Is Assassination an Option?

Soon after the September 11 terrorist attacks on New York and Washington, U.S. officials announced that they had evidence linking Osama bin Laden to the attacks. As Americans began to recover from their initial shock, many of them asked, "Why don’t we just get rid of the guy?"

The terrorist tragedy reopened one of the most controversial issues in national security policy: assassination. Few topics raise more passion. Yet, despite the intense emotions assassination raises, assassination rarely gets the kind of dispassionate analysis that we routinely devote to other national security issues. That is what I will do here. When it comes to assassination, four questions are key: What is it? Is it legal? Does it work? And when, if ever, is assassination acceptable?

What Is It?

One reason assassination—or, for that matter, banning assassination—provokes so much disagreement is that people often use the term without a precise definition and thus are really arguing about different things. One needs to be clear. Depending on the definition, one can be arguing about activities that are really quite different.

For example, is killing during wartime assassination? Does assassination refer to killing people of high rank, or can anyone be the target of assassination? Does it matter if a member of the armed forces, a civilian government official, or a hired hand does the killing? Depending on the definition, killing a military leader during a bombing raid might be "assassination" but killing a low-level civilian official with a sniper might not.

For what it is worth, the Merriam-Webster Dictionary defines assassination by referring to the verb assassinate, which is defined as "to injure or destroy unexpectedly and treacherously" or "murder by sudden or secret attack usually for impersonal reasons." In other words, assassination is murder—killing a person—using secrecy or surprise. Assassination stands in contrast to murder without surprise (e.g., a duel). Also, assassination is not murder for personal gain or vengeance; assassinations support the goals of a government, organization, group, or cause.

Although people associate assassinations with prominent people, strictly speaking, assassination knows no rank. Leaders are often the targets of state-sponsored assassination, but history shows that generals, common soldiers, big-time crime bosses, and low-level terrorists have all been targets, too. Also, it does not seem to matter how you kill the target. It does not matter if you use a bomb or a booby trap; as long as you target a particular person, it’s assassination.

For our purposes, assume that assassination is "deliberately killing a particular person to achieve a military or political objective, using the element of surprise to gain an advantage." We can call such a killing "sanctioned assassination" when a government has someone carry out such an action—as opposed to, say, "simple assassination," killing by an individual acting on his own. Then the question is, should we allow the United States to sanction such activities? And, if we allow the government to sanction assassination, when and how should do it?

Is It Legal?

You might be surprised to learn that there are no international laws banning assassination. The closest thing to a prohibition is the 1973 Convention on the Prevention and Punishment of Crimes Against Internationally Protected Persons, Including Diplomatic Agents. This treaty (which the United States signed) bans attacks against heads of state while they conduct formal functions, heads of government while they travel abroad, and diplomats while they perform their duties.

The Protected Persons Convention was intended to ensure that governments could function and negotiate even during war. Without it, countries might start a war (or get drawn into one) and then find themselves unable to stop because there was no leader at home to make the decision to do so and because their representatives were getting picked off on their way to cease-fire negotiations.

But other than these narrow cases, the Protected Persons Convention says nothing about prohibiting assassination. Even then it applies only to officials representing bona fide governments and "international organizations of an intergovernmental character." So presumably the convention shields the representatives of the United Nations, the World Trade Organization, the International Red Cross, and, probably, the PLO. It does not protect bosses of international crime syndicates or the heads of terrorist groups such as Al Qaeda.

Another treaty that some might construe as an assassination ban is the Hague Convention on the "laws and customs" of war. The Hague Convention states that "the right of belligerents to adopt means of injuring the enemy is not unlimited." (This was a bold statement in 1907, when the convention was signed.)

The Hague Convention tried to draw a sharp line between combatants and noncombatants; combatants were entitled to the convention’s protections but were also obliged to obey its rules. For example, the Hague Convention tried to distinguish combatants by requiring them to wear a "fixed distinctive emblem recognizable at a distance." Wear the emblem while fighting, and you are entitled to be treated as a POW if captured; fail to follow the dress code, and you might be hanged as a mere bandit.

