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 The other chapters are also useful and not really outdated, I think. The appendices are both outdated and all too likely to be misused.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

Appendix B
Risk Comparison Tables and Figures

WARNING NOTES

number 1
Since the data in these tables and figures have been calculated or assembled by others, their technical accuracy cannot be guaranteed. CMA makes no representations or warranties concerning their accuracy.
number 2
Some of the data are old and need to be updated. For some risks, this can make a significant difference, either because the risk itself has changed significantly or because its measurement has improved. In general, rates (e.g., number of deaths per million population) tend to change less over time than fatalities (counts).
number 3
It is not always clear what is included in the specific risk entries. For example, do deaths from smoking include cardiovascular disease and emphysema or just lung cancers? What is included in the categories “falling objects” or “toxic gas”? When in doubt, do not use the statistic.
number 4
Most tables of risk comparisons in the literature contain a hodgepodge of risks characterized by different levels of uncertainty. For risks such as driving, where fatalities can be counted, the number is likely to be reliable. But for risks such as radiation or food additives – based not on counting and actuarial statistics but on theoretical modeling and extrapolation – the number is likely to be highly uncertain. In general, data based on theoretical models and extrapolation are more likely to be a target for debate and criticism than data based on counts and actuarial statistics.
number 5
Most risk comparison tables offer only single number risk estimates, with no range or error term.
number 6
Most tables of risk comparisons in the literature have been developed to make a particular point. No matter how often the table has been reprinted, it is important to be sensitive to biases in the calculation of risks.
number 7
A careful risk comparison requires a good deal of background information about data sources, assumptions, and other qualifiers. Just because a risk statistic is published does not mean it is reliable. Published data often take on a life of their own. The original publication may discuss a number of qualifying uncertainties but these may be left out by the next person who reproduces the statistic. This may make the data look more certain, but in fact it makes them far less reliable. Keep in mind, however, that even if a risk estimate in a table is slightly off, it still could prove useful for comparisons in which risks differ by factors of 10, 100 or more.
number 8
The tables are often neither clear nor consistent about the population used to calculate the risk. Within the same table, some risk estimates are based on the entire population (e.g., the United States), while others are based only on the population that is exposed (e.g., only people who hunt or live in tornado-prone regions). Every risk looks more risky if only the most exposed population is considered, less risky if lots of unexposed people are considered.
number 9
Even if the risk comparison data are carefully and accurately reported, they can be misleading. For example, the risk calculation for driving includes many different driving situations. Yet speeding home from a party just before dawn is two orders of magnitude more dangerous than driving to the supermarket. Similarly, the risk of being hit by lightning for people who remain on a golf course during a thunderstorm is much higher than the risk for the U.S. population provided in these tables.
number 10
Risk comparisons raise all the same framing issues as risk quantification generally. That is, there are many different ways to express risk comparison data. Each of these expressions is likely to have a somewhat different impact on the audience.
number 11
The primary intent of these cautionary statements is to warn against casual acceptance of data in comparison tables, to emphasize the importance of acting fairly and responsibly in constructing comparisons, and to indicate the advisability of having someone carefully cross-check the comparison data. Whenever possible, avoid the use of secondary data sources. Track down the original source of the statistic and, if it seems accurate and appropriate, use that number.
number 12
Remember that a useful risk comparison must be accurate and appropriate. Comparing chemical plant risks to voluntary lifestyle choices such as smoking or driving without a seatbelt is seldom appropriate or successful, even if the comparison is technically accurate.

Table B.1
Annual Risk of Death in the United States

CauseRisk Per
Million
Persons
Motor vehicle accidents (total)240.0
Home accidents 110.0
Falls 62.0
Motor vehicle pedestrian collisions 42.0
Drowning 36.0
Fires 28.0
Inhalation and ingestion of objects15.0
Firearms 10.0
Accidental poisoning by gases and vapors 7.7
Accidental poisoning by solids and liquids (not drugs or medicaments) 6.0
Electrocution 5.3
Tornadoes 0.6
Floods 0.6
Lightning 0.5
Tropical cyclones and hurricanes 0.3
Bites and stings by venomous animals and insects0.2

Source: Adapted from Wilson, R. and Crouch, E., Risk/Benefit Analysis, Cambridge: Ballinger, 1982.

Warning!

Use of data in this table for risk comparison purposes can damage your credibility (see text).

