WE'VE GOT MOT RISK THAN OUR BRAINS CAN HANDLE
30 Nov 2002
Source: Washington Post, June 30, 2002.
We've Got More Risk Than Our Brains Can Handle
By Henry Petroski
Risk is ubiquitous. Each of us is born with a life expectancy, and like it or not we play the odds every day. Numbers can be put on the risks we face from cradle to grave, but it is not possible to play life strictly by the numbers. That is why each of us, in responding to risk, makes personal decisions based not simply on quantitative measures but on qualitative interpretations as well. We party on arsenic-impregnated decks. We play golf in thunderstorms. We build multistory houses despite the risk of falling down stairs. These days -- in a climate of heightened awareness of risk, both real and perceived -- many public policy decisions appear to be made in much the same way. Decisions come ultimately from the gut rather than from the computer.
It is telling that the Department of Homeland Security has introduced a color scale -- a qualitative visual gauge -- instead of a more easily remembered numerical scale for the level of alert. The kinds of risks that terrorists present are virtually impossible to quantify. There is simply too much that we do not know about their timetable, their tactics and their technology. There is also very little historical precedent in our country for the kinds of risks we now face.
Numbers will no doubt in time be attached to a host of new risks, but interpreting those numbers and assigning them colors will require judgment and hard decisions. The possibility of a smallpox attack, for example, requires a decision about whether to vaccinate the entire population when enough vaccine becomes available. Because vaccination itself carries risks, the correct policy is not obvious. Recently, an advisory panel recommended vaccinating critical health care and law enforcement personnel against smallpox but, because of the risk of complications, not the general public. Does the real risk of complications outweigh the potential risk associated with a smallpox attack? This is not an easy question to answer. The risks of vaccinating are limited but fairly certain; the cost of an attack could be immense, but its likelihood is completely uncertain. How can they be measured against each other?
The wildfires that continue to burn in several western states are clear evidence that there are risks associated with living too close to nature. Even if those risks could be quantified, the numbers will never tell us exactly where the next fire will burn or how the wind will be blowing when it does. It is doubtful that those who build and live in houses among the trees base their decision on statistics. They choose to live where they do because of the beauty and peacefulness that a sylvan setting affords. Public policy on how to control forest fires -- or whether this should be done at all -- is by all reports not a debate over numbers. It is a debate over environmental philosophy.
Discussions over whether or not airline pilots should be allowed to carry guns in the cockpit have also revolved more around perceptions of risk than quantifiable measures of risk. This is not surprising, for how can the risks associated with armed pilots be determined? There haven't been guns in cockpits, at least officially, and so there are no solid statistics on what happens when there are. There are no numbers on which to base a quantitative argument. Rather, the debate has involved imagined scenarios of what could happen with pistol-packing pilots in control. No doubt policymakers' opinions on the issue can reflect their views on guns at least as much as their concern for passenger safety.
Public debates about risk and other threatening concepts may be accompanied by a lot of numbers and graphs and budgets, but in the final analysis, decisions are often based on political and emotional grounds, if not on confusing logic. A politician can seem to pay as much attention to the numbers from opinion polls as to the numbers from scientific studies. This happens especially when researchers themselves appear to disagree on the fundamental scientific facts behind issues like global warming and its risks. Scientists can also appear to have an odd perspective on risk. According to the Washington Times, a document uncovered recently in the Environmental Protection Agency's files argued that dumping toxic waste into the Potomac was good for fish because it drove them away from fishing areas. Scientists, like policymakers, are nothing if not human.
Qualitative considerations will always trump quantitative risk assessment analyses in the public policy arena. Even when it may appear that policy decisions are based strictly on the numbers, there is always some overarching qualitative principle defining the choice. Numbers, and the formulas that produce them, are not sufficient for decision making. This fact has been clearly articulated by David Billington in his book "The Innovators," which is about pioneers in engineering who put America on the road to modernity.
Billington noted that engineering formulas are not mere expressions of scientific fact. They also have profound social implications relating to safety and economy. In the case of risk, the numbers alone do not say what is acceptable or unacceptable in the public policy sector. That determination involves a judgment based not on hard numbers but on soft emotions.
The point is illustrated by considering the relative risks associated with highway and air travel. Automobile accidents in the United States claim about 40,000 lives a year, not to mention all the injuries suffered in crashes. Commercial aircraft accidents in this country typically have claimed no more than hundreds of lives a year, two orders of magnitude fewer than highway accidents. The numbers partly reflect the fact that many more people drive than fly. Still, the statistical risks are far higher on the road. Yet many white-knuckled fliers prefer to drive their cars carelessly, risking their own lives and the lives of others on the road. Among the reasons for the irrational fear of flying relative to driving is the sense of individual control felt in a car, a clearly emotional response.
Emotions overwhelm numbers all over the map. It remains inexplicable that less risky airplane travel is subject to strict safety procedures, which pilots and flight attendants review before each takeoff, while children riding on school buses are protected mainly by the yellow color of the vehicle in which they ride.
Predictably, airplane travel is perceived to be even more risky since Sept. 11. But on that day terrorists lost their main weapon, which was surprise. Now, enhanced airport security combined with more secure cockpit doors -- and most likely other measures that have not been publicly revealed -- can be assumed to have made air travel less risky. The next terrorist attack will likely not take place on an airplane, though we second-guess terrorists at our peril, and so continued vigilance at airports and in flight remains absolutely necessary. Letting down our guard is to raise the level of risk.
It is human nature to react the way we have to the terrorist attacks. We all alter our perception of risk after something that was thought to be a remote possibility becomes a clear reality. The 1993 truck-bomb attack on the World Trade Center, though relatively unsuccessful, revealed the true risks involved in allowing unfettered parking beneath the tallest building in New York. Closing the parking garage to the public and instituting tight security procedures reduced the risk of another bomb being carried into the building, but attack by a fuel-laden airplane was not a very bright blip on the risk radar screen. In retrospect, of course, the building was at risk from air attack all along.
Being human, public policymakers are as susceptible as laypersons to minimizing the perception of risks that we all wish did not exist. However, remote risks realized, as they were last September, provide a rude awakening to us all. Everyone can now imagine risks where they were not acknowledged before. The trick, of course, is to see all risks for what they are before the fact. Indeed, they must be seen before the fact if there is to be any attempt to quantify them. Unfortunately, when so many risks formerly seen as incredible become credible, we face the danger of sensory and numerical overload.
In the coming months and years, questions of risk will continue to arise. As with most such questions in the past, the answers will probably depend upon the interpretation of numbers rather than on the numbers themselves. Ordering risks will ultimately come down to political and emotional decisions. It is imperative that, no matter what their math skills, our elected leaders and decision makers are well prepared to deal with those components of risk assessment.
Henry Petroski is the A. S. Vesic Professor of Civil Engineering and a professor of history at Duke University. His latest book is "Paperboy: Confessions of a Future Engineer" (Knopf).