New Scientist Article - How To Foil An Assassin

I was going through my (paper) archives the other day, where I've collected a lot of information from various sources - pre-internet predominently - and found a cutting from a 1993 edition of New Scientist, titled 'How to foil an Assassin' authored by Jiri Kubie.

At the time I found it to be extremely interesting, and from a completely different perspective than I was accustomed to - namely a non-tactical mathematical-model.

I managed to find an electronic version of it thankfully - better than a scan of the hardly pristine cuttings I have, and re-typing the article was out of the question!

Here it is:

----------

How to foil an assassin: Top politicians who use too many bodyguards may increase the risk of being assassinated, mathematics shows. This is a powerful lesson for sites where safety is critical.

Everybody runs a risk of being attacked, but most of us can take simple precautions to minimise the risk. For example, we avoid walking in parks or deserted city areas after dark or when other law-abiding citizens are unlikely to be present. But for politicians, high-profile public servants, business executives or, indeed, any well-known person, the risks of being attacked, injured or even killed are much higher. As history shows, assassination attempts on public figures are only too common.

For example, Julius Caesar was not the only Roman emperor to be assassinated. Indeed, it is difficult to think of any Roman leader who died of old age. The 19th century was particularly marked by assassinations, successful and attempted, throughout Europe. In the present century, almost everyone has heard of the assassination of Archduke Franz Ferdinand at Sarajevo which precipitated the First World War. Many people still remember the moment when they heard that President Kennedy had been assassinated in 1963, or that John Lennon had met an untimely death in 1980. In 1981 there were the attempted assassinations of President Reagan and Pope John Paul and the assassination of Egypt's President Sadat; 1984 saw the attack on the leadership of the British Conservative Party in Brighton and the assassination of Indira Gandhi. But, in addition, every year there are countless successful and attempted assassinations of other politicians, business people and public servants in countries as diverse as Britain, Germany, India and Algeria which are quickly forgotten by all but close relatives and friends.

There is no doubt that, compared with most of us, people in public life run a high risk of being attacked. For VIPs, and especially for senior politicians and heads of state, a strategy to minimise the risk is imperative. No security service would be willing to discuss what form such a strategy might take, but using probability theory and mathematical models we can evaluate the strengths and weakness of certain aspects of such strategies.

Model lessons

At Middlesex University, we have developed a simple mathematical model which includes the basic parameters of a possible attack, such as the behaviour of the politician, the assailant or assailants and the available defences. The model is simple because its purpose is not an accurate prediction of how best to protect VIPs, but to see what general lessons we can learn about reducing risks and then apply them to more esoteric forms of risk. For example, the model can easily be extended to consider the levels of risk of large-scale industrial or transport disasters, or of injury while using a controlled pedestrian crossing.

The model reveals that the major components that affect the level of risk for a politician are: the number of potential assailants; the frequency of exposure; and the probability that, given an exposure and an assailant, that person will succeed in inflicting injury. It follows that to decrease the risk, political figures should ensure the number of potential assailants is as low as possible. Numbers, and the determination of adversaries, however, can vary greatly with political and economic climate. And, as history shows, they can be difficult to influence.

One tactic, employed by many a royal or imperial household in the past, is to eliminate possible rivals and those groups thought to be hostile. For example, Queen Elizabeth I beheaded Mary, Queen of Scots in 1567 because she was too great a threat to the English crown. More recent were Stalin's vicious purges in the Soviet Union during the 1930s. Another common strategy is exile or internment. Before 1917, the Russian tsars exiled many a revolutionary, including Lenin, to keep them at bay. In present-day Egypt, the government keeps certain religious and political leaders in detention. A less draconian measure to reduce the exposure of a visiting VIP to potential assailants is to isolate those thought to be unfriendly to the visitor. In the 1950s, for example, the French government 'invited' well-known troublemakers among Russian exiles and emigres to stay in country mansions or comfortable hotels during an important visit by Soviet dignitaries. The visitors were unharmed, but whether that was a consequence of adopting this particular tactic is another question.

On the whole, though, such measures are unacceptable in a democratic society. Furthermore, controlling the number of potential assailants can be counterproductive. Sometimes it may provoke a pre-emptive strike from the threatened faction as Lavrenti Beria found after Joseph Stalin's death in 1953 when he attempted to use his position as chief of secret police to win the struggle for power. His opponents joined forces to accuse him of anti-party activities and engineered his execution before the year was out. In other cases, suppressing or banning opponents can drive troublemakers underground, or lead to their regrouping (as in the banning of neo-Nazis in Germany), both of which make surveillance and control more difficult.

