Related Resources:

• Scientific Organizations
  
  
• Research Journals
  
  
• The Virology Bookshop
  
  

Some Thoughts about Security

by Norman M. Bradburn
Assistant Director
Directorate for Social, Behavioral and Economic Sciences
National Science Foundation

While it is not exactly correct to say now that the only thing we have to fear is fear itself, it is the predominate danger in the near term. Making decisions about security precautions in such an atmosphere must take into account the emotional climate, and we must avoid doing things that are not helpful or have long term negative consequence.

We do know several things from the behavioral sciences that are relevant to our current thinking. Perhaps the most important is that people, even in normal times, overestimate the probability of high visibility, concentrated events and underestimate the probability and risks of distributed and chronic, low visibility events. They also tend to put a higher negative value on the massed loss of lives than on the same number of lives lost when they are widely distributed. These tendencies lead to concentration of preventive measures on the large, concentrated, low probability events to the detriment of preventive measures for events of greater probability but of a more continuous or distributed nature.

Another thing we know is that in panic situations people tend to react in ways that produce more casualties than when there is some order maintained--the familiar problem in reaction to fires. The behavioral problem in these situations is quite complex because what is good for the individual may not be good for the group as a whole, that is, by pushing my way rapidly through the crowd I might improve my chances of survival, but if everyone does it, more people will get crushed or not survive. Accurate information about what is happening and knowledge of what to do in such situations minimizes many of these problems.

A third relevant well-established finding is that it is extremely difficult to maintain vigilance for rare events over long periods of time. This was first noticed in WW II when it was found that those monitoring radar for planes often missed them after periods of continuous monitoring. This led to all sorts of interesting research on sensory deprivation and the importance of variation and activity for accurate perception. Even highly paid security people will miss objects or events that are rare when they have to monitor a scanner for long periods of time. Thus depending on individual human inspection or checking of routine events is unlikely to be reliable over the long haul.

What does this all say to us at this time? First, there is a need to be sure that risks are assessed properly. What is more likely to occur and what is less likely to occur? The probability of most buildings actually being bombed has gone up only very slightly, while the probability of a false alarm (bomb threat) has probably gone up more, although both are still low probability events (more likely to be higher in the near term than the longer term, however). Thus attention ought to focus on the more likely events. (A significant fire is probably even more likely than either of the other two, although it is difficult to assess such low probability events so as to make such comparative judgements).

Second, there is a need to concentrate on actions that reduce panic if something does happen. The planned use of a building PA system, a buddy -system, marking stairwells and mapping evacuation routes well, and having a clear, and clearly communicated emergency plan are all responsive to preventing panic and reducing fear. Beyond evacuating the building, there needs to be a similar action plan for what to do then. This involves the larger community, transportation authorities, etc. Emergency planning needs to be coordinated with them.

Third, there is a need to concentrate on everyday security measures that are not overly dependent on human vigilance. It is not clear what technology has to offer here, but there may be ways to enhance control over who is in a building better than the typical, human-dependent systems.

Fourth, it may not be appropriate to invest resources on trying to decrease the already low probability of someone trying to blow up a "science" building. If a terrorist is going to blow up a building, it is more or less random which building gets picked beyond a few obvious choices of high symbolic value. If the emotional climate changes so that “science” is seen as the enemy or of someone’s way of life such that an attack on our building would have symbolic value worth taking the risk of blowing yourself up, then it might have to be rethought. But it would probably be easier and more cost effective to move to another building that were more defensible that it would be to try to “bomb-proof” most "science" buildings.

We should not get swept away with trying to create a fool-proof system to prevent a low probability event and ignore steps that are more effective. We have to accept the fact that we live in a riskier world that we did, and that there are no fool-proof systems. But in any objective sense the risks are not much greater than they have been . It is our perception of those risks that has changed greatly. We should strive not to overestimate them.