When EPA was established in 1970, there was a clear expectationthat removing chemicals from the air, water, and soil would reducecancer rates. Now, almost three decades later, scientists arealmost uniform in their opinion that chemicals in the environmentare associated with only a tiny proportion of cancer. Moreover,there is no evidence that EPA's efforts have had any effect oncancer rates.
The great hopes for reducing cancer by reducing environmentalexposures to chemicals have not been realized because they werebased on a wildly incorrect notion about the importance ofenvironmental exposures in causing cancer. The notion arose from amisunderstanding of the word "environment" coupled with clearincentives to perpetuate and publicize the misunderstanding.
John Higginson, the first director of the World HealthOrganization's International Agency for Research on Cancer, iscredited with coining the expression that the environment causesupwards of 90 percent of all cancers As he used the word,"environment" included everything with which people comein contact, "Environment is what surrounds people--and impingesupon them.... the air you breathe, the culture you live in...thechemicals with which you come in contact."  He explained that many people misunderstoodthe word "environment" to mean chemicals, and he underscored thestrong incentive for some people and organizations to have itmisunderstood in that way.
A lot of confusion has arisen...because most people...have usedthe word `environment' purely to mean chemicals....
The ecological movement, I suspect, found the extreme viewconvenient because of the fear of cancer. If they could possiblymake people believe that cancer is going to result from pollutionthat would facilitate the cleaning up of the water, the air,whatever it was... People would love to be able to prove thatcancer is due to the general environment or pollution. It would beso easy to say `let us regulate everything to zero exposure and wehave no more cancer.' The concept is so beautiful that it willoverwhelm a mass of facts to the contrary. 
Higginson was right. The idea that environmental pollutants arethe cause of much of human cancer is very attractive. In the firstplace, it explains the inexplicable. Although 25 percent of peoplein the United States will develop cancer at some time during theirlives, and about 20 percent of all deaths are caused by cancer, thecauses of most cancers remain unknown. Being told that the causeslurk in environmental chemicals provides an explanation for theoccurrence of cancer. Secondly, and even more satisfyingly, ifchemical causes can be identified and eliminated, cancer ratesshould fall.
The "Environment" and Cancer
The evidence that the environment is a significant cause ofcancer was never strong, but scientific reviews of informationabout causes of cancer had little influence on public policy until1981. In that year, Sir Richard Doll and Richard Peto publishedtheir encyclopedic analysis of the causes of cancer, whichinvestigated the evidence for the widely discussed "cancerepidemic" and catalogued the causes of cancer in the United States.
Doll and Peto documented that cancer deaths were increasingamong elderly people who, because of reductions in deaths frominfectious diseases, heart diseases, and accidents, lived to theadvanced ages where cancer is common. Even so, the percentages of elderly people who died from cancerwere not increasing. When allowances were made for the increasednumber of cancer deaths expected in an aging population and forincreases in lung cancers, the age-adjusted death rate from allcancers had not changed over the four decades from 1933, whencollection of national data began, through the 1970s, the latestyears for which data were available when Doll and Peto did theirresearch. There was no evidence for a surge in overall cancerrates.
During the four decades for which Doll and Peto had data,mortality from some cancers--especially pulmonary cancers--hadincreased, and mortality from other cancers--especially stomachcancers--had decreased. Increased rates of pulmonary cancersfollowed increases in smoking; reduced stomach cancer ratesaccompanied changes in food preservation that reduced the amountsof ingested natural toxic materials and of meats preserved bysmoking and salting. Age-adjusted mortality rates from cancers atother body sites remained nearly constant from the 30s through themid-70s.
More recently, Susan S. Devesa and her colleagues at theNational Cancer Institute compared cancer rates in the UnitedStates in the 1975 through 1979 period to rates in 1987 through1991. Cancer incidence rates increased 19 percent in men and 12percent in women, with almost all the increases being accounted forby higher prostate cancer rates in men and higher breast and lungcancer rates in women. Devesa et al. concluded "Improveddetection appears to account for most of the increases in breastcancer among women and prostate cancer among men. On the otherhand, cigarette smoking is the major determinant of the rise inlung cancer among women." 
