The science of strategic starvation: how food has been weaponized throughout history

On May 2, 1941, a group of senior German state secretaries gathered in Berlin to solve a logistics problem. The agenda was food — how to feed an army of three million men advancing east into the Soviet Union. Their conclusion, minuted with the flatness of a procurement note: the Wehrmacht would eat from occupied territory, and “x million people” would starve as a consequence. The x was not uncertainty. It was an acknowledgment that nobody in the room cared about the number. Three weeks later, the Economic Policy Guidelines for Barbarossa filled in the variable: “many tens of millions” would become surplus to requirements.

No one objected. No one resigned. The meeting moved to the next item.

What set that gathering apart was not the strategy. Commanders have been making the identical calculation for as long as armies have besieged cities — from Caesar’s Gaul to the Saudi-led coalition’s Yemen. The Germans were unusual only in writing it down. Which raises a question that isn’t moral but analytical: why does the most effective weapon in military history keep being classified as a side-effect?

The logic of the empty stomach

The strategic case for starvation is not complicated. A starving population cannot fight, cannot produce, cannot resist. Unlike bombardment, starvation requires no ammunition, leaves infrastructure intact for the conqueror’s use, and produces surrender without the attacker absorbing casualties. The besieged city, once taken, still has its granaries, its roads, its workshops — everything bombardment would have wrecked. The weapon destroys the defenders while preserving the prize.

Caesar understood this at Alesia in 52 BCE. Facing a hilltop fortress held by roughly eighty thousand Gauls under Vercingetorix, he chose not to storm it. He built. The result — an inner ring of fortifications stretching some eighteen kilometres, studded with twenty-three redoubts, backed by a longer outer line facing the Gallic relief army — was not a feat of aggression. It was a feat of patience. When the relief force attacked from outside and the garrison sortied from within, the double line held. But the decisive blow was neither battle. Caesar’s own account in De Bello Gallico records the endgame: food exhausted, sorties beaten back, Vercingetorix rode out to surrender. The Romans lost a fraction of the men a direct assault would have cost.

Eighty years earlier, Scipio Aemilianus applied the same logic at Numantia. Celtiberian fighters had bloodied Roman armies in open battle for twenty years. Scipio brought sixty thousand men — outnumbering the defenders roughly fifteen to one — and instead of attacking, built a wall. Towers. Ditches. He blocked the river with booms so nothing could float downstream. Then he waited. Appian records what happened inside over the following months: the Numantines ate leather, then boiled it, then ate the dead. When survivors finally stumbled out to surrender, they were barely recognizable as human — emaciated, matted, nails grown into claws. Most never came out. They burned the city and killed themselves rather than capitulate.

The caloric arithmetic is what makes siege warfare rational rather than merely cruel. A population of eighty thousand requires roughly forty to sixty tonnes of grain per day — about half to three-quarters of a kilogram per person, based on what scholars of ancient Mediterranean diet estimate for daily consumption. That supply chain is fragile. It requires open roads, functioning markets, agricultural hinterland, carts and pack animals moving on a schedule. An attacker who can sever it doesn’t need to breach a wall. Biology does the rest. The besieging army, with supply lines open behind it, simply waits while the defenders’ bodies consume themselves — muscle tissue first, over the course of weeks, then organ tissue, then the immune system’s capacity to fight off infection. The timeline is predictable, the outcome near-certain, and the besieger’s losses are limited to whatever the starving can manage in increasingly feeble sorties. The besieger does not need to risk a single soldier scaling a wall. He only needs to maintain his perimeter and wait.

And the weapon compounds. A siege ending in starvation warns the next city. After Alesia, the defenders of Uxellodunum — who knew what had happened on that hilltop — surrendered faster. Each use reduces the cost of the next, because the besieger’s reputation travels ahead of his army and does part of the work before the first trench is dug.

Starvation is not the weapon of the primitive commander. It is the weapon of the commander who can count.

The industrialisation of hunger

The ancient siege worked through physical encirclement — stone walls around a city, ditches dug by hand, soldiers standing watch. It required a besieger to invest a specific place with a specific force. The twentieth century discovered you could achieve the same result without walls. What changed was not the principle — control sustenance to break a population — but the scale and means of control. Administrative management of grain extraction, railway schedules, port access, and distribution bureaucracies could starve populations not across a hillside but across a continent. You didn’t need to surround a city. You needed to control the paperwork governing who ate and who didn’t.