Alas, maintaining this definition of a "combatant" proved a losing battle throughout the twentieth century. Guerrilla warfare transformed civilians into soldiers. Strategic bombing transformed civilians into targets. Headquarters staff, defense ministers, and civilian commanders in chief today are all more likely to wear suits than uniforms. Teenage paramilitary soldiers in Liberia are lucky to have a pair of Levis to go along with their AK-47s, let alone fatigues or insignia. That is why, practically speaking, a "combatant" today is anyone who is part of a military chain of command.

Yet the Hague Convention may be more interesting not for what it prohibits but for what it permits. The closest the convention comes to banning assassination is when it prohibits signatories from killing or wounding "treacherously individuals belonging to the hostile nation or army." But when it refers to "treachery," it is referring to fighting under false pretenses (e.g., flying the enemy’s flag or wearing his uniform to lure him to death). The Hague Convention specifically permits "ruses of war." Snipers, land mines, deception, camouflage, and other sneaky tactics are okay. In fact, one might even argue that, since the convention prohibits indiscriminate killing, state-sanctioned assassination—the most precise and deliberate killing of all—during war is exactly what the treaty calls for.

The third international agreement that is relevant to assassination is the Charter of the United Nations, which allows countries to use military force in the name of self-defense. If a country can justify a war as "defensive," it can kill any person in the enemy’s military chain of command that it can shoot, bomb, burn, or otherwise eliminate. And it can use whatever "ruses of war" it needs to get the job done. As a result, the main legal constraints on sanctioned assassination other than domestic law, which makes murder a crime in almost all countries, are rules that nations impose on themselves.

The U.S. government adopted such a ban in 1976, when President Ford—responding to the scandal that resulted when the press revealed CIA involvement in several assassinations—issued Executive Order 11905. This order prohibited what it called "political assassination" and essentially reaffirmed an often-overlooked ban that Director of Central Intelligence Richard Helms had adopted for the CIA four years earlier. Jimmy Carter reaffirmed the ban in 1978 with his own Executive Order 12036. Ronald Reagan went even further in 1981; his Executive Order 12333 banned assassination in toto. This ban on assassination remains in effect today.

Even so, there has been a disconnect between our policy and practice. The United States has tried to kill foreign leaders on several occasions since 1976, usually as part of a larger military operation.

For example, in 1986, U.S. Air Force and Navy planes bombed Libya after a Libyan terrorist attack against a nightclub frequented by American soldiers in Berlin. One of the targets was Muammar Qaddafi’s tent. During Desert Storm in 1991, we bombed Saddam Hussein’s official residences and command bunkers. After the United States linked Osama bin Laden to terrorist bombings of U.S. embassies in Kenya and Tanzania in 1998, we launched a cruise missile attack at one of his bases in Afghanistan.

In each case, U.S. officials insisted that our forces were merely aiming at "command and control" nodes or at a building linked to military operations or terrorist activities. In each case, however, the same officials admitted off the record that they would not have been upset if Qaddafi, Saddam, or bin Laden had been killed in the process.

More recently, according to press reports, presidents have also approved so-called lethal covert operations—operations in which there is a good chance that an unfriendly foreign official might be killed. For example, the press reported a CIA-backed covert operation to topple Saddam in 1996 that probably would have killed him in the process, given the record of Iraqi leadership successions (no one has left office alive). After the September 11 terrorist strikes on New York and Washington, former Clinton officials leaked word to reporters that the CIA had trained Pakistani commandos in 1999 to snatch bin Laden. Given the record of such operations, bin Laden would likely not have survived.

In short, the unintended result of banning assassinations has been to make U.S. leaders perform verbal acrobatics to explain how they have tried to kill someone in a military operation without really trying to kill him. One has to wonder about the wisdom of any policy that allows officials to do something but requires them to deny that they are doing it. We would be better off simply doing away with the prohibition, at least as it applies to U.S. military operations.

Does It Work?

The effectiveness of assassination has depended much on its objectives. Most (but not all) attempts to change the course of large-scale political and diplomatic trends have failed. Assassination has been more effective in achieving small, specific goals.

Indeed, past U.S. assassination attempts have had great difficulty in even achieving the minimal level of success: killing the intended target. According to the available information, every U.S. effort to kill a high-ranking official since World War II outside a full-scale war has failed. This record is so poor that it would be hard to find an instrument of national policy that has been less successful in achieving its objectives than assassination (although price controls or election reform may come in a close second).

