![]()
CHAPTER 1
TRUE WARNINGS AND FALSE ALARMS
Cries of alarm are raised so often over new technological risks that journalists, politicians, and the public have thrown up their collective hands in frustration over their inability to distinguish real hazards from false alarms. Which warnings deserve costly governmental solutions? Which should be provisionally ignored?
Public policy decisions in industrial societies increasingly focus on possible technological and environmental threats to human health (Beck 1992; Stern and Fineberg 1996). These may accompany the introduction of new technologies, like the fear of harmful radiation from proliferating cell phone towers. They may be late-recognized consequences of old technologies, such as the increased incidence of breast cancer now associated with estrogen replacement therapy. Engineering âsolutionsâ thought to alleviate one hazard may introduce another, as when the chlorination of drinking water, which effectively stemmed gastrointestinal illness, was discovered to produce carcinogenic byproducts (Graham and Wiener 1995). With technology continuing to advance and intrude into more diverse domains, risk issues become ever more salient.
Often, if the underlying science has enough time to mature, we can reach confident conclusions about alleged risks. Today we know that chlorofluorocarbons really are damaging the stratospheric ozone layer. Sixty-hertz electromagnetic fields from transmission lines probably are not causing significant increases in leukemia. Human actions almost certainly have raised global temperatures. Silicone breast implants do not cause degenerative disease.
But new risks appear faster than we can scientifically evaluate the old ones. Are endocrine disrupters, the class of synthetic chemicals that possibly distort normal hormonal processes in our bodies, a cause for concern? What about genetically modified foods? Do cell phones cause brain cancer? Does their use while driving increase the likelihood of traffic accidents? Will resistant strains of malaria soon overwhelm the drugs available to attack the microbe? Are we changing peopleâs personalities in harmful ways by overprescribing Prozac for adults and Ritalin for children?
Here, I address a specific problem in risk policy: How can we determine, at a fairly early stage, whether or not to take a new warning seriously? Early recognition is important because of the potential risk to our health and wealth. If we act quickly and the warning turns out to be false, money is wasted, people are needlessly frightened, and risk regulators lose their credibility. If we act too slowly and the hazard is real, we suffer environmental damage or detrimental health effects, perhaps even catastrophic losses.
This policy dilemma leads to an empirical question: Are there hallmarks of early warnings that predict when alleged risks will ultimately turn out to be true hazards rather than false alarms? The identification of such hallmarks is my goal here. Such benchmarks may not allow us to make perfect predictions about which warnings we should heed and which we should ignore. However, the hallmarks may reduce the number of times we respond late to true hazards, and reduce the time and money we devote to false alarms.
Fortunately, Edward Lawless (1977) compiled an excellent collection of public warnings from 1948â1971 for a study of products and processes that some people thought posed serious hazards to the public or the environment. Today, 30â50 years later, we can evaluate whether these warnings identified serious risks or were false alarms.
Lawless never intended his compilation for the purpose to which I put it here, but through his collection he has become my steady companion throughout this project. Since his work appears recurrently in these pages, I introduce him at the outset.
THE LAWLESS COLLECTION
In 1977 Edward Lawless published Technology and Social Shock. The book presents numerous contemporary case studies of public concern about technological and environmental hazards. It describes âepisodes of public alarm over technology [and the environment]ââsocial shocksââthat have inspired major news stories in the media in recent yearsâ (xi). The book was âviewed as a pioneering effort, a forerunner of many further studies and more intensive analysesâ (xi).
Lawless and his team developed more than 40 detailed case studies, intending them as a representative sample of the kinds of technical concerns then in the news. Some are dramatic, some mundane. Three involve reproduction and genetics, 12 are about food and medicines, 16 involve radiation and environmental problems, and several concern projects of the federal government, particularly related to military technology. A few cases are brief historical notes about past technological and medical problems (Krebiozen, the supposed cancer cure) or environmental problems without serious threats to human health (corn leaf blight), or purely intellectual controversies (the authenticity of an antique statue). Dropping these cases reduces the number of relevant health warnings to 31 (Table 1-1).
A typical case is 10 pages long and follows a uniform format. Each case includes historical background, the key events that raised the alarm, the major actors, the disposition (if any) in terms of public or private policy, a summary evaluation, and a bibliography. These 31 case studies of alleged hazards are the basis for this book (See Appendix 1 for a summary of them.)
Table 1-1. Selected Substances and Processes Generating Public Health Warnings, 1947â71
| Oral contraceptives | DMSO (dimethyl sulfoxide) | Enzyme detergents |
| Contaminated cranberries | Shoe fluoroscopes | NTA (nitrilotriacetic acid) in detergents |
|
DES (diethylstilbestrol)
in livestock | Medical x rays | Plutonium at Rocky Flats,
Colorado |
| Cyclamate | Radiation from defective televisions | Radioactive waste stored in Kansas |
| MSG (monosodium glutamate) | Smog in Donora, Pennsylvania | Nuclear test on Amchitka Island, Alaska |
| Botulism | Mercury pollution from industry | Poison gas released at Dugway Proving Ground in Utah |
| Fish protein concentrate | Mercury in tuna | Nerve gas disposal |
Fluoridation of water
supply | DDT (dichloro diphenyl trichloroethane) | ELF (extremely low-frequency radiation) at Project Sanguine |
| Salk polio vaccine | Asbestos | Chemical mace |
| Thalidomide | Taconite pollution | Injuries on synthetic turf |
| HCP (hexachlorophene) | | |
Source: Lawless 1977.