The Nazi Hunger Plan, formalized in the weeks following that May 1941 meeting, was the most explicitly documented instance of starvation as state policy. Herbert Backe, its principal architect, projected the deliberate starvation of roughly thirty million Soviet citizens to redirect agricultural output to the Wehrmacht during Barbarossa. The framing was zero-sum logistics in its purest form: occupied Soviet territory would produce the grain, the German army would consume it, and the people who had grown it would do without. The planning documents did not euphemize. They laid out a resource-allocation problem in which Soviet lives were the variable to be eliminated.

Implementation followed invasion immediately. Official rations for Soviet prisoners of war were set at 2,200 calories per day for those performing forced labor — a figure confirmed in Wehrmacht records held by the US Holocaust Memorial Museum. That was already insufficient for sustained life under hard labor. But even this inadequate ration was a fiction. In practice, prisoners received far less — many survived on roughly 700 calories a day, a rate of intake that produces organ failure within weeks. Slow starvation by another name. Of 5.7 million Soviet soldiers captured, approximately 3.3 million died, the vast majority from deliberate starvation, exposure, and disease born of nutritional collapse. By February 1942, two million of the 3.3 million captured to that point were already dead. Snyder’s Bloodlands documents a broader figure: 4.2 million Soviet civilians starved by German occupation between 1941 and 1944.

The Hunger Plan was not atrocity committed in the fog of war. It was pre-war resource planning with Soviet lives placed on the expendable side of a logistics spreadsheet.

Stalin’s government performed a parallel calculation in Ukraine the decade before, though without the Germans’ helpful candor about it. Soviet grain requisitions extracted roughly 4.3 million tonnes from Ukraine in 1932 — during a harvest that was below average, disrupted by forced collectivization, but not at famine levels. No drought occurred. The harvest, without the requisitions, would have been sufficient to sustain the population. This isn’t contested in serious scholarship: Kulchytsky’s decades of archival work on the famine, synthesized in his major monograph, and the broader historiographic consensus confirm that weather cannot explain what happened. When local officials reported that procurement quotas were causing starvation, Moscow didn’t reduce quotas. Moscow increased enforcement — dispatching brigades to confiscate hidden grain, blacklisting entire villages from trade goods, imposing internal passport controls that trapped starving peasants where they stood.

Approximately 3.9 million Ukrainians died in the resulting famine, according to the most rigorous demographic analysis — a collaborative study by the Ptoukha Institute of Demographic and Social Studies in Kyiv and the University of North Carolina at Chapel Hill. Ninety percent of the deaths concentrated in 1933, after winter stores were gone.

The Holodomor was not agricultural failure. It was an extraction policy that removed food a population could have survived on — famine as political discipline against a nation whose separate identity Moscow found threatening. The Soviet state did not need to build walls around Ukraine. It controlled the grain elevators, the railways, the internal border checkpoints. The administrative state was the siege engine.

Bengal in 1943 completed the twentieth century’s pattern. A 2019 study by Mishra et al., published in Geophysical Research Letters, analyzed soil moisture data across six major Indian famines from 1870 to 2016 and found the 1943 Bengal famine was the only one with no drought association whatsoever. Rainfall was above average. The famine that killed between two and three million people resulted, the study concluded, entirely from the failure of policy — no drought, no natural mechanism.

British denial policies confiscated rice stocks and boats from coastal Bengal to prevent their use by advancing Japanese forces — scorched-earth measures that destroyed local food distribution networks. London imposed wartime grain import restrictions that blocked Bengal from purchasing rice on the open market. And when the crisis became undeniable in late 1943, Churchill’s response was dismissal. At a November War Cabinet discussion of Bengal, Secretary of State for India Leopold Amery recorded Churchill opening with what Amery described as “a preliminary flourish on Indians breeding like rabbits and being paid a million a day for doing nothing” — then turning to his shipping minister for a view, as if Bengal were a minor scheduling problem. The exchange is documented in Madhusree Mukerjee’s Churchill’s Secret War.

The structural decision was plain: colonial subjects ranked below the European theatre in the resource hierarchy. Defenders of Churchill point to shipping constraints, Japanese naval interdiction in the Bay of Bengal, and grain shipments that eventually arrived. But constraints are the product of priorities. Other theatres received ships. Bengal did not. That wasn’t fate. It was allocation — and allocation is a choice.

