Was the 20th century the most violent in history?

The numbers are staggering. World War I killed roughly 20 million people. World War II killed roughly 70 million, including 6 million Jews in the Holocaust. Stalin's purges and forced famines killed an estimated 15-20 million Soviet citizens. Mao's Great Leap Forward produced a famine that killed 30-45 million Chinese. The killing fields of Cambodia claimed 1.5-2 million. The Rwandan genocide killed 800,000 in a hundred days. Add the Korean War, the Vietnam War, the Bangladesh genocide, the Nigerian Civil War, and dozens of smaller conflicts, and the twentieth century's death toll from violence and politically caused famine reaches well over 200 million people.

It is natural, confronted with these figures, to conclude that the twentieth century was the most violent in human history. Many historians and public intellectuals have done exactly that. The British historian Eric Hobsbawm titled his history of the twentieth century The Age of Extremes. The theologian Hans Kung called it "the most violent century in human history." The assessment seems obvious.

But is it correct? The answer depends entirely on how you measure violence -- and the measurement choices themselves reveal deep assumptions about what counts, who counts, and what kinds of violence we consider significant.

Absolute Numbers vs. Per Capita Rates

The most straightforward way to assess the violence of a century is to count total deaths. By this measure, the twentieth century almost certainly wins. There were simply more people alive to be killed. The world's population grew from roughly 1.6 billion in 1900 to 6.1 billion in 2000, meaning there were far more potential victims than in any previous century. The absolute numbers of war deaths, genocide victims, and politically caused famine deaths are unmatched.

But many demographers and historians argue that absolute numbers are a misleading measure. A conflict that kills 1 million people in a population of 10 million is proportionally far more devastating than one that kills 10 million in a population of 1 billion. If we measure violence as a rate -- deaths per capita, or as a proportion of the population -- the picture shifts significantly.

By per capita measures, the twentieth century was violent but not uniquely so. The Mongol conquests of the thirteenth century killed an estimated 30-40 million people, a staggering proportion of the world's population at the time -- possibly as high as 10-15% of all humans alive. The An Lushan Rebellion in eighth-century China may have killed 36 million people, representing an enormous proportion of the Tang Dynasty's population. The Taiping Rebellion in nineteenth-century China killed 20-30 million. The Atlantic slave trade killed an estimated 10-15 million people over four centuries, a cumulative toll that devastated entire regions of West and Central Africa.

Steven Pinker, in The Better Angels of Our Nature (2011), uses per capita rates to argue that violence has actually declined over the long arc of human history. Pinker compiles data from archaeological, historical, and modern sources to show that the likelihood of a person dying violently was far higher in pre-state societies than in the modern world. Hunter-gatherer societies, he argues, had homicide rates an order of magnitude higher than even the most violent modern nations. Tribal warfare, raiding, and feuding produced steady attrition that, measured as a proportion of population, far exceeded the death rates of twentieth-century warfare.

The Critique of Pinker's Thesis

Pinker's argument has been influential and controversial in roughly equal measure. His critics have raised several significant objections.

First, the archaeological and ethnographic data on pre-state violence is sparse and contested. Much of Pinker's evidence for high rates of violence in pre-state societies comes from a relatively small number of archaeological sites and ethnographic studies of contemporary hunter-gatherer groups, many of whom had already been affected by contact with state societies. Extrapolating from these samples to general conclusions about all of human prehistory involves substantial assumptions.

Second, Pinker's definition of violence is narrow. He focuses primarily on interpersonal violence (homicide, assault) and organized warfare. He gives less attention to structural violence -- the systematic ways in which social structures harm people by preventing them from meeting their basic needs. The concept, developed by Johan Galtung in 1969, encompasses deaths from poverty, malnutrition, preventable disease, and lack of access to medical care. If structural violence is included, the toll of the modern world looks very different. The number of people who die each year from preventable poverty-related causes -- estimated by the World Health Organization at roughly 15 million -- dwarfs the annual death toll from warfare and homicide combined.

Third, Pinker has been accused of selection bias in his choice of data and his framing of the question. Critics like John Gray, in Straw Dogs (2002), argue that the apparent decline in violence may be a temporary artifact of the post-World War II international order rather than a long-term trend. The existence of nuclear weapons has made great-power war catastrophically risky, but this is a contingent fact of technology, not a change in human nature. Remove the nuclear deterrent and the "long peace" might prove to be neither long nor peaceful.

