Why Famines Are Always Political

On September 14, 1943, a boat carrying rice arrived at the Calcutta docks from Burma. It was not allowed to unload. The British colonial administration had classified the Bengal rice trade as a military matter and imposed procurement controls that effectively redirected food supplies toward the war effort and away from civilian consumption. By the end of 1943, between two and three million Bengalis were dead. In the same period, food exports from India to Britain continued. Winston Churchill, when briefed on the situation, reportedly asked why Gandhi had not yet died.

The Bengal famine is the most extensively documented case for what Amartya Sen argued in his 1981 book “Poverty and Inequality” and later in “Development as Freedom”: that famines do not occur because food is absent. They occur because people lack the entitlements — the rights, purchasing power, or political standing — to access food that exists. This is not a comfortable argument, because it transforms famine from a natural catastrophe into a political crime. The evidence for it is overwhelming, and the implications are ones that powerful institutions have spent decades resisting.

The Food Was There

Begin with the simplest fact. In virtually every major famine of the modern era, food existed within or near the affected region during the period of mass starvation. This is not a quirk of particular cases. It is the characteristic feature of famine that distinguishes it from the theoretical starvation scenario — the island with no food — that politicians invoke when they wish to depoliticize the event.

During the Irish Famine of 1845-1852, food exports from Ireland to Britain continued throughout the worst years of starvation. In 1847, the peak year of mortality, Ireland exported approximately 730,000 gallons of butter and 822,681 gallons of other provisions to Britain, along with substantial quantities of cattle and grain. The exports were smaller than in pre-famine years, but they continued. The people starving in County Mayo did not lack food in Ireland. They lacked money to buy it and political power to commandeer it.

The Holodomor — the Soviet famine of 1932-33 that killed between 3.5 and five million Ukrainians — is another case where the food existed. Soviet grain procurement targets were set deliberately high by Stalin’s government, knowing that meeting them would leave insufficient calories for the Ukrainian peasantry. When villagers hid grain to feed their families, teams of urban workers — the aktiv — were sent to search homes and seize it. The famine was not incidental to Soviet agricultural policy. It was instrumental to it: a mechanism for breaking peasant resistance to collectivization and eliminating Ukrainian nationalism as a political force.

The Ethiopian famine of 1984-85, which generated the Band Aid response and became the defining humanitarian image of its decade, similarly occurred in a context where food existed — in other regions of Ethiopia, in international warehouses, in transit through ports. The famine was concentrated in Tigray and Wollo, where the Derg government was simultaneously fighting a military counterinsurgency. Food aid was systematically withheld from rebel-held areas and used as a weapon of war. The hunger was real; the cause was not crop failure but a calculated political decision.

Entitlement Failure and the Market Mechanism

Sen’s entitlement framework explains the mechanism with elegant simplicity. People have entitlements to food through three channels: they can produce it directly, they can exchange other goods or labor for it, or they can receive it as a social transfer. Famine occurs when all three channels fail simultaneously. But the failure of these channels is always socially structured — it falls along lines of power, property, and political standing.

The Bengal famine illustrates exchange entitlement failure with particular clarity. Bengali agricultural laborers who had no land and no significant possessions could only access food through labor exchange — by selling their work for wages or food. When wartime inflation drove food prices up faster than wages, and when British procurement simultaneously reduced the availability of food for civilian purchase, the exchange rate between their labor and a kilogram of rice collapsed. They were not poor because there was no food; they were dying because the terms of trade had moved catastrophically against them, largely as a result of government policy.

This distinction matters enormously for policy. If famine is caused by food shortage, the solution is supply-side: grow more, import more, increase agricultural productivity. If famine is caused by entitlement failure, the supply-side solution misses the point entirely. You can increase food supply without changing the distribution mechanism that determines who gets it. The Green Revolution — which dramatically increased agricultural yields across Asia and Africa beginning in the 1960s — reduced the frequency of famine but did not eliminate it, precisely because production increases that flowed primarily to large commercial farmers did not necessarily improve the entitlements of landless laborers.

The policy implication that follows from Sen’s framework is politically explosive: preventing famine requires redistribution, not just production. It requires ensuring that the poorest and most politically marginal populations have secure entitlements to food through some combination of land rights, minimum wage guarantees, and social transfer payments. Every one of these mechanisms requires political will to create and maintain. Every one of them has historically faced fierce opposition from propertied classes who benefit from the existing entitlement structure.

Why Democracies Don’t Have Famines

Sen’s most striking empirical claim, stated in “Development as Freedom,” is that no substantial famine has ever occurred in a functioning democracy with a free press. The claim is essentially correct, and its logic is straightforward: democratic governments face electoral punishment for allowing mass starvation, and a free press creates the information environment necessary for that punishment to occur.

