Click to download report.

Alex de Waal

War and famine, two fearsome horsemen, have long ridden side by side. Armed conflict disrupts food systems, destroys livelihoods, displaces people, and leaves those who do not flee both terrified and unsure when they will eat their next meal.

News stories and scholarly articles on conflict and hunger are usually pessimistic, assuming both are inevitable parts of the human condition. But a review of trends offers cause for optimism: a potential end to famine and conflict-induced starvation by 2030. This chapter examines those trends, identifies vulnerable populations, explores the complex relationship between conflict and hunger, and underscores what must be done to eliminate famine for good.

Invisible Victims

When famine or acute hunger occurs today, it is usually the result of armed conflict. The best estimate for the number of people currently affected by conflict is 172 million (CRED 2013). Although refugees are more visible, 87 percent of those affected by conflict are actually residents who do not flee their homes—and who tend to fare even worse than those displaced (CRED 2013). Beyond the reach of aid agencies, they suffer in silence.

Victims of violence in ostensibly peaceable countries are a more vast and less visible group of hungry people. They include victims of violent crime, gang violence, brutality by state enforcement bodies, and intimate partner violence—which together account for the vast majority of violence globally (Geneva Declaration 2011). Of the estimated 780,000 people who died worldwide from violence and its immediate effects each year between 2004 and 2009, 66 percent were killed in nonconflict settings (mainly due to crimes), 27 percent died from hunger and disease due to conflict, and just 7 percent died as a direct consequence of war. Of the 14 countries with annual rates of violent deaths of more than 30 per 100,000, just six were engaged in war. The other eight—with El Salvador topping the list—suffer high rates of violent crime. The impact of all of these forms of violence on development is major and severe; their victims are poorer, more vulnerable, and hungrier than others (World Bank 2011).

Survivors of war are another under-recognized population that is vulnerable to food insecurity. Violent deeds live on, not only in the psychological trauma suffered by survivors and their family members, but also in basic well-being. Recent studies from Uganda on the long-term impact of war wounds and trauma show that affected households are hungrier, sicker, and less well off than others (Mazurana et al. 2014). Meeting the needs of survivors constitutes another vast and often overlooked welfare and food policy challenge.

The End of Calamitous Famines

While more must be done to address the unique situations these invisible groups face, great progress has been made. Yet we are often so focused on the problems of the present, it is easy to overlook vast changes that have occurred over the long term. For example, the historic declines in all kinds of violence (Pinker 2012) and the reduction in the frequency and lethality of armed conflict (Human Security Report Project 2013) are often obscured by crises of the moment.

Much the same is true for famine. Indeed it is all too easy to overlook historic, but unheralded achievements of the last 50 years: the elimination of calamitous famines (those that cause more than 1 million deaths) and the reduction almost to a vanishing point of great famines, or those that cause more than 100,000 deaths (Howe and Devereux 2004).

Until the middle of the 20th century, the drumbeat of starvation was constant, with millions dying every decade. Between 1870 and 2014, 106 episodes of famine and mass starvation each killed 100,000 people or more (Mallory 1926; Newman 1990; Devereux 2000; Dyson and Ó Gráda 2002).

The trends are striking (Figures 3.1 and 3.2). During the 20th century, the death toll from great famines zigzagged, ranging from a 10-year high of 27 million in 1900–1909; to more than 15 million in each of the 1920s, 1940s, 1950s, and 1960s; to a low of 1.4 million during the 1990s. In the 21st century thus far, the death toll is near 600,000.

Click to enlarge.

Taking a closer look at the history behind the graphs, we see famines associated with the Age of Empire from the 1870s to World War I (Hobsbawm 1989). Famines killed tens of millions in South Asia and China, millions in Africa, and smaller numbers in Brazil. The causes: drought and havoc wreaked by imperial conquest and predation, including practices such as dismantling local production systems and imposing a regime of forced labor to produce export crops such as rubber and cotton. With the passing of the most ruthless era of imperial expansion, these famines, also known as “Late Victorian Holocausts,” ceased (Davis 2002).