Given the difficulties of reducing the number of assailants, a politician should be exposed to them as rarely as possible. One way to achieve this is to limit public appearances, especially during times of national or international tension when the number of potential assailants increases. Taken a stage further, they can appear only before invited, screened, searched or known audiences. In some societies this is refined to the ultimate: the politician is never seen by the public but only heard (and sometimes only heard of). Appearances are confined to an inner circle of close colleagues. Stalin successfully used this ploy in the early 1950s, as did Chairman Mao in the mid-1970s. Isolation may be effective for decreasing the risk, but is not always foolproof. Political figures can be at risk from potential assailants within their inner circle, as in the case of the attempted assassination of Hitler in 1944 by Colonel Claus von Stauffenberg.

Double acts

Isolation is also not feasible in democracies, because public exposure of politicians is an important part of the political process. An alternative approach for vulnerable people is to employ doubles. The press has frequently carried stories suggesting that Winston Churchill used doubles during the Second World War. This obviously decreases the risk to the politician, while conceding to public demands; what it does for the hapless double can only be imagined.

Since exposure is unavoidable, top politicians and heads of state need an effective and efficient means of neutralising the threat. This means employing defences or safeguards to reduce the probability of a successful attack. The defences can be either passive, in that they attempt to stop an attack being successful, or active in that they are designed to prevent an assailant from even initiating an attack.

There are many examples of passive defences, from armoured limousines, fortified homes and bulletproof vests to particularly tall bodyguards acting as a human shield around the VIP. An increasing number of politicians and business people use passive defences which, while effective, have limitations. Overconfidence and carelessness about keeping exposure to a minimum, for example, can negate the passive defences. Bulletproof vests and the like can reduce mobility, making evasive action more difficult. A person employing passive defences may also present an enhanced target. A VIP on foot, for example, may be attacked with small arms, whereas one in an armoured limousine is more likely to be attacked with a bomb, mine or missile. The chances of surviving such an attack are much lower.

Active defence falls into two categories: self-defence and external defence provided by other people. The first involves varying daily routines, making lives less predictable, taking lessons in unarmed combat or carrying weapons. Active self-defence has an enormous advantage in that the protection is provided by the person with the greatest interest in avoiding attack - the politician. However, politicians, heads of state and many other public figures must, by definition, lead a routine and predictable life; they have certain regular obligations and duties to which they must attend. Many of them are not trained (or are untrainable) in even the simplest aspects of self-defence. Therefore, if their office warrants it, or they can afford it, they rely on external sources of defence. External active defence is an important, if not the most important, part of protecting VIPs. It is nearly always achieved by hiring bodyguards and other security agents, who must have certain characteristics, the two most important being loyalty and the ability to identify and neutralise potential assailants.

Loyalty can come either from the professionalism of the bodyguards, as is generally the case in Europe and the US, or from family or group relationships, as is usual in many African countries. However, it is not easy to screen bodyguards for loyalty, and loyalties can change. There are many well-documented cases where bodyguards turn from protectors to assailants and attack the person they are meant to protect, such as the assassination in 1984 of Indira Gandhi by two of her bodyguards or, in June last year, when President Boudiaf of Algeria was shot by his own guard.

The ability of the bodyguards to identify and to neutralise potential assailants depends upon training, equipment, general attitudes and numbers. The better they are trained, the greater their chances of success. From watching the television news, we can deduce that most bodyguards have equipment well chosen for their tasks. When in action they bristle with radios and other communications equipment, automatic pistols and other handguns. Ideally, bodyguards should protect the person at all costs. In practice, however, the bodyguards in the West (the professionals as opposed to the family members) would not be expected to sacrifice their lives, but they should be experienced enough to take calculated risks.

Getting the numbers right

Identifying the optimum number of bodyguards to reduce the risk to a minimum is less straightforward. Intuition suggests that highest protection would be provided by a large army of well-trained and well-equipped bodyguards. Experience, however, indicates that this simple argument has two fallacies. First, it assumes that the guards are infallible, and never cause injury to their charges by the accidental discharge of their weapons. Secondly, it assumes that their loyalty is absolute, and that guards never turn against the person they protect.