The mechanism is not mysterious. When people are starving, they vote against the government that presided over their starvation. A government that cannot be voted out has no electoral incentive to prevent starvation, and many cases where it has active incentives to allow or cause it — as a weapon against political opponents, as a mechanism for clearing land, or simply as an acceptable cost of other policy objectives. The British in Bengal were not democratically accountable to Bengalis. The Soviets were not accountable to Ukrainians. The Derg was not accountable to Tigrayans. In each case, the famine-causing government faced zero electoral consequences from the people dying.

The free press component is equally important and equally structural. The Irish Famine was reported extensively by journalists in British newspapers, and this reporting did generate some political response — including the founding of relief organizations and Parliamentary debates. But the reporting occurred at a remove from power, in a political system where the victims had no votes. The coverage was sufficient to produce sympathy but not sufficient to produce the policy changes that would have saved lives, because sympathy without political agency is charity, and charity is discretionary in a way that electoral punishment is not.

When famines have occurred in democratic contexts — the Bengal famine took place under British colonial rule, in which Bengalis had no meaningful democratic representation — the democratic protection breaks down precisely because the affected population is excluded from the democratic mechanism. This is why colonial famines are so common and why post-independence famines in formerly colonized countries have been dramatically less common despite similar or worse agricultural conditions. The mechanism is not food. It is accountability.

The Colonial Famine Engine

The relationship between colonialism and famine deserves direct statement, because it is still politically contested in ways that the evidence does not support. Colonial rule did not merely fail to prevent famines; it actively generated the conditions for them.

The mechanisms were multiple and overlapping. Colonial administrators typically replaced subsistence-oriented agriculture with cash crop agriculture oriented toward export markets. This increased agricultural productivity in the narrow sense — output per acre for the exported commodity — while destroying the food security buffer that subsistence agriculture provided. When the export crop failed or prices fell, colonial populations suddenly had neither subsistence crops nor cash to buy food.

The colonial tax system compounded this. Most colonial administrations required taxes to be paid in cash rather than kind, which forced subsistence farmers into cash crop production or wage labor whether or not market conditions made this rational for them. A farmer who must pay a cash tax must sell something, which means he must participate in the cash economy, which means his food security is now linked to market prices over which he has no control. When those prices moved against him — as they did during the wave of agricultural commodity deflation in the 1870s and 1880s — the result was mass impoverishment and, in drought years, famine.

Mike Davis documented this pattern exhaustively in “Late Victorian Holocausts,” published in 2001, examining the great famines of 1876-79 and 1896-1902 that killed between 30 and 60 million people across India, China, Brazil, and Africa. Davis’s central argument is that these famines — which coincided with El Niño-driven droughts — were not natural disasters amplified by poverty. They were political disasters caused by the simultaneous imposition of colonial free-market ideology (no interference with food prices, no government grain reserves, mandatory food exports) during a period of climate stress that pre-colonial systems had previously managed through their own risk-spreading mechanisms.

The pre-colonial states of India had maintained grain reserves, regulated food prices in dearth years, and organized public works employment for the rural poor — the precise interventions that Victorian colonial ideology prohibited on the grounds that they would interfere with market mechanisms. The market mechanisms proceeded efficiently. Tens of millions died.

The Persistence of Political Famine

The twentieth century’s famines were overwhelmingly political in their immediate causation: the Holodomor, the Bengal famine, the Chinese famine of 1959-61 (caused by Mao’s Great Leap Forward), the Cambodian famine under the Khmer Rouge, the North Korean famines of the 1990s, the Yemeni famine of the 2010s caused by Saudi bombing of food infrastructure. The pattern is consistent enough that it should have ceased to be debatable decades ago.

Yet the natural disaster framing persists, for obvious reasons. It is more comfortable to attribute famine to drought, flood, or crop failure than to attribute it to political decisions made by identifiable governments. The natural disaster framing permits sympathy without accountability — we can feel badly for victims of drought in a way that does not require us to confront the governments whose policies transformed climate stress into mass death. The natural disaster framing also permits the kind of depoliticized humanitarian response — Send food! — that avoids asking who prevented the food from reaching people in the first place.

The Amartya Sen framework is uncomfortable precisely because it demands that we ask the political question. When people are starving, the first question is not where to get more food. The first question is: who controls the food that already exists, and why don’t the starving people have access to it? That question has a political answer in every documented case of modern famine.

The conclusion that follows is not particularly complicated, though it is routinely ignored in favor of agricultural technology solutions and emergency food aid programs. Famines end when governments become accountable to the people who are starving. They do not end when yields improve or logistics get more efficient, though both of those things help at the margin. They end when the people who die from hunger have the political power to punish the governments that let them die — and not one moment before. Every other approach is treating the symptom while the disease continues untreated.