During what historian Eric Hobsbawm (1996) called the “Age of Extremes” from World War I to the end of the Cold War, calamitous famines were caused by totalitarian systems: German and Japanese militarism, Stalinism, and Maoism. Wartime leaders routinely used starvation as a weapon.

Forced collectivization in Ukraine and southern Russia in 1932–1933—a possibly genocidal campaign known to Ukrainians as the “Holodomor”—was perhaps the most terrible example of famine as state policy (Conquest 1987). Had the Nazi Hunger Plan to starve 20–30 million Belorussians, Poles, and Ukrainians been fully carried out, it would have been worse still (Lowe 2012). Asian war famines killed many millions from 1936 to 1945 in Bengal, China, Indonesia, and Vietnam.

After World War II, Communist policies caused horrific famines. Thirty million people died in the Chinese famine of 1958–1962, caused by Mao Zedong’s Great Leap Forward (Becker 1996). The Khmer Rouge starved 1.5 million Cambodians in the 1970s (Kiernan 2008). These calamitous famines ended along with “faminogenic” regimes, such as totalitarian governments and wars of extermination (Marcus 2003).

Click to enlarge.

The last great Communist famines were in Ethiopia in 1983–1985, when collectivization and hunger as a weapon of war collided with drought, killing up to 1 million (de Waal 1997), and in North Korea in 1996–1997, when a food crisis killed 500,000–600,000 (Goodkind, West, and Johnson 2011).

In the 20th century, Europe and Asia accounted for the vast majority of famine deaths (Figure 3.2). Only two African famines in the last 100 years—Biafra and Ethiopia—have killed as many as 1 million each. Since famine has disappeared from Europe and mostly vanished from Asia, it has lost most of its menace.

And finally, the downward slope of the famines graph (Figure 3.3) contrasts with the upward slope of world population, which rose from about 1.7 billion in 1900 to 7.3 billion today. This surely refutes the pessimism of the early 19th century scholar and cleric Reverend Thomas Malthus, who feared that world population was outpacing the food supply. More than two centuries ago he wrote that “gigantic inevitable famine stalks in the rear [of population growth], and with one mighty blow, could level the population” (Malthus 1798, 140). In fact, the converse is actually happening.

Positive Developments

The end of the Cold War, the adoption of international human rights norms, and the rise of globalization are among the key factors that make it possible to eliminate famine for the first time in history. Governments no longer wield the grotesque sovereign privilege to starve their people and tell the rest of the world to mind its own business. Unparalleled global prosperity and interconnectedness, the legitimacy of international concern over domestic violations, and far more information-sharing mean people are less likely to starve in silence because their rulers, or the international community, do not know what is going on.

And the single most important reason an end to famine is within reach? China, once the “land of famine” (Mallory 1926), which suffered more than 80 million famine deaths between 1870 and 1970, or more than half of the global total of 149 million such deaths (World Peace Foundation 2015), has been free of that scourge for half a century.

Most trends are moving in the right direction. In 2013, the Centre for Research on the Epidemiology of Disasters (CRED) reported “encouraging news in terms of lower death rates, indicating that periods of stabilization and humanitarian efforts have succeeded in saving lives” (CRED 2013, 23–24).

Warning Signs

Malnutrition trends, however, have not been so favorable, with global levels of acute malnutrition (wasting or nutritional edema) rising since 2008, according to CRED’s People Affected by Conflict report. This recent upswing coincides with another worrying trend: the reduction in wars has also stalled (Apps 2015; PS21 2015). According to the global think tank, Project for the Study of the 21st Century, the number of conflicts and conflict-related deaths has increased from an all-time low in 2006, but remains well below long-term averages. Between 2013 and 2014, the 20 most conflict-afflicted countries saw violent fatalities rise by 28.7 percent, from 127,134 to 163,562. Syria is by far the largest contributor, with more than 70,000 deaths in that one year alone. These numbers are still low by historic standards, but they show that much more must be done to win the battle against war and hunger.

