The serious nature of threats can be rationally assessed. It depends on their likelihood and the danger they pose to life and property. I assume that when designing and implementing policies to reduce the threat of terrorism, governments face a trade-off between reduction of the threat and the costs of such actions. Costs encompass monetary (government and private sector) and less visible sacrifices, such as delays and reductions in civil liberties.
Since perfect security from terrorism is impossible without infinite resources, governments ought to choose the most cost-effective solution that reduces the threat to an acceptable level. This is the efficient, ‘rational’ response, with which we can compare actual policy. That governments aim to work according to similar principles is implied by the US National Security Strategy statement (in relation to preventive war): "The greater the threat, the greater the risk of inaction"1.
The causes of misperception
Four potential reasons for the misperception of the terrorist threat can be identified in the literature. First, in the face of ignorance, people tend to assess probabilities by asking whether a readily available example comes to mind (the ‘availability heuristic’)2. In the aftermath of a terrorist attack, therefore, people are likely to overestimate the likelihood of another attack and so expend greater efforts to reduce the threat.
Second, it has long been observed that people have difficulty in determining the appropriate response to any low-probability risk, even if they are made aware of the probability. Psychologists Kahneman and Tversky showed that individuals consistently tend towards low-probability outcomes when calculating their response to a risk.3 People appear to have an in-built tendency to overreact to low-probability threats (such as terrorism) by devoting more resources to countering them than their probability warrants.
Third, when determining the appropriate level of response to a risk, people have been shown to disregard probability to a large extent if the possible outcome is particularly emotive.4 When strong emotions are involved (as is the case with terrorism because of the vividness of the outcome and the associated levels of outrage and fear), the resources that people are willing to devote to avoiding the threat do not vary significantly with its probability. People therefore have a tendency to demand very large efforts to reduce highly emotive terrorist threats, be they likely or unlikely.5
Fourth, people show a disproportionate fear of (and thus demand a disproportionate response to) threats that have certain characteristics.6 These include being new or unfamiliar; catastrophic rather than chronic (for example, resulting in many deaths at once rather than spread over time); or indiscriminate and involuntary.7
The 1987 King’s Cross fire in London, in which 31 people died, displays all of these so-called ‘dread’ characteristics. The report of the inquiry into the fire questioned the £300 million (US$553.2m) spent in response to the disaster on equipment such as fireproof doors and metal escalators. It said the money might better have been spent putting smoke detectors into every house in the country to reduce the danger of house fires, which kill approximately 500 people a year in the UK.8
For some threats, therefore, people are likely to expect - and perhaps demand - a government response different from what would be the ‘rational’ course of action. In addition, decision-makers in government might themselves be susceptible to the same misperceptions of risk as the general public.
The threat of terrorism in US after 11 September 2001 would appear particularly susceptible to overperception of risk because: it has a low probability; there is a recent and vivid example in people’s minds; it elicits high levels of emotion; it is new; and its effects are potentially catastrophic and indiscriminate.
If the allocation of resources in the ‘war on terrorism’ resulted from misperceptions of the terrorist threat to the US, one would expect to see certain patterns. Overall, one would expect the resources spent on protection against terrorism to be too great relative to the risk posed to life and property. Within the spending one would expect to see a bias in resource allocation towards threats that have a recent and vivid example and away from equally dangerous and known threats that are not in the public consciousness.
If it is the public and not the government that is affected by threat misperception, the government’s response will aim to reassure as well as protect the public. Thus resources might be directed to highly visible activities in favour of ‘behind the scenes’ work, or might prioritise ease of response over efficacy. Any of these responses to the terrorist threat to the US could be considered a misallocation of resources because they do not represent a ‘rational’, efficient response.9
The US government’s response to the terrorist threat has two components. "[W]hile the National Strategy for Homeland Security focuses on preventing terrorist attacks within the United States, the National Strategy for Combating Terrorism [also] focuses on identifying and defusing threats before they reach [the US] borders," says The National Strategy for Combating Terrorism.10 Within homeland security there are a number of key examples which indicate that US counter-terrorism policy has been affected by threat misperception.
First, the US response to the terrorist threat after 11 September 2001 has been much more extensive than the UK’s. This cannot wholly be explained by the size of the UK or the fact that it already had a higher level of anti-terrorist preparedness.
Furthermore, the risk of terrorism facing the US and UK must, especially since Operation ‘Iraqi Freedom’ in 2003, be considered comparable. Americans seem to be much more worried about terrorism than Britons.11 This must be partly explained by the fact that the threat of terrorism in US is new, whereas the UK has been living intermittently with Irish terrorism since the 19th century.