According to the Church Committee investigations of the 1970s, the CIA supported assassins trying to kill Patrice Lumumba of the Congo in 1961 and repeatedly tried to assassinate Fidel Castro between 1961 and 1963. In addition, American officials were either privy to plots or encouraged coups that caused the death of a leader (Rafael Trujillo of the Dominican Republic in 1961, Ngo Dinh Diem of South Vietnam in 1963, General René Schneider of Chile in 1970, and, later, President Salvador Allende in 1973). And, as noted, in recent years the United States has tried to do away with Qaddafi, Saddam, and bin Laden.

What is notable about this record is that it is remarkably free of success. Castro, Qaddafi, Saddam, and (at least at this writing) bin Laden all survived. (As this is being written, U.S. forces are hunting bin Laden as part of the larger war against the Taliban in Afghanistan.) What is more, Qaddafi continued to support terrorism (e.g., the bombing of Pan Am flight 103). Saddam has managed to outlast the terms of two presidents who wanted to eliminate him (George Bush and Bill Clinton), while continuing to support terrorism—and developing weapons of mass destruction.

One might have predicted this dismal record just by considering why American leaders have resorted to the assassination option. More often than not, assassination is the option when nothing seems to work but officials think that they need to do something. When diplomacy is ineffective and war seems too costly, assassination becomes the fallback—but without anyone asking whether it will accomplish anything.

This seems to have been the thinking behind the reported U.S. covert operation to eliminate Saddam in the mid-1990s. Despite a series of provocations—an assassination attempt against former president George Bush, violence against Shi’ite Muslims and Kurds, and violations of U.N. inspection requirements—the Clinton administration was unwilling to wage a sustained, full-scale war against him. Diplomacy was also failing, as the United States was unable to hold together the coalition that won Desert Storm. Covert support to Saddam’s opponents in the military was the alternative. It was an utter failure.

True, some other countries have been more successful in that they have killed their target. For example, after the terrorist attack on Israeli athletes in the 1972 Munich Olympics, Israeli special services tracked down and killed each of the Palestinian guerrillas who took part (they also killed an innocent Palestinian in a case of mistaken identity). In 1988 Israeli commandos killed Khalil Al-Wazir, a lieutenant of Yasser Arafat’s, in a raid on PLO headquarters in Tunisia. More recently, Israel has killed specifically targeted Palestinian terrorist leaders—for example, Yechya Ayyash, who was killed with a booby-trapped cell phone.

Other countries have also attempted assassinations with some degree of tactical success. During the Cold War, the KGB was linked to several assassinations. Most recently, the Taliban regime in Afghanistan was suspected of being involved in the assassination of Ahmed Shah Massoud, the leader of the Northern Alliance opposition.

But even "successful" assassinations have often left the sponsor worse off, not better. The murder of Diem sucked the United States deeper into a misconceived policy. The assassination of Abraham Lincoln (carried out by a conspiracy some believe to have links to the Confederate secret service) resulted in Reconstruction. German retribution against Czech civilians after the 1942 assassination of Nazi prefect Reinhard Heydrich by British-sponsored resistance fighters was especially brutal. The 1948 assassination of Mohandas Gandhi by Hindu extremists led to violence that resulted in the partition of India.

In short, assassination has usually been unreliable in shaping large-scale political trends the way the perpetrators intended (though the assassination of Yitzhak Rabin by a Zionist extremist in 1995 may be the exception). When it accomplishes anything beyond simply killing the target, it is usually by depriving an enemy of the talents of some uniquely skilled individual. For example, in 1943 U.S. warplanes shot down an aircraft known to be carrying Admiral Isoroku Yamamoto—the architect of Japan’s early victories in the Pacific. His loss hurt the Japanese war effort. The same could be said of the loss of Massoud to the Northern Alliance.

The problem is, picking off a talented individual is almost always harder than it looks. One paradox of modern warfare is that, although it is not that hard to kill many people, it can be very difficult to kill a particular person. One has to know exactly where the target will be at a precise moment. This is almost always hard, especially in wartime.

Should We Do It, and If So, How?

