The 2,194th Meeting of the Society

September 9, 2005 at 8:00 PM

Powell Auditorium at the Cosmos Club

Life in the Age of Risk Management

Kimberly Thompson

Department of Health Policy and Management, Harvard School of Public Health

About the Lecture

This presentation suggests that we’re now in the Age of Risk Management, a time in human history where we recognize that life is full of risks, choices often involve tough trade-offs, and good data and risk analysis play a critical role in the decisions we make as individuals and collectively. Relying on both insight and humor, this talk emphasizes the importance of understanding systems as a requirement for effective management, and the ability to analytically characterize variability (i.e., real differences that exist between individuals), uncertainty (i.e., the unknown), and dynamics (i.e., changes with time) essential skills. Using a number of real examples ranging from airbags and airplanes to vaccines and video games, Professor Thompson will demonstrate the significant interactions between engineering and behavior, the necessary but not sufficient role of economic analysis in public policy, and the critical role of improved risk communication and education for all stakeholders.

About the Speaker

KIMBERLY THOMPSON is an Associate Professor of Risk Analysis and Decision Science at the Harvard School of Public Health. She holds a M.S. from the Massachusetts Institute of Technology and a Sc.D. from the Harvard School of Public Health. Her research interests and teaching focus on the issues related to developing and applying quantitative methods for risk assessment and risk management, and consideration of the public policy implications associated with including uncertainty and variability in risk characterization. She created and directs the KidsRisk Project, which aims to empower kids, parents, policy makers, and others to make better decisions when managing children’s risks. This work builds on Professor Thompson’s long-standing interest in the issues related to variability in risk for sensitive sub-populations, particularly children, and the potential risk tradeoffs associated with policies designed to protect them. Professor Thompson continues to explore the implications of using different analytical tools for structuring information, a theme that is central to her teaching. Building on her broad training, Professor Thompson also focuses on the characterization of information and communication of risks. She developed a guide to help consumers take charge of health information that appears in her book called Risk in Perspective: Insight and Humor in the Age of Risk Management.


President Robert Hershey called the 2,194th meeting to order at 8:17 pm September 9, 2005.

The minutes of the 2,192nd meeting were read and approved.

Mr. Hershey introduced the speaker of the evening, Ms. Kimberly M. Thompson of the Harvard School of Public Health, where she is director of the Kids Risk Project. Ms. Thompson observed that it was a pleasure to be speaking in the John Wesley Powell room; she attended John Wesley Powell High School in Mesa, Arizona, where her interest in science was kindled.

Ms. Thompson spoke on “Life in the Age of Risk Management.” She posed a question of whether we are living in a state of fear, and quoted several observations of Michael Crichton, who said that “... all reality is media reality,” and that the politico-legal-media complex is dedicated to promoting fear.

Ms. Thompson said we are living in a time when life is full of risks, choices often involve tough trade-offs, and good data and risk analysis play a critical role in decisions, both individual and collective. These are the characteristics of the age of risk management.

At this time, we have the benefits of enormous advancements in science and technology. We have high-quality information about risks, a large spectrum of choices, computational tools, and better understanding of problems and solutions. Success still depends on our ability to understand and communicate how things work and to effectively manage variability and uncertainty.

There is great good news about risks. Since 1900, life expectancy has gone from less than 50 years to over 75 now.

The big questions about risks are, which risks are big and which are small, what can and should we do about them, are we investing the right resources and spending wisely, what data do we need, how do we deal with challenges of scale, and how do we effectively communicate. Communication regarding risks especially needs our attention.

Risk analysis uses mathematical models to characterize information – what can happen, how likely is it, and if it happens, what are the consequences. The models are used to evaluate options and weigh trade-offs – what can we do, what happens if we do it, what is the best option. Finally, we need to communicate risk information.

She distinguished variability from uncertainty. Variability is the degree of heterogeneity or variability in a population. Uncertainty is ignorance about a poorly characterized phenomenon.

She discussed three examples of risk problems – the effectiveness of airbags in cars, the mortality risk to people on the ground from crashing airplanes, and the dynamics of managing the risks of polio.

The air bag question had its roots in the concern about the 40,000 deaths a year in car crashes. There was a time, a short time, when a car would not start if the seat belt was not fastened. People did not like that, and it did not last long. After that, rates of seat belt use were poor, and interest in a passive method developed. This interest led several errors regarding airbags. The fact that airbags would kill people was overlooked. A compliance test was needed to show whether airbags worked, and the standardization involved in the test led to ignoring variability human physique and behavior. Not everyone is a 50th percentile male and not everyone sits still in the seat facing forward all the time, like a dummy. Also, the early experience with seat belts led to pessimism regarding seat belt use, and estimates of the effectiveness of airbags inadvertently included people whose outcomes should have been credited to seat belts.

From this several insights were drawn. Technologies perform differently in controlled versus uncontrolled settings. Efforts should be made to quantify the risks of safety technologies. Distributional issues should be examined. We need to guard against overconfidence and we need to be aware of political realities.

It is an irony that airbags are an engineering solution to a behavioral problem and that the result of children being killed by airbags necessitates a new behavioral solution of putting kids in the back. However, we have learned that both the engineering and the behavior matter.

She turned to the problem of airplanes crashing on people on the ground. A paper in 1992 indicated the risk of this was four in a million, which is above the one in a million threshold used to identify actionable risks. This was, however, the risk of it happening in a lifetime, not in a year. A closer look at the data now reveals some interesting facts. The data, representing 1975 to 1985, are not representative of the present; the rates have gone down substantially. The published rates included people killed on the job on the airport tarmac. Most important, the risk is hundreds of times higher close to major airports. This leads to very different ideas about what should be done about the risk. It also shows that, for most of us, the risk is negligible.

The story about polio is one of a brilliant success of vaccines. Polio peaked in 1952 at 21,000 cases. Major headlines on April 12, 1955 said, “The Vaccine Works,” and “[It is] Safe, Effective and Potent.” The decrease in the disease after that was dramatic, and for some time now we have been within striking distance of eradication.

Oral vaccine, a weakened-virus vaccine, causes polio at a very low rate. In 1979, wild cases of polio virtually disappeared. After that we decided the risk of polio from oral vaccine was unacceptable and we went back to using the killed-virus, injectable vaccine.

Since the oral vaccine spreads through populations on its own and is easier to administer, it is the preferred method if we want to eradicate polio totally. However, the choice is difficult because some people will get polio from the vaccine, and there is still no cure. There is also presumably a greater risk that oral vaccine will mutate or evolve into another, more dangerous strain.

Thus we see how concerns change over time. Sometimes choices get tougher as the original problem is reduced. The devil is in the details. The best options for some people will not be the best for others.

Recently, it has been argued that video games are good for kids – that violence in games does not produce violent behavior because incidence of violence among 15 to 17-year-olds has declined while video game sales have increased. Looking at the curve, we saw a different picture. There was a steep increase in reported violent incidents from the late 1980's to 1994. Then violence declined, but it has not gone below the previous rate and appears to be leveling off there.

In concluding, Ms. Thompson emphasized the need for good science to realize the potential of risk analysis and get the most out of life in the age of risk management.

Mr. Hershey announced the next meeting, invited people to apply for membership in the Society, and invited everyone to enjoy the social hour. He adjourned the 2,194th meeting at 9:54 pm to the social hour.

Attendance: 38
Weather: Clear, mild, beautiful
Temperature: 15°C
Respectfully submitted,

Ronald O. Hietala,
Recording secretary