In the first week of March 2026, Iran closed the Strait of Hormuz. Qatar stopped loading liquefied gas at Ras Laffan. And within days, four thousand miles away, Bangladesh shut down five of its six urea fertiliser plants and left a single factory running.

Nothing failed in a field. No harvest was lost. A shipping lane closed in the Gulf and the machines that feed Bangladesh's soil went quiet, because the gas that runs them was rationed to keep the power on and the cargoes that would refill the tanks were suddenly in doubt. That is how modern hunger begins. Not with a drought. With a decision, taken somewhere else, about who gets the gas.

This is a piece about a thing most people have backwards. We picture famine as an absence, a year the rain did not come, the fields bare, the granaries empty. That picture is almost always wrong. The hunger that is coming, the hunger that has arrived before and will arrive again, happens with the warehouses full. It is not made by the weather. It is made by people, in markets and ministries. It is made on purpose, or at least with full knowledge of the result. And it keeps recurring. Once you see that, you cannot unsee it and you start to notice who is standing where when it happens.

The clock already started

The thing about a fertiliser shortage is that it does not show up as hunger for the better part of a year. That lag is the most important fact in the whole story, because it is what makes the misery predictable.

In Bangladesh the dominant rice crop is Boro, the irrigated dry-season harvest and it is more than half of the country's rice. It goes into the ground around the middle of December and comes out from April to June. Fertiliser withheld now is not a price on a shelf, it is a yield that will not exist next spring. The plants went dark in March. The monsoon Aman crop that is planted through the summer takes the first hit, with its harvest late in 2026. The big Boro crop takes the larger one, planted in December and reaped in the spring of 2027. The damage is not a risk. It is on a calendar. You can mark the dates.

Bangladesh grows only about forty per cent of the urea it needs. The rest is imported and more than half of it routes through the same Gulf that just closed. So the shortfall hits twice, once through the idle home plants and once through the import bill and both taps were turned at the same moment by the same shipping lane.

Why five per cent becomes thirty

Here is the counterintuitive engine under everything. A small shortfall does not make a small price rise. It makes a large one.

The reason is that the demand for food does not bend. People have to eat, so when there is less, they do not buy proportionally less, they bid the price up to hold on to what they have. Economists call the demand "extremely inelastic" and the word does real work. Roberts and Schlenker, in a study published in the American Economic Review, found that diverting roughly five per cent of the world's calories into ethanol pushed global food prices up by about thirty per cent. Five per cent in, thirty per cent up. That is not a typo and it is not a worst case. It is the measured behaviour of an inelastic market and it is the whole reason a modest disruption can become a catastrophe at the till.

This is old knowledge dressed in new data. Three centuries ago Gregory King observed that a shortfall in the harvest raised the price of grain far more than the shortfall itself. The market has not changed its nature. A five per cent gap is a thirty per cent price and a thirty per cent price is a death sentence for anyone already spending most of what they have on food.

It was never the weather

If the engine is inelastic demand, the trigger is almost never the one we assume. We blame the sky. The record says blame the policy.

Run the tape back to the food crisis of 2007 and 2008. Global rice output that year was not down. It was a record. And the price of rice still rose somewhere between 117 and 149 per cent. The economists who took it apart afterwards did not find a missing harvest. They found a panic. Derek Headey attributed about 61 per cent of the rice price spike to export restrictions imposed by just four countries, India, Vietnam, Egypt and China. Martin and Anderson, working the same period, concluded that trade policy alone explained 45 per cent of the rice price increase and about 30 per cent of the wheat. The phrase the literature settled on for the rice spike was man-made.

It was not one or two panicked governments either. Across the 2007 to 2011 stretch, of 105 countries surveyed, 33 of them, very nearly a third, reached for export restrictions. Each one was acting rationally to protect its own people. Together they built the fire. A country that bans exports to keep its grain at home removes that grain from the world market, the world price climbs, the next country sees the climb and bans its own exports and the spiral feeds itself. The harvest was fine. The fear was the famine.

The gun is loaded and pointed at the floor

So where is that reflex right now. The encouraging answer and the dangerous one, is that it is dormant.

India spent 2023 locking down its rice. It banned non-basmati white rice exports, it banned broken rice, it slapped duties on parboiled. Then the harvests came in strong and the warehouses filled and India unwound all of it. Non-basmati and parboiled were freed in the autumn of 2024. Broken rice was released in March 2025. As of now the world's largest rice exporter has its doors open.

