Preventive and Preemptive Actions against Terrorists

Fighting against terrorism is different from fighting conventional wars against nation-states. The difference may be a matter of degree, especially when the terrorism is large-scale and state-sponsored or state-supported, as it often is. But it is real nonetheless, especially when the terrorists are willing, sometimes even anxious to die for that cause.¹ Deterrence is less effective under these circumstances, and preventive mechanisms must assume a larger role. In the following pages we shall consider several preventive strategies that have been employed, with varying success, against individual terrorists and terrorist groups. We shall also examine some measures that are likely to be tried if the threat of terrorism increases, as it almost certainly will. Since the threat of terrorism has driven much of the current debate and even changed its terms, it must be a central focus in any attempt to construct a jurisprudence of preemption and prevention.

Preventive Detention of Potential Terrorists

One preventive mechanism that has been widely used during “emergencies” has been the preventive detention of individuals or groups believed likely to engage in future acts of terrorism. Although the decision to detain a particular individual on preventive grounds represents the sort of micro action discussed in Chapter 1, sometimes it is made in the context of macro actions, such as the war against terrorism, and can thus be seen as part of an overall preventive strategy, particularly when it involves large groups of detainees, as distinguished from individual suspects.

Before we return to the Middle East, we should briefly touch on the experiences of Great Britain and the United States, especially since the British approach to preventive detention has served as the basis for similar actions by Israel, Jordan, and other Middle Eastern nations.

The English Experience Great Britain has long practiced preventive detention during wartime and other national security emergencies. During both world wars the British government promulgated regulations explicitly authorizing the preventive detention of certain dangerous persons. In 1915 Regulation 14B was proclaimed. This wartime grant of extraordinary power authorized the home secretary to detain anyone of hostile origin or association whose internment was deemed expedient to the securing of “public safety and defence of the realm.”² During the Second World War a similar regulation was enacted. Pursuant to this power, thousands of people, ranging from Jewish refugees who had escaped Nazi Germany to homegrown British fascists, were interned. Special retention powers were also granted to the government in its fight against Northern Irish terrorism, and hundreds were interned in accordance with this power.³

Following the 9/11 attacks, Britain enacted an antiterrorism law that authorized the indefinite preventive detention of foreign nationals designated as terror suspects. In December 2004 the House of Lords ruled that such unlimited detention without trial was incompatible with the European Convention on Human Rights, but the British government refused to release the nine Muslim detainees it was holding without having pressed criminal charges on the ground that they were “a significant threat to our security.”⁴ (The details of the reason for detention are secret.)⁵ The detainees are now appealing to the European Court of Human Rights.⁶

The terrorist attacks in London on July 7, 2005, stimulated the British Government to reconsider its antiterrorism laws. As Prime Minister Tony Blair asserted following the London bombings: “We have, in my judgment, not been tough enough or effective enough in sending a strong signal across the community that we are not going to tolerate people engaging in extremism or propagating it or inciting it.”⁷ Blair noted that after the bombings the public awareness of terrorism changed: “I tried for several months before the election to get tougher on terrorism legislation through. People said this was scaremongering and so on. People don’t say that now.”⁸ He observed that “[v]irtually every country in Europe, following terrorist acts, has been toughening up their legislation.”⁹

The prime minster’s current twelve-point antiterrorism plan includes the deportation and detention of terror suspects, as well as a ban on organizations supporting terrorism. Blair stated that deportation will be ordered against “those fostering hatred, advocating violence and validating such violence.”¹⁰ Acknowledging that his plans on the issue of deportation might conflict with current human rights laws, Blair stated that the human rights acts should be “amended if necessary”¹¹ to facilitate deportations.

Under the Blair proposal, the detention extensions would be issued by judges with “special security clearance” who are authorized to consider evidence that is “currently inadmissible in normal courts,”¹² such as from certain phone-taps. The director of a British civil rights group reported: “The thought of secret hearings where once again the accused will never hear the case against them fills me with dread.”¹³ Defending the proposed piece of legislation, Britain’s lord chancellor, Lord Falconer, stated: “We need to debate the three months and we need to try to build a consensus around what the right period of time is. . . . But what is being suggested is not any form of internment, just a sensible period to detain suspects while sensible investigation is going on.”¹⁴

Finally, the plan to criminalize “indirect incitement” and “glorification of terrorism” was criticized by human rights groups as an infringement of peaceful expression. Holly Cartner, Europe and Central Asia director at Human Rights Watch, said: “Directly inciting violence is already a crime in Britain. . . . These overly broad new offenses will have a chilling effect on free speech in the classroom, the newsroom and the mosque.”¹⁵ (Australia has begun to follow suit; its prime minister has urged the passage of a law that would expand the definition of sedition “to include statements that ‘urge disaffection’ toward the government, or that promote ‘ill will or hostility’ among groups.”)¹⁶

The American Experience In the United States too preventive detention was practiced on a relatively small scale prior to World War II. During the Civil War, President Lincoln suspended the writ of habeas corpus, thereby making it possible for his military advisers to confine individuals believed to be dangerous.¹⁷

When the Supreme Court ruled that “the writ could not be suspended, except by Act of Congress,” Lincoln persuaded Congress to give him the authority to detain dangerous enemies.¹⁸ In a subsequent case, the military decided not only to detain a civilian named Lambdin Milligan, who was arrested for organizing an uprising against the Union, but also ordered his execution. In an appeal that came before it after the war had ended, the Supreme Court held that the civil courts of the Union state of Indiana had been open and “needed no bayonets” to protect them, meaning that it had been unconstitutional to try Milligan before a military commission. Recognizing that Milligan was arrested in wartime, when passions run high and “considerations of safety” are deemed all-important, the Court concluded that the framers of our Constitution “foresaw that troublous times would arise, when rulers and people would become restive under restraint, and seek by sharp and decisive measures to accomplish ends deemed just and proper; and that the principles of constitutional liberty would be in peril, unless established by irrepealable law. . . .”

It went on to say: “This nation . . . has no right to expect that it will always have wise and humane rulers, sincerely attached to the principles of the Constitution. Wicked men, ambitious of power, with hatred of liberty and contempt of law, may fill the place once occupied by Washington and Lincoln, and if this right [to suspend provisions of the Constitution during the great exigencies of government] is conceded, and the calamities of war again befall us, the dangers to human liberty are frightful to contemplate.”

Having delivered itself of this bold rhetoric about “irrepealable law,” the Supreme Court then proceeded to suggest that the right to bail could be suspended during emergencies: “If it was dangerous, in the distracted condition of affairs, to leave Milligan unrestrained of his liberty, . . . the law said to arrest him, confine him closely, render him powerless to do further mischief; and then . . . try him according to the course of the common law.”¹⁹

This view of preventive confinement during emergencies was reaffirmed and strengthened by Justice Oliver Wendell Holmes in a case growing out of a private war between Colorado coal miners and owners, which led to a declaration of local martial law by the governor. In addition to suppressing newspapers, deposing civil magistrates, and closing all saloons, the governor suspended habeas corpus and ordered the arrest of certain “objectionable characters.” One of these “characters,” a leader of the miners, was detained without bail for two and a half months and sued the governor after his release. Though Holmes need never have reached the legality of the detention, the Civil War veteran went out of his way to justify the governor’s action. Employing “logic” for which he surely would have chastised first-year Harvard law students, Holmes argued that since a governor can order soldiers to “kill persons who resist” efforts to put down a rebellion, it certainly follows that “he may use the milder measure of seizing the bodies of those whom he considers to stand in the way of restoring peace.”²⁰

That is where the law stood on December 7, 1941, when Japanese forces bombed Pearl Harbor, throwing Hawaii into turmoil and generating fear of attack in our West Coast cities. Within hours the governor of Hawaii, at the insistence of the army, declared martial law, suspended habeas corpus, ordered the civil courts closed, and empowered military tribunals to try all criminal cases.²¹ It wasn’t until after the war (and the restoration of habeas corpus by the president) that the Supreme Court decided that Congress, in authorizing martial law in Hawaii, had not intended to permit the “supplanting of courts by military tribunals.”²² By that time thousands of person-days of illegal imprisonment had already been served.

