North Korea and the Science of Death Tolls

How do you estimate casualties that haven’t happened?

North Korean tanks in Seoul during the Korean War.
North Korean combatants plunge together with the tank unit in Seoul during the Korean War.  (Korean Central News Agency / Korea News Service via AP)

Talk of war with North Korea to remove its nuclear capabilities remains active and anxious. The Trump administration has made clear its determination to deal with a threat that might cause terrible damage to North Korea’s neighbors, and even the continental United States. But this has to be set against the immediate death and destruction that would unavoidably follow the first shots in a new Korean war. One Pentagon estimate is reported to warn that even without nuclear weapons being used, the first days of such a war could see fatalities of 20,000 a day in South Korea. Defense Secretary James Mattis, has warned that a conflict of this magnitude would be “catastrophic” and “probably the worst kind of fighting in most people’s lifetimes.”

The practice of estimating deaths that haven’t happened yet has an uneven track record in the United States, even when the deaths being forecast were only American. In part, this is because policymakers need to be warned about the worst consequences of war, and not just the most likely. It is far harder to calculate the impacts on enemy forces, and harder still to assess the impact of civilians caught up in the fight.  So all predictions need to be treated with care. Yet if the North Korean crisis spins into another military conflict on the peninsula, the U.S. would be involved with a kind of war, with all the associated costs, beyond anything contemplated since the end of the Cold War.

For most of the Cold War, of course, a prospective Third World War was expected to lead to a nuclear armageddon and even the end of civilization as we have come to know it. The numbers of those killed in such a conflict would probably exceed the millions lost in the first two world wars. Even the so-called limited wars of Korea in the 1950s and Vietnam in the 1960s saw American death tolls (including from outside the war zone) of 54,487 and 58,193, respectively. But since then, our expectations  have been scaled back. This began with the conflict with Iraq following its invasion of Kuwait, just as the Cold War concluded. In the run-up to Operation Desert Storm, which liberated Kuwait by the end of 1991, expectations were that the casualties would be comparable to, if a bit lower, than what resulted in Korea and Vietnam; one widely cited assessment from the fall of 1990 warned of up to 30,000 American dead.

The analysis behind these warnings tended to be derived from models based on a clash between NATO and the Warsaw Pact. They also assumed a direct assault on Iraqi positions. Yet even assuming a less reckless strategy, Pentagon briefings in December 1990 anticipated as many as 7,000 American deaths out of total casualties of 20,000. Those who focused less on the total numbers of forces arrayed against each other and more on the qualitative advantages enjoyed by the United States and its allies tended to be more optimistic—and accurate. In the event 383 U.S. servicemen and women died during the operation.

This was a turning point in American expectations of war, at least in terms of the costs to their own side. Iraqi military losses were much higher. In the succeeding regular wars, the losses were small for Americans and their allies. NATO fought Serbia over Kosovo in 1999 using an air campaign. There were no combat losses, although two were killed in an accident involving an Apache helicopter. The removal of the Taliban government in Afghanistan in 2001 cost 12 American lives, while the initial occupation of Iraq in 2003 resulted in 138 military dead.

One  reason for the lower death tolls is that modern Western armies are better at looking after their soldiers. Until the middle of the last century, “disease and non-battle injury” was the major cause of death for soldiers deployed to war, often more significant than actual battle. Modern U.S. soldiers are healthier and inoculated against disease as they move into unfamiliar territory. Body armor provides much better protection, and if soldiers are injured in battle they get much improved treatment on the spot and are then speedily evacuated to a field hospital. And this is not just true of the American armed forces. Between 1946 and 2008 in all conflicts, there was a 50 percent decline in known battle deaths.

Of course, the major reason casualties are lower is that a third world war has been avoided. The prospect of mutual destruction that would result from nuclear exchanges war has encouraged caution whenever there was a risk of a conflict among the great powers. The absence of such a conflict since 1945 became known as the Long Peace. Instead, the more recent wars fought by the United States and its allies have not been industrial-scale conflicts between similarly armed states, but rather less-equal struggles against lesser powers, unable to cope with the weight of American firepower and the speed and accuracy with which it can be delivered.

