Posted: 2001
This page is categorized as:    link to Outrage Management index
Hover here for
Article SummaryThis manual on how to use risk comparisons and risk statistics was commissioned to help chemical plant managers explain air emissions to their neighbors. Chapter III on risk comparisons, especially, is still relevant. Later research hasn’t borne out all its seat-of-the-pants conclusions, but the advice at the end of the chapter about the worst risk comparisons holds firm – in my terms these comparisons fail (especially when people are outraged) because they try to compare the hazard of high-outrage and low-outrage risks. The other chapters are also useful and not really outdated, I think. The appendices are both outdated and all too likely to be misused. They’re what the client originally wanted most. Vincent Covello, Paul Slovic, and I wrote the rest of the manual to soften them.

Risk Communication, Risk Statistics,
and Risk Comparisons:
A Manual for Plant Managers

Washington, DC: Chemical Manufacturers Association, 1988

Part IV
Concrete Examples of Risk Comparisons

The risk comparison examples provided below are based on the following scenario:

Imagine that you are the manager of an ethylene oxide plant in Evanston, a small town in the midwestern United States. The results of a recent risk assessment show that the risk of emissions of ethylene oxide from your plant is 0.007 additional cancers per 3,500 people per year. Your job is to present this information to the community by using various risk comparison techniques.

The purpose of the examples is to illustrate points made in the text. The examples have been organized according to the rank of the comparison.

First-Rank Risk Comparisons (first choice – most acceptable)

a. Comparisons of the same risk at two different times.

  • “Health risks from emissions of ethylene oxide at our plant are 40 percent less than a year ago, when we installed exhaust scrubbers. With more equipment coming in, we expect to reduce the risk another 40 percent by the end of next year.”
  • “Despite the extremely low health risks to the community from emissions of ethylene oxide at our plant, we are still looking for ways to lower these levels further. These are some of the plans we have under way to accomplish this: [provide specifics]. As we implement these steps, we will keep you and the community informed of our progress. We will also continue to monitor our workers and keep track of health statistics within the community to ensure that the risks posed by our plant to our workers and to the community remain in the future as low as, if not lower than, they are today. Since some of you may have further questions about these and other matters concerning our plant operations, as plant manager, I am providing my work and home phone numbers so you can call me. I will do my best to supply you with answers to your questions as quickly as possible.”

b. Comparisons with a standard.

  • “Emissions of ethylene oxide from our plant are half the levels permitted by the U.S. Environmental Protection Agency and by our state’s Department of Environmental Protection.”
  • “Emissions of ethylene oxide from our plant are five times lower than the U.S. Environmental Protection Agency’s safety standard.”
  • “Plant emissions of ethylene oxide are five times below what was permitted under the old EPA standard, and two times below the level established by the new, stricter EPA standard.”

c. Comparisons with different estimates of the same risk.

  • “Laboratory studies on rats and mice suggest that current exposure to ethylene oxide may cause seven cancers in 1,000 generations of residents in this city. This estimate is the maximum that would occur under worst-case conditions. Actual health effects from exposure to ethylene oxide are likely to be lower.”
  • “Let me try to put this number into the context of other numbers. We’ve said that our worst case prediction is seven-thousandths of one extra cancer within the next 70 years from our plant’s emissions of ethylene oxide. Now, no one ever gets seven-thousandths of a cancer. A better way to see the effect is that if 130 different communities the same size as Evanston had a plant just like this one, 129 of those towns would see no effect on their cancer rate. One of the 130 Evanstons might have a single extra cancer.”
  • “Our best estimate of the risk is 0.001 cancers per 3,500 persons using what we believe are realistic assumptions. This estimate is based on work done by our own scientists and by researchers at Evanston University. However, you should be aware that the state Department of Environmental Protection (DEP) has calculated a worst-case risk estimate of 0.007 cancers per 3,500 persons. DEP made the assumption that all individuals living in Evanston would be exposed to emissions of ethylene oxide 24 hours a day for 70 years. This formula gave DEP a human-lifetime dose. DEP then took the best available laboratory information for ethylene oxide – data obtained from studies on the laboratory mice most likely to develop cancer in response to ethylene oxide – and calculated first the lowest dose that caused adverse health effects in mice and then the equivalent dose in humans. On the basis of these and other pieces of information, DEP concluded that the maximum cancer risk to people in the community is 0.007 cancers per 3,500 persons over 70 years.”

    Note: In the example provided above and in several examples provided below, decimals and fractions are used. Since decimals and fractions can sometimes be difficult to understand, consider converting these decimals and fractions to whole numbers, such as seven potential cancer cases per 3,500,000 people, or two cancer cases per million people.