Illustrative Verbal Interpretation:

Every year approximately 60 persons per million die from falls in the United States. In a city of 100,000 persons, we could expect approximately 6 persons to die from falls annually. In the United States as a whole, we could expect approximately 15,000 deaths from falls per year.

Table B.2
Annual Risk of Death in the United States

HazardTotal Number
of Deaths
Risk Per
Million Persons
All causes 1,973,003 9000.0
Heart Disease 757,075 3400.0
Cancer 351,055 1600.0
Motor vehicle accidents 46,200 210.0
Work Accidents 33,400 150.0
Homicides 20,465 93.0
Falls 16,300 74.0
Drowning 8,100 37.0
Fires, burns 6,500 30.0
Poisoning by solids or liquids 3,800 17.0
Suffocation, ingested objects 2,900 13.0
Firearms, sporting 2,400 11.0
Railroads 1,989 9.0
Civil aviation 1,757 8.0
Water transport 1,725 7.0
Poisoning by gases 1,700 7.0
Pleasure boating 1,446 6.0
Lightning 124 0.5
Hurricanes 93 0.4
Tornadoes 91 0.4
Bites and Stings 48 0.2

Source: Adapted from Atallah, S., “Assessing and Managing Industrial Risk,” Chemical Engineering, September 8, 1980: 99–103.

2002 Note: Several numbers were incorrect in the original CMA publication and have been corrected here.

Warning!

Use of data in this table for risk comparison purposes can damage your credibility (see text).

Illustrative Verbal Interpretation:

Every year approximately 1,500 persons die in pleasure boating accidents in the United States. This represents six deaths per million persons. Of course, since everyone in the United States is not exposed to this risk, the rate per million boaters would be higher.

Table B.3
Risk Comparisons (Involuntary Risks Only)

RiskRisk of Death/
Person/Year
Influenza 1 in 5000
Leukemia 1 in 12,500
Struck by an automobile (United Kingdom)1 in 16,600
Struck by an automobile (United States)1 in 20,000
Floods (United States)1 in 455,000
Tornadoes (Midwest United States)1 in 455,000
Earthquakes (California)1 in 588,000
Bites of venomous creatures (United Kingdom)1 in 5 million
Lightning (United Kingdom)1 in 10 million
Falling aircraft (United States)1 in 10 million
Release from nuclear power plant  
    At site boundary (United States)1 in 10 million
    At one kilometer (United Kingdom)1 in 10 million
Flooding of dike (the Netherlands)1 in 10 million
Explosion, pressure vehicle (United States)1 in 20 million
Falling aircraft (United Kingdom)1 in 50 million
Meteorite 1 in 100 billion

Adapted from Dinman, B.D., “The Reality and Acceptance of Risk,” Journal of the American Medical Association, Vol. 244 (11): 1126–1128, 1980.

Warning!

Use of data in this table for risk comparison purposes can damage your credibility (see text).

Illustrative Verbal Interpretation:

The risk of tornadoes in the tornado-prone midwestern United States is 1 death per 455,000 persons per year, or about 2.2 deaths per million persons. This is much greater than the risk across the U.S. as a whole (.4 deaths per million persons per year – See Table B.2).

Table B.4
Estimated Loss of Life Expectancy
Due to Various Causes

CauseDays  CauseDays 
Cigarette smoking (male)2250Job with radiation exposure40
Heart disease2100Falls39
Being 30% overweight1300Accidents to Pedestrians37
Being a coal miner1100Safest job (accidents)30
Cancer980Fire (burns)27
Being 20% Overweight900Generation of energy24
Cigarette smoking (female)800Illicit drugs (U.S. average)18
Stroke520Poison (solid, liquid)17
Living in unfavorable state500Suffocation13
Cigar smoking330Firearms accidents11
Dangerous job (accidents)300Natural radiation  8
Pipe smoking220Poisonous gases  7
Increasing food intake 100 calories/day210Medical X rays  6
Motor vehicle accidents207Coffee  6
Pneumonia (influenza)141Oral contraceptives  5
Alcohol (U.S. average)130Accidents to bicycles  5
Accidents in home  95All catastrophes combined  3.5
Suicide  95Diet drinks  2
Diabetes  95Reactor accidents (UCS)  2*
Being murdered (homicide)  90Reactor accidents (NRC)  0.02*
Legal drug misuse  90PAP test-4
Average job (accidents)  74Smoke alarm in home-10
Drowning  41Air bags in car-50
Mobile coronary care units-125

Source: Adapted from Cohen, B. and Lee, I. “A Catalog of Risks.” Health Physics, 36, June, 1979, 707–722.

