The Granary Is Full. The People Are Starving.
You are standing at a port warehouse in a famine year. The building is stocked with grain, the loading cranes idle, the ledgers balanced. You can smell the sacks through the corrugated walls. Outside, people are dying. This is not a hypothetical designed to shock. It is, with some variation in the specifics, a pattern that has repeated itself across multiple famines in the last century: Bengal in the 1940s, Ethiopia in the 1980s, parts of sub-Saharan Africa since. The food was there. The dying happened anyway.
The question most people ask is: how? The honest answer requires scrapping almost everything the word "famine" tends to conjure.
Why "Not Enough Food" Is Usually the Wrong Diagnosis
For most of recorded history, famine was understood as a supply event: drought came, crops failed, people starved. That framing is not wrong in every case, but the economist Amartya Sen spent decades demonstrating that it is catastrophically incomplete. His core argument, developed through close study of the Bengal famine of 1943, is that people starve not because food disappears from a region, but because they lose what he called "entitlements": the combination of income, assets, legal rights, and social position that allows a person to command food in a market or through other means.
Bengal is the case that makes the argument undeniable. Aggregate food availability in the province that year was not dramatically below normal. The famine killed roughly three million people. What had changed was the economic position of the rural poor: wartime inflation had eroded wages, speculation had driven rice prices sharply upward, and the colonial administration had prioritised urban and military supply chains. Landless labourers found that their daily earnings, already thin, could no longer buy the calories they needed. The food was in the system. They simply could not reach it.
This is the part most guides skip. Famine is less a production problem than a distribution and purchasing-power problem. The distinction matters enormously for how you respond to it.
How Entitlement Collapses: A Worked Example
Consider two smallholder farmers, call them Amara and Bekele, living in the same drought-affected district. Both start the year with comparable plots. When the rains are poor and the harvest comes in at sixty percent of normal, both face hardship.
Amara owns her land outright, has a small savings buffer from a good previous year, and belongs to a local cooperative that can access government relief channels. Her entitlement is stressed but intact. She sells a goat, buys grain at elevated prices, and gets through the lean months.
Bekele is a tenant farmer, paying a fixed rent regardless of harvest. His landlord collects in full. He has no savings and no cooperative membership. When grain prices spike because traders anticipate shortage, his wages as a day labourer fall simultaneously: farmers who are also struggling hire fewer hands. He is caught in a scissors movement, income collapsing and prices rising at the same moment. Within weeks, his ability to buy food has effectively gone. He is in famine conditions while grain sits in the district market. Amara is not.
Same village. Same drought. Radically different outcomes, driven entirely by the structure of entitlements, not by aggregate supply.
The Political Machinery That Keeps the Granaries Locked
If entitlement collapse explains the mechanism, politics explains why the mechanism is so often left to run unchecked. Sen made another observation that has held up stubbornly well: no substantial famine has ever occurred in a functioning multiparty democracy with a free press. The reasoning is not sentimental. Governments facing elections and a press that can report freely have powerful incentives to intervene before mass starvation becomes visible. The political cost is too high to ignore.
Authoritarian governments face no such constraint. The record bears this out in ways that are genuinely difficult to look at. Famines in the twentieth century were frequently entangled with deliberate policy: the Soviet collectivisation famine of the early 1930s, the Chinese famine that followed the Great Leap Forward, the Ethiopian famine that the Derg regime initially denied and then used as a tool of counter-insurgency in Tigray. In each case, grain existed. In each case, political decisions determined who could access it. These were not failures of agriculture. They were failures of power, which is a different thing entirely and a more damning one.
Conflict compounds all of this. When fighting disrupts market supply chains, displaces farmers before harvest, and is used deliberately to blockade populations, entitlement collapse is not a side effect. It is, sometimes, the intent.
Still, even well-intentioned democratic governments can fail here. If relief systems are slow, bureaucratically rigid, or captured by the interests of large agricultural producers rather than the rural poor, the political check weakens considerably. The form of government matters less than the actual accountability of those in power to the people most at risk.
What People Get Wrong About Food Aid
There is a persistent folk remedy that needs to die, or at least be substantially revised: the idea that shipping food into a famine zone is straightforwardly good. It can be. In a genuine acute supply collapse, direct food distribution saves lives and the argument for it is simple.
But in famines driven by entitlement failure, flooding a local market with imported grain can work like a slow poison: it destroys the livelihoods of the farmers who were not starving, undermining local production for years afterward. It can also, if aid flows are not carefully managed, be diverted by the same political actors who created the conditions for famine in the first place. Aid agencies have learned this the hard way across decades of field experience. The response that works is increasingly targeted: cash transfers that restore purchasing power, support for local markets, and protection of the legal entitlements (land rights, labour rights, access to credit) that make people resilient to shocks before they arrive.
Ask yourself why that shift in thinking has not yet fully penetrated public understanding of what famine is. The answer, probably, is that a photograph of a food shipment is easier to publish than a photograph of a property rights reform.
The catch: that gap between what works and what the public believes funds the wrong responses, over and over.
The Uncomfortable Conclusion
Famine, understood properly, is almost always a political and economic failure before it is an agricultural one. The global food system produces, by most credible estimates, enough calories to feed every person alive. Famine persists because entitlements are distributed with grotesque unevenness, because political systems that would face consequences for mass starvation are absent in the places most at risk, and because the people who die in famines are, without exception, the people who were already poor, landless, and marginal before the crisis began.
Drought and flood and crop disease are real. They create the pressure. Pressure becomes catastrophe only when the human systems of distribution, power, and accountability have already failed the people at the bottom.
The granary being full is not the reassurance it sounds like. It just means someone else holds the key.