  • “Our worst-case estimate of the risk is seven-thousandths of a cancer per 3,500 persons over the next 70 years. How sure are we that the risk is really this low? The bad news is that we’re not as sure as we’d like to be. Risk assessment is a pretty new science, based on models and assumptions rather than hard data. The good news is that we’re almost certain the risk is actually smaller than our estimate – we’ve instructed our scientists to make every assumption on the cautious side, to provide an extra margin of safety. And here’s a piece of hard information. We’ve been manufacturing ethylene oxide in Evanston for 35 years now. We have continually monitored our employees for signs of adverse health effects associated with exposure to ethylene oxide. In all that time, as far as we know, not a single worker or retiree has had the sort of cancer normally associated with ethylene oxide. Please keep in mind that these workers are exposed to consistently higher levels of emissions than the surrounding population is. Therefore, on the basis of our workers’ experience so far, the risk is zero. There are also people who think our risk estimate is too low. The Evanston chapter of the Sierra Club estimates seven-hundredths of a cancer per 3,500 persons over the next 70 years. That’s 10 times higher than our estimate – but even if they’re right, it’s still an extremely small potential increase in the cancer rate. And we haven’t found anyone with a higher estimate than theirs.”

Second-Rank Risk Comparisons (second choice – less desirable)

a. Comparisons of the risk of doing and not doing something.

  • “If we buy and install the newest and most advanced emission-control equipment available, the worst-case situation is that the maximum total risk will be 0.005 additional cancers per 3,500 persons, a very low number. If we don’t buy new equipment and keep operating the plant with our current pollution-control system, the worst-case situation is that the maximum total risk will be 0.007 additional cancers per 3,500 persons – also a very low number. Please keep in mind that both of these risk estimates are worst-case estimates.”

b. Comparisons of alternative solutions to the same problem.

  • “The maximum health risk from our plant’s emissions of ethylene oxide is 0.007 additional cancers per 3,500 persons. We could switch to producing the only known chemical substitute for ethylene oxide. However, the maximum health risk of emissions of that chemical is 50 times higher.”

c. Comparisons with the same risk as experienced in other places.

  • “We have installed in our plant the most advanced emission control system in the country. Compared with those of older plants, such as the one in Middletown, our emissions are 10 times less.”

    Note: As with any risk comparison, comparisons of “the same risk as experienced in other places” must be handled fairly and carefully. For example, it would clearly be inappropriate to say that the air in Evanston is only 10 percent as polluted as downtown Mexico City.

Third-Rank Risk Comparisons (third choice – even less desirable)

a. Comparisons of average risk with peak risk at a particular time or location.

  • “The risk posed by emissions of ethylene oxide is extremely low, no matter where you live or work in Evanston. However, the risk posed by emissions of ethylene oxide for people living two miles from the plant is 90 percent less than for people living in the nearest home; and the risk for people living in the nearest home is 90 percent less than for people working within the plant gates. And our workers haven’t had a single case of the type of cancer normally thought to be linked to ethylene oxide.”

    Note: It is generally easier to make this type of comparison for exposure data. In many cases, a plant manager can confidently say that one exposure is 90 percent less than another. However, saying the same thing is more difficult in the case of risks, given the multiplicity of factors involved in estimating risks.

b. Comparisons of the risk from one source of a particular adverse effect with the risk from all sources of that same adverse effect.

  • “Let me see whether these numbers will help. Roughly a quarter of all of us get cancer – a disease caused by smoking, diet, heredity, radon in the soil, pollution, and many other factors. Out of 3,500 people, medical data show that one-quarter – or about 875 – are going to get cancer sometime in a lifetime. So here’s the predicted effect of ethylene oxide emissions from our plant on the overall cancer rate. In 129 of 130 hypothetical Evanstons, no effect – that is, no expected increase in cancer rates at all. In the 130th, cancer rates would rise from 875 to 876. Although this is only a tiny increased risk, it is still an increase. If we can find a way to make it even smaller, we should and we will. The most important thing is for all of us in Evanston to work together to find ways to bring down the total cancer rate, that unfortunate 875 out of 3,500. But we at our plant have a special responsibility to be safe neighbors. Much higher risks due to other factors are no reason to ignore a small risk in our facility. Here’s what we’re doing to make sure we keep the risk from our plant as low as it can possibly get… [provide details].”

Fourth-Rank Risk Comparisons (fourth choice – marginally acceptable)

a. Comparisons of risk with cost, or of cost/risk ratio with cost/risk ratio.