Notes: (*) These items assume that all U.S. power is nuclear. UCS stands for the Union of Concerned Scientists, a leading critic of nuclear power. NRC stands for the U.S. Nuclear Regulatory Commission.

Warning!

Use of data in this table for risk comparison purposes can damage your credibility (see text).

Illustrative Verbal Interpretation:

The average coal miner in the United States lives three years less than the national average. Although not indicated in the table, this is presumably due to the increased risk of accidents and disease (e.g., black and brown lung disease). However, other characteristics of coal miners, such as their smoking habits, diet, and access to medical care, may also affect this statistic.

Table B.5
Risks Estimated to Increase the
Probability of Death in Any Year by
One Chance in a Million

ActivityCause of Death
Smoking 1.4 cigarettes cancer, heart disease
Drinking .5 liter of wine cirrhosis of the liver
Spending 1 hour in a coal mine black lung disease
Spending 3 hours in a coal mine accident
Living 2 days in New York or Boston air pollution
Traveling 6 minutes by canoe accident
Traveling 10 miles by bicycle accident
Traveling 300 miles by car accident
Flying 1,000 miles by jet accident
Flying 6,000 miles by jet cancer caused by cosmic radiation
Living 2 months in Denver cancer caused by cosmic radiation
Living 2 months in average stone or brick
building
cancer caused by natural radioactivity
One chest X ray taken in a good hospital cancer caused by radiation
Living 2 months with a cigarette smoker cancer, heart disease
Eating 40 tablespoons of peanut butter liver cancer caused by aflatoxin B
Drinking Miami drinking water for 1 year cancer caused by chloroform
Drinking 30 12 oz cans of diet soda cancer caused by saccharin
Living 5 years at site boundary of a typical
nuclear power plant
cancer caused by radiation
Drinking 1,000 24-oz soft drinks from plastic bottles cancer from acrylonitrile monomer
Living 20 years near a polyvinyl chloride plant cancer caused by vinyl chloride
(1976 standard)
Living 150 years within 20 miles of a nuclear
power plant
cancer caused by radiation
Living 50 years within 5 miles of a nuclear
power plant
cancer caused by radiation
Eating 100 charcoal-broiled steaks cancer from benzopyrene

Source: Adapted from Wilson, R., “Analyzing the Daily Risks of Life.” Technology Review, 81, 1979, pp. 40–46.

Note: These data are based on simple extrapolations from population averages. Some data are based on actuarial statistics (e.g., coal mine accidents) and others are based on theoretical models (e.g., cancers from chlorinated water).

Warning!

Use of data in this table for risk comparison purposes can damage your credibility (see text).

Table B.6
Average Risk of Death
to an Individual from Various
Natural and Human-Caused Accidents

Accident
Type
Total  
Number
Individual Chance
Per Year
Motor Vehicle 55,7911 in 4,000
Falls 17,827 1 in 10,000
Fires and Hot Substances 7,4511 in 25,000
Drowning 6,1811 in 30,000
Firearms 2,3091 in 100,000
Air Travel 1,7781 in 100,000
Falling Objects 1,2711 in 160,000
Electrocution 1,1481 in 160,000
Lightning 1601 in 2,000,000
Tornadoes 911 in 2,500,000
Hurricanes 931 in 2,500,000
All Accidents111,9921 in 1,600

Source: Nuclear Regulatory Commission, Reactor Safety Study, WASH–1400 (NUREG/74/104), Washington, D.C., 1975.

Warning!

Use of data in this table for risk comparison purposes can damage your credibility (see text).

Table B.7
Average Risk of Death from Various Human-Caused and Natural Accidents

Type of EventProbability
of 100 or
More Fatalities
Probability
of 1,000 or
More Fatalities
Human-caused
Airplane Crash 1 in 2 yrs.1 in 2,000 yrs.
Fire1 in 7 yrs.1 in 200 yrs.
Explosion1 in 16 yrs.1 in 120 yrs.
Toxic Gas1 in 100 yrs.1 in 1,000 yrs.
Natural
Tornado1 in 5 yrs.very small
Hurricane1 in 5 yrs.1 in 25 yrs.
Earthquake1 in 20 yrs.1 in 50 yrs.
Meteorite Impact1 in 100,000 yrs.1 in 1 million yrs.