The precision of modern weapons leads to expectations that enemy losses can also be kept down. There is less excuse for collateral damage, when civilians die in strikes supposedly directed against combatants. We have become used to the idea that Western forces will experience much lower levels of losses, still painful and poignant for those affected. In the past, the numbers of casualties were so huge that they numbed the senses. Now it is possible to consider the circumstances and human meaning of every death and injury.

When wars were fought over the most vital national interests, individual sacrifices were part of a great national effort; with more discretionary wars it becomes more of a challenge to justify a relatively small number of casualties. When the U.S. lost 241 in the bombing of the marine barracks in Beirut in 1983, and 19 rangers in Mogadishu in Somalia a decade later, it abandoned both operations within months. This presumed casualty-intolerance explains the tactics adopted by American adversaries. They may not be able to defeat U.S. forces in actual battle, but they can catch them out in ambushes, improvised bombs, and indirect fire. After the successful and relatively smooth early occupations of Afghanistan and Iraq in the 2000s, the U.S. and its allies found themselves coping with insurgencies that resulted in regular casualties. These have now reached 2,404 Americans for Afghanistan and 4,256 for Iraq. They have not led Washington to abandon efforts in those countries, but they have certainly led to caution when it comes to contemplating further interventions that might turn into long-term commitments.

For those who live in countries being consumed by war, there is no easy escape. When insurgencies and civil wars take hold, the impact becomes severe and enduring, as the fighting becomes routine and persistent. With the collapse of infrastructure, disease and malnutrition become big killers. Epidemics of sexual assaults follow armies as they move through populated areas. Those seeking to flee put themselves through terrible hardships, becoming either internally displaced or full refugees. War may just be the worst of many bad things afflicting a country that combine to make life progressively more miserable, including oppressive governments and natural disasters. Casualties accumulate inexorably over time. The numbers of Iraqis that have suffered violent deaths since the 2003 invasion is conservatively put at 268,000, with up to 200,000 of those civilians. Early attempts to record the deaths in the Syrian civil war faltered as it became impossible to keep an accurate count. Some estimates now put the numbers as high as 450,000. The combined toll for the wars in Afghanistan and Pakistan since 2001 was put last year at 173,000 dead and more than 183,000 seriously wounded. The War on Terror is, however, only the latest in a long line of conflicts to have afflicted Afghanistan since the 1970s.

So while the West may have come to expect relatively small numbers of casualties in modern wars, that is not the experience of those living in the battle zones. That was also true during what we may have to call the First Korea War from 1950 to 1953. The effect on the civilian population was dire. “The number of Korean dead, injured or missing by war’s end,” according to Charles Armstrong, “approached three million, 10 percent of the overall population.” This was a vicious war, with atrocities conducted by both sides and unremitting bombing by the United States Air Force of the North’s cities. This was a war that began with a surprise attack launched by the communist North against the South. It might have succeeded, were it not for a last-ditch American intervention.

If this is recalled in Pyongyang, leaders there might think twice before trying again. Certainly if they want to avoid the total destruction threatened by President Trump, they will hope to achieve a quick conventional victory. One possibility is that they will seek to wreck havoc in the South Korean capital, Seoul, by means of a massive artillery barrage, supported by cyber-assaults and perhaps chemical weapons, encouraging the population to flee (and so snarling up South Korean reinforcements) as a land invasion crosses the Demilitarized Zone to occupy the deserted city. It is this prospect that leads to casualty estimates of 20,000 a day. For Pyongyang this would be an enormous gamble, because the invading force would be vulnerable to air strikes and would probably also suffer huge losses. More likely than a quick victory would be early escalation, at its very worst leading to nuclear use. Should the U.S. decide to strike first, either to pre-empt North Korean action or to degrade its nuclear capacity, then it might be able to limit the damage. But it would be a brave military planner who promised that this could be done without a heavy cost as the North retaliated.

When wars come they rarely follow the course anticipated in advance. The immediate costs of conflict have often been overestimated in the past, even as the long-term costs have been underestimated. But this war would be against a country that has been preparing for it for over 60 years. The safest assumption must be that in its scale and intensity a Second Korean War would be unlike anything experienced in recent decades, especially by Western countries. It would represent a sharp reversion to an old norm, and away from any hopes that we have learnt to contain the costs of war.