The Bengal famine: contested historiography

Churchill's culpability remains actively disputed. The International Churchill Society and the Hillsdale Churchill Project argue wartime shipping losses, Japanese naval activity, and grain that did reach India demonstrate genuine constraints rather than deliberate neglect. Against this: the Mishra et al. finding that no natural drought caused the famine; Amartya Sen's entitlement framework in Poverty and Famines (1981), showing Bengal's aggregate food supply was adequate and the famine resulted from distributional collapse driven by wartime inflation and speculation; and documentary evidence that London repeatedly denied or delayed food import requests while supplying other theatres. The scholarly weight favors policy failure as the primary cause. Whether that failure amounts to deliberate weaponization or catastrophic negligence within wartime constraints remains contested — and politically charged.

The twentieth century proved that modernity did not civilize the weapon. It scaled it. The Hunger Plan killed millions without a commander drawing a blade. The Holodomor starved a nation through extraction quotas and railway schedules. Bengal showed you could kill two million people through the administrative act of not sending ships. Bureaucracy gave starvation something it had never previously possessed: deniability.

Biafra and the birth of the televised famine

Then, for the first time, the world watched.

The Nigerian federal government’s blockade of Biafra beginning in 1967 was comprehensive — seaports sealed, airfields closed, mail intercepted, foreign currency frozen, overland routes controlled by federal troops. As federal forces advanced through 1968, Biafran territory shrank and supply options dwindled to a handful of airstrips that operated under fire. Unlike the Soviet Union in 1932 or the British Empire in 1943, Nigeria’s senior leadership dispensed with euphemism. Federal Commissioner for Finance Chief Obafemi Awolowo and the wartime cabinet were explicit: the strategy was to use control of food and resources to break Biafran resistance, and that strategy was avowed as policy. De St. Jorre, whose account of the conflict drew on access to both sides, and Stremlau, whose study incorporated direct interviews with federal leadership including Gowon, both document that the blockade’s starvation effect was understood and accepted as a deliberate instrument of the war. Federal policy excluded food aid that might prolong Biafran resistance — even supplies clearly marked for civilian relief were blocked on the grounds that they would sustain the enclave.

The killing was vast in scale. Estimates of daily deaths at the blockade’s peak — drawn from Red Cross reports and contemporary journalism from 1968 and 1969 — range from three thousand to ten thousand, the majority children under five. Kwashiorkor, the protein deficiency that produces the distinctive bloated abdomen in starving children, became so widespread that relief workers used it as a triage category rather than a diagnosis. Total civilian deaths from starvation over the course of the conflict: somewhere between half a million and two million.

The range itself tells a story. Nobody was counting precisely, because nobody with the power to stop it considered the count a priority.

What made Biafra different was not the tactic but the medium through which it became visible. Television cameras — still relatively new as a tool for war coverage — brought images of emaciated children into Western living rooms in real time. The distended bellies, the enormous eyes, the listless stares became iconic. They created the visual grammar of famine coverage that persists to this day and provided the founding impulse for Médecins Sans Frontières, established in 1971 by French doctors who had served with the Red Cross in Biafra and concluded that bearing witness without speaking publicly was complicity in the silence.

But the international response to visible mass starvation was not accountability. It was charity. Churches and humanitarian organizations mounted the Biafran airlift — flying food into besieged airstrips at night, under Nigerian anti-aircraft fire, in cargo planes that were often barely airworthy. The airlift was extraordinary as a feat of logistics and courage. It also changed nothing strategically. The blockade continued. The starving continued to starve, though some of them starved slightly more slowly. Nobody delivered justice. Nigeria faced zero legal consequences for a blockade that killed hundreds of thousands of civilians. And it couldn’t have faced consequences — in 1967, no treaty law prohibited starvation as a method of warfare. The government invoked its sovereign right to suppress a secessionist rebellion, and existing international law provided neutral states no basis to break an announced blockade.

The legal reform Biafra eventually provoked — Article 54 of Additional Protocol I to the Geneva Conventions, prohibiting attacks on objects indispensable to civilian survival — arrived in 1977. Seven years after the war ended. After the dead were buried and the cameras left.

