Price Controls, Rationing, and the Knowledge Problem: The 1984–85 Famine as Policy Failure

What the cameras saw

In October 1984, the BBC’s Michael Buerk filed a report from Korem in northern Ethiopia that began with the line that has anchored the standard memory of the event ever since: “Dawn, and as the sun breaks through the piercing chill of night on the plain outside Korem, it lights up a biblical famine, now, in the 20th century.”¹ The footage was shocking, the response was massive, Band Aid raised £8 million, USA for Africa followed with “We Are the World,” and roughly $1.7 billion in international relief poured into Ethiopia between 1984 and 1986.²

The death toll, even with the relief, was catastrophic: estimates range from 400,000 to over 1,000,000 dead from famine-related causes between 1983 and 1985, with the central estimates clustering around 600,000.³ What made this number politically explosive then, and historically clarifying now, was not the famine itself — Ethiopia had suffered famines before — but the fact that a government that had taken power on the promise of feeding the country had, ten years on, presided over a famine larger than the one it had displaced its predecessor for failing to address.

The argument of this article is that the 1984–85 famine was a multi-causal event in which drought was the trigger and policy was the accelerant, and that the policy contribution was substantial enough that calling the famine “natural” is dishonest, while calling it purely “man-made” is also dishonest in a different direction. The honest accounting requires separating the causal contributions: how much of the death toll came from the rain that did not fall, and how much from the prices that were not allowed to move, the grain that was procured at gunpoint, the wars that closed the roads, and the resettlement that pulled people from their land. The article makes the case that the policy contribution was at least equal to, and probably larger than, the drought contribution. Both are real. The series is built on the second.

The drought as trigger

The drought was real. Rains failed in significant parts of northern Ethiopia — especially Wollo, Tigray, and parts of Eritrea — through 1982, 1983, and into 1984. The Belg (short) rains of 1984 were poor; the Kiremt (main) rains of 1984 failed catastrophically in the affected areas. Crop production in Tigray fell to approximately 25 percent of normal levels; in Wollo, similar or worse.⁴ This is not a contested fact.

What is contested is what the drought meant for famine. Amartya Sen’s classic argument, developed in Poverty and Famines and applied specifically to the Ethiopian case by Alex de Waal and others, is that famines almost never occur from absolute food shortage alone.⁵ What they occur from is the collapse of entitlements — the ways by which different groups of people get access to food. A pastoralist’s entitlement is his livestock, which he sells to buy grain. A farm labourer’s entitlement is his wages. A smallholder’s entitlement is his crop, plus what he can sell or borrow against. When the entitlement collapses — when livestock prices fall, when wages disappear, when the crop fails and there is nothing to borrow against — the household is in famine even if food is physically available elsewhere in the country.

The Ethiopian famine of 1984–85 was, on this account, a textbook entitlement-collapse famine. There was food in Ethiopia. There was even food in the south of Ethiopia, in significant volumes — the 1984 harvest in the south and west was below normal but not catastrophic.⁶ The famine occurred in the north not because the country lacked food but because the people in the north lacked the means to get to the food. Their crops had failed, their livestock had died, and the system that should have moved food from surplus to deficit areas — the market, the AMC, the road network, the railroads — was broken in specific, identifiable ways.

The drought was the shock. The famine was what the shock interacted with.

What it interacted with: the AMC

The first interaction is the one this series has already named: the Agricultural Marketing Corporation. The AMC’s mandate, established by the 1976 proclamation, was to purchase grain from peasant farmers at fixed government prices substantially below market levels and distribute it through state channels to urban consumers and the military.⁷ The fixed procurement price for teff in 1984–85, as already noted, was 42 birr per 100 kg at the farm level — about a quarter of what the open market would pay.⁸

What this did, predictably, was three things. First, it transferred real income from rural farmers to urban consumers — a redistribution that was central to the Derg’s political strategy of maintaining urban support. Second, it created a permanent disincentive for farmers to produce a marketable surplus, because the surplus would be procured at a loss; this depressed agricultural output in normal years, leaving less buffer for bad years. Third, and most damaging in famine conditions, it required farmers in drought-affected areas to continue meeting their AMC quotas even when their own families were running out of food. The Wikipedia summary, drawn from de Waal’s extensive documentation, is direct: “The very low fixed price of grain served as a disincentive to production, and some peasants had to buy grain on the open market in order to meet their AMC quota. Citizens in Wollo, which continued to be stricken with drought, were required to provide a ‘famine relief tax’ to the AMC until 1984.”⁹

The phrase “famine relief tax” deserves attention. It captures the perversity of the system precisely. The Derg’s policy, in drought-affected Wollo, was to extract grain from peasants who were running out of food, in the form of a tax payable in grain, to fund relief operations elsewhere. The grain extraction continued, in some areas, until late in 1984 — after the famine had already begun, after the international press was filming the death camps. The policy did not adjust because the policy’s information-system had no good way to know that adjustment was needed; the AMC bureaucracy was structured to deliver quotas, not to receive feedback from the people the quotas were destroying. This is the Hayekian knowledge-problem in its most lethal form.