New Wars, New Famines

Today’s famines are “complex humanitarian emergencies,” caused mostly by armed conflict and exacerbated by natural disasters or international policies (Keen 2008). These “new wars” (Kaldor 1999) involve not only state armies and insurgents, but also paramilitaries and ethnic militia, criminal gangs, mercenaries, and international forces. Most new wars are civil wars, which increasingly spill over borders, disrupt livelihoods and food systems, and force people to flee. They tend to be less lethal than old wars, both in violence and in hunger (Human Security Report Project 2013). But they are often intractable and display persistent, seemingly patternless violence from which no one is safe.

In previous eras, governments and rebels controlled humanitarian access. They either permitted it and protected aid workers, or they blocked access. Today, humanitarian workers face greater personal dangers as they navigate a more dangerous micro-terrain of warfare, village by village. Under these circumstances, getting food aid to those in need demands exceptional skills, and the riskier conditions can result in “new famines” (Devereux 2007). A selection of cases illustrates how such famines arise:

  • Sudan. In 2003–2004, armed conflict between the Sudanese military and various rebel groups in Darfur led to an estimated 200,000 civilian deaths through hunger, disease, and displacement (US GAO 2006). An extreme case occurred in the small village of Keilak in April 2004, where a visiting UN team found that overall death rates were more than 40 times the standard threshold for declaring an emergency. This episode was thankfully isolated. And as soon as the UN raised the alarm, the local commander’s siege was ended and aid provided.
  • Somalia. The famine in Somalia in 2011–2012 was the worst of the century thus far, costing an estimated 250,000 lives (Maxwell and Majid 2015). It was a complex emergency: drought, economic crisis, and war all contributed. Politics complicated the dynamics of the hunger situation. Humanitarian agencies had to contend not only with restrictions on access and the risks of kidnapping and violence, but also US counterterrorism legislation impeded their operations in insurgent-controlled areas. The UN did not raise the alarm until famine conditions were already widespread. In 2014–2015, similar factors resulted in acute hunger in the parts of Syria and Iraq that are controlled by the Islamic State, and the starvation of the Yazidi minority.
  • Democratic Republic of the Congo. Since 1996, wars in the Democratic Republic of the Congo have led to an immense humanitarian disaster, with estimates of deaths ranging as high as 5.4 million people (International Rescue Committee 2008). These deaths are overwhelmingly caused by hunger and disease associated with the collapse of health services and other basic infrastructure, along with disruptions to employment and food markets.
  • Iraq. In the 1990s the Iraqi people suffered from a lethal combination of Saddam Hussein’s depredations, comprehensive sanctions, and Hussein’s use of the food rationing system to reward his loyal followers and thereby maintain his power base (Alnasrawi 2000). Between 250,000 and 500,000 children died of hunger and disease (UNICEF 1999).

Even when food was used as a weapon, recent conflicts did not necessarily result in major famine. For example, in its final offensive against the Tamil Tigers in 2009, the Sri Lankan government withheld aid to a besieged and hungry civilian population (International Crisis Group 2010). Nonetheless, the government’s final victory was swift and the possibility of a protracted Biafra-style siege was foreclosed. At its peak in 2008–2009, the Israeli siege of Gaza involved tight control of basic supplies to the territory. The UN was permitted to transport just a fraction of the assistance it considered necessary for humanitarian purposes (Cook 2012). The deprivation suffered by the Gazans was extreme, but apparently stopped short of crossing the line into mass starvation. Dov Weisglass, an adviser to Ehud Olmert, the Israeli Prime Minister, reportedly said, “The idea is to put the Palestinians on a diet, but not to make them die of hunger” (Urquhart 2006).

Does Hunger Lead to Conflict?

Conflict and hunger are strongly associated. The countries with the lowest levels of food security, according to the 2014 Global Hunger Index, are engaged in or recently emerged from war, including Burundi, Comoros, Eritrea, Sudan and South Sudan, and Timor Leste (von Grebmer et al. 2014).1, 2 More strikingly, while most countries have marked significant 25-year gains in objective hunger measures, a handful have stagnated or deteriorated—notably Burundi, Comoros, Iraq, and Sudan (all conflict-affected) and Swaziland, scarred by the world’s worst HIV/AIDS epidemic (Whiteside and Henry 2011). In contrast, hunger is retreating in Ghana and Rwanda, which are relatively peaceful.