The UK - more accustomed to terror
New threats cause people to respond more than familiar threats. Possibly because, unlike the UK, the US mainland had never been attacked before 11 September 2001, terrorism also elicits a higher level of emotion in Americans than in Britons. A higher emotive response to terrorism also makes Americans more likely to suffer from misperception of the threat. Thus the US overreaction to the threat of terrorism in comparison with the UK can be explained by the former’s misperception of the threat.
Second, US terrorism prevention policies have given high priority to aviation security as a component of border and transportation security, which is one of the National Strategy for Homeland Security’s six "critical mission areas". Costs have included legislation requiring private spending (the Aviation and Transportation Security Act of 2001), direct federal spending and delays incurred by passengers. This is in part because the 11 September attacks brought to light serious aviation security weaknesses (most particularly in internal flights) that demanded attention.
However, it can be argued that concentration on aviation security has left other areas neglected. The Progressive Policy Institute’s generally highly critical ‘Homeland Security Report Card’ graded aviation security "above average" - presumably a great improvement on the pre-11 September situation - indicating a concentration of efforts in this direction. This is in contrast to, for example, chemical production and storage facilities, the contents of which have the potential to cause at least as much harm as aircraft and whose security was graded "below average" by the report. Such installations are not subject to any proposed government legislation.12
The fact that the most vivid and recent terrorist attack on the US was executed by hijacking and that aviation security has attracted more resources than other threats suggests that US policy makers might be subject to risk misperception.
Amy Leyland is an analyst with the Terrorism Intelligence Centre in Canberra, Australia
1 The National Security Strategy of the United States of America, September 2002, www.whitehouse.gov/nsc.
2 A Tversky and D Kahneman, ‘Judgement and Uncertainty: Heuristics and Biases’, in Science No.185 p1131 (1974).
3 D Kahneman and A Tversky, ‘Prospect Theory: an Analysis of Decision Under Risk’, in Econometrica No.47 p125 (1979).
4 For example, consider the following thought experiment: How much would you be prepared to pay to avoid a 1%/90% chance of (a) a $20 fine (an emotion-free outcome) or (b) a ‘short, painful, but not dangerous electric shock’ (a strong emotion outcome)? Rottenstreich and Hsee found that the median amount individuals were willing to pay rose from $1 for a 1% chance of a $20 penalty to $18 to avoid a 99% chance. However, for the electric shock, the difference was much smaller: $7 to avoid a 1% chance and $10 to avoid a 99% chance. This shows that when strong emotions are involved, people are willing to pay almost as much to avoid low-probability risks as they are to avoid higher-probability risks.
5 Cass R Sunstein, ‘Terrorism and Probability Neglect’, in The Journal of Risk and Uncertainty Vol.26 No.2/3, 2003, p122.
6 Paul Slovic and Elke U Weber, ‘Perception of Risk Posed by Extreme Events’, p4 in a paper prepared for discussion at the conference ‘Risk Management Strategies in an Uncertain World’, Palisades, New York, 12-13 April 2002
7 People show a much greater fear of radiation from buried nuclear waste than of flying, for example, even if the threat to life is the same. (ibid.)
8 ‘Something must (not) be done’, The Economist 11 September 2003.
9 It is important to note that it is impossible to reach an objective assessment of the extent to which US anti-terrorism policy has been affected by risk misperception, since to do so would require determination of what would be a truly ‘rational’ response to the threat. We have less access than governments to probabilities of different threats or which policies are most likely to reduce threats, and it is even difficult to get an estimate of the costs of policies since they are often spread between the government, the private sector and the general public. However, certain crude methods are available that allow us to assess whether the US response is likely to have been significantly affected. These include: comparing the US response with other countries’; comparing the implementation of policy with that set out in the National Strategies (which assumes that a National Strategy is less liable to misperception of risk); investigating areas of policy that are criticised by experts (assuming that the ‘experts’ are somehow less subjective); and comparing resources devoted to different areas.
10 The National Strategy for Combating Terrorism, Feb 2003, http://usinfo.state.gov/topical/pol/terror/strategy/
11 ‘Perceptions of risk of terrorist attacks’, in Program on International Policy Attitudes, http://www.americans-world.org/digest/global_issues/terrorism/terrorism_perception.cfm.
12 ‘America at Risk: A Homeland Security Report Card’, July 2003, The Progressive Policy Institute, Washington, DC, http://www.ppionline.org/ p 19.