What it interacted with: the wars

The second interaction is war. In 1984–85, the Derg was fighting active counter-insurgencies in Eritrea (against the EPLF), in Tigray (against the TPLF), and in lower-intensity engagements across multiple other regions. The wars in the north overlapped almost exactly with the drought-affected areas. They affected famine outcomes through several channels.

Roads and railways in the war zones were disrupted, in some cases deliberately. The main road from the Red Sea ports (Massawa, Assab) into the highlands ran through Eritrea and Tigray, exactly where the fighting was; food aid that arrived at the ports could not be reliably moved inland through these areas. The Derg’s military operations in Tigray included episodes that have been documented by human-rights organisations as deliberate destruction of food sources — burning of crops, killing of livestock, restrictions on humanitarian access to TPLF-held areas.¹⁰ Both the Derg and the TPLF have been charged with using food as a weapon during this period, and the charges have credible documentary support on both sides.

The Derg’s resettlement programme, initiated in late 1984, compounded the disaster. Approximately 600,000 people were moved — often forcibly — from drought-affected northern areas to government-held areas in the south and west, on the theory that resettling them on better land would solve the famine.¹¹ The implementation was catastrophic. Conditions in the resettlement camps were poor; mortality rates in transit and on arrival were high; many people died during the resettlement who would have survived if left in place to receive food aid. Estimates of resettlement-related deaths range from 50,000 to over 100,000.¹²

And the villagisation programme, accelerated during the famine period, forcibly relocated peasants from dispersed homesteads into planned villages, on the theory that concentrated populations were easier to provide with services. The disruption to agricultural production in 1985 and 1986 was substantial, contributing to a slow recovery from the famine in many areas.¹³

The Sen test: was the food there?

The decisive test, in the Sen-de Waal framework, is whether food was physically available somewhere in the country at the height of the famine. The answer, repeatedly confirmed in subsequent studies, is yes. Total Ethiopian grain production in 1984 was reduced but not catastrophically; the southern surplus alone could, in principle, have fed the northern deficit if moved.¹⁴ The food aid that arrived from abroad in 1984–86, totalling over 1 million tonnes of grain, was sufficient to feed the famine-affected population — if it could be moved to them.¹⁵

The movement is what failed. The market, suppressed by AMC monopoly and price controls, could not move it. The state, distracted by war and committed to extracting grain from the famine areas, did not move it in time. The transport infrastructure, damaged by war and underinvested in by a regime that had not built rural roads, could not move it. Food sat in Addis warehouses while people died in Wollo camps. This is the empirical signature of an entitlement famine, not an absolute-shortage famine.

The honest counter-case

A serious case must engage the strongest version of the opposing reading, which is the Derg’s own: that the famine was caused by drought and by Western indifference, and that the regime did what any government would have done given the constraints it faced. The argument has three components: that the drought was historically severe; that Western donors deliberately delayed aid to weaken the regime; and that the AMC and the wars, whatever their costs, were responses to existential threats the regime was right to take seriously.

Each component has some truth. The drought was severe by recent historical standards (though not unprecedented in the Ethiopian record). Some Western governments, particularly the Reagan administration in its early phase, were demonstrably reluctant to provide major aid to a Soviet-aligned regime; this delay cost lives.¹⁶ And the regime did face genuine threats from internal insurgencies and from neighbouring states.

The series’s reply is not to deny any of this, but to insist on the multi-causal accounting. The drought, the Western delay, and the security threats are contributing causes. The AMC’s procurement-at-gunpoint, the resettlement programme, the wartime restrictions on humanitarian access, and the continued military expenditure during the famine are also contributing causes, and they are causes that resulted directly from the regime’s policy choices. To attribute the famine entirely to drought or to Western policy is to make the choices of the Derg’s planners disappear from the account, and they did not disappear from the deaths.

The 2007 Red Terror Trials in Ethiopia explicitly addressed the question of whether the famine was a crime, charging senior Derg officials including Mengistu with offences under the Ethiopian Penal Code that included “intentionally creating a grave state of misery, want or famine.”¹⁷ The trials, whatever else they were, established a public record that Ethiopian courts considered the famine an event for which specific policy choices were criminally responsible. Mengistu was convicted in absentia. The conviction does not settle the historical question, but it does establish that the question is not the invention of Western critics.