Martial law in Hawaii, with all its abuses, did not include mass detention on racial grounds of the kind employed on the West Coast between 1942 and 1944. At that time there were about 110,000 Americans of Japanese ancestry living on the West Coast, of whom 70,000 were American citizens.²³ A virulent anti-Japanese hysteria followed the attack on Pearl Harbor. Rumors were circulated that Hawaiians of Japanese ancestry were signaling enemy pilots and submarines, that Japanese-Americans had intentionally infiltrated the power and water companies, and that they had formed sabotage and espionage rings numbering in the thousands. None of these stories proved true. The records of “the Federal Bureau of Investigation and Army and Navy intelligence indicate that there was not a single instance of espionage or sabotage by a resident of Japanese ancestry before, during, and after World War II.”²⁴ The absence of such activities did not, however, satisfy a hysterical population with deep-rooted racial antagonisms. Indeed, the attorney general of California, Earl Warren, expressed the Alice in Wonderland view that it was the very absence of sabotage that was “the most ominous sign in our whole situation.” It convinced him, he said, “that the sabotage . . . the fifth-column activities that we are to get, are timed just like Pearl Harbor” and that the present inaction by the Japanese-Americans was designed to lull us “into a false sense of security.”²⁵

The various intelligence agencies—the FBI and army and navy intelligence—preferred to approach the problem of potential terrorism and espionage “on the basis of the individual, regardless of citizenship, and not on a racial basis.” This was what was done with people of German and Italian extraction on the East Coast. Thousands of aliens “regarded by the Attorney General as dangerous to the national security if permitted to remain at large” were preventively detained on an individual basis. But on the West Coast the prevalent attitude was reflected by General John

DeWitt, head of the Western Defense Command: “A Jap’s a Jap. There is no way to determine their loyalty.” Earl Warren agreed: “We believe that when we are dealing with the Caucasian race we have methods that will test their loyalty. . . . But when we deal with the Japanese . . . we cannot form any opinion that we believe to be sound.”²⁶ The decision was made to confine the entire West Coast population of Japanese-Americans: Accordingly, 109,650 men, women, and children were put in detention camps, where they remained for nearly the entire war.²⁷

The mass confinement of West Coast Japanese-Americans constitutes one of the “purest” examples of preemptive incapacitation in history. The alleged harms sought to be prevented by the massive confinement included espionage and sabotage. It was also claimed that the detention of entire families in isolated camps was designed to prevent attacks against Japanese-Americans by “real” Americans who were angered by Japan’s attack on Pearl Harbor and its continuing war against America.

The preemptive incapacitation was “pure” in the sense that it was based entirely on predictions of the future since none of the detainees had any prior history of espionage or sabotage

Many, though not all, civil libertarians were extremely critical of the detention of the Japanese-Americans on racial grounds.²⁸ Prominent leaders of the American Civil Liberties Union urged President Franklin D. Roosevelt to “constitute a system of hearing boards to test the loyalty” of individual citizens and noncitizens. The justices of the Supreme Court who dissented from the judicial approval that was given to the exclusion and detention orders faulted the government for not treating “these Japanese-Americans on an individual basis by holding investigations and hearings to separate the loyal from the disloyal as was done in the case of persons of German and Italian ancestry.” (Virtually no criticism was ever leveled against the individualized preventive detention of the latter.) Academic criticism centered on our failure to detain Japanese-Americans “on the basis of individual suspicion” and on our unwillingness to adopt a system of graded restrictions—as the British and French did—whereby only the most dangerous were detained and others “were subjected to certain continuing restrictions especially as to their travel.”²⁹

There is no hard evidence that even a single case of espionage or sabotage was actually prevented by the massive detentions, though it is of course possible that there may have been such cases. It is in the nature of preventive confinements that it is difficult to prove that they constituted false positives, especially if the confinement is pervasive and coextensive with the duration of the threat. Since confined people cannot engage in the predicted conduct, it is easy to claim that it worked. And it is difficult for its critics to prove that the same result would have been achieved in the absence of the confinement. The verdict of history, however, is on the side of the critics, and detention of the Japanese-Americans must be counted as perhaps the largest-scale false positive confinement in the history of American preemptive incapacitation.³⁰ It was also the largest-scale racial profiling case in American history, differing in both kind and degree from other instances of racial profiling. By confining all West Coast Japanese-Americans, the U.S. government was essentially saying that all Japanese-Americans were potential spies or terrorists. This is quite different from what happened after September 11, 2001. Following that attack on America, many law enforcement officials came to a quite different conclusion—namely, that all suicide bombers were Muslim extremists, but that very few Muslims were potential terrorists. This led law enforcement officials to focus their attention on Muslims, rather than on Christians, Jews, or atheists, but it did not result in the confinement of all or even a significant percentage of Muslims or even Islamic extremists. Despite the Supreme Court’s ill-advised approval of the Japanese detentions, or perhaps because of it, our nation has never again made the mistake of preemptively confining an entire ethnic or national origin group on the basis of a generalized prediction that it might do harm.³¹

The Israeli Experience

Since its establishment in 1948 Israel has followed a model similar to the one suggested by many of those civil libertarians who disagreed with Franklin Roosevelt’s policy of mass detention based solely on ethnicity: namely, individualized determinations of dangerousness (not loyalty). Israel faced terrorist attacks even before it became a state. During the 1920s and 1930s bands of terrorists attacked Jewish settlements in Palestine, killing hundreds of Jewish civilians, and between 1948 and 1967—before the occupation of the West Bank and Gaza Strip—more than fifteen hundred Israelis were killed by fedayeen terrorists. Since the end of the Six-Day War more than two thousand Israelis have been killed in terrorist attacks. Only a fraction of the terrorist attacks planned by Palestinian terrorists have succeeded because Israel has managed—through intelligence and other preventive measures—to preempt or thwart thousands of such attacks, thereby saving tens of thousands of lives.³²

Although these preventive measures have saved many Israeli lives, they have been costly, in terms of Palestinian lives, liberty, property, and dignity. They have entailed preventive detention, targeted killings, the taking of property for the building of security barriers, and the erection of checkpoints. Even more important, they include the reoccupation of large parts of the West Bank following the renewal of the intifada, beginning in 2000 and 2001, with its spate of suicide bombings against Israeli civilians.³³

One preventive mechanism used by Israel since its inception has been preventive (or administrative) detention. Between 1948 and 1979 it operated an individualized system of preventive detention based on British mandatory law. Most, but certainly not all, of those detained under this approach were Arabs suspected of complicity with terrorism. In 1979 Israel’s Knesset enacted the Emergency Powers (Detention) Law. Both the old and current laws required individual proof of dangerousness and provided some procedure for challenging the government’s allegations, though the procedures were a far cry from those required for criminal prosecution. Under current Israeli law a person suspected of complicity with terrorism can be detained for a renewable period of six months on the basis of a finding of “reasonable cause” by the minister of defense that his detention is required by reasons of “state security or public security.” These vague criteria provide little legislative guidance to the appropriate balance to be struck between false positives and false negatives. Nor do they provide much in the way of evidentiary guidance beyond authorization to “deviate from the rules of evidence” if the president of the court is satisfied that this will be “conducive to the discovery of the truth and the just handling of the case.”³⁴

There are several distinct but often overlapping reasons why a democracy, committed to the rule of law, would employ preventive detention instead of (or in addition to) the traditional criminal process. The first involves individuals who the intelligence agencies are convinced have been guilty of past acts of terrorism, ranging from actually planting bombs to organizing terrorist attacks, but for whom the information on which these conclusions of past guilt are based cannot be introduced at a public trial without compromising ongoing intelligence operators. For example, the evidence may come from an undercover operative who is deep inside the terrorist network. If he were to be revealed as the source, he would likely be killed, or at the very least his continuing value as a spy would be endangered. It may also come from electronic or other high-tech means of surveillance not yet known to the terrorists.³⁵

A closely related situation involves individuals who are believed to be planning a future terrorist attack. Since planning a future attack is already a crime (at least if the planning has reached a certain stage), such future criminals are generally also past criminals and therefore fall within the first criterion discussed above. The difference may be only a matter of degree: In the first category, the individual may be suspected of having already caused multiple deaths, whereas in the second category, the past crime may be inchoate in nature, but the future crimes may risk massive loss of life.

In the above cases, a democratic society has essentially four options: It can continue to rely exclusively on the conventional criminal law, an option that will require it to choose in every case whether to reveal the secret information or to forgo prosecution. No democracy faced with significant threats of mass victimization has selected this “pure” option.

The second option is to change the existing criminal law and adapt it to the new reality. This could entail, for example, a watering down of the hearsay rules, which generally require testimony by the person who claims to have firsthand knowledge of the events, so that he can be cross-examined. The hearsay rules could be changed to allow an intelligence operative to testify that a secret informer who has proved reliable in the past told him that the individual at issue had committed specified past acts of terrorism or was in the process of planning future terrorist acts. This is currently permitted in applications for search warrants and in grand jury and some other proceedings, but not at criminal trials. (Even hearsay testimony could in some cases endanger secret intelligence sources because the nature of the information might provide a clue to its source.) In 1970, while studying Israel’s system of preventive (or administrative) detention, I put this option to a high-ranking Israeli legal official. This is what he said in response: “We are very proud of our civil liberties. It would be absurd to wreck our entire judicial system to accommodate a few wartime security cases.” But would you really have to wreck the system? I asked. Couldn’t you just change some of the rules of evidence? “The rules of evidence lie at the center of our civil liberties, and the right to confront your accuser is the heart of any fair system of evidence. If we created a rule allowing into evidence [hearsay testimony regarding] the invisible-ink message and the agent’s report,³⁶ there would be virtually nothing left to the right of confrontation. I would rather see us act completely lawlessly in a few security cases than a little lawlessly in every case.” This official felt strongly enough to say that he would “resign in protest” if Israel ever changed its rules to allow hearsay evidence in the general run of cases.³⁷

A variation of this option would be to change the existing criminal law only in terrorist cases while preserving the hearsay and other protections in all other cases. One problem is that the U.S. Supreme Court recently held that at least certain aspects of the hearsay rules are mandated by the confrontation clause of the Sixth Amendment, which provides that in all criminal prosecutions, the accused must “be confronted with the witness against him.” This constitutional protection could not be changed without amending the Bill of Rights, though the Supreme Court has also sustained watered-down protections in the case of terrorists captured abroad and held as enemy combatants.³⁸

The third option is to do what Great Britain and Israel have done—that is, to enact an explicit preventive detention law authorizing the noncriminal detention of individuals who are thought likely to engage in future acts of terrorism. This option has the virtue of explicitness. A good preventive detention law—and neither Great Britain’s nor Israel’s is particularly well crafted—should provide specific and narrow criteria for permitting this extraordinary remedy. It should also provide procedural safeguards that are as rigorous as the reality of fighting terrorism permits. Finally, it should provide assurances that this mechanism, which is vulnerable to abuses, can never be used to serve the political or ideological ends of those in power. In May 2005 the Israeli government ordered the preventive detention of several Jewish extremists who were opposed to the scheduled unilateral withdrawal from the Gaza Strip. Although the authorities claimed that the detention was justified by the threat of violent opposition to the controversial withdrawal, some Israelis, especially on the hard right, believed it was as much a political attempt to silence the opposition.³⁹Without a specific past criminal act to point to, those who administer preventive detention, especially against political enemies, will always be subject to charges of misuse of this antiterrorist measure.