That is not safety. That is a loaded weapon resting on the floor. Every restriction India lifted, it can reimpose in an afternoon and history says it will the moment prices climb far enough to frighten Delhi about its own streets. The 2026 fertiliser shock is precisely the kind of thing that starts the climb. The lesson of 2008 is not that someone was reckless. It is that thirty rational governments, each defending its own, will between them manufacture the disaster none of them wanted. The doors are open today. They have hinges.

The warehouses are full

Now the fact that should stop you cold. None of this is happening because the world is short of food. The world has never been less short of food.

In May 2026 the United Nations food price index sat at 130.8, around eighteen per cent below its 2022 peak. The 2025 cereal harvest was a record, just under three billion tonnes and revised upward since. Global grain stocks are at a record. The ratio of stocks to use, the number that tells you how much cushion the system is carrying, is the highest it has been since 2017 and 2018, comfortably above the level anyone would call tight. The agency's own word for the supply situation is comfortable.

Hold those two pictures in your head at once. Record harvests, record stocks, full warehouses. And a fertiliser system one shipping lane away from buckling, with the export-ban reflex resting on its hinges. The danger does not sit on top of scarcity. It sits on top of plenty. That is the part that should frighten you more, not less, because it means the abundance is no protection at all. The thing that turns plenty into hunger is not the size of the harvest. It is who can command it.

The buffer nobody will spend

If there were any doubt that a five per cent shortfall is a small thing physically, look at where the grain already goes.

By the projections of the OECD and the FAO, of all the cereals the world grows, only about forty per cent is eaten directly by people. A third goes to animal feed. The remaining quarter and more goes to biofuel and other industrial uses. So roughly sixty per cent of the grain humanity produces never goes straight into a human mouth. On top of that, the FAO's long-standing estimate is that the food lost and wasted every year is equal to more than half the world's annual cereal crop.

Set a five per cent shortfall against that. It vanishes. It is a rounding error inside the grain we burn in engines and feed to livestock and throw away. The calories to cover it exist many times over, sitting in slack that the rich world would barely feel. The reason a five per cent gap becomes a famine anyway is not that the food is not there. It is that price does not allocate by need. It allocates by who can pay. The shortfall is not absorbed where the waste is. It is routed to where the poverty is. That is a choice the system makes by default and calling it a law of nature is the most useful lie the comfortable ever told themselves.

Hunger is a question of entitlement

The economist Amartya Sen spent his life on this and reduced it to a single, devastating correction. Famines are rarely about the supply of food. They are about the ability to command it. Bengal in 1943 starved more than a million people with food in the province. What collapsed was not the harvest. It was the purchasing power of the poor, their wages against the price, their claim on what was there.

The mechanism that decides who starves is income and it is brutally simple. In the poorest countries a household spends around forty per cent of what it has on food and at the very bottom, in places like Nigeria, the figure runs near sixty. In a rich country like Britain or the United States it is under ten. So when the same global price rises by a third, it lands on a British shopper as an irritating month and on a Bangladeshi labourer as the arithmetic of which of his children eats. One price. Two completely different events, divided by nothing but the share of your money that food already takes. The market does not see hunger. It sees bids and it fills the high ones first.

Demand destruction is a euphemism for the dying

There is a phrase the markets use when a price gets high enough that demand finally falls. They call it demand destruction. In oil it means idled factories and emptier roads and nobody mourns it. In food it means something the word is built to hide.

Food demand is inelastic precisely because you cannot destroy it without destroying the consumer. So when the price climbs until demand finally drops, what has actually happened is that people have been removed from the market. It comes in tiers. First people trade down, protein out of the diet, the cheaper grain, the worse calories. Then they cut quantity, smaller portions, skipped meals. Then they sell what they have to buy food now, the livestock, the tools and finally the seed stock, which is to say they eat next year's harvest to survive this year's price. That last tier loops straight back to where we started, because a farmer with no seed and no fertiliser plants less and the shortfall deepens into the next season by his own hand. And at the end of the ladder is the tier the euphemism exists to avoid naming.

So understand what the price chart is showing you. The spike is the inelasticity, demand refusing to bend. The roll-over at the top, the relief, the moment the market clears, is the demand finally being destroyed, which is to say the poorest being priced out and removed. The price curve and the mortality curve are the same line, read two different ways. A trader watching for the top is watching for the moment the hungry are fully gone from the market. That is not a metaphor. It is what the chart is made of.

The chokepoint is bigger than calories

There is one more turn of the screw and it is the one that should end any comfort drawn from the full warehouses. The gas that makes the fertiliser does not only make the fertiliser. It makes the things that move food from the warehouse to the plate and a famine of distribution looks much the same as a famine of supply to the person who cannot eat.