Deaths from cancer increased less, three percent and six percentamong men and women, respectively, between the late `70s and `80s.Those increases were "driven mostly by continuing increases in lungcancer mortality , while death rates for the majority of cancerswere steady or declining."  In fact,clear decreases in cancer mortality were apparent among both menand women under 55.
One conclusion from Devesa et al. is especiallyimportant to any discussion of environmental causes of cancer:
Increasing exposure to general environmental hazards seemsunlikely to have had a major impact on the overall trends in cancerrates, in agreement with the conclusion reached in a recentinvestigation of mortality trends in England and Wales, althoughrising rates from certain tumors have been clearly influenced bychanging exposures to tobacco smoking, HIV infection, and sunlightexposure. 
Reduced mortality rates among younger age groups usually presagereduced rates as they age. For that reason, it could be expectedthat the reduced cancer mortality in people under 55 in the studyby Devesa et al. will accompany them as they age, and thatcancer mortality for all ages will decrease.
In November of last year, Philip Cole and Brad Rodu of theUniversity of Alabama in Birmingham published an analysis showingthat overall death rates from cancer began to decrease in 1990 andthat the rate of decrease is accelerating.  Better medical care has reduced mortality inpeople who develop cancer, and decreased smoking has reduced thenumber of people who develop lung cancer as well as fatal cancersat other body sites. In addition, Cole and Rodu state thatreductions in exposures to carcinogenic substances in the workplaceand reductions in alcohol consumption and exposures to sunlighthave contributed to decreases in cancer mortality.
Cole and Rodu don't mention "environmental chemicals" in theirdiscussion of causes of cancer. There's good reason. They havelittle to do with cancer. Smoking is the major culprit, and lungcancer rates have been declining since 1990 as a result ofdecreased smoking that began about 1965. For cancer at other sites,Cole and Rodu cite NCI analyses that show mortality rates fromthose cancers "been declining since at least the early 1970s."
Exposures Associated with Cancer
In their 1981 study, Doll and Peto concluded that "pollution,"their term for chemicals in the environment, was associated withabout two percent of all cancers (see table 1). These results werefar different from the conclusions that had been reached by peoplewho equated the word "environment" with "chemicals,"  but Doll and Peto's work conclusions havebecome accepted wisdom, and few people argue today that "cleaningup the environment" is going to make much difference in cancerrates.
EPA itself presented data about the minor importance ofenvironmental exposures on cancer rates in its 1987 reportUnfinished Business. EPA's scientific and technicalmanagers' estimates of the numbers of cancers expected fromenvironmental hazards agreed closely with Doll and Peto's. EPAestimated that pollution caused 1 to 3 percent of cancer  compared to Doll and Peto's estimates of 2percent. Indeed, EPA and Doll and Peto's estimates for all thecauses of cancer that might be regulated agree, and they are alllow (see table 1).  Subsequently, Icalculated that if EPA's estimates of cancer risks fromenvironmental exposures were correct and if its regulatory programswere 100 percent successful in controlling those exposures, theagency could eliminate between 0.25 and 1.3 percent of all cancers.
Percentages of Cancer Deaths Attributed to VariousFactors
Percentage of all cancer deaths
Source of estimate
Doll and EPAb Willettc Ames
Petoa et al.d
Factor or class of factors
Diet 35 (10-70)e - 32 (20-42) 20-40
Tobacco 30 (25-40) - - 35
Infection 10 (1->10)f - - -
Reproductive and sexual
behavior 7 (1-13) - - -
Occupation 4 (2-8) 1-4 - 5
Alcohol 3 (2-4) - - -
(natural radiation) 3 (2-4) 3-6 - -
Pollution 2 (<1-5) 1-3 - -
Food additives 1 (-5-2) - - -
Medicines and medical
procedures 1 (0.5-3) - - -
products <1 (<1-2) <1 - -
Unknown ?f - - -
a. Richard Doll and Richard Peto, Journal of the NationalCancer Institute 66 (1981):1191-1308.
b. Environmental Protection Agency, Unfinished Business(USEPA: Washington, DC, 1987); tabulated in Michael Gough,Environmental Science and Technology 23(1989):925-930.
c. Walter C. Willett, Environmental Health Perspectives103 (1995) Supplement 8:165-170.
d. Bruce N. Ames, Lois Swirsky Gold, and Walter C. Willett,Proceedings of the National Academy of Sciences, USA 92(1995):5258-5265.
e. The best estimate is presented, followed by the "range ofacceptable estimates."
f. "Very uncertain" according to Doll and Peto.