The British-Israeli approach, despite its explicitness and accountability, still requires a compromise of the principle articulated by Justice Robert Jackson: “[T]he jailing of persons by the courts because of anticipated but as yet uncommitted crimes” cannot be reconciled with “traditional American law.”⁴⁰

In order to avoid compromising this oft-stated (and, in practice, oft-compromised) principle, the United States has selected the fourth option. Our government does not explicitly acknowledge that we are employing preventive detention. Indeed, a federal law enacted in reaction to the detention of Japanese-Americans explicitly provides that “[n]o citizen shall be imprisoned or otherwise detained by the United States except pursuant to an act of Congress.”⁴¹To avoid this prohibition, our government stretches existing law to achieve the same result, but with far less accountability. In the aftermath of 9/11 we dusted off the anachronistic material witness law that was designed to permit the short-term detention of crucial witnesses to a crime who were proved to be likely to flee the jurisdiction. Federal law enforcement agents swept up hundreds of foreign nationals,⁴² many of them Arabs and Muslims, and declared each to be a “material witness.” But to what crimes were they alleged to be witnesses? Often, it was the crimes they themselves were suspected of committing or planning!⁴³ This misuse of the material witness law was plainly designed to circumvent the absence of an explicit preventive detention law.

Another existing set of laws that was stretched to achieve preventive detention was the immigration rules, which authorize the short-term detention of aliens who are out of status or do not have proper papers. These laws were designed to facilitate the quick deportation of such aliens in cases in which that was the appropriate remedy. But in the aftermath of 9/11 they were used to effectuate the preventive detention of specific individuals—almost all Muslims and Arabs—who were thought to pose a future danger.

In 1971 I wrote an article entitled “ ‘Stretch Points’ of Liberty.”⁴⁴ In the process of researching the article, I interviewed the deputy attorney general of the United States—the person then responsible for planning any response to a national emergency. I asked whether if terrorism were to strike the United States, he would recommend invoking “extraordinary powers of temporary detention.” He told me, “We wouldn’t have to. There is enough play at the joints of our existing criminal law—enough flexibility—so that if we really felt that we had to pick up the leaders of a violent uprising, we could. We would find something to charge them with, and we would be able to hold them that way for a while.”

He was right, at least as a factual matter. Every legal system has its “stretch points,” its flexible areas capable of expansion and contraction depending on the exigencies of the situation. The stretch points in our system include broad police and prosecutorial discretion, vaguely defined offenses (such as disorderly conduct), inchoate crimes (which may also be vaguely defined, like conspiracy), denial of pretrial release (which can sometimes result in confinement exceeding a year), and the material witness and immigration laws previously mentioned. Some systems employ such devices as common law (judge-made) crimes, ex post facto (after the fact) legislation, and emergency powers, to achieve similar results.

It is typical of the U.S. government, in this area and in others, to opt for what one commission labeled “the way of the hypocrite.”⁴⁵We loudly proclaim our commitment to maintaining certain restrictions on governmental action—in this case the prohibition against preventive detention—while at the same time accomplishing the ends of this mechanism by discreetly stretching existing law.

Another reason why a democracy might employ preventive detention is to provide for a period of detention outside the restrictive criminal process sufficient to permit interrogation of the detainee for purposes of obtaining crucial intelligence information about future terrorist acts being planned by others. Interrogation within the criminal process is closely regulated by statutory and constitutional law, most particularly the so-called Miranda rule. In theory, no interrogation of a confined criminal suspect may occur in the absence of his lawyer.⁴⁶ In practice, this restriction is often avoided by a number of well-known subterfuges. Moreover, the Supreme Court recently ruled that the privilege against self-incrimination does not come into play until the prosecution seeks to introduce evidence against a defendant at a criminal trial.⁴⁷ At that point—and only at that point—must the court decide if the evidence was the product of an improper interrogation. This decision provides intelligence-gathering agencies with considerable leeway in interrogating individuals in order to secure intelligence information that they have no intention of ever using in a criminal trial. Still, there must be time to conduct the interrogation, and under current law a criminal suspect may not be detained for purposes of interrogation beyond a relatively brief period of time. Preventive detention seeks to remedy that problem by authorizing noncriminal detention for much longer periods of time, thereby enabling the kind of interrogation that would not be permissible in the criminal process.⁴⁸

The preventive detention of suspected future terrorists is likely to continue wherever terrorism is perceived as a serious threat.⁴⁹ It is essential that democratic nations committed to the rule of law begin to develop a jurisprudence regulating this increasingly important preventive mechanism. It is fair to say that to date no nation has satisfied this important need.

Targeted Preemption of Terrorists Detention (preferably after a fullblown criminal trial) is the tactic of choice against suspected terrorists, but it is not always feasible to capture suspected, or even self-proclaimed, terrorists or others who pose an immediate and serious danger to a state. Many states have opted, under these circumstances, for a more drastic form of preventive incapacitation—namely, targeted killing. Again, Israel has been among the most public in its willingness to address this issue (which it calls focused preemption) directly and to try to incorporate it within the rule of law. The United States has also employed targeted killings, but thus far without any real attempt to justify it under the law.

Over the past several years, nearly all those who have been targeted for killing by Israel have been terrorists or terrorist commanders and leaders, but in earlier years, preemptive or preventive assassinations were also directed against scientists, weapons providers, and others who were in the business of harming Israel by providing its enemies with weapons of mass destruction.

In an earlier microversion of its later macroattack against the Iraqi nuclear reactor, Israel tried by various means to prevent German scientists, some of them former Nazis, from helping Egypt develop weapons of mass destruction for use against the Jewish state. According to a historical account of Israel’s intelligence services by Ian Black and Benny Morris:

[The Mossad] believed that the German scientists were working on weapons that threatened the very existence of Israel. . . . The Mossad in that period [1956–1961] conducted only intelligence-gathering. . . . In September 1961 . . . a first assessment on the development of ground-to-ground missiles in Egypt [was produced]. A second Intelligence Branch evaluation in October 1962 predicted that about 100 rockets could be operational within a year to eighteen months. . . . The worrying reports about the scale of the Egyptian programme were reinforced catastrophically when a [defector] claimed that the Egyptians were preparing to fit their missiles with warheads containing radioactive waste in an operation codenamed Ibis I. Even more seriously, a project called Cleopatra was geared to producing nuclear warheads.⁵⁰

Several targeted attacks on individuals who were working on these weapons of mass destruction ensued, and the Egyptian program never really got off the ground.⁵¹

In the leadup to the attack on the Iraqi nuclear reactor, some targeted attacks were also tried, but without success. Diplomatic efforts failed as well, leading Israel to conclude that its only realistic option was the air assault on the Osirak reactor.

In at least one case, bribery and threats were tried first. But when they failed, the weapons maker was killed. Dr. Gerald Bull, a leading expert in barrel ballistics, was helping the Iraqis develop “a supergun capable of launching shells containing nuclear, chemical, or biological warheads from Iraq directly into Israel. The supergun’s barrel was 487 feet long, composed of thirty-two tons of steel supplied by British firms to Iraq. Late in 1989 a prototype had been test-fired at a gunnery range at Mosul in northern Iraq. Saddam Hussein had ordered three of the weapons to be built at a cost of $20 million. Bull was retained as a consultant at $1 million. The project was code-named Babylon.”⁵² First Israel tried “to buy his expertise.” But each time “Bull had made clear his distaste for the Jewish state.”⁵³ Threats followed, but also to no avail. Finally, Bull was killed, and the supergun project ended.

In addition to these targeted attacks against weapons providers, Israel used targeted killings against terrorist leaders, especially following the attack against Israeli athletes at the Munich Olympics of 1972 and the decision by the West German government to free the captured terrorists.⁵⁴ This decision was one of many by European governments to free captured terrorists or not to apprehend them in the first place. This led Israel to conclude that the only effective measures it could take against terrorist leaders operating in foreign countries was to kill them.