Fourth, Pinker's focus on per capita rates can obscure the absolute scale of suffering. A smaller proportion of a much larger population can still represent an enormous number of individual deaths. The 6 million Jews killed in the Holocaust represented roughly two-thirds of European Jewry, a proportion that makes it one of the most devastating genocides in history by any measure. The 45 million who may have died in the Great Leap Forward famine represent an absolute scale of human suffering that is difficult to comprehend regardless of China's total population.

The Industrialization of Killing

What does make the twentieth century distinctive, even if it was not the most violent per capita, is the industrialization of killing. Previous centuries produced mass death, but the twentieth century produced it with unprecedented efficiency, scale, and bureaucratic organization.

The Holocaust is the paradigmatic example. The Nazi state deployed the full apparatus of modern bureaucracy -- records systems, transportation infrastructure, chemical manufacturing, and industrial engineering -- to the task of systematic mass murder. The death camps were not sites of frenzied violence but of carefully planned, bureaucratically managed killing. Adolf Eichmann's trial in Jerusalem prompted Hannah Arendt's famous phrase "the banality of evil" precisely because the killing was carried out by ordinary bureaucrats, not raving fanatics.

The Soviet Gulag system similarly industrialized mass imprisonment and forced labor, producing millions of deaths through systematic neglect, overwork, and malnutrition. The Great Leap Forward combined ideological fervor with the organizational capacity of a modern state to produce the deadliest famine in history. In each case, the machinery of the modern state -- its capacity to organize, mobilize, and direct human activity -- was turned to destructive ends with an efficiency that pre-modern states could not match.

The development of nuclear weapons added a new dimension: the capacity for truly civilization-ending violence. The bombings of Hiroshima and Nagasaki killed roughly 200,000 people, but they represented only a fraction of what nuclear arsenals could accomplish. By the 1980s, the United States and the Soviet Union possessed enough nuclear weapons to kill hundreds of millions of people and potentially trigger a "nuclear winter" that would collapse agricultural production worldwide. For the first time in history, a single decision could produce planetary-scale destruction.

Colonial Violence and the Question of What Counts

Any assessment of the twentieth century's violence must also reckon with what is excluded from standard counts. The violence of colonialism, while not unique to the twentieth century, reached its peak in the early decades and continued in various forms throughout. The German genocide of the Herero and Nama peoples in Southwest Africa (1904-1908) killed an estimated 65,000 of 80,000 Herero -- a proportional death rate that exceeds the Holocaust. The Belgian exploitation of the Congo Free State under Leopold II, which continued into the twentieth century, killed an estimated 10 million Congolese through forced labor, violence, and disease.

These events are often excluded from standard accounts of twentieth-century violence, which tend to focus on Europe, Russia, and East Asia. The result is a distorted picture that treats the violence of the world wars and totalitarian regimes as exceptional while treating colonial violence as a background condition. This is not merely a statistical oversight; it reflects deep assumptions about whose suffering matters and what kinds of violence are considered historically significant.

The Algerian War of Independence (1954-1962), the Mau Mau Uprising in Kenya (1952-1960), the Indonesian mass killings of 1965-1966 (which killed an estimated 500,000 to 1 million people), and the Bangladesh genocide of 1971 (which killed an estimated 300,000 to 3 million) all contributed to the century's toll. Whether these are counted as part of the century's violence or treated as separate, localized events depends on the framing of the question.

Total War and Civilian Casualties

Another distinctive feature of twentieth-century violence is the blurring of the distinction between combatants and civilians. In World War I, roughly 10-15% of casualties were civilians. By World War II, that figure had risen to roughly 60-65%. In wars since 1945, civilian casualties have typically exceeded military casualties, sometimes by wide margins.

This shift reflects the rise of "total war" -- the mobilization of entire societies and economies for military conflict, which made civilian infrastructure and populations legitimate military targets. The strategic bombing campaigns of World War II deliberately targeted cities and civilian populations. The firebombing of Tokyo on March 9-10, 1945, killed roughly 100,000 people in a single night. The bombing of Dresden, Hamburg, and other cities produced similar, if smaller, death tolls. These were not accidental civilian deaths but deliberate attacks on population centers, justified by the logic of total war.