Cancer, the Environment, and Regulations
There are about 530,000 cancer deaths annually in the UnitedStates. According to EPA's estimate, pollution causes 1 to 3percent of cancers, or between 5,300 and 16,000 cancer deathsannually. According to my estimate, EPA regulation, if it worksperfectly, can prevent between 0.25 and 1.3 percent of cancers, orbetween 1,300 and 7,000 cancer deaths. Environmental exposuresaccount for only a tiny fraction of cancers, and EPA regulations,if perfect, can reduce cancer rates by no more than about 1percent. Any good from such regulation may be offset becauseregulations of pesticides are likely to drive up food costs. Dietsrich in fresh fruits and vegetables protects against cancer, and pesticides, by loweringproduction costs and increasing storage and shelf life, make fruitsand vegetables less expensive and more available.
EPA regulations are expensive when compared to regulations fromother agencies. On average, EPA regulations cost fully 330-timesmore to save a year of life than do Federal Aviation Administrationregulations  (table 2). What's more,most calculations of the effectiveness of EPA regulations are basedon estimates of human risk extrapolated from the results of animaltests. Because of EPA's risk estimation procedures, thoseestimates, uncertain as they are, are weighted to overestimate thebenefits from regulations. In contrast, FAA's estimates are basedon analyses of real accidents involving real people.
Median Cost per Life-Year Saved by Regulations from FederalRegulatory Agencies
Agency_______________________________ Cost per Life-YearSaved
Federal Aviation Administration $ 23,000
Consumer Product Safety Commission 68,000
National Highway Traffic Safety Administration 78,000
Occupational Safety and Health Administration 88,000
Environmental Protection Agency 7,600,000
Source: Tammy O. Tengs, et al., Risk Analysis15(1995):369-390.
It's well known that "wealthier is healthier;" that, on average,wealthier people live longer, healthier lives. Regulations costmoney, and the money that is spent on them is not available toindividuals for the purchase of better food, housing,transportation, schooling, medical care, and recreation, all ofwhich can contribute to longer, healthier lives. In recent years,economists have estimated the impact of regulatory costs on lifeexpectancy. Ralph Keeney estimates that every $7.5 million dollarsspent to meet regulatory costs is associated with one prematuredeath,  and Kip Viscusi estimatesthat every $50 million in regulatory costs is associated with onepremature death.  Loud objectionshave been raised to these estimates, but I think that no one hasargued that spending to meet regulation-generated costs makespeople healthier.
If Keeney's estimate that $7.5 million in regulatory costs areassociated with one premature death is correct, every life-yearpurchased by $7.6 million in EPA regulatory costs is offset by onepremature death. According to Viscusi's estimate that associatesone premature death with $50 million in regulatory costs, EPAregulations that save about seven years of life are associated withone premature death. The benefits of EPA's regulations are almostcertainly overestimated, meaning that fewer years of life are savedas a result of EPA's regulations.
Fundamental to rethinking EPA's role in cancer prevention is awillingness to address science and policy separately andappropriately. EPA laces its documents with claims that itspolicies are "science-based" and incorporate the "best science."Both claims have enormous political advantages for EPA. Eitherclaim moves EPA activities from the policy arena, where Congressand other non-technically trained organizations and citizens have avoice, and sets them up as issues to be left to "technical" or"science-based" organizations like EPA.
In fact, a set of assumptions, sometimes called "science policychoices," is central to EPA's decisionmaking. Most of the choicesare made on the basis of "prudence" or "necessity to protect humanhealth and the environment." Those are certainly worthy goals; theyare not technical decisions. Absent from the choices areconsideration of whether regulations actually cost lives or whetherthe money spent on EPA regulations would be better spent elsewhere.Those policy decisions should not be reserved to EPA.