A pattern was established that has never broken: visibility produces outrage, outrage produces law, and law arrives too late for anyone it was written about. The dead of Biafra got a treaty provision. The provision did not get them anything.

The paper architecture of prohibition

Biafra’s legacy was supposed to be a legal framework that made starvation impossible to deploy without consequence. What was actually built is worth examining with precision, because the gap between the law’s text and its enforcement architecture is where the weapon survives.

The legal timeline looks impressive on its surface. Article 54 of Additional Protocol I (1977) prohibits attacking, destroying, or rendering useless objects indispensable to civilian survival — crops, livestock, water installations, irrigation works. The Rome Statute’s Article 8(2)(b)(xxv) criminalizes intentional starvation as a war crime in international armed conflicts. Security Council Resolution 2417, adopted unanimously on May 24, 2018, was the first time the Council formally condemned starvation as a method of warfare and explicitly recognized the link between armed conflict and famine. And in December 2019, the Assembly of States Parties unanimously adopted Switzerland’s amendment extending the Rome Statute to cover starvation in non-international armed conflicts — civil wars, insurgencies, the settings in which starvation is most commonly deployed today. The language in the amendment is clear: intentionally using starvation of civilians as a method of warfare by depriving them of objects indispensable to their survival, including willfully impeding relief supplies.

On paper, comprehensively illegal.

In practice, available to anyone powerful enough to use it.

The structural failures are specific and were documented at the time of adoption. First: the 2019 amendment applies only to ICC member states that individually choose to ratify it, and only one year after their ratification. States that decline remain beyond the Court’s reach on this charge. The Opinio Juris analysis published December 7, 2019 — the day after the amendment’s unanimous passage — put the flaw plainly: “A member state would be crazy to voluntarily subject itself to a new crime that it might someday want to commit.” Second: ICC jurisdiction requires either a Security Council referral or crimes on state-party territory. The Council’s five permanent members hold vetoes. Any ally of a permanent member deploying starvation is functionally beyond prosecution, because the Council will not refer its own. Third: proving intent. The legal standard requires demonstrating willful impeding of relief supplies or deliberate deprivation, but blockades and access restrictions are routinely framed as legitimate military operations with unfortunate humanitarian consequences — collateral damage, not strategy. The evidentiary threshold demands internal communications, planning documents, command-level decisions that are, by definition, unavailable during active conflicts.

The German prosecution of former militia members for the starvation siege of Yarmouk camp — the first starvation charge to reach trial anywhere in the world, indictments filed July 2025 — demonstrates the difficulty even under favorable conditions. Against a defeated non-state organization whose documents were captured and whose members could testify, establishing intent still required painstaking reconstruction of the command chain. Against a functioning state with diplomatic leverage and operational security, such evidence is unobtainable.

The result is not weak enforcement of strong law. It is something more specific: mechanisms whose known inadequacies were accepted at the time of adoption. The law was built by the same states whose compliance it would require. The provisions prohibit the practice; the enforcement architecture ensures no powerful state or its ally will face the court that declared it criminal. Not conspiracy — institutional choice, made openly, documented at the time.

Proving "intent to starve": the legal mechanics

International humanitarian law draws a line between siege (potentially lawful if proportionate under IHL), blockade (ambiguous under Protocol I), and deliberate denial of humanitarian relief (clearly criminal but nearly impossible to prove without internal documents). The distinction creates a grey zone that belligerents exploit systematically. A state can close a port, delay inspections on aid shipments, restrict humanitarian corridors one by one — each act claiming a legitimate military purpose. The cumulative effect is mass starvation, visible to satellites and journalists and the United Nations. The intent required for prosecution is visible to no one without access to the room where the decision was made. The Yarmouk prosecution, analyzed by the Lieber Institute at West Point, represents the first attempt to build a circumstantial case for starvation intent against a non-state actor. Whether such a case could ever be constructed against a sovereign state with functioning diplomacy remains untested — and likely untestable.

Yemen, Sudan, and the permanent present

Everything described in the preceding sections is happening now. The weapon is deployed. The law exists. The gap between them has not closed by a single millimetre.

