietly smoothed out before prosecution. Thus policy responses based solely on conviction percentages risk treating symptoms of data absence as if they were signals of justice success or failure.

Still, within the data’s own frame, a pattern emerges: conviction outcomes track less with crime incidence and more with state capacity to shepherd cases through the middle ground between registers and courts.

In Kerala and Mizoram, verdicts arrive more reliably because procedure holds. Investigations tend to close within statutory timelines. Prosecutors are not endlessly rotated. District courts are staffed closer to sanctioned strength. These states invested in trial infrastructure and legal services early, and they keep funding them even when returns are invisible. Their policy problem now is different: ensuring that strong conviction outcomes are not paired with low reporting, especially in domestic and sexual violence, where social barriers still shape what enters the system at all.

In the large northern and eastern states—Uttar Pradesh, Bihar, West Bengal—low conviction reflects structural overload. Police stations absorb more cases than they can investigate with depth. Courts operate with extreme pendency. Evidence thins and cases exhaust themselves. Here, the justice system is not malfunctioning. It is undersized.

The urban growth states of Maharashtra, Karnataka, Telangana and Gujarat face a different constraint. Crime here is increasingly financial and digital: cyber fraud, online extortion, coordinated scams. Police units have adapted faster than courtrooms. Cyber cells and inter-state coordination have improved. But prosecution remains anchored in legal practice built for physical evidence and territorial crime. Judges face technical testimony that moves faster than precedent. Conviction, in this category of crime, remains rare. The institutional lag here is legal and procedural.

Karnataka sits at this intersection. It has infrastructure, training academies and administrative capacity. But outcomes do not yet reflect that advantage. Officials point to policy gaps that rarely feature in public debate: vacancies in prosecution services, slow forensic turnaround, lack of witness support mechanisms, and investigation continuity disrupted by routine transfers. These are not issues resolved by directives. They require sustained budgetary decisions and appointments.

Conviction rates, of course, are not synonymous with justice. But across states, one rule holds: where investigation and prosecution are treated as essential public infrastructure, conviction follows; where they are treated as peripheral services, delay becomes structural.

That is the implication of Siddaramaiah’s warning. Not that policing has failed, but that governance has underinvested in what comes after policing. If the state wants conviction rates to improve, the interventions are known: expand forensic capacity, professionalise prosecution as a career service, protect witnesses, limit disruptive transfers of investigators, and align judicial staffing with population and case load. None of this fits neatly into a campaign slogan. All of it determines whether law becomes outcome.