There is a certain probability that each bodyguard will identify and neutralise a potential assailant, but there is also a certain probability that the bodyguards will harm, intentionally or accidentally, the person they are supposed to protect. Every public figure hopes that the first probability is much higher than the second, but is it? We used our mathematical model to consider this scenario. The risk to the politician is obtained by combining the probabilities of various actions, such as the assailant initiating an attack, the bodyguards neutralising this attack and the bodyguards presenting a direct threat to their charge. In common with other risk assessment analyses, it is difficult to obtain reliable data for these probabilities but we can make assumptions - for example, that the direct threat from the bodyguards is substantially lower than the threat from assailants.

Most of the conclusions are obvious, and it is no surprise to find the model predicting that the better the bodyguards are at recognising and eliminating the threat from the potential assailants the lower the risk. There is, however, one important conclusion that is counterintuitive: the risk of harm to the VIP can be minimised with respect to the number of bodyguards. In other words, there is an optimum number of bodyguards, and employing more or fewer than this increases the risk. If there are fewer than the optimum, the risk will increase because the bodyguards may be unable to deal with assailants; if there are more than the optimum, the risk will also be greater because of the threat, accidental or intentional, from the bodyguards themselves. The optimum depends on the assumptions for the various probabilities, but it could be as low as two or three in stable countries with a popular leader.

What about a more complex strategy such as using two groups of bodyguards? The majority, trained to recognise potential assailants, would be unarmed, or armed only with weapons that could not harm their charge. Only a smaller group, trained to recognise and eliminate potential assailants, would be armed with powerful weapons. The model tells us that the minimum risk can be decreased by such an approach, and it can be further decreased by increasing the number of bodyguards in the unarmed group. We do not know how the security services operate, but observing the 'walkabouts' of US presidents, for example, suggests that this may be their strategy.

Our model is not sufficiently detailed to predict the optimum strategy for dealing with people prepared to blow up whole buildings or roads to target their VIP, as happened last year in the assassinations of Giovanni Falcone and Paolo Borsellino, two of Italy's leading anti-Mafia judges. Nor can it differentiate between the optimum strategies to counteract the threat from, for example, professional assassins or unbalanced individuals. Nevertheless, there are several conclusions which could be important for other risk-assessment analyses.

First, the risk to top politicians cannot be completely eliminated, but it can be managed and minimised. Secondly, there are various defences or safeguards available, and the politician should take advantage of many of these as possible. Thirdly, increased 'protection' does not necessarily mean that the risk will be decreased too. Finally, safety is not necessarily improved by increasing the number of bodyguards.

The model can be applied to considering the overall risk from activities such as large-scale industrial undertakings. The parameters then would be the nature of the activity, the event or events leading to potential disasters and the available safeguards. We would also need to include the probabilities of, on the one hand, the safeguards system being able to identify and control the potential disasters, and on the other, of these safeguards initiating the events which could lead to disasters.

In the case of nuclear reactors, for example, we can draw the following conclusions. First, the risk of accidents cannot be eliminated completely. This is recognised by the nuclear power industry and its regulators, whose aim is to make the risk of accidents as low, and thus the reactors as safe, as is reasonably achievable. Secondly, a variety of protection levels (or safeguards) should be employed to minimise the occurrence of faults and to maximise the chances of ensuring that these faults do not lead to large-scale accidents. For this reason, nuclear reactors are designed with passive and active safeguards, and separate dedicated systems are used to identify and to neutralise possible faults which could lead to large-scale disasters.

Thirdly, increasing the number of safeguards does not necessarily mean that the risk of large-scale accidents will decrease. If a fault occurs, a safeguard must either interact with the system it protects or isolate that part of the system to prevent the fault affecting other parts of the reactor. But a malfunctioning safeguards system can itself lead to accidents if it interacts with other parts of the reactor, while a system that isolates the fault may exacerbate accidents by hindering the operation of other safety systems which need to interact with the part of the reactor affected. Too many safeguards, therefore, could result in an overprotected and less safe nuclear reactor.

The nuclear industry is well aware of this paradox in considering the risk and safety of reactors, but not politicians and the general public. The calls for the absolute safety of, or zero risk from, various activities such as industrial enterprises, air and rail travel or food processing may be politically expedient and popular with the electorate, but are technologically impossible and intellectually bankrupt.

Jiri Kubie used to work for the Central Electricity Generating Board and Nuclear Electric. He is professor of mechanical engineering at Middlesex University, London."

----------

Mick

thanks for the article Mick