  • “During the next year, our plant will spend more than $2 million to reduce our already small emissions even further. This new investment will hurt us economically but will reduce the risk of cancer in the community by more than 25 percent when fully operational.”

b. Comparisons of risk with benefit.

  • “If we stopped producing ethylene oxide today, many more people here and throughout the United States might die than could possibly be affected by emissions from our Evanston plant. Ethylene oxide is the best sterilizing agent used by hospitals today. No equivalent substitute for. ethylene oxide is available. Continued production of this product will contribute to saving many lives and will ensure that the surgical instruments that doctors and hospitals use are free from infectious agents.”

c. Comparisons of occupational with environmental risks.

  • “One way to look at the data is to compare the risks of emissions of ethylene oxide to plant neighbors with the risks to plant employees. We have been operating this plant for 35 years, with an average employment of 400 people. We therefore have about 10,000 person-years of worker exposure to ethylene oxide at this plant. Health monitoring at our plant indicates that the average workplace concentration of ethylene oxide is 0.5 ppm, a dose 200 times higher than that in the community. The primary health concern about ethylene oxide is its potential for causing certain types of brain cancer. We have not had a single case of brain cancer in our work force. Moreover, the overall incidence of cancer in our employees is lower than that of the U.S. population as a whole. Nor has Evanston’s health department documented any brain cancers among our workers. On the basis of this information, I believe that the health risk posed by the plant to the community is insignificant.”

d. Comparisons with other risks from the same source, such as the same facility or the same risk agent.

  • “I believe that our ethylene oxide emissions do not pose a significant health risk to the community. I also believe that our emissions pose a much less serious problem than our hazardous waste problem, which is daily becoming more serious because the repositories in our state are filled and none are being built.”

e. Comparisons with other specific causes of the same disease, illness, or injury.

  • “One way to look at the cancer risk from emissions of ethylene oxide in our community is to compare the risk with the cancer risk from the X-rays you get during a health checkup. One chest X-ray per year presents a risk of developing cancer that is twice that of developing cancer from our plant’s emissions of ethylene oxide.”

Fifth-Rank Risk Comparisons (last choice – rarely acceptable
– use with extreme caution!)

The comparisons provided below are placed in the bottom rank because they disregard distinctions that the public considers important in judging the acceptability of a risk. Unless there is already a high level of trust between you and your audience, this sort of comparison is likely not only to fail but to provoke outrage.

There is seldom if ever a compelling reason to present fifth-rank risk comparisons. But if you think that you must use one, precede the comparison with an acknowledgment of its limitations. For example, you might want to say: “A final way to get some perspective on the risk of ethylene oxide emissions is by comparing it to some of the risks that we all face in our daily lives, such as being struck by lightning and driving. My purpose in making such a comparison is only to put the size of the risk in context. I recognize that such comparisons are like comparing apples and oranges. Still, I think the comparison can help us all understand and gain some perspective on the size of the risk we are talking about….”

If you find this too laborious an introduction, your best strategy may be to find a less irrelevant or more legitimate comparison.

a. Comparisons of unrelated risks. (Warning: Using these comparisons may severely damage your credibility.)

  • “To help gain some perspective on this risk for ethylene oxide, it may be useful to compare it to other risks that we are exposed to as a result of contaminants in our food supply. For example, the risk of death by salmonella poisoning from poultry bought at the local supermarket is at least five times greater than the risk of cancer from the highest exposure to ethylene oxide in this community.”
  • “You may be wondering, ‘But what does that mean to me as a resident of this community? What’s the risk to me and my family?’ First let me tell you that I am convinced that there is no threat to the health or safety of any member of our community at these extremely low exposure levels. However, I recognize that the data still may be troubling. So it would probably be helpful to put these levels of risk from exposure to ethylene oxide into the context of other risks that you’re exposed to in your daily lives. For example, the risk of death from lightning is at least 140 times greater than the risk of cancer from the highest exposure to ethylene oxide. Hurricanes and tornadoes also pose a risk about 140 times greater. Insect bites pose a risk about 70 times greater. The additional 0.007 risk is about the same as the additional risk you would incur spending four hours in Denver rather than at sea level because of Denver’s high altitude and higher radiation level.”

Again, the general rule of thumb in using risk comparisons is: Select from the highest-ranking risk comparisons whenever possible. When you have no choice but to use a low-ranking risk comparison, do so cautiously, being aware that it could well backfire.

Copyright © 1988 by Chemical Manufacturers Association

For more on outrage management:    link to Outrage Management index
      Comment or Ask      Read the comments
Contact information page:    Peter M. Sandman

Website design and management provided by SnowTao Editing Services.