India spends heavily on disaster preparedness for its citizens, yet a HelpAge India study finds the people most likely to die in the next cyclone are the ones the system was never actually built to see.

Why "The Elderly Are Vulnerable" Is the Wrong Sentence

• and Has Been Since 2005 India's disaster management architecture was built around a category error and it has repeated that error for two decades. The National Disaster Management Act, 2005, like most of the frameworks that followed it, was designed to answer "what kind of hazard is this"
• cyclone, flood, heatwave - not "what kind of person is standing in its path." For twenty years, that has meant treating "the elderly" as a single risk group, the same way a 1970s census once treated "the poor" as one undifferentiated mass before disaggregation by caste, gender and region became standard practice. HelpAge India's new study, drawing on 2,224 older persons across ten states, finally does to disaster policy what those earlier census reforms did to poverty policy - and the finding is uncomfortable: two people can share an age bracket and face entirely different odds of surviving the same cyclone. The report's own example makes the pattern unmistakable. An uneducated, unskilled, unpensioned older widow faces categorically higher risk from a cyclone than an able-bodied older man who lives with family - not because she is older, but because widowhood, poverty, illiteracy and social isolation stack on top of age itself. This is what the study calls an "Intersectional Place Perspective," and it should sound familiar to anyone who has studied why the 2008 National Action Plan on Climate Change struggled to protect its most vulnerable citizens: a policy built for the median case systematically fails everyone who isn't the median case. India has made this exact mistake before, in poverty measurement, in health access, in disaster relief itself - and each time, the fix was the same: stop designing for "the group" and start designing for the specific, overlapping disadvantages within it. What makes this recurrence worth naming is where the pattern finally breaks in the data itself. Odisha's village-panchayat cyclone warning system - cited in the same report as a genuine success - worked precisely because it operated at a scale small enough to know who the isolated widow in the village actually was, not just how many "elderly" lived there. That is the entire argument for a people-centric shift in one sentence: hazard-centric systems count heads, people-centric systems know names. The report's headline number - India's elderly population doubling from 149 million to 347 million by 2050 - is not really the urgent part of this story. The urgent part is that unless disaster policy learns to disaggregate "elderly" the way it eventually learned to disaggregate "poor," a doubling population means a doubling of exactly the invisible, unpensioned, unaccompanied cases this report finally has the data to name. For the exam, this is the sharper version of a familiar GS1 theme: vulnerability is not a demographic category, it is an intersection - and India's disaster institutions are only now, twenty years late, starting to build for that.