Some of these killings were explicitly preventive—the targets were directly involved in the ongoing planning of future terrorist attacks, and their deaths would abort or at least confound these plans—but others were carried out for deterrent purposes. According to Black and Morris, “The Munich massacre marked a turning-point in Israel’s war against Palestinian terrorism. Golda Meir decided that the time had come for a wholesale vengeance, not just for its own sake but as a deterrent.”⁵⁵

These operations provided the background for the controversial Israeli decision to employ targeted killings during the Palestinian terrorism campaign that began in the fall of 2000. Targeted killings were adopted as a tactic, especially in the cities, towns, and refugee camps of the Gaza Strip, where Israeli soldiers did not operate freely. Hamas did, however, have a relatively free hand in Gaza. Many terrorist acts originated in Gaza, and Israel had good intelligence about terrorist groups and individuals in that area. Sometimes, but only rarely, it was feasible to arrest terrorists who were planning or implementing imminent attacks. In most cases, however, capture was not feasible because the terrorists deliberately hid among civilians in densely populated urban areas, where Israeli soldiers could not safely operate without massive military support and at great risk to themselves as well as to Palestinian civilians. In those cases, Israel opted for another preventive measure, the targeted killing of terrorists, including commanders, operatives, and suicide bombers. Generally the attacks took place from the air, but sometimes they employed exploding cell phones and other technological ways of delivering a lethal explosive against a terrorist combatant, like the Engineer, a man named Yehiya Ayash, who was the chief bomb maker for Hamas until Israeli agents targeted him by planting an explosive charge in his cell phone. If these combatants—and they are combatants, under any reasonable definition of that term—had been captured, arrested, or otherwise detained, there would be no basis for criticism since every government is empowered to take reasonable and proportional steps to protect its civilians and soldiers from terrorist attacks. The criticism arises because this preventive mechanism involves the targeting and killing of suspected terrorists. Since targeting, no matter how carefully implemented, can never be perfect, especially in the large urban centers where terrorists hide, innocent bystanders are sometimes killed or injured along with the terrorists. The complaint is therefore twofold: First, even if, as is often the case, the only ones killed are suspected terrorists (sometimes along with their bodyguards, who are also combatants), they are being “executed” without the “due process” of a trial; second, innocent people are too often killed or injured along with suspected terrorists.

As for the first complaint, critics of this practice argue that terrorism is a law enforcement issue, since terrorists are not technically at war with Israel. It would follow from this argument—were it sound—that suspected terrorists must be treated precisely the same as suspected rapists or robbers. They must be arrested and tried, not hunted down and killed. They may be killed only in self-defense—if they pose an immediate danger to the arresting officers. If they are killed, this constitutes extrajudicial execution, which is illegal.

The extrajudicial execution argument is flawed, however, since all military killings are by their nature extrajudicial. Indeed, it is judicial executions that I, and many other civil libertarians, strongly oppose, precisely because there is always the alternative of lengthy, even life imprisonment. The person to be judicially executed is, by definition, already in captivity and no longer a direct danger. With regard to targeted killings, there is generally no other viable option other than allowing the terrorist to continue killing without any effort to stop him. Targeted killings of terrorists are more akin to the killing of a dangerous felon who is fleeing from arrest or to killing in self-defense, both of which are extrajudicial but entirely lawful. The relevant question to be asked of targeted killings is whether they are lawful and moral, not whether they are extrajudicial. This involves a number of factors, including the evidence that the targeted suspect is in fact a terrorist involved in ongoing operations, the imminence and likelihood that these terrorist operations will succeed, the availability of other less lethal alternatives, and the possibility that others will be killed or injured in the targeted attack (in other words, how well targeted will the attack be).

As for the question of whether terrorists are to be treated as ordinary criminals or as combatants who are violating the laws of war, it is interesting to note that when terrorists are captured, many refuse to be treated as ordinary criminals, demanding POW status and insisting that they are part of a military or paramilitary insurgency. On the flip side, those who capture them often refuse to treat them as combatants entitled to prisoner of war status. The reality is that terrorists involved in an ongoing insurgency or campaign of terror are in a hybrid status that may well justify treating them as combatants for purposes of targeted killings but not as prisoners of war once they are captured.⁵⁶ As Professor Yoram Din-stein has put it, “However, a person is not allowed to wear simultaneously two caps: the hat of a civilian and the helmet of a soldier. The person who engages in military raids by night, while purporting to be an innocent civilian by day, is neither a civilian nor a lawful combatant. He is an unlawful combatant. He is a combatant in the sense he can be lawfully targeted by an enemy, but he cannot claim privileges appertaining to lawful combatancy.”⁵⁷

The second major criticism of targeted killings focuses on the risk to what Israel calls uninvolved individuals. Because terrorists and their commanders deliberately hide among civilians (often, but not always, with the consent or even encouragement of these “uninvolved individuals”), there is almost always some risk of “collateral damage” (to use an antiseptic term calculated to diminish the suffering of innocent civilians). This is a legitimate concern in evaluating the costs and benefits of targeted killings of terrorists, and it is a concern that Israel takes very seriously. According to Israeli military authorities, approximately two hundred terrorists have been successfully targeted, while approximately one hundred uninvolved civilians were unintentionally killed during these operations.⁵⁸Palestinian authorities obviously disagree with these figures, but they provide some rough quantification of the ratio between true and false positives. Israel also claims that hundreds, perhaps thousands of deaths to its own citizens have been prevented by these targeted killings.

In December 2003 I studied this issue at close range during a three-week visit to Israel. I met with many of those responsible for implementing the policy, as well as with critics. I watched as a high-intensity television camera, mounted on a drone, zeroed in on the apartment of a terrorist. The Israeli commander had been given authorization by the government to kill the terrorist if he got into his car alone or with other terrorists, provided that the street was empty and the likelihood of killing or injuring uninvolved civilians was minimal. The commander could easily have killed the terrorist by blowing up the house he was in, but the decision had been made not to target the house or even the apartment because there may have been others in the house and their status as combatants or noncombatants was uncertain. I watched as the camera focused on the house and the nearly empty streets. People could be seen as large dots, but it could not be determined (at least by my untrained eye) whether they were men or women, children or adults. Cars could easily be identified, and even their type—jeep, small car, large car, truck—could be determined. It was early evening, and infrared cameras could pick up the heat from the engines of moving vehicles and determine whether an engine was in the front or the back of the vehicle. I was permitted to watch for only a few minutes, and no action was taken while I was watching because the target remained in the house. He could not be arrested because Israeli soldiers and police could not safely enter Gaza City, as they could certain cities and towns on the West Bank, where arrests are used more frequently than targeted preemptions.

The next day I learned that the terrorist target did leave his home and got into his car later in the evening with another terrorist and that a missile was fired at the car, but the terrorist escaped with only minor injuries. Several others were injured as well. There were no deaths.

Earlier that day I watched a video of a previous successful attack in which two terrorists, on their way to plant explosives, were targeted and killed. The video showed a clean attack on their car in an isolated area. Two explosions can be seen. The first is of the Israeli rocket hitting the engine in the front of the car; the second is an explosion from the trunk of the car, where the explosives were being kept.

A few days before the failed preemptive attack, I had been allowed to meet with the several members of a high-level committee that had been established to consider policy questions regarding the technique of targeted preemption. The committee members included men and women with varying expertise. Among them were an eminent professor of philosophy at Tel Aviv University, a distinguished professor of international and human rights law at Bar-Ilan Law School, several lawyers of the IDF, a general who headed the military college, and a few field commanders in charge of implementing the policy of preemptive targeting in the Gaza Strip. Sometimes a mathematician specializing in probability theory attended the meetings. These men and women considered various possible scenarios, some real, others hypothetical, and discussed the policy options. Typical and not so typical cases were discussed.

The simplest case involves the ticking bomb terrorist who, alone, is on his way to a civilian population center with a suicide bomb belt strapped around his waist. Hard, confirmed intelligence makes it almost certain that unless he is stopped before he reaches his target, he will blow himself up in a crowd, killing and injuring dozens. He is in a car driving toward the target. In half an hour he will reach the population center, with its heavy traffic.⁵⁹ The decision must now be made whether to blow up his car with an air-to-ground missile. Trying to stop him in any other way—say, by arrest—will endanger the arresting officers and will almost surely cause his death anyhow since he is likely to detonate himself rather than disarm the bomb.

What are the possible objections to ordering the rocket attack under these circumstances? First, the intelligence may be wrong, and an innocent person may be killed. Error is always a possibility, even in classic self-defense cases. The man coming at you with a gun may be an actor; the gun may be unloaded; he may decide not to fire, or he may miss. The classic response to this possibility compares the likelihood of a false negative error to the likelihood of a false positive error. If the self-defense action taken appeared reasonable under the circumstances, if the person claiming self-defense had a reasonable apprehension of imminent harm and had no other reasonable way to avoid it, he should prevail in a claim of self-defense even if it turns out he was mistaken. If the alleged attacker is culpable, if he in fact was intent on harming the alleged victim, then society prefers a false positive error (mistakenly killing the culpable assailant) to a false negative error (mistakenly allowing the culpable assailant to harm the victim). Even if it turns out that the alleged assailant was not culpable (he was an actor pretending to be a killer), we excuse the mistake if it was reasonable under the circumstances because we recognize the right of a person to defend his life against reasonably perceived risks (though it cannot be said that we actually prefer a false positive to a false negative error in the context of two nonculpable people, one of whom unintentionally appears to be culpable).