RETROSPECTIVE RISK ASSESSMENT
I ask if, from todayâs perspective, each warning proved true or false. (This is not the same as asking if the warning was wise or foolish at the time, a question to which I return in the last chapter.) I approach this question as a sociologist and a technical generalist, making no attempt to substitute my own assessment for that of experts in the field of the alleged hazard. I obtained these expert assessments from the scientific and government literature and from interviews.
No scientific judgment is absolute. Objections or qualifications can be raised for every risk assessment (see Chapter 4). The technical controversy over fluoridation, for example, has continued for decades. Some competent scientists still insist that adding fluoride to drinking water is imprudent and without beneficial effect, while most experts maintain the practice is not harmful and reduces the incidence of dental cavities. If there is still considerable scientific uncertainty or dispute over the validity of a warning, I consider appraisals from opposing experts. However, I have not taken the existence of dissenting opinion per se to mean that the validity of an alleged hazard has not been determined. With sufficient searching, one can find experts to defend virtually every one of the Lawless warnings, whether because they really believe it or because they want to show intellectual virtuosity. I avoid agnosticism on that basis alone. In general, I have accepted recent scientific positions of orthodox and authoritative bodies such as the National Academy of Sciences, even when there are some dissenters.
For some cases, there remains sufficient scientific uncertainty that warnings still cannot be judged true or false on empirical grounds. For example, a half-century after the first warnings about very low levels of ionizing radiation, the harmful effect, if any, of low doses of x rays remains unknown (see Chapter 4). Nonetheless, it has become accepted regulatory wisdomâvirtually without dissentâthat human exposure to x rays should be avoided unless there are overriding benefits. This is the basis for the U.S. Nuclear Regulatory Commissionâs requirement that public exposure to sources of ionizing radiation shall be âas low as reasonably achievableâ (Title 10, Code of Federal Regulations, Part 20). In the absence of empirical evidence, I take such a regulatory posture as a societal judgment that a warning is or is not valid from todayâs perspective. Thus, even without compelling empirical evidence, I regard early warnings against unnecessary exposure to x rays as valid. (See Chapter 6 for a full description of the basis on which I coded the warnings as true or false.)
HYPOTHETICAL HALLMARKS
My search for hallmarks of true-versus-false warnings is guided by four hypotheses. Each suggests a potential clue to the validity of an early alarm.
⢠True warnings are more likely than false alarms to reach the news media from reputable scientific sources operating in conventional scientific ways.
⢠False alarms are more likely than true warnings to have sponsors with biases against the producer of the alleged hazard.
⢠Hyped media coverage is more likely to indicate a false alarm than a true warning.
⢠A âderivative warningââone arising from a popular social issueâis more likely to indicate a false alarm than a true warning.
True Warnings Are More Likely than False Alarms to Reach the News Media from Reputable Scientific Sources Operating in Conventional Scientific Ways
My first hypothesis speaks to the source of the first public warningâthe person or organization providing information to the first journalist reporting about an alleged hazard. This hypothesis represents an âinternalistâ view of science; that is, the view that despite personal failings and biases among scientists, the conventional scientific enterprise usually produces reasonably objective information about hazards. According to this view, science eventually resolves many technical controversies through the weight of evidence and sound reasoning (Engelhardt and Caplan 1987). While unorthodox views of maverick scientists occasionally turn out to be correct, even revolutionary, more often they are wrong.
Early warnings about fluoroscopic shoe-fitting machinesâx ray machines used to measure peopleâs feet before they purchased shoesâ were based on widely accepted scientific evidence about the cancer-causing properties of high doses of ionizing radiation. In contrast, early warnings about fluoridation came from politically active laypeople in Wisconsin objecting to ârat poisonâ in the water (Mazur 1981).
Rather than attempt to judge on substantive grounds whether or not an early (often contentious) warning was scientifically sound at the time it was made, I adopt the internalist perspective that valid science is conducted by professional Ph.D. or M.D. level scientists working in recognized research institutions and reported in recognized publications, usually after peer review. Thus I code public warnings according to whether or not they are based on reputable science.
False Alarms Are More Likely Than True Warnings to Have Sponsors with Biases against the Producer of the Alleged Hazard
This and the remaining hypotheses reflect âexternalistâ thinking; that is, technical claims are strongly influenced by the biases and self-interests of their proponents (Martin 1991; Krimsky and Golding 1992). People who dislike a corporation are likely to perceive its products and processes negatively, even as hazardous, whereas company executives see them in a positive light. Warnings originating with disinterested parties are, according to this hypothesis, more likely to be valid than those originating with individuals or groups with a history of antipathy toward agents promoting or profiting from the alleged hazard. By the same token, assurances coming from the producer of a suspected hazard are less credible than those from an independent expert.
Personal biases may remain invisible, but we can often identify in professional affiliations and biographical material clear indications that a warning is correlated with a broader pattern of hostility against the source of the alleged hazard. For example, prominent opponents of fluoridation carried strong ideological biases against big government and big corporations. They were especially biased against the aluminum industry, which they regarded as foisting fluoride compounds, waste products from aluminum manufacture, onto unsuspecting communities (Mazur 1981). In contrast, warnings about fluoroscopic shoe-fitting machines came from scientists with no prior grievances against manufacturers of shoes or fluoroscopes.
Hyped Media Coverage Is More Likely to Indicate a False Alarm Than a True Warning
In many cases, public and governmental concerns over an alleged hazard increase or decrease with the amount of media attention. Unfortunately, the amount of news coverage given to various hazards bears little relation to their measurable or calculable risks (Mazur 1984, 1987; Singer and Endreny 1994). Perhaps, then, hints at the validity of an early warning may be found in the ma...
