The most appalling failure of a criminal justice system is to convict and imprison an innocent person. That is why prosecutors bear a heavy burden. They must convince 12 people to agree unanimously that the accused is guilty beyond all reasonable doubt. The principle that, in a criminal trial, the defendant is "innocent until proven guilty" makes a lot of sense. Does it also make sense when the "defendant" is a new pesticide intended for use on food crops? Or a new pharmaceutical product, or a new food preservative? Should we presume that a proposed new product, intended for use by millions of people, is safe and effective unless it can be proved, beyond any reasonable doubt, that it is harmful?
Unfortunately, a combination of factors, to be discussed below, combine to produce just that effect for a wide variety of materials and devices. The regulatory process, implemented by such agencies as the EPA (Environmental Protection Agency), the FDA (Food and Drug Administration), and OSHA (Occupational Safety and Health Administration) has never been very powerful, and has rarely, if ever, dealt with anything but the most obvious hazards. But, over the past several decades, a period during which there has been a proliferation of new substances and devices, it has been weakened to the point where the protection provided is minimal. Let's look at the problems, root causes, and possible solutions associated with one important product.
Well over 200 million Americans (and perhaps more than two billion people worldwide) use cell phones (also known as mobile phones). Even low-probability hazards associated with cell phones could harm many tens of thousands of people annually. One might have assumed that, before such devices were allowed on the market, the most painstaking research would have been done to identify potential problems. Not so!
Can the high frequency (roughly 1 or 2 GHz) signals generated by cell phones harm users? Many laboratory studies have demonstrated significant biological effects on animal (including human) organisms, from radiation at energy levels below the specified standards. These include effects on skin cells, reproductive systems, eyes, sleep, and brain cells [1-7]. The mechanisms are not well understood. Some of these effects are harmful, e.g., brain cell deaths in rats. In addition to the laboratory experiments, there are a number of epidemiological studies [8-10] correlating cell phone use with brain cancers and other illnesses. But there are also many laboratory studies showing no causal relations between radiation exposure and hypothesized effects on various organisms, and many epidemiological studies yielding negative results. If, indeed cell phone radiation can cause real problems, what would we expect to see in the way of research results?
Obviously the results of some experiments and statistical studies would indicate harm. But we would also expect that many studies would have negative results; various hypothesized harms would not exist—or might exist but not be detectable by particular experimental methods. Many epidemiological studies would not arrive at statistically significant results, sometimes even where there are real cause and effect relations. Some experiments or studies would produce apparent results, but, would be defective in various ways. There would be both false positives and false negatives. (I.e., some defective studies would find harm where it didn't exist, and other defective studies would fail to identify real harms.) Given that funding for such research is limited, some important approaches would not be taken, and some significant initial work would not be followed up properly.
This is precisely what we see today. Laboratory studies make a very strong case that cell phone radiation does have real, measurable, biological effects, and some of these effects are intrinsically harmful. While some epidemiological studies indicate correlations between long-term cell phone use and some diseases—brain cancer is the most prominent—there is much room for controversy, as other studies show no such correlations.
Superficially, the picture would not look much different if it were the case that cell phones caused little or no harm. There is a lot of uncertainty, many unanswered questions, much that is not known about electrical effects in biological organisms and about how they might be impacted by externally produced electromagnetic fields with various properties. How can such a murky situation, where the consequences of bad decisions can be severe, best be resolved, and how is it being handled by the various entities involved, i.e., the cell phone industry, regulatory agencies, the medical profession, and the news media? Here is where the issue of burden of proof comes into play.
It would have been nice if, before cell phones had been sold to hundreds of millions of people, extensive studies had been made by unbiased scientists and engineers to explore the possibilities of harmful effects, with the results weighed, along with the potential benefits, by impartial agencies operating openly and accountable to the public. Why this didn't happen is part of a broader problem.
The regulatory agencies supposed to protect the public against dangerous products and processes are unable to do the job. William Hubbard, a former high level FDA administrator, describing how the FDA does not have the resources to carry out its mission, states, "They are part of a pattern of neglect by officials of both parties in the White House and Congress" . He points out that "Food inspections have dropped from a robust 50,000 in 1972 to about 5,000 today".
Apart from using political influence, backed by campaign finance leverage, to weaken regulatory agencies in general (e.g., via inadequate funding, and the appointment of administrators beholden to industry), corporate interests have used carrots and sticks to influence individual researchers. The carrots are generally in the form of research contracts and consultancies. The sticks are in the form of pressure on employers. Targets of such attacks were Henry Lai and Narendra Singh, whose research, starting in 1995, indicated non-thermal damage to DNA by cell phone type radiation . Their work was dismissed by industry spokespeople as out of line with mainline research. While some industry sponsored scientists were unable to reproduce their results, other scientists have reported similar findings, most recently a German team .
George Carlo is a scientist who changed his position in the controversy over cell phone radiation. He headed a multi-million dollar 6-year research project funded by the cell phone industry. For the most part, no problems were found. But there were some indications of possible harm, which he thought were warning signs meriting further investigation. In a 1999 letter to the AT&T Chairman, Carlo pointed out a number of troublesome findings of the project and expressed concern that the industry did not seem to be responding appropriately . He subsequently became a vocal critic of the industry's position.