Self-defense is appropriate even if the risk of death is less than certain, as the following hypothetical case illustrates. An assailant points a six-shooter at your head and tells you that there is only one bullet in the barrel and that he is going to play Russian roulette with you. If the gun fires a bullet, you are dead, but if it does not, then you are free. You believe he is telling the truth. (Perhaps you have seen him play this dangerous game with others.) Surely you should have the right to kill him if that is the only way to prevent him from exposing you to a nearly 17 percent chance of being killed.

In the context of armed conflict, the law is even more supportive of lethal action. All that is required under the laws of war is a reasonable basis for concluding that the target is a combatant, as distinguished from a noncombatant. In a traditional war, a combatant is defined as a uniformed soldier. Under the accepted laws of war, once the target’s status as a combatant is reasonably determined, he can be killed even in his sleep and regardless of his role in the military.⁶⁰ This would seem highly immoral, especially in extreme cases. If the soldiers of Nation A come upon a group of unarmed and sleeping army cooks from Nation B, under existing law they can slaughter them without even giving them a chance to surrender. No moral army would actually do this, but the traditional laws of war are based on clear lines and sharp demarcations between combatants and noncombatants.

These lines become blurred in the context of terrorism, but a ticking bomb terrorist headed toward a city with the bomb strapped around his waist certainly should qualify as a combatant under any reasonable standard. He is armed, intent on engaging in imminent lethal action, and his intended targets are noncombatants. Nor is he simply an individual criminal in the traditional sense of that word. He is part of a military, quasi-military, paramilitary, or insurgency group that is functionally closer to a military unit than to a criminal gang. He is an unlawful combatant engaged in military conflict. In the context of military conflict, there is no requirement of imminence or of the absence of other alternatives, such as arrest. As we shall see, preemptive self-defense is generally deemed more acceptable in the military than in the civilian context. Only an absolutist who believes that preemption should never be permissible (or a selective moralist with a political ax to grind against the nation employing preemption) would argue that blowing up the car of this ticking bomb terrorist should be deemed a violation of the laws of war, of human rights, or of morality. Reasonable people could argue against this action by reference to policy considerations, such as “It will only cause more terrorism,” or “It will only contribute to the cycle of violence.” But equally reasonable people could argue that taking the life of one probable terrorist to prevent the probable deaths of more than one innocent civilian is a good policy. In a democracy such decisions are properly made by the government, so long as they are lawful and reasonable. Nearly every democracy would opt for the preventive killing of a probable terrorist over a probable terrorist attack on its civilians.⁶¹ So would any reasonable moralist.

This first, relatively simple case is not typical of the difficult situations faced by the Israeli military, though there have been many such cases (including the one involving the explosives in the trunk of the car).⁶² More complex cases involve difficult variations on this paradigm. Consider the following case that was presented to the Israeli government in 2002. Salah Shehada, the leader of the military wing of Hamas, who was known to be responsible for planning and ordering numerous successful suicide bombings, was in the planning stages of yet other attacks that were “unprecedented in size and scope,” according to Israeli intelligence.⁶³ They included at least the following: “rigging a truck with 600 kilograms of explosives to blow up the recently constructed Gush Katif bridge, the sole route used by hundreds of Israelis daily to travel to and from their communities . . . massacr[ing] residents of Gush Katif communities by carrying out an attack during the celebrations currently under way there, and . . . planning to have suicide bombers infiltrate Israel for a number of bomb attacks to be carried out in populated areas in Beersheba. . . . abduct[ing] soldiers and civilians to gain the release of Palestinians in Israeli jails, and plann[ing] to rig a boat with explosives to be blown up on one of the beaches used by residents of the Gush Katif communities.”

He was also using “young Palestinian children” to “carry out suicide attacks against Israeli targets, arming them with homemade weapons.” This intelligence was gathered “from Palestinians affiliated with the Hamas in the Gaza Strip, who operated under Shehadeh’s command.”⁶⁴ At least one of the massive attacks was believed to be imminent. He was “the ultimate ticking time bomb,” according to an Israeli spokesman.⁶⁵ “We did not target him simply to retaliate or as punishment. We did it as a preemptive operation. . . . He planned to send people to carry out a massacre tonight in Gaza.”⁶⁶ Arresting him was not feasible. If it were, it would have been the option of choice.” ‘The optimal move is to arrest a terrorist, because then you get information out of him,’ a senior Prime Minister’s Office official said. ‘The next option is to kill him, because we would rather prevent terror attacks than react after the fact.’ ”⁶⁷

Killing him, it was reasonably believed, would thwart the planned future attacks. The government therefore decided that he was an appropriate target under the principles of the first case, so long as he could be killed without undue risk to other noncombatants. The problem was that “he constantly changed his living quarters, moving from safe house to safe house, chosen purposely in densely populated residential areas, from where he continued to give orders for attacks to be carried out.”⁶⁸ He often traveled and slept with his wife at his side. Although his wife was fully aware of, and presumably supported, his activities, she was regarded as an “uninvolved person.” The military commanders sought permission from the government to attack him with a rocket when he was alone with his wife, arguing that he was a sufficiently important military target to justify the killing of one noncombatant, especially since that particular noncombatant was far from a “total innocent,” having allowed herself to be used as a “human shield” by her combatant husband. After considerable debate, the decision was made to authorize the “hit,” but only if it was clear that he could not be found alone and only if there would likely be no other casualties among noncombatants. Three days before they ultimately bombed him they had “called off a strike . . . when they discovered he was with family members.”⁶⁹ When they did bomb him, they were not aware that there were civilians close by. The rocket killed not only the terrorist commander and his top aide but also his wife, his fourteen-year-old daughter, two other relatives, and several neighbors who were noncombatants. The undue number of civilian casualties was apparently caused by a breakdown in intelligence, and the reaction from the Israeli media and public was extremely critical of the action.” We did not intend to cause this number of civilian casualties,” an Israeli military official was reported as saying. “If we had known this would be the result, we would not have taken this action today.”⁷⁰ Yossi Sarid, who at the time chaired the “dovish” Meretz Party, said that though the assassination was justified, “the timing was inappropriate, the assassination method was wrong, it was a kind of act of terror.”⁷¹ But if the intelligence that he was planning a massive attack that very night was accurate, a postponement might have cost the lives of even more Israeli civilians than the number of innocent Palestinian lives that were lost by the preemptive killing. Such is the nature of tragic choices caused by ticking bomb terrorists who deliberately hide among their own civilians.

Another variation on the theme of targeting suspected terrorists involves the targeting of individuals who are the religious or political leaders of terrorist groups. In 2004, Israel killed two Hamas leaders in quick succession. One was Sheikh Ahmed Yassin, and the other was Dr. Abdel Aziz Rantisi. Both were killed with minimal “collateral damage,” yet their targeting was controversial and widely condemned. I wrote the following article about these targeted killings, comparing Israel’s action with those of the United States and Great Britain:

The United States Army was recently given a highly specific military order. According to the top U.S. commander in Iraq, Lt. General Ricardo Sanchez, the mission is to kill radical Shi’ite Cleric Muqtada al-Sadr.

This order to target al-Sadr for extrajudicial killing is perfectly legitimate and lawful under the laws of war. Al-Sadr is a combatant, and it is proper to kill a combatant during an ongoing war unless he surrenders first. It doesn’t matter whether the combatant is a cook or bomb-maker, a private or a general. Nor does it matter whether he wears an army uniform, a three-piece suit, or a kaffiyeh. So long as he is in the chain of command, he is an appropriate target, regardless of whether he is actually engaged in combat at the time he is killed or is fast asleep. Of course, his killing would be extrajudicial. Military attacks against combatants are not preceded by jury trials or judicial warrants.

Al-Sadr fits squarely into any reasonable definition of combatant. He leads a militia that has declared war on American and coalition forces, as well as on civilians, both foreign and Iraqi. He is at the top of the chain of command, and it is he who presses the on-off button for the killings. Like Osama bin Laden and Mullah Omar Mohammed, he is a proper military target, so long as he can be killed without disproportional injury to noncombatants.

If American forces can capture him, they are permitted that option as well, but they are not required—under the laws of war—to endanger the lives of their soldiers in order to spare al-Sadr’s life. Indeed, unless al-Sadr were to surrender, it is entirely lawful for American troops to kill him rather than to capture him—if it were decided that this was tactically advantageous.

Although U.S. commanders mentioned capture along with killing as an option, it may well be preferable not to capture al-Sadr, for fear that his imprisonment would stimulate even more hostage-taking in an attempt to exchange hostages for al-Sadr. The order to kill or capture al-Sadr may well be a euphemism for “kill him unless he surrenders first” (as Saddam Hussein did).