The Saudi-led coalition’s blockade of Yemen, beginning in 2015, closed Sanaa airport, seized the country’s oil-producing regions, and repeatedly shut Hodeidah port — Yemen’s primary food import point, through which roughly seventy percent of the country’s food supply entered. Coalition-imposed delays on UN-verified commercial ships — vessels already cleared for humanitarian cargo — stretched for weeks, with ships sitting idle while food rotted in their holds. By March 2022, according to IPC reporting, over 17.4 million Yemenis experienced high levels of acute food insecurity — in a country of roughly thirty million people. Both sides weaponized food: the Houthis diverted humanitarian aid and hiked food prices to finance their war effort; the coalition attempted to starve Houthi-controlled areas into capitulation.

The UN described Yemen as potentially the deadliest famine in decades.

The accountability outcome: zero prosecutions. The UN Group of Eminent International and Regional Experts documented violations by all parties to the conflict, building a record that could — in theory — have supported future prosecution. In October 2021, under sustained Saudi diplomatic pressure, the Human Rights Council voted not to renew the group’s mandate — 21 states against, 7 abstaining. The only international body investigating war crimes in Yemen was dismantled at the behest of one of the states committing them.

In Sudan, the pattern is more naked still. The Yale Humanitarian Research Lab reported in March 2026, using satellite imagery and NASA Harvest data, that the Rapid Support Forces systematically razed 41 agricultural communities surrounding El-Fasher between March and June 2024 — the opening months of an eighteen-month siege. Ten communities were attacked more than once — one was razed at least seven times. The targets were not military infrastructure. They were farms. Livestock corrals destroyed, farmers displaced or killed, food production capacity eliminated during the precise window when crops needed to go into the ground. Two-thirds of the targeted communities showed no subsequent pattern of life in satellite imagery — depopulated.

The UN Independent International Fact-Finding Mission reported separately in September 2025 that the RSF used starvation as a method of warfare — including deliberate deprivation of food, medicine, and relief supplies — potentially amounting to crimes against humanity including extermination. The Sudanese Armed Forces bear equal responsibility: systematic blocking of humanitarian access, over 84 humanitarian workers killed between April 2023 and April 2025. Over 25 million Sudanese civilians — nearly half the country — face acute food insecurity, according to UN experts reporting in June 2024.

Resolution 2417 exists. The Rome Statute amendment exists. The accountability outcome: zero prosecutions. No referral. No investigation. No arrest warrants.

The UN Human Rights Office called on both parties in Sudan to stop using starvation as a weapon. The call was ignored.

The international response in both Yemen and Sudan follows the Biafra template exactly — humanitarian aid plus legal condemnation, minus enforcement. Statements are issued. Reports are filed. Special sessions are convened. The machinery of outrage works perfectly. The machinery of accountability does not work at all. Fifty-five years of legal development separate Biafra from El-Fasher. Fifty-five years of identical results.

The German officials who gathered on May 2, 1941, were not exceptional in their willingness to deploy hunger as strategy. They were exceptional only in their willingness to write it down. What the post-1977 legal architecture has actually prohibited is not the weapon itself but the candor of those meeting minutes — the naked, minuted documentation of intent that modern prosecution requires for conviction and that modern states have learned never to produce.

The contemporary version operates with better paperwork. Blockades accompanied by humanitarian exceptions that go undelivered. Denial policies reframed as military necessity. Aid corridors opened and closed at tactical intervals — opened when the cameras are present, closed when they leave. Diplomatic pressure that targets not belligerents but the accountability bodies mandated to investigate them. Each individual act deniable. The cumulative effect is mass starvation attributed to complexity, to the fog of war, to the tragic difficulty of delivering aid in conflict zones. The method is the Hunger Plan’s. The paperwork is different.

The problem was never that enforcement has gaps. The problem is that starvation works — has always worked, from Alesia to El-Fasher — and the legal regime built to prohibit it was constructed by states who chose, through the mechanisms they adopted, to ensure that those who deploy it need never face the court that declared it criminal.

The language has changed. “X million will starve” became “we are doing everything we can.”

The arithmetic has not.

Between the 1941 Hunger Plan and a 2024 satellite image of burned Sudanese farmland, the distance is not moral progress. It is paperwork.

Descargo de responsabilidad de Gen AI

Algunos contenidos de esta página han sido generados y/o editados con la ayuda de una IA Generativa.