The connection to money

This article has so far written about grain prices, procurement quotas, and resettlement, with money offstage. The connection deserves explicit naming. The AMC’s losses were funded by state-bank credit, which is to say by central-bank advances. The military expenditure that absorbed 46 percent of the government budget was funded the same way. The state-enterprise sector that was haemorrhaging losses through this period was funded the same way. The accumulating fiscal deficit was, in every meaningful sense, the monetary deficit — the gap between what the state was spending and what its revenues could fund was filled by the printing press.

This is why the 1984–85 famine belongs in a monetary history, not just a social or political one. The fixed prices that destroyed the AMC’s procurement system were the same fixed prices that suppressed the inflation signal in the official statistics. The deficit that financed the AMC’s losses was the same deficit that fed the parallel exchange rate’s accelerating premium. The wars that closed the food routes were the same wars that absorbed the fiscal expenditure that produced the monetary expansion. The famine was, in a precise causal sense, the real-economy expression of a monetary disorder that the official numbers had been pretending did not exist.

The next article turns to the place where the disorder was visible — the parallel exchange rate — and what it was telling anyone willing to look.

---

Footnotes

1. Michael Buerk, BBC News report from Korem, broadcast 23 October 1984. Transcript widely reproduced; see Suzanne Franks, Reporting Disasters: Famine, Aid, Politics and the Media (Hurst, 2013). ↩

2. David Rieff, A Bed for the Night: Humanitarianism in Crisis (Simon & Schuster, 2002), chapter 3, on the 1984–86 Ethiopia aid response. ↩

3. Alex de Waal, Evil Days: Thirty Years of War and Famine in Ethiopia (Human Rights Watch, 1991), 4–5, gives the 400,000–1,000,000 range. Mesfin Wolde-Mariam’s later estimates and the work of the Tigray Health Bureau cluster around the higher end of that range. ↩

4. Alex de Waal, Famine Crimes: Politics and the Disaster Relief Industry in Africa (James Currey, 1997), chapter 5. ↩

5. Amartya Sen, Poverty and Famines: An Essay on Entitlement and Deprivation (Clarendon Press, 1981); de Waal, Famine Crimes, chapter 1. ↩

6. De Waal, Evil Days, chapter 2, documenting that production in southern and western Ethiopia in 1984 was reduced but not catastrophic, while production in the northern famine areas had collapsed. ↩

7. Proclamation No. 105/1976. Discussed in Dessalegn Rahmato, Agrarian Reform in Ethiopia (Scandinavian Institute of African Studies, 1984). ↩

8. “Agriculture in Ethiopia,” Wikipedia, https://en.wikipedia.org/wiki/Agriculture_in_Ethiopia. ↩

9. “1983–1985 famine in Ethiopia,” Wikipedia, https://en.wikipedia.org/wiki/1983%E2%80%931985_famine_in_Ethiopia. ↩

10. De Waal, Evil Days, chapters 3 and 7, documenting both sides’ use of food as a weapon. African Rights, Food and Power in Sudan: A Critique of Humanitarianism (African Rights, 1997), chapter 2, contextualises the Ethiopian case. ↩

11. Jason W. Clay and Bonnie K. Holcomb, Politics and the Ethiopian Famine, 1984–1985 (Cultural Survival, 1985); Jason W. Clay, Sandra Steingraber, and Peter Niggli, The Spoils of Famine: Ethiopian Famine Policy and Peasant Agriculture (Cultural Survival, 1988). ↩

12. De Waal, Evil Days, 211–227, on resettlement mortality estimates. ↩

13. Christopher Clapham, Transformation and Continuity in Revolutionary Ethiopia (Cambridge University Press, 1988), 184–192, on villagisation. ↩

14. De Waal, Evil Days, chapter 2; Sen, Poverty and Famines, applied to Ethiopia. ↩

15. Rieff, A Bed for the Night, chapter 3; Edward Clay and Olav Stokke, eds., Food Aid and Human Security (Frank Cass, 2000), chapter 4. ↩

16. David A. Korn, Ethiopia, the United States, and the Soviet Union (Croom Helm, 1986), chapter 7. ↩

17. Federal High Court of Ethiopia, Red Terror Trial judgement, 12 December 2006; see Firew Kebede Tiba, “The Mengistu Genocide Trial in Ethiopia,” Journal of International Criminal Justice 5, no. 2 (2007): 513–528. The famine-related charges are discussed in Marshet Tadesse Tessema, “Ethiopia’s 1984/85 famine and the Red Terror Trials,” Third World Quarterly 44, no. 4 (2023). ↩