Fund the fever: paying for help before the outbreak peaks

Indian companies spent ₹7,150 crore on healthcare through CSR in 2023–24. Almost all of it was planned a year ahead and pointed at a map. Humanitarian funders have learned to release money on a trigger, days before a crisis peaks. Health philanthropy in India can learn the same trick.

By Ishan Kashyap, Founder & Editor · 13 October 2026 · 11 min read

About this essay. It follows part one, which argued that pharmacy sales could warn of outbreaks weeks before official counts. The CSR head in the opening scene is a composite. Every number is real and sourced at the end.

The camp that arrived on schedule

In March, a CSR head we will call Kavita signs off next year’s health plan. It is a good plan. There will be four health camps near the company’s plant, an eye-screening drive and a contribution to a district hospital. The board approves it in April. The money is committed, the NGO partner is contracted, the dates are fixed.

In September, a fever sweeps through a district three hundred kilometres away. In part one of this series, we imagined a chemist there called Imran, selling paracetamol twice as fast as usual, and a district officer called Dr Meera, hearing about it three weeks late. There is no line in Kavita’s budget for that district, and no way to move money in less than a board cycle. In November, her first health camp opens, on schedule, near the plant. It is well run. Nobody there has dengue.

Nobody in this story did anything wrong. The money simply moved on the calendar, and the fever moved on its own.

Illness decides weekly, in a lane. Philanthropy decides annually, in a boardroom.

The money, and where it goes

The sums are not small. Indian companies spent ₹34,908.75 crore on CSR in 2023–24, according to the government’s National CSR Portal, and ₹7,150.81 crore of it, about a fifth, went to healthcare. The Bain and Dasra India Philanthropy Report puts total private philanthropy at about ₹1.43 lakh crore in 2024–25.

The same report puts total social-sector spending at about ₹27 lakh crore, of which roughly 95% is public money. That is the most important number in this essay. Private money cannot outspend the state on health. Its advantage has to come from somewhere else: from being faster, more flexible and better targeted than a government budget can be. At the moment, most of it is none of those things.

Part of the reason is written into the law. Section 135 of the Companies Act says a company “shall give preference to the local area and areas around it where it operates” when spending CSR money. The ministry has since clarified that this preference is “only directory and not mandatory”, and that companies should balance it against national priorities. But habits formed around the plant gate are hard to shift, and a CSR committee is naturally more comfortable funding projects its employees can visit.

The other reason is the calendar. CSR is planned and reported by financial year. Multi-year projects are allowed, but their plans are fixed in advance. Nothing in the standard cycle is designed to respond to a signal that arrives on a Tuesday in September.

What the humanitarians learned

Disaster funders faced the same problem, and some of them solved it. On 4 July 2020, a global flood forecast gave a better than even chance of severe flooding along Bangladesh’s Jamuna river in the middle of the month. The UN’s Central Emergency Response Fund released $5.2 million against that forecast. Within four hours of the trigger, the agencies had authority to spend it. The UN called it the fastest allocation in the fund’s history and the first made before peak flooding.

The World Food Programme sent 4,500 taka by mobile money, about two weeks of food spending, to more than 23,000 poor households, four or five days before the river peaked. Researchers who compared them with similar households that did not receive the money found that recipients were 36% less likely to go a day without eating, and more likely to evacuate their families and livestock. A later review by the UN agencies found that this early support reached people at half the cost of the previous year’s response, US$13 a person against US$26, and about 100 days sooner.

This is now a system, not a single experiment. The Start Network, a group of aid agencies, disburses money from its Start Fund within 72 hours of an alert, and in 2025 sent £6.4 million to projects that acted before a crisis. The Red Cross and Red Crescent movement lets national societies pre-agree up to 500,000 Swiss francs for an early action protocol, valid for up to five years, released automatically when a forecast crosses a pre-defined trigger.

Health has been slower to learn. An analysis of cholera vaccine allocations in Africa, published in 2025, found that 77.4% were made reactively, after outbreaks began, and that governments took an average of almost 300 days to request doses. When money and supplies wait for proof, they arrive after the peak.

India already pays on triggers

The idea is not foreign to India either. In 2024, about 50,000 women in the informal sector, organised by the Self-Employed Women’s Association, were covered by an extreme heat insurance product designed by Climate Resilience for All, underwritten by Swiss Re and delivered by ICICI Lombard. In May, after temperatures passed 40°C, each woman received a small cash payment, and over the season 92% of them received insurance payouts triggered automatically by temperature data, without filing a claim. The largest were about US$20 a person, in the hardest-hit district. Nobody convened a meeting to decide that it was hot.