This is the most complex issue, of course. The morality of sanctioned assassination depends mainly on whether and when one can justify murder. Most religions and agnostic philosophies agree that individuals have the right to kill in self-defense when faced with immediate mortal danger. This principle is codified in American law. And, as we have seen, even international law seems to allow killing—even killing specific individuals—when it can be justified as armed self-defense.

Although most Americans do not like the idea of deliberate killing, they do not completely reject it, either. Most would agree that their government should be allowed to kill (or, more precisely, allow people to kill in its behalf) in at least two situations.

One situation is when a police officer must eliminate an immediate threat to public safety—for example, shooting an armed robber or apprehending a suspect who has proven dangerous in the past and who resists arrest. The other situation is when soldiers go to war to defend the country from attack. In addition, many—but not all—Americans believe that the government should be allowed to kill in the case of capital crimes.

It is probably not a coincidence that the U.S. Constitution also envisions these three—and only these three—situations in which the federal government might take a life: policing, going to war, and imposing capital punishment. Logically, then, assassination must fit into one of these three tracks. Assassination can be considered a police act, in which case it must follow the rules for protecting accused criminals. Or it can be considered a military act, in which case it must follow the rules that control how the United States wages war. Or it can be considered capital punishment, in which case it must follow the rules of due process.

Given this, when would we want to allow government to kill a particular foreign national? Clearly we should not use assassination as a form of de facto capital punishment. Unless the intended target presents a clear and immediate threat, there is always time to bring a suspect to justice, where we could guarantee due process. Similarly, although police should be able to protect themselves and others while making an arrest, we would not want police to pursue their targets with the expectation that they would routinely kill them.

The only time we should consider assassination is when we need to eliminate a clear, immediate, lethal threat from abroad. In other words, assassination is a military option. We need to understand it as such because the United States will face more situations in which it must decide whether it is willing, in effect, to go to war to kill a particular individual and how it will target specific individuals during wartime. Two factors make this scenario likely.

First, technology often makes it hard for one not to target specific people. Weapons are so accurate today that, when one programs their guidance systems, you aim not just for a neighborhood, or a building in the neighborhood, but for a particular room in a particular building. In effect, even bombing and long-range missile attacks have become analogous to sniping. You cannot always be sure you will hit your target—just as snipers often miss and sometimes hit the wrong target—but you still must aim at specific people.

Second, the nature of the threats we face today will likely require us to target specific individuals. Terrorist organizations today use modern communications to organize themselves as worldwide networks. These networks consist of small cells that can group and regroup as needed to prepare for a strike. This is how the bin Laden organization has operated. Seeing how successful these tactics have been, many armies will likely often adopt a similar approach. To defeat such networked organizations, our military forces will need to move quickly, find the critical cells in a network, and destroy them. This inevitably will mean identifying specific individuals and killing them—in other words, assassination.

But when we do so, we should be clear in our own minds that, when the United States tries to assassinate someone, we are going to war—with all the risks and costs that war brings. These include, for example, diplomatic consequences, the danger of escalation, the threat of retaliation against our own leaders, the threat of retaliation against American civilians, and so on.

Because assassination is an act of war, such activities should always be considered a military operation. American leaders need to resist the temptation to use intelligence organizations for this mission. Intelligence organizations are outside the military chain of command. Intelligence operatives are not expected to obey the rules of war and thus are not protected by those rules. At the same time, intelligence organizations are also not law enforcement organizations. In many situations, having intelligence organizations kill specific individuals looks too much like a death sentence without due process.

Indeed, there is reason to question whether intelligence organizations are even technically qualified for assassination. In every publicly known case in which the CIA has considered killing a foreign leader, the agency has outsourced the job. In most cases, it has recruited a foreign intelligence service or military officials with better access. In some of the attempts to kill Castro, the CIA recruited Mafia hit men. Even in the more recent reported cases of lethal covert actions, foreigners would have done the actual killing. It is hard to maintain control and quality when you subcontract assassination services—as the record shows.

The United States did not ask for the threats we currently face, and killing on behalf of the state will always be the most controversial, most distasteful policy issue of all. That is why we need to use blunt language and appreciate exactly what we are proposing. Sugarcoating the topic only hides the tough issues we need to decide as a country. But if we do need to target specific people for military attack, it is important that we get it right.