The trend has continued in subsequent conflicts. The Korean War killed an estimated 2-3 million civilians, many in the extensive bombing campaigns that destroyed most of North Korea's cities. The Vietnam War killed roughly 2 million Vietnamese civilians. In the post-Cold War era, conflicts in the Democratic Republic of Congo, Syria, Yemen, and elsewhere have continued to produce enormous civilian casualties.

Different Frameworks for Measuring Violence

The question of whether the twentieth century was the most violent depends fundamentally on the framework used to measure violence. Different frameworks produce different answers, and the choice of framework is itself a consequential decision.

Total deaths: The twentieth century wins by this measure, with well over 200 million deaths from war, genocide, and politically caused famine. But this measure is sensitive to population size and conflates the scale of destruction with its intensity.

Deaths per capita: The twentieth century was violent but probably not the most violent by proportional measures. The Mongol conquests, several Chinese dynastic conflicts, and various pre-state conflicts may have produced higher per capita death rates. But the data for earlier periods is sparse and contested.

Genocide as a category: The twentieth century saw the emergence of genocide as both a concept and a systematic practice. The Holocaust, the Armenian genocide, the Rwandan genocide, and the Cambodian genocide represent a distinctive form of violence that is arguably unprecedented in its bureaucratic organization and ideological motivation.

Structural violence: If deaths from poverty, preventable disease, and systemic deprivation are included, the picture changes dramatically. The annual toll of structural violence in the modern world exceeds the annual toll of direct violence by an order of magnitude. But counting structural violence requires defining it, and the boundaries of the concept are contested.

Duration and intensity: Short, intense bursts of killing may produce different social and psychological effects than chronic, lower-level violence. The Rwandan genocide killed 800,000 people in a hundred days -- a rate of killing that exceeds anything in the historical record. But chronic low-level violence, as in the eastern Congo over the past two decades, can cumulatively produce enormous death tolls while receiving far less international attention.

The Weight of the Question

The question of which century was "most violent" may be less important than it appears. The framing implies a competition that is both unresolvable and somewhat grotesque. Whether the twentieth century killed more people per capita than the thirteenth century does not change the experience of the people who were killed.

What is more useful is understanding why the twentieth century produced the particular forms of violence it did -- industrialized killing, total war, nuclear weapons, ideological genocide -- and what conditions made these possible. The answer lies in the intersection of technology, ideology, and state capacity. The technologies of the industrial age enabled killing on an unprecedented scale. The ideologies of the twentieth century -- fascism, communism, nationalism, imperialism -- provided the motivation. And the modern state, with its unprecedented capacity to organize, mobilize, and direct human activity, provided the means.

The lesson is not that the twentieth century was uniquely violent but that modernity creates new possibilities for violence alongside its possibilities for welfare. The same state capacity that can provide universal healthcare and public education can also organize genocide. The same technology that can feed billions can also incinerate cities. The question is not which century was worst but what combination of technology, ideology, and institutional power will produce in the future.

Sources

• Arendt, Hannah. Eichmann in Jerusalem: A Report on the Banality of Evil. Viking Press, 1963.
• Galtung, Johan. "Violence, Peace, and Peace Research." Journal of Peace Research, vol. 6, no. 3, 1969, pp. 167-191.
• Gray, John. Straw Dogs: Thoughts on Humans and Other Animals. Granta Books, 2002.
• Hobsbawm, Eric. The Age of Extremes: The Short Twentieth Century, 1914-1991. Michael Joseph, 1994.
• Leitenberg, Milton. "Deaths in Wars and Conflicts in the 20th Century." Peace Studies Program Occasional Paper, Cornell University, 2006.
• Pinker, Steven. The Better Angels of Our Nature: Why Violence Has Declined. Viking, 2011.
• Rummel, R.J. Death by Government. Transaction Publishers, 1994.
• Shaw, Martin. War and Genocide: Organized Killing in Modern Society. Polity Press, 2003.
• Snyder, Timothy. Bloodlands: Europe Between Hitler and Stalin. Basic Books, 2010.
• Valentino, Benjamin A. Final Solutions: Mass Killing and Genocide in the Twentieth Century. Cornell University Press, 2004.