Start with carbon dioxide. The CO2 the food industry runs on is not pulled from the air. It is captured as a byproduct of ammonia production, the same reaction that makes nitrogen fertiliser, with natural gas as the principal input. So when a fertiliser plant goes dark over gas, the CO2 goes with it and CO2 turns out to be load-bearing across the whole food system. It stuns pigs and poultry at slaughter. It fills the modified-atmosphere packaging that gives fresh meat, salad and bread their shelf life. It carbonates the drinks, freezes the cold chain as dry ice and does duty in hospitals as medical gas.

Britain ran this exact experiment in the autumn of 2021 and it is worth telling in full because it is the whole argument in miniature. In September 2021 the American firm CF Industries shut both its UK fertiliser plants, at Ince in Cheshire and Billingham on Teesside, because the gas price had made production unprofitable. Those two plants alone supplied around sixty per cent of Britain's food-grade CO2, with Billingham accounting for perhaps a third of national capacity on its own, some 750 tonnes a day. Within days an online grocer pulled frozen lines for want of dry ice, the pig industry warned of a slaughter backlog and the government, facing a food-supply emergency caused by a fertiliser price, stepped in with a taxpayer subsidy reportedly worth tens of millions of pounds to restart Billingham for the sole purpose of making CO2, followed by a brokered pricing deal to hold supply through to the end of January 2022. A cost in the gas market had walked straight into the meat aisle in under a week.

Then Britain did the telling thing. It dismantled the capacity rather than secure it. CF permanently closed the Ince plant in 2022, idled ammonia production at Billingham that August in favour of importing ammonia and by July 2023 proposed to shut the Billingham ammonia plant for good. So the country that had just learned how much of its food system hung on two fertiliser plants responded by closing them, trading domestic capacity for imported ammonia and longer supply lines. The lesson was right there and the opposite lesson was the one that stuck.

The plastic tells the same story from another angle. The polyethylene and polypropylene that most food and medical packaging is made of begin as ethylene and propylene, cracked out of natural-gas liquids in steam crackers, with ethane the dominant feedstock and gas the process heat. So a gas shock hits packaging twice, once through the lost CO2 and once through the resin and Europe's crackers have been cutting rates and closing under exactly that pressure. The wrapper is as exposed as the contents.

This is why the full warehouses are no comfort. A modern food system does not only need calories grown. It needs them slaughtered, chilled, wrapped, gassed and shipped and every one of those verbs runs back to the same gas, the same chokepoint, the same shipping lane in the Gulf. The harvest can be a record and the shelves can still empty, because the thing that breaks is not the field. It is everything between the field and the mouth.

The same petrochemical stream reaches further still, into the medicines it also feeds but that is a second piece and it deserves its own.

It is not a collapse, it is a sorting

We have been waiting for the wrong catastrophe. We imagined the four horsemen riding in, the dramatic end, the systems going down all at once. That is not how it works and the food story is the clearest proof.

Civilisation does not fall. It contracts. It sheds. The lights stay on and the shelves stay stocked for everyone who can still pay, while the world quietly ends for the ones who cannot. A good year for the people holding the ledger and an apocalypse for the people outside the line, in the same country, in the same week, divided only by which side of the price they stand on. That is not a forecast. It is the daily condition of a fair part of the planet already and the template is on the news most nights if you are willing to read it as a template and not an exception.

Which brings us to the uncomfortable part, the one the polite version of this argument leaves out. Many will die, yet fortunes will be made. That is not an accident of the system. It is the system doing what it does, scarcity priced by entitlement paying the people standing furthest from the hunger. The only distinction that carries any weight is between the money made alongside the dying and the money made because of it. A long position in fertiliser transfers wealth and rides a curve someone else bent. An export ban, a hoard, a cornered market, those bend the curve and there the deaths are the mechanism of the gain and not a regrettable side effect. Same spreadsheet. Different ledger. Most of the people who profit here will never have to know which one they were on.

What this asks of the reader

Ireland has been hit far more gently than Bangladesh, so far. That phrase, so far, is the whole of the warning. The thing to brace against was never a single dramatic event you could stockpile against. It is a line, drawn by price and entitlement, that moves inward and the only question that matters is which side of it your people are on when it passes. You cannot tin your way onto the right side of an entitlement line. What keeps people on the live side of it is what they are tied to and what claims they can still make, which is relational and political long before it is material.

So the work is not preparation for a siege. It is refusing the lie that turns plenty into famine and calling it weather. It is naming, early and plainly, before the famine images make the story cheap, that the hunger now arriving is a decision and not a disaster, that the calories exist, that the warehouses are full and that what is killing people is the quiet machinery that routes a five per cent shortfall to the weakest and books the difference as profit. The harvest is fine. It was always going to be the fear and the ledger and the open doors with their hinges.