It is clear that conflict is the main cause of persistent severe hunger. Might hunger—whether in the form of famine, chronic malnutrition, or general deprivation—also be a factor that drives conflict? Possibly, but it’s less likely.

Synthesizing more than a decade of controversy about the causes of armed conflict—a debate widely known as “greed or grievance” (Collier and Hoeffler 2004)—the World Bank’s 2011 World Development Report concluded that there is no simple causal explanation for conflict (World Bank 2011). Conflict has many sources (Box 3.1). Many economic factors make countries vulnerable to lapsing into civil war. The good news is that as governance improved over the past few decades, conflict, poverty, and hunger have all consistently declined. Unfortunately, the trajectories have been uneven. Worse yet, more recent evidence suggests that progress has stalled.

Box 3.1 Multiple Sources of Conflict

The causes of conflict are complex, nonlinear, and mediated by a host of factors, including political institutions and economic structures (World Bank 2011). A spectrum of threats to human security—assaults on community, livelihoods, and basic welfare—provide a common thread that explains why people resort to violence. Hunger certainly figures among the causes, but how it plays out varies by context.

Such complexity means we must be exceptionally careful when drawing conclusions about any one driver of conflict. In countries such as Afghanistan, Colombia, Congo, Sudan, or Yemen, patterns of violence are turbulent. Like the waters of a fast-flowing mountain stream, they are chaotic from one moment to the next, but retain a recognizable structure over time. If a researcher takes the data for violence in any particular country over any particular period of time, feeds them into a computer, and looks for correlations with weather patterns, market prices, malnutrition levels—or any other indicator—some correlation will always arise. However, most of these associations fade away upon closer scrutiny (Buhaug et al. 2014). Too often, though, researchers’ disclaimers about the uncertainty of their findings are set aside when their findings are synthesized or popularized.

While major famine and war have virtually disappeared in East Asia and Southeast Asia, the Middle East has seen an increase in both armed violence and hunger over the last five years. Africa—the poorest and most conflict-prone continent—faces the greatest risk. Many affected countries are prone to authoritarianism and violent competition, due to “resource curse” economies that rely heavily on the exploitation of natural resources, especially minerals, to speed up economic growth (Kaldor, Karl, and Said 2007). The “great African land-grab” (Cotula 2013)—in which local elites and foreign corporations are taking the land from millions of smallholders—is contributing to deep human insecurity and grievance, which has led to both nonviolent and violent resistance in countries as diverse as Ethiopia and Sierra Leone.

Hunger's Destabilizing Effect

Hunger is somehow different from other human stresses. Food and famine strike a deep emotive chord, even among people who have never personally faced starvation. Around the world, people believe that a government that cannot feed its people has forfeited its legitimacy.

High bread prices famously brought out revolutionary mobs in Paris in 1789 (Grove 1998; Neely 2007). The 1943 Bengal famine discredited the British Raj, undermining its promise to prevent starvation (Drèze 1991). Famine in Wollo Province undermined the rule of Ethiopia’s Emperor Haile Selassie in 1973–1974, and Sudanese President Jaafar Nimeiri’s failure to provide drought relief in 1985 helped bring down his government (Article 19 1990; de Waal 1997). After Cyclone Bhola, the deadliest storm in the last 100 years, struck East Bengal in 1970, the slow and inadequate response of Pakistan’s Ayub Khan government to hunger and deprivation helped mobilize the Bangladesh independence movement (Sommer and Mosley 1972; Hossain 2010).

More recently, food protests helped bring down the Haitian government in 2008, and food price rises coincided with protests during the Arab Spring of 2011 (Brinkman and Hendrix 2011). Protests are more likely to lead to political violence in fragile states (World Bank 2011). The pathways from food scarcity to protest are complex and unique to each case, but a common thread runs through these examples. Food security is not only an essential component of human well-being, but also a foundation for political stability. Governments jeopardize food security at their peril.