Source: Nuclear Regulatory Commission, Reactor Safety Study, WASH–1400 (NUREG/74/104), Washington, D.C., 1975.

Warning!

Use of data in this table for risk comparison purposes can damage your credibility (see text).

Table B.8
Ranking of Possible Cancer Risks
from Common Substances

RankingRisk Source
0.2   PCBs (daily dietary intake): exposure through industrial residues
0.3   DDE/DDT (daily dietary intake): exposure through pesiticide residues; DDE is a by-product of DDT
1      Tap water (1 liter a day): contains chloroform, a by-product of chlorination
3     Cooked bacon (100 g/about 15 slices a day): contains dimethylnitrosamine, a preservative by-product
4      Contaminated well water (1 liter a day): from worst well in Silicon Valley; contains trichloroethylene
4      EDB (daily dietary intake): exposure through pesticide and other residues in grains and grain products
8      Swimming pool (1 hour a day for a child): exposure to chloroform by swallowing chlorinated water
30      Peanut butter (32 g/2 tablespoons a day): contains aflatoxin, a natural mold
30      Comfrey herb tea (1 cup a day): contains symphytine, a natural pesticide
60      Diet cola (12 ounces a day): contains saccharin
100      Raw mushroom (1 a day): contains hydrazines, natural pesticides
100      Dried basil (1 g of dried leaf): contains estragole, a natural pesticide
300      Phenacetin pill (average dose): ingredient in pain reliever
600      Indoor air (homes) (14 hours a day): formaldehyde emitted from furniture, carpets, and wall coverings
2,800      Beer (12 ounces a day): contains ethyl alcohol
4,700      Wine (250 ml/8 ounces a day): contains ethyl alchohol
5,800      Formaldehyde (6.1 mg/worker’s average daily intake): exposure through inhalation
16,000      Phenobarbitol (1 pill a day): a sleeping pill
140,000      EDB (150 mg/worker’s daily intake at high exposure): exposure through inhalation; worker’s maximum legal exposure

Source: Adapted from Ames, B.N., Magaw, R., and Gold, L.S., “Ranking Possible Carcinogenic Hazards,” Science, 1987, Vol. 236, (17 April 1987), 27 1–285; and J. Tierney (1988), “Not to Worry…,” Hippocrates, January/February 1988, pp. 29–38.

Warning!

Use of data in this table for risk comparison purposes can damage your credibility (see text).

Illustrative Verbal Interpretation.

The risk from industrial formaldehyde is 5,800 times greater than the risk from tap water.

Figure B.1
Health Risk Ladder – Annual Number
of Deaths per Million People

Source: Adapted from Schultz, W., G. McClelland, B. Hurd, and J. Smith (1986), Improving Accuracy and Reducing Costs of Environmental Benefits Assessment, Vol. IV. Boulder: University of Colorado, Center for Economic Analysis.

Warning!

Use of data in this table for risk comparison purposes can damage your credibility (see text).

Figure B.2
Upper Bound Estimates of Deaths
for Different Energy Systems

(Image not available)

Source: Adapted from Inhaber, H. (1979), “Risks with energy from conventional and non-conventional sources.” Science, 203, 1979, 718–723. Also Inhaber, H., Risk of Energy Production, Report No. AECB 119/rev. 3, 4th edition. Ottawa: Atomic Energy Control Board, 1979.

Figure B.3
Comparisons of Different Sources
of Radiation Exposure

Source: Adapted from the National Radiological Protection Board (1986), Living with Radiation, London: HMSO.

Warning!

Use of data in this table for risk comparison purposes can damage your credibility (see text).

Figure B.4
The Causes of Cancer:
Quantitative Estimates of the
Avoidable Risk of Cancer in the U.S.

Note: Due to rounding error, percentage figures do not add up to one hundred percent.

Source: Adapted from Doll, R. and Peto, R. (1981), “The Causes of Cancer: Quantitative Estimates of the Avoidable Risk of Cancer in the U.S. Today.” Journal of the National Cancer Institute, 1981, Vol. 66, 1191–1308.

Warning!

Use of data in this table for risk comparison purposes can damage your credibility (see text).

Figure B.5
Radon Risk Charts

(Image not available)

Source: Adapted from Smith, V.K., W.D. Desvousges, and A. Fisher (1987), Communicating Radon Risk Effectively: A Mid-Course Evaluation, Report No. CR–811075. Washington, D.C.: U.S. Environmental Protection Agency, Office of Policy Analysis.

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.