The world seems to understand and accept the American decision to target al-Sadr for killing, as it accepts our belated decision to try to kill bin Laden and Mullah Omar Mohammed. There has been little international condemnation of America’s policy of extrajudicial killing of terrorist leaders. Indeed, the predominant criticism has been that we didn’t get bin Laden and Mullah Omar Mohammed before September 11.⁷²

How then to explain the world’s very different reaction to Israel’s decision to target terrorist leaders, such as Sheikh Ahmed Yassin and Dr. Abdel Aziz Rantisi, the former leaders of Hamas. Surely, there is no legal or moral difference between Yassin and Rantisi on the one hand, and al-Sadr and bin Laden on the other. Yassin and Rantisi both personally ordered terrorist attacks against Israeli civilians, approved them in advance, and praised them when they succeeded.

Each was responsible for hundreds of civilian deaths and was involved in ordering and planning more terrorist attacks at the times of their timely deaths. They were terrorist commanders, just as al-Sadr was. They were both killed, along with their military bodyguards, in a manner that minimized civilian casualties, despite the fact that they generally—and unlawfully—hid among civilians, using them as human shields. Israel waited until they, and their fellow terrorist guards, were alone and then targeted them successfully. There was no realistic possibility of capturing them alive, since they had sworn to die fighting; and any attempt to extirpate them from the civilians among whom they were hiding would have resulted in numerous civilian casualties. (Israel does try to capture terrorist commanders in the West Bank, where it has large numbers of troops on the ground, but it employs targeted killings in Gaza, where it has a far more limited military presence.)

Reasonable people can disagree about whether the decision to target Yassin, Rantisi, al-Sadr, bin Laden, or any other terrorist is tactically wise or unwise, or whether it will have the effect of reducing or increasing the dangers to civilians. But no reasonable argument can be made that the decision to target these combatants—these terrorist commanders—is unlawful under the laws of war or under international law.

British Foreign Secretary Jack Straw was simply wrong when he declared that targeted assassinations of this kind—specifically referring to the killing of Yassin and Rantisi—are unlawful and in violation of international law. And he knows it, because his own government has authorized the killing of terrorist leaders who threaten British interests.

I challenge Straw to distinguish Israel’s killing of Yassin and Rantisi from the coalition’s targeting of al-Sadr, Saddam Hussein and his sons, Osama bin Laden, and Mullah Omar Mohammed.

He could not do so. Any claims that Hamas is divided into military and political (or religious ) wings is belied by the fact that Yassin and Rantisi both ordered the military wing of Hamas to engage in acts of terrorism and approved specific murderous acts in advance.

If Straw cannot distinguish these situations, then does he disapprove of the American policy of killing al-Sadr? If British troops were to have al-Sadr—or, for that matter, bin Laden—in their sights, would they have to hold their fire because Straw has told them it would be illegal to pull the trigger?

We have a right to know the answers to these questions, since American and British troops are supposedly operating under the same rules of engagement. Or would Straw simply (and honestly) say he is not applying the same rules to Israel as he is to his own nation and its military allies?

The international community cannot retain credibility if it continues to apply a different, and more demanding, standard to Israel than it does to more powerful nations.^73*

The debate over targeted killings persists. As is typical of the Israeli legal system—unlike the American system—the issue of targeted killings of suspected terrorists, like so many other contentious ones, has been brought before the courts⁷⁴ by Israeli human rights groups. The case had been fully briefed when Israel, as part of its cease-fire with the Palestinians, decided to suspend targeted killings, and the case is now on hold, but the controversy—within Israel and around the rest of the world—persists. Russia announced, following the hostage takings and deaths in the Beslan school in September 2004, that its military would henceforth engage in preemptive actions against Chechnyan terrorists, presumably including targeted killings of their leaders and commanders. Both candidates in the American presidential election of 2004 pledged to kill Osama bin Laden and other terrorists who endanger the United States.

In October 2004 a debate erupted following the publication in the British medical journal Lancet of a study suggesting that approximately a hundred thousand Iraqi civilians had died as a result of the American invasion of Iraq. These numbers were hotly disputed, and the actual figures will never be known.⁷⁵ An organization called Iraq Body Count has placed the number of civilian dead at between 27,115 and 30,559 as of November 28, 2005. It is beyond dispute that a large number of collateral deaths were caused by American and British bombings and other military actions—far greater than the number of deaths caused by Israel’s policy of targeted killings. Some of the coalition bombings were precisely targeted; others less so. The issue should not be framed in terms of whether targeted killings should ever be permitted—clearly they should under some circumstances—but rather in terms of the proper criteria for employing this extraordinary means of preventive self-defense. It also should be framed by reference to the available alternatives.

A report issued in November 2004 by a group under the direction of Professor Philip Heymann of Harvard Law School proposed that targeted killings of terrorists who pose a real and imminent threat should be authorized by the law, provided that specified rigorous criteria and procedures are satisfied. The report concluded:

[T]argeted killings against known terrorists have become a real and accepted option within the United States as the only reasonably effective way of reaching a hostile target. The targeted killing of a terrorist could prevent a planned attack and could serve as a deterrent to terrorist groups or to individuals who may be inclined to join them. Targeted killing, in some contexts, would also improve domestic morale because it shows progress against a specific terrorist enemy. One can readily imagine the impact of the known death of Osama bin Laden on the sense of security in the United States. Still, the most basic purpose is trying to stop the next attack. The targeted killing of a leader or a critical member of a terrorist group may temporarily incapacitate the organization and at least delay terrorist activity. Given the other options available, killing a terrorist receiving shelter in a hostile state would be far less costly in terms of lives and money than invading that state. Where a threat is imminent, targeted killing—which does not require extensive evidence gathering for trial or preparation for full-scale invasion—also provides needed speed.⁷⁶

The report observed that the existing executive order, which again represents the “way of the hypocrite,” mandates an absolute prohibition on anyone “acting on behalf of the United States” participating in any assassination. Despite this paper prohibition, the United States has targeted Osama bin Laden and several other terrorist leaders for assassination. We actually assassinated the terrorist Qaed Salim Sinan al-Harethi by bombing a car in which he—along with five others, including an American citizen—was driving in Yemen. The current executive order is, according to the report, “basically no standard at all.”⁷⁷

Accordingly the report proposes a “three-pronged test” for targeted killings: “[T]o be ‘necessary’ means that there is no other reasonable alternative, that targeted killing is a practice of last resort; to be ‘reasonably imminent’ means that the development of an alternative (capture, arrest, etc.) would not eliminate a real likelihood of imminently threatened, lethal attack or would be inordinately dangerous to U.S. or allied personnel; and, finally, to be preventive, the targeted killing can only be for prospective purposes, rather than as retribution for previous bad acts under those standards.”⁷⁸

These criteria, while generally sensible, lack the specificity that is required to focus accountability on decision makers at various levels of responsibility. I propose some such specificity in Chapter 7.⁷⁹

On January 25, 2005, Israel announced that it would stop the targeted killings of Palestinian terrorists as part of a truce or cease-fire.⁸⁰ Hamas agreed in turn to suspend its attacks against Israeli civilians. Both sides reserved the right to resume their actions if the other side broke the cease-fire. In late September 2005, following the Israeli withdrawal from Gaza, Hamas conducted rocket attacks against the town of Sderot, resulting in several injuries. In response, Israel launched air strikes in Gaza against Hamas targets. Shortly thereafter, Hamas announced that it was ending the rocket attacks—at least for the time being. On November 14, 2005, Hamas threatened to renew its terrorist attacks against Israel after the Israeli Army shot and killed a Hamas commander in a raid in Nablus.

Preemption and Prevention of Biochemical Attacks In January 2005 it was reported that a CIA-affiliated think tank was “predicting that America is ‘likely’ to be hit by bio-terrorist attacks at some point in the next 15 years. ‘Bioterrorism . . . appears particularly suited to the smaller, better-informed groups. Indeed, the bioterrorist’s laboratory could well be the size of a household kitchen, and the weapon built there could be smaller than a toaster. Terrorist use of biological agents is therefore likely, and the range of options will grow. Because the recognition of anthrax, smallpox or other diseases is typically delayed, under a “nightmare scenario”an attack could be well under way before authorities would be cognizant of it.’ ”⁸¹

A leading expert on bioterrorism, who serves as a consultant to the Pentagon, warned that although biological agents have the “potential to be much more life-threatening [than nuclear weapons], we are not prepared to prevent or control this threat.” Nor is America’s “fragmented and stressed” health care system prepared for a biological attack because it is “structurally unfit” for any adequate response.⁸² According to a feature article, “Remain Calm,” in the November 14, 2005, issue of New York magazine, Dr. Irwin Redlener, head of the National Center for Disaster Preparedness at Columbia, has observed that “[f]our years after 9/11, we are, as a nation, extraordinarily, inexplicably unprepared to deal with a major catastrophic event.” The inadequate response to Hurricane Katrina in September 2005 only confirms these concerns.⁸³ On November 1, 2005, President Bush announced a $7.1 billion plan to prepare for a possible avian flu pandemic, following the death of sixty-two people in sixteen countries from this disease. “Our country has been given fair warning of this danger . . . and time to prepare,” the president said. “It’s my responsibility . . . to take measures now to protect the American people.” We may receive no warning, however, of a secret biological attack.