Standing Firm

Despite failing to keep hunger at bay, some governments carry on regardless. The Myanmar government presided over devastation caused by Cyclone Nargis in 2008, which drowned an estimated 138,000 people and left nearly 2 million without shelter, drinking water, or basic food items (Guha-Sapir and Vogt 2009). Fearing an increased international presence during a referendum on the new constitution, Myanmar refused to initiate or allow substantive aid into the country for two weeks (Zarni 2015). Rulers such as China’s Mao Zedong, North Korea’s Kim Jong-Il, and Ethiopia’s Mengistu Haile Mariam stayed in power, oblivious to human suffering. They even used deprivation and controlled food supplies to consolidate power (Becker 1996; Natsios 2001; de Waal 1997). While failing to address severe hunger and deprivation does not necessarily lead to the downfall of governments, a sound food security policy is good political insurance.

From these threads we can underline the generally positive conclusion that hunger may pose less of a threat to peace than it has in the past. And there is no reason why natural disasters must cause either famine or political crisis (Box 3.2).

Box 3.2 Global Warming, Conflict, and Hunger

There is no question that climate-related disasters are increasing in number and severity (Guha-Sapir, Hoyois, and Below 2014). As a consequence, should we expect to see more conflicts, and therefore more hunger?

One area in which methodological sobriety is much needed is conflict and the environment. The big picture on climate change, hunger, and conflict is encouraging thus far. Over the past five decades, even as global warming has advanced and natural disasters have become more frequent and damaging, both wars and hunger have decreased.

At the country level, despite fears that the coming century would be marked by “water wars,” transboundary river basins ranging from the Indus to the Jordan, in fact, show more cooperation than conflict (Islam and Susskind 2013). The March 2015 agreement between Ethiopia, Egypt, and Sudan that defined principles for sharing the Nile waters illustrates this positive turn of events.

At the local level, the picture is not encouraging. East African data show that extreme rainfall variation in either direction—both too much or too little—increases conflict risks (Raleigh and Kniverton 2012). Fluctuations in livestock prices and changes in local seasonal migrations, which are both influenced by rainfall, are associated with risks of violence (Maystadt, Calderone, and You 2014; Maystadt and Ecker 2014). But to extrapolate these findings to wider predictions is not warranted. The effects are arguably well within the turbulent parameters of normal local conflict and should not be seen as harbingers of bigger wars to come.

Recent attempts to identify climate change as a driver of large-scale armed conflict have been effectively criticized, with the plea that connections are complicated (Raleigh, Linke, and O’Loughlin 2014). Recent studies disagree on both the magnitude of the impact of climate change on conflict and the direction of the effect. One noted that “research to date has failed to converge on a specific and direct association between climate and violent conflict” (Buhaug et al. 2014, 394–395).

To understand the dynamics of extreme weather and conflict, we should not lose sight of the role of political leaders. For example, did the protracted drought in Syria from 2006–2010 help spark the conflict that erupted in 2011? One researcher, Francesca de Châtel, argues that government policies, including bureaucrats’ long-term mismanagement of natural resources, were to blame (de Châtel 2014). Their focus on dam construction and irrigation projects in the northeast meant that small farmers in that area were neglected, impoverished, and angry.

Another problem was the government’s failure to respond to the humanitarian crisis and food price increases. This was one of many grievances that triggered protests in March 2011. De Châtel vigorously challenges those who see a strong causal link, writing: “The possible role of climate change in this chain of events is not only irrelevant, it is also an unhelpful distraction and a damaging alibi for the Assad regime’s failings” (de Châtel 2014, 532). Other climate researchers’ findings on Syria detail the gravity of drought and groundwater depletion, and suggest that these elements might have contributed to the 2011 unrest (Kelley et al. 2015).

The relationship between environmental disasters and conflict is highly nuanced. According to researchers at London’s Overseas Development Institute, “natural disasters reduce some conflict drivers while exacerbating others” (Harris, Keen, and Mitchell 2013). Disaster and government failures to respond can deepen existing societal tensions, while disruption presents economic opportunities for criminal activities. Crisis often offers opportunities to advance militant or divisive political agendas. And disasters can make conflicts more likely by altering the balance of power or a warring party’s ability to misappropriate aid.

Looking Ahead

The last decades of the 20th century saw the end of calamitous famines, which kill one million or more people. But what will it take to eliminate famine and acute hunger, or starvation, by 2030?