States are doing it too. In August 2024, Nagaland signed a parametric disaster insurance agreement with SBI General Insurance covering 2024 to 2027, with the premium paid by the state. Its first claim, of about ₹1.07 crore, was processed in March 2025 after the monsoon. Tellingly, an earlier pilot in 2021 to 2023 had never paid out, because its threshold was set too high and the data behind it was inaccurate. A trigger is only as good as the signal under it.

The idea: a fever fund

Put these pieces together and a different model of health philanthropy appears. Instead of deciding in March where to hold camps in November, a group of funders agrees in March what they will do when illness rises, and where it rises becomes the only open question.

The menu matters as much as the trigger. Much of what an outbreak needs is cheap and hard to waste. In the 2019–21 national family health survey, the most recent to report it, only 60.6% of young children with diarrhoea received ORS, 30.5% received zinc, and only about a quarter received both, even though both are inexpensive and recommended. An estimated 102,678 children under five died of diarrhoeal disease in India in 2017. Our inference is that a fever fund which did nothing but put ORS and zinc into homes in the right districts a few weeks early would justify itself, and that if a trigger fired falsely, ORS in a cupboard is a small price.

Does the law allow it?

Probably, with care. The Companies Act’s list of eligible CSR activities includes health care and, since 2019, disaster management. In March 2020, the ministry told companies that spending on COVID-19 was eligible CSR and that those items could be “interpreted liberally”. The CSR Rules also allow “ongoing projects” lasting up to three years beyond the year they start, with unspent money moved to a separate account within 30 days of the year’s end and spent within three years.

In our reading, a multi-year standby programme with defined triggers, defined uses and named implementing partners could be structured as an ongoing project. That is an interpretation, not legal advice. Any board considering it should take advice first. The point is that the obstacle is mostly habit, not law.

Measure it like a marketer

Funders should hold a fever fund to a standard most health camps never face: proof that it changed something. This is where marketing analytics earns its place. A marketer who wants to know whether a campaign worked does not just count how many people saw it, but compares regions that got it with similar regions that did not. The same discipline applies here, and our guide to measuring marketing returns describes the toolkit.

In a pilot, the fund could roll out to some eligible districts in the first year and others in the second, and compare hospital admissions, deaths, household spending on treatment and the number of days between the first signal and the first response. India already has a precedent for paying on verified results. The Utkrisht impact bond in Rajasthan paid for improvements in the quality of private maternity care, judged by an independent assessor, which estimated that facilities meeting the standard would avert about 13,449 maternal and newborn deaths by 2023. That is a modelled estimate, not a count. But it shows that Indian funders can be persuaded to pay for outcomes rather than activities.

What could go wrong

Three risks deserve a straight answer.

The trigger misses. Nagaland’s first pilot is the warning: a threshold set badly, on poor data, pays nothing and destroys trust. Thresholds should be tested against several past seasons before money depends on them, and revised every year.

The signal misses the poorest. Sales only see people who can buy. A fund that listens only to chemists will hear the middle class first. That is why the second signal, from community health workers and the official platform, is not optional, and why thresholds in the poorest districts should be set lower.

Private money crowds out public duty. Outbreak response is the state’s job, and with 95% of social-sector money coming from government, it will stay that way. A fever fund should buy speed and flexibility at the margin, working through district health teams rather than around them, not build a parallel system.

Six questions for a CSR committee

RavenArc tests every decision against six questions. Applied to a fever fund, they make a short agenda for the next CSR committee meeting.

The same September, replayed

Run Kavita’s year again. In March, alongside the camps near the plant, her board commits a share of the health budget to a pooled fever fund with four other companies and two foundations. The board approves rules, not projects.

In September, the pharmacy signal in Imran’s district jumps for two weeks running, and community health workers report a cluster of fevers. Both signals agree. The trigger fires on a Tuesday. By Friday, an NGO already contracted for that district has ORS and zinc kits moving and a testing camp booked, and Dr Meera’s team knows the help is coming. Kavita hears about it in a two-line email. Her company’s name is on a report that shows days saved and admissions avoided, which is a better story for the annual report than a photograph of a camp.

In November, the camps near the plant still open on schedule. They were never the problem. The problem was that they were the only thing the money could do.

India does not need more health money to do this. It needs some of the money it already has to listen to the chemist, and to move when the counter speaks.

Part one of this series. Three people knew about the fever: how India’s pharmacy sales could warn of outbreaks weeks before the official count.

RavenArc tests every decision against six questions. See the RavenArc Decision Method.