What we know now is far different from what was "known" 27 yearsago. Does the knowledge that the environment plays little role incancer make any difference to EPA and its supporters in their driveto reduce cancer by reducing exposures to chemicals in theenvironment? No. Can we expect that such knowledge will ever make adifference? Yes, if Congress reconsiders EPA's mandate andinvestigates the underpinnings for expectations that regulation ofenvironmental carcinogens will improve public health.
There is no cancer epidemic. Cancer mortality from all cancersother than lung cancers has been dropping since the early 1970s,and lung cancer mortality began dropping in 1990. The contributionof environmental exposures to cancer is small--two percent orless--and regulation of those exposures can reduce cancer mortalityby no more than one percent. Some of that reduction, even ifrealized, might be offset by increased food prices that woulddecrease consumption of fresh fruits and vegetables that are knownto protect against cancer.
The already measurable successes in the fight against cancer,brought about by better treatment coupled with reduced smoking,drinking, and exposures to workplace carcinogens and to sunlight,point the way to continuing progress. In addition, betterunderstanding of the role of diet in cancer prevention and of thecellular and molecular mechanisms of carcinogens are like to havemajor payoffs. To pursue environmental causes in the expectation ofhaving any detectable effects on cancer rates is to chase awill-o'-the-wisp.
 "Cancer andEnvironment: Higginson Speaks Out," Science 205 (1979):1363-1366, at p. 1363.
 "Cancer andEnvironment," at pp. 1363-1364.
 Richard Doll andRichard Peto, "The Causes of Cancer: Quantitative Estimates ofAvoidable Risks of Cancer in the United States Today," Journalof the National Cancer Institute 66 (1981):1193-1308.
 Larry Kessler, astatistician at the Food and Drug Administration, appeared on theAugust 20, 1996, broadcast of NBC's Dateline. He statedthat "The greatest single risk is aging."
 Susan S. Devesa,William J. Blot, B.J. Stone, et al, "Recent Cancer Trendsin the United States," Journal of the National CancerInstitute 87(1995):175-182.
 Devesa etal. 1995 at p. 175.
 Devesa etal. 1995 at p. 181.
 Philip Cole andBrad Rodu, "Declining Cancer Mortality in the United States,"Cancer 78(1996): 2045-2048.
 Cole and Rodu 1996at p. 2046.
 See "Cancer andEnvironment: Higginson Speaks Out."
 U.S.Environmental Protection Agency, Unfinished Business. AComparative Assessment of Environmental Problems. Appendix I:Report of the Cancer Risk Work Group. (USEPA: Washington, DC,February 1987).
 Michael Gough,"Estimating Cancer Mortality. Epidemiological and ToxicologicalMethods Produce Similar Assessments, Environmental Science andTechnology 23 (1989):925-930.
 Michael Gough,"How Much Cancer Can EPA Regulate Away?" Risk Analysis 10(1990):1-6.
 Bruce N. Amesand Lois S. Gold, "The Causes and Prevention of Cancer: The Role ofEnvironment," in The True State of the Planet, ed. RonaldBailey (New York: Free Press, 1995; Bruce N. Ames, Lois S. Gold,and Walter W. Willett, "The Causes and Prevention of Cancer,"Proceedings of the National Academy of Sciences 92(1995):5258-5265; National Research Council, Carcinogens andAnticarcinogens in the Human Diet (Washington: NationalAcademy Press, 1996), pp. 336-337; Walter C. Willett,"Diet, Nutrition, and Avoidable Cancer," Environmental HealthPerspectives 103 (1995) Supplement 8:165-170.
 Tammy O. Tengs,et al. "Five-hundred Life-saving Interventions and TheirCost Effectiveness. Risk Analysis 15 (1995):369-390.
 Ralph L. Keeney,"Mortality Risks Induced by Economic Expenditures," RiskAnalysis 10 (1990): 147-159.
 W. Kip Viscusi,"Mortality Effects of Regulatory Costs and Policy EvaluationCriteria," RAND Journal of Economics 25 (1994):94-109.