Medios de comunicación

” Conquest of Lisbon “, by Roque Gameiro, in Quadros da História de Portugal (“Pictures of the History of Portugal”, 1917). – Wikipedia

Principales fuentes y referencias

Alex J. Kay, “Germany’s Staatssekretäre, Mass Starvation and the Meeting of 2 May 1941,” Journal of Contemporary History, Vol. 41, No. 4, 2006.

Timothy Snyder, Bloodlands: Europe Between Hitler and Stalin, Basic Books, 2010.

United States Holocaust Memorial Museum, “Nazi Persecution of Soviet Prisoners of War,” Holocaust Encyclopedia. https://encyclopedia.ushmm.org/content/en/article/nazi-persecution-of-soviet-prisoners-of-war

United States Holocaust Memorial Museum, “The Treatment of Soviet POWs: Starvation, Disease, and Shootings, June 1941–January 1942,” Holocaust Encyclopedia. https://encyclopedia.ushmm.org/content/en/article/the-treatment-of-soviet-pows-starvation-disease-and-shootings-june-1941january-1942

Stanislav Kulchytsky, The Famine of 1932–1933 in Ukraine: An Anatomy of the Holodomor, translated by Ali Kinsella, Canadian Institute of Ukrainian Studies Press, Edmonton and Toronto, 2018.

Omelian Rudnytskyi, Nataliia Levchuk, Pavlo Shevchuk, Alla Kovbasiuk, and Oleh Wolowyna, “Demography of a Man-Made Human Catastrophe: The Case of Massive Famine in Ukraine 1932–1933,” Canadian Studies in Population, Vol. 42, No. 1-2, 2015.

Vimal Mishra, Amar Deep Tiwari, Saran Aadhar, Reepal Shah, Mu Xiao, D.S. Pai, and Dennis Lettenmaier, “Drought and Famine in India, 1870–2016,” Geophysical Research Letters, Vol. 46, No. 4, 2019. https://agupubs.onlinelibrary.wiley.com/doi/full/10.1029/2018GL081477

Madhusree Mukerjee, Churchill’s Secret War: The British Empire and the Ravaging of India During World War II, Basic Books, 2010.

Amartya Sen, Poverty and Famines: An Essay on Entitlement and Deprivation, Oxford University Press, 1981.

John de St. Jorre, The Brothers’ War: Biafra and Nigeria, Houghton Mifflin, Boston, 1972.

John J. Stremlau, The International Politics of the Nigerian Civil War, 1967–1970, Princeton University Press, 1977.

Caesar, De Bello Gallico, Book VII.

Appian, Roman History, Book VI (The Wars in Spain).

Protocol Additional to the Geneva Conventions of 12 August 1949 (Protocol I), Article 54, 1977.

Rome Statute of the International Criminal Court, Article 8(2)(b)(xxv) and Article 8(2)(e)(xix) (2019 amendment).

United Nations Security Council Resolution 2417, adopted May 24, 2018. https://docs.un.org/en/S/res/2417(2018)

“The Rome Statute’s Flawed Amendment Regime — Starvation in NIAC Edition,” Opinio Juris, December 7, 2019. http://opiniojuris.org/2019/12/07/the-rome-statutes-flawed-amendment-regime-starvation-in-niac-edition/

“Prosecuting the Starvation War Crime in Germany: The Yarmouk Case,” Lieber Institute, West Point. https://lieber.westpoint.edu/prosecuting-starvation-war-crime-germany-yarmouk-case/

IPC, “Yemen: 17 million people experiencing high levels of acute food insecurity,” Alert Issue 58, March 2022. https://www.ipcinfo.org/ipcinfo-website/alerts-archive/issue-58/en/

United Nations Human Rights Council, vote on renewal of the Group of Eminent International and Regional Experts on Yemen, October 7, 2021.

Humanitarian Research Lab, Yale School of Public Health, “Rapid Support Forces intentionally targeted agricultural communities to starve people in El-Fasher,” Special Report, March 2026. https://ysph.yale.edu/news-article/hrl-report-rsf-intentionally-targeted-agricultural-communities-el-fasher/

United Nations Independent International Fact-Finding Mission for the Sudan, “A war of atrocities: Sudan civilians deliberately targeted,” September 2025. https://www.ohchr.org/en/press-releases/2025/09/war-atrocities-sudan-civilians-deliberately-targeted-un-fact-finding-mission