That position, shared by the regulatory agencies, was, and remains, that there is nothing to worry about, because there is no clear proof that cell phone radiation can cause significant harm. Neither industry nor regulatory agencies concede that non-ionizing radiation, the type associated with cell phones, can have any biological effects other than those due to heating (i.e., as in a microwave oven). Their attitude is that cell phones must be safe, since those who suspect otherwise have not provided bullet proof evidence to the contrary.
An insider critic of this prevailing view, Norbert Hankin, Chief EMF Scientist for the EPA, stated :
The U.S. Federal Communications Commission, (FCC's) exposure guidelines are considered protective of effects arising from a thermal mechanism but not from all possible mechanisms. Therefore, the generalization by many that the guidelines protect human beings from harm by any or all mechanisms is not justified.Hankin alludes to the substantial body of experimental evidence exemplified above.
We have here an adversarial situation. On one side, by far the most powerful, is the industry, which vigorously denies that there is anything to worry about, and that no real action need be taken unless there is iron-clad proof of harm. On the other side, are individual scientists who argue that there are indeed effects of cell phone radiation, some harmful, but that we do not have enough knowledge at this time to quantify the extent of that harm. The result of this imbalance is that hundreds of millions of people, including a great many children, are, unknowingly, the subjects of a great experiment.
What we need is a major research effort to develop a much greater understanding of the interactions between electromagnetic fields and biological organisms. If the mechanisms involved were reasonably well understood, then it might be possible to find engineering solutions to whatever problems were found. But success is very unlikely if most researchers in the field are biased by financial and/or career considerations, as is the case now.
Rather than trying the aforementioned approach, the cell phone industry seems to be following the trail blazed by other industries faced with charges that their products are unsafe. There are two components. One is to argue that the burden of proof is on those who suggest that the product might be unsafe, and the other is to do everything possible to find fault with work done to generate such proof, so as to stave off or delay regulatory action and law suits. The idea was pioneered by the tobacco companies and refined to an even higher degree by pharmaceutical corporations. It has led to the birth of what has been called the "doubt" industry. These are companies specializing in providing science-based arguments to cast doubt on evidence that a particular product is unsafe (or not efficacious) .
On the face of it, the idea of systematically looking for errors in scientific work is perfectly reasonable, and, in fact, is an integral part of the scientific process. The problem is that in this instance the goal is not to facilitate the search for valid answers, but rather to raise all sorts of side issues that require researchers to spend time and effort to deal with. Here is an example of such a service provided by a company in this field to a pair of pharmaceutical firms, with a product that the FDA was trying to terminate. After briefly summarizing on its website what they did for their clients in this case, The Weinberg Group proudly concluded :
This led to an extensive process with a written appeal from the first decision to the Commissioner and leading to 10 additional years of sales prior to the ultimate cancellation of the drug.
In addition to simply criticizing the work of others, another technique for obscuring the issues is the design of epidemiological studies biased to obtain certain results. In the case of cell phones, such subtle matters as defining what constitutes a heavy user of cell phones, and exactly what effects on users are measured can be manipulated to affect the results.
Coming up with solutions about product safety is not easy, as it is often necessary to deal with a lot of uncertainty about the science involved. What is needed is a substantial body of scientists and engineers who can do the necessary research and who can evaluate products and processes in an objective manner, free of conflicts of interest. In order to develop such a body, we need to increase very substantially research and even some development by laboratories in the public domain, that are not controlled by private interests. Regulatory agencies should be funded to operate major laboratories staffed by first-rate people. NIST (National Institute for Science and Technology) is another valuable asset that should be expanded. The Office of Technology Assessment, which for two decades served the US Congress, should be revived . NSF funding should be expanded to make universities less dependent on corporate support.
There are fundamental philosophical differences over the extent to which technologies that endanger only voluntary users should be regulated, but surely it is not acceptable to permit deception about the nature and extent of the risks. In my view, a fundamental obligation of government is to develop and disseminate accurate information about such risks. One important benefit of this would be to facilitate civil court procedures that would encourage providers of such technology to do their best to minimize hazards and to disseminate accurate information about them. Another benefit would be to stimulate and facilitate efforts to reduce risks by technological means.
In the case of cell phones, there are many conceivable approaches to risk reduction. To what extent can radiation power be reduced? The use of headsets to minimize radiation reaching the brain is already well known. What are the relative benefits of acoustic, wired, wireless, and fiber optic connections to such headsets? Can some sort of shielding be employed? Is there some way to direct radiation away from the user? Is it feasible to develop wave cancellation techniques (analogous to noise reducing headphones)? Could efficient "loudspeakers" and more sensitive microphones be developed so that users need not hold their units close to their heads? Are there modulation methods that reduce harmful effects? Note that, unless there are clearly impartial organizations to evaluate fixes such as these, it would not be sensible to believe that they work. Of course, even without any technological breakthrus, people can choose to use cell phones only for essential purposes, perhaps even substituting pagers in many situations. (With respect to the cell phone I just bought, I expect to use it very sparingly—probably less than two hours per year.) A very disturbing aspect of this problem is the increasingly intensive use of cell phones by children, especially teenagers.
I have discussed radiation produced by individual cell phone transmitters. But there are also questions about radiation from cell phone base stations. The latter can affect people who are not cell phone users, so that the threshold for acceptable risk is much lower. Similar questions about non-ionizing electromagnetic radiation arise in connection with cordless phones, wireless computer connections, and, at the low end of the spectrum, with good old 60 Hz electric power. None of these have been properly resolved.
Comments can be sent to me at unger(at)cs(dot)columbia(dot)edu
Return to Ends and Means