A stampede at a Durga Puja pavilion in north Kolkata killed at least 17 people Saturday evening, including a seven-year-old girl, and injured 25 others. West Bengal Chief Minister Suvendu Adhikari confirmed the toll, noting that only six bodies had been identified at the time of his statement. Police attributed the disaster to heavy crowds and rainfall at the temporary pandal — a pavilion housing idols of the goddess Durga — saying an announcement triggered commotion as the waiting crowd pushed forward and fell. The mechanics are depressingly familiar. A massive crowd funneled into a temporary structure with limited exits. Rain compressed the timeline. An announcement — likely about weather or closure — created a surge. People fell, and others trampled them. This is not a freak event. It is the predictable output of a system that moves millions of people through ad hoc infrastructure with minimal crowd-flow engineering. Prime Minister Modi called the incident "saddening" and announced compensation of 200,000 rupees ($2,106) per death and 50,000 rupees ($527) per injury. Interior Minister Amit Shah pledged medical support. These are the standard post-disaster gestures — grief expressed, small payments disbursed, the structural problem untouched. The death toll from Indian crowd disasters forms its own grim ledger. In 2024, 121 people died at a Hindu gathering in Uttar Pradesh. Earlier in 2025, Kumbh Mela killed 30. In Bengaluru, 11 cricket fans were crushed during celebrations. In Tamil Nadu, 36 died at a political rally. Each event triggers the same cycle: shock, compensation, promises, then nothing until the next one. The core failure is infrastructure and regulation for temporary mass-gathering sites. Pandals are erected for days or weeks, draw tens of thousands, and operate with crowd management that would be illegal at a mid-size European music festival. Entry and exit flows are uncontrolled. Capacity limits are unenforced or nonexistent. Emergency egress planning is minimal. Rain — entirely predictable during monsoon season — is treated as an unforeseeable complication. Compensation of $2,106 per death is not a policy response. It is a receipt. The cost of a life is set low enough that it creates no fiscal incentive to prevent the next disaster. Until crowd-management standards carry enforcement teeth and liability consequences for organizers and local authorities, the mathematics of this problem do not change. The question is not whether another stampede will happen at an Indian mass gathering. It is whether any institutional actor faces sufficient cost — political, legal, or financial — to alter the infrastructure before it does. The current evidence says no.