The strategy for anticipating a biological or chemical attack, from an enemy nation or a terrorist group, is in some respects the same as for other weapons of mass destruction or conventional mass casualty attacks: effective intelligence; preventive destruction of the capacity to develop, deploy, and deliver the weapons; deterrence by threat of massive retaliation; and other diplomatic, economic, political and military options. But there are important differences as well. Steps can be taken to neutralize or diminish the threat of at least some biological and chemical weapons. These steps include a program of widespread inoculation capable of preventing much of the damage sought to be wrought by those contemplating a biological or chemical attack. Other more reactive steps, which, as we shall see, also have prevention components, would focus on developing plans and procedures for an immediate and effective first response, including the use of gas masks and sealed rooms to protect against the deadly chemicals, a quarantine program designed to stanch the spread of contagious disease, a pharmaceutical response designed to minimize the effects of the germ or chemical weapons, the widespread availability of detoxification facilities capable of cleaning the body of toxic materials, and other combinations of public health and conventional medical and pharmaceutical responses.

Some of these steps, such as the construction of detox facilities, are relatively cost-free, at least in terms of human life. They simply require expenditures of money. Israel has already begun to construct facilities that include powerful cleansing showers in the event of a chemical attack.⁸⁴ New York City has installed “extensive monitoring systems to catch problems before they occur.”⁸⁵ These problems include chemical attacks. A BioWatch system registers biological dangers. Wind patterns are checked to determine the likely paths of anthrax if it were to be released into the air. One expert recently concluded that “an outdoor anthrax release over a city would not be difficult at all.”⁸⁶ Other steps, such as widespread availability of injections or pills for use following an attack, may cost some lives, since almost any powerful drug produces some serious side effects, including death. Even gas masks can create a hazard to life when improperly used, as they inevitably are by some. When Iraqi Scuds rained down on Israeli cities during the First Gulf War, and it was feared that they contained chemical warheads, several people died from putting on their gas masks. It turned out that the Scuds contained only conventional explosives and preventive measures were unnecessary, but more Israelis died from the gas masks than from the missiles.⁸⁷

If a biological or chemical attack is confirmed, then the benefit of a pharmaceutical response outweighs the costs, since the number of lives to be saved by the injections, pills, or other measures will greatly exceed the number of lives lost from the pharmaceuticals themselves.

The same cannot necessarily be said for a program of widespread or universal inoculation in anticipation of a biological or chemical attack. If the attack is relatively certain and immediate, and if the inoculation is very effective, the benefit will exceed the cost. But if this attack is uncertain and distant, the number of people who will inevitably die from receiving the inoculation may well be greater than the risk of attack discounted by the unlikelihood that it will actually occur.

Much useful thought has gone into the calculation of costs and benefits in the context of the feared weaponization of smallpox.⁸⁸ It is uncertain whether any of America’s enemies actually have access to the smallpox virus and are capable of weaponizing it. But recent advances in synthesizing long molecules of DNA have made it more likely that “the technique might be used to make the genome of the smallpox virus”⁸⁹ and weaponizing it. What is known is that if terrorists can manage to introduce smallpox into our borders, the devastation may be catastrophic since most Americans are not currently immunized against that killer disease. What is also known is that an extremely effective inoculation that could neutralize the threat with a high degree of success is available. But that success would come at a considerable cost, since a small but not insignificant number of people who receive the inoculation will suffer serious permanent and devastating medical consequences, and a smaller number will die.⁹⁰

According to a recent report by a group of doctors at the University of Michigan, “We estimate that a vaccination strategy directed at people aged 1 to 29 years would result in approximately 1600 serious adverse events and 190 deaths. Vaccinating people aged 1 to 65 would result in approximately 4,600 serious adverse events and 285 deaths.”⁹¹

Because the threat of a smallpox attack is deemed to be small—despite the catastrophic effect of such an attack were it to occur in the absence of widespread inoculation—the current approach apparently taken by public health and terrorist experts is to prepare for the possibility that inoculation may be prudent at some point in the future if the likelihood of an attack were to increase, but that no program of widespread preventive inoculation is warranted now.

An alternative to widespread inoculation, with its inevitable costs in human life and health, would be more limited targeting for preventive inoculation only of specific categories of such people as first responders, soldiers, public health workers, those assigned to handle any quarantines, and others. It might also include much larger groups, such as all medical personnel, all residents of particularly vulnerable cities, all residents of an entire nation, and all people with particular susceptibilities, such as the very old and the very young, or even only those who chose to accept the risks and benefits of being inoculated (though if the germ is highly contagious, individual choice might not be a viable option). For each such group, the balance of risks might be somewhat different, but for each such category a probabilistic determination would have to be made and an appropriate balance struck. Whether this balance should be struck by legislative, judicial, administrative, or other institutions and whether it should be a matter of federal or local decision making involve important issues of democratic theory.

Another alternative is the quick availability of inoculation, coupled with quarantine, at the first sign of a smallpox attack. Public health experts have devised a ring strategy for reducing the spread of smallpox or other highly contagious dangerous diseases. An appropriate balance must be struck between too much and too little inoculation.⁹²

Whether inoculation is employed either before or after a biological attack, quarantine will be part of any effective response to a highly contagious disease. The United States, however, has no quarantine law on the books designed to cope with a biological attack. Most state laws are anachronistic both medically and legally. They were enacted in response to natural outbreaks of such contagious diseases as tuberculosis, which no longer pose significant public health hazards.

Were a highly contagious and deadly virus to be transmitted through the air filtration (or air-conditioning or heating) system of a large, fully occupied skyscraper (such as the Empire State Building), a terrible choice of evil problems would be presented. Should the building be sealed off and all the people in it be quarantined (at least until they can be tested)? If the decision to do so were to be the best public health response to the crisis, there might be no legal basis under any specific statute for issuing and enforcing such an order, especially against those who would argue that they had not yet been exposed to the virus, but who reasonably feared exposure if required to remain in quarantine alongside exposed and contagious individuals.⁹³ I have little doubt that the courts would in fact sustain any reasonable public health order, citing the emergency and perhaps martial law as declared by the proper authorities. But, to paraphrase an old saw, “Martial law is to law as martial art is to art.”⁹⁴ It is essentially a lawless assertion of power labeled as “necessity,” and to quote an ancient aphorism, “Necessity knows no law.” (In every jurisdiction, there is at least one judge known to the bar as “necessity” because he too “knows no law.”) It would be far better to enact specific quarantine laws in advance of any biological attack. They should address the issues likely to be faced in the event of such an attack. A leading scholar of quarantine laws has concluded that “both law and science have changed since the time when quarantine was a standard tool against infectious disease. A court giving unquestioning approval to quarantine in modern circumstances would be utilizing an anachronistic rationale.” But this was written in 1985, in the context of the AIDS epidemic. It did not take into account the weaponization of contagious disease. The time has come for experts to rewrite the laws of quarantine, making them both more responsive to current needs and more protective of civil liberties.⁹⁵

On a broader level, the legal structure for dealing with biological or chemical warfare or terrorism is woefully inadequate. It must be addressed before an attack occurs if chaos is to be reduced and the role of law preserved, and the criteria for requiring people to accept costs and risks—whether of inoculation, quarantine, or any other burden—must become part of an overall approach to or jurisprudence of preemptive and preventive governmental actions.

Prior Restraint of Expression

Throughout history, expression—speech, writing, religious practices, art, and assembly—has been restrained on the basis of predictions that if not prevented, it would lead directly, or indirectly, to harms ranging from undercutting the authority of the church to overthrowing the government, rape, even genocide. Opponents of prior restraint or censorship often argue that there is no empirical relationship between the restraining expression and the predicted harms. For example, many civil libertarians point to the absence of evidence that pornography causes (or contributes substantially to) the incidence of rape (or related sexist evils).⁹⁶ Others argue that even if there is a direct causal relationship between certain kinds of pornography (for example, violent sexual portrayals) and certain evils (such as rape, harassment, and devaluation of women), the cost of prior restraint or censorship is too high.⁹⁷

There can be no question that some kinds of expression contribute significantly to some kinds of evils. The International Criminal Tribunal for Rwanda found a direct causal relationship between certain radio broadcasts and genocide,⁹⁸ just as the Nuremberg Tribunal had found a relationship between the anti-Semitic diatribes of Der Stürmer and the genocide against the Jews. The International Criminal Tribunal for Rwanda convicted three men of genocide, conspiracy to commit genocide, and other crimes for allegedly inciting violence against the Tutsis. One of the defendants was charged with helping “to set up the Radio Télévision Libre des Mille Collines (RTLM) in order to promote the ideology of Hutu extremism. The programmes broadcast by RTLM allegedly promoted ethnic division and incited the murder and persecution of persons of Tutsi origin.” Another was accused of “broadcasting messages on national radio that incited ethnic hatred and murders of the Tutsi people.” A third was “alleged to have consented to the publication of material in the Kangura Newspaper which promoted ethnic hatred and incited mass killings and/or serious bodily or mental harm to the Tutsi.” In convicting the three, the court noted that the case “raises important principles concerning the role of the media, which have not been addressed at the level of international criminal justice since Nuremberg. The power of the media to create and destroy fundamental human values comes with great responsibility. Those who control such media are accountable for its consequences.” Two of the defendants received life sentences, and the third thrity-five years in prison. All three have appealed.⁹⁹