Two tasks stand out for eliminating conflict-related hunger. First, we need stronger mechanisms to prevent and resolve conflicts. With wars becoming fewer and less lethal, the long-term trends for violent conflict are actually encouraging (Human Security Report Project 2013). But progress appears to have stalled, and the challenges of the day—for example in South Sudan, Syria, and Yemen—are formidable. Second, we must activate the international emergency relief system to dispatch large-scale food aid where it is needed most.

Political Commitment

We need political leadership to strengthen international food security policy. This depends on political decisions in western capitals—and these are not always easy to make. While the United Nations and the European Union can do much to unlock a humanitarian response, the role of the US government is pivotal, especially in politically controversial cases. With its continuing cereal surpluses, which are the backbone of global food aid, its agenda-setting role at the UN Security Council, and its power to impose financial and legal sanctions on those who violate its counterterrorism legislation, the United States remains the “veto holder” with regard to global famine response.

When the first signs of famine in North Korea became apparent to the world in 1997, a vigorous debate took place in US newspapers. Some argued that it would be wrong to condition aid on policy change by the regime, notwithstanding its culpability for the famine. Others implicitly made the case for starving North Korea into collapse, claiming that aid would ultimately be used to support the military apparatus of a government hostile to both its own people and to the United States. In the pro-aid group, Andrew Natsios, who later became administrator of the US Agency for International Development (USAID) in 2001–2006, observed that US aid helped make the regime more open to the international community and that no totalitarian dictator has ever been overturned during or after a famine (Natsios 2001).

Natsios was tasked with making good on President George W. Bush’s promise of “no famine on my watch.” Perhaps USAID’s most remarkable but under-recognized act in those years was initiating a relief program for Darfur in September 2003—six months before the humanitarian crisis became headline news. Natsios made this decision well aware that he would be open to the same critique made when the United States aided North Korea. Regardless, he did the right thing. Food aid undoubtedly saved many thousands of Darfurian lives.

While the United Nations and powerful governments can predict and stop major food crises, ultimately the decision is always political. Faced with an imminent famine in Somalia in 2011, the US government failed to override its antipathy to al-Shabab and waited until famine was well advanced to authorize assistance (Maxwell and Majid 2015). Behind the scenes, US counterterrorism legislation made it impossible for UN agencies and nongovernmental organizations to operate in al-Shabab-controlled areas. To do so would be to risk being accused of supporting a terrorist organization. Only when the United Nations moved to declare famine in Somalia was the United States ready to respond and allow others to do so without automatically running afoul of its prohibition on supporting terrorism.

The lesson is clear: Political commitment at the highest levels to prevent famine, no matter what the political context, is needed. Countries in need should be aided, regardless of their standing with any other government.

In closing, while the elimination of calamitous famines and the foreseeable end of great famines are tremendous achievements, the work of conquering acute and chronic hunger is not finished. Economic development, better food policy, conflict resolution, and international humanitarian response will all continue to play roles in this venture. Unless armed conflicts can be reduced—and preferably ended—and the many invisible victims of violence can be reached with better humanitarian action and welfare policy, the gains will not last.

Note: The views expressed in this chapter are those of the author. They do not necessarily reflect the views of IFPRI, Welthungerhilfe, or Concern Worldwide.


  1. While the analysis in previous GHI reports focused on comparisons with hunger levels in 1990, the analysis in this year’s report centers on comparisons with hunger levels in 2000. Many countries experienced fluctuations between 1990 and 2015, and making comparisons with 2000 captures more recent trends. Back
  2. The regional and global aggregates for each component indicator are calculated as population-weighted averages, using the indicator values reported in Appendix B. Provisional estimates for under-nourishment for Burundi, Comoros, the Democratic Republic of the Congo, Eritrea, Libya, Papua New Guinea, Somalia, and Syria were used in the calculation of the global and regional aggregates only, but are not reported in Appendix B. These estimates are based on previously published undernourishment data and provisional estimates provided by FAO in 2014 for the sake of regional and global aggregation only. The regional and global GHI scores are calculated using the regional and global aggregates for each indicator and the revised formula described in Chapter 1. Back