UP’s Law and Order After 2017: What Changed Under Yogi Adityanath
Uttar Pradesh’s reputation for weak law and order was not invented for election season. For years the state’s size, politics and criminal networks produced a familiar picture: frequent riots and curfews, land and mining mafias, politically protected gangsters, and crime rates that often put UP near the top of national tables in absolute numbers. When Yogi Adityanath became chief minister in March 2017, his government made “zero tolerance” the centre of its pitch. Nine years later, the claim of a transformation is partly supported by National Crime Records Bureau (NCRB) data and partly inflated by political language. The record is better understood as a sharp shift in policing style than as a simple story of crime disappearing.
The starting point matters. In 2016, UP recorded about 4,889 murders and 4,816 rapes. Riots that year numbered 8,016. The state accounted for a large share of India’s killings and crimes against women, not only because it is India’s most populous state but because organised crime and public-order failures were widely seen as routine. Business groups often listed security as a reason to stay away. The government that followed treated that diagnosis as its first brief.
NCRB figures show a decline in several violent categories after 2017, and a clearer improvement in the most recent years. Murders fell from 4,324 in 2017 to 3,491 in 2022, 3,206 in 2023 and 3,218 in 2024. The 2024 murder rate was 1.3 per lakh population against a national rate of 1.9. Rape cases declined from 4,246 in 2017 to 3,209 in 2024. Robbery and dacoity dropped steeply from 2017 levels of 4,089 robberies and 263 dacoities. Kidnapping for ransom has been near zero in recent official tables. A Times of India review of five violent categories between 2022 and 2024 found UP outperforming the national trend in murder, attempt to murder, rape, kidnapping and rioting. Government statements go further, bundling those offences and claiming an 85 per cent fall in “heinous” crime against 2016. That figure depends on the basket and the baseline; the direction of travel in the core violent offences is nonetheless downward.
Rioting is the most contested metric. All-riot cases were 8,990 in 2017 and 8,908 in 2018 before falling to 5,714 in 2019, 4,478 in 2022, 3,160 in 2023 and 2,610 in 2024. That is a large reduction, especially after 2022, and steeper than the national decline over the same recent period. The government’s stronger claim is different: that there have been no major communal riots or prolonged curfews since 2017, against 815 incidents and 192 deaths in 2012–17 and 616 incidents and 121 deaths in 2007–11. NCRB’s separate communal-riot category, introduced later, often records zero or near-zero for UP in recent years. Independent monitors have counted communal clashes that official classification does not label as communal riots. The honest distinction is between the old pattern of large, prolonged communal breakdowns and a later period in which major riots became rare, even if local violence and thousands of riot cases of other kinds did not vanish.
The method that defined the Yogi years is the encounter-and-asset campaign. Official tallies through 2025–26 put encounter operations in the range of 15,000 to 17,000, with roughly 250 to 289 people killed, more than 11,000 injured—often shot in the legs, a practice critics call “half-encounters”—and more than 30,000 arrested. Pre-2017 encounter deaths over a comparable stretch were far lower. Parallel action under the Gangsters Act produced tens of thousands of cases and arrests, and the government claims seizures of illegal assets worth more than ₹14,000 crore. Demolitions of properties linked to accused persons became a public symbol of the same campaign. Police recruitment rose, including women personnel. UP-112 response times improved. Later “Operation Conviction” figures were used to argue that courts, not only police firing, were now delivering outcomes.
Supporters say this broke the old protection economy. Notorious gangsters lost networks, property and political cover. Everyday fear of extortion and kidnapping receded in many districts, and that change is cited as a reason investment became easier to sell. The public inside the state has often backed a hard line against known criminals. That popularity does not settle the legal question. Human-rights groups, some retired IPS officers and court observations have described many shootings as extra-judicial. Allegations include staged encounters, rewards announced after the fact, and a disproportionate impact on Muslims and some OBC groups. The Supreme Court’s 2014 guidelines on encounters were meant to stop exactly this pattern. The National Human Rights Commission has reviewed large numbers of official encounter reports and rarely found foul play; critics call that record a sign of weakness, not proof of cleanliness. The Allahabad High Court has said firing at non-vital body parts cannot be treated as standard policing and has ordered investigations in specific cases.
Data caveats are necessary. UP’s absolute crime totals remain high because the population is huge. Better FIR registration, including online systems, can raise counts even when actual violence falls. Total cognizable crime has not always moved in the same direction as murder or dacoity; cybercrime and petty offences complicate the picture. Crimes against women are still large in raw numbers even when rates have at times been below the national average. High-profile cases such as Hathras and Unnao showed that statistical improvement does not guarantee consistent justice. Land disputes continue to drive a large share of homicides.
National context also matters. Violent crime declined in several categories across India after 2014. UP’s recent performance is notable because it often did better than that national trend, not because the rest of the country stood still. Comparing Yogi’s years only with the last Samajwadi Party term is politically convenient; comparing rates with other large states and with UP’s own population share is more useful.
What really changed, then, is not that crime ended. Organised crime became costlier. Large communal riots and long curfews became rare. Several violent-crime rates fell relative to UP’s own mid-2010s peak and, lately, relative to India. Police were given political cover to act against previously insulated networks. The price of that shift was a policing model that treats deterrence, injury and demolition as everyday tools, and that has repeatedly tested the boundary between law enforcement and extra-judicial punishment. The NCRB tables, encounter logs and court rebukes belong in the same account. Without all three, the transformation story is either a slogan or a denunciation, not an explanation.
About The Author