There are various mechanisms by which the relationship between expression and action can operate. Sometimes it is simply informational. Facts that facilitate previously decided-upon goal are communicated. For example, one nation that wishes to attack another is given secret information. It could be anything from attack plans to vulnerable areas to the recipe for producing a nuclear weapon. Sometimes the information is largely emotional: incitements to kill; dehumanization of the enemy; religious commands to kill. For example, the material that appeared in Der Stürmer or in some of the sermons that are commonly broadcast on Arab and Islamic television stations may well have contributed to the willingness of readers or listeners to kill. Sometimes there is a combination of factors. For example, the media in Rwanda combined emotional incitements to kill Tutsis, with precise information about the whereabouts of particular Tutsi leaders. Some accounts of the genocide suggest that it was common for members of militias in search of Tutsis to set up roadblocks and listen to their transistor radios, which broadcast messages to try to rile up listeners’ hatred.¹⁰⁰

The constitutional law of the United States generally prohibits governmental censorship of “speech,”¹⁰¹ but various exceptions have always been recognized. Within five years of the ratification of the First Amendment, Congress enacted the Alien and Sedition Laws, which punished seditious speech that was critical of the administration of President John Adams.

Prior restraint of speech, as contrasted with subsequent punishment of speech, has rarely been authorized by the courts. In the famous Pentagon Papers case, the Supreme Court drew a sharp distinction between the two mechanisms of controlling speech that is alleged to be dangerous.”Prior restraints require an unusually heavy justification under the First Amendment,” Justice Byron White wrote in his concurring opinion, “but failure by the Government to justify prior restraints does not measure its constitutional entitlement to a conviction for criminal publication. That the Government mistakenly chose to proceed by injunction does not mean that it could not successfully proceed in another way.”¹⁰²

The rationale of the court was that distinguished newspapers like the New York Times and the Washington Post would remain accountable after publication for violating any laws that would result from publishing impermissible material. In other words, because they could be deterred, there was no basis for seeking to prevent violation of the law by the extraordinary and disfavored mechanism of prior restraint. But those who publish on the Internet might not be as amenable to deterrence since many of them are not traditional publishers, with return addresses and deep pockets. For purposes of deciding whether deterrence or prevention is the more appropriate mechanism of control, these “fly-by-night publishers” have been compared with nondeterrable terrorists.

At the Practising Law Institute’s Communications Law 2002 Conference, panelists “struggled to reconcile recent court decisions allowing speech restrictions with their notions of prior restraint doctrine, finding new issues arising as more and more communication is accessed through the Internet.” One case the panel examined arose after the American Coalition of Life Activists “distributed posters and posted on the Internet names and other identifying information of doctors who perform abortions, crossing out those who had been killed or injured.”¹⁰³

Recently there have been efforts at self-policing by Web logs and other Internet operators, particularly in the context of assuring credibility: “There have even been fledgling attempts to create ethical guidelines, like the ones found at Cyberjournalist.net. Defenders of the status quo argue that ethics rules are not necessary in the blogosphere because truth emerges through ‘collaboration,’ and that bias and conflicts of interest are rooted out by ‘transparency.’ But ‘collaboration’ is a haphazard way of defending against dishonesty and slander, and blogs are actually not all that transparent. . . . Bloggers may need to institutionalize ethics policies to avoid charges of hypocrisy. But the real reason for an ethical upgrade is that it is the right way to do journalism, online or offline.”¹⁰⁴

These efforts do not, however, address the problems of irremediable harms that could be caused by irresponsible bloggers, such as death and disclosure of state secrets or terrorist targets. If terrorism were to increase, and if the Internet were to become an important tool for the communication of dangerous information, and if those employing the Internet were deemed undeterrable by after-the-fact sanctions, then the push for preventive tools, including expanded notions of prior restraint, would increase. This is the time to begin serious consideration of the conflicting claims of liberty and security in this important context.

As terrorism increases, I predict that there will be increasing calls for prior censorship of speech that is believed to incite suicide bombers and others who target civilians. The American constitutional law of incitement as an exception to the First Amendment has always varied with the perceived dangers of the times. During World War I, handing out leaflets against the draft was analogized to falsely shouting fire in a crowded theater. During the McCarthy era, being an active member of the Communist Party was deemed to be prohibited advocacy of violent overthrow of the government. These were unjustified constraints on freedom of speech since there was no reasonable basis for concluding that these expressions constituted a clear and present danger of serious harm.

The same cannot necessarily be said about charismatic imams inciting young believers to become suicide bombers.¹⁰⁵ The competitive marketplace of ideas is not always open to these zealots, who are often not exposed to the general media and live controlled religious lives in which the imam’s word is law. The culture of death coupled with the promise of a glorious afterlife for martyrs may be irresistible to some.

It is only a matter of time before serious proposals for prior restraint of this kind of direct incitement will be offered.¹⁰⁶ Already some imams have been threatened with deportation from France, Australia, and Britain for their incitements.¹⁰⁷ Deportation is one form of preventive restraint on speech. Others will not be long in coming. We must be prepared to confront this challenge to freedom of expression and religion.

Other Preventive Mechanisms

There are other mechanisms of prevention that are more specific to particular locations or disputes and warrant only brief mention. The military “occupation” of an area from which terrorism originates may have preventive components. The U.S. presence in Afghanistan and Iraq, and the Israeli occupation of the West Bank and Gaza, were sought to be justified on terrorist prevention grounds. Opponents claim that it is the occupations that stimulate, or at least contribute to, the terrorism. Israel points to the fact that it largely ended its day-to-day occupation of the major West Bank cities shortly after the Oslo agreements were signed in 1993 and that it reoccupied some, but not all, only after the onslaught of suicide bombing that followed Arafat’s refusal to agree to the offers made at Camp David and Taba in 2000 and 2001. Reoccupation became necessary as a means of controlling and preventing the terrorism that was emanating from the West Bank cities. The Palestinian perspective is of course quite different, and this is not the place to try to resolve these differences. I mention this dispute only to point out the allegedly preventive nature of the reoccupation.

Related to the above is the construction by Israel of the security fence that it claims is also designed to prevent terrorist attacks and that it sees as a more passive substitute mechanism for occupation. It is, of course, the location of the fence rather than its construction alone that has generated so much controversy. Again, this is not the place to address that controversy, but rather to list the mechanism of a security fence within the catalog of preventive tactics that have been employed in an effort to control terrorism.

Either with or without security fences, checkpoints control the flow of human and cargo traffic and can serve as a preventive measure against terrorism. But these same checkpoints impede legitimate traffic and make it difficult for innocent people to get to work, to hospitals, and to families. Striking the proper balance between the legitimate preventive needs and the illegitimate harassment effects of such checkpoints is never without considerable controversy. Fences and checkpoints are of course part of the larger issue of border controls. In the United States there are large, uncontrolled borders to both the north and the south, as well as large stretches of unpatrolled shorelines. Efforts to control these vulnerable borders, both by the government and by citizens’ groups, generate considerable controversy because it is never easy to separate legitimate concerns over terrorism with less legitimate concerns about limiting the influx of Mexicans and others seeking to come to the United States for better lives.

Another controversial preventive tactic is the sting or the scam, in which law enforcement agents pretend to be terrorists, arms dealers, or other facilitators of terror. They actively seek to encourage the members of terrorist cells who are in deep cover to surface by offering their services and wares. Related to this tactic is the recruitment of informers from within the ranks of terrorists. A more passive variation on this mechanism is the planting of spies or electronic spying devices among terrorists or potential terrorists.

Other preventive mechanisms include mass roundups of suspects; sweeps for weapons; profiling; random stops, frisks, and searches; preventive arrests; material witness detentions; dog sniffs; and sophisticated chemical detection devices.

Only the human imagination, current technology, the law, and morality limit what steps could be taken in an effort to prevent terrorism.¹⁰⁸ The mechanisms described above are intended to be illustrative of those currently in use and those likely to be tried if terrorism increases. Each of these preventive mechanisms presents perils and prospects that must be thoughtfully considered on an ongoing basis. But they should not be considered only on an ad hoc, case by case basis. They must become part of a developing jurisprudence of early intervention, not only in the context of terrorism but in the context of other serious harms as well. Only the greatest of these harms can ever justify resort to the preventive mechanism of full-scale war—such as that which the United States waged against Iraq in March 2003—since such a war imposes the highest of costs on those attacked and often on those attacking. It is to that war that we now turn.

^* On December 4, 2005, the New York Times reported the targeted killing, apparently by the CIA, of Hamza Rabia, the number-three man in al-Qaida. Mohammed Khan and Douglas Jehl, “Attack Kills a Top Leader of Al Qaeda, Pakistan Says,” New York Times, December 4, 2005, p. 24.