Risk Analysis

Dale H. Easley


The (small) chance of dying in an aircrash is one of the prices that society agrees to pay for rapid convenient global transportation. Some risks, including nuclear power generation, have caused fewer deaths but provoked greater calls for regulation, whereas others, such as automobiles, cause more deaths but arouse less concern. 1

Each year in my environmental geology class, I ask students how many buy bottled water. (In New Orleans, there is a lot of concern over the water supply because public supplies come from the Mississippi River.) Most hands go up. I then ask them how many have driven 70mph or more on the expressway in the last two weeks. Again, most of the hands go up. That they initially fail to see any relation between the two questions demonstrates how selective we humans can be in choosing which risks we attempt to reduce. And in the U.S., we seem particularly torn between wanting the freedom to take risks while wanting the government to shield us from risks.

At the federal level, we have created a number of government agencies to deal with perceived hazards to our society. The Department of Defense is the big spender---historically, warfare has been an obvious health risk. For domestic risks, we have created OSHA to oversee on-the-job safety, the USDA for food safety, the EPA for pollution, the CDC for epidemics and public health, etc. Each of these agencies makes regulations based on risk, primarily using one of three main approaches, utility, rights, or technology:

  • The utility-based approach for risk reduction typically uses cost-benefit analysis. Regulations are aimed at maximizing the tradeoff between benefit to society and cost due to imposition of the regulations. For example, during the Carter administration speed limits on interstate highways were reduced in order to reduce consumption of gasoline. A drop of thousands of annual highway fatalities resulted. Speed limits were later raised because the cost to interstate commerce of the slow speeds was believed to outweigh the potential increased loss of life.2
  • The rights-based approach incorporates the non-economic principle of justice. For example, exposure to pollution should not simply be an economic decision but should protect the young, old, and weak. This idea was also applied to non-humans through the Endangered Species Act.
  • The technology-based approach considers what is possible technologically at the time of enforcement. For example, this approach may require that best-management-practices be adopted. A polluter might be required to reduce admissions to the level the current technology makes practical. As technology changes, the regulatory level may be reconsidered.

On the personal level, however, people are less likely to look at the statistics and costs of hazards when ranking risks. Researchers 3 have found three main reasons why we rank risks as we do:

  • dreadfulness: how awful it seems to people, such as a nuclear bomb planted by terrorists or kids getting cancer,
  • level of understanding: how well people understand the hazard, such as smoking, where increased education has lowered smoking rates, and
  • number of people exposed: who might be affected at a given time. For example, a plane crash kills more than a car crash on a per crash basis, even though car crashes kill far more in any given year.
So, to continue with our automobile example, a car crash is not particularly dreadful, perhaps because it is so familiar, we understand well how it happens, and it only affects a few people at any given time. Therefore, we tend, at least on an emotional level, to rank it fairly low as a hazard of concern, even though statistically car crashes are major killers.

At a government level, we've addressed the risk from car crashes through safer roads, mandatory car seats for kids, and seat belts for adults. But in Louisiana we've done away with helmet laws for motorcyclists. (I only ask that for the good of society, if the motorcyclist leaves his helmet at home that he first fill out an organ-donor card.) This apparent contradiction in risk policy is an example of the balancing act needed to create regulations that take into account the expert opinion based on statistical data while addressing the concerns and desires for freedom of citizens. Statistical evidence shows that riding a motorcycle is one of the most dangerous activities available. Yet seldom is anyone hurt other than the rider himself. In my opinion, as long as there is sufficient education about the risks for you to make an informed choice, the risk to others of your behavior would be the major factor in defining whether or not the behavior should be regulated.


1This essay is based on Risk Analysis and Management, by M.~Granger Morgan, Scientific American, July 1993, p 32-41, the source of the opening quotation.

2Due to safer cars, seatbelt requirements, better road safety, etc., the highway fatality rates did not rise back as high as their previous rates.

3Experimental psychologists Baruch Fischhoff of Carnegie Mellon University, and Paul Slovic and Sarah Lichenstein of Decision Research in Eugene, OR.

File translated from TEX by T TH, version 2.86.
On 19 Jan 2001, 13:38.